From d215f014b3526a9898d87bfb4b866287f0864cc9 Mon Sep 17 00:00:00 2001 From: Claudio Imbrenda Date: Fri, 28 Aug 2026 13:54:32 +0200 Subject: [PATCH 001/694] KVM: s390: Fix dirty marking in adapter_indicators_set*() When the indicator and/or summary bits are set in the guest, the accessed page was only marked dirty in KVM if the access was performed using the slow path; accesses through the new kvm_arch_set_irq_inatomic fast inject path would not mark the page as dirty. Fix by adding/moving the missing calls to mark_page_dirty(). Note that for the inatomic path set_page_dirty{,_lock}() is not needed as the page stays pinned; the unpin path correctly marks it as dirty. Opportunistically reorder the local variables to be in reverse Christmas tree order and refactor to use guard(). Fixes: 1e95e3bc6b05 ("KVM: s390: Enable adapter_indicators_set to use mapped pages") Signed-off-by: Claudio Imbrenda Message-ID: <20260828115439.145885-2-imbrenda@linux.ibm.com> --- arch/s390/kvm/s390/interrupt.c | 77 +++++++++++++++++----------------- 1 file changed, 39 insertions(+), 38 deletions(-) diff --git a/arch/s390/kvm/s390/interrupt.c b/arch/s390/kvm/s390/interrupt.c index 0381ae981703..49b4e233e791 100644 --- a/arch/s390/kvm/s390/interrupt.c +++ b/arch/s390/kvm/s390/interrupt.c @@ -2984,61 +2984,58 @@ static int adapter_indicators_set(struct kvm *kvm, struct s390_io_adapter *adapter, struct kvm_s390_adapter_int *adapter_int) { - unsigned long bit; - int summary_set, idx; struct s390_map_info *ind_info, *summary_info; - void *map; struct page *ind_page, *summary_page; - unsigned long flags; + unsigned long bit; + int summary_set; + void *map; ind_page = NULL; - spin_lock_irqsave(&adapter->maps_lock, flags); - ind_info = get_map_info(adapter, adapter_int->ind_addr); + scoped_guard(spinlock_irqsave, &adapter->maps_lock) { + ind_info = get_map_info(adapter, adapter_int->ind_addr); + if (ind_info) { + map = page_address(ind_info->page); + bit = get_ind_bit(ind_info->addr, adapter_int->ind_offset, adapter->swap); + set_bit(bit, map); + } + } if (!ind_info) { - spin_unlock_irqrestore(&adapter->maps_lock, flags); ind_page = pin_map_page(kvm, adapter_int->ind_addr, 0); if (!ind_page) return -1; - idx = srcu_read_lock(&kvm->srcu); map = page_address(ind_page); bit = get_ind_bit(adapter_int->ind_addr, adapter_int->ind_offset, adapter->swap); set_bit(bit, map); - mark_page_dirty(kvm, adapter_int->ind_gaddr >> PAGE_SHIFT); set_page_dirty_lock(ind_page); - srcu_read_unlock(&kvm->srcu, idx); unpin_user_page(ind_page); - } else { - map = page_address(ind_info->page); - bit = get_ind_bit(ind_info->addr, adapter_int->ind_offset, adapter->swap); - set_bit(bit, map); - spin_unlock_irqrestore(&adapter->maps_lock, flags); } + scoped_guard(srcu, &kvm->srcu) + mark_page_dirty(kvm, gpa_to_gfn(adapter_int->ind_gaddr)); - spin_lock_irqsave(&adapter->maps_lock, flags); - summary_info = get_map_info(adapter, adapter_int->summary_addr); + scoped_guard(spinlock_irqsave, &adapter->maps_lock) { + summary_info = get_map_info(adapter, adapter_int->summary_addr); + if (summary_info) { + map = page_address(summary_info->page); + bit = get_ind_bit(summary_info->addr, adapter_int->summary_offset, + adapter->swap); + summary_set = test_and_set_bit(bit, map); + } + } if (!summary_info) { - spin_unlock_irqrestore(&adapter->maps_lock, flags); summary_page = pin_map_page(kvm, adapter_int->summary_addr, 0); if (!summary_page) return -1; - idx = srcu_read_lock(&kvm->srcu); map = page_address(summary_page); bit = get_ind_bit(adapter_int->summary_addr, adapter_int->summary_offset, adapter->swap); summary_set = test_and_set_bit(bit, map); - mark_page_dirty(kvm, adapter_int->summary_gaddr >> PAGE_SHIFT); set_page_dirty_lock(summary_page); - srcu_read_unlock(&kvm->srcu, idx); unpin_user_page(summary_page); - } else { - map = page_address(summary_info->page); - bit = get_ind_bit(summary_info->addr, adapter_int->summary_offset, - adapter->swap); - summary_set = test_and_set_bit(bit, map); - spin_unlock_irqrestore(&adapter->maps_lock, flags); } + scoped_guard(srcu, &kvm->srcu) + mark_page_dirty(kvm, gpa_to_gfn(adapter_int->summary_gaddr)); return summary_set ? 0 : 1; } @@ -3048,26 +3045,29 @@ static int adapter_indicators_set_fast(struct kvm *kvm, struct kvm_s390_adapter_int *adapter_int, int setbit) { + struct s390_map_info *ind_info, *summary_info; unsigned long bit; int summary_set; - struct s390_map_info *ind_info, *summary_info; void *map; - spin_lock(&adapter->maps_lock); + guard(srcu)(&kvm->srcu); + guard(spinlock)(&adapter->maps_lock); + ind_info = get_map_info(adapter, adapter_int->ind_addr); - if (!ind_info) { - spin_unlock(&adapter->maps_lock); + if (!ind_info) return -EWOULDBLOCK; - } + map = page_address(ind_info->page); bit = get_ind_bit(ind_info->addr, adapter_int->ind_offset, adapter->swap); - if (setbit) + if (setbit) { set_bit(bit, map); - summary_info = get_map_info(adapter, adapter_int->summary_addr); - if (!summary_info) { - spin_unlock(&adapter->maps_lock); - return -EWOULDBLOCK; + mark_page_dirty(kvm, gpa_to_gfn(adapter_int->ind_gaddr)); } + + summary_info = get_map_info(adapter, adapter_int->summary_addr); + if (!summary_info) + return -EWOULDBLOCK; + map = page_address(summary_info->page); bit = get_ind_bit(summary_info->addr, adapter_int->summary_offset, adapter->swap); @@ -3077,7 +3077,8 @@ static int adapter_indicators_set_fast(struct kvm *kvm, summary_set = test_and_set_bit(bit, map); else summary_set = test_and_clear_bit(bit, map); - spin_unlock(&adapter->maps_lock); + mark_page_dirty(kvm, gpa_to_gfn(adapter_int->summary_gaddr)); + return summary_set ? 0 : 1; } From ae12d2f9c119639a142d92c4a37c30277e8576da Mon Sep 17 00:00:00 2001 From: Claudio Imbrenda Date: Fri, 28 Aug 2026 13:54:33 +0200 Subject: [PATCH 002/694] KVM: s390: Fix compile warning for kvm_s390_update_cmma_dirty() The parameter "old" should be marked as const, to prevent compile-time warnings. Fixes: d487a24041c2 ("KVM: s390: Prepare gmap for a second KVM implementation") Signed-off-by: Claudio Imbrenda Message-ID: <20260828115439.145885-3-imbrenda@linux.ibm.com> --- arch/s390/kvm/s390/s390.c | 2 +- arch/s390/kvm/s390/s390.h | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/s390/kvm/s390/s390.c b/arch/s390/kvm/s390/s390.c index b0839e887221..8f7e09d7d049 100644 --- a/arch/s390/kvm/s390/s390.c +++ b/arch/s390/kvm/s390/s390.c @@ -5766,7 +5766,7 @@ static long cmma_d_count_pte(union pte *ptep, gfn_t gfn, gfn_t next, struct dat_ return 0; } -void kvm_s390_update_cmma_dirty(struct kvm *kvm, struct kvm_memory_slot *old) +void kvm_s390_update_cmma_dirty(struct kvm *kvm, const struct kvm_memory_slot *old) { const struct dat_walk_ops ops = { .pte_entry = cmma_d_count_pte, }; diff --git a/arch/s390/kvm/s390/s390.h b/arch/s390/kvm/s390/s390.h index d284a263ba70..aa0d1d062f8d 100644 --- a/arch/s390/kvm/s390/s390.h +++ b/arch/s390/kvm/s390/s390.h @@ -472,7 +472,7 @@ int __kvm_s390_mprotect_many(struct gmap *gmap, gpa_t gpa, u8 npages, unsigned i unsigned long bits); bool kvm_arch_setup_async_pf(struct kvm_vcpu *vcpu); -void kvm_s390_update_cmma_dirty(struct kvm *kvm, struct kvm_memory_slot *old); +void kvm_s390_update_cmma_dirty(struct kvm *kvm, const struct kvm_memory_slot *old); int kvm_s390_vm_stop_migration(struct kvm *kvm); From faff4c8ff3dbec6d71b89dd281a0d17c4478fa44 Mon Sep 17 00:00:00 2001 From: Claudio Imbrenda Date: Fri, 28 Aug 2026 13:54:34 +0200 Subject: [PATCH 003/694] KVM: s390: Fix _gaccess_shadow_fault() In some circumstances, it is possible that the page of nested guest memory that is being shadowed is not present at all in the parent guest gmap. dat_entry_walk() will not find any leaf entry and return with -ENOENT, which will erroneously be propagated all the way to userspace. Fix by manually calling gmap_link() on the memory of the nested guest that is being shadowed if the mapping was not already present. Fixes: e38c884df921 ("KVM: s390: Switch to new gmap") Signed-off-by: Claudio Imbrenda Message-ID: <20260828115439.145885-4-imbrenda@linux.ibm.com> --- arch/s390/kvm/s390/gaccess.c | 13 +++++++++++++ 1 file changed, 13 insertions(+) diff --git a/arch/s390/kvm/s390/gaccess.c b/arch/s390/kvm/s390/gaccess.c index e5c064f263df..405345ccc4f1 100644 --- a/arch/s390/kvm/s390/gaccess.c +++ b/arch/s390/kvm/s390/gaccess.c @@ -1589,12 +1589,25 @@ static inline int ___gaccess_shadow_fault(struct kvm_vcpu *vcpu, struct gmap *sg parent = READ_ONCE(sg->parent); if (!parent) return -EAGAIN; +retry: scoped_guard(spinlock, &parent->children_lock) { if (READ_ONCE(sg->parent) != parent) return -EAGAIN; sg->invalidated = false; rc = _gaccess_do_shadow(vcpu->arch.mc, sg, saddr, walk); } + if (rc == -ENOENT) { + struct kvm_memory_slot *slot; + struct guest_fault *entries; + + entries = get_entries(walk); + slot = kvm_vcpu_gfn_to_memslot(vcpu, entries[LEVEL_MEM].gfn); + if (!slot) + return PGM_ADDRESSING; + rc = gmap_link(vcpu->arch.mc, parent, entries + LEVEL_MEM, slot); + if (!rc) + goto retry; + } if (!rc) kvm_s390_release_faultin_array(vcpu->kvm, walk->raw_entries, false); return rc; From 00c0ae5e438615bde828a3b9f33912960f4e6511 Mon Sep 17 00:00:00 2001 From: Claudio Imbrenda Date: Fri, 28 Aug 2026 13:54:35 +0200 Subject: [PATCH 004/694] KVM: s390: Refactor dat_set_slot() Refactor dat_set_slot(), _dat_slot_pte(), _dat_slot_crste(). Now they only take a struct kvm_s390_mmu_cache as priv. For dat_delete_slot(), mc is NULL, as no allocations should take place. This is needed as a prerequisite to move gmap DAT table setup from kvm_arch_commit_memory_region() to kvm_arch_prepare_memory_region(). Signed-off-by: Claudio Imbrenda Message-ID: <20260828115439.145885-5-imbrenda@linux.ibm.com> --- arch/s390/kvm/gmap/dat.c | 29 +++++++++-------------------- arch/s390/kvm/gmap/dat.h | 10 ++++------ arch/s390/kvm/gmap/kvm_mmu.c | 4 ++-- 3 files changed, 15 insertions(+), 28 deletions(-) diff --git a/arch/s390/kvm/gmap/dat.c b/arch/s390/kvm/gmap/dat.c index 24547e39fab2..dcedd5479d82 100644 --- a/arch/s390/kvm/gmap/dat.c +++ b/arch/s390/kvm/gmap/dat.c @@ -846,19 +846,12 @@ long dat_reset_skeys(union asce asce, gfn_t start) } #endif /* KVM_S390_MANAGES_S390_GUEST */ -struct slot_priv { - unsigned long token; - struct kvm_s390_mmu_cache *mc; -}; - static long _dat_slot_pte(union pte *ptep, gfn_t gfn, gfn_t next, struct dat_walk *walk) { - struct slot_priv *p = walk->priv; - union crste dummy = { .val = p->token }; union pte new_pte, pte = READ_ONCE(*ptep); union pgste pgste; - new_pte = _PTE_TOK(dummy.tok.type, dummy.tok.par); + new_pte = walk->priv ? _PTE_EMPTY : _PTE_TOK(_DAT_TOKEN_PIC, PGM_ADDRESSING); /* Table entry already in the desired state. */ if (pte.val == new_pte.val) @@ -875,10 +868,9 @@ static long _dat_slot_pte(union pte *ptep, gfn_t gfn, gfn_t next, struct dat_wal static long _dat_slot_crste(union crste *crstep, gfn_t gfn, gfn_t next, struct dat_walk *walk) { union crste new_crste, crste = READ_ONCE(*crstep); - struct slot_priv *p = walk->priv; + struct kvm_s390_mmu_cache *mc = walk->priv; - new_crste.val = p->token; - new_crste.h.tt = crste.h.tt; + new_crste = mc ? _CRSTE_EMPTY(crste.h.tt) : _CRSTE_HOLE(crste.h.tt); /* Table entry already in the desired state. */ if (crste.val == new_crste.val) @@ -902,7 +894,10 @@ static long _dat_slot_crste(union crste *crstep, gfn_t gfn, gfn_t next, struct d if (!crste.h.fc && !crste.h.i) return 0; /* Split (install a lower level table), and handle things there. */ - return dat_split_crste(p->mc, crstep, gfn, walk->asce, false); + if (mc) + return dat_split_crste(mc, crstep, gfn, walk->asce, false); + /* A large page should never cross memslots boundaries */ + return -EINVAL; } static const struct dat_walk_ops dat_slot_ops = { @@ -910,16 +905,10 @@ static const struct dat_walk_ops dat_slot_ops = { .crste_ops = { _dat_slot_crste, _dat_slot_crste, _dat_slot_crste, _dat_slot_crste, }, }; -int dat_set_slot(struct kvm_s390_mmu_cache *mc, union asce asce, gfn_t start, gfn_t end, - u16 type, u16 param) +int dat_set_slot(struct kvm_s390_mmu_cache *mc, union asce asce, gfn_t start, gfn_t end) { - struct slot_priv priv = { - .token = _CRSTE_TOK(0, type, param).val, - .mc = mc, - }; - return _dat_walk_gfn_range(start, end, asce, &dat_slot_ops, - DAT_WALK_IGN_HOLES | DAT_WALK_ANY, &priv); + DAT_WALK_IGN_HOLES | DAT_WALK_ANY, mc); } static void pgste_set_unlock_multiple(union pte *first, int n, union pgste *pgstes) diff --git a/arch/s390/kvm/gmap/dat.h b/arch/s390/kvm/gmap/dat.h index e452c141b841..90389d47ba4e 100644 --- a/arch/s390/kvm/gmap/dat.h +++ b/arch/s390/kvm/gmap/dat.h @@ -547,8 +547,7 @@ long dat_reset_skeys(union asce asce, gfn_t start); unsigned long dat_get_ptval(struct page_table *table, struct ptval_param param); void dat_set_ptval(struct page_table *table, struct ptval_param param, unsigned long val); -int dat_set_slot(struct kvm_s390_mmu_cache *mc, union asce asce, gfn_t start, gfn_t end, - u16 type, u16 param); +int dat_set_slot(struct kvm_s390_mmu_cache *mc, union asce asce, gfn_t start, gfn_t end); #if KVM_S390_MANAGES_S390_GUEST int dat_set_prefix_notif_bit(union asce asce, gfn_t gfn); @@ -973,16 +972,15 @@ static inline int get_level(union crste *crstep, union pte *ptep) return ptep ? TABLE_TYPE_PAGE_TABLE : crstep->h.tt; } -static inline int dat_delete_slot(struct kvm_s390_mmu_cache *mc, union asce asce, gfn_t start, - unsigned long npages) +static inline int dat_delete_slot(union asce asce, gfn_t start, unsigned long npages) { - return dat_set_slot(mc, asce, start, start + npages, _DAT_TOKEN_PIC, PGM_ADDRESSING); + return dat_set_slot(NULL, asce, start, start + npages); } static inline int dat_create_slot(struct kvm_s390_mmu_cache *mc, union asce asce, gfn_t start, unsigned long npages) { - return dat_set_slot(mc, asce, start, start + npages, _DAT_TOKEN_NONE, 0); + return dat_set_slot(mc, asce, start, start + npages); } static inline bool crste_is_ucas(union crste crste) diff --git a/arch/s390/kvm/gmap/kvm_mmu.c b/arch/s390/kvm/gmap/kvm_mmu.c index b08b8229bb6f..4c8054e18490 100644 --- a/arch/s390/kvm/gmap/kvm_mmu.c +++ b/arch/s390/kvm/gmap/kvm_mmu.c @@ -111,10 +111,10 @@ void s390_kvm_mmu_commit_memory_region(struct kvm *kvm, kvm_s390_update_cmma_dirty(kvm, old); switch (change) { case KVM_MR_DELETE: - rc = dat_delete_slot(mc, kvm->arch.gmap->asce, old->base_gfn, old->npages); + rc = dat_delete_slot(kvm->arch.gmap->asce, old->base_gfn, old->npages); break; case KVM_MR_MOVE: - rc = dat_delete_slot(mc, kvm->arch.gmap->asce, old->base_gfn, old->npages); + rc = dat_delete_slot(kvm->arch.gmap->asce, old->base_gfn, old->npages); if (rc) break; fallthrough; From 19192a4043277af380f83d550a55b92eeeaac7c6 Mon Sep 17 00:00:00 2001 From: Claudio Imbrenda Date: Fri, 28 Aug 2026 13:54:36 +0200 Subject: [PATCH 005/694] KVM: s390: Move all code into s390_kvm_mmu_prepare_memory_region() Move all code from s390_kvm_mmu_commit_memory_region() into s390_kvm_mmu_prepare_memory_region(). This allows the function to fail gracefully if needed. The previous behaviour was to print a warning and continue execution with page tables inconsistent with the memslots. Fixes: e38c884df921 ("KVM: s390: Switch to new gmap") Signed-off-by: Claudio Imbrenda Message-ID: <20260828115439.145885-6-imbrenda@linux.ibm.com> --- arch/s390/kvm/gmap/kvm_mmu.c | 89 +++++++++++++++--------------------- arch/s390/kvm/gmap/kvm_mmu.h | 4 -- arch/s390/kvm/s390/s390.c | 1 - 3 files changed, 36 insertions(+), 58 deletions(-) diff --git a/arch/s390/kvm/gmap/kvm_mmu.c b/arch/s390/kvm/gmap/kvm_mmu.c index 4c8054e18490..c2ffb5e59ec6 100644 --- a/arch/s390/kvm/gmap/kvm_mmu.c +++ b/arch/s390/kvm/gmap/kvm_mmu.c @@ -47,6 +47,9 @@ int s390_kvm_mmu_prepare_memory_region(struct kvm *kvm, struct kvm_memory_slot *new, enum kvm_mr_change change) { + struct kvm_s390_mmu_cache *mc __free(kvm_s390_mmu_cache) = NULL; + int rc = 0; + if (kvm_is_ucontrol(kvm) && new && new->id < KVM_USER_MEM_SLOTS) return -EINVAL; @@ -61,6 +64,10 @@ int s390_kvm_mmu_prepare_memory_region(struct kvm *kvm, * and munmap() stuff in this slot after doing this call at any * time. */ + if (change != KVM_MR_MOVE && change != KVM_MR_CREATE) { + WARN(1, "Unknown KVM MR CHANGE: %d\n", change); + return -EINVAL; + } if (new->userspace_addr & ~PAGE_MASK) return -EINVAL; if ((new->base_gfn + new->npages) * PAGE_SIZE > kvm->arch.mem_limit) @@ -69,65 +76,41 @@ int s390_kvm_mmu_prepare_memory_region(struct kvm *kvm, return -EINVAL; } - if (!kvm_s390_is_migration_mode(kvm)) - return 0; - - /* - * Turn off migration mode when: - * - userspace creates a new memslot with dirty logging off, - * - userspace modifies an existing memslot (MOVE or FLAGS_ONLY) and - * dirty logging is turned off. - * Migration mode expects dirty page logging being enabled to store - * its dirty bitmap. - */ - if (change != KVM_MR_DELETE && - !(new->flags & KVM_MEM_LOG_DIRTY_PAGES)) - WARN(kvm_s390_vm_stop_migration(kvm), - "Failed to stop migration mode"); - - return 0; -} - -void s390_kvm_mmu_commit_memory_region(struct kvm *kvm, - struct kvm_memory_slot *old, - const struct kvm_memory_slot *new, - enum kvm_mr_change change) -{ - struct kvm_s390_mmu_cache *mc __free(kvm_s390_mmu_cache) = NULL; - int rc = 0; - - guard(mutex)(&kvm->slots_arch_lock); + if (kvm->arch.migration_mode) { + /* + * Turn off migration mode when: + * - userspace creates a new memslot with dirty logging off, + * - userspace modifies an existing memslot (MOVE or FLAGS_ONLY) + * and dirty logging is turned off. + * Migration mode expects dirty page logging being enabled to + * store its dirty bitmap. + */ + if (change != KVM_MR_DELETE && + !(new->flags & KVM_MEM_LOG_DIRTY_PAGES)) + WARN(kvm_s390_vm_stop_migration(kvm), + "Failed to stop migration mode"); + } if (change == KVM_MR_FLAGS_ONLY) - return; - - mc = kvm_s390_new_mmu_cache(); - if (!mc) { - rc = -ENOMEM; - goto out; + return 0; + if (change != KVM_MR_DELETE) { + /* Enough capacity to add a new memslot */ + mc = kvm_s390_new_mmu_cache(); + if (!mc) + return -ENOMEM; } - scoped_guard(write_lock, &kvm->mmu_lock) { kvm_s390_update_cmma_dirty(kvm, old); - switch (change) { - case KVM_MR_DELETE: + if (change == KVM_MR_DELETE || change == KVM_MR_MOVE) rc = dat_delete_slot(kvm->arch.gmap->asce, old->base_gfn, old->npages); - break; - case KVM_MR_MOVE: - rc = dat_delete_slot(kvm->arch.gmap->asce, old->base_gfn, old->npages); - if (rc) - break; - fallthrough; - case KVM_MR_CREATE: + if (!rc && (change == KVM_MR_MOVE || change == KVM_MR_CREATE)) rc = dat_create_slot(mc, kvm->arch.gmap->asce, new->base_gfn, new->npages); - break; - case KVM_MR_FLAGS_ONLY: - break; - default: - WARN(1, "Unknown KVM MR CHANGE: %d\n", change); - } } -out: - if (rc) - pr_warn("failed to commit memory region\n"); + /* + * Can only be triggered if dat_{create,delete}_slot() found an + * internal inconsistency or if the mmu cache ran out of memory; + * both should be impossible. + */ + KVM_BUG_ON(rc, kvm); + return rc; } diff --git a/arch/s390/kvm/gmap/kvm_mmu.h b/arch/s390/kvm/gmap/kvm_mmu.h index cdbd390bd33c..43cde61bae03 100644 --- a/arch/s390/kvm/gmap/kvm_mmu.h +++ b/arch/s390/kvm/gmap/kvm_mmu.h @@ -10,9 +10,5 @@ int s390_kvm_mmu_prepare_memory_region(struct kvm *kvm, const struct kvm_memory_slot *old, struct kvm_memory_slot *new, enum kvm_mr_change change); -void s390_kvm_mmu_commit_memory_region(struct kvm *kvm, - struct kvm_memory_slot *old, - const struct kvm_memory_slot *new, - enum kvm_mr_change change); #endif /* ARCH_KVM_GMAP_KVM_MMU_H */ diff --git a/arch/s390/kvm/s390/s390.c b/arch/s390/kvm/s390/s390.c index 8f7e09d7d049..eca4a4359ab2 100644 --- a/arch/s390/kvm/s390/s390.c +++ b/arch/s390/kvm/s390/s390.c @@ -5781,7 +5781,6 @@ void kvm_arch_commit_memory_region(struct kvm *kvm, struct kvm_memory_slot *old, const struct kvm_memory_slot *new, enum kvm_mr_change change) { - s390_kvm_mmu_commit_memory_region(kvm, old, new, change); } /** From f3a557067d57ce6ae98d485c8009b223e16f5f36 Mon Sep 17 00:00:00 2001 From: Claudio Imbrenda Date: Fri, 28 Aug 2026 13:54:37 +0200 Subject: [PATCH 006/694] KVM: s390: Add missing srcu in kvm_s390_set_irq_state() Like kvm_s390_inject_vcpu(), kvm_s390_set_irq_state() also needs the kvm->srcu or the slots lock when performing the Store status operation. Fix by taking kvm->srcu in kvm_s390_set_irq_state(). Fixes: ba5c1e9b6cee ("KVM: s390: interrupt subsystem, cpu timer, waitpsw") Fixes: 062e44a9319f ("KVM: s390: Use srcu in kvm_arch_vcpu_unlocked_ioctl()") Signed-off-by: Claudio Imbrenda Message-ID: <20260828115439.145885-7-imbrenda@linux.ibm.com> --- arch/s390/kvm/s390/interrupt.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/s390/kvm/s390/interrupt.c b/arch/s390/kvm/s390/interrupt.c index 49b4e233e791..f892f4307883 100644 --- a/arch/s390/kvm/s390/interrupt.c +++ b/arch/s390/kvm/s390/interrupt.c @@ -3229,9 +3229,9 @@ int kvm_s390_set_irq_state(struct kvm_vcpu *vcpu, void __user *irqstate, int len break; } } - if (storestatus) { - n = kvm_s390_store_status_unloaded(vcpu, KVM_S390_STORE_STATUS_NOADDR); + scoped_guard(srcu, &vcpu->kvm->srcu) + n = kvm_s390_store_status_unloaded(vcpu, KVM_S390_STORE_STATUS_NOADDR); return r ? r : n; } From 27554b9505ddfc0aeab466aeb60929dfa17284c7 Mon Sep 17 00:00:00 2001 From: Claudio Imbrenda Date: Fri, 28 Aug 2026 13:54:38 +0200 Subject: [PATCH 007/694] KVM: s390: Fix potential races in dat skey functions When dat_cond_set_storage_key() finds a large page, it will conditionally set the storage key in absolute memory using large_crste_to_phys() to get the absolute address. There is a race window between dat_entry_walk() and large_crste_to_phys(): the large page could have been split concurrently, and large_crste_to_phys() might be called with a crste that does not designate a large page, leading to crashes. Similar issues were also present in dat_set_storage_key(). dat_get_storage_key() and dat_reset_reference_bit() did instead check for a potential concurrent splitting of the large page, but then handled it incorrectly. Fix by performing a READ_ONCE on the crste pointer, checking and using the result, instead of dereferencing the pointer again. In case a race is detacted, try dat_entry_walk() again. Fixes: 8e03e8316eb2 ("KVM: s390: KVM page table management functions: storage keys") Signed-off-by: Claudio Imbrenda Message-ID: <20260828115439.145885-8-imbrenda@linux.ibm.com> --- arch/s390/kvm/gmap/dat.c | 36 ++++++++++++++++++++++++++++-------- 1 file changed, 28 insertions(+), 8 deletions(-) diff --git a/arch/s390/kvm/gmap/dat.c b/arch/s390/kvm/gmap/dat.c index dcedd5479d82..ff80d02c9f56 100644 --- a/arch/s390/kvm/gmap/dat.c +++ b/arch/s390/kvm/gmap/dat.c @@ -621,17 +621,20 @@ int dat_get_storage_key(union asce asce, gfn_t gfn, union skey *skey) union pte *ptep; int rc; +again: skey->skey = 0; rc = dat_entry_walk(NULL, gfn, asce, DAT_WALK_ANY, TABLE_TYPE_PAGE_TABLE, &crstep, &ptep); if (rc) return rc; if (!ptep) { - union crste crste; + union crste crste = READ_ONCE(*crstep); - crste = READ_ONCE(*crstep); - if (!crste.h.fc || !crste.s.fc1.pr) + if (!crste_leaf(crste) && !crste.h.i) + goto again; + if (!crste.s.fc1.pr) return 0; + skey->skey = page_get_storage_key(large_crste_to_phys(crste, gfn)); return 0; } @@ -662,13 +665,20 @@ int dat_set_storage_key(struct kvm_s390_mmu_cache *mc, union asce asce, gfn_t gf union pte *ptep; int rc; +again: rc = dat_entry_walk(mc, gfn, asce, DAT_WALK_LEAF_ALLOC, TABLE_TYPE_PAGE_TABLE, &crstep, &ptep); if (rc) return rc; if (!ptep) { - page_set_storage_key(large_crste_to_phys(*crstep, gfn), skey.skey, !nq); + union crste crste = READ_ONCE(*crstep); + + /* A large page has been split concurrently, try again */ + if (!crste_leaf(crste)) + goto again; + + page_set_storage_key(large_crste_to_phys(crste, gfn), skey.skey, !nq); return 0; } @@ -718,15 +728,22 @@ int dat_cond_set_storage_key(struct kvm_s390_mmu_cache *mmc, union asce asce, gf union pte *ptep; int rc; +again: rc = dat_entry_walk(mmc, gfn, asce, DAT_WALK_LEAF_ALLOC, TABLE_TYPE_PAGE_TABLE, &crstep, &ptep); if (rc) return rc; if (!ptep) { + union crste crste = READ_ONCE(*crstep); + + /* A large page has been split concurrently, try again */ + if (!crste_leaf(crste)) + goto again; if (!oldkey) oldkey = &prev; - return page_cond_set_storage_key(large_crste_to_phys(*crstep, gfn), skey, oldkey, + + return page_cond_set_storage_key(large_crste_to_phys(crste, gfn), skey, oldkey, nq, mr, mc); } @@ -768,7 +785,7 @@ int dat_reset_reference_bit(union asce asce, gfn_t gfn, union skey *skey) int rc; skey->skey = 0; - +again: rc = dat_entry_walk(NULL, gfn, asce, DAT_WALK_ANY, TABLE_TYPE_PAGE_TABLE, &crstep, &ptep); if (rc) return rc; @@ -776,9 +793,12 @@ int dat_reset_reference_bit(union asce asce, gfn_t gfn, union skey *skey) if (!ptep) { union crste crste = READ_ONCE(*crstep); - if (!crste.h.fc || !crste.s.fc1.pr) + /* A large page has been split concurrently, try again */ + if (!crste_leaf(crste) && !crste.h.i) + goto again; + if (!crste.s.fc1.pr) return 0; - skey->skey = page_reset_referenced(large_crste_to_phys(*crstep, gfn)) << 1; + skey->skey = page_reset_referenced(large_crste_to_phys(crste, gfn)) << 1; return 0; } old = pgste_get_lock(ptep); From 4ca00a9154f998116fba9a32cce5bd938d228065 Mon Sep 17 00:00:00 2001 From: Claudio Imbrenda Date: Fri, 28 Aug 2026 13:54:39 +0200 Subject: [PATCH 008/694] KVM: s390: Fix race in _destroy_pages_crste() Use READ_ONCE() in _destroy_pages_crste() to read the crste, avoid dereferencing the pointer multiple times. Fixes: a2c17f9270cc ("KVM: s390: New gmap code") Signed-off-by: Claudio Imbrenda Message-ID: <20260828115439.145885-9-imbrenda@linux.ibm.com> --- arch/s390/kvm/gmap/gmap.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/arch/s390/kvm/gmap/gmap.c b/arch/s390/kvm/gmap/gmap.c index 4968330e9553..3f3fa864cc36 100644 --- a/arch/s390/kvm/gmap/gmap.c +++ b/arch/s390/kvm/gmap/gmap.c @@ -994,11 +994,13 @@ static long _destroy_pages_pte(union pte *ptep, gfn_t gfn, gfn_t next, struct da static long _destroy_pages_crste(union crste *crstep, gfn_t gfn, gfn_t next, struct dat_walk *walk) { phys_addr_t origin, cur, end; + union crste crste; - if (!crstep->h.fc || !crstep->s.fc1.pr) + crste = READ_ONCE(*crstep); + if (!crste.h.fc || !crste.s.fc1.pr) return 0; - origin = crste_origin_large(*crstep); + origin = crste_origin_large(crste); cur = ((max(gfn, walk->start) - gfn) << PAGE_SHIFT) + origin; end = ((min(next, walk->end) - gfn) << PAGE_SHIFT) + origin; for ( ; cur < end; cur += PAGE_SIZE) From 65e05ec252a9b79e75930d3c4dd42d8877db04c5 Mon Sep 17 00:00:00 2001 From: Anthony Krowiak Date: Tue, 18 Aug 2026 15:33:49 -0400 Subject: [PATCH 009/694] s390/vfio-ap: fix KVM GISC and page leak when queue removed from host config Three related problems exist in the handling of KVM interrupt and page resources when a queue is removed from the host's AP configuration while assigned to a mediated device (mdev). Problem 1: ~~~~~~~~~ AP_RESPONSE_Q_NOT_AVAIL not handled in vfio_ap_mdev_reset_queue() When the AP bus removes a queue device whose adapter or domain has been removed from the host's AP configuration, vfio_ap_mdev_remove_queue() is called. If the queue is still in the host's AP configuration at that point, it calls vfio_ap_mdev_reset_queue(), which issues a PQAP(ZAPQ). Since the adapter is already gone from the host configuration, ap_zapq() returns AP_RESPONSE_Q_NOT_AVAIL (0x01). This response code is not handled in vfio_ap_mdev_reset_queue()'s switch statement and falls through to the default case, which issues a WARN but does not call vfio_ap_free_aqic_resources(). As a result, if IRQ handling was enabled for the queue by the guest, the KVM GISC registration and the pinned guest page holding the notification indicator byte (NIB) are both leaked. This is fixed by adding AP_RESPONSE_Q_NOT_AVAIL to the same case as AP_RESPONSE_DECONFIGURED and AP_RESPONSE_CHECKSTOPPED in vfio_ap_mdev_reset_queue(). Like those response codes, Q_NOT_AVAIL indicates the queue is not operational and no further reset attempts are possible; the correct action is to free the IRQ resources immediately. Problem 2: ~~~~~~~~~ AP_RESPONSE_Q_NOT_AVAIL not handled in apq_status_check() In vfio_ap_mdev_reset_queue(), there are four cases that indicate a queue reset has not yet completed, in which case apq_reset_check() is queued to a work queue to verify completion of the reset operation. This function uses the PQAP(TAPQ) function to get the queue's status and calls apq_status_check() to verify whether the reset has completed, failed or needs to be executed again. As described in Problem #1 above, apq_reset_check() does not specifically check for AP_RESPONSE_Q_NOT_AVAIL, thereby potentially leaking KVM GISC registration and the pinned guest page holding the NIB. This is fixed by adding a case statement for AP_RESPONSE_Q_NOT_AVAIL to apq_status_check() and returning -ENODEV for that case. The caller, apq_reset_check() will then check for this return code and call vfio_ap_free_aqic_resources() to prevent the leak. Problem 3: ~~~~~~~~~ vfio_ap_free_aqic_resources() leaks saved_isc when kvm is NULL vfio_ap_free_aqic_resources() guards the call to kvm_s390_gisc_unregister() with: if (q->saved_isc != VFIO_AP_ISC_INVALID && !WARN_ON(!(q->matrix_mdev && q->matrix_mdev->kvm))) If matrix_mdev->kvm is NULL -- which can happen when vfio_ap_mdev_unset_kvm() has already run and cleared kvm before a subsequent cleanup path reaches this function -- the WARN_ON fires and the entire block is skipped. This leaves q->saved_isc set to a non-invalid value, creating a potential double-free on any subsequent call to this function. When kvm is NULL the KVM guest is already torn down, so kvm_s390_gisc_unregister() need not and cannot be called; however, q->saved_isc must always be cleared. Fix this by separating the kvm_s390_gisc_unregister() call from the q->saved_isc reset. The WARN_ON now guards only the genuinely impossible case of matrix_mdev being NULL. A NULL kvm is handled gracefully by skipping only the unregister call, and q->saved_isc = VFIO_AP_ISC_INVALID is set unconditionally whenever saved_isc was not already invalid. Additionally, add an else clause to the host-config check in vfio_ap_mdev_remove_queue() to call vfio_ap_free_aqic_resources() directly when the queue is not in the host's AP configuration. This serves as a backstop: when the AP bus fires the driver .remove callback after an adapter is removed from the host config, the queue is by definition no longer addressable, so vfio_ap_mdev_reset_queue() would always return Q_NOT_AVAIL. The else clause handles this case directly without the unnecessary ap_zapq() call, and ensures cleanup occurs even if kvm has already been set to NULL by a prior call to vfio_ap_mdev_unset_kvm(). Fixes: b9bd10c43456d ("s390/vfio-ap: do not reset queue removed from host config") Cc: stable@vger.kernel.org Signed-off-by: Anthony Krowiak Reviewed-by: Matthew Rosato Acked-by: Halil Pasic Signed-off-by: Claudio Imbrenda Message-ID: <20260818193349.1877940-2-akrowiak@linux.ibm.com> --- drivers/s390/crypto/vfio_ap_ops.c | 18 ++++++++++++++---- 1 file changed, 14 insertions(+), 4 deletions(-) diff --git a/drivers/s390/crypto/vfio_ap_ops.c b/drivers/s390/crypto/vfio_ap_ops.c index 940c0ff668be..4db878c18f41 100644 --- a/drivers/s390/crypto/vfio_ap_ops.c +++ b/drivers/s390/crypto/vfio_ap_ops.c @@ -277,9 +277,9 @@ static void vfio_ap_free_aqic_resources(struct vfio_ap_queue *q) { if (!q) return; - if (q->saved_isc != VFIO_AP_ISC_INVALID && - !WARN_ON(!(q->matrix_mdev && q->matrix_mdev->kvm))) { - kvm_s390_gisc_unregister(q->matrix_mdev->kvm, q->saved_isc); + if (q->saved_isc != VFIO_AP_ISC_INVALID) { + if (!WARN_ON(!q->matrix_mdev) && q->matrix_mdev->kvm) + kvm_s390_gisc_unregister(q->matrix_mdev->kvm, q->saved_isc); q->saved_isc = VFIO_AP_ISC_INVALID; } if (q->saved_iova && !WARN_ON(!q->matrix_mdev)) { @@ -1935,6 +1935,8 @@ static int apq_status_check(int apqn, struct ap_queue_status *status) * a value indicating a reset needs to be performed again. */ return -EAGAIN; + case AP_RESPONSE_Q_NOT_AVAIL: + return -ENODEV; default: WARN(true, "failed to verify reset of queue %02x.%04x: TAPQ rc=%u\n", @@ -1961,6 +1963,10 @@ static void apq_reset_check(struct work_struct *reset_work) ret = apq_status_check(q->apqn, &status); if (ret == -EIO) return; + if (ret == -ENODEV) { + vfio_ap_free_aqic_resources(q); + return; + } if (ret == -EBUSY) { pr_notice_ratelimited(WAIT_MSG, elapsed, AP_QID_CARD(q->apqn), @@ -2004,6 +2010,7 @@ static void vfio_ap_mdev_reset_queue(struct vfio_ap_queue *q) break; case AP_RESPONSE_DECONFIGURED: case AP_RESPONSE_CHECKSTOPPED: + case AP_RESPONSE_Q_NOT_AVAIL: vfio_ap_free_aqic_resources(q); break; default: @@ -2528,12 +2535,15 @@ void vfio_ap_mdev_remove_queue(struct ap_device *apdev) /* * If the queue is not in the host's AP configuration, then resetting * it will fail with response code 01, (APQN not valid); so, let's make - * sure it is in the host's config. + * sure it is in the host's config. If it is not, free the KVM GISC + * resources. */ if (test_bit_inv(apid, (unsigned long *)matrix_dev->info.apm) && test_bit_inv(apqi, (unsigned long *)matrix_dev->info.aqm)) { vfio_ap_mdev_reset_queue(q); flush_work(&q->reset_work); + } else { + vfio_ap_free_aqic_resources(q); } done: From d12ce6bce5ec5175c3581e01c71e7a5abb286d9b Mon Sep 17 00:00:00 2001 From: Steffen Eiden Date: Wed, 12 Aug 2026 17:55:19 +0200 Subject: [PATCH 010/694] s390/uv: Fix loop condition in uv_find_secrets Systems with more than 85 UV secrets got -ENOENT for any secret past the first page. Fix this by setting the start index at the beginning of the loop in uv_find_secret() and not at the end. First test if there are more secrets left by comparing start_idx with list->next_secret_idx, and then set the start index to the next secret index. Fixes: 7c9137af2042 ("s390/uv: Retrieve UV secrets support") Acked-by: Claudio Imbrenda Reviewed-by: Christoph Schlameuss Signed-off-by: Steffen Eiden Signed-off-by: Claudio Imbrenda Message-ID: <20260812-uv_secrets_fix-v3-1-a85bd29e0666@linux.ibm.com> --- arch/s390/kernel/uv.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/arch/s390/kernel/uv.c b/arch/s390/kernel/uv.c index dc14ebc0105b..52003c463fad 100644 --- a/arch/s390/kernel/uv.c +++ b/arch/s390/kernel/uv.c @@ -846,11 +846,14 @@ int uv_find_secret(const u8 secret_id[UV_SECRET_ID_LEN], struct uv_secret_list *list, struct uv_secret_list_item_hdr *secret) { - u16 start_idx = 0; + u16 start_idx; u16 list_rc; int ret; + list->next_secret_idx = 0; + do { + start_idx = list->next_secret_idx; uv_list_secrets(list, start_idx, &list_rc, NULL); if (list_rc != UVC_RC_EXECUTED && list_rc != UVC_RC_MORE_DATA) { if (list_rc == UVC_RC_INV_CMD) @@ -861,7 +864,6 @@ int uv_find_secret(const u8 secret_id[UV_SECRET_ID_LEN], ret = find_secret_in_page(secret_id, list, secret); if (ret == 0) return ret; - start_idx = list->next_secret_idx; } while (list_rc == UVC_RC_MORE_DATA && start_idx < list->next_secret_idx); return -ENOENT; From f47190b08b71e8482072978373ee88cb2dfbdaf4 Mon Sep 17 00:00:00 2001 From: Steffen Eiden Date: Wed, 12 Aug 2026 17:55:20 +0200 Subject: [PATCH 011/694] s390/uv: Prevent potential out-of-bounds read When the system has more than 85 secrets, the uv_secret_list struct array only holds up to 85 items per page, resulting in an out of bounds read in find_secret_in_page if the targeted secret is in the next page or not stored at all. Fix this by looping over the number of stored secrets which is the per sub-list count of stored secrets and not the overall count. Fixes: 7c9137af2042 ("s390/uv: Retrieve UV secrets support") Signed-off-by: Steffen Eiden Reviewed-by: Christoph Schlameuss Signed-off-by: Claudio Imbrenda Message-ID: <20260812-uv_secrets_fix-v3-2-a85bd29e0666@linux.ibm.com> --- arch/s390/kernel/uv.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/s390/kernel/uv.c b/arch/s390/kernel/uv.c index 52003c463fad..8ea9dd7704ff 100644 --- a/arch/s390/kernel/uv.c +++ b/arch/s390/kernel/uv.c @@ -825,7 +825,7 @@ static int find_secret_in_page(const u8 secret_id[UV_SECRET_ID_LEN], { u16 i; - for (i = 0; i < list->total_num_secrets; i++) { + for (i = 0; i < list->num_secr_stored; i++) { if (memcmp(secret_id, list->secrets[i].id, UV_SECRET_ID_LEN) == 0) { *secret = list->secrets[i].hdr; return 0; From b5965c4221165045e19736794ca5f48ae3d67142 Mon Sep 17 00:00:00 2001 From: Pei Xiao Date: Thu, 13 Aug 2026 10:48:12 +0800 Subject: [PATCH 012/694] mtd: block2mtd: Fix divide error when erase_size is zero MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The erase size is parsed from the "block2mtd" module parameter and can be set to zero. add_device() then evaluates if (size % erase_size) with a zero divisor, which triggers a divide error: divide error: 0000 [#1] PREEMPT SMP PTI RIP: 0010:add_device drivers/mtd/devices/block2mtd.c:296 [inline] RIP: 0010:block2mtd_setup2+0x592/0xda0 drivers/mtd/devices/block2mtd.c:459 Call Trace: block2mtd_setup+0x27/0xe0 drivers/mtd/devices/block2mtd.c:476 param_attr_store+0x214/0x310 kernel/params.c:589 module_attr_store+0x65/0x90 kernel/params.c:904 kernfs_fop_write_iter+0x3a4/0x540 fs/kernfs/file.c:345 ... Reject a zero erase size before performing the modulo operation so the existing "erasesize must be a divisor of device size" error path reports the invalid argument and frees the device. Fixes: ea6d833a3fdd ("mtd: block2mtd: check device size") Reported-by: syzbot+b320a4d5f65a61dbbf89@syzkaller.appspotmail.com Closes: https://lore.kernel.org/lkml/6a7b58ba.ac361c09.22ff0a.0045.GAE@google.com/ Suggested-by: Jörn Engel Cc: stable@vger.kernel.org Signed-off-by: Pei Xiao Signed-off-by: Miquel Raynal --- drivers/mtd/devices/block2mtd.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/mtd/devices/block2mtd.c b/drivers/mtd/devices/block2mtd.c index 03e80b2c4f5a..349fa07be314 100644 --- a/drivers/mtd/devices/block2mtd.c +++ b/drivers/mtd/devices/block2mtd.c @@ -293,7 +293,7 @@ static struct block2mtd_dev *add_device(char *devname, int erase_size, } size = bdev_nr_bytes(bdev); - if ((long)size % erase_size) { + if (!erase_size || (long)size % erase_size) { pr_err("erasesize must be a divisor of device size\n"); goto err_free_block2mtd; } From 63d6cace2c4a7f36c1cb44f1bb5e0f3ef19e5c7f Mon Sep 17 00:00:00 2001 From: Han Xu Date: Thu, 13 Aug 2026 08:07:24 -0500 Subject: [PATCH 013/694] mtd: spinand: Enable QE on all dies MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The QUAD ENABLE (QE) bit is stored in a per-die configuration register on some SPI-NAND devices. When a device contains multiple dies, updating the QE bit only on the currently selected die can leave the remaining dies operating in non-quad mode.   Iterate over all targets and update the QE setting on each die during initialization to ensure consistent quad I/O operation across the entire device. Tested on ISSI IS38SMW04G8B. Fixes: 7529df465248 ("mtd: nand: Add core infrastructure to support SPI NANDs") Cc: stable@vger.kernel.org Signed-off-by: Han Xu Signed-off-by: Miquel Raynal --- drivers/mtd/nand/spi/core.c | 22 ++++++++++++++++++++-- 1 file changed, 20 insertions(+), 2 deletions(-) diff --git a/drivers/mtd/nand/spi/core.c b/drivers/mtd/nand/spi/core.c index 8bf9301f25e7..bea352dd82d7 100644 --- a/drivers/mtd/nand/spi/core.c +++ b/drivers/mtd/nand/spi/core.c @@ -281,8 +281,26 @@ static int spinand_init_cfg_cache(struct spinand_device *spinand) static int spinand_init_quad_enable(struct spinand_device *spinand, bool enable) { - return spinand_upd_cfg(spinand, CFG_QUAD_ENABLE, - enable ? CFG_QUAD_ENABLE : 0); + struct nand_device *nand = spinand_to_nand(spinand); + unsigned int target; + int ret; + + /* + * QE is a per-die setting on some devices. Program each target + * individually when enabling or disabling quad I/O mode. + */ + for (target = 0; target < nand->memorg.ntargets; target++) { + ret = spinand_select_target(spinand, target); + if (ret) + return ret; + + ret = spinand_upd_cfg(spinand, CFG_QUAD_ENABLE, + enable ? CFG_QUAD_ENABLE : 0); + if (ret) + return ret; + } + + return 0; } static int spinand_ecc_enable(struct spinand_device *spinand, From 26300879cd8e4612dce66bb8f6ff7cd35fcf3e59 Mon Sep 17 00:00:00 2001 From: Karl Mehltretter Date: Sun, 16 Aug 2026 18:30:27 +0200 Subject: [PATCH 014/694] mtd: core: avoid double-free of OTP NVMEM device If factory OTP setup fails after the user OTP NVMEM device has been registered, mtd_otp_nvmem_add() unregisters the user device but leaves mtd->otp_user_nvmem set. On an error, the caller unregisters it again. For -EOPNOTSUPP, registration continues and normal teardown unregisters it again. Clear mtd->otp_user_nvmem after unregistering it. Fixes: e0489f6e221f ("mtd: core: fix error path for nvmem provider") Fixes: fe0b8213c012 ("mtd: core: Don't fail mtd_otp_nvmem_add() if OTP is unsupported") Assisted-by: Codex:gpt-5.6-sol Signed-off-by: Karl Mehltretter Signed-off-by: Miquel Raynal --- drivers/mtd/mtdcore.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/mtd/mtdcore.c b/drivers/mtd/mtdcore.c index 16629382a787..6158452be24d 100644 --- a/drivers/mtd/mtdcore.c +++ b/drivers/mtd/mtdcore.c @@ -1083,6 +1083,7 @@ static int mtd_otp_nvmem_add(struct mtd_info *mtd) err: nvmem_unregister(mtd->otp_user_nvmem); + mtd->otp_user_nvmem = NULL; /* Don't report error if OTP is not supported. */ if (err == -EOPNOTSUPP) return 0; From 12b31d1acd20b3185bb5f0748db0cf6874c8e9d1 Mon Sep 17 00:00:00 2001 From: Karl Mehltretter Date: Sun, 16 Aug 2026 18:30:58 +0200 Subject: [PATCH 015/694] mtd: core: call _get_device() with the master MTD __get_mtd_device() calls the master's _get_device() callback with mtd, which may be a partition. __put_mtd_device() passes the master. This breaks gluebi partitions because gluebi_get_device() uses container_of() and therefore requires the master MTD. Pass the master to _get_device(), matching _put_device(). Fixes: 46b5889cc2c5 ("mtd: implement proper partition handling") Assisted-by: Codex:gpt-5.6-sol Signed-off-by: Karl Mehltretter Signed-off-by: Miquel Raynal --- drivers/mtd/mtdcore.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/mtd/mtdcore.c b/drivers/mtd/mtdcore.c index 6158452be24d..491a27d0315f 100644 --- a/drivers/mtd/mtdcore.c +++ b/drivers/mtd/mtdcore.c @@ -1333,7 +1333,7 @@ int __get_mtd_device(struct mtd_info *mtd) int err; if (master->_get_device) { - err = master->_get_device(mtd); + err = master->_get_device(master); if (err) return err; } From 39b975ff208610301bc400460ec245caeffa4ee0 Mon Sep 17 00:00:00 2001 From: Karl Mehltretter Date: Mon, 24 Aug 2026 03:47:03 +0200 Subject: [PATCH 016/694] mtd: cfi_cmdset_0001: shrink do_write_buffer() stack frame arm32 allmodconfig fails to build with gcc: drivers/mtd/chips/cfi_cmdset_0001.c:1883:1: error: the frame size of 1296 bytes is larger than 1280 bytes [-Werror=frame-larger-than=] With MTD_MAP_BANK_WIDTH_32 a map_word is 32 bytes, and with KASAN_STACK every temporary gets its own redzoned slot. do_write_buffer() builds a map_word for each of the twelve commands it issues, and those temporaries dominate its frame. Issue the commands through a small noinline helper that takes the command as a plain value and builds the map_word in its own frame. The sequence of reads and writes to the chip is unchanged. The helper must not be inlined, and reusing one map_word local is not enough: clang cannot assume the callee does not observe the local, so it still returns each CMD() into a per-call-site temporary, and the frame grows (1760 -> 1824 bytes in cfi_intelext_writev(), into which clang inlines do_write_buffer()). Frame sizes with MTD_MAP_BANK_WIDTH_32 and KASAN_STACK: gcc 15.2 do_write_buffer() 1296 -> 520 clang 21 cfi_intelext_writev() 1760 -> 896 The helper itself takes 168 (gcc) / 160 (clang) bytes. Reported-by: Andy Shevchenko Closes: https://lore.kernel.org/all/20260408211407.2295175-1-andriy.shevchenko@linux.intel.com/ Suggested-by: Miquel Raynal Link: https://lore.kernel.org/all/87ik9cfm6g.fsf@bootlin.com/ Assisted-by: LLM Signed-off-by: Karl Mehltretter Signed-off-by: Miquel Raynal --- drivers/mtd/chips/cfi_cmdset_0001.c | 42 ++++++++++++++++++----------- 1 file changed, 26 insertions(+), 16 deletions(-) diff --git a/drivers/mtd/chips/cfi_cmdset_0001.c b/drivers/mtd/chips/cfi_cmdset_0001.c index 6049ba2d6bcb..60d82a1acb01 100644 --- a/drivers/mtd/chips/cfi_cmdset_0001.c +++ b/drivers/mtd/chips/cfi_cmdset_0001.c @@ -1716,12 +1716,24 @@ static int cfi_intelext_write_words (struct mtd_info *mtd, loff_t to , size_t le } +/* + * Keep noinline: inlined, the map_word temporaries put do_write_buffer() over + * the frame-size limit with MTD_MAP_BANK_WIDTH_32 and KASAN_STACK. + */ +static noinline void __xipram cfi_write_cmd(struct map_info *map, + unsigned long cmd, unsigned long adr) +{ + struct cfi_private *cfi = map->fldrv_priv; + + map_write(map, CMD(cmd), adr); +} + static int __xipram do_write_buffer(struct map_info *map, struct flchip *chip, unsigned long adr, const struct kvec **pvec, unsigned long *pvec_seek, int len) { struct cfi_private *cfi = map->fldrv_priv; - map_word status, write_cmd, datum; + map_word status, datum; unsigned long cmd_adr; int ret, wbufsize, word_gap, words; const struct kvec *vec; @@ -1740,9 +1752,6 @@ static int __xipram do_write_buffer(struct map_info *map, struct flchip *chip, if (is_LH28F640BF(cfi)) cmd_adr = adr; - /* Let's determine this according to the interleave only once */ - write_cmd = (cfi->cfiq->P_ID != P_ID_INTEL_PERFORMANCE) ? CMD(0xe8) : CMD(0xe9); - mutex_lock(&chip->mutex); ret = get_chip(map, chip, cmd_adr, FL_WRITING); if (ret) { @@ -1759,7 +1768,7 @@ static int __xipram do_write_buffer(struct map_info *map, struct flchip *chip, So we must check here and reset those bits if they're set. Otherwise we're just pissing in the wind */ if (chip->state != FL_STATUS) { - map_write(map, CMD(0x70), cmd_adr); + cfi_write_cmd(map, 0x70, cmd_adr); chip->state = FL_STATUS; } status = map_read(map, cmd_adr); @@ -1767,21 +1776,22 @@ static int __xipram do_write_buffer(struct map_info *map, struct flchip *chip, xip_enable(map, chip, cmd_adr); printk(KERN_WARNING "SR.4 or SR.5 bits set in buffer write (status %lx). Clearing.\n", status.x[0]); xip_disable(map, chip, cmd_adr); - map_write(map, CMD(0x50), cmd_adr); - map_write(map, CMD(0x70), cmd_adr); + cfi_write_cmd(map, 0x50, cmd_adr); + cfi_write_cmd(map, 0x70, cmd_adr); } chip->state = FL_WRITING_TO_BUFFER; - map_write(map, write_cmd, cmd_adr); + cfi_write_cmd(map, (cfi->cfiq->P_ID != P_ID_INTEL_PERFORMANCE) ? 0xe8 : 0xe9, + cmd_adr); ret = WAIT_TIMEOUT(map, chip, cmd_adr, 0, 0); if (ret) { /* Argh. Not ready for write to buffer */ map_word Xstatus = map_read(map, cmd_adr); - map_write(map, CMD(0x70), cmd_adr); + cfi_write_cmd(map, 0x70, cmd_adr); chip->state = FL_STATUS; status = map_read(map, cmd_adr); - map_write(map, CMD(0x50), cmd_adr); - map_write(map, CMD(0x70), cmd_adr); + cfi_write_cmd(map, 0x50, cmd_adr); + cfi_write_cmd(map, 0x70, cmd_adr); xip_enable(map, chip, cmd_adr); printk(KERN_ERR "%s: Chip not ready for buffer write. Xstatus = %lx, status = %lx\n", map->name, Xstatus.x[0], status.x[0]); @@ -1800,7 +1810,7 @@ static int __xipram do_write_buffer(struct map_info *map, struct flchip *chip, } /* Write length of data to come */ - map_write(map, CMD(words), cmd_adr ); + cfi_write_cmd(map, words, cmd_adr); /* Write data */ vec = *pvec; @@ -1837,7 +1847,7 @@ static int __xipram do_write_buffer(struct map_info *map, struct flchip *chip, *pvec_seek = vec_seek; /* GO GO GO */ - map_write(map, CMD(0xd0), cmd_adr); + cfi_write_cmd(map, 0xd0, cmd_adr); chip->state = FL_WRITING; ret = INVAL_CACHE_AND_WAIT(map, chip, cmd_adr, @@ -1845,7 +1855,7 @@ static int __xipram do_write_buffer(struct map_info *map, struct flchip *chip, chip->buffer_write_time, chip->buffer_write_time_max); if (ret) { - map_write(map, CMD(0x70), cmd_adr); + cfi_write_cmd(map, 0x70, cmd_adr); chip->state = FL_STATUS; xip_enable(map, chip, cmd_adr); printk(KERN_ERR "%s: buffer write error (status timeout)\n", map->name); @@ -1858,8 +1868,8 @@ static int __xipram do_write_buffer(struct map_info *map, struct flchip *chip, unsigned long chipstatus = MERGESTATUS(status); /* reset status */ - map_write(map, CMD(0x50), cmd_adr); - map_write(map, CMD(0x70), cmd_adr); + cfi_write_cmd(map, 0x50, cmd_adr); + cfi_write_cmd(map, 0x70, cmd_adr); xip_enable(map, chip, cmd_adr); if (chipstatus & 0x02) { From 200c32e3cd53132fbf99f1e2a5751f556755cf6d Mon Sep 17 00:00:00 2001 From: Menachem Adin Date: Thu, 27 Aug 2026 08:11:56 +0300 Subject: [PATCH 017/694] mtd: mtd_intel_dg: reset poll counter for each erase The non-posted erase polling counter is initialized only once per MTD erase request. Large requests therefore share the polling budget across all 4K erase commands and can fail with -ETIME even though no individual command timed out. Reset the counter for each 4K erase command so every operation gets the intended completion timeout. Cc: stable@vger.kernel.org Fixes: a1c940cbf505 ("drm/xe/nvm: add support for non-posted erase") Signed-off-by: Menachem Adin Signed-off-by: Alexander Usyskin Signed-off-by: Miquel Raynal --- drivers/mtd/devices/mtd_intel_dg.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/mtd/devices/mtd_intel_dg.c b/drivers/mtd/devices/mtd_intel_dg.c index a566e86eb5e3..55a945c1a67a 100644 --- a/drivers/mtd/devices/mtd_intel_dg.c +++ b/drivers/mtd/devices/mtd_intel_dg.c @@ -387,7 +387,7 @@ idg_erase(struct intel_dg_nvm *nvm, u8 region, loff_t from, u64 len, u64 *fail_a void __iomem *base2 = nvm->base2; void __iomem *base = nvm->base; const u32 block = 0x10; - u32 iter = 0; + u32 iter; u32 reg; u64 i; @@ -396,6 +396,7 @@ idg_erase(struct intel_dg_nvm *nvm, u8 region, loff_t from, u64 len, u64 *fail_a iowrite32(region << 24 | block, base + NVM_ERASE_REG); if (nvm->non_posted_erase) { /* Wait for Erase Done */ + iter = 0; reg = ioread32(base2 + NVM_DEBUG_REG); while (!(reg & NVM_NON_POSTED_ERASE_DONE) && ++iter < NVM_NON_POSTED_ERASE_DONE_ITER) { From b890e6163761ddb580cc74f43e15f8bbde15647e Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Nuno=20S=C3=A1?= Date: Mon, 31 Aug 2026 16:07:17 +0100 Subject: [PATCH 018/694] mtd: spinand: fix NULL pointer dereference with no ECC engine MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit When "nand-no-ecc-engine" is set in DT, nanddev_get_ecc_engine() takes the NAND_ECC_ENGINE_TYPE_NONE path and returns success while leaving nand->ecc.engine NULL. The SPI-NAND code nevertheless dereferences it unconditionally to test for a pipelined engine, so probing such a device oopses immediately. Rather than open-coding the test three times, add a nand_ecc_is_pipelined() helper to the NAND core that folds the NULL check into the integration comparison, and use it everywhere. Future callers then cannot reintroduce the problem. Fixes: f9d7c7265bcf ("mtd: spinand: Create direct mapping descriptors for ECC operations") Cc: stable@vger.kernel.org Signed-off-by: Nuno Sá Signed-off-by: Miquel Raynal --- drivers/mtd/nand/ecc.c | 7 +++++++ drivers/mtd/nand/spi/core.c | 8 +++----- include/linux/mtd/nand.h | 2 ++ 3 files changed, 12 insertions(+), 5 deletions(-) diff --git a/drivers/mtd/nand/ecc.c b/drivers/mtd/nand/ecc.c index 6ccdff3fc913..1d809cbd47e6 100644 --- a/drivers/mtd/nand/ecc.c +++ b/drivers/mtd/nand/ecc.c @@ -159,6 +159,13 @@ int nand_ecc_finish_io_req(struct nand_device *nand, } EXPORT_SYMBOL(nand_ecc_finish_io_req); +bool nand_ecc_is_pipelined(const struct nand_device *nand) +{ + return nand->ecc.engine && + nand->ecc.engine->integration == NAND_ECC_ENGINE_INTEGRATION_PIPELINED; +} +EXPORT_SYMBOL(nand_ecc_is_pipelined); + /* Define default OOB placement schemes for large and small page devices */ static int nand_ooblayout_ecc_sp(struct mtd_info *mtd, int section, struct mtd_oob_region *oobregion) diff --git a/drivers/mtd/nand/spi/core.c b/drivers/mtd/nand/spi/core.c index bea352dd82d7..4fb769581595 100644 --- a/drivers/mtd/nand/spi/core.c +++ b/drivers/mtd/nand/spi/core.c @@ -526,8 +526,7 @@ static int spinand_read_from_cache_op(struct spinand_device *spinand, else rdesc->info.op_tmpl = &rdesc->info.primary_op_tmpl; - if (nand->ecc.engine->integration == NAND_ECC_ENGINE_INTEGRATION_PIPELINED && - req->mode != MTD_OPS_RAW) + if (nand_ecc_is_pipelined(nand) && req->mode != MTD_OPS_RAW) rdesc->info.op_tmpl->data.ecc = true; else rdesc->info.op_tmpl->data.ecc = false; @@ -621,8 +620,7 @@ static int spinand_write_to_cache_op(struct spinand_device *spinand, wdesc = spinand->dirmaps[req->pos.plane].wdesc; - if (nand->ecc.engine->integration == NAND_ECC_ENGINE_INTEGRATION_PIPELINED && - req->mode != MTD_OPS_RAW) + if (nand_ecc_is_pipelined(nand) && req->mode != MTD_OPS_RAW) wdesc->info.op_tmpl->data.ecc = true; else wdesc->info.op_tmpl->data.ecc = false; @@ -1279,7 +1277,7 @@ static int spinand_create_dirmap(struct spinand_device *spinand, struct spi_mem_dirmap_desc *desc; bool enable_ecc = false, secondary_op = false; - if (nand->ecc.engine->integration == NAND_ECC_ENGINE_INTEGRATION_PIPELINED) + if (nand_ecc_is_pipelined(nand)) enable_ecc = true; if (spinand->cont_read_possible && spinand->op_templates->cont_read_cache) diff --git a/include/linux/mtd/nand.h b/include/linux/mtd/nand.h index 09c8c93e4dba..6936180b6ea5 100644 --- a/include/linux/mtd/nand.h +++ b/include/linux/mtd/nand.h @@ -305,6 +305,8 @@ int nand_ecc_prepare_io_req(struct nand_device *nand, struct nand_page_io_req *req); int nand_ecc_finish_io_req(struct nand_device *nand, struct nand_page_io_req *req); +bool nand_ecc_is_pipelined(const struct nand_device *nand); + bool nand_ecc_is_strong_enough(struct nand_device *nand); #if IS_REACHABLE(CONFIG_MTD_NAND_CORE) From 636fe10d8f3559210ff63108a208d39ea2e9c3bd Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Nuno=20S=C3=A1?= Date: Mon, 31 Aug 2026 16:07:18 +0100 Subject: [PATCH 019/694] mtd: spinand: fix zero oobavail when no ECC engine is used MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Commit 00c15b78b4b4 ("mtd: spinand: Allow the case where there is no ECC engine") made the OOB free bytes count conditional on having an ECC engine so that probing would not fail when none is requested. However mtd->oobavail is still assigned from ret just after that block, and ret is 0 there, so the device ends up advertising no available OOB bytes at all. mtd_oobavail() returns mtd->oobavail for MTD_OPS_AUTO_OOB, so a zero value makes every automatic OOB access fail with -EINVAL. JFFS2 fares worse: it keeps its cleanmarker in the OOB area on NAND and refuses to mount outright, reporting "inconsistent device description". No ECC engine also means no ooblayout was ever installed, which would make the count return -ENOTSUPP, so install the same fallback layout the on-die path already uses before counting unconditionally. Fixes: 00c15b78b4b4 ("mtd: spinand: Allow the case where there is no ECC engine") Cc: stable@vger.kernel.org Signed-off-by: Nuno Sá Signed-off-by: Miquel Raynal --- drivers/mtd/nand/spi/core.c | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-) diff --git a/drivers/mtd/nand/spi/core.c b/drivers/mtd/nand/spi/core.c index 4fb769581595..43bbe2629b85 100644 --- a/drivers/mtd/nand/spi/core.c +++ b/drivers/mtd/nand/spi/core.c @@ -1996,12 +1996,17 @@ static int spinand_init(struct spinand_device *spinand) goto err_cleanup_ecc_engine; } - if (nand->ecc.engine) { - ret = mtd_ooblayout_count_freebytes(mtd); - if (ret < 0) - goto err_cleanup_ecc_engine; + if (!nand->ecc.engine) { + if (spinand->eccinfo.ooblayout) + mtd_set_ooblayout(mtd, spinand->eccinfo.ooblayout); + else + mtd_set_ooblayout(mtd, &spinand_noecc_ooblayout); } + ret = mtd_ooblayout_count_freebytes(mtd); + if (ret < 0) + goto err_cleanup_ecc_engine; + mtd->oobavail = ret; /* Propagate ECC information to mtd_info */ From 37adc9c5789c76db3b1ee7aeec3999b8503020d0 Mon Sep 17 00:00:00 2001 From: Mehmet Fide Date: Tue, 1 Sep 2026 09:39:06 +0200 Subject: [PATCH 020/694] mtd: rawnand: vf610_nfc: fix reads on chips with more than 64 bytes of OOB The controller transfers 64 spare bytes per page and the driver only implements the matching 64-byte ECC layout, so attach_chip() shrinks mtd->oobsize when the chip provides more. That clamp does not survive: nand_scan_tail() runs nanddev_init() after ->attach_chip(), and it restores mtd->oobsize from the memory organization, which still holds the value detected from the chip. The driver then transfers writesize plus the chip's full OOB size, the hardware ECC parity ends up at a different offset than the layout the controller was set up for, and every ECC-protected read fails with -EBADMSG. Measured on a Colibri VF61 (MX30LF4G28AC, 2048-byte pages, 112 bytes of OOB): with the clamp lost, UBI cannot read the erase counter headers of the pages U-Boot has just written, and the on-flash bad block table written by an older kernel reads back with ECC errors, so the board does not boot. Kernels before commit a7ab085d7c16 ("mtd: rawnand: Initialize the nand_device object") are not affected because nothing overwrote the clamp there, which is why the same chip works with a v4.4 kernel and with U-Boot, whose copy of this driver has no memory organization to restore the value from. Edward Karpicz reported that the clamp no longer takes effect on this chip; see the link below. Instead of modifying the memory organization, keep the detected OOB size and give the driver its own mtd_ooblayout_ops: the same layout the NAND core uses for large pages, but computed on the first 64 OOB bytes instead of the whole OOB, so the ECC bytes stay where U-Boot and the old kernels put them. The data paths transfer writesize plus those 64 bytes, as the controller always has. Since mtd->oobsize now reports the chip's real spare size, fill the tail of oob_poi with 0xff after the 64 transferred bytes on ECC page reads: the core may copy the full mtd->oobsize from it, which would otherwise expose whatever the buffer held before. 0xff also matches what a raw read returns from flash, since the write path only ever programs the first 64 spare bytes. Reported-by: Edward Karpicz Link: https://community.toradex.com/t/colibri-vf50-vf61-on-the-current-bsp-mainline-u-boot-v2026-07-and-linux-6-18-lts/30735 Suggested-by: Miquel Raynal Fixes: a7ab085d7c16 ("mtd: rawnand: Initialize the nand_device object") Cc: stable@vger.kernel.org Signed-off-by: Mehmet Fide Signed-off-by: Miquel Raynal --- drivers/mtd/nand/raw/vf610_nfc.c | 72 ++++++++++++++++++++++++++------ 1 file changed, 60 insertions(+), 12 deletions(-) diff --git a/drivers/mtd/nand/raw/vf610_nfc.c b/drivers/mtd/nand/raw/vf610_nfc.c index 9940681810cf..1c3e7b167e53 100644 --- a/drivers/mtd/nand/raw/vf610_nfc.c +++ b/drivers/mtd/nand/raw/vf610_nfc.c @@ -505,6 +505,11 @@ static int vf610_nfc_exec_op(struct nand_chip *chip, check_only); } +static unsigned int vf610_nfc_spare_size(struct mtd_info *mtd) +{ + return min_t(unsigned int, mtd->oobsize, 64); +} + static inline int vf610_nfc_correct_data(struct nand_chip *chip, uint8_t *dat, uint8_t *oob, int page) { @@ -522,7 +527,7 @@ static inline int vf610_nfc_correct_data(struct nand_chip *chip, uint8_t *dat, return ecc_count; nfc->data_access = true; - nand_read_oob_op(&nfc->chip, page, 0, oob, mtd->oobsize); + nand_read_oob_op(&nfc->chip, page, 0, oob, vf610_nfc_spare_size(mtd)); nfc->data_access = false; /* @@ -530,7 +535,7 @@ static inline int vf610_nfc_correct_data(struct nand_chip *chip, uint8_t *dat, * at least less then half of the ECC strength. */ return nand_check_erased_ecc_chunk(dat, nfc->chip.ecc.size, oob, - mtd->oobsize, NULL, 0, + vf610_nfc_spare_size(mtd), NULL, 0, flips_threshold); } @@ -551,7 +556,7 @@ static int vf610_nfc_read_page(struct nand_chip *chip, uint8_t *buf, { struct vf610_nfc *nfc = chip_to_nfc(chip); struct mtd_info *mtd = nand_to_mtd(chip); - int trfr_sz = mtd->writesize + mtd->oobsize; + int trfr_sz = mtd->writesize + vf610_nfc_spare_size(mtd); u32 row = 0, cmd1 = 0, cmd2 = 0, code = 0; int stat; @@ -577,11 +582,16 @@ static int vf610_nfc_read_page(struct nand_chip *chip, uint8_t *buf, */ vf610_nfc_rd_from_sram(buf, nfc->regs + NFC_MAIN_AREA(0), mtd->writesize, false); - if (oob_required) + if (oob_required) { + unsigned int spare = vf610_nfc_spare_size(mtd); + vf610_nfc_rd_from_sram(chip->oob_poi, nfc->regs + NFC_MAIN_AREA(0) + mtd->writesize, - mtd->oobsize, false); + spare, false); + /* Not transferred, and never written: reads back erased */ + memset(chip->oob_poi + spare, 0xff, mtd->oobsize - spare); + } stat = vf610_nfc_correct_data(chip, buf, chip->oob_poi, page); @@ -599,7 +609,7 @@ static int vf610_nfc_write_page(struct nand_chip *chip, const uint8_t *buf, { struct vf610_nfc *nfc = chip_to_nfc(chip); struct mtd_info *mtd = nand_to_mtd(chip); - int trfr_sz = mtd->writesize + mtd->oobsize; + int trfr_sz = mtd->writesize + vf610_nfc_spare_size(mtd); u32 row = 0, cmd1 = 0, cmd2 = 0, code = 0; u8 status; int ret; @@ -740,6 +750,49 @@ static void vf610_nfc_init_controller(struct vf610_nfc *nfc) } } +/* + * With 64 byte OOB chips the core's large page layout matches what + * U-Boot uses, and on chips with more, U-Boot and older kernels clamped + * mtd->oobsize to 64. Modifying the OOB size is no longer possible, the + * actual chip geometry must be respected, so to avoid breaking those + * existing setups use our own layout: the core's large page one, + * computed over the first 64 spare bytes only. + */ +static int vf610_nfc_ooblayout_ecc(struct mtd_info *mtd, int section, + struct mtd_oob_region *oobregion) +{ + struct nand_device *nand = mtd_to_nanddev(mtd); + unsigned int total_ecc_bytes = nand->ecc.ctx.total; + + if (section || !total_ecc_bytes) + return -ERANGE; + + oobregion->length = total_ecc_bytes; + oobregion->offset = vf610_nfc_spare_size(mtd) - oobregion->length; + + return 0; +} + +static int vf610_nfc_ooblayout_free(struct mtd_info *mtd, int section, + struct mtd_oob_region *oobregion) +{ + struct nand_device *nand = mtd_to_nanddev(mtd); + unsigned int total_ecc_bytes = nand->ecc.ctx.total; + + if (section) + return -ERANGE; + + oobregion->length = vf610_nfc_spare_size(mtd) - total_ecc_bytes - 2; + oobregion->offset = 2; + + return 0; +} + +static const struct mtd_ooblayout_ops vf610_nfc_ooblayout_ops = { + .ecc = vf610_nfc_ooblayout_ecc, + .free = vf610_nfc_ooblayout_free, +}; + static int vf610_nfc_attach_chip(struct nand_chip *chip) { struct mtd_info *mtd = nand_to_mtd(chip); @@ -770,12 +823,7 @@ static int vf610_nfc_attach_chip(struct nand_chip *chip) return -ENXIO; } - /* Only 64 byte ECC layouts known */ - if (mtd->oobsize > 64) - mtd->oobsize = 64; - - /* Use default large page ECC layout defined in NAND core */ - mtd_set_ooblayout(mtd, nand_get_large_page_ooblayout()); + mtd_set_ooblayout(mtd, &vf610_nfc_ooblayout_ops); if (chip->ecc.strength == 32) { nfc->ecc_mode = ECC_60_BYTE; chip->ecc.bytes = 60; From 68fe2faf5c69d0e21f94cd695d28f0ba677d484f Mon Sep 17 00:00:00 2001 From: Mehmet Fide Date: Tue, 1 Sep 2026 09:39:07 +0200 Subject: [PATCH 021/694] mtd: rawnand: vf610_nfc: fix false bitflips on reads of erased pages When the ECC engine fails to decode a page, the driver re-reads the OOB area with the engine bypassed, but runs the erased-page check for the data area on the buffer left in the controller SRAM by the failed transfer. That buffer does not hold what is on the flash: the failing engine writes a bogus single-bit "correction" into it. In the 60-byte ECC mode the all-0xff content of an erased page always decodes to the same error location, so every erased page shows one stale zero bit at data offset 0x5FD, which the erased-page check then reports as a corrected bitflip. Edward Karpicz discovered this behaviour and identified the offset on a Colibri VF61; the analysis and the fix build on his finding. Measured with an instrumented driver on a Colibri VF50 (MX30LF1G18AC, 32-bit ECC): reading a 126 MiB partition with nanddump increased the corrected counter by 18035, exactly one per erased page, while raw reads of the same pages return clean 0xff. A v4.4 kernel on the VF61 (MX30LF4G28AC) accumulates the same false counts, so the behaviour follows the controller rather than the chip or the driver generation. Neither the Vybrid reference manual nor the published mask set errata (VFXXX_2N02G) document it. The 45-byte ECC mode is not affected. Restoring the known byte is not enough: on pages that fail to decode with content other than all-0xff the engine writes its correction wherever the syndrome points (measured at a different offset on such a page), so the check has to run on what the flash holds. Re-read the data area with the ECC engine bypassed, exactly as already done for the OOB area. The corrected counter then stays at zero on both boards. Reported-by: Edward Karpicz Link: https://community.toradex.com/t/colibri-vf50-vf61-on-the-current-bsp-mainline-u-boot-v2026-07-and-linux-6-18-lts/30735 Signed-off-by: Mehmet Fide Signed-off-by: Miquel Raynal --- drivers/mtd/nand/raw/vf610_nfc.c | 15 +++++++++++---- 1 file changed, 11 insertions(+), 4 deletions(-) diff --git a/drivers/mtd/nand/raw/vf610_nfc.c b/drivers/mtd/nand/raw/vf610_nfc.c index 1c3e7b167e53..1b9b370adfab 100644 --- a/drivers/mtd/nand/raw/vf610_nfc.c +++ b/drivers/mtd/nand/raw/vf610_nfc.c @@ -519,6 +519,7 @@ static inline int vf610_nfc_correct_data(struct nand_chip *chip, uint8_t *dat, u8 ecc_status; u8 ecc_count; int flips_threshold = nfc->chip.ecc.strength / 2; + int ret; ecc_status = vf610_nfc_read(nfc, ecc_status_off) & 0xff; ecc_count = ecc_status & ECC_STATUS_ERR_COUNT; @@ -526,15 +527,21 @@ static inline int vf610_nfc_correct_data(struct nand_chip *chip, uint8_t *dat, if (!(ecc_status & ECC_STATUS_MASK)) return ecc_count; + /* The failed decode leaves a bogus correction in SRAM; re-read without ECC */ nfc->data_access = true; - nand_read_oob_op(&nfc->chip, page, 0, oob, vf610_nfc_spare_size(mtd)); + ret = nand_read_page_op(&nfc->chip, page, 0, dat, mtd->writesize); + if (!ret) + ret = nand_read_oob_op(&nfc->chip, page, 0, oob, + vf610_nfc_spare_size(mtd)); nfc->data_access = false; + if (ret) + return ret; /* - * On an erased page, bit count (including OOB) should be zero or - * at least less then half of the ECC strength. + * Run the erased-page check with the driver's historic threshold + * of half the ECC strength. */ - return nand_check_erased_ecc_chunk(dat, nfc->chip.ecc.size, oob, + return nand_check_erased_ecc_chunk(dat, mtd->writesize, oob, vf610_nfc_spare_size(mtd), NULL, 0, flips_threshold); } From 21f027016b1290d13c30b198ae7a00e6b3d1d5a5 Mon Sep 17 00:00:00 2001 From: Runyu Xiao Date: Wed, 2 Sep 2026 15:05:42 +0800 Subject: [PATCH 022/694] mtd: rawnand: cadence: Initialize IRQ state before requesting IRQ The Cadence NAND interrupt handler uses both the IRQ lock and completion object. Registering the IRQ before initializing them leaves a window in which a pending interrupt can access uninitialized synchronization state. Initialize them before registering the handler. Fixes: ec4ba01e894d ("mtd: rawnand: Add new Cadence NAND driver to MTD subsystem") Cc: stable@vger.kernel.org Assisted-by: Codex:GPT-5 Signed-off-by: Runyu Xiao Signed-off-by: Miquel Raynal --- drivers/mtd/nand/raw/cadence-nand-controller.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/mtd/nand/raw/cadence-nand-controller.c b/drivers/mtd/nand/raw/cadence-nand-controller.c index d53b35a8b3cb..2085632631c0 100644 --- a/drivers/mtd/nand/raw/cadence-nand-controller.c +++ b/drivers/mtd/nand/raw/cadence-nand-controller.c @@ -3143,6 +3143,9 @@ static int cadence_nand_init(struct cdns_nand_ctrl *cdns_ctrl) goto free_buf_desc; } + spin_lock_init(&cdns_ctrl->irq_lock); + init_completion(&cdns_ctrl->complete); + if (devm_request_irq(cdns_ctrl->dev, cdns_ctrl->irq, cadence_nand_isr, IRQF_SHARED, "cadence-nand-controller", cdns_ctrl)) { @@ -3151,9 +3154,6 @@ static int cadence_nand_init(struct cdns_nand_ctrl *cdns_ctrl) goto free_buf; } - spin_lock_init(&cdns_ctrl->irq_lock); - init_completion(&cdns_ctrl->complete); - ret = cadence_nand_hw_init(cdns_ctrl); if (ret) goto disable_irq; From 3b55f350c68a0aceff108f47f9d31f47ebffaf7b Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Mon, 7 Sep 2026 14:10:24 +0200 Subject: [PATCH 023/694] bpf: Fix bpf_skb_change_tail wrt csum partial skbs Cilium generates ICMP "frag needed" replies from BPF when a LB DSR packet exceeds the egress MTU. The reply is built by first trimming the packet down to target size via bpf_skb_change_tail(), and then pushing the ICMP error headers in front of it. The trim is rejected for skbs which carry a checksum offload, e.g. TCP packets aggregated by GRO on ingress where tcp_gro_complete() leaves the skb as CHECKSUM_PARTIAL. __bpf_skb_min_len() raises the minimum length to the end of the L4 checksum field, so a trim to 42 bytes bails out with -EINVAL given a min_len of 52 in this case, and due to that the ICMP generator fails. This is not the case if GRO is turned off. Fix this bpf_skb_change_tail() restriction and drop the checksum offload when the new length no longer covers the checksum field. The BPF program rewrites the skb into an ICMP error and computes the checksum itself anyway. Fixes: 5293efe62df8 ("bpf: add bpf_skb_change_tail helper") Reported-by: Tom Hadlaw Reported-by: Yusuke Suzuki Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/r/20260907121025.1923656-1-daniel@iogearbox.net Signed-off-by: Alexei Starovoitov --- net/core/filter.c | 11 +++++------ 1 file changed, 5 insertions(+), 6 deletions(-) diff --git a/net/core/filter.c b/net/core/filter.c index 61940e753552..8513167a858a 100644 --- a/net/core/filter.c +++ b/net/core/filter.c @@ -3961,12 +3961,6 @@ static u32 __bpf_skb_min_len(const struct sk_buff *skb) if (offset > 0) min_len = offset; } - if (skb->ip_summed == CHECKSUM_PARTIAL) { - offset = skb_checksum_start_offset(skb) + - skb->csum_offset + sizeof(__sum16); - if (offset > 0) - min_len = offset; - } return min_len; } @@ -3983,6 +3977,11 @@ static int bpf_skb_grow_rcsum(struct sk_buff *skb, unsigned int new_len) static int bpf_skb_trim_rcsum(struct sk_buff *skb, unsigned int new_len) { + if (skb->ip_summed == CHECKSUM_PARTIAL && + new_len < skb_checksum_start_offset(skb) + skb->csum_offset + + sizeof(__sum16)) + skb->ip_summed = CHECKSUM_NONE; + return __skb_trim_rcsum(skb, new_len); } From 15e2565f1c43771af0bc5324971cabaad79ac286 Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Mon, 7 Sep 2026 14:10:25 +0200 Subject: [PATCH 024/694] selftests/bpf: Add test for bpf_skb_change_tail on csum partial skbs Add a test which builds an ICMP error out of a TCP segment. A tcx prog on the client's egress side trims the first data segment down to the target size and pushes the ICMP error headers in front of it to then reflect the packet back to the sender. # LDLIBS=-static PKG_CONFIG='pkg-config --static' ./vmtest.sh -- ./test_progs -t tc_change_tail [...] #509 tc_change_tail:OK #510 tc_change_tail_pmtu:OK Summary: 2/0 PASSED, 0 SKIPPED, 0/0 FAILED Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/r/20260907121025.1923656-2-daniel@iogearbox.net Signed-off-by: Alexei Starovoitov --- .../bpf/prog_tests/tc_change_tail_pmtu.c | 125 +++++++++++++++++ .../bpf/progs/test_tc_change_tail_pmtu.c | 129 ++++++++++++++++++ 2 files changed, 254 insertions(+) create mode 100644 tools/testing/selftests/bpf/prog_tests/tc_change_tail_pmtu.c create mode 100644 tools/testing/selftests/bpf/progs/test_tc_change_tail_pmtu.c diff --git a/tools/testing/selftests/bpf/prog_tests/tc_change_tail_pmtu.c b/tools/testing/selftests/bpf/prog_tests/tc_change_tail_pmtu.c new file mode 100644 index 000000000000..7acdbd5757a9 --- /dev/null +++ b/tools/testing/selftests/bpf/prog_tests/tc_change_tail_pmtu.c @@ -0,0 +1,125 @@ +// SPDX-License-Identifier: GPL-2.0 + +#include + +#include "test_progs.h" +#include "network_helpers.h" +#include "test_tc_change_tail_pmtu.skel.h" + +#define CLIENT_NS "tc-change-tail-cli-ns" +#define SERVER_NS "tc-change-tail-srv-ns" +#define CLIENT_IP "192.168.1.1" +#define SERVER_IP "192.168.1.2" + +#define TEST_PMTU 1000 +#define TEST_MSS_MAX (TEST_PMTU - 20 - 20) +#define TIMEOUT_MS 3000 +#define XFER_BYTES 8192 + +void test_tc_change_tail_pmtu(void) +{ + LIBBPF_OPTS(bpf_tcx_opts, tcx_opts); + int mss_before = 0, mss_after = 0, ifindex, port; + int srv_fd = -1, srv_conn_fd = -1, cli_fd = -1; + struct test_tc_change_tail_pmtu *skel = NULL; + struct nstoken *nstoken = NULL; + static char buf[XFER_BYTES]; + socklen_t optlen; + ssize_t bytes; + size_t total; + + if (!ASSERT_OK(make_netns(CLIENT_NS), "make client ns")) + return; + if (!ASSERT_OK(make_netns(SERVER_NS), "make server ns")) + goto out_client_ns; + + nstoken = open_netns(CLIENT_NS); + if (!ASSERT_OK_PTR(nstoken, "open client ns")) + goto out; + SYS(out, "ip link add veth1 type veth peer name veth2 netns " SERVER_NS); + SYS(out, "ip -4 addr add " CLIENT_IP "/24 dev veth1"); + SYS(out, "ip link set veth1 up"); + ifindex = if_nametoindex("veth1"); + if (!ASSERT_NEQ(ifindex, 0, "if_nametoindex")) + goto out; + close_netns(nstoken); + nstoken = NULL; + + nstoken = open_netns(SERVER_NS); + if (!ASSERT_OK_PTR(nstoken, "open server ns")) + goto out; + SYS(out, "ip -4 addr add " SERVER_IP "/24 dev veth2"); + SYS(out, "ip link set veth2 up"); + srv_fd = start_server(AF_INET, SOCK_STREAM, SERVER_IP, 0, TIMEOUT_MS); + if (!ASSERT_OK_FD(srv_fd, "start server")) + goto out; + close_netns(nstoken); + nstoken = NULL; + + skel = test_tc_change_tail_pmtu__open_and_load(); + if (!ASSERT_OK_PTR(skel, "open and load skeleton")) + goto out; + + port = get_socket_local_port(srv_fd); + if (!ASSERT_GE(port, 0, "get server port")) + goto out; + + skel->bss->server_port = port; + skel->bss->pmtu = TEST_PMTU; + + nstoken = open_netns(CLIENT_NS); + if (!ASSERT_OK_PTR(nstoken, "open client ns")) + goto out; + + skel->links.change_tail_icmp = + bpf_program__attach_tcx(skel->progs.change_tail_icmp, ifindex, + &tcx_opts); + if (!ASSERT_OK_PTR(skel->links.change_tail_icmp, "attach tcx")) + goto out; + + cli_fd = connect_to_fd(srv_fd, TIMEOUT_MS); + if (!ASSERT_OK_FD(cli_fd, "connect to server")) + goto out; + srv_conn_fd = accept(srv_fd, NULL, NULL); + if (!ASSERT_OK_FD(srv_conn_fd, "accept connection")) + goto out; + if (!ASSERT_OK(settimeo(srv_conn_fd, TIMEOUT_MS), "set server timeout")) + goto out; + + optlen = sizeof(mss_before); + if (!ASSERT_OK(getsockopt(cli_fd, IPPROTO_TCP, TCP_MAXSEG, &mss_before, + &optlen), "get mss before")) + goto out; + + bytes = send(cli_fd, buf, sizeof(buf), 0); + if (!ASSERT_EQ(bytes, (ssize_t)sizeof(buf), "send data")) + goto out; + + for (total = 0; total < sizeof(buf); total += bytes) { + bytes = recv(srv_conn_fd, buf, sizeof(buf), 0); + if (bytes <= 0) + break; + } + + ASSERT_EQ(total, sizeof(buf), "receive data"); + ASSERT_OK(skel->data->change_tail_ret, "change tail"); + ASSERT_OK(skel->bss->adjust_room_ret, "adjust room"); + ASSERT_TRUE(skel->bss->icmp_sent, "icmp sent"); + + optlen = sizeof(mss_after); + if (!ASSERT_OK(getsockopt(cli_fd, IPPROTO_TCP, TCP_MAXSEG, &mss_after, + &optlen), "get mss after")) + goto out; + + ASSERT_LT(mss_after, mss_before, "mss reduced"); + ASSERT_LE(mss_after, TEST_MSS_MAX, "mss below pmtu"); +out: + close(srv_conn_fd); + close(cli_fd); + close(srv_fd); + test_tc_change_tail_pmtu__destroy(skel); + close_netns(nstoken); + remove_netns(SERVER_NS); +out_client_ns: + remove_netns(CLIENT_NS); +} diff --git a/tools/testing/selftests/bpf/progs/test_tc_change_tail_pmtu.c b/tools/testing/selftests/bpf/progs/test_tc_change_tail_pmtu.c new file mode 100644 index 000000000000..5c4c07545bc9 --- /dev/null +++ b/tools/testing/selftests/bpf/progs/test_tc_change_tail_pmtu.c @@ -0,0 +1,129 @@ +// SPDX-License-Identifier: GPL-2.0 + +#include +#include + +#include +#include +#include +#include +#include +#include + +#include +#include + +#define ICMP_SAMPLE_LEN (sizeof(struct iphdr) + 8) +#define ICMP_HDRS_LEN (sizeof(struct iphdr) + sizeof(struct icmphdr)) + +__be16 server_port = 0; +__u16 pmtu = 0; + +long change_tail_ret = 1; +long adjust_room_ret = 0; +bool icmp_sent = false; +bool icmp_err = false; + +static __always_inline __sum16 csum_fold(__wsum csum) +{ + csum = (csum & 0xffff) + (csum >> 16); + csum = (csum & 0xffff) + (csum >> 16); + + return (__sum16)~csum; +} + +SEC("tc/egress") +int change_tail_icmp(struct __sk_buff *skb) +{ + __u8 smac[ETH_ALEN], dmac[ETH_ALEN]; + void *data, *data_end; + struct icmphdr *icmp; + struct ethhdr *eth; + struct tcphdr *tcp; + __be32 saddr, daddr; + struct iphdr *ip; + __wsum csum; + + if (icmp_sent || icmp_err) + return TCX_PASS; + + data = (void *)(long)skb->data; + data_end = (void *)(long)skb->data_end; + + eth = data; + if ((void *)(eth + 1) > data_end) + return TCX_PASS; + if (eth->h_proto != bpf_htons(ETH_P_IP)) + return TCX_PASS; + + ip = (void *)(eth + 1); + if ((void *)(ip + 1) > data_end) + return TCX_PASS; + if (ip->ihl != 5 || ip->protocol != IPPROTO_TCP) + return TCX_PASS; + + tcp = (void *)(ip + 1); + if ((void *)(tcp + 1) > data_end) + return TCX_PASS; + if (tcp->dest != server_port) + return TCX_PASS; + if (bpf_ntohs(ip->tot_len) <= sizeof(*ip) + tcp->doff * 4) + return TCX_PASS; + + __builtin_memcpy(smac, eth->h_source, ETH_ALEN); + __builtin_memcpy(dmac, eth->h_dest, ETH_ALEN); + saddr = ip->saddr; + daddr = ip->daddr; + + change_tail_ret = bpf_skb_change_tail(skb, ETH_HLEN + ICMP_SAMPLE_LEN, 0); + if (change_tail_ret) { + icmp_err = true; + return TCX_PASS; + } + + adjust_room_ret = bpf_skb_adjust_room(skb, ICMP_HDRS_LEN, + BPF_ADJ_ROOM_MAC, + BPF_F_ADJ_ROOM_NO_CSUM_RESET); + if (adjust_room_ret) { + icmp_err = true; + return TCX_DROP; + } + + data = (void *)(long)skb->data; + data_end = (void *)(long)skb->data_end; + + eth = data; + ip = (void *)(eth + 1); + icmp = (void *)(ip + 1); + if ((void *)icmp + sizeof(*icmp) + ICMP_SAMPLE_LEN > data_end) { + icmp_err = true; + return TCX_DROP; + } + + __builtin_memcpy(eth->h_dest, smac, ETH_ALEN); + __builtin_memcpy(eth->h_source, dmac, ETH_ALEN); + + __builtin_memset(icmp, 0, sizeof(*icmp)); + icmp->type = ICMP_DEST_UNREACH; + icmp->code = ICMP_FRAG_NEEDED; + icmp->un.frag.mtu = bpf_htons(pmtu); + + __builtin_memset(ip, 0, sizeof(*ip)); + ip->version = 4; + ip->ihl = 5; + ip->ttl = 64; + ip->protocol = IPPROTO_ICMP; + ip->tot_len = bpf_htons(ICMP_HDRS_LEN + ICMP_SAMPLE_LEN); + ip->saddr = daddr; + ip->daddr = saddr; + + csum = bpf_csum_diff(NULL, 0, (__be32 *)icmp, + sizeof(*icmp) + ICMP_SAMPLE_LEN, 0); + icmp->checksum = csum_fold(csum); + csum = bpf_csum_diff(NULL, 0, (__be32 *)ip, sizeof(*ip), 0); + ip->check = csum_fold(csum); + icmp_sent = true; + return bpf_redirect(skb->ifindex, BPF_F_INGRESS); +} + +char _license[] SEC("license") = "GPL"; From 86ad6489e38c76c06c0b4a4229391a794023b5e0 Mon Sep 17 00:00:00 2001 From: Julian Braha Date: Sun, 16 Aug 2026 15:22:07 +0100 Subject: [PATCH 025/694] ARC: cleanup dead ARC_CANT_LLSC option in Kconfig Currently, the ARC_CANT_LLSC config option can never be enabled. It is also not referenced anywhere else. So let's remove this dead code. Note that I have only compile-tested these changes, because I do not have the hardware to runtime test. This dead code was found by kconfirm, a static analysis tool for Kconfig. Signed-off-by: Julian Braha Signed-off-by: Vineet Gupta --- arch/arc/Kconfig | 4 ---- 1 file changed, 4 deletions(-) diff --git a/arch/arc/Kconfig b/arch/arc/Kconfig index 2ed7186c81c5..80e61175bf52 100644 --- a/arch/arc/Kconfig +++ b/arch/arc/Kconfig @@ -331,13 +331,9 @@ config ARC_FPU_SAVE_RESTORE for floating point exceptions and rounding modes. These are preserved across task context switch when enabled. -config ARC_CANT_LLSC - def_bool n - config ARC_HAS_LLSC bool "Insn: LLOCK/SCOND (efficient atomic ops)" default y - depends on !ARC_CANT_LLSC if ISA_ARCV2 From 04eb7b5e043fbccf83d27a9be5c3973c2dbb110f Mon Sep 17 00:00:00 2001 From: Anthony Iliopoulos Date: Thu, 20 Aug 2026 15:13:39 +0200 Subject: [PATCH 026/694] arc: remove unused profile.h includes None of the code in these files uses declarations provided by profile.h. Remove the unnecessary header includes across the various files to clean up dependencies and reduce include bloat. Signed-off-by: Anthony Iliopoulos Signed-off-by: Vineet Gupta --- arch/arc/kernel/smp.c | 1 - 1 file changed, 1 deletion(-) diff --git a/arch/arc/kernel/smp.c b/arch/arc/kernel/smp.c index 2d99dffed0ce..ee210a94ce45 100644 --- a/arch/arc/kernel/smp.c +++ b/arch/arc/kernel/smp.c @@ -12,7 +12,6 @@ #include #include #include -#include #include #include #include From d12c6a6f0c8fd5b5b41120f7be319f743dae3b4f Mon Sep 17 00:00:00 2001 From: blaze <1466528493@qq.com> Date: Wed, 9 Sep 2026 12:42:59 +0800 Subject: [PATCH 027/694] arc: kernel: Fix clk reference leak in show_cpuinfo() In show_cpuinfo(), clk_get() acquires a reference to the CPU clock, but clk_put() is never called to release it. Since show_cpuinfo() is the seq_file show callback for /proc/cpuinfo, this causes a clock reference count leak every time /proc/cpuinfo is read. The frequency is extracted via clk_get_rate() into a local variable, so the clock reference can be safely released immediately after. Fix this by calling clk_put(cpu_clk) right after clk_get_rate(). Fixes: 98f2892607c5 ("ARC: [cpuinfo] Add clk info in /proc/cpuinfo") Signed-off-by: blaze <1466528493@qq.com> Signed-off-by: Vineet Gupta --- arch/arc/kernel/setup.c | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/arc/kernel/setup.c b/arch/arc/kernel/setup.c index 7b6a9beba9db..8bebd0e6fc2f 100644 --- a/arch/arc/kernel/setup.c +++ b/arch/arc/kernel/setup.c @@ -593,6 +593,7 @@ static int show_cpuinfo(struct seq_file *m, void *v) cpu_id); } else { freq = clk_get_rate(cpu_clk); + clk_put(cpu_clk); } if (freq) seq_printf(m, "CPU speed\t: %lu.%02lu Mhz\n", From e4a62833adff6ef0fe7c0b90393204fe3c26b5c5 Mon Sep 17 00:00:00 2001 From: Weiming Shi Date: Wed, 9 Sep 2026 12:08:08 +0800 Subject: [PATCH 028/694] bpf: Disallow bpf_skb_pull_data() for LWT_SEG6LOCAL An LWT_SEG6LOCAL program can invalidate its cached SRH with bpf_lwt_seg6_adjust_srh() and then call bpf_skb_pull_data(). The latter may reallocate skb->head, leaving the per-CPU SRH pointer dangling. Post-program SRH validation then writes through that pointer. Disallow bpf_skb_pull_data() for LWT_SEG6LOCAL programs so the verifier rejects this unsafe helper combination. Other LWT program types continue to expose the helper through lwt_out_func_proto(). Fixes: 004d4b274e2a ("ipv6: sr: Add seg6local action End.BPF") Reported-by: co+adfca3e91be95776@bugs.sh Suggested-by: Alexei Starovoitov Signed-off-by: Weiming Shi Signed-off-by: Daniel Borkmann Reviewed-by: Emil Tsalapatis Closes: https://lore.kernel.org/all/GCy0KRM2IcQGoJQTjJEU9D0maBxXzEDHuQpq@bugs.sh/ Link: https://lore.kernel.org/bpf/DL9COXZQXX4V.1FN45QO2Q77ZH@gmail.com/ Link: https://lore.kernel.org/bpf/20260909040807.3885815-2-bestswngs@gmail.com --- net/core/filter.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/net/core/filter.c b/net/core/filter.c index 8513167a858a..2a84f9d01131 100644 --- a/net/core/filter.c +++ b/net/core/filter.c @@ -9044,6 +9044,8 @@ static const struct bpf_func_proto * lwt_seg6local_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog) { switch (func_id) { + case BPF_FUNC_skb_pull_data: + return NULL; #if IS_ENABLED(CONFIG_IPV6_SEG6_BPF) case BPF_FUNC_lwt_seg6_store_bytes: return &bpf_lwt_seg6_store_bytes_proto; From d22c3e0088e85be8131f7a9283f759cdbb20726d Mon Sep 17 00:00:00 2001 From: Sven Schnelle Date: Wed, 9 Sep 2026 11:29:53 +0200 Subject: [PATCH 029/694] selftests/ftrace: Fix unique symbol check in kprobe_non_uniq_symbol.tc The current regex also matches symbols in modules, which makes the test fail on s390 where name_show is present only once in the kernel, but also multiple times in modules: 000001b1401cdc20 t name_show 000001b0c05e6c40 t name_show [mdev] 000001b0c0495f30 t name_show [i2c_core] Fix this by changing the regular expression to only match the function name. Link: https://lore.kernel.org/all/20260909092954.2200558-1-svens@linux.ibm.com/ Fixes: 03b80ff8023a ("selftests/ftrace: Add new test case which checks non unique symbol") Signed-off-by: Sven Schnelle Reviewed-by: Steven Rostedt Signed-off-by: Masami Hiramatsu (Google) --- .../selftests/ftrace/test.d/kprobe/kprobe_non_uniq_symbol.tc | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/testing/selftests/ftrace/test.d/kprobe/kprobe_non_uniq_symbol.tc b/tools/testing/selftests/ftrace/test.d/kprobe/kprobe_non_uniq_symbol.tc index bc9514428dba..07b1177c1634 100644 --- a/tools/testing/selftests/ftrace/test.d/kprobe/kprobe_non_uniq_symbol.tc +++ b/tools/testing/selftests/ftrace/test.d/kprobe/kprobe_non_uniq_symbol.tc @@ -6,7 +6,7 @@ SYMBOL='name_show' # We skip this test on kernel where SYMBOL is unique or does not exist. -if [ "$(grep -c -E "[[:alnum:]]+ t ${SYMBOL}" /proc/kallsyms)" -le '1' ]; then +if [ "$(grep -c -E "[[:alnum:]]+ t ${SYMBOL}$" /proc/kallsyms)" -le '1' ]; then exit_unsupported fi From 57a78ad2305f299bc3f933ed36b3d402df04784a Mon Sep 17 00:00:00 2001 From: David Howells Date: Wed, 9 Sep 2026 09:06:31 +0100 Subject: [PATCH 030/694] block: Fix start and length check added to iov_iter_extract_bvecs() Commit 14b007e17881 added an address check using iter_iov_addr() and a length check using iter_iov_len() to iov_iter_extract_bvecs(), but these cannot be used so and are unsafe in this circumstance as the functions have hardwired assumptions about the iterator type. They should only be used with ITER_UBUF or ITER_IOVEC-type iterators; they shouldn't be used with ITER_BVEC, ITER_KVEC, ITER_FOLIOQ, ITER_XARRAY or ITER_DISCARD iterators. This proves to be a problem for cachefiles as an iterator of type ITER_FOLIOQ is passed and iter_iov_addr() and iter_iov_len() both malfunction because iter->__iov in iter_iov() is not pointing to an iovec array. Fix this by using iov_iter_alignment() instead. Fixes: 14b007e17881 ("block: validate user space vectors during extraction") Signed-off-by: David Howells Link: https://patch.msgid.link/1667275.1788941191@warthog.procyon.org.uk Reviewed-by: Keith Busch Reviewed-by: Christoph Hellwig cc: Hannes Reinecke cc: Christoph Hellwig cc: Jens Axboe cc: Alexander Viro cc: Paulo Alcantara cc: netfs@lists.linux.dev cc: linux-block@vger.kernel.org cc: linux-fsdevel@vger.kernel.org Signed-off-by: Christian Brauner (Amutable) --- lib/iov_iter.c | 18 ++++++++++++++++-- 1 file changed, 16 insertions(+), 2 deletions(-) diff --git a/lib/iov_iter.c b/lib/iov_iter.c index 6665372ecf71..2072c04e99d0 100644 --- a/lib/iov_iter.c +++ b/lib/iov_iter.c @@ -1921,15 +1921,29 @@ ssize_t iov_iter_extract_bvecs(struct iov_iter *iter, struct bio_vec *bv, unsigned short max_vecs, unsigned mem_align_mask, iov_iter_extraction_t extraction_flags) { - unsigned long start = (unsigned long)iter_iov_addr(iter); unsigned short entries_left = max_vecs - *nr_vecs; unsigned short nr_pages, i = 0; size_t left, offset, len; struct page **pages; ssize_t size; - if ((start | iter_iov_len(iter)) & mem_align_mask) + /* + * DMA engines typically have both memory address and length alignment + * requirements, so check these against the alignment mask. For UBUF, + * IOVEC and KVEC, only the current segment will be extracted from; for + * everything else we might extract from multiple segments, so we need + * to check those too. + */ + if (likely(iter_is_ubuf(iter) || + iter_is_iovec(iter) || + iov_iter_is_kvec(iter))) { + unsigned long start = (unsigned long)iter_iov_addr(iter); + + if ((start | iter_iov_len(iter)) & mem_align_mask) + return -EINVAL; + } else if (iov_iter_alignment(iter) & mem_align_mask) { return -EINVAL; + } /* * Move page array up in the allocated memory for the bio vecs as far as From d59ac79915daa567c69aa93d458d41844676e92a Mon Sep 17 00:00:00 2001 From: Disha Goel Date: Fri, 3 Jul 2026 20:37:42 +0530 Subject: [PATCH 031/694] selftests/filesystems: fix missing and stale TARGETS entries filesystems/eventfd, filesystems/open_tree_ns and filesystems/xattr were never added to TARGETS when introduced. filesystems/openat2 was moved from selftests/openat2/ but the TARGETS entry was never updated, leaving a stale entry pointing at a directory that no longer exists. Fix this by adding the four missing subdirectories to TARGETS and removing the stale openat2 entry. Link: https://lore.kernel.org/20260703150742.58991-1-disgoel@linux.ibm.com Fixes: 7c37857fc23a ("selftests: add eventfd selftests") Fixes: b8f7622aa6e3 ("selftests/open_tree: add OPEN_TREE_NAMESPACE tests") Fixes: 7e28fef5d4db ("selftests/xattr: path-based AF_UNIX socket xattr tests") Fixes: fe087927046c ("selftests: move openat2 tests to selftests/filesystems/") Signed-off-by: Disha Goel Reviewed-by: Christian Brauner (Amutable) Cc: "Darrick J. Wong" Cc: Jan Kara Cc: Jeff Layton Cc: Shuah Khan Cc: Wen Yang Signed-off-by: Andrew Morton Link: https://patch.msgid.link/20260904183659.B81CD1F00A3D@smtp.kernel.org Signed-off-by: Christian Brauner (Amutable) --- tools/testing/selftests/Makefile | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/tools/testing/selftests/Makefile b/tools/testing/selftests/Makefile index 2d960626750e..330d061f6366 100644 --- a/tools/testing/selftests/Makefile +++ b/tools/testing/selftests/Makefile @@ -35,8 +35,11 @@ TARGETS += fchmodat2 TARGETS += filesystems TARGETS += filesystems/binderfs TARGETS += filesystems/epoll +TARGETS += filesystems/eventfd TARGETS += filesystems/failfs TARGETS += filesystems/fat +TARGETS += filesystems/openat2 +TARGETS += filesystems/open_tree_ns TARGETS += filesystems/overlayfs TARGETS += filesystems/statmount TARGETS += filesystems/mount-notify @@ -47,6 +50,7 @@ TARGETS += filesystems/empty_mntns TARGETS += filesystems/fsmount_ns TARGETS += filesystems/fscontext_ns TARGETS += filesystems/mntns_cleanup +TARGETS += filesystems/xattr TARGETS += firmware TARGETS += fpu TARGETS += ftrace @@ -103,7 +107,6 @@ TARGETS += prctl TARGETS += proc TARGETS += pstore TARGETS += ptrace -TARGETS += openat2 TARGETS += rdma TARGETS += resctrl TARGETS += riscv From 1abd643f3783ea8f8e273c18697ff0413aa92dc7 Mon Sep 17 00:00:00 2001 From: Christian Brauner Date: Wed, 9 Sep 2026 11:03:18 +0200 Subject: [PATCH 032/694] fs/ntfs3: use d_instantiate_new() in ntfs_create_inode() and murder syzbot's "WARNING in do_new_mount" saga ntfs_create_inode() creates a new inode via ntfs_new_inode(). It hashes it with insert_inode_locked() and so it's marked as I_NEW until unlock_new_inode(). ntfs 3 calls d_instantiate() in between though... Since the dentry was already hashed by the lookup before the create any path walk finds it without touching the parent's i_rwsem and so can lock the inode. If the inode is a directory unlock_new_inode() calls lockdep_annotate_inode_mutex_key() and marks i_rwsem with the i_mutex_dir_key class. That resets the count and the owner of a lock somebody else may already hold by now... syzbot has been spamming us with the same godforsaken bug "WARNING in do_new_mount" since 2023. I can't take it anymore so I went looking. Afaict, syzbot's executor chdirs into a freshly mounted ntfs3 image, creates a directory and then mounts some pseudofs on it. Everytime the mkdir() takes longer than syzbot waits mount() runs concurrently: mkdir("./sys") mount(NULL, "./sys", "sysfs") ntfs_create_inode() d_instantiate() user_path_at() finds the dentry do_lock_mount() inode_lock(inode) namespace_lock() unlock_new_inode() lockdep_annotate_inode_mutex_key() init_rwsem(&inode->i_rwsem) unlock_mount() inode_unlock(inode) The mount side then releases a lock that according to the rwsem nobody holds: DEBUG_RWSEMS_WARN_ON((rwsem_owner(sem) != current) && ...): count = 0x0, magic = 0xffff888043a854e8, owner = 0x0, curr 0xffff888000244880, list empty WARNING: CPU: 0 PID: 5346 at kernel/locking/rwsem.c:1368 __up_write Call Trace: inode_unlock include/linux/fs.h:877 [inline] unlock_mount fs/namespace.c:2892 [inline] do_new_mount_fc fs/namespace.c:3828 [inline] do_new_mount+0x777/0xa40 fs/namespace.c:3887 On PREEMPT_RT the same thing shows up as DEBUG_LOCKS_WARN_ON(rt_mutex_owner(lock) != current) WARNING: kernel/locking/rtmutex_common.h:193 at rt_mutex_slowunlock The up_write() underflows the reset count. A following inode_lock() on that directory then never returns. A path walk into the new directory racing with the mkdir() corrupts the lock the same way via inode_lock_shared() in lookup_slow(). Switch to d_instantiate_new() and drop the trailing unlock_new_inode(). All error paths bail out before that point with I_NEW still set and keep using discard_new_inode(). May we never see this fscking bug report again. Link: https://patch.msgid.link/20260909-work-ntfs3-d_instantiate_new-v1-1-2db697162ce8@kernel.org Fixes: 82cae269cfa9 ("fs/ntfs3: Add initialization of super block") Reviewed-by: Jan Kara Cc: stable@vger.kernel.org # v5.15+ Reported-by: syzbot+2a13ad6914e6fcec716c@syzkaller.appspotmail.com Closes: https://lore.kernel.org/6a9beced.a5e650b3.26d8a.000b.GAE@google.com Signed-off-by: Christian Brauner (Amutable) --- fs/ntfs3/inode.c | 7 ++----- 1 file changed, 2 insertions(+), 5 deletions(-) diff --git a/fs/ntfs3/inode.c b/fs/ntfs3/inode.c index 56b4f6469a28..4ac26c80bd34 100644 --- a/fs/ntfs3/inode.c +++ b/fs/ntfs3/inode.c @@ -1866,10 +1866,10 @@ int ntfs_create_inode(struct mnt_idmap *idmap, struct inode *dir, goto out6; /* - * Call 'd_instantiate' after inode->i_op is set + * Call 'd_instantiate_new' after inode->i_op is set * but before finish_open. */ - d_instantiate(dentry, inode); + d_instantiate_new(dentry, inode); /* Set original time. inode times (i_ctime) may be changed in ntfs_init_acl. */ inode_set_atime_to_ts(inode, ni->i_crtime); @@ -1917,9 +1917,6 @@ int ntfs_create_inode(struct mnt_idmap *idmap, struct inode *dir, if (!fnd) ni_unlock(dir_ni); - if (!err) - unlock_new_inode(inode); - return err; } From 88ce88e933e40f4ab4b81a5e9d0a10328583f593 Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Thu, 10 Sep 2026 23:35:07 +0200 Subject: [PATCH 033/694] bpf: Add KF_PERFMON kfunc flag Tracing related BPF helpers e.g. under bpf_base_func_proto() are gated behind CAP_PERFMON. However, the same is currently not true for kfuncs and they are accessible via plain CAP_BPF. Add a new KF_PERFMON flag which can be used such that check_kfunc_call() ensures env->allow_ptr_leaks is permitted. This follows similar pattern to existing KF_DESTRUCTIVE flag. The rejection returns -EPERM to match the other CAP_PERFMON gates in the verifier, that is, check_ptr_to_btf_access() and check_ptr_to_map_access(), which report the very same policy to user space. Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/r/20260910213510.49358-1-daniel@iogearbox.net Signed-off-by: Alexei Starovoitov --- Documentation/bpf/kfuncs.rst | 10 ++++++++++ include/linux/btf.h | 1 + kernel/bpf/verifier.c | 14 ++++++++++++++ 3 files changed, 25 insertions(+) diff --git a/Documentation/bpf/kfuncs.rst b/Documentation/bpf/kfuncs.rst index 85f73e0bbd0f..6691fe8a32c3 100644 --- a/Documentation/bpf/kfuncs.rst +++ b/Documentation/bpf/kfuncs.rst @@ -486,6 +486,16 @@ Example usage in BPF program: /* note that the last argument is omitted */ bpf_task_work_schedule_signal(task, &work->tw, &arrmap, task_work_callback); +2.5.10 KF_PERFMON flag +---------------------- + +The KF_PERFMON flag is used for kfuncs that can expose kernel memory or kernel +addresses to the BPF program, for example by reading through a pointer that the +verifier does not check. Calling such a kfunc requires CAP_PERFMON, or +CAP_SYS_ADMIN, in the same way that the equivalent BPF helpers are gated in +bpf_base_func_proto(). A program loaded with CAP_BPF alone is rejected at load +time. + 2.6 Registering the kfuncs -------------------------- diff --git a/include/linux/btf.h b/include/linux/btf.h index 89d5a5c4f117..7c62ea17b116 100644 --- a/include/linux/btf.h +++ b/include/linux/btf.h @@ -80,6 +80,7 @@ #define KF_ARENA_ARG2 (1 << 15) /* kfunc takes an arena pointer as its second argument */ #define KF_IMPLICIT_ARGS (1 << 16) /* kfunc has implicit arguments supplied by the verifier */ #define KF_SPINLOCK_SAFE (1 << 17) /* kfunc is allowed inside bpf_spin_lock-ed region */ +#define KF_PERFMON (1 << 18) /* kfunc requires CAP_PERFMON */ /* * Tag marking a kernel function as a kfunc. This is meant to minimize the diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 72a3f5998dd2..939e535a3442 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -11356,6 +11356,11 @@ static bool is_kfunc_destructive(struct bpf_call_arg_meta *meta) return meta->kfunc_flags & KF_DESTRUCTIVE; } +static bool is_kfunc_perfmon(struct bpf_call_arg_meta *meta) +{ + return meta->kfunc_flags & KF_PERFMON; +} + static bool is_kfunc_rcu(struct bpf_call_arg_meta *meta) { return meta->kfunc_flags & KF_RCU; @@ -13834,6 +13839,15 @@ static int check_kfunc_call(struct bpf_verifier_env *env, struct bpf_insn *insn, return -EACCES; } + if (is_kfunc_perfmon(&meta) && !env->allow_ptr_leaks) { + verbose(env, "%s is allowed only to CAP_PERFMON and CAP_SYS_ADMIN\n", + func_name); + operation = bpf_diag_fmt(env, "kfunc %s", func_name); + bpf_diag_policy(env, insn_idx, operation, "the kfunc requires CAP_PERFMON", + "Load the program with CAP_PERFMON, or avoid the kfunc."); + return -EPERM; + } + sleepable = bpf_is_kfunc_sleepable(&meta); if (sleepable && !in_sleepable(env)) { verbose(env, "program must be sleepable to call sleepable kfunc %s\n", func_name); From 81c975aae375d2053fa2926ee3730f02e713140c Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Thu, 10 Sep 2026 23:35:08 +0200 Subject: [PATCH 034/694] bpf: Require CAP_PERFMON for kfuncs reading memory Mark fault-safe probe reading kfuncs as KF_PERFMON, similarly as we do for the old-style BPF helper equivalents. bpf_rdonly_cast() is included in this list as well as it returns PTR_TO_MEM | MEM_RDONLY | PTR_UNTRUSTED for an unchecked object and is using fault-safe BPF_PROBE_MEM. Note that only the void form of bpf_rdonly_cast() produced a register that was readable without CAP_PERFMON. For a struct type id the kfunc returns PTR_TO_BTF_ID | PTR_UNTRUSTED, whose dereference has always been gated in check_ptr_to_btf_access(). The flag is not conditional on the type id, so for the latter it only moves the rejection from the dereference to the call itself, which is the better place to report it anyway. The bpf_stream_vprintk() and bpf_stream_print_stack() kfuncs are marked as well. The former ends up in the same bpf_bprintf_prepare() as the bpf_snprintf() helper, where %pks, %pus and %pI4 read through a program- supplied address and %pB resolves one into a symbol. The latter walks the stack and prints each instruction pointer via %pS. Lastly, bpf_get_kmem_cache() takes a raw scalar address that the verifier does not constrain and dereferences the page/slab metadata for it. Field reads on the returned pointer are still blocked (PTR_TO_BTF_ID | PTR_UNTRUSTED -> check_ptr_to_btf_access() results in -EPERM without the CAP), but the NULL/non-NULL result is observable. Reported-by: STAR Labs SG Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/r/20260910213510.49358-2-daniel@iogearbox.net Signed-off-by: Alexei Starovoitov --- kernel/bpf/helpers.c | 60 ++++++++++++++++++++++---------------------- 1 file changed, 30 insertions(+), 30 deletions(-) diff --git a/kernel/bpf/helpers.c b/kernel/bpf/helpers.c index b3cc5c8fc875..712dca5a2c5b 100644 --- a/kernel/bpf/helpers.c +++ b/kernel/bpf/helpers.c @@ -4883,7 +4883,7 @@ BTF_ID(func, bpf_cgroup_release_dtor) BTF_KFUNCS_START(common_btf_ids) BTF_ID_FLAGS(func, bpf_cast_to_kern_ctx, KF_FASTCALL) -BTF_ID_FLAGS(func, bpf_rdonly_cast, KF_FASTCALL) +BTF_ID_FLAGS(func, bpf_rdonly_cast, KF_FASTCALL | KF_PERFMON) BTF_ID_FLAGS(func, bpf_rcu_read_lock) BTF_ID_FLAGS(func, bpf_rcu_read_unlock) BTF_ID_FLAGS(func, bpf_dynptr_slice, KF_RET_NULL) @@ -4920,26 +4920,26 @@ BTF_ID_FLAGS(func, bpf_wq_set_callback, KF_IMPLICIT_ARGS) BTF_ID_FLAGS(func, bpf_wq_start) BTF_ID_FLAGS(func, bpf_preempt_disable) BTF_ID_FLAGS(func, bpf_preempt_enable) -BTF_ID_FLAGS(func, bpf_iter_bits_new, KF_ITER_NEW) +BTF_ID_FLAGS(func, bpf_iter_bits_new, KF_ITER_NEW | KF_PERFMON) BTF_ID_FLAGS(func, bpf_iter_bits_next, KF_ITER_NEXT | KF_RET_NULL) BTF_ID_FLAGS(func, bpf_iter_bits_destroy, KF_ITER_DESTROY) -BTF_ID_FLAGS(func, bpf_copy_from_user_str, KF_SLEEPABLE) -BTF_ID_FLAGS(func, bpf_copy_from_user_task_str, KF_SLEEPABLE) -BTF_ID_FLAGS(func, bpf_get_kmem_cache) +BTF_ID_FLAGS(func, bpf_copy_from_user_str, KF_SLEEPABLE | KF_PERFMON) +BTF_ID_FLAGS(func, bpf_copy_from_user_task_str, KF_SLEEPABLE | KF_PERFMON) +BTF_ID_FLAGS(func, bpf_get_kmem_cache, KF_PERFMON) BTF_ID_FLAGS(func, bpf_iter_kmem_cache_new, KF_ITER_NEW | KF_SLEEPABLE) BTF_ID_FLAGS(func, bpf_iter_kmem_cache_next, KF_ITER_NEXT | KF_RET_NULL | KF_SLEEPABLE) BTF_ID_FLAGS(func, bpf_iter_kmem_cache_destroy, KF_ITER_DESTROY | KF_SLEEPABLE) BTF_ID_FLAGS(func, bpf_local_irq_save) BTF_ID_FLAGS(func, bpf_local_irq_restore) #ifdef CONFIG_BPF_EVENTS -BTF_ID_FLAGS(func, bpf_probe_read_user_dynptr) -BTF_ID_FLAGS(func, bpf_probe_read_kernel_dynptr) -BTF_ID_FLAGS(func, bpf_probe_read_user_str_dynptr) -BTF_ID_FLAGS(func, bpf_probe_read_kernel_str_dynptr) -BTF_ID_FLAGS(func, bpf_copy_from_user_dynptr, KF_SLEEPABLE) -BTF_ID_FLAGS(func, bpf_copy_from_user_str_dynptr, KF_SLEEPABLE) -BTF_ID_FLAGS(func, bpf_copy_from_user_task_dynptr, KF_SLEEPABLE) -BTF_ID_FLAGS(func, bpf_copy_from_user_task_str_dynptr, KF_SLEEPABLE) +BTF_ID_FLAGS(func, bpf_probe_read_user_dynptr, KF_PERFMON) +BTF_ID_FLAGS(func, bpf_probe_read_kernel_dynptr, KF_PERFMON) +BTF_ID_FLAGS(func, bpf_probe_read_user_str_dynptr, KF_PERFMON) +BTF_ID_FLAGS(func, bpf_probe_read_kernel_str_dynptr, KF_PERFMON) +BTF_ID_FLAGS(func, bpf_copy_from_user_dynptr, KF_SLEEPABLE | KF_PERFMON) +BTF_ID_FLAGS(func, bpf_copy_from_user_str_dynptr, KF_SLEEPABLE | KF_PERFMON) +BTF_ID_FLAGS(func, bpf_copy_from_user_task_dynptr, KF_SLEEPABLE | KF_PERFMON) +BTF_ID_FLAGS(func, bpf_copy_from_user_task_str_dynptr, KF_SLEEPABLE | KF_PERFMON) #endif #ifdef CONFIG_DMA_SHARED_BUFFER BTF_ID_FLAGS(func, bpf_iter_dmabuf_new, KF_ITER_NEW | KF_SLEEPABLE) @@ -4947,26 +4947,26 @@ BTF_ID_FLAGS(func, bpf_iter_dmabuf_next, KF_ITER_NEXT | KF_RET_NULL | KF_SLEEPAB BTF_ID_FLAGS(func, bpf_iter_dmabuf_destroy, KF_ITER_DESTROY | KF_SLEEPABLE) #endif BTF_ID_FLAGS(func, __bpf_trap) -BTF_ID_FLAGS(func, bpf_strcmp); -BTF_ID_FLAGS(func, bpf_strcasecmp); -BTF_ID_FLAGS(func, bpf_strncasecmp); -BTF_ID_FLAGS(func, bpf_strchr); -BTF_ID_FLAGS(func, bpf_strchrnul); -BTF_ID_FLAGS(func, bpf_strnchr); -BTF_ID_FLAGS(func, bpf_strrchr); -BTF_ID_FLAGS(func, bpf_strlen); -BTF_ID_FLAGS(func, bpf_strnlen); -BTF_ID_FLAGS(func, bpf_strspn); -BTF_ID_FLAGS(func, bpf_strcspn); -BTF_ID_FLAGS(func, bpf_strstr); -BTF_ID_FLAGS(func, bpf_strcasestr); -BTF_ID_FLAGS(func, bpf_strnstr); -BTF_ID_FLAGS(func, bpf_strncasestr); +BTF_ID_FLAGS(func, bpf_strcmp, KF_PERFMON); +BTF_ID_FLAGS(func, bpf_strcasecmp, KF_PERFMON); +BTF_ID_FLAGS(func, bpf_strncasecmp, KF_PERFMON); +BTF_ID_FLAGS(func, bpf_strchr, KF_PERFMON); +BTF_ID_FLAGS(func, bpf_strchrnul, KF_PERFMON); +BTF_ID_FLAGS(func, bpf_strnchr, KF_PERFMON); +BTF_ID_FLAGS(func, bpf_strrchr, KF_PERFMON); +BTF_ID_FLAGS(func, bpf_strlen, KF_PERFMON); +BTF_ID_FLAGS(func, bpf_strnlen, KF_PERFMON); +BTF_ID_FLAGS(func, bpf_strspn, KF_PERFMON); +BTF_ID_FLAGS(func, bpf_strcspn, KF_PERFMON); +BTF_ID_FLAGS(func, bpf_strstr, KF_PERFMON); +BTF_ID_FLAGS(func, bpf_strcasestr, KF_PERFMON); +BTF_ID_FLAGS(func, bpf_strnstr, KF_PERFMON); +BTF_ID_FLAGS(func, bpf_strncasestr, KF_PERFMON); #if defined(CONFIG_BPF_LSM) && defined(CONFIG_CGROUPS) BTF_ID_FLAGS(func, bpf_cgroup_read_xattr, KF_RCU) #endif -BTF_ID_FLAGS(func, bpf_stream_vprintk, KF_IMPLICIT_ARGS | KF_SPINLOCK_SAFE) -BTF_ID_FLAGS(func, bpf_stream_print_stack, KF_IMPLICIT_ARGS | KF_SPINLOCK_SAFE) +BTF_ID_FLAGS(func, bpf_stream_vprintk, KF_IMPLICIT_ARGS | KF_SPINLOCK_SAFE | KF_PERFMON) +BTF_ID_FLAGS(func, bpf_stream_print_stack, KF_IMPLICIT_ARGS | KF_SPINLOCK_SAFE | KF_PERFMON) BTF_ID_FLAGS(func, bpf_task_work_schedule_signal, KF_IMPLICIT_ARGS) BTF_ID_FLAGS(func, bpf_task_work_schedule_resume, KF_IMPLICIT_ARGS) BTF_ID_FLAGS(func, bpf_dynptr_from_file) From f9191460cd805e84eace0884c75c67ec719fa1d8 Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Thu, 10 Sep 2026 23:35:09 +0200 Subject: [PATCH 035/694] bpf: Require CAP_PERFMON for untrusted read-only memory reads Marking bpf_rdonly_cast() KF_PERFMON CAP-limits one producer of PTR_TO_MEM | MEM_RDONLY | PTR_UNTRUSTED, but not the type itself. A global subprogram argument tagged __arg_untrusted results in the same register with no kfunc call. Reported-by: STAR Labs SG Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/r/20260910213510.49358-3-daniel@iogearbox.net Signed-off-by: Alexei Starovoitov --- kernel/bpf/verifier.c | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 939e535a3442..8058f684a9ea 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -6453,6 +6453,15 @@ static int check_mem_access(struct bpf_verifier_env *env, int insn_idx, struct b return -EACCES; } + if (rdonly_untrusted && !env->allow_ptr_leaks) { + verbose(env, "%s access is allowed only to CAP_PERFMON and CAP_SYS_ADMIN\n", + reg_type_str(env, reg->type)); + bpf_diag_policy(env, insn_idx, "read from untrusted read-only memory", + "the access requires CAP_PERFMON", + "Load the program with CAP_PERFMON, or avoid dereferencing untrusted pointers."); + return -EPERM; + } + /* * Accesses to untrusted PTR_TO_MEM are done through probe * instructions, hence no need to check bounds in that case. From a903f145a8914d9ecba009b398d135154e4a5d94 Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Thu, 10 Sep 2026 23:35:10 +0200 Subject: [PATCH 036/694] selftests/bpf: Add tests for the KF_PERFMON gates Add test cases where each one loads with CAP_BPF alone and checks that the program is correctly rejected. # LDLIBS=-static PKG_CONFIG='pkg-config --static' ./vmtest.sh -- ./test_progs -t verifier_kfunc_perfmon [...] #627/1 verifier_kfunc_perfmon/rdonly_cast_noperfmon:OK #627/2 verifier_kfunc_perfmon/rdonly_cast_noperfmon @unpriv:OK #627/3 verifier_kfunc_perfmon/probe_read_kernel_dynptr_noperfmon:OK #627/4 verifier_kfunc_perfmon/probe_read_kernel_dynptr_noperfmon @unpriv:OK #627/5 verifier_kfunc_perfmon/stream_vprintk_noperfmon:OK #627/6 verifier_kfunc_perfmon/stream_vprintk_noperfmon @unpriv:OK #627/7 verifier_kfunc_perfmon/get_kmem_cache_noperfmon:OK #627/8 verifier_kfunc_perfmon/get_kmem_cache_noperfmon @unpriv:OK #627/9 verifier_kfunc_perfmon/arg_untrusted_read_noperfmon:OK #627/10 verifier_kfunc_perfmon/arg_untrusted_read_noperfmon @unpriv:OK #627 verifier_kfunc_perfmon:OK Summary: 1/10 PASSED, 0 SKIPPED, 0/0 FAILED Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/r/20260910213510.49358-4-daniel@iogearbox.net Signed-off-by: Alexei Starovoitov --- .../selftests/bpf/prog_tests/verifier.c | 2 + .../bpf/progs/verifier_kfunc_perfmon.c | 75 +++++++++++++++++++ 2 files changed, 77 insertions(+) create mode 100644 tools/testing/selftests/bpf/progs/verifier_kfunc_perfmon.c diff --git a/tools/testing/selftests/bpf/prog_tests/verifier.c b/tools/testing/selftests/bpf/prog_tests/verifier.c index 64ac49ad67e6..ec9e3907c75d 100644 --- a/tools/testing/selftests/bpf/prog_tests/verifier.c +++ b/tools/testing/selftests/bpf/prog_tests/verifier.c @@ -53,6 +53,7 @@ #include "verifier_iterating_callbacks.skel.h" #include "verifier_jeq_infer_not_null.skel.h" #include "verifier_jit_convergence.skel.h" +#include "verifier_kfunc_perfmon.skel.h" #include "verifier_ld_ind.skel.h" #include "verifier_ldsx.skel.h" #include "verifier_leak_ptr.skel.h" @@ -216,6 +217,7 @@ void test_verifier_int_ptr(void) { RUN(verifier_int_ptr); } void test_verifier_iterating_callbacks(void) { RUN(verifier_iterating_callbacks); } void test_verifier_jeq_infer_not_null(void) { RUN(verifier_jeq_infer_not_null); } void test_verifier_jit_convergence(void) { RUN(verifier_jit_convergence); } +void test_verifier_kfunc_perfmon(void) { RUN(verifier_kfunc_perfmon); } void test_verifier_load_acquire(void) { RUN(verifier_load_acquire); } void test_verifier_ld_ind(void) { RUN(verifier_ld_ind); } void test_verifier_ldsx(void) { RUN(verifier_ldsx); } diff --git a/tools/testing/selftests/bpf/progs/verifier_kfunc_perfmon.c b/tools/testing/selftests/bpf/progs/verifier_kfunc_perfmon.c new file mode 100644 index 000000000000..76c39ef30e96 --- /dev/null +++ b/tools/testing/selftests/bpf/progs/verifier_kfunc_perfmon.c @@ -0,0 +1,75 @@ +// SPDX-License-Identifier: GPL-2.0 + +#include +#include +#include "bpf_misc.h" + +void *user_ptr; +char dynptr_buf[8]; +u64 kaddr; + +extern struct kmem_cache *bpf_get_kmem_cache(u64 addr) __ksym; + +SEC("socket") +__success +__caps_unpriv(CAP_BPF) +__failure_unpriv +__msg_unpriv("bpf_rdonly_cast is allowed only to CAP_PERFMON and CAP_SYS_ADMIN") +int rdonly_cast_noperfmon(void *ctx) +{ + char *p = bpf_rdonly_cast(0, 0); + + return p[0x7fff]; +} + +SEC("socket") +__success +__caps_unpriv(CAP_BPF) +__failure_unpriv +__msg_unpriv("bpf_probe_read_kernel_dynptr is allowed only to CAP_PERFMON and CAP_SYS_ADMIN") +int probe_read_kernel_dynptr_noperfmon(void *ctx) +{ + struct bpf_dynptr dptr; + + bpf_dynptr_from_mem(dynptr_buf, sizeof(dynptr_buf), 0, &dptr); + bpf_probe_read_kernel_dynptr(&dptr, 0, sizeof(dynptr_buf), user_ptr); + return 0; +} + +SEC("socket") +__success +__caps_unpriv(CAP_BPF) +__failure_unpriv +__msg_unpriv("bpf_stream_vprintk is allowed only to CAP_PERFMON and CAP_SYS_ADMIN") +int stream_vprintk_noperfmon(void *ctx) +{ + bpf_stream_printk(BPF_STDOUT, "%pB", (void *)kaddr); + return 0; +} + +SEC("socket") +__success +__caps_unpriv(CAP_BPF) +__failure_unpriv +__msg_unpriv("bpf_get_kmem_cache is allowed only to CAP_PERFMON and CAP_SYS_ADMIN") +int get_kmem_cache_noperfmon(void *ctx) +{ + return !!bpf_get_kmem_cache(kaddr); +} + +__weak int subprog_untrusted_read(void *p __arg_untrusted) +{ + return *(char *)p; +} + +SEC("socket") +__success +__caps_unpriv(CAP_BPF) +__failure_unpriv +__msg_unpriv("rdonly_untrusted_mem access is allowed only to CAP_PERFMON and CAP_SYS_ADMIN") +int arg_untrusted_read_noperfmon(void *ctx) +{ + return subprog_untrusted_read(0); +} + +char _license[] SEC("license") = "GPL"; From 2936aed9b02162df7fbe08fcd6bfbb3270ad9f95 Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Thu, 10 Sep 2026 20:56:04 +0200 Subject: [PATCH 037/694] bpf: Clear scalar delta on narrowing stack spill check_stack_write_fixed_off() breaks the scalar link on a narrowing spill by zeroing the id of the destination stack slot, but leaves the delta in place. save_register_state() has just copied the source register verbatim, so the slot keeps the BPF_ADD_CONST delta of a register it is no longer linked to. This is the same shape as the case addressed in 1b327732c846 ("bpf: Clear delta when clearing reg id for non-{add,sub} ops"). Unlike the latter, no miscomputation seems reachable, so mainly consistency. Signed-off-by: Daniel Borkmann Acked-by: Eduard Zingerman Link: https://lore.kernel.org/r/20260910185604.44949-1-daniel@iogearbox.net Signed-off-by: Alexei Starovoitov --- kernel/bpf/verifier.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 8058f684a9ea..1b0b1fb62878 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -3582,7 +3582,7 @@ static int check_stack_write_fixed_off(struct bpf_verifier_env *env, save_register_state(env, state, spi, reg, size); /* Break the relation on a narrowing spill. */ if (!reg_value_fits) - state->stack[spi].spilled_ptr.id = 0; + clear_scalar_id(&state->stack[spi].spilled_ptr); } else if (!reg && !(off % BPF_REG_SIZE) && is_bpf_st_mem(insn) && env->bpf_capable) { struct bpf_reg_state *tmp_reg = &env->fake_reg[0]; From b0b3dc66529676228cb938cbcad66920f735c223 Mon Sep 17 00:00:00 2001 From: Jiayuan Chen Date: Thu, 10 Sep 2026 20:22:55 +0800 Subject: [PATCH 038/694] bpf: Fix divide-by-zero in btf_struct_walk() When an access goes past the struct and the last member is a flexible array, btf_struct_walk() folds the offset back into a single element with (off - moff) % t->size, but never checks that the element type has a size. BTF takes an empty struct, so this in program BTF /* event could be empty */ struct event { #ifdef HAVE_TIMESTAMP __u64 ts; #endif }; struct batch { int nr; struct event events[]; }; divides by zero at prog load time. Getting there needs a PTR_TO_BTF_ID that is not MEM_ALLOC, e.g. a plain read of a local kptr stashed in a map from a sleepable program. Oops: divide error: 0000 [#1] SMP KASAN PTI RIP: 0010:btf_struct_walk+0x53f/0x1570 Call Trace: btf_struct_access+0x42a/0xcd0 check_ptr_to_btf_access+0x4dc/0x1160 check_mem_access+0x3a45/0x8740 check_load_mem+0x36a/0xd10 do_check_common+0x3ef0/0xb210 bpf_check+0x6d3b/0x8580 bpf_prog_load+0xf7c/0x2720 __sys_bpf+0xa83/0x3690 __x64_sys_bpf+0xc7/0x150 x64_sys_call+0x1f3f/0x27e0 do_syscall_64+0xe5/0x610 entry_SYSCALL_64_after_hwframe+0x76/0x7e Reject a zero-sized element type. The fixed array path in the same function already bails out on the same thing: btf_struct_walk() ... /* skip empty array */ if (moff == mtrue_end) continue; msize /= total_nelems; Fixes: 9c5f8a1008a1 ("bpf: Support variable length array in tracing programs") Signed-off-by: Jiayuan Chen Acked-by: Eduard Zingerman Link: https://lore.kernel.org/r/20260910122316.186384-1-jiayuan.chen@linux.dev Signed-off-by: Alexei Starovoitov --- kernel/bpf/btf.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/kernel/bpf/btf.c b/kernel/bpf/btf.c index 9f33e95d5741..d67a169ba2bb 100644 --- a/kernel/bpf/btf.c +++ b/kernel/bpf/btf.c @@ -7187,7 +7187,7 @@ static int btf_struct_walk(struct bpf_verifier_log *log, const struct btf *btf, if (btf_type_is_int(t)) return WALK_SCALAR; - if (!btf_type_is_struct(t)) + if (!btf_type_is_struct(t) || !t->size) goto error; off = (off - moff) % t->size; From f77d21245710690f3fb02d19d8dd4dc17ee58c2c Mon Sep 17 00:00:00 2001 From: Jiayuan Chen Date: Thu, 10 Sep 2026 20:22:56 +0800 Subject: [PATCH 039/694] selftests/bpf: Test BTF walk into a flexible array of zero-sized elements The program stashes a bpf_obj_new() object whose type ends with a flexible array of empty structs, then reads it back as an untrusted kptr. Without the previous patch this divides by zero in btf_struct_walk() instead of being rejected. # ./test_progs -t verifier_btf_flex_array ... #602 verifier_btf_flex_array:OK Summary: 1/1 PASSED, 0 SKIPPED, 0/0 FAILED Signed-off-by: Jiayuan Chen Link: https://lore.kernel.org/r/20260910122316.186384-2-jiayuan.chen@linux.dev Signed-off-by: Alexei Starovoitov --- .../selftests/bpf/prog_tests/verifier.c | 2 + .../bpf/progs/verifier_btf_flex_array.c | 56 +++++++++++++++++++ 2 files changed, 58 insertions(+) create mode 100644 tools/testing/selftests/bpf/progs/verifier_btf_flex_array.c diff --git a/tools/testing/selftests/bpf/prog_tests/verifier.c b/tools/testing/selftests/bpf/prog_tests/verifier.c index ec9e3907c75d..8b439e194bcc 100644 --- a/tools/testing/selftests/bpf/prog_tests/verifier.c +++ b/tools/testing/selftests/bpf/prog_tests/verifier.c @@ -23,6 +23,7 @@ #include "verifier_bpf_trap.skel.h" #include "verifier_bswap.skel.h" #include "verifier_btf_ctx_access.skel.h" +#include "verifier_btf_flex_array.skel.h" #include "verifier_btf_unreliable_prog.skel.h" #include "verifier_call_large_imm.skel.h" #include "verifier_cfg.skel.h" @@ -187,6 +188,7 @@ void test_verifier_bpf_get_stack(void) { RUN(verifier_bpf_get_stack); } void test_verifier_bpf_trap(void) { RUN(verifier_bpf_trap); } void test_verifier_bswap(void) { RUN(verifier_bswap); } void test_verifier_btf_ctx_access(void) { RUN(verifier_btf_ctx_access); } +void test_verifier_btf_flex_array(void) { RUN(verifier_btf_flex_array); } void test_verifier_btf_unreliable_prog(void) { RUN(verifier_btf_unreliable_prog); } void test_verifier_call_large_imm(void) { RUN(verifier_call_large_imm); } void test_verifier_cfg(void) { RUN(verifier_cfg); } diff --git a/tools/testing/selftests/bpf/progs/verifier_btf_flex_array.c b/tools/testing/selftests/bpf/progs/verifier_btf_flex_array.c new file mode 100644 index 000000000000..59b84261f622 --- /dev/null +++ b/tools/testing/selftests/bpf/progs/verifier_btf_flex_array.c @@ -0,0 +1,56 @@ +// SPDX-License-Identifier: GPL-2.0 + +#include +#include + +#include "bpf_experimental.h" +#include "bpf_misc.h" + +struct test_empty_event {}; + +struct test_flex_batch { + int nr; + struct test_empty_event events[]; +}; + +struct map_value { + struct test_flex_batch __kptr *batch; +}; + +struct { + __uint(type, BPF_MAP_TYPE_ARRAY); + __type(key, int); + __type(value, struct map_value); + __uint(max_entries, 1); +} batches SEC(".maps"); + +SEC("syscall") +__description("btf walk into flexible array of zero-sized elements") +__failure __msg("access beyond struct test_flex_batch at off 4 size 1") +int stash_and_peek(void *ctx) +{ + struct test_flex_batch *b, *old; + struct map_value *v; + int key = 0; + + v = bpf_map_lookup_elem(&batches, &key); + if (!v) + return 0; + + b = bpf_obj_new(struct test_flex_batch); + if (!b) + return 0; + b->nr = 1; + + old = bpf_kptr_xchg(&v->batch, b); + if (old) + bpf_obj_drop(old); + + b = v->batch; + if (!b) + return 0; + + return b->nr + *(char *)&b->events[0]; +} + +char _license[] SEC("license") = "GPL"; From 01b245ba016d44861690594e10f67e026ce8552f Mon Sep 17 00:00:00 2001 From: Jiayuan Chen Date: Thu, 10 Sep 2026 19:26:26 +0800 Subject: [PATCH 040/694] bpf: Fix out-of-bounds read of sk_protocol in bpf_sock_destroy() sk_protocol lives in struct sock, not in struct sock_common. A timewait or request sock handed to bpf_sock_destroy() by the tcp iterator is neither, so reading sk->sk_protocol runs past the object: ================================================================== BUG: KASAN: slab-out-of-bounds in bpf_sock_destroy+0xc7/0xe0 Read of size 2 at addr ffff8881047d11b4 by task test_progs/428 Tainted: [W]=WARN Call Trace: dump_stack_lvl+0x91/0xf0 print_report+0xd1/0x630 kasan_report+0xf3/0x130 __asan_report_load2_noabort+0x14/0x30 bpf_sock_destroy+0xc7/0xe0 bpf_prog_c3dd61f9d9cd9f37_iter_tcp6_timewait+0x9f/0xb7 bpf_iter_run_prog+0x538/0xde0 bpf_iter_tcp_seq_show+0x26b/0x4b0 bpf_seq_read+0x424/0x1210 vfs_read+0x197/0xe40 ksys_read+0x119/0x240 __x64_sys_read+0x72/0xc0 x64_sys_call+0x647/0x27e0 do_syscall_64+0xe5/0x610 entry_SYSCALL_64_after_hwframe+0x76/0x7e Only check sk_protocol on full socks. tcp_abort() already knows how to deal with TIME_WAIT and NEW_SYN_RECV socks. Also fix the comment, it never matched the code. Fixes: 4ddbcb886268 ("bpf: Add bpf_sock_destroy kfunc") Reported-by: Xiang Mei (Microsoft) Closes: https://lore.kernel.org/bpf/20260702224519.800135-1-xmei5@asu.edu/ Signed-off-by: Jiayuan Chen Reviewed-by: Kuniyuki Iwashima Link: https://lore.kernel.org/r/20260910112634.152195-1-jiayuan.chen@linux.dev Signed-off-by: Alexei Starovoitov --- net/core/filter.c | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-) diff --git a/net/core/filter.c b/net/core/filter.c index 2a84f9d01131..cae43b999162 100644 --- a/net/core/filter.c +++ b/net/core/filter.c @@ -12913,8 +12913,9 @@ __bpf_kfunc_start_defs(); * @sock: Pointer to socket to be destroyed * * Return: - * On error, may return EPROTONOSUPPORT, EINVAL. - * EPROTONOSUPPORT if protocol specific destroy handler is not supported. + * On error, may return EOPNOTSUPP, or whatever the protocol specific + * destroy handler returns. + * EOPNOTSUPP if protocol specific destroy handler is not supported. * 0 otherwise */ __bpf_kfunc int bpf_sock_destroy(struct sock_common *sock) @@ -12926,8 +12927,12 @@ __bpf_kfunc int bpf_sock_destroy(struct sock_common *sock) * Supporting protocols will need to acquire sock lock in the BPF context * prior to invoking this kfunc. */ - if (!sk->sk_prot->diag_destroy || (sk->sk_protocol != IPPROTO_TCP && - sk->sk_protocol != IPPROTO_UDP)) + if (!sk->sk_prot->diag_destroy) + return -EOPNOTSUPP; + + if (sk_fullsock(sk) && + sk->sk_protocol != IPPROTO_TCP && + sk->sk_protocol != IPPROTO_UDP) return -EOPNOTSUPP; return sk->sk_prot->diag_destroy(sk, ECONNABORTED); From eaab8cab451b9502ce224cd202550375b894a467 Mon Sep 17 00:00:00 2001 From: Jiayuan Chen Date: Thu, 10 Sep 2026 19:27:28 +0800 Subject: [PATCH 041/694] tcp: Skip cond_resched() in inet_csk_listen_stop() under BPF context bpf_sock_destroy() runs from the tcp iterator, under rcu_read_lock(). If the sock is a listener that still has children in its accept queue, tcp_abort() ends up in inet_csk_listen_stop() and the cond_resched() there trips the debug check: BUG: sleeping function called from invalid context at net/ipv4/inet_connection_sock.c:1523 in_atomic(): 0, irqs_disabled(): 0, non_block: 0, pid: 628, name: test_progs preempt_count: 0, expected: 0 RCU nest depth: 1, expected: 0 locks held by test_progs/628: 3, last CPU#3: #0: ffff8881158cee18 (&p->lock){+.+.}-{4:4}, at: bpf_seq_read+0x56/0x1210 #1: ffff8881106bb858 (sk_lock-AF_INET6){+.+.}-{0:0}, at: bpf_iter_tcp_seq_show+0x32b/0x4b0 #2: ffffffffb435af20 (rcu_read_lock){....}-{1:3}, at: bpf_iter_run_prog+0x46b/0xde0 CPU: 3 UID: 0 PID: 628 Comm: test_progs Tainted: G W 7.2.0+ #65 PREEMPT Tainted: [W]=WARN Call Trace: dump_stack_lvl+0xc1/0xf0 dump_stack+0x10/0x20 __might_resched+0x3d2/0x610 inet_csk_listen_stop+0x7b/0xbf0 tcp_abort+0x23b/0x3b0 bpf_sock_destroy+0xfc/0x140 bpf_prog_448133d24601754f_iter_tcp6_server+0x81/0x8a bpf_iter_run_prog+0x538/0xde0 bpf_iter_tcp_seq_show+0x26b/0x4b0 bpf_seq_read+0x424/0x1210 vfs_read+0x197/0xe40 ksys_read+0x119/0x240 __x64_sys_read+0x72/0xc0 x64_sys_call+0x647/0x27e0 do_syscall_64+0xe5/0x610 entry_SYSCALL_64_after_hwframe+0x76/0x7e RIP: 0033:0x7fad39b28aca RSP: 002b:00007ffc381c61c0 EFLAGS: 00000246 ORIG_RAX: 0000000000000000 RAX: ffffffffffffffda RBX: 00007ffc381c6a88 RCX: 00007fad39b28aca RDX: 0000000000000032 RSI: 00007ffc381c6250 RDI: 0000000000000014 RBP: 00007ffc381c61e0 R08: 0000000000000000 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000003 R13: 0000000000000000 R14: 000055f077c1bbb0 R15: 00007fad3a0f3000 The commit that added the kfunc already guards lock_sock() in tcp_abort() and udp_abort() with has_current_bpf_ctx(), but missed the listener path. Do the same for the cond_resched(). The loop runs inside the iterator's rcu_read_lock(), it must not reschedule or report a quiescent state there. Fixes: 4ddbcb886268 ("bpf: Add bpf_sock_destroy kfunc") Signed-off-by: Jiayuan Chen Link: https://lore.kernel.org/r/20260910112736.153710-1-jiayuan.chen@linux.dev Signed-off-by: Alexei Starovoitov --- net/ipv4/inet_connection_sock.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_sock.c index 6257459bcee2..6a30f1138454 100644 --- a/net/ipv4/inet_connection_sock.c +++ b/net/ipv4/inet_connection_sock.c @@ -1520,7 +1520,8 @@ void inet_csk_listen_stop(struct sock *sk) local_bh_enable(); sock_put(child); - cond_resched(); + if (!has_current_bpf_ctx()) + cond_resched(); } if (queue->fastopenq.rskq_rst_head) { /* Free all the reqs queued in rskq_rst_head. */ From 8036d3a5a6589ef0721d15e7c03976bc3996f7a8 Mon Sep 17 00:00:00 2001 From: Jiayuan Chen Date: Thu, 10 Sep 2026 19:28:02 +0800 Subject: [PATCH 042/694] selftests/bpf: Test bpf_sock_destroy() on TIME_WAIT and listener socks Add two subtests. tcp_timewait: the client shuts down first and the server closes after it, so the client sock ends up in TIME_WAIT. A tcp iterator then finds the timewait sock by the cookie it inherited from the client sock and destroys it. Iterate once more to make sure it is gone. Without the first fix bpf_sock_destroy() reads past the timewait sock and KASAN complains. tcp_listen_pending: connect to a listener but never accept, so the child sits in the accept queue, then destroy the listener. Without the second fix the cond_resched() in inet_csk_listen_stop() trips the might_sleep check under rcu_read_lock(). ./test_progs -a sock_destroy #444/1 sock_destroy/tcp_client:OK #444/2 sock_destroy/tcp_server:OK #444/3 sock_destroy/tcp_listen_pending:OK #444/4 sock_destroy/tcp_timewait:OK #444/5 sock_destroy/udp_client:OK #444/6 sock_destroy/udp_server:OK #444/7 sock_destroy/trace_tcp_destroy_sock:OK #444 sock_destroy:OK Summary: 1/7 PASSED, 0 SKIPPED, 0/0 FAILED Signed-off-by: Jiayuan Chen Link: https://lore.kernel.org/r/20260910112810.153801-1-jiayuan.chen@linux.dev Signed-off-by: Alexei Starovoitov --- .../selftests/bpf/prog_tests/sock_destroy.c | 121 ++++++++++++++++++ .../selftests/bpf/progs/sock_destroy_prog.c | 30 +++++ 2 files changed, 151 insertions(+) diff --git a/tools/testing/selftests/bpf/prog_tests/sock_destroy.c b/tools/testing/selftests/bpf/prog_tests/sock_destroy.c index 9c11938fe597..78d642a02bdb 100644 --- a/tools/testing/selftests/bpf/prog_tests/sock_destroy.c +++ b/tools/testing/selftests/bpf/prog_tests/sock_destroy.c @@ -1,4 +1,5 @@ // SPDX-License-Identifier: GPL-2.0 +#include #include #include @@ -110,6 +111,122 @@ static void test_tcp_server(struct sock_destroy_prog *skel) close(serv); } +static void test_tcp_listen_pending(struct sock_destroy_prog *skel) +{ + int serv = -1, clien = -1, accept_serv = -1, n, serv_port; + struct pollfd pfd = { .events = POLLIN }; + char buf[1]; + + serv = start_server(AF_INET6, SOCK_STREAM, NULL, 0, 0); + if (!ASSERT_GE(serv, 0, "start_server")) + goto cleanup; + serv_port = get_socket_local_port(serv); + if (!ASSERT_GE(serv_port, 0, "get_sock_local_port")) + goto cleanup; + skel->bss->serv_port = (__be16)serv_port; + + /* + * Connect but never accept, so the child sits in the accept queue + * of the listener. Wait until it's actually there. + */ + clien = connect_to_fd(serv, 0); + if (!ASSERT_GE(clien, 0, "connect_to_fd")) + goto cleanup; + pfd.fd = serv; + if (!ASSERT_EQ(poll(&pfd, 1, -1), 1, "poll listener")) + goto cleanup; + + /* Run iterator program that destroys server sockets. */ + start_iter_sockets(skel->progs.iter_tcp6_server); + + accept_serv = accept(serv, NULL, NULL); + if (!ASSERT_LT(accept_serv, 0, "accept on destroyed listener")) + goto cleanup; + ASSERT_EQ(errno, EINVAL, "error code on destroyed listener"); + + /* The unaccepted child was reset along with the listener. */ + n = recv(clien, buf, sizeof(buf), 0); + if (!ASSERT_LT(n, 0, "client recv on reset child")) + goto cleanup; + ASSERT_EQ(errno, ECONNRESET, "error code on reset child"); + +cleanup: + if (clien != -1) + close(clien); + if (accept_serv != -1) + close(accept_serv); + if (serv != -1) + close(serv); +} + +static void test_tcp_timewait(struct sock_destroy_prog *skel) +{ + int serv = -1, clien = -1, accept_serv = -1, n; + struct timeval tv = {}; + char buf[1]; + + serv = start_server(AF_INET6, SOCK_STREAM, NULL, 0, 0); + if (!ASSERT_GE(serv, 0, "start_server")) + goto cleanup; + + clien = connect_to_fd(serv, 0); + if (!ASSERT_GE(clien, 0, "connect_to_fd")) + goto cleanup; + + accept_serv = accept(serv, NULL, NULL); + if (!ASSERT_GE(accept_serv, 0, "serv accept")) + goto cleanup; + + /* + * Active close from the client, then close the server side. Once + * recv() sees EOF the server FIN has been processed and the client + * sock is in TIME_WAIT. Block without timeout so a loaded CI box + * can't race us. + */ + if (!ASSERT_OK(setsockopt(clien, SOL_SOCKET, SO_RCVTIMEO, &tv, + sizeof(tv)), "clear rcvtimeo")) + goto cleanup; + if (!ASSERT_OK(shutdown(clien, SHUT_WR), "client shutdown")) + goto cleanup; + + /* + * Make sure the server has seen the client FIN before it closes, + * so the two FINs never cross. + */ + n = recv(accept_serv, buf, sizeof(buf), 0); + if (!ASSERT_EQ(n, 0, "server recv EOF")) + goto cleanup; + + close(accept_serv); + accept_serv = -1; + + /* block until return EOF */ + n = recv(clien, buf, sizeof(buf), 0); + if (!ASSERT_EQ(n, 0, "client recv EOF")) + goto cleanup; + + /* Run iterator program that destroys the timewait client sock. */ + skel->bss->tw_found = 0; + start_iter_sockets(skel->progs.iter_tcp6_timewait); + if (!ASSERT_EQ(skel->bss->tw_found, 1, "timewait sock found")) + goto cleanup; + + ASSERT_OK(skel->bss->tw_destroy_err, "destroy timewait sock"); + + /* The destroyed timewait sock must be gone. */ + skel->bss->tw_found = 0; + start_iter_sockets(skel->progs.iter_tcp6_timewait); + ASSERT_EQ(skel->bss->tw_found, 0, "timewait sock destroyed"); + +cleanup: + if (clien != -1) + close(clien); + if (accept_serv != -1) + close(accept_serv); + if (serv != -1) + close(serv); +} + static void test_udp_client(struct sock_destroy_prog *skel) { int serv = -1, clien = -1, n = 0; @@ -204,6 +321,10 @@ void test_sock_destroy(void) test_tcp_client(skel); if (test__start_subtest("tcp_server")) test_tcp_server(skel); + if (test__start_subtest("tcp_listen_pending")) + test_tcp_listen_pending(skel); + if (test__start_subtest("tcp_timewait")) + test_tcp_timewait(skel); if (test__start_subtest("udp_client")) test_udp_client(skel); if (test__start_subtest("udp_server")) diff --git a/tools/testing/selftests/bpf/progs/sock_destroy_prog.c b/tools/testing/selftests/bpf/progs/sock_destroy_prog.c index 9e0bf7a54cec..0a8887543218 100644 --- a/tools/testing/selftests/bpf/progs/sock_destroy_prog.c +++ b/tools/testing/selftests/bpf/progs/sock_destroy_prog.c @@ -7,6 +7,8 @@ #include "bpf_tracing_net.h" __be16 serv_port = 0; +int tw_found = 0; +int tw_destroy_err = 0; int bpf_sock_destroy(struct sock_common *sk) __ksym; @@ -100,6 +102,34 @@ int iter_tcp6_server(struct bpf_iter__tcp *ctx) return 0; } +SEC("iter/tcp") +int iter_tcp6_timewait(struct bpf_iter__tcp *ctx) +{ + struct sock_common *sk_common = ctx->sk_common; + __u64 *val; + int key = 0; + + if (!sk_common) + return 0; + + if (sk_common->skc_family != AF_INET6) + return 0; + + if (!bpf_skc_to_tcp_timewait_sock(sk_common)) + return 0; + + val = bpf_map_lookup_elem(&tcp_conn_sockets, &key); + if (!val) + return 0; + /* The timewait sock inherits the cookie of the closed client sock. */ + if (bpf_get_socket_cookie(sk_common) != *val) + return 0; + + tw_found++; + tw_destroy_err = bpf_sock_destroy(sk_common); + + return 0; +} SEC("iter/udp") int iter_udp6_client(struct bpf_iter__udp *ctx) From 79a71cc2568f4b5d42284da2aa26f3b4f47ce01b Mon Sep 17 00:00:00 2001 From: Jim Mattson Date: Wed, 2 Sep 2026 11:47:11 -0700 Subject: [PATCH 043/694] KVM: x86/pmu: Move Intel PMU global MSRs to intel_is_valid_msr() Commit c85cdc1cc1ea ("KVM: x86/pmu: Move handling PERF_GLOBAL_CTRL and friends to common x86") moved the existence check for the following Intel PMU MSRs to kvm_pmu_is_valid_msr(): - MSR_CORE_PERF_GLOBAL_STATUS - MSR_CORE_PERF_GLOBAL_CTRL - MSR_CORE_PERF_GLOBAL_OVF_CTRL That commit deemed these MSRs valid whenever pmu->version > 1. It intended to share the check with AMD PerfMonV2 because both vendor implementations require version 2 or greater for global PMU controls. However, as noted in the commit message, AMD uses different MSR indices for its global PMU registers. Commit 4a2771895ca6 ("KVM: x86/svm/pmu: Add AMD PerfMonV2 support") subsequently added AMD PerfMonV2 support and set pmu->version = 2. Because kvm_pmu_is_valid_msr() validated the Intel MSRs whenever pmu->version > 1, KVM incorrectly permitted AMD guests with PerfMonV2 to access these Intel MSRs without a #GP. Move the validation of these Intel MSRs to intel_is_valid_msr() and remove the common switch statement from kvm_pmu_is_valid_msr(). AMD already validates its own global PMU MSRs in amd_is_valid_msr(). Fixes: 4a2771895ca6 ("KVM: x86/svm/pmu: Add AMD PerfMonV2 support") Signed-off-by: Jim Mattson Reviewed-by: Like Xu Reviewed-by: Sandipan Das Link: https://patch.msgid.link/20260902184711.138538-1-jmattson@google.com Signed-off-by: Sean Christopherson --- arch/x86/kvm/pmu.c | 8 -------- arch/x86/kvm/vmx/pmu_intel.c | 3 +++ 2 files changed, 3 insertions(+), 8 deletions(-) diff --git a/arch/x86/kvm/pmu.c b/arch/x86/kvm/pmu.c index a7d60c8785cd..d2fd47ee5ec8 100644 --- a/arch/x86/kvm/pmu.c +++ b/arch/x86/kvm/pmu.c @@ -823,14 +823,6 @@ void kvm_pmu_deliver_pmi(struct kvm_vcpu *vcpu) bool kvm_pmu_is_valid_msr(struct kvm_vcpu *vcpu, u32 msr) { - switch (msr) { - case MSR_CORE_PERF_GLOBAL_STATUS: - case MSR_CORE_PERF_GLOBAL_CTRL: - case MSR_CORE_PERF_GLOBAL_OVF_CTRL: - return kvm_pmu_has_perf_global_ctrl(vcpu_to_pmu(vcpu)); - default: - break; - } return kvm_pmu_call(msr_idx_to_pmc)(vcpu, msr) || kvm_pmu_call(is_valid_msr)(vcpu, msr); } diff --git a/arch/x86/kvm/vmx/pmu_intel.c b/arch/x86/kvm/vmx/pmu_intel.c index bfa8612fb450..70a8c4816135 100644 --- a/arch/x86/kvm/vmx/pmu_intel.c +++ b/arch/x86/kvm/vmx/pmu_intel.c @@ -187,6 +187,9 @@ static bool intel_is_valid_msr(struct kvm_vcpu *vcpu, u32 msr) int ret; switch (msr) { + case MSR_CORE_PERF_GLOBAL_STATUS: + case MSR_CORE_PERF_GLOBAL_CTRL: + case MSR_CORE_PERF_GLOBAL_OVF_CTRL: case MSR_CORE_PERF_FIXED_CTR_CTRL: return kvm_pmu_has_perf_global_ctrl(pmu); case MSR_IA32_PEBS_ENABLE: From f13368e0acffd6d6289629705cb821d921104725 Mon Sep 17 00:00:00 2001 From: Sean Christopherson Date: Wed, 26 Aug 2026 09:14:30 -0700 Subject: [PATCH 044/694] KVM: selftests: Use __GLIBC__, not _GNU_SOURCE, to detect actual glibc Use __GLIBC__ in the hardware disable test to detect when selftests are being built/linked against glibc and thus pthread_attr_setaffinity_np() is (hopefully) available. As pointed out by Sashiko and Hisam, _GNU_SOURCE is effectively a "request" macro to enable functionality, whereas __GLIBC__ is an announcement of support and selftests' idiomatic way of guarding code that's specific to glibc. Fixes: 496779b54943 ("KVM: selftests: Pre-set threads affinity in hardware disable test when possible") Reported-by: Sashiko Bot Closes: https://lore.kernel.org/all/20260731201140.5AF0C1F00AC4@smtp.kernel.org Suggested-by: Hisam Mehboob Link: https://patch.msgid.link/20260826161430.714316-1-seanjc@google.com Signed-off-by: Sean Christopherson --- tools/testing/selftests/kvm/hardware_disable_test.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/tools/testing/selftests/kvm/hardware_disable_test.c b/tools/testing/selftests/kvm/hardware_disable_test.c index 43a36ef3ead8..1c20892d6782 100644 --- a/tools/testing/selftests/kvm/hardware_disable_test.c +++ b/tools/testing/selftests/kvm/hardware_disable_test.c @@ -37,7 +37,7 @@ static void *run_vcpu(void *arg) struct kvm_vcpu *vcpu = arg; struct kvm_run *run = vcpu->run; -#ifndef _GNU_SOURCE +#ifndef __GLIBC__ kvm_sched_setaffinity(0, sizeof(cpu_set_t), &threads_cpu_set); #endif @@ -51,7 +51,7 @@ static void *sleeping_thread(void *arg) { int fd; -#ifndef _GNU_SOURCE +#ifndef __GLIBC__ kvm_sched_setaffinity(0, sizeof(cpu_set_t), &threads_cpu_set); #endif @@ -71,7 +71,7 @@ static void run_test(u32 run) u32 i, j; TEST_ASSERT_EQ(pthread_attr_init(&attr), 0); -#ifdef _GNU_SOURCE +#ifdef __GLIBC__ TEST_ASSERT_EQ(pthread_attr_setaffinity_np(&attr, sizeof(cpu_set_t), &threads_cpu_set), 0); #endif From 8cd280282d1239bca68f8c3632ef1ac8556a396b Mon Sep 17 00:00:00 2001 From: Sean Christopherson Date: Thu, 6 Aug 2026 14:46:18 -0700 Subject: [PATCH 045/694] KVM: Never clear KVM_REQ_VM_DEAD from a vCPU's requests Use kvm_test_request() instead of kvm_check_request() when querying KVM_REQ_VM_DEAD, i.e. don't clear KVM_REQ_VM_DEAD, as the entire purpose of KVM_REQ_VM_DEAD is to prevent the vCPU from enterring the guest ever again, even if userspace insists on redoing KVM_RUN. Ensuring KVM_REQ_VM_DEAD is never cleared will allow relaxing KVM's rule that ioctls can't be invoked on dead VMs, to only disallow ioctls if the VM is bugged, i.e. if KVM hit a KVM_BUG_ON(). Opportunistically add compile-time assertions to guard against clearing KVM_REQ_VM_DEAD through the standard APIs. Reviewed-by: Kai Huang Acked-by: Marc Zyngier Link: https://patch.msgid.link/20260806214618.82180-1-seanjc@google.com Signed-off-by: Sean Christopherson --- arch/arm64/kvm/arm.c | 2 +- arch/x86/kvm/mmu/mmu.c | 2 +- arch/x86/kvm/vmx/tdx.c | 2 +- arch/x86/kvm/x86.c | 2 +- include/linux/kvm_host.h | 9 +++++++-- 5 files changed, 11 insertions(+), 6 deletions(-) diff --git a/arch/arm64/kvm/arm.c b/arch/arm64/kvm/arm.c index 8b080804bc90..62c81d6ae8ee 100644 --- a/arch/arm64/kvm/arm.c +++ b/arch/arm64/kvm/arm.c @@ -1129,7 +1129,7 @@ static int kvm_vcpu_suspend(struct kvm_vcpu *vcpu) static int check_vcpu_requests(struct kvm_vcpu *vcpu) { if (kvm_request_pending(vcpu)) { - if (kvm_check_request(KVM_REQ_VM_DEAD, vcpu)) + if (kvm_test_request(KVM_REQ_VM_DEAD, vcpu)) return -EIO; if (kvm_check_request(KVM_REQ_SLEEP, vcpu)) diff --git a/arch/x86/kvm/mmu/mmu.c b/arch/x86/kvm/mmu/mmu.c index 064ecc33b926..8e62476e477b 100644 --- a/arch/x86/kvm/mmu/mmu.c +++ b/arch/x86/kvm/mmu/mmu.c @@ -5058,7 +5058,7 @@ static int kvm_tdp_page_prefault(struct kvm_vcpu *vcpu, gpa_t gpa, if (signal_pending(current)) return -EINTR; - if (kvm_check_request(KVM_REQ_VM_DEAD, vcpu)) + if (kvm_test_request(KVM_REQ_VM_DEAD, vcpu)) return -EIO; cond_resched(); diff --git a/arch/x86/kvm/vmx/tdx.c b/arch/x86/kvm/vmx/tdx.c index b272c20586a7..6c842e9191a5 100644 --- a/arch/x86/kvm/vmx/tdx.c +++ b/arch/x86/kvm/vmx/tdx.c @@ -1995,7 +1995,7 @@ static int tdx_handle_ept_violation(struct kvm_vcpu *vcpu) if (kvm_vcpu_has_events(vcpu) || signal_pending(current)) break; - if (kvm_check_request(KVM_REQ_VM_DEAD, vcpu)) { + if (kvm_test_request(KVM_REQ_VM_DEAD, vcpu)) { ret = -EIO; break; } diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index 79468ddfe473..a137dc6dd8c6 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -8060,7 +8060,7 @@ static int vcpu_enter_guest(struct kvm_vcpu *vcpu) bool req_immediate_exit = false; if (kvm_request_pending(vcpu)) { - if (kvm_check_request(KVM_REQ_VM_DEAD, vcpu)) { + if (kvm_test_request(KVM_REQ_VM_DEAD, vcpu)) { r = -EIO; goto out; } diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h index 03bfc92864b6..cf7fe835c4ad 100644 --- a/include/linux/kvm_host.h +++ b/include/linux/kvm_host.h @@ -2324,13 +2324,18 @@ static inline bool kvm_test_request(int req, struct kvm_vcpu *vcpu) return test_bit(req & KVM_REQUEST_MASK, (void *)&vcpu->requests); } -static inline void kvm_clear_request(int req, struct kvm_vcpu *vcpu) +static __always_inline void kvm_clear_request(int req, struct kvm_vcpu *vcpu) { + BUILD_BUG_ON(req == KVM_REQ_VM_DEAD); + clear_bit(req & KVM_REQUEST_MASK, (void *)&vcpu->requests); } -static inline bool kvm_check_request(int req, struct kvm_vcpu *vcpu) +static __always_inline bool kvm_check_request(int req, struct kvm_vcpu *vcpu) { + /* Once a VM is dead, it needs to stay dead. */ + BUILD_BUG_ON(req == KVM_REQ_VM_DEAD); + if (kvm_test_request(req, vcpu)) { kvm_clear_request(req, vcpu); From 77be3641f3e3a56e42a5ed889372ef395a933f3c Mon Sep 17 00:00:00 2001 From: Mikhail Gavrilov Date: Thu, 3 Sep 2026 16:51:35 +0500 Subject: [PATCH 046/694] debugfs: don't warn about uninitialized debugfs for an error parent Since commit c3a280ff728a ("debugfs: warn if file creation failed due to uninitialized debugfs") every boot with CONFIG_REF_TRACKER=y and CONFIG_DEBUG_FS=y prints two errors before the root filesystem is mounted: debugfs: Unable to create file 'net_refcnt@(____ptrval____)', debugfs is not initialized yet debugfs: Unable to create file 'net_notrefcnt@(____ptrval____)', debugfs is not initialized yet Nothing is actually wrong. Both files show up under /sys/kernel/debug/ref_tracker/ once the system is up. The kernel is reporting an error for a condition the caller has already accounted for. net_ns_init() runs directly from start_kernel(), before any initcall, and calls ref_tracker_dir_init() for init_net's two trackers. debugfs_init() is a core_initcall, so debugfs cannot possibly be up at that point. That is by design: ref_tracker_dir_debugfs() is documented as safe to call again later, and net/core/net_namespace.c has a late_initcall() that re-registers both directories once debugfs exists. ref_tracker also states that intent to debugfs. ref_tracker_debug_dir is initialised to ERR_PTR(-ENOENT) and only gets a real dentry in a late_initcall, so the early call hands debugfs_create_file() a parent that is already an error. debugfs_start_creating() honours that and returns the parent error, but only after the new pr_err() has fired. Move the IS_ERR(parent) check above the debugfs_initialized() test. A caller passing an error parent is propagating an earlier failure, which is the pattern debugfs documents and which the warning is not aimed at. A caller passing a valid or NULL parent too early - the case the warning was added for - still gets it. One behaviour change: an early caller with an error parent now gets PTR_ERR(parent) back instead of -ENOENT. All callers of these interfaces are documented to ignore the return value. Fixes: c3a280ff728a ("debugfs: warn if file creation failed due to uninitialized debugfs") Link: https://lore.kernel.org/all/6d1dc775f7d5e754d734907514534054f682bac5.1781171918.git.yk@y-koj.net/ Signed-off-by: Mikhail Gavrilov Tested-by: Yohei Kojima Link: https://patch.msgid.link/20260903115135.63210-1-mikhail.v.gavrilov@gmail.com Signed-off-by: Danilo Krummrich --- fs/debugfs/inode.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/fs/debugfs/inode.c b/fs/debugfs/inode.c index e054e62919ec..a4d08bd3743b 100644 --- a/fs/debugfs/inode.c +++ b/fs/debugfs/inode.c @@ -368,6 +368,9 @@ static struct dentry *debugfs_start_creating(const char *name, if (!debugfs_enabled) return ERR_PTR(-EPERM); + if (IS_ERR(parent)) + return parent; + if (!debugfs_initialized()) { pr_err("Unable to create file '%s', debugfs is not initialized yet\n", name); @@ -376,9 +379,6 @@ static struct dentry *debugfs_start_creating(const char *name, pr_debug("creating file '%s'\n", name); - if (IS_ERR(parent)) - return parent; - error = simple_pin_fs(&debug_fs_type, &debugfs_mount, &debugfs_mount_count); if (error) { From 0d7823cd4cda35f2060a685fa276ff7709915abc Mon Sep 17 00:00:00 2001 From: Siddharth Chintamaneni Date: Wed, 2 Sep 2026 17:14:13 +0000 Subject: [PATCH 047/694] bpf: Allow terminal gotox instructions check_subprogs() treats gotox as a direct jump and validates its reserved zero offset. When gotox is the final instruction, this produces a synthetic successor one instruction past the end of the subprogram and rejects an otherwise valid program. Skip direct-offset validation for gotox and accept it as a non-fallthrough terminal instruction. Its actual targets remain validated from the instruction-array jump table during CFG construction. Fixes: 493d9e0d6083 ("bpf, x86: add support for indirect jumps") Signed-off-by: Siddharth Chintamaneni Reviewed-by: Anton Protopopov Link: https://lore.kernel.org/r/20260902171414.96165-1-sidchintamaneni@gmail.com Signed-off-by: Alexei Starovoitov --- kernel/bpf/verifier.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 1b0b1fb62878..ddba53eaa333 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -3042,6 +3042,8 @@ static int check_subprogs(struct bpf_verifier_env *env) subprog[cur_subprog].exit_idx = i; goto next; } + if (insn_is_gotox(&insn[i])) + goto next; off = i + bpf_jmp_offset(&insn[i]) + 1; if (off < subprog_start || off >= subprog_end) { verbose(env, "jump out of range from insn %d to %d\n", i, off); @@ -3061,7 +3063,8 @@ static int check_subprogs(struct bpf_verifier_env *env) */ if (code != (BPF_JMP | BPF_EXIT) && code != (BPF_JMP32 | BPF_JA) && - code != (BPF_JMP | BPF_JA)) { + code != (BPF_JMP | BPF_JA) && + !insn_is_gotox(&insn[i])) { verbose(env, "last insn is not an exit or jmp\n"); bpf_diag_program_structure( env, i, "subprogram can fall through", From ac781acaef4904b9be64a7b5755c778e5d76ede6 Mon Sep 17 00:00:00 2001 From: Siddharth Chintamaneni Date: Wed, 2 Sep 2026 17:14:14 +0000 Subject: [PATCH 048/694] selftests/bpf: Test terminal gotox instructions Add tests that place gotox at the end of the main program and a subprogram, with each jump-table target preceding the gotox instruction. This tests gotox as a valid non-fallthrough terminal instruction. Signed-off-by: Siddharth Chintamaneni Reviewed-by: Anton Protopopov Link: https://lore.kernel.org/r/20260902171414.96165-2-sidchintamaneni@gmail.com Signed-off-by: Alexei Starovoitov --- .../selftests/bpf/progs/verifier_gotox.c | 48 +++++++++++++++++++ 1 file changed, 48 insertions(+) diff --git a/tools/testing/selftests/bpf/progs/verifier_gotox.c b/tools/testing/selftests/bpf/progs/verifier_gotox.c index 5b18c9a27717..0e27c2c79c57 100644 --- a/tools/testing/selftests/bpf/progs/verifier_gotox.c +++ b/tools/testing/selftests/bpf/progs/verifier_gotox.c @@ -47,6 +47,54 @@ DEFINE_SIMPLE_JUMP_TABLE_PROG(reserved_field_src_reg, BPF_REG_1, 0, 0, __fa DEFINE_SIMPLE_JUMP_TABLE_PROG(reserved_field_non_zero_off, BPF_REG_0, 1, 0, __failure __msg("BPF_JA|BPF_X uses reserved fields")) DEFINE_SIMPLE_JUMP_TABLE_PROG(reserved_field_non_zero_imm, BPF_REG_0, 0, 1, __failure __msg("BPF_JA|BPF_X uses reserved fields")) +#define DEFINE_TERMINAL_GOTOX_PROG(NAME, BASE) \ + __naked void NAME(void) \ + { \ + asm volatile (" \ + .pushsection .jumptables,\"\",@progbits; \ +jt0_%=: \ + .quad ret0_%= - " BASE "; \ + .size jt0_%=, 8; \ + .global jt0_%=; \ + .popsection; \ + \ + r0 = jt0_%= ll; \ + r0 = *(u64 *)(r0 + 0); \ + goto end_%=; \ +ret0_%=: \ + r0 = 0; \ + exit; \ +end_%=: \ + .8byte %[gotox_r0]; \ +" : \ + : __imm_insn(gotox_r0, BPF_RAW_INSN(BPF_JMP | BPF_JA | BPF_X, \ + BPF_REG_0, 0, 0, 0)) \ + : __clobber_all); \ + } + +SEC("socket") +__success __retval(0) +DEFINE_TERMINAL_GOTOX_PROG(jump_table_terminal_gotox, "socket") + +static __noinline __used +DEFINE_TERMINAL_GOTOX_PROG(terminal_gotox_subprog1, ".text") + +static __noinline __used int terminal_gotox_subprog2(void) +{ + return 0; +} + +SEC("socket") +__success __retval(0) +__naked void jump_table_terminal_gotox_subprog(void) +{ + asm volatile (" \ + call terminal_gotox_subprog1; \ + call terminal_gotox_subprog2; \ + exit; \ +" ::: __clobber_all); +} + /* * Gotox is forbidden when there is no jump table loaded * which points to the sub-function where the gotox is used From 85136bf22404474a815fc0ed26ec0d1cbc1bc3f9 Mon Sep 17 00:00:00 2001 From: "Jose Fernandez (Anthropic)" Date: Wed, 9 Sep 2026 17:51:04 +0000 Subject: [PATCH 049/694] bpf: Avoid soft lockup in __htab_map_lookup_and_delete_batch() __htab_map_lookup_and_delete_batch() has no rescheduling point. The batch count bounds how many entries are copied out, not how many buckets are visited, so one BPF_MAP_LOOKUP_BATCH call can walk the map end to end. The empty-bucket fast path is worse: it stays inside a single rcu_read_lock() / bpf_disable_instrumentation() section for any run of consecutive empty buckets. That holds up on small maps, but it falls apart at scale. On a 144-CPU arm64 host running a CONFIG_PREEMPT_NONE kernel, periodic BPF_MAP_LOOKUP_BATCH calls against an LRU hash map with 16,777,216 buckets held a CPU inside the batch op for 77+ seconds and triggered the soft lockup watchdog. Commit 75134f16e7dd ("bpf: Add schedule points in batch ops") fixed this same problem in the generic batch ops, but not in this htab-native path, which every htab-based hash map variant uses for its lookup[_and_delete] batch ops. Complete that fix here. Leave the critical section after 64 consecutive empty buckets, call cond_resched_tasks_rcu_qs(), and resume at the saved bucket cursor. No locks are held at that point, and resuming from the cursor is already the function's behavior for non-empty buckets. Add the same call to the per-bucket loop after copy_to_user(), where every lock has been dropped. cond_resched_rcu() is not enough here: sleeping with bpf_prog_active elevated makes tracing programs on that CPU silently skip their invocations. Plain cond_resched() is not enough either. It is a no-op under PREEMPT and PREEMPT_LAZY, the only models arm64 and x86 have offered since commit 7dadeaa6e851 ("sched: Further restrict the preemption modes"). It is also never a Tasks RCU quiescent state, in any model: the reschedule counts as a preemption. The walking task stays a holdout and stalls every synchronize_rcu_tasks() caller, ftrace and BPF trampoline teardown included, until the syscall returns [1]. cond_resched_tasks_rcu_qs() is the usual tool for that [2]. It reports the quiescent state at each yield and still reschedules as cond_resched() does on PREEMPT_NONE and PREEMPT_VOLUNTARY kernels. Fixes: 057996380a42 ("bpf: Add batch ops to all htab bpf map") Cc: "Paul E. McKenney" Cc: Rik van Riel Link: https://lore.kernel.org/bpf/20260715215314.44423f47@fangorn/ [1] Link: https://lore.kernel.org/bpf/9d444098-7c03-4163-af12-bd0a79a51443@paulmck-laptop/ [2] Assisted-by: LLM Signed-off-by: Jose Fernandez (Anthropic) Signed-off-by: Josef Bacik Reviewed-by: Rik van Riel Link: https://lore.kernel.org/r/20260909-b4-htab-batch-resched-v2-1-0cb529d8f95a@toxicpanda.com Signed-off-by: Alexei Starovoitov --- kernel/bpf/hashtab.c | 24 +++++++++++++++++++++--- 1 file changed, 21 insertions(+), 3 deletions(-) diff --git a/kernel/bpf/hashtab.c b/kernel/bpf/hashtab.c index 6f331c80130d..a72dc5b9f184 100644 --- a/kernel/bpf/hashtab.c +++ b/kernel/bpf/hashtab.c @@ -1772,6 +1772,12 @@ static int htab_lru_percpu_map_lookup_and_delete_elem(struct bpf_map *map, flags); } +/* + * Max consecutive empty buckets to walk in one RCU + + * instrumentation-disabled section before rescheduling. + */ +#define HTAB_BATCH_EMPTY_RESCHED 64 + static int __htab_map_lookup_and_delete_batch(struct bpf_map *map, const union bpf_attr *attr, @@ -1793,6 +1799,7 @@ __htab_map_lookup_and_delete_batch(struct bpf_map *map, unsigned long flags = 0; bool locked = false; struct htab_elem *l; + u32 empty_cnt = 0; struct bucket *b; int ret = 0; @@ -1971,12 +1978,21 @@ __htab_map_lookup_and_delete_batch(struct bpf_map *map, } next_batch: - /* If we are not copying data, we can go to next bucket and avoid - * unlocking the rcu. + /* + * If we are not copying data, we can go to next bucket and avoid + * unlocking the rcu. Bound the walk though: after + * HTAB_BATCH_EMPTY_RESCHED consecutive empty buckets, fully exit + * the critical section (no locks are held here) and reschedule. */ if (!bucket_cnt && (batch + 1 < htab->n_buckets)) { batch++; - goto again_nocopy; + if (++empty_cnt < HTAB_BATCH_EMPTY_RESCHED) + goto again_nocopy; + empty_cnt = 0; + rcu_read_unlock(); + bpf_enable_instrumentation(); + cond_resched_tasks_rcu_qs(); + goto again; } rcu_read_unlock(); @@ -1990,11 +2006,13 @@ __htab_map_lookup_and_delete_batch(struct bpf_map *map, } total += bucket_cnt; + empty_cnt = 0; batch++; if (batch >= htab->n_buckets) { ret = -ENOENT; goto after_loop; } + cond_resched_tasks_rcu_qs(); goto again; after_loop: From 75f8cf22463d82bb1fb0239a3d485fc8f4c8ef03 Mon Sep 17 00:00:00 2001 From: Jiayuan Chen Date: Thu, 3 Sep 2026 18:09:20 +0800 Subject: [PATCH 050/694] bpf: Fix out-of-bounds read of rtt_min in sock_ops A sockops prog reading skops->rtt_min never checks the sk type: on the tcp_conn_request() path sock_ops->sk is a request_sock (non-full), and the ctx rewrite casts it to a tcp_sock (full) and reads rtt_min past the end of the request_sock, returning dirty adjacent memory. SEC("sockops") int prog(struct bpf_sock_ops *skops) { switch (skops->op) { case BPF_SOCK_OPS_RWND_INIT: leak = skops->rtt_min; /* reads the request_sock OOB */ ... } } For instance one such read returned rtt_min=0xffff8881, the high half of a leaked kernel pointer. Guarding that cast is exactly what SOCK_OPS_GET_FIELD() does -- it checks is_locked_tcp_sock and returns 0 when sock_ops->sk is not a locked full socket. Every other tcp_sock field in sock_ops goes through it; rtt_min is the only one open-coded, so it skips the check. Read rtt_min through SOCK_OPS_GET_FIELD() too. rtt_min is a bit special: it is a struct minmax and we only want the current min, so pass rtt_min.s[0].v. That is equivalent to the old hand-computed offset offsetof(struct tcp_sock, rtt_min) + sizeof_field(struct minmax_sample, t) (s[0] sits at rtt_min + 0 and .v at + sizeof(.t), i.e. what minmax_get() returns), so the loaded field is unchanged and only the full-sock guard is added. The two BUILD_BUG_ON()s that protected the hand-computed offset are no longer needed. Before patch: 0: r1 = *(u64 *)(r1 +0) ; r1 = skops->sk 1: r1 = *(u32 *)(r1 +2324) ; ((tcp_sock *)sk)->rtt_min.s[0].v After patch: 0: *(u64 *)(r1 +56) = r9 1: r9 = *(u8 *)(r1 +50) ; is_locked_tcp_sock 2: if r9 == 0 goto pc+4 ; not a locked full sock -> 0 3: r9 = *(u64 *)(r1 +56) 4: r1 = *(u64 *)(r1 +0) ; r1 = skops->sk 5: r1 = *(u32 *)(r1 +2324) ; rtt_min.s[0].v 6: goto pc+2 7: r9 = *(u64 *)(r1 +56) 8: r1 = 0 Fixes: 44f0e43037d3 ("bpf: Add support for reading sk_state and more") Reported-by: VEGA Signed-off-by: Jiayuan Chen Reviewed-by: Emil Tsalapatis Link: https://lore.kernel.org/r/20260903100921.113374-1-jiayuan.chen@linux.dev Signed-off-by: Alexei Starovoitov --- net/core/filter.c | 13 +------------ 1 file changed, 1 insertion(+), 12 deletions(-) diff --git a/net/core/filter.c b/net/core/filter.c index cae43b999162..532405988fd9 100644 --- a/net/core/filter.c +++ b/net/core/filter.c @@ -11106,18 +11106,7 @@ static u32 sock_ops_convert_ctx_access(enum bpf_access_type type, break; case offsetof(struct bpf_sock_ops, rtt_min): - BUILD_BUG_ON(sizeof_field(struct tcp_sock, rtt_min) != - sizeof(struct minmax)); - BUILD_BUG_ON(sizeof(struct minmax) < - sizeof(struct minmax_sample)); - - *insn++ = BPF_LDX_MEM(BPF_FIELD_SIZEOF( - struct bpf_sock_ops_kern, sk), - si->dst_reg, si->src_reg, - offsetof(struct bpf_sock_ops_kern, sk)); - *insn++ = BPF_LDX_MEM(BPF_W, si->dst_reg, si->dst_reg, - offsetof(struct tcp_sock, rtt_min) + - sizeof_field(struct minmax_sample, t)); + SOCK_OPS_GET_FIELD(rtt_min, rtt_min.s[0].v, struct tcp_sock); break; case offsetof(struct bpf_sock_ops, bpf_sock_ops_cb_flags): From 7d70a0b02d262971201fdd1e221586bdd95c2910 Mon Sep 17 00:00:00 2001 From: Zhixing Chen Date: Thu, 3 Sep 2026 18:43:58 +0800 Subject: [PATCH 051/694] bpf: Use kvfree() in xdp_test_run_teardown() xdp_test_run_setup() allocates xdp->frames and xdp->skbs with kvmalloc_array(). The setup error path already releases both arrays with kvfree(), while the normal teardown path still uses kfree(). Use kvfree() in xdp_test_run_teardown() as well, so the release helper matches the allocator on both paths. Signed-off-by: Zhixing Chen Reviewed-by: Emil Tsalapatis Link: https://lore.kernel.org/r/20260903104358.29228-1-running910@gmail.com Signed-off-by: Alexei Starovoitov --- net/bpf/test_run.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/net/bpf/test_run.c b/net/bpf/test_run.c index 5d51f6cb7d15..513354e928cb 100644 --- a/net/bpf/test_run.c +++ b/net/bpf/test_run.c @@ -205,8 +205,8 @@ static void xdp_test_run_teardown(struct xdp_test_data *xdp) { xdp_unreg_mem_model(&xdp->mem); page_pool_destroy(xdp->pp); - kfree(xdp->frames); - kfree(xdp->skbs); + kvfree(xdp->frames); + kvfree(xdp->skbs); } static bool frame_was_changed(const struct xdp_page_head *head) From 1c21452d02eec2f008e2c5535820f85adbd7587a Mon Sep 17 00:00:00 2001 From: Pu Lehui Date: Sat, 5 Sep 2026 02:11:39 +0000 Subject: [PATCH 052/694] bpf: Fix UAF due to concurrent consumption of ttrace lists in alloc_bulk Syzkaller repeatedly triggered UAF splats related to nodes in waiting_for_gp_ttrace within the bpf memalloc: BUG: KASAN: slab-use-after-free in llist_del_first+0x85/0x110 lib/llist.c:61 Read of size 8 at addr ffff8881572cd080 by task syz.4.470/5112 ... llist_del_first+0x85/0x110 lib/llist.c:61 alloc_bulk+0x193/0x460 kernel/bpf/memalloc.c:229 bpf_mem_refill+0x386/0x560 kernel/bpf/memalloc.c:436 Freed by task 14: ... __free_rcu kernel/bpf/memalloc.c:281 [inline] __free_rcu_tasks_trace+0x48/0xd0 kernel/bpf/memalloc.c:291 rcu_tasks_invoke_cbs+0x1ec/0x3e0 kernel/rcu/tasks.h:571 rcu_tasks_one_gp+0x13d/0x220 kernel/rcu/tasks.h:621 rcu_tasks_kthread+0xf3/0x120 kernel/rcu/tasks.h:651 The reason is that the UAF occurs after the RCU Tasks Trace GP expires: when the __free_rcu() callback runs, there is no synchronization protecting llist_del_all() against concurrent alloc_bulk() operating on waiting_for_gp_ttrace, leading to the race condition below: CPU0 CPU1 __free_rcu (RCU Tasks Trace callback) alloc_bulk llist_del_first(&c->waiting_for_gp_ttrace) entry = smp_load_acquire(&head->first); do { if (entry == NULL) return NULL; free_all(llist_del_all(&c->waiting_for_gp_ttrace)) llist_for_each_safe(pos, t, llnode) free_one(pos); next = READ_ONCE(entry->next); <-- trigger UAF } while (!try_cmpxchg(&head->first, &entry, next)); In addition, there is also a theoretical race condition on the free_by_rcu_ttrace list. This race requires two preconditions: an in-flight Tasks Trace GP keeping c->call_rcu_ttrace_in_progress == 1, and concurrent cross-CPU frees repopulating c->free_by_rcu_ttrace with new nodes. Under these conditions, the following scenario triggers UAF: // CPU0 // irq work is still busy (on PREEMPT_RT) alloc_bulk() llist_del_first(&c->free_by_rcu_ttrace) entry = smp_load_acquire(&head->first); do { if (entry == NULL) return NULL; // CPU1 bpf_mem_alloc_destroy() WRITE_ONCE(c->draining, true) // wait for CPU0 irq_work_sync() // CPU2 do_call_rcu_ttrace(tgt(CPU0)) if (c->draining) { llist_del_all(&c->free_by_rcu_ttrace) free_all() } // CPU0 continue next = READ_ONCE(entry->next); <-- trigger UAF while (!try_cmpxchg(&head->first, &entry, next)); Fix this by introducing a raw spinlock to synchronize the concurrent consumption on waiting_for_gp_ttrace and free_by_rcu_ttrace. Fixes: 04fabf00b4d3 ("bpf: Allow reuse from waiting_for_gp_ttrace list.") Suggested-by: Alexei Starovoitov Suggested-by: Hou Tao Signed-off-by: Pu Lehui Acked-by: Hou Tao Link: https://lore.kernel.org/r/20260905021139.4116529-1-pulehui@huaweicloud.com Signed-off-by: Alexei Starovoitov --- kernel/bpf/memalloc.c | 50 ++++++++++++++++++++++++------------------- 1 file changed, 28 insertions(+), 22 deletions(-) diff --git a/kernel/bpf/memalloc.c b/kernel/bpf/memalloc.c index e9662db7198f..8a8f088e83e6 100644 --- a/kernel/bpf/memalloc.c +++ b/kernel/bpf/memalloc.c @@ -119,6 +119,7 @@ struct bpf_mem_cache { struct llist_head waiting_for_gp_ttrace; struct rcu_head rcu_ttrace; atomic_t call_rcu_ttrace_in_progress; + raw_spinlock_t lock; }; struct bpf_mem_caches { @@ -214,25 +215,24 @@ static void alloc_bulk(struct bpf_mem_cache *c, int cnt, int node, bool atomic) gfp = __GFP_NOWARN | __GFP_ACCOUNT; gfp |= atomic ? GFP_NOWAIT : GFP_KERNEL; - for (i = 0; i < cnt; i++) { - /* - * For every 'c' llist_del_first(&c->free_by_rcu_ttrace); is - * done only by one CPU == current CPU. Other CPUs might - * llist_add() and llist_del_all() in parallel. - */ - obj = llist_del_first(&c->free_by_rcu_ttrace); - if (!obj) - break; - add_obj_to_free_list(c, obj); - } - if (i >= cnt) - return; + /* + * c->lock serializes concurrent llist_del_first() against + * llist_del_all() in __free_rcu() and do_call_rcu_ttrace(). + */ + scoped_guard(raw_spinlock_irqsave, &c->lock) { + for (i = 0; i < cnt; i++) { + obj = llist_del_first(&c->free_by_rcu_ttrace); + if (!obj) + break; + add_obj_to_free_list(c, obj); + } - for (; i < cnt; i++) { - obj = llist_del_first(&c->waiting_for_gp_ttrace); - if (!obj) - break; - add_obj_to_free_list(c, obj); + for (; i < cnt; i++) { + obj = llist_del_first(&c->waiting_for_gp_ttrace); + if (!obj) + break; + add_obj_to_free_list(c, obj); + } } if (i >= cnt) return; @@ -279,8 +279,12 @@ static int free_all(struct bpf_mem_cache *c, struct llist_node *llnode, bool per static void __free_rcu(struct rcu_head *head) { struct bpf_mem_cache *c = container_of(head, struct bpf_mem_cache, rcu_ttrace); + struct llist_node *llnode; - free_all(c, llist_del_all(&c->waiting_for_gp_ttrace), !!c->percpu_size); + scoped_guard(raw_spinlock_irqsave, &c->lock) + llnode = llist_del_all(&c->waiting_for_gp_ttrace); + + free_all(c, llnode, !!c->percpu_size); atomic_set(&c->call_rcu_ttrace_in_progress, 0); } @@ -300,7 +304,8 @@ static void do_call_rcu_ttrace(struct bpf_mem_cache *c) if (atomic_xchg(&c->call_rcu_ttrace_in_progress, 1)) { if (unlikely(READ_ONCE(c->draining))) { - llnode = llist_del_all(&c->free_by_rcu_ttrace); + scoped_guard(raw_spinlock_irqsave, &c->lock) + llnode = llist_del_all(&c->free_by_rcu_ttrace); free_all(c, llnode, !!c->percpu_size); } return; @@ -535,6 +540,7 @@ int bpf_mem_alloc_init(struct bpf_mem_alloc *ma, int size, bool percpu) c->objcg = objcg; c->percpu_size = percpu_size; c->tgt = c; + raw_spin_lock_init(&c->lock); init_refill_work(c); prefill_mem_cache(c, cpu); } @@ -557,7 +563,7 @@ int bpf_mem_alloc_init(struct bpf_mem_alloc *ma, int size, bool percpu) c->objcg = objcg; c->percpu_size = percpu_size; c->tgt = c; - + raw_spin_lock_init(&c->lock); init_refill_work(c); prefill_mem_cache(c, cpu); } @@ -609,7 +615,7 @@ int bpf_mem_alloc_percpu_unit_init(struct bpf_mem_alloc *ma, int size) c->objcg = objcg; c->percpu_size = percpu_size; c->tgt = c; - + raw_spin_lock_init(&c->lock); init_refill_work(c); prefill_mem_cache(c, cpu); } From 490a83d6386eec1d29f470c8d7331677fb46c3b7 Mon Sep 17 00:00:00 2001 From: Geliang Tang Date: Tue, 8 Sep 2026 17:08:32 +0800 Subject: [PATCH 053/694] bpf, sockmap: Fix self-redirect copied_seq double-counting When a BPF stream_verdict program redirects an skb back to the same socket (self-redirect with BPF_F_INGRESS), sk_psock_verdict_apply() calls tcp_eat_skb() which advances tcp_sk->copied_seq. However, the skb is then delivered to the socket's psock ingress queue and later read by tcp_bpf_recvmsg_parser(), which also advances copied_seq via the copied_from_self accounting path. This double-counting causes copied_seq to advance by 2x the actual data length, triggering: TCP recvmsg seq # bug 2: copied BF2E806, seq BF2E7FD, \ rcvnxt BF2E806, fl 0 WARNING: net/ipv4/tcp.c:2745 at tcp_recvmsg_locked+0x72b/0x2640 Call Trace: tcp_recvmsg+0x10a/0x500 sock_recvmsg+0x168/0x1d0 __sys_recvfrom+0x19a/0x2a0 __x64_sys_recvfrom+0xe4/0x1f0 do_syscall_64+0xf7/0x530 entry_SYSCALL_64_after_hwframe+0x77/0x7f cleanup rbuf bug: copied BF2E806 seq BF2E806 rcvnxt BF2E806 WARNING: net/ipv4/tcp.c:1609 at tcp_cleanup_rbuf+0xf2/0x1c0 Call Trace: tcp_recvmsg_locked+0x8d1/0x2640 tcp_recvmsg+0x10a/0x500 sock_recvmsg+0x168/0x1d0 __sys_recvfrom+0x19a/0x2a0 __x64_sys_recvfrom+0xe4/0x1f0 do_syscall_64+0xf7/0x530 entry_SYSCALL_64_after_hwframe+0x77/0x7f Fix this by converting self-redirect verdict to __SK_PASS at the beginning of sk_psock_verdict_apply(). This bypasses the __SK_REDIRECT case entirely (which calls sk_psock_eat_skb), letting the __SK_PASS path queue the skb to the psock ingress queue. The data is then read via tcp_bpf_recvmsg_parser(), which advances copied_seq exactly once through copied_from_self. Cross-socket redirects continue through __SK_REDIRECT with sk_psock_eat_skb() unchanged. Fixes: e5c6de5fa025 ("bpf, sockmap: Incorrectly handling copied_seq") Suggested-by: Jakub Sitnicki Suggested-by: Jiayuan Chen Signed-off-by: Geliang Tang Reviewed-by: Emil Tsalapatis Reviewed-by: Jiayuan Chen Link: https://lore.kernel.org/r/1a8e797a1b26e2f695aaac22ac644c2862f63466.1788858299.git.tanggeliang@kylinos.cn Signed-off-by: Alexei Starovoitov --- net/core/skmsg.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/net/core/skmsg.c b/net/core/skmsg.c index 2521b643fa05..df385a5a961e 100644 --- a/net/core/skmsg.c +++ b/net/core/skmsg.c @@ -1000,6 +1000,10 @@ static int sk_psock_verdict_apply(struct sk_psock *psock, struct sk_buff *skb, int err = 0; u32 len, off; + if (verdict == __SK_REDIRECT && skb_bpf_ingress(skb) && + skb_bpf_redirect_fetch(skb) == psock->sk) + verdict = __SK_PASS; + switch (verdict) { case __SK_PASS: err = -EIO; From 953824e508b27d12837e32ef37ef6248e1f6fc7a Mon Sep 17 00:00:00 2001 From: Masoud Aghasi Date: Thu, 3 Sep 2026 09:27:34 +0100 Subject: [PATCH 054/694] bpf: Fix u32 overflow issue in map batch operations Several map batch operation implementations such as generic_map_lookup_batch() use calculations in the form of "values + cp * map->value_size" to compute the desired userspace memory address for reading or writing. This can overflow the u32 type (the result of "cp * map->value_size") when the map size exceeds 4GB. generic_map_lookup_batch() may corrupt values for some keys in userspace memory, and in some cases it mismatches values for some keys while still reporting success. Other batch operations may fail to delete or update some keys, or the syscall may return unexpected errors. Add size_t casts to prevent the affected offset and size calculations from overflowing. Fixes: cb4d03ab499d ("bpf: Add generic support for lookup batch op") Fixes: aa2e93b8e58e ("bpf: Add generic support for update and delete batch ops") Fixes: 057996380a42 ("bpf: Add batch ops to all htab bpf map") Signed-off-by: Masoud Aghasi Link: https://lore.kernel.org/r/20260903082734.623904-1-maghasi@disroot.org Signed-off-by: Alexei Starovoitov --- kernel/bpf/hashtab.c | 8 ++++---- kernel/bpf/syscall.c | 10 +++++----- 2 files changed, 9 insertions(+), 9 deletions(-) diff --git a/kernel/bpf/hashtab.c b/kernel/bpf/hashtab.c index a72dc5b9f184..4f495dcbf670 100644 --- a/kernel/bpf/hashtab.c +++ b/kernel/bpf/hashtab.c @@ -1997,10 +1997,10 @@ __htab_map_lookup_and_delete_batch(struct bpf_map *map, rcu_read_unlock(); bpf_enable_instrumentation(); - if (bucket_cnt && (copy_to_user(ukeys + total * key_size, keys, - key_size * bucket_cnt) || - copy_to_user(uvalues + total * value_size, values, - value_size * bucket_cnt))) { + if (bucket_cnt && (copy_to_user(ukeys + (size_t)total * key_size, keys, + (size_t)key_size * bucket_cnt) || + copy_to_user(uvalues + (size_t)total * value_size, values, + (size_t)value_size * bucket_cnt))) { ret = -EFAULT; goto after_loop; } diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c index c7bc9ba9b331..853b47f81384 100644 --- a/kernel/bpf/syscall.c +++ b/kernel/bpf/syscall.c @@ -2036,7 +2036,7 @@ int generic_map_delete_batch(struct bpf_map *map, for (cp = 0; cp < max_count; cp++) { err = -EFAULT; - if (copy_from_user(key, keys + cp * map->key_size, + if (copy_from_user(key, keys + (size_t)cp * map->key_size, map->key_size)) break; @@ -2098,9 +2098,9 @@ int generic_map_update_batch(struct bpf_map *map, struct file *map_file, for (cp = 0; cp < max_count; cp++) { err = -EFAULT; - if (copy_from_user(key, keys + cp * map->key_size, + if (copy_from_user(key, keys + (size_t)cp * map->key_size, map->key_size) || - copy_from_user(value, values + cp * value_size, value_size)) + copy_from_user(value, values + (size_t)cp * value_size, value_size)) break; err = bpf_map_update_value(map, map_file, key, value, @@ -2179,12 +2179,12 @@ int generic_map_lookup_batch(struct bpf_map *map, if (err) goto free_buf; - if (copy_to_user(keys + cp * map->key_size, key, + if (copy_to_user(keys + (size_t)cp * map->key_size, key, map->key_size)) { err = -EFAULT; goto free_buf; } - if (copy_to_user(values + cp * value_size, value, value_size)) { + if (copy_to_user(values + (size_t)cp * value_size, value, value_size)) { err = -EFAULT; goto free_buf; } From ef1fb82f12186dd26153b14d9fbcf4ec98db81b3 Mon Sep 17 00:00:00 2001 From: Donggeun Yoo Date: Mon, 7 Sep 2026 22:06:23 +0900 Subject: [PATCH 055/694] bpf, arm64: set up the frame pointer for the exception callback A program acting as exception boundary saves all callee-saved registers, so build_prologue() takes the exception_cb path and never calls push_callee_regs(). That is the only place find_used_callee_regs() runs, and with it the only place ctx->fp_used is set, so the callback prologue does not emit the mov x25, sp that points BPF_REG_FP at the frame the callback runs on. x25 keeps whatever it held when bpf_throw() was called. If the throw came from a subprogram that uses its own BPF stack, that is the subprogram's frame pointer, and since the subprogram never returns it never restores x25 either. Stack accesses through BPF_REG_FP are rewritten to be stack pointer relative, so those still land in the callback's own frame. Materializing the register does not: a callback that passes the address of a local variable to a helper hands over an address in the dead subprogram's frame. That address is below the callback's stack pointer by then, and the helper's own call chain covers it, so the helper can write over its own return address. 0x1234 below is the value the helper was asked to store: pc : 0x1234 lr : 0x1234 Call trace: 0x1234 (P) bpf_test_run+0x188/0x3e0 bpf_prog_test_run_skb+0x47c/0x998 __sys_bpf+0xbdc/0xdd8 Kernel panic - not syncing: Oops: Fatal exception in interrupt Set ctx->fp_used on the exception callback path so that the existing code further down sets x25 from the stack pointer. The epilogue restores it from the main program's save area along with the other callee-saved registers, as it already does. x86 sets the frame pointer for the callback from the argument it is passed, and powerpc computes it from the stack pointer. Fixes: 5d4fa9ec5643 ("bpf, arm64: Avoid blindly saving/restoring all callee-saved registers") Acked-by: Xu Kuohai Signed-off-by: Donggeun Yoo Link: https://lore.kernel.org/r/20260907130624.611942-2-donggeunyoo.kernel@gmail.com Signed-off-by: Alexei Starovoitov --- arch/arm64/net/bpf_jit_comp.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/arch/arm64/net/bpf_jit_comp.c b/arch/arm64/net/bpf_jit_comp.c index c18e005a41db..c5f55d6161fe 100644 --- a/arch/arm64/net/bpf_jit_comp.c +++ b/arch/arm64/net/bpf_jit_comp.c @@ -600,6 +600,8 @@ static int build_prologue(struct jit_ctx *ctx, bool ebpf_from_cbpf) * 12 registers are on the stack */ emit(A64_SUB_I(1, A64_SP, A64_FP, 96), ctx); + /* The callback may use its own BPF stack, set up fp for it. */ + ctx->fp_used = true; } /* Stack must be multiples of 16B */ From 26a43a5f8c31b9132dc40a449d0f6c7ecfaa1f40 Mon Sep 17 00:00:00 2001 From: Donggeun Yoo Date: Mon, 7 Sep 2026 22:06:24 +0900 Subject: [PATCH 056/694] selftests/bpf: cover the exception callback using its own BPF stack The existing exception tests do not reach a callback that materializes BPF_REG_FP into a register. They either throw from the main program, where BPF_REG_FP already holds the value the callback needs, or use a callback whose only stack accesses are frame pointer relative, which the arm64 JIT rewrites to be stack pointer relative. Add a test that throws from a subprogram using its own BPF stack, with a callback that hands the address of a local variable to bpf_probe_read_kernel(). The helper and the callback have to name the same slot for the value read back to be the one the helper stored. Signed-off-by: Donggeun Yoo Link: https://lore.kernel.org/r/20260907130624.611942-3-donggeunyoo.kernel@gmail.com Signed-off-by: Alexei Starovoitov --- .../selftests/bpf/prog_tests/exceptions.c | 1 + .../testing/selftests/bpf/progs/exceptions.c | 30 +++++++++++++++++++ 2 files changed, 31 insertions(+) diff --git a/tools/testing/selftests/bpf/prog_tests/exceptions.c b/tools/testing/selftests/bpf/prog_tests/exceptions.c index 3588d6f97fd4..639866ce09a9 100644 --- a/tools/testing/selftests/bpf/prog_tests/exceptions.c +++ b/tools/testing/selftests/bpf/prog_tests/exceptions.c @@ -55,6 +55,7 @@ static void test_exceptions_success(void) RUN_SUCCESS(exception_ext, 0); RUN_SUCCESS(exception_ext_mod_cb_runtime, 35); RUN_SUCCESS(exception_throw_subprog, 1); + RUN_SUCCESS(exception_throw_subprog_stack_cb, 0x1234); RUN_SUCCESS(exception_assert_nz_gfunc, 1); RUN_SUCCESS(exception_assert_zero_gfunc, 1); RUN_SUCCESS(exception_assert_neg_gfunc, 1); diff --git a/tools/testing/selftests/bpf/progs/exceptions.c b/tools/testing/selftests/bpf/progs/exceptions.c index c8d716fbd419..91c81971e58c 100644 --- a/tools/testing/selftests/bpf/progs/exceptions.c +++ b/tools/testing/selftests/bpf/progs/exceptions.c @@ -212,6 +212,36 @@ int exception_throw_subprog(struct __sk_buff *ctx) return 0; } +u64 exception_cb_stack_src = 0x1234; + +/* + * The address handed to the helper has to be this callback's own stack + * slot, not one from a frame that is already gone. + */ +__noinline int exception_cb_stack(u64 cookie) +{ + volatile u64 val = 0xdead; + + bpf_probe_read_kernel((void *)&val, sizeof(val), &exception_cb_stack_src); + return val; +} + +/* Throws from a subprogram that has a stack of its own. */ +__noinline static int throwing_subprog_stack(struct __sk_buff *ctx) +{ + volatile u64 pad[4] = {}; + + bpf_throw(pad[0]); + return 0; +} + +SEC("tc") +__exception_cb(exception_cb_stack) +int exception_throw_subprog_stack_cb(struct __sk_buff *ctx) +{ + return throwing_subprog_stack(ctx); +} + __noinline int assert_nz_gfunc(u64 c) { volatile u64 cookie = c; From 692f32609a30f75ca3401e25b504bfd06bd5662a Mon Sep 17 00:00:00 2001 From: Sean Anderson Date: Mon, 17 Aug 2026 12:22:11 -0400 Subject: [PATCH 057/694] pinctrl: meson: Fix typo in s4 group name One of the i2c pin groups has some junk at the end. The name should be i2c2_scl_h1, and indeed that's the name used by i2c2_pins3 in meson-s4.dtsi. Fixes: 775214d389c25 ("pinctrl: meson: add pinctrl driver support for Meson-S4 Soc") Signed-off-by: Sean Anderson Reviewed-by: Neil Armstrong Signed-off-by: Linus Walleij --- drivers/pinctrl/meson/pinctrl-meson-s4.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/pinctrl/meson/pinctrl-meson-s4.c b/drivers/pinctrl/meson/pinctrl-meson-s4.c index 872948699e9f..365dafe457a9 100644 --- a/drivers/pinctrl/meson/pinctrl-meson-s4.c +++ b/drivers/pinctrl/meson/pinctrl-meson-s4.c @@ -854,7 +854,7 @@ static const char * const i2c1_groups[] = { static const char * const i2c2_groups[] = { "i2c2_sda_d", "i2c2_scl_d", "i2c2_sda_h8", "i2c2_scl_h9", - "i2c2_sda_h0", "i2c2_scl_h1l," + "i2c2_sda_h0", "i2c2_scl_h1", }; static const char * const i2c3_groups[] = { From 51ae99659469edba2e83931efa26b77d36ca02c2 Mon Sep 17 00:00:00 2001 From: Sarah Emery Date: Fri, 28 Aug 2026 17:58:17 +0200 Subject: [PATCH 058/694] pinctrl: generic: serialise pinctrl_generic_dt_node_to_map() pinctrl_generic_add_group() documents that the caller must take care of locking, and pinmux_generic_add_function() needs it too, but pinctrl_generic_dt_node_to_map() calls them without holding pctldev->mutex, and the core caller in create_pinctrl() does not take it either. The driver core calls pinctrl_bind_pins() before probing a device, so two devices that reference the same pin controller can run pinctrl_generic_dt_node_to_map() on one pctldev at the same time. Both `add` functions take the new selector from pctldev->num_groups or pctldev->num_functions, and radix_tree_insert() at that index. Two racing callers can read the same selector before either has inserted, so the second insert collides and fails: k1-pinctrl d401e000.pinctrl: error -EEXIST: error adding function pcie2-0-cfg k1-pinctrl d401e000.pinctrl: does not have pin group pcie0-0-cfg.pcie0-0-pins leaving one consumer without its pin configuration. This was hit on a SpacemiT K3 board, where PCIe devices probe in parallel against the single shared pin controller. Take pctldev->mutex across the whole function, so that the groups and the function referring are in a single critical section. Fixes: 43722575e5cd ("pinctrl: add generic functions + pins mapper") Signed-off-by: Sarah Emery Signed-off-by: Linus Walleij --- drivers/pinctrl/pinctrl-generic.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/drivers/pinctrl/pinctrl-generic.c b/drivers/pinctrl/pinctrl-generic.c index fd6bdb74028a..4277c8748513 100644 --- a/drivers/pinctrl/pinctrl-generic.c +++ b/drivers/pinctrl/pinctrl-generic.c @@ -3,8 +3,10 @@ #define pr_fmt(fmt) "generic pinconfig core: " fmt #include +#include #include #include +#include #include #include @@ -196,6 +198,8 @@ static int pinctrl_generic_dt_node_to_map(struct pinctrl_dev *pctldev, int ngroups = 0; int ret; + guard(mutex)(&pctldev->mutex); + *maps = NULL; *num_maps = 0; From e6c5d6c589764a41218cbd81bd7d6c9b906e2cf0 Mon Sep 17 00:00:00 2001 From: Conor Dooley Date: Mon, 31 Aug 2026 11:50:58 +0100 Subject: [PATCH 059/694] pinctrl: mpfs-mssio: fix width of unused bank voltage setting The bank voltages are only 4 bits wide, so when a pin was unused the driver was not correctly interpreting it as being at zero volts, because the driver's value for unused had two extra set bits. CC: stable@vger.kernel.org Fixes: 488d704ed7b7 ("pinctrl: add polarfire soc mssio pinctrl driver") Signed-off-by: Conor Dooley Signed-off-by: Linus Walleij --- drivers/pinctrl/microchip/pinctrl-mpfs-mssio.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/pinctrl/microchip/pinctrl-mpfs-mssio.c b/drivers/pinctrl/microchip/pinctrl-mpfs-mssio.c index ea1026a0d22c..dafca82f3e54 100644 --- a/drivers/pinctrl/microchip/pinctrl-mpfs-mssio.c +++ b/drivers/pinctrl/microchip/pinctrl-mpfs-mssio.c @@ -86,7 +86,7 @@ static struct mpfs_pinctrl_bank_voltage mpfs_pinctrl_bank_voltages[8] = { { .uv = 1800000, .val = 4 }, { .uv = 2500000, .val = 6 }, { .uv = 3300000, .val = 8 }, - { .uv = 0, .val = 0x3f }, // pin unused + { .uv = 0, .val = 0xf }, // pin unused }; static int mpfs_pinctrl_get_drive_strength_ma(u32 drive_strength) From 8039b75af5808572ff3656eaed07d348aed3989a Mon Sep 17 00:00:00 2001 From: Conor Dooley Date: Mon, 31 Aug 2026 11:50:59 +0100 Subject: [PATCH 060/694] pinctrl: mpfs-mssio: use correct regmap function to set bank voltage regmap_assign_bits() is not the correct function to use for an RMW operation, as it maps to regmap_set_bits() or regmap_clear_bits() and the former will never zero a bit. Use regmap_update_bits() instead, which will actually set the bank voltages to what have been requested. CC: stable@vger.kernel.org Fixes: 488d704ed7b7 ("pinctrl: add polarfire soc mssio pinctrl driver") Signed-off-by: Conor Dooley Signed-off-by: Linus Walleij --- drivers/pinctrl/microchip/pinctrl-mpfs-mssio.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/pinctrl/microchip/pinctrl-mpfs-mssio.c b/drivers/pinctrl/microchip/pinctrl-mpfs-mssio.c index dafca82f3e54..92d38e5336de 100644 --- a/drivers/pinctrl/microchip/pinctrl-mpfs-mssio.c +++ b/drivers/pinctrl/microchip/pinctrl-mpfs-mssio.c @@ -156,10 +156,10 @@ static void mpfs_pinctrl_set_bank_voltage(struct mpfs_pinctrl *pctrl, unsigned i u32 val = FIELD_PREP(MPFS_PINCTRL_BANK_VOLTAGE_MASK, bank_voltage); if (pin < MPFS_PINCTRL_BANK2_START) - regmap_assign_bits(pctrl->sysreg_regmap, MPFS_PINCTRL_MSSIO_BANK4_CFG_CR, + regmap_update_bits(pctrl->sysreg_regmap, MPFS_PINCTRL_MSSIO_BANK4_CFG_CR, MPFS_PINCTRL_BANK_VOLTAGE_MASK, val); else - regmap_assign_bits(pctrl->sysreg_regmap, MPFS_PINCTRL_MSSIO_BANK2_CFG_CR, + regmap_update_bits(pctrl->sysreg_regmap, MPFS_PINCTRL_MSSIO_BANK2_CFG_CR, MPFS_PINCTRL_BANK_VOLTAGE_MASK, val); } From 44b8a0bf5f96e8393312fc34b956766882341f16 Mon Sep 17 00:00:00 2001 From: Runyu Xiao Date: Thu, 27 Aug 2026 16:26:56 +0800 Subject: [PATCH 061/694] mtd: spi-nor: core: Fix mutex leak in spi_nor_rww_start_exclusive() Use the same guard(mutex) pattern as the other RWW helpers so nor->lock is released on both the busy and successful return paths. Fixes: 03e7bb864d9a ("mtd: spi-nor: use scope-based mutex cleanup helpers") Cc: stable@vger.kernel.org Signed-off-by: Runyu Xiao Reviewed-by: Miquel Raynal Reviewed-by: Tudor Ambarus [mw: rephrased commit message] Signed-off-by: Michael Walle --- drivers/mtd/spi-nor/core.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/mtd/spi-nor/core.c b/drivers/mtd/spi-nor/core.c index ccf4396cdcd0..8bc117b46e02 100644 --- a/drivers/mtd/spi-nor/core.c +++ b/drivers/mtd/spi-nor/core.c @@ -1310,7 +1310,7 @@ static bool spi_nor_rww_start_exclusive(struct spi_nor *nor) { struct spi_nor_rww *rww = &nor->rww; - mutex_lock(&nor->lock); + guard(mutex)(&nor->lock); if (rww->ongoing_io || rww->ongoing_rd || rww->ongoing_pe) return false; From ba13f1f32d96d7ce9069ae4f32404ecde8da89c4 Mon Sep 17 00:00:00 2001 From: Meijing Zhao Date: Tue, 15 Sep 2026 20:46:41 +0800 Subject: [PATCH 062/694] mm: memblock: add missing HugeTLB flag name Commit 7d163a75f821 ("memblock: make HugeTLB bootmem allocation work with KHO") added MEMBLOCK_RSRV_HUGETLB but did not add the corresponding entry to flagname[]. As a result, memblock debugfs cannot report the flag by name. Add the missing RSV_HUGETLB entry. Fixes: 7d163a75f821 ("memblock: make HugeTLB bootmem allocation work with KHO") Signed-off-by: Meijing Zhao Signed-off-by: Wandun Chen Link: https://lore.kernel.org/lkml/20260821020910.3428585-2-zhaomeijing100@gmail.com/ Link: https://patch.msgid.link/20260915124641.2498280-1-chenwandun1@gmail.com Signed-off-by: Mike Rapoport (Microsoft) --- mm/memblock.c | 1 + 1 file changed, 1 insertion(+) diff --git a/mm/memblock.c b/mm/memblock.c index 021db49eb7fc..e6fdf95a0534 100644 --- a/mm/memblock.c +++ b/mm/memblock.c @@ -2886,6 +2886,7 @@ static const char * const flagname[] = { [ilog2(MEMBLOCK_RSRV_NOINIT)] = "RSV_NIT", [ilog2(MEMBLOCK_RSRV_KERN)] = "RSV_KERN", [ilog2(MEMBLOCK_KHO_SCRATCH)] = "KHO_SCRATCH", + [ilog2(MEMBLOCK_RSRV_HUGETLB)] = "RSV_HUGETLB", }; static int memblock_debug_show(struct seq_file *m, void *private) From aaad136d56d91252517272b68cd533e5714698d5 Mon Sep 17 00:00:00 2001 From: Myeonghun Pak Date: Sat, 1 Aug 2026 01:35:50 +0900 Subject: [PATCH 063/694] RISC-V: KVM: Synchronize hrtimer callback during teardown The non-Sstc hrtimer callback clears next_set before its final uses of the enclosing vCPU. If teardown observes next_set as false while the callback is still running, kvm_riscv_vcpu_timer_cancel() skips hrtimer_cancel() and kvm_destroy_vcpus() can free the vCPU before the callback enters kvm_riscv_vcpu_set_interrupt(). A guest can arm the timer with SBI TIME and request shutdown with SBI legacy shutdown or SRST. A VMM that honors KVM_EXIT_SYSTEM_EVENT and destroys the VM supplies the teardown side of the race; no post-launch host ioctl is needed to arm or request teardown. On upstream master 62cc90241548, generic KASAN reported: BUG: KASAN: slab-use-after-free in do_raw_spin_lock Write of size 4 at addr ff60000005e58898 kvm_riscv_vcpu_set_interrupt kvm_riscv_vcpu_hrtimer_expired __hrtimer_run_queues hrtimer_interrupt The object was allocated by KVM_CREATE_VCPU and freed concurrently by: kvm_destroy_vcpus kvm_arch_destroy_vm kvm_destroy_vm __fput For deterministic validation, I added mdelay(1000) immediately after the existing next_set = false assignment. This only widens the existing post-clear callback window. A no-delay trace build naturally reached the callback-after-teardown-start/before-deinit ordering in 12 of 200 runs, but 1,500 stock-kernel stress iterations did not produce a KASAN report, so natural reproduction is timing-sensitive. Always invoke hrtimer_cancel() for an initialized timer. Preserve the existing -EINVAL result when the timer is no longer set, but only after synchronizing with a running callback. With this patch, hrtimer_cancel() blocked for the full widened callback window before vCPU destruction. KASAN reported no error in 100 fixed-and-widened runs or 200 fix-only timing-sweep runs. Fixes: 3a9f66cb25e1 ("RISC-V: KVM: Add timer functionality") Cc: stable@vger.kernel.org Assisted-by: OpenAI:GPT-5.6 Signed-off-by: Myeonghun Pak Reviewed-by: Anup Patel Link: https://lore.kernel.org/r/20260731163550.46991-1-mhun512@gmail.com Signed-off-by: Anup Patel --- arch/riscv/kvm/vcpu_timer.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/arch/riscv/kvm/vcpu_timer.c b/arch/riscv/kvm/vcpu_timer.c index ae53133c7ab0..a2cd277a4059 100644 --- a/arch/riscv/kvm/vcpu_timer.c +++ b/arch/riscv/kvm/vcpu_timer.c @@ -61,10 +61,13 @@ static enum hrtimer_restart kvm_riscv_vcpu_hrtimer_expired(struct hrtimer *h) static int kvm_riscv_vcpu_timer_cancel(struct kvm_vcpu_timer *t) { - if (!t->init_done || !t->next_set) + if (!t->init_done) return -EINVAL; hrtimer_cancel(&t->hrt); + + if (!t->next_set) + return -EINVAL; t->next_set = false; return 0; From 52c6b7d20d3e791a9e75aa2be2a467990154c7cd Mon Sep 17 00:00:00 2001 From: Yicong Yang Date: Tue, 4 Aug 2026 21:40:18 +0800 Subject: [PATCH 064/694] RISC-V: KVM: Fix the conversion between vsip and hvip Per AIA spec 1.0 Section 6.3.2, the interrupt numbers 13-63 shares same bit position between related VS shadow CSRs and hypervisor CSRs. So there's a shift only for SSI, STI and SEI interrupt. Currently the KVM always does a shift for all the interrupts (include LCOFI with number 13) when doing the conversion between vsip and hvip. Fix this by only doing shift the SSI, STI and SEI. Add wrappers for doing the conversion between vsip and hvip. Fixes: 16b0bde9a37c ("RISC-V: KVM: Add perf sampling support for guests") Signed-off-by: Yicong Yang Reviewed-by: Anup Patel Link: https://lore.kernel.org/r/20260804134018.85497-1-yang.yicong@picoheart.com Signed-off-by: Anup Patel --- arch/riscv/include/asm/csr.h | 20 ++++++++++++++++---- arch/riscv/kvm/vcpu.c | 3 +-- arch/riscv/kvm/vcpu_onereg.c | 8 +++----- tools/arch/riscv/include/asm/csr.h | 20 ++++++++++++++++---- 4 files changed, 36 insertions(+), 15 deletions(-) diff --git a/arch/riscv/include/asm/csr.h b/arch/riscv/include/asm/csr.h index 6c823361be86..621f84d1898f 100644 --- a/arch/riscv/include/asm/csr.h +++ b/arch/riscv/include/asm/csr.h @@ -188,12 +188,24 @@ #define HGATP_MODE_SHIFT HGATP32_MODE_SHIFT #endif -/* VSIP & HVIP relation */ +/* + * VSIP & HVIP relation + * + * The bit positions are same between VSIP and HVIP for interrupt + * numbers 13-63, where there's a shift for the SSI, STI and SEI. + */ #define VSIP_TO_HVIP_SHIFT (IRQ_VS_SOFT - IRQ_S_SOFT) -#define VSIP_VALID_MASK ((_AC(1, UL) << IRQ_S_SOFT) | \ +#define VSIP_BIAS_MASK ((_AC(1, UL) << IRQ_S_SOFT) | \ (_AC(1, UL) << IRQ_S_TIMER) | \ - (_AC(1, UL) << IRQ_S_EXT) | \ - (_AC(1, UL) << IRQ_PMU_OVF)) + (_AC(1, UL) << IRQ_S_EXT)) +#define VSIP_NO_BIAS_MASK (_AC(1, UL) << IRQ_PMU_OVF) +#define VSIP_VALID_MASK (VSIP_BIAS_MASK | VSIP_NO_BIAS_MASK) +#define vsip_to_hvip(_vsip) ((((_vsip) & VSIP_BIAS_MASK) << \ + VSIP_TO_HVIP_SHIFT) | \ + ((_vsip) & VSIP_NO_BIAS_MASK)) +#define hvip_to_vsip(_hvip) ((((_hvip) >> VSIP_TO_HVIP_SHIFT) & \ + VSIP_BIAS_MASK) | \ + ((_hvip) & VSIP_NO_BIAS_MASK)) /* AIA CSR bits */ #define TOPI_IID_SHIFT 16 diff --git a/arch/riscv/kvm/vcpu.c b/arch/riscv/kvm/vcpu.c index e062ca19f9d8..7f1636c44d41 100644 --- a/arch/riscv/kvm/vcpu.c +++ b/arch/riscv/kvm/vcpu.c @@ -491,8 +491,7 @@ bool kvm_riscv_vcpu_has_interrupts(struct kvm_vcpu *vcpu, u64 mask) bool ret; raw_spin_lock_irqsave(&vcpu->arch.irqs_pending_lock, flags); - ie = ((vcpu->arch.guest_csr.vsie & VSIP_VALID_MASK) - << VSIP_TO_HVIP_SHIFT) & (unsigned long)mask; + ie = vsip_to_hvip(vcpu->arch.guest_csr.vsie) & (unsigned long)mask; ie |= vcpu->arch.guest_csr.vsie & ~IRQ_LOCAL_MASK & (unsigned long)mask; ret = vcpu->arch.irqs_pending[0] & ie; diff --git a/arch/riscv/kvm/vcpu_onereg.c b/arch/riscv/kvm/vcpu_onereg.c index 841f2cf87168..f16e25098af1 100644 --- a/arch/riscv/kvm/vcpu_onereg.c +++ b/arch/riscv/kvm/vcpu_onereg.c @@ -272,7 +272,7 @@ static int kvm_riscv_vcpu_general_get_csr(struct kvm_vcpu *vcpu, if (reg_num == KVM_REG_RISCV_CSR_REG(sip)) { kvm_riscv_vcpu_flush_interrupts(vcpu); - *out_val = (csr->hvip >> VSIP_TO_HVIP_SHIFT) & VSIP_VALID_MASK; + *out_val = hvip_to_vsip(csr->hvip); *out_val |= csr->hvip & ~IRQ_LOCAL_MASK; } else *out_val = ((unsigned long *)csr)[reg_num]; @@ -293,10 +293,8 @@ static int kvm_riscv_vcpu_general_set_csr(struct kvm_vcpu *vcpu, reg_num = array_index_nospec(reg_num, regs_max); - if (reg_num == KVM_REG_RISCV_CSR_REG(sip)) { - reg_val &= VSIP_VALID_MASK; - reg_val <<= VSIP_TO_HVIP_SHIFT; - } + if (reg_num == KVM_REG_RISCV_CSR_REG(sip)) + reg_val = vsip_to_hvip(reg_val); ((unsigned long *)csr)[reg_num] = reg_val; diff --git a/tools/arch/riscv/include/asm/csr.h b/tools/arch/riscv/include/asm/csr.h index 21d8cee04638..8df64314d613 100644 --- a/tools/arch/riscv/include/asm/csr.h +++ b/tools/arch/riscv/include/asm/csr.h @@ -163,12 +163,24 @@ #define HGATP_MODE_SHIFT HGATP32_MODE_SHIFT #endif -/* VSIP & HVIP relation */ +/* + * VSIP & HVIP relation + * + * The bit positions are same between VSIP and HVIP for interrupt + * numbers 13-63, where there's a shift for the SSI, STI and SEI. + */ #define VSIP_TO_HVIP_SHIFT (IRQ_VS_SOFT - IRQ_S_SOFT) -#define VSIP_VALID_MASK ((_AC(1, UL) << IRQ_S_SOFT) | \ +#define VSIP_BIAS_MASK ((_AC(1, UL) << IRQ_S_SOFT) | \ (_AC(1, UL) << IRQ_S_TIMER) | \ - (_AC(1, UL) << IRQ_S_EXT) | \ - (_AC(1, UL) << IRQ_PMU_OVF)) + (_AC(1, UL) << IRQ_S_EXT)) +#define VSIP_NO_BIAS_MASK (_AC(1, UL) << IRQ_PMU_OVF) +#define VSIP_VALID_MASK (VSIP_BIAS_MASK | VSIP_NO_BIAS_MASK) +#define vsip_to_hvip(_vsip) ((((_vsip) & VSIP_BIAS_MASK) << \ + VSIP_TO_HVIP_SHIFT) | \ + ((_vsip) & VSIP_NO_BIAS_MASK)) +#define hvip_to_vsip(_hvip) ((((_hvip) >> VSIP_TO_HVIP_SHIFT) & \ + VSIP_BIAS_MASK) | \ + ((_hvip) & VSIP_NO_BIAS_MASK)) /* AIA CSR bits */ #define TOPI_IID_SHIFT 16 From 8ae12ccaec6ec74945d8c1ef39f2c1b8df779abc Mon Sep 17 00:00:00 2001 From: Xie Bo Date: Mon, 10 Aug 2026 09:21:15 +0800 Subject: [PATCH 065/694] RISC-V: KVM: Serialize IMSIC attributes with vCPU migration KVM device ioctls are not serialized against KVM_RUN. As a result, kvm_riscv_aia_imsic_rw_attr() can snapshot the physical CPU and HGEI of an IMSIC VS-file before a concurrent vCPU migration releases it. The HGEI can then be allocated to another vCPU before imsic_vsfile_rw() uses the stale tuple. A GET or SET attribute may consequently access the new owner's interrupt file. Serialize the entire IMSIC attribute operation with the target vCPU mutex. This prevents the VS-file from being migrated and recycled until the attribute access completes. Acquire the mutex killably so that the device ioctl remains interruptible while waiting for KVM_RUN to finish. Fixes: db8b7e97d613 ("RISC-V: KVM: Add in-kernel virtualization of AIA IMSIC") Cc: stable@vger.kernel.org Signed-off-by: Xie Bo Reviewed-by: Anup Patel Link: https://lore.kernel.org/r/20260810-imsic-attr-race-v2-1-00ed95ad321e@ultrarisc.com Signed-off-by: Anup Patel --- arch/riscv/kvm/aia_imsic.c | 11 +++++++++-- 1 file changed, 9 insertions(+), 2 deletions(-) diff --git a/arch/riscv/kvm/aia_imsic.c b/arch/riscv/kvm/aia_imsic.c index c1af23e79ae0..f17dc559a709 100644 --- a/arch/riscv/kvm/aia_imsic.c +++ b/arch/riscv/kvm/aia_imsic.c @@ -965,9 +965,14 @@ int kvm_riscv_aia_imsic_rw_attr(struct kvm *kvm, unsigned long type, if (!vcpu) return -ENODEV; + if (mutex_lock_killable(&vcpu->mutex)) + return -EINTR; + imsic = vcpu->arch.aia_context.imsic_state; - if (!imsic) - return -ENODEV; + if (!imsic) { + rc = -ENODEV; + goto out_unlock; + } isel = KVM_DEV_RISCV_AIA_IMSIC_GET_ISEL(type); read_lock_irqsave(&imsic->vsfile_lock, flags); @@ -991,6 +996,8 @@ int kvm_riscv_aia_imsic_rw_attr(struct kvm *kvm, unsigned long type, rc = imsic_vsfile_rw(vsfile_hgei, vsfile_cpu, imsic->nr_eix, isel, write, val); +out_unlock: + mutex_unlock(&vcpu->mutex); return rc; } From ed54fdb460a6e81d5f8f38388d1f10b385dd4a58 Mon Sep 17 00:00:00 2001 From: Xie Bo Date: Mon, 10 Aug 2026 13:15:43 +0800 Subject: [PATCH 066/694] RISC-V: KVM: Release unused page after MMU invalidation If an MMU invalidation races with a G-stage fault, the fault handler skips installing the page but leaves ret set to zero. As a result, kvm_release_faultin_page() treats the page as used and can unnecessarily mark it dirty. Track the invalidation retry separately and release the page as unused, while preserving the existing return value so that the vCPU retries the fault. Fixes: 2ed90cb0938a ("KVM: RISC-V: Retry fault if vma_lookup() results become invalid") Cc: stable@vger.kernel.org Signed-off-by: Xie Bo Reviewed-by: Anup Patel Link: https://lore.kernel.org/r/20260810051544.3953925-2-xb@ultrarisc.com Signed-off-by: Anup Patel --- arch/riscv/kvm/mmu.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/arch/riscv/kvm/mmu.c b/arch/riscv/kvm/mmu.c index 6035b5ec9503..d189fd58d7bf 100644 --- a/arch/riscv/kvm/mmu.c +++ b/arch/riscv/kvm/mmu.c @@ -627,6 +627,7 @@ int kvm_riscv_mmu_map(struct kvm_vcpu *vcpu, struct kvm_memory_slot *memslot, int ret; kvm_pfn_t hfn; bool is_hugetlb; + bool unused = false; bool writable; unsigned int vma_pageshift; gfn_t gfn = gpa >> PAGE_SHIFT; @@ -719,8 +720,10 @@ int kvm_riscv_mmu_map(struct kvm_vcpu *vcpu, struct kvm_memory_slot *memslot, write_lock(&kvm->mmu_lock); - if (mmu_invalidate_retry(kvm, mmu_seq)) + if (mmu_invalidate_retry(kvm, mmu_seq)) { + unused = true; goto out_unlock; + } /* * Check if we are backed by a THP and thus use block mapping if @@ -743,7 +746,8 @@ int kvm_riscv_mmu_map(struct kvm_vcpu *vcpu, struct kvm_memory_slot *memslot, kvm_err("Failed to map in G-stage\n"); out_unlock: - kvm_release_faultin_page(kvm, page, ret && ret != -EEXIST, writable); + kvm_release_faultin_page(kvm, page, + unused || (ret && ret != -EEXIST), writable); write_unlock(&kvm->mmu_lock); return ret; } From f41fb17143df890855d3de980f8eb92dfd595817 Mon Sep 17 00:00:00 2001 From: Xie Bo Date: Mon, 10 Aug 2026 13:15:44 +0800 Subject: [PATCH 067/694] RISC-V: KVM: Propagate interrupted G-stage faults __kvm_faultin_pfn() reports an interrupted host page fault with KVM_PFN_ERR_SIGPENDING. RISC-V currently handles it as a generic error PFN and returns -EFAULT. Return -EINTR for the signal-pending sentinel so callers can distinguish an interrupted fault from an invalid userspace mapping. Do not log the expected interruption as a vCPU exit error. Fixes: 9d05c1fee837 ("RISC-V: KVM: Implement stage2 page table programming") Cc: stable@vger.kernel.org Signed-off-by: Xie Bo Reviewed-by: Anup Patel Link: https://lore.kernel.org/r/20260810051544.3953925-3-xb@ultrarisc.com Signed-off-by: Anup Patel --- arch/riscv/kvm/mmu.c | 2 ++ arch/riscv/kvm/vcpu_exit.c | 2 +- 2 files changed, 3 insertions(+), 1 deletion(-) diff --git a/arch/riscv/kvm/mmu.c b/arch/riscv/kvm/mmu.c index d189fd58d7bf..3e955d808743 100644 --- a/arch/riscv/kvm/mmu.c +++ b/arch/riscv/kvm/mmu.c @@ -708,6 +708,8 @@ int kvm_riscv_mmu_map(struct kvm_vcpu *vcpu, struct kvm_memory_slot *memslot, vma_pageshift, current); return 0; } + if (is_sigpending_pfn(hfn)) + return -EINTR; if (is_error_noslot_pfn(hfn)) return -EFAULT; diff --git a/arch/riscv/kvm/vcpu_exit.c b/arch/riscv/kvm/vcpu_exit.c index 88e0c369b354..8d36eb8abce1 100644 --- a/arch/riscv/kvm/vcpu_exit.c +++ b/arch/riscv/kvm/vcpu_exit.c @@ -283,7 +283,7 @@ int kvm_riscv_vcpu_exit(struct kvm_vcpu *vcpu, struct kvm_run *run, } /* Print details in-case of error */ - if (ret < 0) { + if (ret < 0 && ret != -EINTR) { kvm_err("VCPU exit error %d\n", ret); kvm_err("SEPC=0x%lx SSTATUS=0x%lx HSTATUS=0x%lx\n", vcpu->arch.guest_context.sepc, From b3d346838ec65fac7fd83f5dbcedd13cadfffddb Mon Sep 17 00:00:00 2001 From: Zongmin Zhou Date: Wed, 26 Aug 2026 15:50:09 +0800 Subject: [PATCH 068/694] KVM: riscv: Fix NACL hfence entry update order The SBI v3.0 specification (section 15.1.2) requires a nested HFENCE entry to be populated as follows: 1) find an unused entry with Config.Pending == 0 2) update the Page_Number and Page_Count words 3) update the Config word with Config.Pending set __kvm_riscv_nacl_hfence() writes the Config word first, so the SBI implementation (or NACL hardware) can observe a pending entry with pnum/pcount values left over from the previous use of that entry, resulting in incorrect TLB flush ranges. Write pnum and pcount first and the Config word last. Since the consumer is an external agent on coherent shared memory, use WRITE_ONCE() to stop the compiler from reordering the stores and smp_wmb() to make the parameter words globally visible before the Pending bit is set. Fixes: d466c19cead5 ("RISC-V: KVM: Add common nested acceleration support") Signed-off-by: Zongmin Zhou Reviewed-by: Anup Patel Link: https://lore.kernel.org/r/20260826075009.68952-1-min_halo@163.com Signed-off-by: Anup Patel --- arch/riscv/kvm/nacl.c | 20 ++++++++++++++++---- 1 file changed, 16 insertions(+), 4 deletions(-) diff --git a/arch/riscv/kvm/nacl.c b/arch/riscv/kvm/nacl.c index 9aff03c4f667..a5cda9a65156 100644 --- a/arch/riscv/kvm/nacl.c +++ b/arch/riscv/kvm/nacl.c @@ -42,12 +42,24 @@ void __kvm_riscv_nacl_hfence(void *shmem, } } - entp = shmem + SBI_NACL_SHMEM_HFENCE_ENTRY_CONFIG(i); - *entp = cpu_to_lelong(control); + /* + * Per SBI v3.0 section 15.1.2, the Page_Number and Page_Count + * words must be updated before the Config word with its Pending + * bit set. WRITE_ONCE() stops the compiler from reordering the + * stores and smp_wmb() makes the parameter words globally + * visible to the SBI implementation (or NACL hardware) before + * the Pending bit is set. + */ entp = shmem + SBI_NACL_SHMEM_HFENCE_ENTRY_PNUM(i); - *entp = cpu_to_lelong(page_num); + WRITE_ONCE(*entp, cpu_to_lelong(page_num)); entp = shmem + SBI_NACL_SHMEM_HFENCE_ENTRY_PCOUNT(i); - *entp = cpu_to_lelong(page_count); + WRITE_ONCE(*entp, cpu_to_lelong(page_count)); + + /* Ensure the parameter words are visible before the Pending bit */ + smp_wmb(); + + entp = shmem + SBI_NACL_SHMEM_HFENCE_ENTRY_CONFIG(i); + WRITE_ONCE(*entp, cpu_to_lelong(control)); } int kvm_riscv_nacl_enable(void) From b7749531a9b195f4fd7db92ca3cc49bda1a4dc8d Mon Sep 17 00:00:00 2001 From: Zongmin Zhou Date: Wed, 26 Aug 2026 14:41:17 +0800 Subject: [PATCH 069/694] RISC-V: KVM: Fix sdata leak and stale snapshot_addr in snapshot_set_shmem A guest may call SBI_PMU_SNAPSHOT_SET_SHMEM repeatedly. Each call overwrites kvpmu->sdata without freeing the old buffer (memory leak), and if a later kvm_vcpu_write_guest() fails, the error path frees sdata but leaves snapshot_addr stale. A subsequent SBI_PMU_COUNTER_START then passes the INVALID_GPA check and crashes the host with a NULL buffer in kvm_vcpu_read_guest(). Fix this by clearing the previously installed snapshot area before installing a new one, which keeps sdata and snapshot_addr consistent. The SBI spec suggests a single invocation but defines no error code for repeated calls, so KVM must tolerate them. Fixes: c2f41ddbcdd756 ("RISC-V: KVM: Implement SBI PMU Snapshot feature") Cc: stable@vger.kernel.org Signed-off-by: Zongmin Zhou Reviewed-by: Anup Patel Link: https://lore.kernel.org/r/20260826064117.58029-1-min_halo@163.com Signed-off-by: Anup Patel --- arch/riscv/kvm/vcpu_pmu.c | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/arch/riscv/kvm/vcpu_pmu.c b/arch/riscv/kvm/vcpu_pmu.c index 6ff741ee7803..cc0c138651bc 100644 --- a/arch/riscv/kvm/vcpu_pmu.c +++ b/arch/riscv/kvm/vcpu_pmu.c @@ -454,6 +454,14 @@ int kvm_riscv_vcpu_pmu_snapshot_set_shmem(struct kvm_vcpu *vcpu, unsigned long s } } + /* + * Clear any previously installed snapshot area to avoid leaking + * the old sdata and to keep sdata/snapshot_addr consistent if + * the re-install fails below. + */ + if (kvpmu->snapshot_addr != INVALID_GPA) + kvm_pmu_clear_snapshot_area(vcpu); + kvpmu->sdata = kzalloc(snapshot_area_size, GFP_ATOMIC | __GFP_ACCOUNT); if (!kvpmu->sdata) { sbiret = SBI_ERR_FAILURE; From 8b3fd1a8b305321171602bfa7c41212441cf69e4 Mon Sep 17 00:00:00 2001 From: SeungJu Cheon Date: Tue, 25 Aug 2026 17:37:17 +0900 Subject: [PATCH 070/694] RISC-V: KVM: Preserve firmware counter value across stop/start Firmware events accumulate in kvpmu->fw_event[].value while running, but counter stop only clears fw_event[].started without saving the value back to pmc->counter_val. A subsequent counter start without SBI_PMU_START_FLAG_SET_INIT_VALUE reloads the stale counter_val into fw_event[].value, losing all events counted so far. Save fw_event[].value into counter_val when actually stopping a running counter, and remove the now redundant synchronization from the snapshot path. Fixes: badc386869e2c ("RISC-V: KVM: Support firmware events") Signed-off-by: SeungJu Cheon Reviewed-by: Anup Patel Link: https://lore.kernel.org/r/20260825083719.643970-2-suunj1331@gmail.com Signed-off-by: Anup Patel --- arch/riscv/kvm/vcpu_pmu.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/arch/riscv/kvm/vcpu_pmu.c b/arch/riscv/kvm/vcpu_pmu.c index cc0c138651bc..5dfe2054d12b 100644 --- a/arch/riscv/kvm/vcpu_pmu.c +++ b/arch/riscv/kvm/vcpu_pmu.c @@ -683,10 +683,12 @@ int kvm_riscv_vcpu_pmu_ctr_stop(struct kvm_vcpu *vcpu, unsigned long ctr_base, goto out; } - if (!kvpmu->fw_event[fevent_code].started) + if (!kvpmu->fw_event[fevent_code].started) { sbiret = SBI_ERR_ALREADY_STOPPED; - - kvpmu->fw_event[fevent_code].started = false; + } else { + kvpmu->fw_event[fevent_code].started = false; + pmc->counter_val = kvpmu->fw_event[fevent_code].value; + } } else if (pmc->perf_event) { if (pmc->started) { /* Stop counting the counter */ @@ -704,9 +706,7 @@ int kvm_riscv_vcpu_pmu_ctr_stop(struct kvm_vcpu *vcpu, unsigned long ctr_base, } if (snap_flag_set && !sbiret) { - if (pmc->cinfo.type == SBI_PMU_CTR_TYPE_FW) - pmc->counter_val = kvpmu->fw_event[fevent_code].value; - else if (pmc->perf_event) + if (pmc->perf_event) pmc->counter_val += perf_event_read_value(pmc->perf_event, &enabled, &running); /* From 057dd2639ceae79adced5d8fe52c32d562edcb3a Mon Sep 17 00:00:00 2001 From: SeungJu Cheon Date: Tue, 25 Aug 2026 17:37:18 +0900 Subject: [PATCH 071/694] RISC-V: KVM: Report snapshot write failure to the guest If kvm_vcpu_write_guest() fails while updating the PMU snapshot area on counter stop, the guest may receive SBI_SUCCESS without the snapshot being updated, leaving stale data in shared memory. Return SBI_ERR_FAILURE when the snapshot write fails. Fixes: c2f41ddbcdd7 ("RISC-V: KVM: Implement SBI PMU Snapshot feature") Signed-off-by: SeungJu Cheon Reviewed-by: Anup Patel Link: https://lore.kernel.org/r/20260825083719.643970-3-suunj1331@gmail.com Signed-off-by: Anup Patel --- arch/riscv/kvm/vcpu_pmu.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/arch/riscv/kvm/vcpu_pmu.c b/arch/riscv/kvm/vcpu_pmu.c index 5dfe2054d12b..8be87f879473 100644 --- a/arch/riscv/kvm/vcpu_pmu.c +++ b/arch/riscv/kvm/vcpu_pmu.c @@ -735,9 +735,10 @@ int kvm_riscv_vcpu_pmu_ctr_stop(struct kvm_vcpu *vcpu, unsigned long ctr_base, } } - if (shmem_needs_update) - kvm_vcpu_write_guest(vcpu, kvpmu->snapshot_addr, kvpmu->sdata, - sizeof(struct riscv_pmu_snapshot_data)); + if (shmem_needs_update && + kvm_vcpu_write_guest(vcpu, kvpmu->snapshot_addr, kvpmu->sdata, + sizeof(struct riscv_pmu_snapshot_data))) + sbiret = SBI_ERR_FAILURE; out: retdata->err_val = sbiret; From c7e2cc38c56142cdab25e6f73602a8222bf9479b Mon Sep 17 00:00:00 2001 From: SeungJu Cheon Date: Tue, 25 Aug 2026 17:37:19 +0900 Subject: [PATCH 072/694] RISC-V: KVM: Fix perf-backed counter accounting across stop and read pmu_ctr_read() adds the event count returned by perf_event_read_value() to counter_val, which can accumulate the same count repeatedly across reads. kvm_riscv_vcpu_pmu_ctr_stop() also leaves counter_val stale by not folding the current event count into it. Make reads of perf-backed counters side-effect free, and use perf_event_pause() when stopping a counter to fold the current event count into counter_val while resetting it. This preserves the counter value across stop/start and lets the snapshot path use counter_val directly. Fixes: 0cb74b65d2e5 ("RISC-V: KVM: Implement perf support without sampling") Signed-off-by: SeungJu Cheon Reviewed-by: Anup Patel Link: https://lore.kernel.org/r/20260825083719.643970-4-suunj1331@gmail.com Signed-off-by: Anup Patel --- arch/riscv/kvm/vcpu_pmu.c | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-) diff --git a/arch/riscv/kvm/vcpu_pmu.c b/arch/riscv/kvm/vcpu_pmu.c index 8be87f879473..e095d1ff439f 100644 --- a/arch/riscv/kvm/vcpu_pmu.c +++ b/arch/riscv/kvm/vcpu_pmu.c @@ -270,12 +270,13 @@ static int pmu_ctr_read(struct kvm_vcpu *vcpu, unsigned long cidx, return -EINVAL; pmc->counter_val = kvpmu->fw_event[fevent_code].value; + *out_val = pmc->counter_val; } else if (pmc->perf_event) { - pmc->counter_val += perf_event_read_value(pmc->perf_event, &enabled, &running); + *out_val = pmc->counter_val + + perf_event_read_value(pmc->perf_event, &enabled, &running); } else { return -EINVAL; } - *out_val = pmc->counter_val; return 0; } @@ -653,7 +654,6 @@ int kvm_riscv_vcpu_pmu_ctr_stop(struct kvm_vcpu *vcpu, unsigned long ctr_base, { struct kvm_pmu *kvpmu = vcpu_to_pmu(vcpu); int i, pmc_index, sbiret = 0; - u64 enabled, running; struct kvm_pmc *pmc; int fevent_code; bool snap_flag_set = flags & SBI_PMU_STOP_FLAG_TAKE_SNAPSHOT; @@ -691,8 +691,11 @@ int kvm_riscv_vcpu_pmu_ctr_stop(struct kvm_vcpu *vcpu, unsigned long ctr_base, } } else if (pmc->perf_event) { if (pmc->started) { - /* Stop counting the counter */ - perf_event_disable(pmc->perf_event); + /* + * Stop the counter and fold the live count into counter_val. + * Reset the event value to avoid redundant accumulation. + */ + pmc->counter_val += perf_event_pause(pmc->perf_event, true); pmc->started = false; } else { sbiret = SBI_ERR_ALREADY_STOPPED; @@ -706,9 +709,6 @@ int kvm_riscv_vcpu_pmu_ctr_stop(struct kvm_vcpu *vcpu, unsigned long ctr_base, } if (snap_flag_set && !sbiret) { - if (pmc->perf_event) - pmc->counter_val += perf_event_read_value(pmc->perf_event, - &enabled, &running); /* * The counter and overflow indices in the snapshot region are w.r.to * cbase. Modify the set bit in the counter mask instead of the pmc_index From 41e81f7e3ef96594fb840445343c0ee7723aa550 Mon Sep 17 00:00:00 2001 From: Tan Chi Date: Mon, 14 Sep 2026 11:11:46 +0800 Subject: [PATCH 073/694] RISC-V: KVM: Fix HSM hart status error propagation kvm_sbi_hsm_vcpu_get_status() returns SBI_ERR_INVALID_PARAM when the requested hart does not exist. However, the HART_STATUS case returns from the SBI handler without storing this error in retdata->err_val. As a result, a guest querying the status of a non-existent hart observes SBI_SUCCESS instead of SBI_ERR_INVALID_PARAM. Use the common SBI error handling path for HART_STATUS after saving a valid hart state in retdata->out_val. This preserves the returned error when kvm_sbi_hsm_vcpu_get_status() fails. Fixes: bae0dfd74e01 ("RISC-V: KVM: Modify SBI extension handler to return SBI error code") Cc: stable@vger.kernel.org Assisted-by: LLM Signed-off-by: Tan Chi Reviewed-by: Anup Patel Link: https://lore.kernel.org/r/20260914031146.446157-1-tanchi25@mails.ucas.ac.cn Signed-off-by: Anup Patel --- arch/riscv/kvm/vcpu_sbi_hsm.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/riscv/kvm/vcpu_sbi_hsm.c b/arch/riscv/kvm/vcpu_sbi_hsm.c index f26207f84bab..06a15629c26b 100644 --- a/arch/riscv/kvm/vcpu_sbi_hsm.c +++ b/arch/riscv/kvm/vcpu_sbi_hsm.c @@ -95,9 +95,9 @@ static int kvm_sbi_ext_hsm_handler(struct kvm_vcpu *vcpu, struct kvm_run *run, ret = kvm_sbi_hsm_vcpu_get_status(vcpu); if (ret >= 0) { retdata->out_val = ret; - retdata->err_val = 0; + ret = 0; } - return 0; + break; case SBI_EXT_HSM_HART_SUSPEND: switch (lower_32_bits(cp->a0)) { case SBI_HSM_SUSPEND_RET_DEFAULT: From 8cd92f77ae4f5371a7d581f8324c24919670b304 Mon Sep 17 00:00:00 2001 From: Fuad Tabba Date: Fri, 21 Aug 2026 07:44:42 +0100 Subject: [PATCH 074/694] KVM: arm64: vgic-its: Free the caches when GITS_BASER changes A guest that disables the ITS and re-points or shrinks GITS_BASER with VALID still set keeps the devices and collections it mapped against the old table, as KVM frees them only when VALID is cleared. The contents of the table are IMPLEMENTATION DEFINED, so a write that gives GITS_BASER a different address or size may lose whatever the old value described. Free the list whenever the stored value changes, and drop the translation cache with it. The cache is not empty just because the ITS is disabled: its->enabled is written under the cmd_lock, while vgic_its_resolve_lpi() tests it under the its_lock, so an injection can still cache an entry after the ITS was disabled. Hence the invalidation inside the its_lock section. Test for a change rather than a write: its_restore_enable() rewrites GITS_BASER from its probe-time cache on resume, and KVM reports GITS_TYPER.HCC as 0, so nothing re-maps the boot CPU's collection afterwards. Fixes: 36d6961c2b481 ("KVM: arm/arm64: vgic-its: Free caches when GITS_BASER Valid bit is cleared") Suggested-by: Marc Zyngier Link: https://lore.kernel.org/all/87ecg9owwa.wl-maz@kernel.org/ Signed-off-by: Fuad Tabba Reviewed-by: Marc Zyngier Link: https://patch.msgid.link/20260821064445.615838-2-fuad.tabba@linux.dev Signed-off-by: Oliver Upton --- arch/arm64/kvm/vgic/vgic-its.c | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-) diff --git a/arch/arm64/kvm/vgic/vgic-its.c b/arch/arm64/kvm/vgic/vgic-its.c index 9e782a4fea7e..ab89b0138efd 100644 --- a/arch/arm64/kvm/vgic/vgic-its.c +++ b/arch/arm64/kvm/vgic/vgic-its.c @@ -1658,7 +1658,7 @@ static void vgic_mmio_write_its_baser(struct kvm *kvm, unsigned long val) { const struct vgic_its_abi *abi = vgic_its_get_abi(its); - u64 entry_size, table_type; + u64 old, entry_size, table_type; u64 reg, *regptr, clearbits = 0; /* When GITS_CTLR.Enable is 1, we ignore write accesses. */ @@ -1681,7 +1681,9 @@ static void vgic_mmio_write_its_baser(struct kvm *kvm, return; } - reg = update_64bit_reg(*regptr, addr & 7, len, val); + old = *regptr; + + reg = update_64bit_reg(old, addr & 7, len, val); reg &= ~GITS_BASER_RO_MASK; reg &= ~clearbits; @@ -1691,7 +1693,8 @@ static void vgic_mmio_write_its_baser(struct kvm *kvm, *regptr = reg; - if (!(reg & GITS_BASER_VALID)) { + /* The ITS driver rewrites an unchanged GITS_BASER on resume. */ + if (reg != old) { /* Take the its_lock to prevent a race with a save/restore */ mutex_lock(&its->its_lock); switch (table_type) { @@ -1702,6 +1705,8 @@ static void vgic_mmio_write_its_baser(struct kvm *kvm, vgic_its_free_collection_list(kvm, its); break; } + /* A concurrent injection may have cached a translation. */ + vgic_its_invalidate_cache(its); mutex_unlock(&its->its_lock); } } From 30908e7272479b6453745022abb9f2eb9c9933f7 Mon Sep 17 00:00:00 2001 From: Fuad Tabba Date: Fri, 21 Aug 2026 07:44:43 +0100 Subject: [PATCH 075/694] Revert "KVM: arm64: vgic-its: Don't save collections the table cannot hold" This reverts commit 9b10fb74e4b661543d188701bd4d024fc5c18f58. Freeing the collections when GITS_BASER changes removes the state this check rejected: vgic_its_cmd_handle_mapi(), vgic_its_cmd_handle_mapc() and vgic_its_restore_cte() all validate the ID against the current table before allocating, and the table can no longer change under the list. What remains is a collection whose entry is not backed by a memslot, which the write fails on anyway, so the check costs a save userspace should be able to issue reliably and buys nothing. The reverted commit credited the check with bounding the walk as well. It stays bounded without it: collection IDs are unique and each is below the table's capacity, so the list cannot be longer than the table. Suggested-by: Marc Zyngier Link: https://lore.kernel.org/all/87ecg9owwa.wl-maz@kernel.org/ Signed-off-by: Fuad Tabba Reviewed-by: Marc Zyngier Link: https://patch.msgid.link/20260821064445.615838-3-fuad.tabba@linux.dev Signed-off-by: Oliver Upton --- arch/arm64/kvm/vgic/vgic-its.c | 3 --- 1 file changed, 3 deletions(-) diff --git a/arch/arm64/kvm/vgic/vgic-its.c b/arch/arm64/kvm/vgic/vgic-its.c index ab89b0138efd..313bf9e802bf 100644 --- a/arch/arm64/kvm/vgic/vgic-its.c +++ b/arch/arm64/kvm/vgic/vgic-its.c @@ -2546,9 +2546,6 @@ static int vgic_its_save_collection_table(struct vgic_its *its) max_size = GITS_BASER_NR_PAGES(baser) * SZ_64K; list_for_each_entry(collection, &its->collection_list, coll_list) { - if (!vgic_its_check_id(its, baser, collection->collection_id, NULL)) - return -EINVAL; - ret = vgic_its_save_cte(its, collection, gpa); if (ret) return ret; From cc5d96036e01ac330d24b2f0c336d60f82ab4930 Mon Sep 17 00:00:00 2001 From: Fuad Tabba Date: Fri, 21 Aug 2026 07:44:44 +0100 Subject: [PATCH 076/694] KVM: arm64: vgic-its: Skip unreachable devices instead of failing the save vgic_its_save_device_tables() aborts with -EINVAL when a device's entry falls outside the device table, which a guest can arrange on its own: an indirect table lets it clear an L1 entry's valid bit without touching GITS_BASER. That fails a save userspace should be able to issue reliably. Skip the device instead, and point the saved DTE chain past it, as commit ad1e686e2378d ("KVM: arm64: vgic-its: Point saved ITEs at the next valid entry") does for ITEs. compute_next_devid_offset() takes the next device off the list whether or not it was saved, so the predecessor would otherwise point at an entry the save never wrote. Restore follows that offset while it stays inside the table being scanned: within an L2 block, or anywhere in a flat table. Both need userspace to remove a memslot under the table, since dropping an L1 entry takes the whole block with it and scan_its_table() stops at the block boundary. Fixes: 57a9a117154c9 ("KVM: arm64: vgic-its: Device table save/restore") Suggested-by: Marc Zyngier Link: https://lore.kernel.org/all/86bjaz5s6v.wl-maz@kernel.org/ Signed-off-by: Fuad Tabba Reviewed-by: Marc Zyngier Link: https://patch.msgid.link/20260821064445.615838-4-fuad.tabba@linux.dev Signed-off-by: Oliver Upton --- arch/arm64/kvm/vgic/vgic-its.c | 32 +++++++++++++++++++------------- 1 file changed, 19 insertions(+), 13 deletions(-) diff --git a/arch/arm64/kvm/vgic/vgic-its.c b/arch/arm64/kvm/vgic/vgic-its.c index 313bf9e802bf..0904ae850c35 100644 --- a/arch/arm64/kvm/vgic/vgic-its.c +++ b/arch/arm64/kvm/vgic/vgic-its.c @@ -2024,18 +2024,22 @@ static int vgic_its_attr_regs_access(struct kvm_device *dev, return ret; } -static u32 compute_next_devid_offset(struct list_head *h, +static u32 compute_next_devid_offset(struct vgic_its *its, u64 baser, struct its_device *dev) { - struct its_device *next; - u32 next_offset; + struct its_device *next = dev; - if (list_is_last(&dev->dev_list, h)) - return 0; - next = list_next_entry(dev, dev_list); - next_offset = next->device_id - dev->device_id; + /* + * Point at the next device vgic_its_save_device_tables() saves. It + * sorts device_list first, so the subtraction cannot underflow. + */ + list_for_each_entry_continue(next, &its->device_list, dev_list) { + if (vgic_its_check_id(its, baser, next->device_id, NULL)) + return min_t(u32, next->device_id - dev->device_id, + VITS_DTE_MAX_DEVID_OFFSET); + } - return min_t(u32, next_offset, VITS_DTE_MAX_DEVID_OFFSET); + return 0; } static u32 compute_next_eventid_offset(struct list_head *h, struct its_ite *ite) @@ -2276,17 +2280,18 @@ static int vgic_its_restore_itt(struct vgic_its *its, struct its_device *dev) * vgic_its_save_dte - Save a device table entry at a given GPA * * @its: ITS handle + * @baser: GITS_BASER the caller is saving against * @dev: ITS device * @ptr: GPA */ -static int vgic_its_save_dte(struct vgic_its *its, struct its_device *dev, - gpa_t ptr) +static int vgic_its_save_dte(struct vgic_its *its, u64 baser, + struct its_device *dev, gpa_t ptr) { u64 val, itt_addr_field; u32 next_offset; itt_addr_field = dev->itt_addr >> 8; - next_offset = compute_next_devid_offset(&its->device_list, dev); + next_offset = compute_next_devid_offset(its, baser, dev); val = (1ULL << KVM_ITS_DTE_VALID_SHIFT | ((u64)next_offset << KVM_ITS_DTE_NEXT_SHIFT) | (itt_addr_field << KVM_ITS_DTE_ITTADDR_SHIFT) | @@ -2385,15 +2390,16 @@ static int vgic_its_save_device_tables(struct vgic_its *its) int ret; gpa_t eaddr; + /* Don't fail a save that userspace must be able to issue. */ if (!vgic_its_check_id(its, baser, dev->device_id, &eaddr)) - return -EINVAL; + continue; ret = vgic_its_save_itt(its, dev); if (ret) return ret; - ret = vgic_its_save_dte(its, dev, eaddr); + ret = vgic_its_save_dte(its, baser, dev, eaddr); if (ret) return ret; } From 6f182db39fb00548c26d689163ceb2fe4a802793 Mon Sep 17 00:00:00 2001 From: Fuad Tabba Date: Fri, 21 Aug 2026 07:44:45 +0100 Subject: [PATCH 077/694] KVM: arm64: selftests: Add ITS table save tests Cover the two ways a guest can leave a table that KVM_DEV_ARM_ITS_SAVE_TABLES has to cope with: a GITS_BASER write that changes the table, which drops the collections it described, and a device whose L2 block the guest invalidated, which the save skips. Each case then resets and restores, which is what the save exists for. Both fail without the preceding patches. The first save returns -EINVAL, and the second saves a DTE pointing 8192 entries ahead, at an entry the save never wrote. Assisted-by: Antigravity:gemini-3.1-pro Signed-off-by: Fuad Tabba Reviewed-by: Marc Zyngier Link: https://patch.msgid.link/20260821064445.615838-5-fuad.tabba@linux.dev Signed-off-by: Oliver Upton --- tools/testing/selftests/kvm/Makefile.kvm | 1 + .../selftests/kvm/arm64/vgic_its_save.c | 441 ++++++++++++++++++ 2 files changed, 442 insertions(+) create mode 100644 tools/testing/selftests/kvm/arm64/vgic_its_save.c diff --git a/tools/testing/selftests/kvm/Makefile.kvm b/tools/testing/selftests/kvm/Makefile.kvm index 96bab7002d39..6a1482e3a286 100644 --- a/tools/testing/selftests/kvm/Makefile.kvm +++ b/tools/testing/selftests/kvm/Makefile.kvm @@ -189,6 +189,7 @@ TEST_GEN_PROGS_arm64 += arm64/stage2_block_transitions TEST_GEN_PROGS_arm64 += arm64/vcpu_width_config TEST_GEN_PROGS_arm64 += arm64/vgic_init TEST_GEN_PROGS_arm64 += arm64/vgic_irq +TEST_GEN_PROGS_arm64 += arm64/vgic_its_save TEST_GEN_PROGS_arm64 += arm64/vgic_lpi_stress TEST_GEN_PROGS_arm64 += arm64/vgic_v5 TEST_GEN_PROGS_arm64 += arm64/vpmu_counter_access diff --git a/tools/testing/selftests/kvm/arm64/vgic_its_save.c b/tools/testing/selftests/kvm/arm64/vgic_its_save.c new file mode 100644 index 000000000000..864da01539f3 --- /dev/null +++ b/tools/testing/selftests/kvm/arm64/vgic_its_save.c @@ -0,0 +1,441 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * vgic_its_save - KVM_DEV_ARM_ITS_SAVE_TABLES against tables a guest broke. + * + * Both cases are reachable by a guest on its own, and neither may fail a save + * that userspace has to be able to issue: + * + * - Changing GITS_BASER drops the collections it described, so the save + * writes nothing but the terminating invalid entry. + * - A device the device table can no longer address is skipped, and the saved + * DTE chain skips it too rather than pointing at an entry never written. + * + * Both cases then reset and restore, which is what the save exists for. + * + * Copyright (c) 2026 Google LLC + * Author: Fuad Tabba + */ + +#include +#include +#include +#include + +#include "kvm_util.h" +#include "gic.h" +#include "gic_v3.h" +#include "gic_v3_its.h" +#include "processor.h" +#include "ucall.h" +#include "vgic.h" + +#define TEST_MEMSLOT_INDEX 1 + +/* All three ITS table entry sizes are 8 bytes in ABI 0. */ +#define ESZ 8 +#define ENTRIES_PER_PAGE (SZ_64K / ESZ) + +/* CTE and DTE layout, mirroring KVM's KVM_ITS_* in arch/arm64/kvm/vgic/vgic.h */ +#define CTE_VALID_MASK BIT_ULL(63) +#define DTE_VALID_MASK BIT_ULL(63) +#define DTE_NEXT_SHIFT 49 +#define DTE_NEXT_MASK GENMASK_ULL(62, 49) + +/* L1 entry of an indirect table: valid bit plus a 64K aligned L2 address. */ +#define L1E_VALID_MASK BIT_ULL(63) +#define L1E_ADDR_MASK GENMASK_ULL(51, 16) + +#define GITS_BASER_PAGES_MASK GENMASK_ULL(7, 0) + +#define POISON 0xdeadbeefdeadbeefULL + +/* The collection table starts at two pages and is shrunk to one. */ +#define COLL_TBL_PAGES 2 +#define COLL_TBL_SZ (COLL_TBL_PAGES * SZ_64K) + +/* One more collection than the shrunken table can hold. */ +#define NR_COLLECTIONS (ENTRIES_PER_PAGE + 1) + +/* Two devices, one per L2 block of the indirect device table. */ +#define DEVICE_A_ID 0 +#define DEVICE_B_ID ENTRIES_PER_PAGE + +/* + * its_send_mapd_cmd() encodes ilog2(itt_size) - 1 as num_eventid_bits, and + * vgic_its_restore_itt() scans BIT_ULL(num_eventid_bits) * ESZ, so the size + * handed to MAPD has to match the ITT allocated for it. + */ +#define ITT_EVENTID_BITS 13 +#define ITT_MAPD_SIZE BIT_ULL(ITT_EVENTID_BITS + 1) +#define ITT_SZ (BIT_ULL(ITT_EVENTID_BITS) * ESZ) + +static struct kvm_vm *vm; +static struct kvm_vcpu *vcpu; +static int its_fd; +static gpa_t gpa_base; + +static struct test_data { + gpa_t device_table; + gpa_t collection_table; + gpa_t cmdq_base; + void *cmdq_base_va; + + gpa_t lpi_prop_table; + gpa_t lpi_pend_table; + + void *device_l1_va; + gpa_t device_l2[2]; + gpa_t itt_tables; +} test_data; + +static unsigned long its_baser_offset(unsigned int type) +{ + int i; + + for (i = 0; i < GITS_BASER_NR_REGS; i++) { + unsigned long offset = GITS_BASER + (i * sizeof(u64)); + u64 baser = readq_relaxed(GITS_BASE_GVA + offset); + + if (GITS_BASER_TYPE(baser) == type) + return offset; + } + + GUEST_FAIL("Couldn't find an ITS BASER of type %u", type); + return -1; +} + +static void its_set_enable(bool enable) +{ + u32 ctlr = readl_relaxed(GITS_BASE_GVA + GITS_CTLR); + + if (enable) + ctlr |= GITS_CTLR_ENABLE; + else + ctlr &= ~GITS_CTLR_ENABLE; + + writel_relaxed(ctlr, GITS_BASE_GVA + GITS_CTLR); +} + +/* + * Shrink the collection table to a single page, leaving VALID set. BASER + * writes are ignored while the ITS is enabled. + */ +static void guest_shrink_coll_table(void) +{ + unsigned long offset = its_baser_offset(GITS_BASER_TYPE_COLLECTION); + u64 baser; + + its_set_enable(false); + + baser = readq_relaxed(GITS_BASE_GVA + offset); + baser &= ~GITS_BASER_PAGES_MASK; + writeq_relaxed(baser, GITS_BASE_GVA + offset); +} + +static void guest_baser_change(void) +{ + u32 coll_id; + + gic_init(GIC_V3, 1); + gic_rdist_enable_lpis(test_data.lpi_prop_table, SZ_64K, + test_data.lpi_pend_table); + + its_init(test_data.collection_table, COLL_TBL_SZ, + test_data.device_table, SZ_64K, + test_data.cmdq_base, SZ_64K); + + for (coll_id = 0; coll_id < NR_COLLECTIONS; coll_id++) + its_send_mapc_cmd(test_data.cmdq_base_va, 0, coll_id, true); + + guest_shrink_coll_table(); + + GUEST_DONE(); +} + +/* Turn the already installed device table into an indirect one. */ +static void guest_make_device_table_indirect(void) +{ + unsigned long offset = its_baser_offset(GITS_BASER_TYPE_DEVICE); + u64 baser; + + its_set_enable(false); + + baser = readq_relaxed(GITS_BASE_GVA + offset); + writeq_relaxed(baser | GITS_BASER_INDIRECT, GITS_BASE_GVA + offset); + + its_set_enable(true); +} + +static void guest_unreachable_device(void) +{ + u64 *l1; + + gic_init(GIC_V3, 1); + gic_rdist_enable_lpis(test_data.lpi_prop_table, SZ_64K, + test_data.lpi_pend_table); + + its_init(test_data.collection_table, SZ_64K, + test_data.device_table, SZ_64K, + test_data.cmdq_base, SZ_64K); + + guest_make_device_table_indirect(); + + /* Both L2 blocks present, so both MAPDs are in range. */ + l1 = test_data.device_l1_va; + l1[0] = L1E_VALID_MASK | (test_data.device_l2[0] & L1E_ADDR_MASK); + l1[1] = L1E_VALID_MASK | (test_data.device_l2[1] & L1E_ADDR_MASK); + + its_send_mapd_cmd(test_data.cmdq_base_va, DEVICE_A_ID, + test_data.itt_tables, ITT_MAPD_SIZE, true); + its_send_mapd_cmd(test_data.cmdq_base_va, DEVICE_B_ID, + test_data.itt_tables + ITT_SZ, ITT_MAPD_SIZE, true); + + /* + * Drop the block holding device B. No ITS command and no GITS_BASER + * write is involved, so nothing tells KVM the device is now + * unreachable. + */ + l1[1] = 0; + + GUEST_DONE(); +} + +static void run_guest(void) +{ + struct ucall uc; + + vcpu_run(vcpu); + switch (get_ucall(vcpu, &uc)) { + case UCALL_DONE: + break; + case UCALL_ABORT: + REPORT_GUEST_ASSERT(uc); + break; + default: + TEST_FAIL("Unexpected ucall: %lu", uc.cmd); + } +} + +static int save_tables(void) +{ + return __kvm_device_attr_set(its_fd, KVM_DEV_ARM_VGIC_GRP_CTRL, + KVM_DEV_ARM_ITS_SAVE_TABLES, NULL); +} + +static u64 its_reg_get(unsigned long offset) +{ + u64 val; + + kvm_device_attr_get(its_fd, KVM_DEV_ARM_VGIC_GRP_ITS_REGS, offset, + &val); + return val; +} + +static void its_reg_set(unsigned long offset, u64 val) +{ + kvm_device_attr_set(its_fd, KVM_DEV_ARM_VGIC_GRP_ITS_REGS, offset, + &val); +} + +/* + * What a migration target does with the saved tables, in the order + * Documentation/virt/kvm/devices/arm-vgic-its.rst gives: the GITS_ registers + * first, then the tables. The reset in between clears GITS_BASER.Valid, + * which is why the registers have to be written back before the restore. + */ +static void reset_and_restore_tables(void) +{ + u64 baser[GITS_BASER_NR_REGS]; + int ret, i; + + for (i = 0; i < GITS_BASER_NR_REGS; i++) + baser[i] = its_reg_get(GITS_BASER + (i * sizeof(u64))); + + ret = __kvm_device_attr_set(its_fd, KVM_DEV_ARM_VGIC_GRP_CTRL, + KVM_DEV_ARM_ITS_CTRL_RESET, NULL); + TEST_ASSERT(!ret, "Expected the reset to succeed, got ret %d errno %d", + ret, errno); + + for (i = 0; i < GITS_BASER_NR_REGS; i++) + its_reg_set(GITS_BASER + (i * sizeof(u64)), baser[i]); + + ret = __kvm_device_attr_set(its_fd, KVM_DEV_ARM_VGIC_GRP_CTRL, + KVM_DEV_ARM_ITS_RESTORE_TABLES, NULL); + TEST_ASSERT(!ret, "Expected the restore to succeed, got ret %d errno %d", + ret, errno); +} + +static void poison_range(gpa_t base, size_t size) +{ + u64 *entry = addr_gpa2hva(vm, base); + size_t i; + + for (i = 0; i < size / ESZ; i++) + entry[i] = POISON; +} + +static void setup_memslot(size_t sz) +{ + size_t pages = sz / vm->page_size; + + gpa_base = ((vm_compute_max_gfn(vm) + 1) * vm->page_size) - sz; + vm_userspace_mem_region_add(vm, VM_MEM_SRC_ANONYMOUS, gpa_base, + TEST_MEMSLOT_INDEX, pages, 0); +} + +static gpa_t alloc_64k(size_t nr) +{ + size_t pages_per_64k = vm_calc_num_guest_pages(vm->mode, SZ_64K); + gpa_t gpa = vm_phy_pages_alloc(vm, nr * pages_per_64k, gpa_base, + TEST_MEMSLOT_INDEX); + + TEST_ASSERT(IS_ALIGNED(gpa, SZ_64K), + "Allocation at 0x%lx is not 64K aligned, GITS_BASER cannot address it", + gpa); + return gpa; +} + +static void map_to_guest(gpa_t gpa, size_t nr) +{ + size_t pages_per_64k = vm_calc_num_guest_pages(vm->mode, SZ_64K); + + virt_map(vm, gpa, gpa, nr * pages_per_64k); +} + +static void setup_common(void) +{ + test_data.cmdq_base = alloc_64k(1); + map_to_guest(test_data.cmdq_base, 1); + test_data.cmdq_base_va = (void *)test_data.cmdq_base; + + test_data.lpi_prop_table = alloc_64k(1); + test_data.lpi_pend_table = alloc_64k(1); +} + +static void teardown(void) +{ + close(its_fd); + kvm_vm_free(vm); + memset(&test_data, 0, sizeof(test_data)); +} + +/* + * A GITS_BASER write that changes the table drops the collections it + * described. The save then has an empty list, so it writes the terminating + * invalid entry and nothing else. + */ +static void test_baser_change_drops_collections(void) +{ + u64 *cte; + int ret, i; + + pr_info("Testing that a GITS_BASER change drops the collections\n"); + + vm = vm_create_with_one_vcpu(&vcpu, guest_baser_change); + setup_memslot((4 + COLL_TBL_PAGES) * SZ_64K); + its_fd = vgic_its_setup(vm); + + test_data.device_table = alloc_64k(1); + test_data.collection_table = alloc_64k(COLL_TBL_PAGES); + setup_common(); + + sync_global_to_guest(vm, test_data); + run_guest(); + + /* Anything KVM writes is then the only thing that changed. */ + poison_range(test_data.collection_table, COLL_TBL_SZ); + + ret = save_tables(); + TEST_ASSERT(!ret, "Expected the save to succeed, got %d errno %d", + ret, errno); + + cte = addr_gpa2hva(vm, test_data.collection_table); + + /* + * Finding the terminator at the head of the table is also what proves + * the reads below landed in the saved table rather than elsewhere. + */ + TEST_ASSERT(le64toh(cte[0]) == 0, + "CTE 0: expected the terminating invalid entry, got 0x%llx", + (unsigned long long)le64toh(cte[0])); + + for (i = 1; i < COLL_TBL_SZ / ESZ; i++) + TEST_ASSERT(cte[i] == POISON, + "CTE %d: expected it untouched, got 0x%llx", + i, (unsigned long long)cte[i]); + + reset_and_restore_tables(); + + teardown(); +} + +/* + * A device whose L2 block the guest dropped is skipped by the save, and the + * DTE chain skips it too: left alone, the surviving device would point at an + * entry the save never wrote. + */ +static void test_unreachable_device_skipped(void) +{ + u64 dte; + int ret; + + pr_info("Testing that an unreachable device is skipped by the save\n"); + + vm = vm_create_with_one_vcpu(&vcpu, guest_unreachable_device); + setup_memslot(9 * SZ_64K); + its_fd = vgic_its_setup(vm); + + test_data.device_table = alloc_64k(1); + test_data.collection_table = alloc_64k(1); + test_data.device_l2[0] = alloc_64k(1); + test_data.device_l2[1] = alloc_64k(1); + test_data.itt_tables = alloc_64k(2); + setup_common(); + + map_to_guest(test_data.device_table, 1); + test_data.device_l1_va = (void *)test_data.device_table; + + sync_global_to_guest(vm, test_data); + run_guest(); + + poison_range(test_data.device_l2[0], SZ_64K); + poison_range(test_data.device_l2[1], SZ_64K); + + ret = save_tables(); + TEST_ASSERT(!ret, "Expected the save to succeed, got %d errno %d", + ret, errno); + + dte = le64toh(*(u64 *)addr_gpa2hva(vm, test_data.device_l2[0])); + + /* Device A is still reachable, so it is saved. */ + TEST_ASSERT(dte & DTE_VALID_MASK, + "Device A: expected a valid DTE, got 0x%llx", + (unsigned long long)dte); + + /* + * Device B is the only device after it and was skipped, so nothing + * follows A in the saved chain. + */ + TEST_ASSERT(FIELD_GET(DTE_NEXT_MASK, dte) == 0, + "Device A: expected no next device, got offset %llu", + (unsigned long long)FIELD_GET(DTE_NEXT_MASK, dte)); + + /* And nothing was written into the block the guest dropped. */ + TEST_ASSERT(*(u64 *)addr_gpa2hva(vm, test_data.device_l2[1]) == POISON, + "Device B: expected its entry untouched"); + + reset_and_restore_tables(); + + teardown(); +} + +int main(void) +{ + TEST_REQUIRE(kvm_supports_vgic_v3()); + + test_baser_change_drops_collections(); + test_unreachable_device_skipped(); + + pr_info("All ok!\n"); + return 0; +} From 38b70fc453c3112f1a62583b89903ae41116cc27 Mon Sep 17 00:00:00 2001 From: "Lorenzo Stoakes (ARM)" Date: Tue, 1 Sep 2026 18:28:59 +0100 Subject: [PATCH 078/694] KVM: arm64: Fix spurious warning for benign stage 2 teardown race kvmtool was used to establish an L1 guest with 8 CPUs and 8 GiB of RAM, an L2 guest with 4 CPUs and 4 GiB of RAM and an L3 guest with 2 CPUs and 2 GiB of RAM, all of which was then exited. Under memory pressure in the L0 host warnings were observed due to migration triggered by compaction: WARNING: arch/arm64/kvm/mmu.c:336 at __unmap_stage2_range+0x64/0x80, CPU#5: kcompactd0/66 Which was, in turn, triggered by an MMU notifier for the host invalidation: mmu_notifier_invalidate_range_start() -> ... -> kvm_mmu_notifier_invalidate_range_start() -> kvm_mmu_unmap_gfn_range() -> kvm_unmap_gfn_range() -> kvm_nested_s2_unmap() -> kvm_stage2_unmap_range() -> __unmap_stage2_range() -> stage2_apply_range() <- -EINVAL, triggering a WARN_ON() Racing with L0's teardown of stage 2 page tables: exit_mm() -> mmput() -> __mmput() -> exit_mmap() -> mmu_notifier_release() -> ... -> kvm_mmu_notifier_release() -> kvm_flush_shadow_all() -> kvm_arch_flush_shadow_all() -> kvm_free_stage2_pgd() -> [ acquire kvm->mmu_lock for write ] -> mmu->pgt = NULL [ among other tasks ] -> [ release kvm->mmu_lock for write ] It turns out there is a benign race resulting in a spurious warning: Thread A - notify: migration | Thread B - notify: release -------------------------------|--------------------------------- < kvm->mmu_lock held > | stage2_apply_range() | get mmu->pgt, check !NULL | ... | kvm_arch_flush_shadow_all() cond_resched_rwlock_write(); | < contend, sleep kvm->mmu_lock > < drop kvm->mmu_lock > | < acquire kvm->mmu_lock> | ... | kvm_free_stage2_pgd() | mmu->pgt = NULL | < invalidate MMU > | ... | < release kvm->mmu_lock > [ scheduled ] | stage2_apply_range() | < loop to next > | get, mmu->pgt, check !NULL | is NULL, return -EINVAL | __unmap_stage2_range() | WARN_ON(-EINVAL) <--- entirely spurious - the race was handled correctly. Fix the spurious warning by updating stage2_apply_range() to no longer treat concurrent PGT teardown on lock release as an error - whether the walker is tearing down page tables or doing something else this is a legitimate reason to abort the operation without error. This keeps the warning in place for all other circumstances. In practice only __unmap_stage2_range() actually does anything with the error so this only impacts that. Fixes: ec14c272408a ("KVM: arm64: nv: Unmap/flush shadow stage 2 page tables") Cc: stable@vger.kernel.org Reviewed-by: Yuan Yao Reviewed-by: Marc Zyngier Signed-off-by: Lorenzo Stoakes (ARM) Link: https://patch.msgid.link/20260901-kvm-arm-nested-virt-fix-v3-1-b154676f7e4c@kernel.org Signed-off-by: Oliver Upton --- arch/arm64/kvm/mmu.c | 15 ++++++++++++--- 1 file changed, 12 insertions(+), 3 deletions(-) diff --git a/arch/arm64/kvm/mmu.c b/arch/arm64/kvm/mmu.c index 9ba86450fe4a..2d44cd6a5aed 100644 --- a/arch/arm64/kvm/mmu.c +++ b/arch/arm64/kvm/mmu.c @@ -59,27 +59,36 @@ static phys_addr_t stage2_range_addr_end(phys_addr_t addr, phys_addr_t end) * long will also starve other vCPUs. We have to also make sure that the page * tables are not freed while we released the lock. */ -static int stage2_apply_range(struct kvm_s2_mmu *mmu, phys_addr_t addr, +static int stage2_apply_range(struct kvm_s2_mmu *mmu, phys_addr_t start, phys_addr_t end, int (*fn)(struct kvm_pgtable *, u64, u64), bool resched) { struct kvm *kvm = kvm_s2_mmu_to_kvm(mmu); + bool lock_dropped = false; + phys_addr_t addr = start; int ret; u64 next; do { struct kvm_pgtable *pgt = mmu->pgt; + /* + * We may be raced on PGT teardown when we release the + * kvm->mmu_lock. That's fine as the PGT is legitimately no + * longer present. + */ if (!pgt) - return -EINVAL; + return lock_dropped ? 0 : -EINVAL; next = stage2_range_addr_end(addr, end); ret = fn(pgt, addr, next - addr); if (ret) break; - if (resched && next != end) + if (resched && next != end) { cond_resched_rwlock_write(&kvm->mmu_lock); + lock_dropped = true; + } } while (addr = next, addr != end); return ret; From 4c74e233cdedd11592775fae2a6243e67ca3f891 Mon Sep 17 00:00:00 2001 From: "Lorenzo Stoakes (ARM)" Date: Tue, 1 Sep 2026 18:29:00 +0100 Subject: [PATCH 079/694] KVM: arm64: nv: Fix null ptr deref on nested wp/unmap, teardown race Commit 7270cc9157f4 ("KVM: arm64: nv: Handle VNCR_EL2 invalidation from MMU notifiers") introduced VNCR_EL2 invalidation in both kvm_nested_s2_unmap() and kvm_nested_s2_wp(). However at the point of this being performed concurrent stage 2 teardown of a nested guest can cause kvm->arch.mmu.pgt to be set to NULL. This happens in kvm_flush_shadow_all() -> kvm_arch_flush_shadow_all() -> kvm_free_stage2_pgd() and is performed under the kvm->mmu_lock. Commit ec14c272408a ("KVM: arm64: nv: Unmap/flush shadow stage 2 page tables") introduced the teardown of the entire nested MMU range, which then invokes stage2_apply_range() with resched=true: mmu_notifier_invalidate_range_start() -> ... -> kvm_mmu_notifier_invalidate_range_start() -> kvm_mmu_unmap_gfn_range() -> kvm_unmap_gfn_range() -> kvm_nested_s2_unmap() -> kvm_stage2_unmap_range() -> __unmap_stage2_range() -> stage2_apply_range() This means that stage2_apply_range() can drop the kvm->mmu_lock and thus concurrent progress can be made in lockstep with kvm_arch_flush_shadow_all(). If kvm_arch_flush_shadow_all() advances ahead of stage2_apply_range() and completes its operation it guarantees a NULL pointer deref. Since kvm_free_stage2_pgd() is performed under the kvm->mmu_lock this will either be observed NULL or not and serialised against kvm_free_stage2_pgd(). Resolve the issue by abstracting the invalidation to a new function, kvm_invalidate_vncr_ipa_all(), and check that the pgt is non-NULL before dereferencing it. Fixes: 7270cc9157f4 ("KVM: arm64: nv: Handle VNCR_EL2 invalidation from MMU notifiers") Cc: stable@vger.kernel.org Reviewed-by: Marc Zyngier Signed-off-by: Lorenzo Stoakes (ARM) Tested-by: Jonathan Davies Link: https://patch.msgid.link/20260901-kvm-arm-nested-virt-fix-v3-2-b154676f7e4c@kernel.org Signed-off-by: Oliver Upton --- arch/arm64/kvm/nested.c | 15 +++++++++++++-- 1 file changed, 13 insertions(+), 2 deletions(-) diff --git a/arch/arm64/kvm/nested.c b/arch/arm64/kvm/nested.c index 3c4fc566eafc..a0808391a456 100644 --- a/arch/arm64/kvm/nested.c +++ b/arch/arm64/kvm/nested.c @@ -1260,6 +1260,17 @@ void kvm_handle_s1e2_tlbi(struct kvm_vcpu *vcpu, u32 inst, u64 val) invalidate_vncr_va(vcpu->kvm, &scope); } +static void kvm_invalidate_vncr_ipa_all(struct kvm *kvm) +{ + struct kvm_pgtable *pgt = kvm->arch.mmu.pgt; + + lockdep_assert_held_write(&kvm->mmu_lock); + + /* if the mmu lock was dropped, pgt teardown may have raced. */ + if (pgt) + kvm_invalidate_vncr_ipa(kvm, 0, BIT(pgt->ia_bits)); +} + void kvm_nested_s2_wp(struct kvm *kvm) { int i; @@ -1276,7 +1287,7 @@ void kvm_nested_s2_wp(struct kvm *kvm) kvm_stage2_wp_range(mmu, 0, kvm_phys_size(mmu)); } - kvm_invalidate_vncr_ipa(kvm, 0, BIT(kvm->arch.mmu.pgt->ia_bits)); + kvm_invalidate_vncr_ipa_all(kvm); } void kvm_nested_s2_unmap(struct kvm *kvm, bool may_block) @@ -1295,7 +1306,7 @@ void kvm_nested_s2_unmap(struct kvm *kvm, bool may_block) kvm_stage2_unmap_range(mmu, 0, kvm_phys_size(mmu), may_block); } - kvm_invalidate_vncr_ipa(kvm, 0, BIT(kvm->arch.mmu.pgt->ia_bits)); + kvm_invalidate_vncr_ipa_all(kvm); } void kvm_nested_s2_flush(struct kvm *kvm) From 2a2eb10795a1e495aebc7f829ccecb72c05b4fd9 Mon Sep 17 00:00:00 2001 From: Fuad Tabba Date: Tue, 25 Aug 2026 09:59:45 +0100 Subject: [PATCH 080/694] KVM: arm64: Validate the SVE vector length in pkvm_vcpu_init_sve() pkvm_vcpu_init_sve() clamps only the upper bound of the host-provided sve_max_vl, so an invalid vector length reaches sve_state_size_from_vl() and the WARN_ON() there, which is fatal at EL2. The existing !sve_state_size test rejects such a length, but only after the macro has run. Check sve_vl_valid() before deriving the state size. A valid length cannot yield a zero size, so the !sve_state_size test goes with it. Fixes: 5db1bef93342 ("KVM: arm64: Track SVE state in the hypervisor vcpu structure") Reported-by: Stefan Teodorescu Reviewed-by: Marc Zyngier Signed-off-by: Fuad Tabba Link: https://patch.msgid.link/20260825085948.1674721-2-fuad.tabba@linux.dev Signed-off-by: Oliver Upton --- arch/arm64/kvm/hyp/nvhe/pkvm.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/arch/arm64/kvm/hyp/nvhe/pkvm.c b/arch/arm64/kvm/hyp/nvhe/pkvm.c index 459bd9eb7e4b..627f13fc98d4 100644 --- a/arch/arm64/kvm/hyp/nvhe/pkvm.c +++ b/arch/arm64/kvm/hyp/nvhe/pkvm.c @@ -459,14 +459,15 @@ static int pkvm_vcpu_init_sve(struct pkvm_hyp_vcpu *hyp_vcpu, struct kvm_vcpu *h /* Limit guest vector length to the maximum supported by the host. */ sve_max_vl = min(READ_ONCE(host_vcpu->arch.sve_max_vl), kvm_host_sve_max_vl); - sve_state_size = sve_state_size_from_vl(sve_max_vl); sve_state = kern_hyp_va(READ_ONCE(host_vcpu->arch.sve_state)); - if (!sve_state || !sve_state_size) { + if (!sve_vl_valid(sve_max_vl) || !sve_state) { ret = -EINVAL; goto err; } + sve_state_size = sve_state_size_from_vl(sve_max_vl); + ret = hyp_pin_shared_mem(sve_state, sve_state + sve_state_size); if (ret) goto err; From a1b3c788ad31837e348075e93dbba3f447492776 Mon Sep 17 00:00:00 2001 From: Fuad Tabba Date: Tue, 25 Aug 2026 09:59:46 +0100 Subject: [PATCH 081/694] KVM: arm64: Do not clear VM-wide SVE feature on vCPU init failure pkvm_vcpu_init_sve() clears KVM_ARM_VCPU_SVE in kvm->arch.vcpu_features when it fails, but vcpu_has_sve() tests KVM_ARCH_FLAG_GUEST_HAS_SVE, which is left set. Later vCPUs on that VM then skip the SVE setup and register with a NULL sve_state, which the guest's first FP access hands to sve_load_state(). Return the error without touching vcpu_features. Fixes: 5db1bef93342 ("KVM: arm64: Track SVE state in the hypervisor vcpu structure") Signed-off-by: Fuad Tabba Link: https://patch.msgid.link/20260825085948.1674721-3-fuad.tabba@linux.dev Signed-off-by: Oliver Upton --- arch/arm64/kvm/hyp/nvhe/pkvm.c | 13 ++++--------- 1 file changed, 4 insertions(+), 9 deletions(-) diff --git a/arch/arm64/kvm/hyp/nvhe/pkvm.c b/arch/arm64/kvm/hyp/nvhe/pkvm.c index 627f13fc98d4..4857a11d4292 100644 --- a/arch/arm64/kvm/hyp/nvhe/pkvm.c +++ b/arch/arm64/kvm/hyp/nvhe/pkvm.c @@ -450,7 +450,7 @@ static int pkvm_vcpu_init_sve(struct pkvm_hyp_vcpu *hyp_vcpu, struct kvm_vcpu *h unsigned int sve_max_vl; size_t sve_state_size; void *sve_state; - int ret = 0; + int ret; if (!vcpu_has_feature(vcpu, KVM_ARM_VCPU_SVE)) { vcpu_clear_flag(vcpu, VCPU_SVE_FINALIZED); @@ -461,24 +461,19 @@ static int pkvm_vcpu_init_sve(struct pkvm_hyp_vcpu *hyp_vcpu, struct kvm_vcpu *h sve_max_vl = min(READ_ONCE(host_vcpu->arch.sve_max_vl), kvm_host_sve_max_vl); sve_state = kern_hyp_va(READ_ONCE(host_vcpu->arch.sve_state)); - if (!sve_vl_valid(sve_max_vl) || !sve_state) { - ret = -EINVAL; - goto err; - } + if (!sve_vl_valid(sve_max_vl) || !sve_state) + return -EINVAL; sve_state_size = sve_state_size_from_vl(sve_max_vl); ret = hyp_pin_shared_mem(sve_state, sve_state + sve_state_size); if (ret) - goto err; + return ret; vcpu->arch.sve_state = sve_state; vcpu->arch.sve_max_vl = sve_max_vl; return 0; -err: - clear_bit(KVM_ARM_VCPU_SVE, vcpu->kvm->arch.vcpu_features); - return ret; } static int vm_copy_id_regs(struct pkvm_hyp_vcpu *hyp_vcpu) From 0d62fbf34d8fe7a3ff56692d39fbb8e6e9bf15fb Mon Sep 17 00:00:00 2001 From: Fuad Tabba Date: Tue, 25 Aug 2026 09:59:47 +0100 Subject: [PATCH 082/694] KVM: arm64: Key unpin_host_sve_state() on the state it unpins unpin_host_sve_state() gates on the VM's SVE feature bit, but what it unpins is the state pkvm_vcpu_init_sve() pinned. A vCPU that completed init has sve_state set exactly when that bit is set, so the two agree. Gate on sve_state, which is what is being unpinned. Signed-off-by: Fuad Tabba Link: https://patch.msgid.link/20260825085948.1674721-4-fuad.tabba@linux.dev Signed-off-by: Oliver Upton --- arch/arm64/kvm/hyp/nvhe/pkvm.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/arm64/kvm/hyp/nvhe/pkvm.c b/arch/arm64/kvm/hyp/nvhe/pkvm.c index 4857a11d4292..453d3fa4a420 100644 --- a/arch/arm64/kvm/hyp/nvhe/pkvm.c +++ b/arch/arm64/kvm/hyp/nvhe/pkvm.c @@ -398,10 +398,10 @@ static void unpin_host_sve_state(struct pkvm_hyp_vcpu *hyp_vcpu) { void *sve_state; - if (!vcpu_has_feature(&hyp_vcpu->vcpu, KVM_ARM_VCPU_SVE)) + sve_state = hyp_vcpu->vcpu.arch.sve_state; + if (!sve_state) return; - sve_state = hyp_vcpu->vcpu.arch.sve_state; hyp_unpin_shared_mem(sve_state, sve_state + vcpu_sve_state_size(&hyp_vcpu->vcpu)); } From 4f16c5fc8dc4c5596e3777ab9f449a54e3f85fd5 Mon Sep 17 00:00:00 2001 From: Fuad Tabba Date: Tue, 25 Aug 2026 09:59:48 +0100 Subject: [PATCH 083/694] KVM: arm64: Derive GUEST_HAS_SVE from the SVE feature bit at EL2 pkvm_init_features_from_host() takes KVM_ARCH_FLAG_GUEST_HAS_SVE and KVM_ARM_VCPU_SVE from the host separately, but pkvm_vcpu_init_sve() tests the bit while vcpu_has_sve() reads the flag. A host that sets the flag without the bit gets a vCPU with a NULL sve_state that the world switch loads the guest's SVE state from. Derive the flag from the bit, and drop the protected path's copy of the host's flag, which is dead code since protected VMs are not allowed SVE. Fixes: 41d6028e28bd ("KVM: arm64: Convert the SVE guest vcpu flag to a vm flag") Signed-off-by: Fuad Tabba Link: https://patch.msgid.link/20260825085948.1674721-5-fuad.tabba@linux.dev Signed-off-by: Oliver Upton --- arch/arm64/kvm/hyp/nvhe/pkvm.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/arch/arm64/kvm/hyp/nvhe/pkvm.c b/arch/arm64/kvm/hyp/nvhe/pkvm.c index 453d3fa4a420..bb3e0dc0676e 100644 --- a/arch/arm64/kvm/hyp/nvhe/pkvm.c +++ b/arch/arm64/kvm/hyp/nvhe/pkvm.c @@ -360,7 +360,7 @@ static void pkvm_init_features_from_host(struct pkvm_hyp_vm *hyp_vm, const struc if (test_bit(KVM_ARCH_FLAG_WRITABLE_IMP_ID_REGS, &host_arch_flags)) hyp_vm->kvm.arch.midr_el1 = host_kvm->arch.midr_el1; - return; + goto out; } if (kvm_pkvm_ext_allowed(kvm, KVM_CAP_ARM_MTE)) @@ -379,13 +379,14 @@ static void pkvm_init_features_from_host(struct pkvm_hyp_vm *hyp_vm, const struc if (kvm_pkvm_ext_allowed(kvm, KVM_CAP_ARM_PTRAUTH_GENERIC)) set_bit(KVM_ARM_VCPU_PTRAUTH_GENERIC, allowed_features); - if (kvm_pkvm_ext_allowed(kvm, KVM_CAP_ARM_SVE)) { + if (kvm_pkvm_ext_allowed(kvm, KVM_CAP_ARM_SVE)) set_bit(KVM_ARM_VCPU_SVE, allowed_features); - kvm->arch.flags |= host_arch_flags & BIT(KVM_ARCH_FLAG_GUEST_HAS_SVE); - } bitmap_and(kvm->arch.vcpu_features, host_kvm->arch.vcpu_features, allowed_features, KVM_VCPU_MAX_FEATURES); +out: + __assign_bit(KVM_ARCH_FLAG_GUEST_HAS_SVE, &kvm->arch.flags, + kvm_vcpu_has_feature(kvm, KVM_ARM_VCPU_SVE)); } static void unpin_host_vcpu(struct kvm_vcpu *host_vcpu) From 64dc6f1db7e620f2e9337bb181f305fb0561da79 Mon Sep 17 00:00:00 2001 From: Karl Mehltretter Date: Sat, 29 Aug 2026 07:48:55 +0200 Subject: [PATCH 084/694] KVM: arm64: Return -EINVAL for an empty SMCCC filter range at base 0 kvm_smccc_set_filter() only rejects a range if its inclusive end, base + nr_functions - 1, is below base. That catches an empty range (nr_functions == 0) at every nonzero base, but at base 0 the end wraps to U32_MAX and KVM tries to insert [0, U32_MAX], which overlaps the reserved Arm Architecture Calls ranges. KVM_ARM_VM_SMCCC_FILTER then returns -EEXIST instead of the -EINVAL that the smccc_filter selftest expects for an empty range. Reject a zero function count explicitly. Tested with a userspace reproducer on an arm64 VHE host under QEMU TCG: EEXIST before, EINVAL after. Fixes: 821d935c87bc ("KVM: arm64: Introduce support for userspace SMCCC filtering") Assisted-by: LLM Signed-off-by: Karl Mehltretter Reviewed-by: Steffen Eiden Reviewed-by: Fuad Tabba Tested-by: Fuad Tabba Link: https://patch.msgid.link/20260829054856.70549-2-kmehltretter@gmail.com Signed-off-by: Oliver Upton --- arch/arm64/kvm/hypercalls.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/arch/arm64/kvm/hypercalls.c b/arch/arm64/kvm/hypercalls.c index b11b8821c9fb..dfa25bb6f25d 100644 --- a/arch/arm64/kvm/hypercalls.c +++ b/arch/arm64/kvm/hypercalls.c @@ -185,7 +185,8 @@ static int kvm_smccc_set_filter(struct kvm *kvm, struct kvm_smccc_filter __user start = filter.base; end = start + filter.nr_functions - 1; - if (end < start || filter.action >= NR_SMCCC_FILTER_ACTIONS) + if (!filter.nr_functions || end < start || + filter.action >= NR_SMCCC_FILTER_ACTIONS) return -EINVAL; mutex_lock(&kvm->arch.config_lock); From 0a46eb5719fa57cc9d06025dcd8ba271643dee61 Mon Sep 17 00:00:00 2001 From: Karl Mehltretter Date: Sat, 29 Aug 2026 07:48:56 +0200 Subject: [PATCH 085/694] KVM: arm64: selftests: Test empty SMCCC filter range at base 0 test_invalid_nr_functions() only checks an empty range at PSCI_0_2_FN64_CPU_ON, which KVM's end < start check happens to catch. It never exercised base 0, where the inclusive end wraps to U32_MAX instead. Add the base 0 case. Without the preceding fix it fails with EEXIST. Assisted-by: LLM Signed-off-by: Karl Mehltretter Reviewed-by: Steffen Eiden Reviewed-by: Fuad Tabba Tested-by: Fuad Tabba Link: https://patch.msgid.link/20260829054856.70549-3-kmehltretter@gmail.com Signed-off-by: Oliver Upton --- tools/testing/selftests/kvm/arm64/smccc_filter.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/tools/testing/selftests/kvm/arm64/smccc_filter.c b/tools/testing/selftests/kvm/arm64/smccc_filter.c index 21e41880261b..a41ed3e016ba 100644 --- a/tools/testing/selftests/kvm/arm64/smccc_filter.c +++ b/tools/testing/selftests/kvm/arm64/smccc_filter.c @@ -140,6 +140,10 @@ static void test_invalid_nr_functions(void) TEST_ASSERT(r < 0 && errno == EINVAL, "Attempt to filter 0 functions should return EINVAL"); + r = __set_smccc_filter(vm, 0, 0, KVM_SMCCC_FILTER_DENY); + TEST_ASSERT(r < 0 && errno == EINVAL, + "Attempt to filter 0 functions at base 0 should return EINVAL"); + kvm_vm_free(vm); } From 3a8c562892b96f35bba1e00d5e455a15963bbb92 Mon Sep 17 00:00:00 2001 From: Fuad Tabba Date: Tue, 8 Sep 2026 12:07:10 +0100 Subject: [PATCH 086/694] KVM: arm64: Transfer the hyp stack pages out of the host stage-2 fix_host_ownership() walks only the linear-map alias of each memblock region, and the per-CPU hyp stack, mapped in the private VA range for its guard page, has none. Walk each stack's VA range with the same walker. Fixes: 1a919b17ef012 ("KVM: arm64: Add guard pages for pKVM (protected nVHE) hypervisor stack") Reported-by: Hiroyuki Katsura Cc: stable@vger.kernel.org Signed-off-by: Fuad Tabba Reviewed-by: Vincent Donnefort Tested-by: Vincent Donnefort Reviewed-by: Marc Zyngier Link: https://patch.msgid.link/20260908110713.1540304-2-fuad.tabba@linux.dev Signed-off-by: Oliver Upton --- arch/arm64/kvm/hyp/nvhe/setup.c | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/arch/arm64/kvm/hyp/nvhe/setup.c b/arch/arm64/kvm/hyp/nvhe/setup.c index 75b00c323310..362f2891cb32 100644 --- a/arch/arm64/kvm/hyp/nvhe/setup.c +++ b/arch/arm64/kvm/hyp/nvhe/setup.c @@ -269,6 +269,16 @@ static int fix_host_ownership(void) return ret; } + /* The stacks sit in the private VA range, not the linear map. */ + for (i = 0; i < hyp_nr_cpus; i++) { + struct kvm_nvhe_init_params *params = per_cpu_ptr(&kvm_init_params, i); + u64 start = params->stack_hyp_va - NVHE_STACK_SIZE; + + ret = kvm_pgtable_walk(&pkvm_pgtable, start, NVHE_STACK_SIZE, &walker); + if (ret) + return ret; + } + return 0; } From 5a8b505ede133fb30ca3b3a19d0db00c08237615 Mon Sep 17 00:00:00 2001 From: Fuad Tabba Date: Tue, 8 Sep 2026 12:07:11 +0100 Subject: [PATCH 087/694] KVM: arm64: Match hyp text by physical address in fix_host_ownership() On a non-hVHE host, fix_host_ownership_walker()'s test for PAGE_HYP_EXEC never matches: KVM_PGTABLE_PROT_UX is cleared at map time and only PX is reported on read-back. Hyp text is therefore donated rather than left read-only in the host stage-2, and the instruction dump in nvhe_hyp_panic_handler() reads a page the host has no access to. Match the text by physical address instead, in a helper a later patch reuses. A test on the permissions would leave any other executable mapping host-readable too. Fixes: 80cbfd7174f31 ("KVM: arm64: Honor UX/PX attributes for EL2 S1 mappings") Signed-off-by: Fuad Tabba Reviewed-by: Vincent Donnefort Tested-by: Vincent Donnefort Reviewed-by: Marc Zyngier Link: https://patch.msgid.link/20260908110713.1540304-3-fuad.tabba@linux.dev Signed-off-by: Oliver Upton --- arch/arm64/kvm/hyp/include/nvhe/mem_protect.h | 1 + arch/arm64/kvm/hyp/nvhe/mem_protect.c | 8 ++++++++ arch/arm64/kvm/hyp/nvhe/setup.c | 2 +- 3 files changed, 10 insertions(+), 1 deletion(-) diff --git a/arch/arm64/kvm/hyp/include/nvhe/mem_protect.h b/arch/arm64/kvm/hyp/include/nvhe/mem_protect.h index 29935c7da1de..cab27f7bd423 100644 --- a/arch/arm64/kvm/hyp/include/nvhe/mem_protect.h +++ b/arch/arm64/kvm/hyp/include/nvhe/mem_protect.h @@ -52,6 +52,7 @@ int __pkvm_host_test_clear_young_guest(u64 gfn, u64 nr_pages, bool mkold, struct int __pkvm_host_mkyoung_guest(u64 gfn, struct pkvm_hyp_vcpu *vcpu); bool addr_is_memory(phys_addr_t phys); +bool addr_is_hyp_text(phys_addr_t phys); int host_stage2_idmap_locked(phys_addr_t addr, u64 size, enum kvm_pgtable_prot prot); int host_stage2_set_owner_locked(phys_addr_t addr, u64 size, u8 owner_id); int kvm_host_prepare_stage2(void *pgt_pool_base); diff --git a/arch/arm64/kvm/hyp/nvhe/mem_protect.c b/arch/arm64/kvm/hyp/nvhe/mem_protect.c index 39aa8911f62c..d026f446bd8e 100644 --- a/arch/arm64/kvm/hyp/nvhe/mem_protect.c +++ b/arch/arm64/kvm/hyp/nvhe/mem_protect.c @@ -450,6 +450,14 @@ bool addr_is_memory(phys_addr_t phys) return !!find_mem_range(phys, &range); } +bool addr_is_hyp_text(phys_addr_t phys) +{ + phys_addr_t start = ALIGN_DOWN(__hyp_pa(__hyp_text_start), PAGE_SIZE); + phys_addr_t end = PAGE_ALIGN(__hyp_pa(__hyp_text_end)); + + return phys >= start && phys < end; +} + static bool is_in_mem_range(u64 addr, struct kvm_mem_range *range) { return range->start <= addr && addr < range->end; diff --git a/arch/arm64/kvm/hyp/nvhe/setup.c b/arch/arm64/kvm/hyp/nvhe/setup.c index 362f2891cb32..bb667cd7080b 100644 --- a/arch/arm64/kvm/hyp/nvhe/setup.c +++ b/arch/arm64/kvm/hyp/nvhe/setup.c @@ -217,7 +217,7 @@ static int fix_host_ownership_walker(const struct kvm_pgtable_visit_ctx *ctx, case PKVM_PAGE_OWNED: set_hyp_state(page, PKVM_PAGE_OWNED); /* hyp text is RO in the host stage-2 to be inspected on panic. */ - if (prot == PAGE_HYP_EXEC) { + if (addr_is_hyp_text(phys)) { set_host_state(page, PKVM_NOPAGE); return host_stage2_idmap_locked(phys, PAGE_SIZE, KVM_PGTABLE_PROT_R); } else { From 2245841401147b8022a5857061b870d82e595352 Mon Sep 17 00:00:00 2001 From: Fuad Tabba Date: Tue, 8 Sep 2026 12:07:12 +0100 Subject: [PATCH 088/694] KVM: arm64: Move the private VA allocation cursor to __io_map_next __io_map_base is the start of the private VA range only until the first allocation from it, after which it is the allocation cursor. Keep it as the start and move the cursor to __io_map_next, for the walk of the range the next patch adds. No functional change intended. Signed-off-by: Fuad Tabba Reviewed-by: Vincent Donnefort Tested-by: Vincent Donnefort Reviewed-by: Marc Zyngier Link: https://patch.msgid.link/20260908110713.1540304-4-fuad.tabba@linux.dev Signed-off-by: Oliver Upton --- arch/arm64/kvm/hyp/nvhe/mm.c | 20 +++++++++++--------- 1 file changed, 11 insertions(+), 9 deletions(-) diff --git a/arch/arm64/kvm/hyp/nvhe/mm.c b/arch/arm64/kvm/hyp/nvhe/mm.c index 3b0bee496bff..422ee57be956 100644 --- a/arch/arm64/kvm/hyp/nvhe/mm.c +++ b/arch/arm64/kvm/hyp/nvhe/mm.c @@ -25,6 +25,7 @@ struct memblock_region hyp_memory[HYP_MEMBLOCK_REGIONS]; unsigned int hyp_memblock_nr; static u64 __io_map_base; +static u64 __io_map_next; struct hyp_fixmap_slot { u64 addr; @@ -50,7 +51,7 @@ static int __pkvm_alloc_private_va_range(unsigned long start, size_t size) hyp_assert_lock_held(&pkvm_pgd_lock); - if (!start || start < __io_map_base) + if (!start || start < __io_map_next) return -EINVAL; /* The allocated size is always a multiple of PAGE_SIZE */ @@ -60,7 +61,7 @@ static int __pkvm_alloc_private_va_range(unsigned long start, size_t size) if (cur > __hyp_vmemmap) return -ENOMEM; - __io_map_base = cur; + __io_map_next = cur; return 0; } @@ -70,7 +71,7 @@ static int __pkvm_alloc_private_va_range(unsigned long start, size_t size) * @size: The size of the VA range to reserve. * @haddr: The hypervisor virtual start address of the allocation. * - * The private virtual address (VA) range is allocated above __io_map_base + * The private virtual address (VA) range is allocated above __io_map_next * and aligned based on the order of @size. * * Return: 0 on success or negative error code on failure. @@ -81,7 +82,7 @@ int pkvm_alloc_private_va_range(size_t size, unsigned long *haddr) int ret; hyp_spin_lock(&pkvm_pgd_lock); - addr = __io_map_base; + addr = __io_map_next; ret = __pkvm_alloc_private_va_range(addr, size); hyp_spin_unlock(&pkvm_pgd_lock); @@ -341,7 +342,7 @@ static int create_fixblock(void) return -EINVAL; hyp_spin_lock(&pkvm_pgd_lock); - addr = ALIGN(__io_map_base, PMD_SIZE); + addr = ALIGN(__io_map_next, PMD_SIZE); ret = __pkvm_alloc_private_va_range(addr, PMD_SIZE); if (ret) goto unlock; @@ -426,6 +427,7 @@ int hyp_create_idmap(u32 hyp_va_bits) */ __io_map_base = start & BIT(hyp_va_bits - 2); __io_map_base ^= BIT(hyp_va_bits - 2); + __io_map_next = __io_map_base; __hyp_vmemmap = __io_map_base | BIT(hyp_va_bits - 3); return __pkvm_create_mappings(start, end - start, start, PAGE_HYP_EXEC); @@ -433,19 +435,19 @@ int hyp_create_idmap(u32 hyp_va_bits) int pkvm_create_stack(phys_addr_t phys, unsigned long *haddr) { - unsigned long addr, prev_base; + unsigned long addr, prev_next; size_t size; int ret; hyp_spin_lock(&pkvm_pgd_lock); - prev_base = __io_map_base; + prev_next = __io_map_next; /* * Efficient stack verification using the NVHE_STACK_SHIFT bit implies * an alignment of our allocation on the order of the size. */ size = NVHE_STACK_SIZE * 2; - addr = ALIGN(__io_map_base, size); + addr = ALIGN(__io_map_next, size); ret = __pkvm_alloc_private_va_range(addr, size); if (!ret) { @@ -461,7 +463,7 @@ int pkvm_create_stack(phys_addr_t phys, unsigned long *haddr) ret = kvm_pgtable_hyp_map(&pkvm_pgtable, addr + NVHE_STACK_SIZE, NVHE_STACK_SIZE, phys, PAGE_HYP); if (ret) - __io_map_base = prev_base; + __io_map_next = prev_next; } hyp_spin_unlock(&pkvm_pgd_lock); From cfe80c3837f93202970b6d8f706f79c5c7396f17 Mon Sep 17 00:00:00 2001 From: Fuad Tabba Date: Tue, 8 Sep 2026 12:07:13 +0100 Subject: [PATCH 089/694] KVM: arm64: Check every private mapping is hyp-owned at pKVM init fix_host_ownership() transfers only what it walks, so a hyp mapping outside the linear map is not manipulated by the walk. Walk the quarter of the VA space holding the private range and the vmemmap once the transfer is done, and fail init unless every valid leaf is hyp-owned: in the vmemmap when the page is memory, and in the host stage-2, where hyp text may instead be mapped without write access. A leaf that is not memory has no vmemmap entry and is checked against the host stage-2 alone. Hyp text is matched by physical address, since the only executable mapping in the range is the Spectre-v3a vectors, whose VA is a private allocation, and an executable mapping of anything else must not be host-readable. The vmemmap can be block-mapped, so the walker checks each page of a leaf. Suggested-by: Will Deacon Signed-off-by: Fuad Tabba Reviewed-by: Vincent Donnefort Tested-by: Vincent Donnefort Reviewed-by: Marc Zyngier Link: https://patch.msgid.link/20260908110713.1540304-5-fuad.tabba@linux.dev Signed-off-by: Oliver Upton --- arch/arm64/kvm/hyp/include/nvhe/mem_protect.h | 1 + arch/arm64/kvm/hyp/include/nvhe/mm.h | 1 + arch/arm64/kvm/hyp/nvhe/mem_protect.c | 12 ++++ arch/arm64/kvm/hyp/nvhe/mm.c | 59 +++++++++++++++++++ arch/arm64/kvm/hyp/nvhe/setup.c | 4 ++ 5 files changed, 77 insertions(+) diff --git a/arch/arm64/kvm/hyp/include/nvhe/mem_protect.h b/arch/arm64/kvm/hyp/include/nvhe/mem_protect.h index cab27f7bd423..ec85a9547120 100644 --- a/arch/arm64/kvm/hyp/include/nvhe/mem_protect.h +++ b/arch/arm64/kvm/hyp/include/nvhe/mem_protect.h @@ -55,6 +55,7 @@ bool addr_is_memory(phys_addr_t phys); bool addr_is_hyp_text(phys_addr_t phys); int host_stage2_idmap_locked(phys_addr_t addr, u64 size, enum kvm_pgtable_prot prot); int host_stage2_set_owner_locked(phys_addr_t addr, u64 size, u8 owner_id); +bool host_stage2_pte_is_hyp_owned(kvm_pte_t pte); int kvm_host_prepare_stage2(void *pgt_pool_base); int kvm_guest_prepare_stage2(struct pkvm_hyp_vm *vm, void *pgd); void kvm_guest_destroy_stage2(struct pkvm_hyp_vm *vm); diff --git a/arch/arm64/kvm/hyp/include/nvhe/mm.h b/arch/arm64/kvm/hyp/include/nvhe/mm.h index 6e83ce35c2f2..31cae95ddb71 100644 --- a/arch/arm64/kvm/hyp/include/nvhe/mm.h +++ b/arch/arm64/kvm/hyp/include/nvhe/mm.h @@ -29,6 +29,7 @@ int __pkvm_create_private_mapping(phys_addr_t phys, size_t size, enum kvm_pgtable_prot prot, unsigned long *haddr); int pkvm_create_stack(phys_addr_t phys, unsigned long *haddr); +int pkvm_check_host_ownership(void); int pkvm_alloc_private_va_range(size_t size, unsigned long *haddr); #endif /* __KVM_HYP_MM_H */ diff --git a/arch/arm64/kvm/hyp/nvhe/mem_protect.c b/arch/arm64/kvm/hyp/nvhe/mem_protect.c index d026f446bd8e..a6a47c1e058b 100644 --- a/arch/arm64/kvm/hyp/nvhe/mem_protect.c +++ b/arch/arm64/kvm/hyp/nvhe/mem_protect.c @@ -641,6 +641,18 @@ int host_stage2_set_owner_locked(phys_addr_t addr, u64 size, u8 owner_id) return ret; } +bool host_stage2_pte_is_hyp_owned(kvm_pte_t pte) +{ + if (kvm_pte_valid(pte)) + return false; + + if (FIELD_GET(KVM_INVALID_PTE_TYPE_MASK, pte) != + KVM_HOST_INVALID_PTE_TYPE_DONATION) + return false; + + return FIELD_GET(KVM_HOST_DONATION_PTE_OWNER_MASK, pte) == PKVM_ID_HYP; +} + #define KVM_HOST_PTE_OWNER_GUEST_HANDLE_MASK GENMASK(15, 0) /* We need 40 bits for the GFN to cover a 52-bit IPA with 4k pages and LPA2 */ #define KVM_HOST_PTE_OWNER_GUEST_GFN_MASK GENMASK(55, 16) diff --git a/arch/arm64/kvm/hyp/nvhe/mm.c b/arch/arm64/kvm/hyp/nvhe/mm.c index 422ee57be956..29ab5ee9d57f 100644 --- a/arch/arm64/kvm/hyp/nvhe/mm.c +++ b/arch/arm64/kvm/hyp/nvhe/mm.c @@ -472,6 +472,65 @@ int pkvm_create_stack(phys_addr_t phys, unsigned long *haddr) return ret; } +static int check_page_ownership(phys_addr_t phys) +{ + kvm_pte_t pte; + bool host_ok; + int ret; + + if (addr_is_memory(phys)) { + struct hyp_page *page = hyp_phys_to_page(phys); + + if (get_hyp_state(page) != PKVM_PAGE_OWNED || + get_host_state(page) != PKVM_NOPAGE) + return -EPERM; + } + + ret = kvm_pgtable_get_leaf(&host_mmu.pgt, phys, &pte, NULL); + if (ret) + return ret; + + /* Hyp text may stay host-readable, see fix_host_ownership_walker(). */ + if (kvm_pte_valid(pte) && addr_is_hyp_text(phys)) + host_ok = !(kvm_pgtable_stage2_pte_prot(pte) & KVM_PGTABLE_PROT_W); + else + host_ok = host_stage2_pte_is_hyp_owned(pte); + + return host_ok ? 0 : -EPERM; +} + +static int check_host_ownership_walker(const struct kvm_pgtable_visit_ctx *ctx, + enum kvm_pgtable_walk_flags visit) +{ + phys_addr_t phys, end; + int ret; + + if (!kvm_pte_valid(ctx->old)) + return 0; + + phys = kvm_pte_to_phys(ctx->old); + end = phys + kvm_granule_size(ctx->level); + for (; phys < end; phys += PAGE_SIZE) { + ret = check_page_ownership(phys); + if (ret) + return ret; + } + + return 0; +} + +int pkvm_check_host_ownership(void) +{ + struct kvm_pgtable_walker walker = { + .cb = check_host_ownership_walker, + .flags = KVM_PGTABLE_WALK_LEAF, + }; + + /* The private range and the vmemmap share one quarter of the VA space. */ + return kvm_pgtable_walk(&pkvm_pgtable, __io_map_base, + BIT(pkvm_pgtable.ia_bits - 2), &walker); +} + static void *admit_host_page(void *arg) { struct kvm_hyp_memcache *host_mc = arg; diff --git a/arch/arm64/kvm/hyp/nvhe/setup.c b/arch/arm64/kvm/hyp/nvhe/setup.c index bb667cd7080b..45ac5f2ba4f7 100644 --- a/arch/arm64/kvm/hyp/nvhe/setup.c +++ b/arch/arm64/kvm/hyp/nvhe/setup.c @@ -334,6 +334,10 @@ void __noreturn __pkvm_init_finalise(void) if (ret) goto out; + ret = pkvm_check_host_ownership(); + if (ret) + goto out; + ret = hyp_ffa_init(ffa_proxy_pages); if (ret) goto out; From 33346f8960c7bb6a3b4e273b5cfe25c5a8be349f Mon Sep 17 00:00:00 2001 From: Marc Zyngier Date: Fri, 11 Sep 2026 17:22:02 +0100 Subject: [PATCH 090/694] KVM: arm64: nv: Fix life cycle of the nested_mmus array The nested_mmus array holds the shadow page tables that are used when a guest is running a nested context. These structures are allocated on VCPU_INIT for whole guest, which implies that they may have to be relocated as the array grows. Should a VCPU_INIT occur whilst a vcpu is actively running an L2 and that the allocation requires relocation, that vcpu will still be running with a pointer to the previous structure, which will have been freed. Fix this by turning the array of structures to an array of pointers, which is now allocated at VM creation, sized to the absolute maximum that KVM can handle. In turn, each VCPU_INIT contributes S2_MMU_PER_VCPU to the pool. No reallocation is ever performed, and the life cycle of each object is much clearer: - the nested_mmus array is allocated in kvm_init_nested(), and freed in kvm_arch_destroy_vm() - s2_mmu structures are allocated in kvm_vcpu_init_nested(), and freed on kvm_arch_flush_shadow_all() Finally, the freeing of vcpu->arch.vncr_array is made consistent rather than being done on some failure paths, but not others. Fixes: 4f128f8e1aaa ("KVM: arm64: nv: Support multiple nested Stage-2 mmu structures") Reported-by: Shen Yongchao Reported-by: Karl Mehltretter Suggested-by: Karl Mehltretter Acked-by: Lorenzo Stoakes (ARM) Link: https://lore.kernel.org/r/20260803224405.41468-1-kmehltretter@gmail.com Signed-off-by: Marc Zyngier Cc: stable@vger.kernel.org Reviewed-by: Wei-Lin Chang Link: https://patch.msgid.link/20260911162203.1919330-2-maz@kernel.org Signed-off-by: Oliver Upton --- arch/arm64/include/asm/kvm_host.h | 2 +- arch/arm64/include/asm/kvm_nested.h | 2 +- arch/arm64/kvm/arm.c | 8 ++- arch/arm64/kvm/nested.c | 93 +++++++++++++---------------- 4 files changed, 51 insertions(+), 54 deletions(-) diff --git a/arch/arm64/include/asm/kvm_host.h b/arch/arm64/include/asm/kvm_host.h index 27fe0cd5b2d7..cd9b9d2462f9 100644 --- a/arch/arm64/include/asm/kvm_host.h +++ b/arch/arm64/include/asm/kvm_host.h @@ -322,7 +322,7 @@ struct kvm_arch { * Stage 2 paging state for VMs with nested S2 using a virtual * VMID. */ - struct kvm_s2_mmu *nested_mmus; + struct kvm_s2_mmu **nested_mmus; size_t nested_mmus_size; int nested_mmus_next; diff --git a/arch/arm64/include/asm/kvm_nested.h b/arch/arm64/include/asm/kvm_nested.h index 1ed708335809..5b8edb2e8a87 100644 --- a/arch/arm64/include/asm/kvm_nested.h +++ b/arch/arm64/include/asm/kvm_nested.h @@ -66,7 +66,7 @@ static inline u64 translate_ttbr0_el2_to_ttbr0_el1(u64 ttbr0) extern bool forward_smc_trap(struct kvm_vcpu *vcpu); extern bool forward_debug_exception(struct kvm_vcpu *vcpu); -extern void kvm_init_nested(struct kvm *kvm); +extern int kvm_init_nested(struct kvm *kvm); extern int kvm_vcpu_init_nested(struct kvm_vcpu *vcpu); extern void kvm_init_nested_s2_mmu(struct kvm_s2_mmu *mmu); extern struct kvm_s2_mmu *lookup_s2_mmu(struct kvm_vcpu *vcpu); diff --git a/arch/arm64/kvm/arm.c b/arch/arm64/kvm/arm.c index 8b080804bc90..b53219e048bd 100644 --- a/arch/arm64/kvm/arm.c +++ b/arch/arm64/kvm/arm.c @@ -236,8 +236,6 @@ int kvm_arch_init_vm(struct kvm *kvm, unsigned long type) mutex_unlock(&kvm->lock); #endif - kvm_init_nested(kvm); - ret = kvm_share_hyp(kvm, kvm + 1); if (ret) return ret; @@ -252,6 +250,10 @@ int kvm_arch_init_vm(struct kvm *kvm, unsigned long type) if (ret) goto err_free_cpumask; + ret = kvm_init_nested(kvm); + if (ret) + goto err_uninit_mmu; + if (is_protected_kvm_enabled()) { /* * If any failures occur after this is successful, make sure to @@ -280,6 +282,7 @@ int kvm_arch_init_vm(struct kvm *kvm, unsigned long type) err_uninit_mmu: kvm_uninit_stage2_mmu(kvm); + kvfree(kvm->arch.nested_mmus); err_free_cpumask: free_cpumask_var(kvm->arch.supported_cpus); err_unshare_kvm: @@ -337,6 +340,7 @@ void kvm_arch_destroy_vm(struct kvm *kvm) kvm_unshare_hyp(kvm, kvm + 1); + kvfree(kvm->arch.nested_mmus); kvm_arm_teardown_hypercalls(kvm); } diff --git a/arch/arm64/kvm/nested.c b/arch/arm64/kvm/nested.c index a0808391a456..2571f177e654 100644 --- a/arch/arm64/kvm/nested.c +++ b/arch/arm64/kvm/nested.c @@ -45,11 +45,15 @@ struct vncr_tlb { */ #define S2_MMU_PER_VCPU 2 -void kvm_init_nested(struct kvm *kvm) +int kvm_init_nested(struct kvm *kvm) { - kvm->arch.nested_mmus = NULL; + kvm->arch.nested_mmus = kvmalloc_objs(struct kvm_s2_mmu *, + KVM_MAX_VCPUS * S2_MMU_PER_VCPU, + GFP_KERNEL_ACCOUNT); kvm->arch.nested_mmus_size = 0; atomic_set(&kvm->arch.vncr_tlb_count, 0); + + return kvm->arch.nested_mmus ? 0 : -ENOMEM; } static int init_nested_s2_mmu(struct kvm *kvm, struct kvm_s2_mmu *mmu) @@ -70,8 +74,9 @@ static int init_nested_s2_mmu(struct kvm *kvm, struct kvm_s2_mmu *mmu) int kvm_vcpu_init_nested(struct kvm_vcpu *vcpu) { struct kvm *kvm = vcpu->kvm; - struct kvm_s2_mmu *tmp; - int num_mmus, ret = 0; + int num_mmus; + + lockdep_assert_held(&kvm->arch.config_lock); if (test_bit(KVM_ARM_VCPU_HAS_EL2_E2H0, kvm->arch.vcpu_features) && !cpus_have_final_cap(ARM64_HAS_HCR_NV1)) @@ -84,51 +89,40 @@ int kvm_vcpu_init_nested(struct kvm_vcpu *vcpu) if (!vcpu->arch.ctxt.vncr_array) return -ENOMEM; - /* - * Let's treat memory allocation failures as benign: If we fail to - * allocate anything, return an error and keep the allocated array - * alive. Userspace may try to recover by initializing the vcpu - * again, and there is no reason to affect the whole VM for this. - */ num_mmus = atomic_read(&kvm->online_vcpus) * S2_MMU_PER_VCPU; if (num_mmus > kvm->arch.nested_mmus_size) { - tmp = kvzalloc_objs(*tmp, num_mmus, GFP_KERNEL_ACCOUNT); + struct kvm_s2_mmu *tmp; + int i, ret = 0; + + tmp = kvzalloc_objs(*tmp, S2_MMU_PER_VCPU, GFP_KERNEL_ACCOUNT); if (!tmp) - return -ENOMEM; + ret = -ENOMEM; - write_lock(&kvm->mmu_lock); - - if (kvm->arch.nested_mmus_size) { - memcpy(tmp, kvm->arch.nested_mmus, - size_mul(sizeof(*tmp), kvm->arch.nested_mmus_size)); - - for (int i = 0; i < kvm->arch.nested_mmus_size; i++) - tmp[i].pgt->mmu = &tmp[i]; + for (i = 0; !ret && i < S2_MMU_PER_VCPU; i++) { + ret = init_nested_s2_mmu(kvm, &tmp[i]); + if (ret) + break; } - swap(kvm->arch.nested_mmus, tmp); + if (ret) { + while (--i >= 0) + kvm_free_stage2_pgd(&tmp[i]); - write_unlock(&kvm->mmu_lock); + kvfree(tmp); + free_page((unsigned long)vcpu->arch.ctxt.vncr_array); + vcpu->arch.ctxt.vncr_array = NULL; + return ret; + } - kvfree(tmp); + guard(write_lock)(&kvm->mmu_lock); + + for (i = 0; i < S2_MMU_PER_VCPU; i++) + kvm->arch.nested_mmus[i + kvm->arch.nested_mmus_size] = &tmp[i]; + + kvm->arch.nested_mmus_size += S2_MMU_PER_VCPU; } - for (int i = kvm->arch.nested_mmus_size; !ret && i < num_mmus; i++) - ret = init_nested_s2_mmu(kvm, &kvm->arch.nested_mmus[i]); - - if (ret) { - for (int i = kvm->arch.nested_mmus_size; i < num_mmus; i++) - kvm_free_stage2_pgd(&kvm->arch.nested_mmus[i]); - - free_page((unsigned long)vcpu->arch.ctxt.vncr_array); - vcpu->arch.ctxt.vncr_array = NULL; - - return ret; - } - - kvm->arch.nested_mmus_size = num_mmus; - return 0; } @@ -742,7 +736,7 @@ void kvm_s2_mmu_iterate_by_vmid(struct kvm *kvm, u16 vmid, write_lock(&kvm->mmu_lock); for (int i = 0; i < kvm->arch.nested_mmus_size; i++) { - struct kvm_s2_mmu *mmu = &kvm->arch.nested_mmus[i]; + struct kvm_s2_mmu *mmu = kvm->arch.nested_mmus[i]; if (!kvm_s2_mmu_valid(mmu)) continue; @@ -784,7 +778,7 @@ struct kvm_s2_mmu *lookup_s2_mmu(struct kvm_vcpu *vcpu) * if S2 translation is disabled. */ for (int i = 0; i < kvm->arch.nested_mmus_size; i++) { - struct kvm_s2_mmu *mmu = &kvm->arch.nested_mmus[i]; + struct kvm_s2_mmu *mmu = kvm->arch.nested_mmus[i]; if (!kvm_s2_mmu_valid(mmu)) continue; @@ -823,7 +817,7 @@ static struct kvm_s2_mmu *get_s2_mmu_nested(struct kvm_vcpu *vcpu) for (i = kvm->arch.nested_mmus_next; i < (kvm->arch.nested_mmus_size + kvm->arch.nested_mmus_next); i++) { - s2_mmu = &kvm->arch.nested_mmus[i % kvm->arch.nested_mmus_size]; + s2_mmu = kvm->arch.nested_mmus[i % kvm->arch.nested_mmus_size]; if (atomic_read(&s2_mmu->refcnt) == 0) break; @@ -1281,7 +1275,7 @@ void kvm_nested_s2_wp(struct kvm *kvm) return; for (i = 0; i < kvm->arch.nested_mmus_size; i++) { - struct kvm_s2_mmu *mmu = &kvm->arch.nested_mmus[i]; + struct kvm_s2_mmu *mmu = kvm->arch.nested_mmus[i]; if (kvm_s2_mmu_valid(mmu)) kvm_stage2_wp_range(mmu, 0, kvm_phys_size(mmu)); @@ -1300,7 +1294,7 @@ void kvm_nested_s2_unmap(struct kvm *kvm, bool may_block) return; for (i = 0; i < kvm->arch.nested_mmus_size; i++) { - struct kvm_s2_mmu *mmu = &kvm->arch.nested_mmus[i]; + struct kvm_s2_mmu *mmu = kvm->arch.nested_mmus[i]; if (kvm_s2_mmu_valid(mmu)) kvm_stage2_unmap_range(mmu, 0, kvm_phys_size(mmu), may_block); @@ -1319,7 +1313,7 @@ void kvm_nested_s2_flush(struct kvm *kvm) return; for (i = 0; i < kvm->arch.nested_mmus_size; i++) { - struct kvm_s2_mmu *mmu = &kvm->arch.nested_mmus[i]; + struct kvm_s2_mmu *mmu = kvm->arch.nested_mmus[i]; if (kvm_s2_mmu_valid(mmu)) kvm_stage2_flush_range(mmu, 0, kvm_phys_size(mmu)); @@ -1328,16 +1322,15 @@ void kvm_nested_s2_flush(struct kvm *kvm) void kvm_arch_flush_shadow_all(struct kvm *kvm) { - int i; - - for (i = 0; i < kvm->arch.nested_mmus_size; i++) { - struct kvm_s2_mmu *mmu = &kvm->arch.nested_mmus[i]; + for (int i = kvm->arch.nested_mmus_size - 1; i >= 0; i--) { + struct kvm_s2_mmu *mmu = kvm->arch.nested_mmus[i]; if (!WARN_ON(atomic_read(&mmu->refcnt))) kvm_free_stage2_pgd(mmu); + + if ((i % S2_MMU_PER_VCPU) == 0) + kvfree(mmu); } - kvfree(kvm->arch.nested_mmus); - kvm->arch.nested_mmus = NULL; kvm->arch.nested_mmus_size = 0; kvm_uninit_stage2_mmu(kvm); } From e5843f4effaa2ffac3e789ecd4456403564961d4 Mon Sep 17 00:00:00 2001 From: Marc Zyngier Date: Fri, 11 Sep 2026 17:22:03 +0100 Subject: [PATCH 091/694] KVM: arm64: nv: Delay freeing of shadow S2 structures until VM destruction We free the shadow S2 structures from kvm_arch_flush_shadow_all(), which is a Bad Idea(tm). Freeing the page tables is fair game (this is what this callback is for), but freeing the container that could still be referenced by another part of the system is not great. Instead, grow separate destructors that gets called when we tear the VM down for good. From there, we can nuke both the individual MMUs as well as the global array that points to them, safe in the knowledge that the vcpus themselves have been destroyed already. Fixes: 4f128f8e1aaac ("KVM: arm64: nv: Support multiple nested Stage-2 mmu structures") Reviewed-by: Lorenzo Stoakes (ARM) Signed-off-by: Marc Zyngier Cc: stable@vger.kernel.org Reviewed-by: Wei-Lin Chang Link: https://patch.msgid.link/20260911162203.1919330-3-maz@kernel.org Signed-off-by: Oliver Upton --- arch/arm64/include/asm/kvm_nested.h | 1 + arch/arm64/kvm/arm.c | 4 ++-- arch/arm64/kvm/nested.c | 15 ++++++++++----- 3 files changed, 13 insertions(+), 7 deletions(-) diff --git a/arch/arm64/include/asm/kvm_nested.h b/arch/arm64/include/asm/kvm_nested.h index 5b8edb2e8a87..586026e85903 100644 --- a/arch/arm64/include/asm/kvm_nested.h +++ b/arch/arm64/include/asm/kvm_nested.h @@ -67,6 +67,7 @@ static inline u64 translate_ttbr0_el2_to_ttbr0_el1(u64 ttbr0) extern bool forward_smc_trap(struct kvm_vcpu *vcpu); extern bool forward_debug_exception(struct kvm_vcpu *vcpu); extern int kvm_init_nested(struct kvm *kvm); +extern void kvm_destroy_nested(struct kvm *kvm); extern int kvm_vcpu_init_nested(struct kvm_vcpu *vcpu); extern void kvm_init_nested_s2_mmu(struct kvm_s2_mmu *mmu); extern struct kvm_s2_mmu *lookup_s2_mmu(struct kvm_vcpu *vcpu); diff --git a/arch/arm64/kvm/arm.c b/arch/arm64/kvm/arm.c index b53219e048bd..eaf583b77193 100644 --- a/arch/arm64/kvm/arm.c +++ b/arch/arm64/kvm/arm.c @@ -282,7 +282,7 @@ int kvm_arch_init_vm(struct kvm *kvm, unsigned long type) err_uninit_mmu: kvm_uninit_stage2_mmu(kvm); - kvfree(kvm->arch.nested_mmus); + kvm_destroy_nested(kvm); err_free_cpumask: free_cpumask_var(kvm->arch.supported_cpus); err_unshare_kvm: @@ -340,7 +340,7 @@ void kvm_arch_destroy_vm(struct kvm *kvm) kvm_unshare_hyp(kvm, kvm + 1); - kvfree(kvm->arch.nested_mmus); + kvm_destroy_nested(kvm); kvm_arm_teardown_hypercalls(kvm); } diff --git a/arch/arm64/kvm/nested.c b/arch/arm64/kvm/nested.c index 2571f177e654..b191365d97cc 100644 --- a/arch/arm64/kvm/nested.c +++ b/arch/arm64/kvm/nested.c @@ -56,6 +56,15 @@ int kvm_init_nested(struct kvm *kvm) return kvm->arch.nested_mmus ? 0 : -ENOMEM; } +void kvm_destroy_nested(struct kvm *kvm) +{ + for (int i = 0; i < kvm->arch.nested_mmus_size; i+= S2_MMU_PER_VCPU) + kvfree(kvm->arch.nested_mmus[i]); + + kvm->arch.nested_mmus_size = 0; + kvfree(kvm->arch.nested_mmus); +} + static int init_nested_s2_mmu(struct kvm *kvm, struct kvm_s2_mmu *mmu) { /* @@ -1322,16 +1331,12 @@ void kvm_nested_s2_flush(struct kvm *kvm) void kvm_arch_flush_shadow_all(struct kvm *kvm) { - for (int i = kvm->arch.nested_mmus_size - 1; i >= 0; i--) { + for (int i = 0; i < kvm->arch.nested_mmus_size; i++) { struct kvm_s2_mmu *mmu = kvm->arch.nested_mmus[i]; if (!WARN_ON(atomic_read(&mmu->refcnt))) kvm_free_stage2_pgd(mmu); - - if ((i % S2_MMU_PER_VCPU) == 0) - kvfree(mmu); } - kvm->arch.nested_mmus_size = 0; kvm_uninit_stage2_mmu(kvm); } From 49d9d295d69d07e850cae35933ba8519e2915f26 Mon Sep 17 00:00:00 2001 From: Fuad Tabba Date: Mon, 14 Sep 2026 10:38:38 +0100 Subject: [PATCH 092/694] KVM: arm64: Don't WARN on an unknown VM ioctl in protected mode kvm_pkvm_ioctl_allowed() WARNs when kvm_get_cap_for_kvm_ioctl() doesn't find the ioctl number in vm_ioctl_caps[], and kvm_arch_vm_ioctl() calls it for every number the generic code doesn't handle, so ioctl(vm_fd, 0xdeadbeef) from userspace taints a pKVM host and panics it under panic_on_warn. The lookup is fed userspace input: return false, and userspace gets the -EINVAL kvm_arch_vm_ioctl() returns for that number on a host without pKVM. Fixes: b12b3b04f6ba0 ("KVM: arm64: Check whether a VM IOCTL is allowed in pKVM") Cc: stable@vger.kernel.org Signed-off-by: Fuad Tabba Reviewed-by: Suzuki K Poulose Link: https://patch.msgid.link/20260914093838.1082637-1-fuad.tabba@linux.dev Signed-off-by: Oliver Upton --- arch/arm64/include/asm/kvm_pkvm.h | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/arch/arm64/include/asm/kvm_pkvm.h b/arch/arm64/include/asm/kvm_pkvm.h index beea00e693a0..cad60569f061 100644 --- a/arch/arm64/include/asm/kvm_pkvm.h +++ b/arch/arm64/include/asm/kvm_pkvm.h @@ -65,8 +65,7 @@ static inline bool kvm_pkvm_ioctl_allowed(struct kvm *kvm, unsigned int ioctl) int r; r = kvm_get_cap_for_kvm_ioctl(ioctl, &ext); - - if (WARN_ON_ONCE(r < 0)) + if (r < 0) return false; return kvm_pkvm_ext_allowed(kvm, ext); From 96e6757cb0674acb86ee8b558ee6bffe0eec0bb0 Mon Sep 17 00:00:00 2001 From: Sebastian Ott Date: Mon, 14 Sep 2026 15:10:13 +0200 Subject: [PATCH 093/694] KVM: selftests: fix steal_time for arm64 with host page size > 4K Fix the following failure when running with 16K host page size: ==== Test Assertion Failure ==== lib/kvm_util.c:991: vm_adjust_num_guest_pages(vm->mode, npages) == npages pid=873 tid=873 errno=0 - Success 1 0x0000000000405a27: vm_mem_add at kvm_util.c:991 2 0x000000000040241f: check_steal_time_uapi at steal_time.c:223 (discriminator 7) 3 (inlined by) main at steal_time.c:539 (discriminator 7) 4 0x00007fff8b57af3b: ?? ??:0 5 0x00007fff8b57b007: ?? ??:0 6 0x0000000000402b6f: _start at ??:? Number of guest pages is not compatible with the host. Try npages=4 Fixes: fc240715fc50 ("KVM: selftests: arm64: Fix steal_time test after UAPI refactoring") Reported-by: Zenghui Yu Link: https://lore.kernel.org/kvmarm/7575a845-a542-4b16-b512-aec3126f97f3@linux.dev/T/#u Signed-off-by: Sebastian Ott Reviewed-by: Zenghui Yu (Huawei) Link: https://patch.msgid.link/20260914131013.60334-1-sebott@redhat.com Signed-off-by: Oliver Upton --- tools/testing/selftests/kvm/steal_time.c | 30 ++++++++++++++++-------- 1 file changed, 20 insertions(+), 10 deletions(-) diff --git a/tools/testing/selftests/kvm/steal_time.c b/tools/testing/selftests/kvm/steal_time.c index bc3c62b72c58..785d19f9ee8f 100644 --- a/tools/testing/selftests/kvm/steal_time.c +++ b/tools/testing/selftests/kvm/steal_time.c @@ -27,6 +27,9 @@ static void *st_gva[NR_VCPUS]; static u64 guest_stolen_time[NR_VCPUS]; +static struct kvm_vm *vm_create_steal_time(u32 nr_vcpus, void *guest_code, + struct kvm_vcpu *vcpus[]); + #if defined(__x86_64__) /* steal_time must have 64-byte alignment */ @@ -210,17 +213,14 @@ static void check_steal_time_uapi(void) u64 st_ipa; int ret; - vm = vm_create_with_one_vcpu(&vcpu, NULL); - struct kvm_device_attr dev = { .group = KVM_ARM_VCPU_PVTIME_CTRL, .attr = KVM_ARM_VCPU_PVTIME_IPA, .addr = (u64)&st_ipa, }; + vm = vm_create_steal_time(1, NULL, &vcpu); vcpu_ioctl(vcpu, KVM_HAS_DEVICE_ATTR, &dev); - vm_userspace_mem_region_add(vm, VM_MEM_SRC_ANONYMOUS, ST_GPA_BASE, 1, 1, 0); - virt_map(vm, ST_GPA_BASE, ST_GPA_BASE, 1); st_ipa = (ulong)ST_GPA_BASE | 1; ret = __vcpu_ioctl(vcpu, KVM_SET_DEVICE_ATTR, &dev); @@ -500,13 +500,27 @@ static void run_vcpu(struct kvm_vcpu *vcpu) } } +static struct kvm_vm *vm_create_steal_time(u32 nr_vcpus, void *guest_code, + struct kvm_vcpu *vcpus[]) +{ + unsigned int gpages; + struct kvm_vm *vm; + + /* Create a VM and an identity mapped memslot for the steal time structure */ + vm = vm_create_with_vcpus(nr_vcpus, guest_code, vcpus); + gpages = vm_calc_num_guest_pages(VM_MODE_DEFAULT, STEAL_TIME_SIZE * nr_vcpus); + vm_userspace_mem_region_add(vm, VM_MEM_SRC_ANONYMOUS, ST_GPA_BASE, 1, gpages, 0); + virt_map(vm, ST_GPA_BASE, ST_GPA_BASE, gpages); + + return vm; +} + int main(int ac, char **av) { struct kvm_vcpu *vcpus[NR_VCPUS]; struct kvm_vm *vm; pthread_t thread; cpu_set_t cpuset; - unsigned int gpages; long stolen_time; long run_delay; bool verbose; @@ -517,11 +531,7 @@ int main(int ac, char **av) /* Set CPU affinity so we can force preemption of the VCPU */ cpu = pin_self_to_any_cpu(); - /* Create a VM and an identity mapped memslot for the steal time structure */ - vm = vm_create_with_vcpus(NR_VCPUS, guest_code, vcpus); - gpages = vm_calc_num_guest_pages(VM_MODE_DEFAULT, STEAL_TIME_SIZE * NR_VCPUS); - vm_userspace_mem_region_add(vm, VM_MEM_SRC_ANONYMOUS, ST_GPA_BASE, 1, gpages, 0); - virt_map(vm, ST_GPA_BASE, ST_GPA_BASE, gpages); + vm = vm_create_steal_time(NR_VCPUS, guest_code, vcpus); ksft_print_header(); TEST_REQUIRE(is_steal_time_supported(vcpus[0])); From 089e4f3c4862ba3f29dff2361caa8084879194fd Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Tue, 1 Sep 2026 22:47:00 +0100 Subject: [PATCH 094/694] KVM: arm64: Fix FGT mapping for HFGITR_EL2.nGCSEPP The encoding to trap mapping currently maps a FGT on OP_GCSPOPX to HFGITR_EL2.nGCSEPP but as per DDI0601 2026-06 this FGT controls trapping of GCSPUSHX and GCSPOPCX, and not the separate GCSPOPX instruction. Update the mapping to reflect the architecture. Fixes: 863ac38984a82 ("KVM: arm64: Add missing HFGITR_EL2 FGT entries to nested virt") Reviewed-by: Leonardo Bras Signed-off-by: Mark Brown Reviewed-by: Lorenzo Stoakes (ARM) Link: https://patch.msgid.link/20260901-arm64-gcs-v20-2-f31750bdfadb@kernel.org Signed-off-by: Oliver Upton --- arch/arm64/kvm/emulate-nested.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/arm64/kvm/emulate-nested.c b/arch/arm64/kvm/emulate-nested.c index 625604019fb3..3806ff0920fe 100644 --- a/arch/arm64/kvm/emulate-nested.c +++ b/arch/arm64/kvm/emulate-nested.c @@ -1445,7 +1445,7 @@ static const struct encoding_to_trap_config encoding_to_fgt[] __initconst = { SR_FGT(OP_AT_S1E1A, HFGITR, ATS1E1A, 1), SR_FGT(OP_COSP_RCTX, HFGITR, COSPRCTX, 1), SR_FGT(OP_GCSPUSHX, HFGITR, nGCSEPP, 0), - SR_FGT(OP_GCSPOPX, HFGITR, nGCSEPP, 0), + SR_FGT(OP_GCSPOPCX, HFGITR, nGCSEPP, 0), SR_FGT(OP_GCSPUSHM, HFGITR, nGCSPUSHM_EL1, 0), SR_FGT(OP_BRB_IALL, HFGITR, nBRBIALL, 0), SR_FGT(OP_BRB_INJ, HFGITR, nBRBINJ, 0), From 7de9a6fb44eae4f05e68c805d58b9c618815adfa Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Tue, 15 Sep 2026 12:34:44 -1000 Subject: [PATCH 095/694] sched_ext: Wait for SCX_OPSS_DISPATCHING before reenqueueing a task ebf1ccff79c4 ("sched_ext: Fix ops.dequeue() semantics") moved the final ops_state store in scx_dispatch_enqueue() after the DSQ unlock so that the custody update and ops.dequeue() precede it. A task can thus be found on a DSQ while still SCX_OPSS_DISPATCHING. The dequeue and core-sched pick paths wait for the state to clear in ops_dequeue() but the reenqueue paths don't. A reenqueue in that window runs ops.enqueue() and sets SCX_OPSS_QUEUED before the dispatch has completed. The dispatcher's final store then overwrites it with SCX_OPSS_NONE and finish_dispatch() drops every later dispatch of the task. Wait for SCX_OPSS_DISPATCHING to clear before dequeueing a task for reenqueue, the same way ops_dequeue() does. Fixes: ebf1ccff79c4 ("sched_ext: Fix ops.dequeue() semantics") Cc: stable@vger.kernel.org # v7.1+ Signed-off-by: Tejun Heo Reviewed-by: Andrea Righi --- kernel/sched/ext/ext.c | 13 +++++++++++++ kernel/sched/ext/internal.h | 1 + kernel/sched/ext/sub.c | 1 + 3 files changed, 15 insertions(+) diff --git a/kernel/sched/ext/ext.c b/kernel/sched/ext/ext.c index 70b711c4de6e..725b890fcab9 100644 --- a/kernel/sched/ext/ext.c +++ b/kernel/sched/ext/ext.c @@ -4404,6 +4404,17 @@ static bool local_task_should_reenq(struct rq *rq, struct task_struct *p, return *reenq_flags & SCX_REENQ_ANY; } +/* + * The dispatcher stores the final ops_state after dropping the DSQ lock, so @p + * can be found on a DSQ while still %SCX_OPSS_DISPATCHING. Reenqueueing @p + * before that store lands would have it clobber the new %SCX_OPSS_QUEUED. + */ +void scx_reenq_wait_dispatching(struct task_struct *p) +{ + if (unlikely(atomic_long_read_acquire(&p->scx.ops_state) == SCX_OPSS_DISPATCHING)) + wait_ops_state(p, SCX_OPSS_DISPATCHING); +} + static u32 reenq_local(struct scx_sched *sch, struct rq *rq, u64 reenq_flags) { LIST_HEAD(tasks); @@ -4447,6 +4458,7 @@ static u32 reenq_local(struct scx_sched *sch, struct rq *rq, u64 reenq_flags) if (!local_task_should_reenq(rq, p, &reenq_flags, &reason)) continue; + scx_reenq_wait_dispatching(p); scx_dispatch_dequeue(rq, p); if (WARN_ON_ONCE(p->scx.flags & SCX_TASK_REENQ_REASON_MASK)) @@ -4570,6 +4582,7 @@ static void reenq_user(struct rq *rq, struct scx_dispatch_q *dsq, u64 reenq_flag } /* @p is on @dsq, its rq and @dsq are locked */ + scx_reenq_wait_dispatching(p); dispatch_dequeue_locked(p, dsq); raw_spin_unlock(&dsq->lock); diff --git a/kernel/sched/ext/internal.h b/kernel/sched/ext/internal.h index 3464e0f113c1..115c96fbf932 100644 --- a/kernel/sched/ext/internal.h +++ b/kernel/sched/ext/internal.h @@ -2078,6 +2078,7 @@ void scx_kick_cpu(struct scx_sched *sch, s32 cpu, u64 flags); u64 __scx_bpf_now(struct rq *rq); void schedule_dsq_reenq(struct scx_sched *sch, struct scx_dispatch_q *dsq, u64 reenq_flags, struct rq *locked_rq); +void scx_reenq_wait_dispatching(struct task_struct *p); int __scx_init_task(struct scx_sched *sch, struct task_struct *p, struct cgroup *cgrp, bool fork); void scx_enable_task(struct scx_sched *sch, struct task_struct *p); diff --git a/kernel/sched/ext/sub.c b/kernel/sched/ext/sub.c index f7aeb1488566..a30c965b175d 100644 --- a/kernel/sched/ext/sub.c +++ b/kernel/sched/ext/sub.c @@ -801,6 +801,7 @@ void scx_reenq_reject(struct rq *rq) if (WARN_ON_ONCE(p->migration_pending)) continue; + scx_reenq_wait_dispatching(p); scx_dispatch_dequeue(rq, p); if (WARN_ON_ONCE(p->scx.flags & SCX_TASK_REENQ_REASON_MASK)) From b52d695d062095327b944acf7daabbc816ab319b Mon Sep 17 00:00:00 2001 From: Stanley Jhu Date: Sat, 12 Sep 2026 21:16:25 +0800 Subject: [PATCH 096/694] scsi: ufs: core: Keep internal commands dispatchable during error handling Commit 08b12cda6c44 ("scsi: ufs: core: Switch to scsi_get_internal_cmd()") switched UFS internal commands to allocate requests on hba->host->pseudo_sdev->request_queue, which shares the host tagset with regular LUNs. During error recovery, ufshcd_err_handling_prepare() calls blk_mq_quiesce_tagset(&hba->host->tag_set), marking all queues in the tagset as quiesced, including pseudo_sdev->request_queue. When ufshcd_verify_dev_init() subsequently issues internal commands (e.g. NOP OUT UPIU) via blk_execute_rq(), blk_mq_run_hw_queue() skips running the quiesced queue, resulting in an unrecoverable circular wait deadlock. Keep quiescing the tagset and unquiesce the pseudo SCSI device on top of that, so internal commands stay dispatchable while the logical units remain quiesced. Re-quiesce the pseudo device before unquiescing the tagset so that quiesce_depth stays balanced. Clock scaling and ufshcd_pause_command_processing() are unaffected: they keep quiescing the whole tagset, internal commands included. Fixes: 08b12cda6c44 ("scsi: ufs: core: Switch to scsi_get_internal_cmd()") Cc: stable@vger.kernel.org Link: https://lore.kernel.org/all/6f78c4bd-a70b-402d-abfd-599091b67674@acm.org/ Signed-off-by: Stanley Jhu Reviewed-by: Bart Van Assche Link: https://patch.msgid.link/20260912131625.2301486-1-stanleyjhu@google.com Signed-off-by: Martin K. Petersen (Oracle) --- drivers/ufs/core/ufshcd.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/drivers/ufs/core/ufshcd.c b/drivers/ufs/core/ufshcd.c index 2ba244cf40ac..54f4e7d7de02 100644 --- a/drivers/ufs/core/ufshcd.c +++ b/drivers/ufs/core/ufshcd.c @@ -6817,11 +6817,17 @@ static void ufshcd_err_handling_prepare(struct ufs_hba *hba) } /* Wait for ongoing ufshcd_queuecommand() calls to finish. */ blk_mq_quiesce_tagset(&hba->host->tag_set); + /* + * Internal commands are submitted on the pseudo SCSI device. Let them + * through so that the error handler can recover the link. + */ + blk_mq_unquiesce_queue(hba->host->pseudo_sdev->request_queue); cancel_work_sync(&hba->eeh_work); } static void ufshcd_err_handling_unprepare(struct ufs_hba *hba) { + blk_mq_quiesce_queue_nowait(hba->host->pseudo_sdev->request_queue); blk_mq_unquiesce_tagset(&hba->host->tag_set); ufshcd_release(hba); if (ufshcd_is_clkscaling_supported(hba)) From c9ee6511332687ea714ad8ab86a53cb837d86eea Mon Sep 17 00:00:00 2001 From: Geert Uytterhoeven Date: Mon, 14 Sep 2026 16:00:01 +0200 Subject: [PATCH 097/694] scsi: ufs: pltfrm: Add quirk for R-Car S4 lacking lanes-per-direction Since commit e72323f3b09f ("scsi: ufs: core: Configure only active lanes during link"), the following error is observed on R-Car S4: ufshcd-renesas e6860000.ufs: Tx lane mismatch [config,reported] [2,1] ufshcd-renesas e6860000.ufs: link startup failed -67 ufshcd-renesas e6860000.ufs: error -ENOLINK: Initialization failed with error -67 ufshcd-renesas e6860000.ufs: probe with driver ufshcd-renesas failed with error -67 R-Car S4 has one UFS lane per direction, as described in section 152.1 of its hardware manual. Without lanes-per-direction, the UFS platform driver defaults to two lanes. Previously, the core used PA_CONNECTEDRXDATALANES and PA_CONNECTEDTXDATALANES to configure the link without checking them against lanes-per-direction, so the missing property did not prevent initialization. While fixing the R-Car S4 DTS is the proper solution, doing only that would still break backwards compatibility with existing DTBs. Hence add a quirk to let lanes-per-direction default to one on R-Car S4. Fixes: e72323f3b09f9c89 ("scsi: ufs: core: Configure only active lanes during link") Reported-by: Koichiro Den Closes: https://lore.kernel.org/20260911073058.253000-1-den@valinux.co.jp Cc: stable@vger.kernel.org # 7.2+ Signed-off-by: Geert Uytterhoeven Link: https://patch.msgid.link/ae0cc2bd764e6dfffce99db3d8b44a55887c508c.1789394185.git.geert+renesas@glider.be Signed-off-by: Martin K. Petersen (Oracle) --- drivers/ufs/host/ufshcd-pltfrm.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/drivers/ufs/host/ufshcd-pltfrm.c b/drivers/ufs/host/ufshcd-pltfrm.c index 5ac7afe75934..169dbc1a75fe 100644 --- a/drivers/ufs/host/ufshcd-pltfrm.c +++ b/drivers/ufs/host/ufshcd-pltfrm.c @@ -206,7 +206,11 @@ static void ufshcd_init_lanes_per_dir(struct ufs_hba *hba) dev_dbg(hba->dev, "%s: failed to read lanes-per-direction, ret=%d\n", __func__, ret); - hba->lanes_per_direction = UFSHCD_DEFAULT_LANES_PER_DIRECTION; + /* Old R-Car S4 DTBs lack "lanes-per-direction = <1>" */ + if (of_device_is_compatible(dev->of_node, "renesas,r8a779f0-ufs")) + hba->lanes_per_direction = 1; + else + hba->lanes_per_direction = UFSHCD_DEFAULT_LANES_PER_DIRECTION; } } From bce07e2f37b5e4a427d36fd6b1c14067b27591db Mon Sep 17 00:00:00 2001 From: Yehyeong Lee Date: Sat, 1 Aug 2026 22:36:35 +0900 Subject: [PATCH 098/694] scsi: libiscsi_tcp: Check the data direction of a Data-In PDU The Data-In branch of iscsi_tcp_hdr_dissect() resolves the ITT to a task and copies the PDU's data segment into that command's scatterlist without asking whether the command was reading. iscsi_tcp_r2t_rsp() in the same file does ask, and rejects an R2T for a command that is not DMA_TO_DEVICE. A target that answers a WRITE command's ITT with a Data-In therefore has the initiator write target-supplied bytes into the pages that write was about to send. Those are the caller's own pinned pages for an O_DIRECT write, and page cache pages for a buffered one. Observed against a test target that emits one 512-byte Data-In naming a 128 KB write's ITT, after the R2T for that write. With O_DIRECT the caller's buffer ends up holding 512 bytes of the target's data while pwrite() returns 131072. Buffered is quieter: pwrite() and fsync() both succeed, nothing is logged, and reading those blocks back returns the target's bytes out of the page cache without a command going on the wire. Check the direction before using the scatterlist, the way the R2T path already does. Cc: stable@vger.kernel.org Signed-off-by: Yehyeong Lee Reviewed-by: Mike Christie Link: https://patch.msgid.link/20260801133635.1986706-1-yhlee@isslab.korea.ac.kr Fixes: a081c13e39b5 ("[SCSI] iscsi_tcp: split module into lib and lld") Signed-off-by: Martin K. Petersen (Oracle) --- drivers/scsi/libiscsi_tcp.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/scsi/libiscsi_tcp.c b/drivers/scsi/libiscsi_tcp.c index 7223bb18b048..d35f93451ee9 100644 --- a/drivers/scsi/libiscsi_tcp.c +++ b/drivers/scsi/libiscsi_tcp.c @@ -480,6 +480,9 @@ static int iscsi_tcp_data_in(struct iscsi_conn *conn, struct iscsi_task *task) int datasn = be32_to_cpu(rhdr->datasn); unsigned total_in_length = task->sc->sdb.length; + if (task->sc->sc_data_direction != DMA_FROM_DEVICE) + return ISCSI_ERR_PROTO; + /* * lib iscsi will update this in the completion handling if there * is status. From f06a44e235ef188689ba23ffc72e9e89b10951a9 Mon Sep 17 00:00:00 2001 From: "Ewan D. Milne" Date: Tue, 15 Sep 2026 13:38:00 -0400 Subject: [PATCH 099/694] scsi: devinfo: Add BLIST_SKIP_IO_HINTS for EMC Symmetrix EMC Symmetrix returns an error on MODE SENSE for page 0Ah subpage 05h because it does not implement the SBC-5 I/O hints. These commands began to be sent as a result of commit 4f53138fffc2 ("scsi: sd: Translate data lifetime information"). Add BLIST_SKIP_IO_HINTS to avoid sending these commands because in large configurations the failed commands are displacing other useful information in internal error logs. Signed-off-by: Ewan D. Milne Link: https://patch.msgid.link/20260915173800.39117-1-emilne@redhat.com Signed-off-by: Martin K. Petersen (Oracle) --- drivers/scsi/scsi_devinfo.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/scsi/scsi_devinfo.c b/drivers/scsi/scsi_devinfo.c index 15ffbe93ac72..88a911b35c94 100644 --- a/drivers/scsi/scsi_devinfo.c +++ b/drivers/scsi/scsi_devinfo.c @@ -161,7 +161,7 @@ static struct { {"DGC", "DISK", NULL, BLIST_SPARSELUN}, /* EMC CLARiiON, no storage on LUN 0 */ {"EMC", "Invista", "*", BLIST_SPARSELUN | BLIST_LARGELUN}, {"EMC", "SYMMETRIX", NULL, BLIST_SPARSELUN | BLIST_LARGELUN | - BLIST_REPORTLUN2 | BLIST_RETRY_ITF}, + BLIST_REPORTLUN2 | BLIST_RETRY_ITF | BLIST_SKIP_IO_HINTS}, {"EMULEX", "MD21/S2 ESDI", NULL, BLIST_SINGLELUN}, {"easyRAID", "16P", NULL, BLIST_NOREPORTLUN}, {"easyRAID", "X6P", NULL, BLIST_NOREPORTLUN}, From 278210c60c6f6958bd2eeaa2120c862683b83d09 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Tue, 15 Sep 2026 22:20:59 +0200 Subject: [PATCH 100/694] scsi: leapraid: Avoid -Wformat-security warning When extra warnings are enabled, the alloc_ordered_workqueue() function cannot be called with a variable name for the format string: drivers/scsi/leapraid/leapraid_os.c: In function 'leapraid_probe': drivers/scsi/leapraid/leapraid_os.c:2062:58: error: format not a string literal and no format arguments [-Werror=format-security] 2062 | alloc_ordered_workqueue(adapter->fw_evt_s.fw_evt_name, 0); | ~~~~~~~~~~~~~~~~~^~~~~~~~~~~~ As the string is only assembled before the call and not used anywhere else, just fold the earlier snprintf() into the alloc_ordered_workqueue() call. Fixes: 5597088c9e79 ("scsi: leapraid: Add new SCSI driver") Signed-off-by: Arnd Bergmann Reviewed-by: Damien Le Moal Link: https://patch.msgid.link/20260915202134.3534708-1-arnd@kernel.org Signed-off-by: Martin K. Petersen (Oracle) --- drivers/scsi/leapraid/leapraid_func.h | 2 -- drivers/scsi/leapraid/leapraid_os.c | 8 +++----- 2 files changed, 3 insertions(+), 7 deletions(-) diff --git a/drivers/scsi/leapraid/leapraid_func.h b/drivers/scsi/leapraid/leapraid_func.h index 4c0b9ca728d8..e8a0815bf95f 100644 --- a/drivers/scsi/leapraid/leapraid_func.h +++ b/drivers/scsi/leapraid/leapraid_func.h @@ -554,7 +554,6 @@ struct leapraid_fw_evt_work { /** * struct leapraid_fw_evt_struct - Firmware event handling structure * - * @fw_evt_name: Name of the firmware event. * @fw_evt_thread: Workqueue used for processing firmware events. * @fw_evt_lock: Spinlock protecting access to the firmware event list. * @fw_evt_list: Linked list of pending firmware events. @@ -565,7 +564,6 @@ struct leapraid_fw_evt_work { */ struct leapraid_fw_evt_struct { u32 leapraid_evt_masks[4]; - char fw_evt_name[48]; struct workqueue_struct *fw_evt_thread; spinlock_t fw_evt_lock; /* protects firmware event */ struct list_head fw_evt_list; diff --git a/drivers/scsi/leapraid/leapraid_os.c b/drivers/scsi/leapraid/leapraid_os.c index ee3242779dfd..507f11862276 100644 --- a/drivers/scsi/leapraid/leapraid_os.c +++ b/drivers/scsi/leapraid/leapraid_os.c @@ -2054,12 +2054,10 @@ static int leapraid_probe(struct pci_dev *pdev, const struct pci_device_id *id) shost->transportt = leapraid_transport_template; shost->unique_id = adapter->adapter_attr.id; - snprintf(adapter->fw_evt_s.fw_evt_name, - sizeof(adapter->fw_evt_s.fw_evt_name), - "fw_event_%s%d", LEAPRAID_DRIVER_NAME, - adapter->adapter_attr.id); adapter->fw_evt_s.fw_evt_thread = - alloc_ordered_workqueue(adapter->fw_evt_s.fw_evt_name, 0); + alloc_ordered_workqueue("fw_event_%s%d", 0, + LEAPRAID_DRIVER_NAME, + adapter->adapter_attr.id); if (!adapter->fw_evt_s.fw_evt_thread) { dev_err(&adapter->pdev->dev, "%s: Failed to create fw event workqueue\n", __func__); From 1f7745fb3580152ca902ef181b605f33cabfb1d0 Mon Sep 17 00:00:00 2001 From: Ran Hongyun Date: Mon, 13 Jul 2026 19:55:25 +0800 Subject: [PATCH 101/694] squashfs: Add dictionary size range check to prevent shift-out-of-bounds When an abnormal SquashFS image (COMP_OPTS flag is 1 but dictionary size is 0) is mounted, and performs shift operations using dictionarysize, the shift exponent is -1, causing a shift-out-of-bounds. Detail as below: squashfs_comp_opts(msblk, buffer, length) squashfs_xz_comp_opts() if (comp_opts) n = ffs(opts->dict_size) - 1;<----opts->dict_size=0, n=-1 if (opts->dict_size != (1 << n) && opts->dict_size != (1 << n) + (1 << (n + 1))) <----shift-out-of-bounds Fix it by adding a dictionary size range check before the shift operation. Fixes: ff750311d30a ("Squashfs: add compression options support to xz decompressor") Signed-off-by: Ran Hongyun Link: https://patch.msgid.link/20260713115525.2661734-1-ranhongyun1@huawei.com Reviewed-by: Phillip Lougher Reviewed-by: Zhihao Cheng Signed-off-by: Christian Brauner (Amutable) --- fs/squashfs/xz_wrapper.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/fs/squashfs/xz_wrapper.c b/fs/squashfs/xz_wrapper.c index 0a4ff3ec9c8c..6610af241449 100644 --- a/fs/squashfs/xz_wrapper.c +++ b/fs/squashfs/xz_wrapper.c @@ -57,10 +57,10 @@ static void *squashfs_xz_comp_opts(struct squashfs_sb_info *msblk, opts->dict_size = le32_to_cpu(comp_opts->dictionary_size); - /* the dictionary size should be 2^n or 2^n+2^(n+1) */ + /* the dictionary size should be positive and 2^n or 2^n+2^(n+1) */ n = ffs(opts->dict_size) - 1; - if (opts->dict_size != (1 << n) && opts->dict_size != (1 << n) + - (1 << (n + 1))) { + if (opts->dict_size <= 0 || (opts->dict_size != (1 << n) && + opts->dict_size != (1 << n) + (1 << (n + 1)))) { err = -EIO; goto out; } From a13f7f5d14af9baf34eb12c25962f8d3542b281d Mon Sep 17 00:00:00 2001 From: Ilya Titov Date: Thu, 3 Sep 2026 12:17:18 +0300 Subject: [PATCH 102/694] pinctrl: sunxi: keep a shadow copy of the data register output latches MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit On Allwinner SoCs, reading a bank's data register returns the pin level, not the output latch, for pins that are muxed as inputs. Writing a GPIO therefore corrupts the output latches of all input-muxed pins in the same bank: the read-modify-write in sunxi_pinctrl_gpio_set() reads back their pin levels and writes those into their latches. This breaks emulated open-drain lines (e.g. a bit-banged I2C bus from i2c-gpio). Such a line is released high by muxing it as input and letting the pull-up raise it, so any concurrent GPIO write in the same bank stores 1 into its latch. Driving the line low afterwards is a non-atomic data-then-mux sequence in sunxi_pinctrl_gpio_direction_output(); if the poisoning write lands between the two steps, the pin actively drives high (push-pull) instead of low. Observed in practice as sporadic glitches on a T507 board bit-banging I2C on port E while other PE GPIOs are toggled. On a scope the failure is unmistakable: on a clock pulse where SCL should fall to GND, the line instead steps *above* its idle high level for the whole low phase — the pad drives a strong push-pull 3.3 V high, higher than the level the pull-up sustains on the loaded bus — before the next transition recovers it. The same can hit SDA, corrupting data instead of clocks. Steps to reproduce on any sunxi board with a bit-banged (i2c-gpio) bus: # background: toggle any other GPIO of the same bank, e.g. line 21 gpioset -c --toggle 100us 21=0 & # foreground: keep the bit-banged bus busy while :; do i2cdetect -y 0x50 0x57; done # watch SCL/SDA with a scope or logic analyzer: sporadic clock-low # phases driven high (above the pull-up level) instead of low The bank spinlock cannot help: the racing write is a perfectly valid whole-register RMW that faithfully writes back what the hardware returned. There are no set/clear registers on this IP to write a single bit atomically. Fix it the same way gpio-mmio handles hardware whose data register read does not return the output latch: keep a shadow copy of each bank's latches, base the read-modify-write on the shadow, and only write the register. The shadow is seeded from the hardware at probe time so pins left in output mode by the bootloader keep their state. Pins that reach output mode through the gpiolib paths write their value (and thereby their shadow bit) before the mux switch in sunxi_pinctrl_gpio_direction_output(); pins muxed to gpio_out directly through a pinmux node bypass that path, so sunxi_pmx_set() refreshes their shadow bit from the latch (readable once the pin is in output mode) to keep them driving their pre-existing level. Seeding the shadow reads the PIO registers at probe time, which requires the bus clock to be enabled. The clock was only requested at the very end of probe, after devm_pinctrl_register() had already claimed the pin hogs described in the device tree - which mux pins, and thus access registers, with the clock still gated. Move the request ahead of both. Boards whose bootloader leaves the PIO clock running are unaffected, which is why the pre-existing hog problem has gone unnoticed since commit 950707c0eb5c ("pinctrl: sunxi: add clock support"). Fixes: df7b34f4c3d2 ("pinctrl: sunxi: Fix gpio_set behaviour") Cc: stable@vger.kernel.org Signed-off-by: Ilya Titov Signed-off-by: Linus Walleij --- drivers/pinctrl/sunxi/pinctrl-sunxi.c | 76 +++++++++++++++++++++------ drivers/pinctrl/sunxi/pinctrl-sunxi.h | 7 +++ 2 files changed, 68 insertions(+), 15 deletions(-) diff --git a/drivers/pinctrl/sunxi/pinctrl-sunxi.c b/drivers/pinctrl/sunxi/pinctrl-sunxi.c index 25489beeb312..2881a83be99c 100644 --- a/drivers/pinctrl/sunxi/pinctrl-sunxi.c +++ b/drivers/pinctrl/sunxi/pinctrl-sunxi.c @@ -837,6 +837,21 @@ static void sunxi_pmx_set(struct pinctrl_dev *pctldev, writel((readl(pctl->membase + reg) & ~mask) | config << shift, pctl->membase + reg); + /* + * A pin muxed to gpio_out directly through a pinmux node bypasses + * sunxi_pinctrl_gpio_set() and drives whatever its output latch + * holds. Now that the pin is in output mode the data register + * reads back the latch, so refresh the shadow to keep such pins + * driving their pre-existing level. + */ + if (config == SUN4I_FUNC_OUTPUT) { + u32 *shadow = &pctl->dat_shadow[pin / PINS_PER_BANK]; + + sunxi_data_reg(pctl, pin, ®, &shift, &mask); + *shadow = (*shadow & ~mask) | + (readl(pctl->membase + reg) & mask); + } + raw_spin_unlock_irqrestore(&pctl->lock, flags); } @@ -1017,21 +1032,29 @@ static int sunxi_pinctrl_gpio_set(struct gpio_chip *chip, unsigned int offset, int value) { struct sunxi_pinctrl *pctl = gpiochip_get_data(chip); - u32 reg, shift, mask, val; + u32 *shadow = &pctl->dat_shadow[offset / PINS_PER_BANK]; + u32 reg, shift, mask; unsigned long flags; sunxi_data_reg(pctl, offset, ®, &shift, &mask); raw_spin_lock_irqsave(&pctl->lock, flags); - val = readl(pctl->membase + reg); - + /* + * Reading the data register returns the pin level, not the output + * latch, for pins muxed as inputs. A read-modify-write based on + * the register would therefore corrupt the latches of input-muxed + * pins in the same bank (e.g. an emulated open-drain I2C line + * released high), making them drive the wrong level once switched + * to output. Base the read-modify-write on a shadow copy of the + * latches instead. + */ if (value) - val |= mask; + *shadow |= mask; else - val &= ~mask; + *shadow &= ~mask; - writel(val, pctl->membase + reg); + writel(*shadow, pctl->membase + reg); raw_spin_unlock_irqrestore(&pctl->lock, flags); @@ -1572,7 +1595,7 @@ int sunxi_pinctrl_init_with_flags(struct platform_device *pdev, struct pinctrl_pin_desc *pins; struct sunxi_pinctrl *pctl; struct pinmux_ops *pmxops; - int i, ret, last_pin, pin_idx; + int i, ret, last_pin, pin_idx, nbanks; struct clk *clk; pctl = devm_kzalloc(&pdev->dev, sizeof(*pctl), GFP_KERNEL); @@ -1610,6 +1633,37 @@ int sunxi_pinctrl_init_with_flags(struct platform_device *pdev, if (!pctl->irq_array) return -ENOMEM; + /* + * The bus clock has to be enabled before the pinctrl device + * registers, as the pin hogs claimed from there access registers. + */ + ret = of_clk_get_parent_count(node); + clk = devm_clk_get_enabled(&pdev->dev, ret == 1 ? NULL : "apb"); + if (IS_ERR(clk)) + return PTR_ERR(clk); + + /* + * Seed the output latch shadow from the hardware so pins the + * bootloader left in output mode keep their state; see + * sunxi_pinctrl_gpio_set() for why a shadow is needed. This must + * happen before the pinctrl device registers, as pin hogs can mux + * pins to gpio_out and thereby update the shadow. + */ + last_pin = pctl->desc->pins[pctl->desc->npins - 1].pin.number; + nbanks = DIV_ROUND_UP(last_pin + 1 - pctl->desc->pin_base, + PINS_PER_BANK); + pctl->dat_shadow = devm_kcalloc(&pdev->dev, nbanks, + sizeof(*pctl->dat_shadow), GFP_KERNEL); + if (!pctl->dat_shadow) + return -ENOMEM; + + for (i = 0; i < nbanks; i++) { + u32 reg, shift, mask; + + sunxi_data_reg(pctl, i * PINS_PER_BANK, ®, &shift, &mask); + pctl->dat_shadow[i] = readl(pctl->membase + reg); + } + ret = sunxi_pinctrl_build_state(pdev); if (ret) { dev_err(&pdev->dev, "dt probe failed: %d\n", ret); @@ -1665,7 +1719,6 @@ int sunxi_pinctrl_init_with_flags(struct platform_device *pdev, if (!pctl->chip) return -ENOMEM; - last_pin = pctl->desc->pins[pctl->desc->npins - 1].pin.number; pctl->chip->owner = THIS_MODULE; pctl->chip->request = gpiochip_generic_request; pctl->chip->free = gpiochip_generic_free; @@ -1699,13 +1752,6 @@ int sunxi_pinctrl_init_with_flags(struct platform_device *pdev, goto gpiochip_error; } - ret = of_clk_get_parent_count(node); - clk = devm_clk_get_enabled(&pdev->dev, ret == 1 ? NULL : "apb"); - if (IS_ERR(clk)) { - ret = PTR_ERR(clk); - goto gpiochip_error; - } - pctl->irq = devm_kcalloc(&pdev->dev, pctl->desc->irq_banks, sizeof(*pctl->irq), diff --git a/drivers/pinctrl/sunxi/pinctrl-sunxi.h b/drivers/pinctrl/sunxi/pinctrl-sunxi.h index 0daf7600e2fb..498e88a19f71 100644 --- a/drivers/pinctrl/sunxi/pinctrl-sunxi.h +++ b/drivers/pinctrl/sunxi/pinctrl-sunxi.h @@ -85,6 +85,7 @@ #define IO_BIAS_MASK GENMASK(3, 0) #define SUN4I_FUNC_INPUT 0 +#define SUN4I_FUNC_OUTPUT 1 #define SUN4I_FUNC_IRQ 6 #define SUN4I_FUNC_DISABLED_OLD 7 #define SUN4I_FUNC_DISABLED_NEW 15 @@ -175,6 +176,12 @@ struct sunxi_pinctrl { int *irq; unsigned *irq_array; raw_spinlock_t lock; + /* + * Output latch shadow, one word per bank. Seeded lockless at + * probe before the pinctrl device registers, protected by @lock + * afterwards. + */ + u32 *dat_shadow; struct pinctrl_dev *pctl_dev; unsigned long flags; u32 bank_mem_size; From 7c431d61b69a3fd0784c20aa4cd0b8fb501b5653 Mon Sep 17 00:00:00 2001 From: ZHOU Jiaxiang Date: Wed, 16 Sep 2026 21:58:21 +0800 Subject: [PATCH 103/694] scsi: block: Fix zones_cond out-of-bounds write on zone report blk_revalidate_disk_zones() sizes the zones_cond array from the disk capacity and zone size, but the index used by blk_revalidate_zone_cond() comes from the device-driven report_zones() walk and is never checked against the array size. A device reporting more zones than fit the array makes blk_zone_set_cond() write out of bounds. One way to reach this is a zone count exceeding 32 bits: both blk_revalidate_zone_args.nr_zones and struct zoned_disk_info.nr_zones are unsigned int, so a disk advertising more than UINT_MAX zones (e.g. 2^32 + 1024 zones of one 512-byte logical block) gets its zone count truncated to a small value, undersizing the array while the report walk keeps counting upward. Check the index against the array size before storing the zone condition, and refuse to revalidate when the zone count does not fit 32 bits. Fixes: 6e945ffb6555 ("block: use zone condition to determine conventional zones") Signed-off-by: ZHOU Jiaxiang Reviewed-by: Damien Le Moal Link: https://patch.msgid.link/7815D1B293A8F55E+20260916135822.32584-2-me@fxti.xyz Signed-off-by: Martin K. Petersen (Oracle) --- block/blk-zoned.c | 15 +++++++++++++-- 1 file changed, 13 insertions(+), 2 deletions(-) diff --git a/block/blk-zoned.c b/block/blk-zoned.c index a5afb842bf35..475aa16bc41a 100644 --- a/block/blk-zoned.c +++ b/block/blk-zoned.c @@ -2018,12 +2018,17 @@ static int disk_revalidate_zone_resources(struct gendisk *disk, struct blk_revalidate_zone_args *args) { struct queue_limits *lim = &disk->queue->limits; + unsigned long long nr_zones; unsigned int pool_size; int ret = 0; args->disk = disk; - args->nr_zones = - DIV_ROUND_UP_ULL(get_capacity(disk), lim->chunk_sectors); + nr_zones = DIV_ROUND_UP_ULL(get_capacity(disk), lim->chunk_sectors); + if (nr_zones > UINT_MAX) { + pr_warn("%s: Too many zones (%llu)\n", disk->disk_name, nr_zones); + return -EINVAL; + } + args->nr_zones = nr_zones; /* Cached zone conditions: 1 byte per zone */ args->zones_cond = kzalloc(args->nr_zones, GFP_NOIO); @@ -2131,6 +2136,12 @@ static int blk_revalidate_zone_cond(struct blk_zone *zone, unsigned int idx, { enum blk_zone_cond cond = zone->cond; + if (idx >= args->nr_zones) { + pr_warn("%s: Zone report index %u exceeds zone count %u\n", + args->disk->disk_name, idx, args->nr_zones); + return -EINVAL; + } + /* Check that the zone condition is consistent with the zone type. */ switch (cond) { case BLK_ZONE_COND_NOT_WP: From b6ec0f79745967c751c85df373062c8d15e45fc4 Mon Sep 17 00:00:00 2001 From: ZHOU Jiaxiang Date: Wed, 16 Sep 2026 21:58:22 +0800 Subject: [PATCH 104/694] scsi: sd_zbc: Reject disks with too many zones sd_zbc_read_zones() computes the number of zones with 64-bit arithmetic and stores the result in the unsigned int nr_zones field of struct zoned_disk_info, silently truncating counts that exceed 32 bits. The truncated count is later used to size per-zone resources, while the device may still report more zones than fit. Moreover, sd_zbc_report_zones() counts the reported zones with a signed int zone_idx, which overflows past INT_MAX. Reject devices reporting more than INT_MAX zones at scan time; such a device is not realistic for any medium that exists today, and accepting it produces inconsistent zone bookkeeping. Fixes: 89d947561077 ("sd: Implement support for ZBC devices") Signed-off-by: ZHOU Jiaxiang Reviewed-by: Damien Le Moal Link: https://patch.msgid.link/C41798AB5AA6BF2B+20260916135822.32584-3-me@fxti.xyz Signed-off-by: Martin K. Petersen (Oracle) --- drivers/scsi/sd_zbc.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/drivers/scsi/sd_zbc.c b/drivers/scsi/sd_zbc.c index 56e455fb5add..456beaf2e769 100644 --- a/drivers/scsi/sd_zbc.c +++ b/drivers/scsi/sd_zbc.c @@ -589,7 +589,7 @@ int sd_zbc_revalidate_zones(struct scsi_disk *sdkp) int sd_zbc_read_zones(struct scsi_disk *sdkp, struct queue_limits *lim, u8 buf[SD_BUF_SIZE]) { - unsigned int nr_zones; + u64 nr_zones; u32 zone_blocks = 0; int ret; @@ -621,6 +621,12 @@ int sd_zbc_read_zones(struct scsi_disk *sdkp, struct queue_limits *lim, goto err; nr_zones = round_up(sdkp->capacity, zone_blocks) >> ilog2(zone_blocks); + if (nr_zones > INT_MAX) { + sd_printk(KERN_ERR, sdkp, "Too many zones (%llu)\n", + nr_zones); + ret = -EINVAL; + goto err; + } sdkp->early_zone_info.nr_zones = nr_zones; sdkp->early_zone_info.zone_blocks = zone_blocks; From ef56085dfd1df3a53ecce8a4ff440cf6d67f430d Mon Sep 17 00:00:00 2001 From: Andre Przywara Date: Mon, 14 Sep 2026 12:07:44 +0200 Subject: [PATCH 105/694] pinctrl: sunxi: A523: fix voltage withstand encoding The Allwinner A523 uses the same GPIO voltage "withstand" programming (setting the input level voltage thresholds) as the previous SoCs, but for some odd reason inverts the encoding of 1.8V vs. 3.3V. Add a new bias voltage type to note this difference, and select it for the A523. At the same time also use the newer "CTL" version, which in addition allows to turn off the withstand programming for I/O voltages other than exact 1.8V or 3.3V (for instance for 2.5V sometimes used for Ethernet PHYs). The A523 has that enable register, but didn't use it so far. This fixes eMMC and reportedly Ethernet operation on some A523 boards. Fixes: 648be4cd9517 ("pinctrl: sunxi: Add support for the Allwinner A523") Signed-off-by: Andre Przywara Tested-by: Per Larsson Tested-by: Juan Manuel Lopez Carrillo Reviewed-by: Chen-Yu Tsai Tested-by: Chen-Yu Tsai # Fixes eMMC on Orange Pi 4A Signed-off-by: Linus Walleij --- drivers/pinctrl/sunxi/pinctrl-sun55i-a523-r.c | 2 +- drivers/pinctrl/sunxi/pinctrl-sun55i-a523.c | 2 +- drivers/pinctrl/sunxi/pinctrl-sunxi.c | 6 ++++++ drivers/pinctrl/sunxi/pinctrl-sunxi.h | 2 ++ 4 files changed, 10 insertions(+), 2 deletions(-) diff --git a/drivers/pinctrl/sunxi/pinctrl-sun55i-a523-r.c b/drivers/pinctrl/sunxi/pinctrl-sun55i-a523-r.c index 462aa1c4a5fa..e27e4945def2 100644 --- a/drivers/pinctrl/sunxi/pinctrl-sun55i-a523-r.c +++ b/drivers/pinctrl/sunxi/pinctrl-sun55i-a523-r.c @@ -26,7 +26,7 @@ static const u8 a523_r_irq_bank_muxes[SUNXI_PINCTRL_MAX_BANKS] = static struct sunxi_pinctrl_desc a523_r_pinctrl_data = { .irq_banks = ARRAY_SIZE(a523_r_irq_bank_map), .irq_bank_map = a523_r_irq_bank_map, - .io_bias_cfg_variant = BIAS_VOLTAGE_PIO_POW_MODE_SEL, + .io_bias_cfg_variant = BIAS_VOLTAGE_PIO_POW_MODE_CTL_INV, .pin_base = PL_BASE, }; diff --git a/drivers/pinctrl/sunxi/pinctrl-sun55i-a523.c b/drivers/pinctrl/sunxi/pinctrl-sun55i-a523.c index b6f78f1f30ac..88d8acd5bc24 100644 --- a/drivers/pinctrl/sunxi/pinctrl-sun55i-a523.c +++ b/drivers/pinctrl/sunxi/pinctrl-sun55i-a523.c @@ -26,7 +26,7 @@ static const u8 a523_irq_bank_muxes[SUNXI_PINCTRL_MAX_BANKS] = static struct sunxi_pinctrl_desc a523_pinctrl_data = { .irq_banks = ARRAY_SIZE(a523_irq_bank_map), .irq_bank_map = a523_irq_bank_map, - .io_bias_cfg_variant = BIAS_VOLTAGE_PIO_POW_MODE_SEL, + .io_bias_cfg_variant = BIAS_VOLTAGE_PIO_POW_MODE_CTL_INV, }; static int a523_pinctrl_probe(struct platform_device *pdev) diff --git a/drivers/pinctrl/sunxi/pinctrl-sunxi.c b/drivers/pinctrl/sunxi/pinctrl-sunxi.c index 2881a83be99c..31cd142ce0f7 100644 --- a/drivers/pinctrl/sunxi/pinctrl-sunxi.c +++ b/drivers/pinctrl/sunxi/pinctrl-sunxi.c @@ -728,6 +728,7 @@ static int sunxi_pinctrl_set_io_bias_cfg(struct sunxi_pinctrl *pctl, { unsigned short bank; unsigned long flags; + bool inverted = false; u32 val, reg; int uV; @@ -766,6 +767,9 @@ static int sunxi_pinctrl_set_io_bias_cfg(struct sunxi_pinctrl *pctl, reg &= ~IO_BIAS_MASK; writel(reg | val, pctl->membase + sunxi_grp_config_reg(pin)); return 0; + case BIAS_VOLTAGE_PIO_POW_MODE_CTL_INV: + inverted = true; + fallthrough; case BIAS_VOLTAGE_PIO_POW_MODE_CTL: val = uV > 1800000 && uV <= 2500000 ? BIT(bank) : 0; @@ -780,6 +784,8 @@ static int sunxi_pinctrl_set_io_bias_cfg(struct sunxi_pinctrl *pctl, fallthrough; case BIAS_VOLTAGE_PIO_POW_MODE_SEL: val = uV <= 1800000 ? 1 : 0; + if (inverted) + val = !val; raw_spin_lock_irqsave(&pctl->lock, flags); reg = readl(pctl->membase + pctl->pow_mod_sel_offset); diff --git a/drivers/pinctrl/sunxi/pinctrl-sunxi.h b/drivers/pinctrl/sunxi/pinctrl-sunxi.h index 498e88a19f71..8bd00c6ff628 100644 --- a/drivers/pinctrl/sunxi/pinctrl-sunxi.h +++ b/drivers/pinctrl/sunxi/pinctrl-sunxi.h @@ -117,8 +117,10 @@ enum sunxi_desc_bias_voltage { * Bias voltage is set through PIO_POW_MOD_SEL_REG * and PIO_POW_MOD_CTL_REG register, as seen on * A100 and D1 SoC, for example. + * Some SoCs invert the encoding for 1.8V vs. 3.3V. */ BIAS_VOLTAGE_PIO_POW_MODE_CTL, + BIAS_VOLTAGE_PIO_POW_MODE_CTL_INV, }; struct sunxi_desc_function { From 1d9bb9c870632adea249cfdec49ac37e6f069164 Mon Sep 17 00:00:00 2001 From: Myeonghun Pak Date: Sun, 13 Sep 2026 00:03:12 -0400 Subject: [PATCH 106/694] pinctrl: single: free the IRQ on domain creation failure pcs_irq_init_chained_handler() requests a shared IRQ on affected SoCs, but its domain creation failure path only removes a chained handler. That does not release the action installed by request_irq(). The probe can continue without interrupt support while leaving the shared IRQ action registered. Use pcs_irq_free() to undo the appropriate type of handler registration. At this point pcs->domain is NULL, so the helper only releases the parent IRQ handler. Then mark the IRQ invalid, as the other initialization error paths already do, to prevent another release from a later probe unwind or remove. This issue was identified during our ongoing static-analysis research while reviewing kernel code. Fixes: 3e6cee1786a1 ("pinctrl: single: Add support for wake-up interrupts") Cc: stable@vger.kernel.org Assisted-by: LLM Co-developed-by: Ijae Kim Signed-off-by: Ijae Kim Signed-off-by: Myeonghun Pak Signed-off-by: Linus Walleij --- drivers/pinctrl/pinctrl-single.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/pinctrl/pinctrl-single.c b/drivers/pinctrl/pinctrl-single.c index 4d5f85b7e6bb..e0eb7240a985 100644 --- a/drivers/pinctrl/pinctrl-single.c +++ b/drivers/pinctrl/pinctrl-single.c @@ -1628,7 +1628,8 @@ static int pcs_irq_init_chained_handler(struct pcs_device *pcs, &pcs_irqdomain_ops, pcs_soc); if (!pcs->domain) { - irq_set_chained_handler(pcs_soc->irq, NULL); + pcs_irq_free(pcs); + pcs_soc->irq = -1; return -EINVAL; } From f033482d76a9f18080c7a40c5f9c678bd7adc8f3 Mon Sep 17 00:00:00 2001 From: Christiano Amora Date: Wed, 16 Sep 2026 10:46:22 -0300 Subject: [PATCH 107/694] Bluetooth: SMP: reject Security Request over BR/EDR Bose QC Ultra Headphones (dual-mode, same public address on both transports) occasionally send an SMP Security Request on the BR/EDR SMP fixed channel right after the ACL link is encrypted. The kernel handles it as if it were an LE link: smp_cmd_security_req() has no transport check, smp_ltk_encrypt() looks up an LTK with the ACL connection's dst_type, and hci_find_ltk() matches the peer's LE LTK because the LE public address type is stored as ADDR_LE_DEV_PUBLIC (0), the same value as BDADDR_BREDR. HCI_OP_LE_START_ENC is then issued on the ACL handle, the controller rejects it with Invalid HCI Command Parameters, and hci_cs_le_start_enc() disconnects the link with HCI_ERROR_AUTH_FAILURE. The headphones drop within a second of connecting, before any profile is up; a manual reconnect works. btmon (MediaTek MT7922, kernel 7.0.12): > HCI Event: Encryption Change (0x08) plen 4 Status: Success (0x00) Handle: 50 Address: BC:87:FA:47:73:5E (Bose Corporation) Encryption: Enabled with AES-CCM (0x02) > ACL Data RX: Handle 50 flags 0x02 dlen 6 BR/EDR SMP: Security Request (0x0b) len 1 Authentication requirement: No bonding, No MITM, SC (0x08) < HCI Command: LE Start Encryption (0x08|0x0019) plen 28 Handle: 50 Address: BC:87:FA:47:73:5E (Bose Corporation) > HCI Event: Command Status (0x0f) plen 4 LE Start Encryption (0x08|0x0019) ncmd 1 Status: Invalid HCI Command Parameters (0x12) < HCI Command: Disconnect (0x01|0x0006) plen 3 Handle: 50 Address: BC:87:FA:47:73:5E (Bose Corporation) Reason: Authentication Failure (0x05) SMP over BR/EDR is limited to cross-transport key derivation; the Security Request procedure (Core Specification Vol 3, Part H, Section 2.4.6, PDU in Section 3.6.7) has no BR/EDR counterpart. Reply with Pairing Failed / Command Not Supported on a non-LE link, before the PDU is parsed, and keep the connection. The reply is sent directly rather than through smp_failure(): rejecting a command on the wrong transport is not an authentication failure, and MGMT_EV_AUTH_FAILED would make bluetoothd disconnect the device. Tested on the affected host (kernel 7.0.12, MediaTek MT7922, Bose QC Ultra) with the patched module built out of tree: 7 days and 49 reconnects without a drop, against 2 drops in the 3 days before the patch. Every disconnect in that week had a userspace or remote reason. Fixes: b5ae344d4c0f ("Bluetooth: Add full SMP BR/EDR support") Assisted-by: LLM Signed-off-by: Christiano Amora Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/smp.c | 17 +++++++++++++++++ 1 file changed, 17 insertions(+) diff --git a/net/bluetooth/smp.c b/net/bluetooth/smp.c index 6091c47cb002..d23f9d0729c4 100644 --- a/net/bluetooth/smp.c +++ b/net/bluetooth/smp.c @@ -2269,6 +2269,23 @@ static u8 smp_cmd_security_req(struct l2cap_conn *conn, struct sk_buff *skb) bt_dev_dbg(hdev, "conn %p", conn); + /* SMP over BR/EDR only covers cross-transport key derivation; the + * Security Request procedure has no BR/EDR counterpart. Reject it + * here, otherwise smp_ltk_encrypt() finds the peer's LE LTK + * (ADDR_LE_DEV_PUBLIC and BDADDR_BREDR are both 0) and issues + * HCI_OP_LE_START_ENC on the ACL handle, which the controller + * rejects and hci_cs_le_start_enc() turns into a disconnect. Reply + * without smp_failure(): this is not an authentication failure, and + * MGMT_EV_AUTH_FAILED would make bluetoothd drop the device. + */ + if (hcon->type != LE_LINK) { + u8 reason = SMP_CMD_NOTSUPP; + + smp_send_cmd(conn, SMP_CMD_PAIRING_FAIL, sizeof(reason), + &reason); + return 0; + } + if (skb->len < sizeof(*rp)) return SMP_INVALID_PARAMS; From f2bbb36426581045a8bf7793da5419b9375e4348 Mon Sep 17 00:00:00 2001 From: Zijun Hu Date: Tue, 15 Sep 2026 19:17:18 -0700 Subject: [PATCH 108/694] Bluetooth: btnxpuart: Fix skb leak in nxp_process_fw_dump() When CONFIG_DEV_COREDUMP=n, hci_devcd_append() returns -EOPNOTSUPP without freeing its skb argument. This leaks the cloned skb and also prevents nxp_set_ind_reset() from being called to perform recovery. Fix by guarding the hci_devcd_append(hdev, skb_clone(skb, GFP_ATOMIC)) call with IS_ENABLED(CONFIG_DEV_COREDUMP). Fixes: 998e447f443f ("Bluetooth: btnxpuart: Add support for HCI coredump feature") Signed-off-by: Zijun Hu Signed-off-by: Luiz Augusto von Dentz --- drivers/bluetooth/btnxpuart.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/drivers/bluetooth/btnxpuart.c b/drivers/bluetooth/btnxpuart.c index 25e7b41b349f..4e23d71d00e8 100644 --- a/drivers/bluetooth/btnxpuart.c +++ b/drivers/bluetooth/btnxpuart.c @@ -1388,9 +1388,11 @@ static int nxp_process_fw_dump(struct hci_dev *hdev, struct sk_buff *skb) msecs_to_jiffies(20000)); } - err = hci_devcd_append(hdev, skb_clone(skb, GFP_ATOMIC)); - if (err < 0) - goto free_skb; + if (IS_ENABLED(CONFIG_DEV_COREDUMP)) { + err = hci_devcd_append(hdev, skb_clone(skb, GFP_ATOMIC)); + if (err < 0) + goto free_skb; + } if (buf_len == 0) { bt_dev_warn(hdev, "==== FW dump complete ==="); From 71af682ba4692c2ed9ace4c3d4ca462ae368c029 Mon Sep 17 00:00:00 2001 From: Lee Jones Date: Tue, 15 Sep 2026 12:08:22 +0000 Subject: [PATCH 109/694] Bluetooth: mgmt: Dequeue pending mesh_send_sync entries on cancel In send_cancel(), pending mesh_tx objects are removed from the hdev->mesh_pending list and freed via mesh_send_complete(). However, if a mesh transmission was already queued onto hdev->cmd_sync_work_list via mesh_next(), the queued entry retains a raw pointer to mesh_tx. When hci_cmd_sync_work later processes the entry, it attempts to execute mesh_send_sync and its destroy callback mesh_send_start_complete using the already freed mesh_tx pointer, leading to a use-after-free. Fix this by invoking hci_cmd_sync_dequeue() for mesh_send_sync on the target mesh_tx before completing it. If the entry is found and dequeued, its destroy callback will complete and free the object; otherwise, mesh_send_complete() is called directly. Additionally, ensure the transmission queue advances after cancellation or errors. In mesh_send_start_complete(), call mesh_next() on error unless err is -ECANCELED, because hci_cmd_sync_dequeue() holds hdev->cmd_sync_work_lock and calling mesh_next() synchronously would deadlock. Instead, advance the queue in send_cancel() once the lock is released and if no transmission is in progress. Fixes: b338d91703fa ("Bluetooth: Implement support for Mesh") Signed-off-by: Lee Jones Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/mgmt.c | 19 +++++++++++++++---- 1 file changed, 15 insertions(+), 4 deletions(-) diff --git a/net/bluetooth/mgmt.c b/net/bluetooth/mgmt.c index ac4864e56ec7..f740e745ae78 100644 --- a/net/bluetooth/mgmt.c +++ b/net/bluetooth/mgmt.c @@ -2316,6 +2316,8 @@ static void mesh_send_start_complete(struct hci_dev *hdev, void *data, int err) hci_dev_clear_flag(hdev, HCI_MESH_SENDING); /* Send Complete Error Code for handle */ mesh_send_complete(hdev, mesh_tx, false); + if (err != -ECANCELED) + mesh_next(hdev, NULL, 0); return; } @@ -2425,19 +2427,28 @@ static int send_cancel(struct hci_dev *hdev, void *data) do { mesh_tx = mgmt_mesh_next(hdev, cmd->sk); - if (mesh_tx) - mesh_send_complete(hdev, mesh_tx, false); + if (mesh_tx) { + if (!hci_cmd_sync_dequeue(hdev, mesh_send_sync, + mesh_tx, NULL)) + mesh_send_complete(hdev, mesh_tx, false); + } } while (mesh_tx); } else { mesh_tx = mgmt_mesh_find(hdev, cancel->handle); - if (mesh_tx && mesh_tx->sk == cmd->sk) - mesh_send_complete(hdev, mesh_tx, false); + if (mesh_tx && mesh_tx->sk == cmd->sk) { + if (!hci_cmd_sync_dequeue(hdev, mesh_send_sync, + mesh_tx, NULL)) + mesh_send_complete(hdev, mesh_tx, false); + } } mgmt_cmd_complete(cmd->sk, hdev->id, MGMT_OP_MESH_SEND_CANCEL, 0, NULL, 0); + if (!hci_dev_test_flag(hdev, HCI_MESH_SENDING)) + mesh_next(hdev, NULL, 0); + return 0; } From e06d549fcd4a0ba381ed67ddf1ab3c7a6ca4314c Mon Sep 17 00:00:00 2001 From: Aldo Ariel Panzardo Date: Tue, 15 Sep 2026 13:04:29 -0300 Subject: [PATCH 110/694] Bluetooth: hci_conn: fix CIS hold ownership on reuse Commit 69997d50ec57 ("Bluetooth: ISO: handle bound CIS cleanup via hci_conn") made hci_bind_cis() and hci_connect_cis() return a connection with one hold for the ISO layer. hci_bind_cis() currently takes that hold only after configuring a CIS, so its BT_CONNECTED and matching BT_BOUND paths return a bare lookup result. Its configuration failure path can likewise call hci_conn_drop() before taking a hold. Take the hold before any state-dependent return or configuration error so every successful return follows the documented ownership contract and every error drop is balanced. hci_connect_cis() also assumes hci_conn_link() always takes a new CIS hold before dropping the one returned by hci_bind_cis(). However, the helper returns an existing link without taking another hold. In that case, preserve the CIS hold for the caller and drop the redundant LE hold because the existing link already owns its parent hold. Returning early also avoids changing an existing CIS back to BT_CONNECT. Fixes: 69997d50ec57 ("Bluetooth: ISO: handle bound CIS cleanup via hci_conn") Cc: stable@vger.kernel.org Signed-off-by: Aldo Ariel Panzardo Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/hci_conn.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) diff --git a/net/bluetooth/hci_conn.c b/net/bluetooth/hci_conn.c index fa72cf8aaa7a..e32bb9c342a5 100644 --- a/net/bluetooth/hci_conn.c +++ b/net/bluetooth/hci_conn.c @@ -2079,6 +2079,8 @@ struct hci_conn *hci_bind_cis(struct hci_dev *hdev, bdaddr_t *dst, cis->conn_timeout = timeout; } + hci_conn_hold(cis); + if (cis->state == BT_CONNECTED) return cis; @@ -2120,7 +2122,6 @@ struct hci_conn *hci_bind_cis(struct hci_dev *hdev, bdaddr_t *dst, return ERR_PTR(-EINVAL); } - hci_conn_hold(cis); cis->state = BT_BOUND; return cis; @@ -2497,6 +2498,12 @@ struct hci_conn *hci_connect_cis(struct hci_dev *hdev, bdaddr_t *dst, return cis; } + /* The existing link already owns the hold on its parent. */ + if (cis->link) { + hci_conn_drop(le); + return cis; + } + link = hci_conn_link(le, cis); hci_conn_drop(cis); if (!link) { From 0fcd4dad555c96e0bd3a1b8c569f989be85c7341 Mon Sep 17 00:00:00 2001 From: Aldo Ariel Panzardo Date: Tue, 15 Sep 2026 13:04:30 -0300 Subject: [PATCH 111/694] Bluetooth: ISO: release unused CIS holds after channel attach hci_bind_cis() and hci_connect_cis() return one hci_conn hold for the ISO layer. A new channel association consumes that hold, which is eventually released by iso_conn_free(). There are two cases where iso_chan_add() does not create an association: it returns success when the socket is already attached to the same iso_conn, and it returns -EBUSY when another socket is attached. The hold returned for the current call is unused in both cases. This occurs when deferred setup calls iso_connect_cis() again for its existing socket, or when another socket attempts to reuse the CIS. Detect the idempotent case while the connection is locked and release the unused hold after iso_chan_add(). Also release it on -EBUSY. Do not drop it for other errors: a newly allocated iso_conn releases the transferred hold when its last temporary reference is put. Fixes: 69997d50ec57 ("Bluetooth: ISO: handle bound CIS cleanup via hci_conn") Cc: stable@vger.kernel.org Signed-off-by: Aldo Ariel Panzardo Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/iso.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/net/bluetooth/iso.c b/net/bluetooth/iso.c index eb99653f33f9..7657c2a0abbf 100644 --- a/net/bluetooth/iso.c +++ b/net/bluetooth/iso.c @@ -496,6 +496,7 @@ static int iso_connect_cis(struct sock *sk) struct hci_dev *hdev; bdaddr_t src, dst; u8 src_type; + bool already_attached; int err; lock_sock(sk); @@ -568,8 +569,14 @@ static int iso_connect_cis(struct sock *sk) goto unlock; } + iso_conn_lock(conn); + already_attached = iso_pi(sk)->conn == conn && conn->sk == sk; + iso_conn_unlock(conn); + err = iso_chan_add(conn, sk, NULL); iso_conn_put(conn); + if (already_attached || err == -EBUSY) + hci_conn_drop(hcon); if (err) goto unlock; From e93fad891c72deb84cae49430163b384ebcc92b1 Mon Sep 17 00:00:00 2001 From: Aldo Ariel Panzardo Date: Tue, 15 Sep 2026 13:03:58 -0300 Subject: [PATCH 112/694] Bluetooth: hci_sock: reject out-of-range OCF values The raw HCI socket security filter has 128 OCF bits per supported OGF, but masks the 10-bit OCF with 127 before looking up the command. An unprivileged socket can therefore submit a reserved OCF that aliases an allowlisted command modulo 128. A conforming controller should reject reserved opcodes. Nevertheless, the security decision must apply to the opcode that will actually be sent, especially since controller-specific behavior is outside the host stack's control. Reject OCF values that cannot be represented by the security filter instead of aliasing them onto an unrelated command. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Cc: stable@vger.kernel.org Signed-off-by: Aldo Ariel Panzardo Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/hci_sock.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/net/bluetooth/hci_sock.c b/net/bluetooth/hci_sock.c index 070ca388f9ac..6413593eee51 100644 --- a/net/bluetooth/hci_sock.c +++ b/net/bluetooth/hci_sock.c @@ -1881,7 +1881,8 @@ static int hci_sock_sendmsg(struct socket *sock, struct msghdr *msg, u16 ocf = hci_opcode_ocf(opcode); if (((ogf > HCI_SFLT_MAX_OGF) || - !hci_test_bit(ocf & HCI_FLT_OCF_BITS, + (ocf > HCI_FLT_OCF_BITS) || + !hci_test_bit(ocf, &hci_sec_filter.ocf_mask[ogf])) && !capable(CAP_NET_RAW)) { err = -EPERM; From b0a6cf99afd57a39598b1beca0e86ef5004980de Mon Sep 17 00:00:00 2001 From: Aldo Ariel Panzardo Date: Tue, 15 Sep 2026 13:03:07 -0300 Subject: [PATCH 113/694] Bluetooth: hci_sock: validate event length before filtering is_filtered_packet() reads the event code from skb->data[0] without first checking that the skb is nonempty. When an opcode filter is configured, it also reads the command opcode at offsets 3 or 4 without checking that a Command Complete or Command Status event is long enough. hci_send_to_sock() invokes the filter before hci_event_packet() validates the event header. A malformed event supplied by a controller or a vhci device can therefore cause an out-of-bounds read. Keep the unmasked event code for the opcode checks. The masked value is needed for the 64-bit event bitmap, but using it to identify command events aliases event codes above 0x3f. In particular, Synchronous Train Complete (0x4f) was treated as Command Status (0x0f) even though its payload has no opcode. Reject actual command events that are too short for the field being inspected. A truncated command event cannot match a configured opcode. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Cc: stable@vger.kernel.org Signed-off-by: Aldo Ariel Panzardo Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/hci_sock.c | 17 ++++++++++++++--- 1 file changed, 14 insertions(+), 3 deletions(-) diff --git a/net/bluetooth/hci_sock.c b/net/bluetooth/hci_sock.c index 6413593eee51..6d56c77741e1 100644 --- a/net/bluetooth/hci_sock.c +++ b/net/bluetooth/hci_sock.c @@ -164,6 +164,7 @@ static bool is_filtered_packet(struct sock *sk, struct sk_buff *skb) { struct hci_filter *flt; int flt_type, flt_event; + u8 event; /* Apply filter */ flt = &hci_pi(sk)->filter; @@ -177,7 +178,11 @@ static bool is_filtered_packet(struct sock *sk, struct sk_buff *skb) if (hci_skb_pkt_type(skb) != HCI_EVENT_PKT) return false; - flt_event = (*(__u8 *)skb->data & HCI_FLT_EVENT_BITS); + if (skb->len < 1) + return true; + + event = *(__u8 *)skb->data; + flt_event = event & HCI_FLT_EVENT_BITS; if (!hci_test_bit(flt_event, &flt->event_mask)) return true; @@ -186,11 +191,17 @@ static bool is_filtered_packet(struct sock *sk, struct sk_buff *skb) if (!flt->opcode) return false; - if (flt_event == HCI_EV_CMD_COMPLETE && + if (event == HCI_EV_CMD_COMPLETE && skb->len < 5) + return true; + + if (event == HCI_EV_CMD_COMPLETE && flt->opcode != get_unaligned((__le16 *)(skb->data + 3))) return true; - if (flt_event == HCI_EV_CMD_STATUS && + if (event == HCI_EV_CMD_STATUS && skb->len < 6) + return true; + + if (event == HCI_EV_CMD_STATUS && flt->opcode != get_unaligned((__le16 *)(skb->data + 4))) return true; From 4c94557dd02569efa6c1072a0439addaef9a5224 Mon Sep 17 00:00:00 2001 From: Aldo Ariel Panzardo Date: Tue, 15 Sep 2026 13:03:32 -0300 Subject: [PATCH 114/694] Bluetooth: ISO: balance the parent hold in hci_bind_bis() hci_conn_link() takes a lifetime reference to its parent with hci_conn_get(), but only takes an operational hold on the child. hci_conn_unlink() later balances both a hold and a reference on the parent. The SCO and CIS paths pass a parent acquired from a connect helper, so it already has a hold. For an additional BIS, hci_bind_bis() obtains the parent from hci_conn_hash_lookup_big(), which returns a bare pointer. Unlinking the child then drops the parent's existing hold and can schedule it for disconnection while its socket is still using it. Take a hold on the parent before linking it and drop that hold if linking fails. A successful link transfers the hold to hci_conn_unlink(). Fixes: fa224d0c094a ("Bluetooth: ISO: Reassociate a socket with an active BIS") Cc: stable@vger.kernel.org Signed-off-by: Aldo Ariel Panzardo Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/hci_conn.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/net/bluetooth/hci_conn.c b/net/bluetooth/hci_conn.c index e32bb9c342a5..96195d2fd10f 100644 --- a/net/bluetooth/hci_conn.c +++ b/net/bluetooth/hci_conn.c @@ -2375,10 +2375,13 @@ struct hci_conn *hci_bind_bis(struct hci_dev *hdev, bdaddr_t *dst, __u8 sid, parent = hci_conn_hash_lookup_big(hdev, conn->iso_qos.bcast.big); if (parent && parent != conn) { + hci_conn_hold(parent); link = hci_conn_link(parent, conn); hci_conn_drop(conn); - if (!link) + if (!link) { + hci_conn_drop(parent); return ERR_PTR(-ENOLINK); + } } return conn; From 6c78a213d9070b610c7f418af2c25b66180b7e37 Mon Sep 17 00:00:00 2001 From: Aldo Ariel Panzardo Date: Tue, 15 Sep 2026 13:02:39 -0300 Subject: [PATCH 115/694] Bluetooth: L2CAP: validate frame length before control and FCS access l2cap_data_rcv() unpacks either a two-byte or four-byte control field without first ensuring that it is present. A short ERTM or streaming-mode frame can therefore cause an out-of-bounds read. There is a second short-frame case when CRC16 is enabled. After the control field is pulled, l2cap_check_fcs() subtracts two from skb->len without checking it. If fewer than two bytes remain, the subtraction wraps; skb_trim() leaves the buffer unchanged and the subsequent FCS load reads past the logical end of the frame. Validate that the frame contains both its control field and, when enabled, its FCS before either field is accessed. Fixes: 1c2acffb76d4 ("Bluetooth: Add initial support for ERTM packets transfers") Fixes: fcc203c30d72 ("Bluetooth: Add support for FCS option to L2CAP") Cc: stable@vger.kernel.org Signed-off-by: Aldo Ariel Panzardo Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/l2cap_core.c | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-) diff --git a/net/bluetooth/l2cap_core.c b/net/bluetooth/l2cap_core.c index 644e31160d55..aaa2a1cd489a 100644 --- a/net/bluetooth/l2cap_core.c +++ b/net/bluetooth/l2cap_core.c @@ -6702,9 +6702,17 @@ static int l2cap_stream_rx(struct l2cap_chan *chan, struct l2cap_ctrl *control, static int l2cap_data_rcv(struct l2cap_chan *chan, struct sk_buff *skb) { struct l2cap_ctrl *control = &bt_cb(skb)->l2cap; - u16 len; + u16 len, min_len; u8 event; + min_len = test_bit(FLAG_EXT_CTRL, &chan->flags) ? + L2CAP_EXT_CTRL_SIZE : L2CAP_ENH_CTRL_SIZE; + if (chan->fcs == L2CAP_FCS_CRC16) + min_len += L2CAP_FCS_SIZE; + + if (skb->len < min_len) + goto drop; + __unpack_control(chan, skb); len = skb->len; From b5dbb41b212c50c095a4dbee3017a84fe94f033b Mon Sep 17 00:00:00 2001 From: Aldo Ariel Panzardo Date: Tue, 15 Sep 2026 12:59:52 -0300 Subject: [PATCH 116/694] Bluetooth: mgmt: fix race in read_unconf_index_list() read_unconf_index_list() counts unconfigured controllers before allocating its response, then checks the device flags again while filling it. hci_dev_list_lock stabilizes list membership, but it does not serialize the per-device flags. During asynchronous controller setup, the worker can set HCI_UNCONFIGURED and clear HCI_SETUP between the two passes. A controller omitted from the allocation count can then become eligible for the fill pass, causing an out-of-bounds write to rp->index[]. Allocate space for every device on hci_dev_list. Since list membership cannot change while hci_dev_list_lock is held, the response remains large enough regardless of flag transitions. The reported count and response length still include only eligible unconfigured controllers. Fixes: 73d1df2a7a10 ("Bluetooth: Add support for Read Unconfigured Index List command") Cc: stable@vger.kernel.org Signed-off-by: Aldo Ariel Panzardo Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/mgmt.c | 8 ++------ 1 file changed, 2 insertions(+), 6 deletions(-) diff --git a/net/bluetooth/mgmt.c b/net/bluetooth/mgmt.c index f740e745ae78..41956cdde982 100644 --- a/net/bluetooth/mgmt.c +++ b/net/bluetooth/mgmt.c @@ -496,13 +496,9 @@ static int read_unconf_index_list(struct sock *sk, struct hci_dev *hdev, read_lock(&hci_dev_list_lock); - count = 0; - list_for_each_entry(d, &hci_dev_list, list) { - if (hci_dev_test_flag(d, HCI_UNCONFIGURED)) - count++; - } + count = list_count_nodes(&hci_dev_list); - rp_len = sizeof(*rp) + (2 * count); + rp_len = sizeof(*rp) + (sizeof(__le16) * count); rp = kmalloc(rp_len, GFP_ATOMIC); if (!rp) { read_unlock(&hci_dev_list_lock); From c774ec8f0a5d02a06d34c27f5a7de7e333b91265 Mon Sep 17 00:00:00 2001 From: Eva Kurchatova Date: Wed, 16 Sep 2026 23:44:31 +0300 Subject: [PATCH 117/694] selftests: cgroup: give the O_TMPFILE open in get_temp_fd() a mode O_TMPFILE, like O_CREAT, needs the third argument. Without it glibc refuses the call at compile time as soon as fortification is on: In function 'open', inlined from 'get_temp_fd' at test_memcontrol.c:33:9: /usr/include/bits/fcntl2.h:52:11: error: call to '__open_missing_mode' declared with attribute error: open with O_CREAT or O_TMPFILE in second argument needs 3 arguments The fortify checks take effect only once the compiler optimises, and cgroup/Makefile builds with "-Wall -pthread" alone, so this goes unnoticed in a plain build. Building the tests with the flags distributions commonly use, -O2 -D_FORTIFY_SOURCE=3, loses test_memcontrol entirely. Fixes: 84092dbcf901 ("selftests: cgroup: add memory controller self-tests") Signed-off-by: Eva Kurchatova Signed-off-by: Tejun Heo --- tools/testing/selftests/cgroup/test_memcontrol.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/testing/selftests/cgroup/test_memcontrol.c b/tools/testing/selftests/cgroup/test_memcontrol.c index 3a84d068fbf3..0ed82347044e 100644 --- a/tools/testing/selftests/cgroup/test_memcontrol.c +++ b/tools/testing/selftests/cgroup/test_memcontrol.c @@ -30,7 +30,7 @@ static int page_size; int get_temp_fd(void) { - return open(".", O_TMPFILE | O_RDWR | O_EXCL); + return open(".", O_TMPFILE | O_RDWR | O_EXCL, 0600); } int alloc_pagecache(int fd, size_t size) From 42d1221d321e55afc7bba9109a77aaf5a817c8a3 Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Mon, 31 Aug 2026 12:27:20 -0700 Subject: [PATCH 118/694] scsi: megaraid_sas: Protect megasas_get_ctrl_info() in megasas_resume() Protect the megasas_get_ctrl_info() call in megasas_resume() with instance->reset_mutex using scoped_guard(). megasas_get_ctrl_info() may release and reacquire instance->reset_mutex. Hence, calling this function without holding instance->reset_mutex is not safe. Fixes: c3b10a55abc9 ("scsi: megaraid_sas: Update controller info during resume") Cc: Kashyap Desai Cc: Sumit Saxena Cc: Shivasharan S Cc: Chandrakanth patil Signed-off-by: Bart Van Assche Link: https://patch.msgid.link/f06b5ee432b21cf293f0663e15b64f75a84b9fd5.1788204406.git.bvanassche@acm.org Signed-off-by: Martin K. Petersen (Oracle) --- drivers/scsi/megaraid/megaraid_sas_base.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/scsi/megaraid/megaraid_sas_base.c b/drivers/scsi/megaraid/megaraid_sas_base.c index b95f187297ae..4f9a53769966 100644 --- a/drivers/scsi/megaraid/megaraid_sas_base.c +++ b/drivers/scsi/megaraid/megaraid_sas_base.c @@ -7886,7 +7886,9 @@ megasas_resume(struct device *dev) goto fail_init_mfi; } - if (megasas_get_ctrl_info(instance) != DCMD_SUCCESS) + scoped_guard(mutex, &instance->reset_mutex) + rval = megasas_get_ctrl_info(instance); + if (rval != DCMD_SUCCESS) goto fail_init_mfi; tasklet_init(&instance->isr_tasklet, instance->instancet->tasklet, From 2f91fc9a96cdab6c03d436246056552e04677636 Mon Sep 17 00:00:00 2001 From: Hemanth Selam Date: Mon, 7 Sep 2026 12:08:56 +0530 Subject: [PATCH 119/694] s390: Fix typos in comments Fix typos in comments, reported by scripts/checkpatch.pl using the misspelling list in scripts/spelling.txt. Only touches comments, no code changes. Assisted-by: Cursor:claude-opus-5 Signed-off-by: Hemanth Selam Signed-off-by: Heiko Carstens --- drivers/s390/block/dasd_3990_erp.c | 6 +++--- drivers/s390/block/dasd_eckd.c | 4 ++-- drivers/s390/char/raw3270.c | 2 +- drivers/s390/char/vmlogrdr.c | 2 +- 4 files changed, 7 insertions(+), 7 deletions(-) diff --git a/drivers/s390/block/dasd_3990_erp.c b/drivers/s390/block/dasd_3990_erp.c index 736459477c19..121cb9ae5c67 100644 --- a/drivers/s390/block/dasd_3990_erp.c +++ b/drivers/s390/block/dasd_3990_erp.c @@ -103,7 +103,7 @@ dasd_3990_erp_int_req(struct dasd_ccw_req * erp) /* first time set initial retry counter and erp_function */ /* and retry once without blocking queue */ - /* (this enables easier enqueing of the cqr) */ + /* (this enables easier enqueuing of the cqr) */ if (erp->function != dasd_3990_erp_int_req) { erp->retries = 256; @@ -302,7 +302,7 @@ dasd_3990_erp_action_4(struct dasd_ccw_req * erp, char *sense) /* first time set initial retry counter and erp_function */ /* and retry once without waiting for state change pending */ - /* interrupt (this enables easier enqueing of the cqr) */ + /* interrupt (this enables easier enqueuing of the cqr) */ if (erp->function != dasd_3990_erp_action_4) { DBF_DEV_EVENT(DBF_INFO, device, "%s", @@ -1078,7 +1078,7 @@ dasd_3990_erp_bus_out(struct dasd_ccw_req * erp) /* first time set initial retry counter and erp_function */ /* and retry once without blocking queue */ - /* (this enables easier enqueing of the cqr) */ + /* (this enables easier enqueuing of the cqr) */ if (erp->function != dasd_3990_erp_bus_out) { erp->retries = 256; erp->function = dasd_3990_erp_bus_out; diff --git a/drivers/s390/block/dasd_eckd.c b/drivers/s390/block/dasd_eckd.c index 8d976dd58a6c..9d1b11020b4e 100644 --- a/drivers/s390/block/dasd_eckd.c +++ b/drivers/s390/block/dasd_eckd.c @@ -2060,7 +2060,7 @@ dasd_eckd_psf_ssc(struct dasd_device *device, int enable_pav, } /* - * Valide storage server of current device. + * Valid storage server of current device. */ static int dasd_eckd_validate_server(struct dasd_device *device, unsigned long flags) @@ -5672,7 +5672,7 @@ static struct dasd_ccw_req *dasd_eckd_build_cp_raw(struct dasd_device *startdev, char *dst; /* - * raw track access needs to be mutiple of 64k and on 64k boundary + * raw track access needs to be multiple of 64k and on 64k boundary * For read requests we can fix an incorrect alignment by padding * the request with dummy pages. */ diff --git a/drivers/s390/char/raw3270.c b/drivers/s390/char/raw3270.c index aa9c4d81225c..6861bf27d15f 100644 --- a/drivers/s390/char/raw3270.c +++ b/drivers/s390/char/raw3270.c @@ -420,7 +420,7 @@ struct raw3270_ua { /* Query Reply structure for Usable Area */ char flags0; char flags1; short w; /* Width of usable area */ - short h; /* Heigth of usavle area */ + short h; /* Height of usavle area */ char units; /* 0x00:in; 0x01:mm */ int xr; int yr; diff --git a/drivers/s390/char/vmlogrdr.c b/drivers/s390/char/vmlogrdr.c index 383e7e2bd69f..005735a89920 100644 --- a/drivers/s390/char/vmlogrdr.c +++ b/drivers/s390/char/vmlogrdr.c @@ -444,7 +444,7 @@ static int vmlogrdr_receive_data(struct vmlogrdr_priv_t *priv) spin_unlock_bh(&priv->priv_lock); /* An rc of 5 indicates that the record was bigger than * the buffer, which is OK for us. A 9 indicates that the - * record was purged befor we could receive it. + * record was purged before we could receive it. */ if (rc == 5) rc = 0; From 4525a911049543c23885a540a788d13be318a486 Mon Sep 17 00:00:00 2001 From: Karl Mehltretter Date: Mon, 7 Sep 2026 07:58:47 +0200 Subject: [PATCH 120/694] s390/pci/docs: Fix sriov_numvfs attribute name The attribute is sriov_numvfs (drivers/pci/iov.c); the document names it sriov_numvf, which does not exist. Use sriov_numvfs. Fixes: de267a7c71ba ("s390/pci: Documentation for zPCI") Assisted-by: LLM Signed-off-by: Karl Mehltretter Reviewed-by: Randy Dunlap Signed-off-by: Heiko Carstens --- Documentation/arch/s390/pci.rst | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/Documentation/arch/s390/pci.rst b/Documentation/arch/s390/pci.rst index 80f4ba193159..565434626fb5 100644 --- a/Documentation/arch/s390/pci.rst +++ b/Documentation/arch/s390/pci.rst @@ -67,7 +67,7 @@ Entries specific to zPCI functions and entries that hold zPCI information. A physical function that currently supports a virtual function cannot be powered off until all virtual functions are removed with: - echo 0 > /sys/bus/pci/devices/DDDD:BB:dd.f/sriov_numvf + echo 0 > /sys/bus/pci/devices/DDDD:BB:dd.f/sriov_numvfs * /sys/bus/pci/devices/DDDD:BB:dd.f/: From 29d9e5835d89223aa913dcf7b942cc1c148bdd25 Mon Sep 17 00:00:00 2001 From: Vineeth Vijayan Date: Thu, 10 Sep 2026 11:32:01 +0200 Subject: [PATCH 121/694] s390/cio: Fix cio_update_schib() to not cache invalid schib When pmcw.dnv is 0, the contents of all SCHIB fields are unpredictable. Zero sch->schib in that case to prevent subsequent code from making decisions based on unpredictable data. Reported-by: William Bezenah Signed-off-by: Vineeth Vijayan Reviewed-by: Peter Oberparleiter Fixes: 8c58a229688c ("s390/cio: Do not unregister the subchannel based on DNV") Signed-off-by: Heiko Carstens --- drivers/s390/cio/cio.c | 11 +++++++---- 1 file changed, 7 insertions(+), 4 deletions(-) diff --git a/drivers/s390/cio/cio.c b/drivers/s390/cio/cio.c index 70dc8cc76594..e1c62eb60cca 100644 --- a/drivers/s390/cio/cio.c +++ b/drivers/s390/cio/cio.c @@ -453,7 +453,8 @@ EXPORT_SYMBOL_GPL(cio_commit_config); /** * cio_update_schib - Perform stsch and update schib if subchannel is valid. * @sch: subchannel on which to perform stsch - * Return zero on success, -ENODEV otherwise. + * Return zero on success, -ENODEV if the subchannel is not operational, + * -EACCES if the subchannel has no valid device. */ int cio_update_schib(struct subchannel *sch) { @@ -462,10 +463,12 @@ int cio_update_schib(struct subchannel *sch) if (stsch(sch->schid, &schib)) return -ENODEV; - memcpy(&sch->schib, &schib, sizeof(schib)); - - if (!css_sch_is_valid(&schib)) + if (!css_sch_is_valid(&schib)) { + memset(&sch->schib, 0, sizeof(sch->schib)); return -EACCES; + } + + memcpy(&sch->schib, &schib, sizeof(schib)); return 0; } From f6f2985eabdb2bfdc82ce90a1ea3ec53ba795f34 Mon Sep 17 00:00:00 2001 From: Vineeth Vijayan Date: Thu, 10 Sep 2026 11:32:02 +0200 Subject: [PATCH 122/694] s390/cio: Check pmcw.dnv before pmcw.ena in I/O entry points The device number valid (dnv) bit in the PMCW must be checked before acting on any other PMCW fields for IO-type subchannels. A subchannel with dnv=0 has no valid device number associated, making it meaningless to evaluate the enabled (ena) state or issue any I/O instruction against it. Reported-by: William Bezenah Signed-off-by: Vineeth Vijayan Reviewed-by: Peter Oberparleiter Fixes: 8c58a229688c ("s390/cio: Do not unregister the subchannel based on DNV") Signed-off-by: Heiko Carstens --- drivers/s390/cio/device_ops.c | 12 ++++++++++++ 1 file changed, 12 insertions(+) diff --git a/drivers/s390/cio/device_ops.c b/drivers/s390/cio/device_ops.c index 61c07b4a0fe8..c1ba4a19368f 100644 --- a/drivers/s390/cio/device_ops.c +++ b/drivers/s390/cio/device_ops.c @@ -142,6 +142,8 @@ int ccw_device_clear(struct ccw_device *cdev, unsigned long intparm) if (!cdev || !cdev->dev.parent) return -ENODEV; sch = to_subchannel(cdev->dev.parent); + if (!sch->schib.pmcw.dnv) + return -ENODEV; if (!sch->schib.pmcw.ena) return -EINVAL; if (cdev->private->state == DEV_STATE_NOT_OPER) @@ -198,6 +200,8 @@ int ccw_device_start_timeout_key(struct ccw_device *cdev, struct ccw1 *cpa, if (!cdev || !cdev->dev.parent) return -ENODEV; sch = to_subchannel(cdev->dev.parent); + if (!sch->schib.pmcw.dnv) + return -ENODEV; if (!sch->schib.pmcw.ena) return -EINVAL; if (cdev->private->state == DEV_STATE_NOT_OPER) @@ -379,6 +383,8 @@ int ccw_device_halt(struct ccw_device *cdev, unsigned long intparm) if (!cdev || !cdev->dev.parent) return -ENODEV; sch = to_subchannel(cdev->dev.parent); + if (!sch->schib.pmcw.dnv) + return -ENODEV; if (!sch->schib.pmcw.ena) return -EINVAL; if (cdev->private->state == DEV_STATE_NOT_OPER) @@ -413,6 +419,8 @@ int ccw_device_resume(struct ccw_device *cdev) if (!cdev || !cdev->dev.parent) return -ENODEV; sch = to_subchannel(cdev->dev.parent); + if (!sch->schib.pmcw.dnv) + return -ENODEV; if (!sch->schib.pmcw.ena) return -EINVAL; if (cdev->private->state == DEV_STATE_NOT_OPER) @@ -548,6 +556,8 @@ int ccw_device_tm_start_timeout_key(struct ccw_device *cdev, struct tcw *tcw, int rc; sch = to_subchannel(cdev->dev.parent); + if (!sch->schib.pmcw.dnv) + return -ENODEV; if (!sch->schib.pmcw.ena) return -EINVAL; if (cdev->private->state == DEV_STATE_VERIFY) { @@ -694,6 +704,8 @@ int ccw_device_tm_intrg(struct ccw_device *cdev) { struct subchannel *sch = to_subchannel(cdev->dev.parent); + if (!sch->schib.pmcw.dnv) + return -ENODEV; if (!sch->schib.pmcw.ena) return -EINVAL; if (cdev->private->state != DEV_STATE_ONLINE) From 9590f4d83880dfb5a81906e48e72779248fbe8f0 Mon Sep 17 00:00:00 2001 From: Vineeth Vijayan Date: Thu, 10 Sep 2026 11:32:03 +0200 Subject: [PATCH 123/694] s390/cio: Guard PMCW field accesses with dnv check When PMCW.DNV is 0, no I/O device is associated with the subchannel. However, several code paths access PMCW fields directly from the cached sch->schib without first invoking the update helper. Add explicit DNV validation before accessing PMCW fields from the cached SCHIB to avoid using invalid data. Reported-by: William Bezenah Signed-off-by: Vineeth Vijayan Reviewed-by: Peter Oberparleiter Fixes: 8c58a229688c ("s390/cio: Do not unregister the subchannel based on DNV") Signed-off-by: Heiko Carstens --- drivers/s390/cio/chp.c | 3 +++ drivers/s390/cio/device.c | 9 +++++---- drivers/s390/cio/device_fsm.c | 3 +++ drivers/s390/cio/device_ops.c | 9 +++++++++ drivers/s390/cio/vfio_ccw_fsm.c | 2 +- 5 files changed, 21 insertions(+), 5 deletions(-) diff --git a/drivers/s390/cio/chp.c b/drivers/s390/cio/chp.c index c890f21a82ce..eaf0527bff6c 100644 --- a/drivers/s390/cio/chp.c +++ b/drivers/s390/cio/chp.c @@ -78,6 +78,9 @@ u8 chp_get_sch_opm(struct subchannel *sch) int opm; int i; + if (!sch->schib.pmcw.dnv) + return 0; + opm = 0; chp_id_init(&chpid); for (i = 0; i < 8; i++) { diff --git a/drivers/s390/cio/device.c b/drivers/s390/cio/device.c index fb591118ecb2..68dd4a62975d 100644 --- a/drivers/s390/cio/device.c +++ b/drivers/s390/cio/device.c @@ -922,7 +922,7 @@ static int ccw_device_move_to_sch(struct ccw_device *cdev, if (!sch_is_pseudo_sch(old_sch)) { spin_lock_irq(&old_sch->lock); - old_enabled = old_sch->schib.pmcw.ena; + old_enabled = old_sch->schib.pmcw.dnv && old_sch->schib.pmcw.ena; rc = 0; if (old_enabled) rc = cio_disable_subchannel(old_sch); @@ -941,7 +941,7 @@ static int ccw_device_move_to_sch(struct ccw_device *cdev, CIO_MSG_EVENT(0, "device_move(0.%x.%04x,0.%x.%04x)=%d\n", cdev->private->dev_id.ssid, cdev->private->dev_id.devno, sch->schid.ssid, - sch->schib.pmcw.dev, rc); + sch->schid.sch_no, rc); if (old_enabled) { /* Try to re-enable the old subchannel. */ spin_lock_irq(&old_sch->lock); @@ -1207,7 +1207,7 @@ static void io_subchannel_quiesce(struct subchannel *sch) cdev = sch_get_cdev(sch); if (cio_is_console(sch->schid)) goto out_unlock; - if (!sch->schib.pmcw.ena) + if (!sch->schib.pmcw.dnv || !sch->schib.pmcw.ena) goto out_unlock; ret = cio_disable_subchannel(sch); if (ret != -EBUSY) @@ -1254,7 +1254,8 @@ static int recovery_check(struct device *dev, void *data) switch (cdev->private->state) { case DEV_STATE_ONLINE: sch = to_subchannel(cdev->dev.parent); - if ((sch->schib.pmcw.pam & sch->opm) == sch->vpm) + if (sch->schib.pmcw.dnv && + (sch->schib.pmcw.pam & sch->opm) == sch->vpm) break; fallthrough; case DEV_STATE_DISCONNECTED: diff --git a/drivers/s390/cio/device_fsm.c b/drivers/s390/cio/device_fsm.c index ab419d40a8a7..b5686c25c83c 100644 --- a/drivers/s390/cio/device_fsm.c +++ b/drivers/s390/cio/device_fsm.c @@ -170,6 +170,9 @@ __recover_lost_chpids(struct subchannel *sch, int old_lpm) int mask, i; struct chp_id chpid; + if (!sch->schib.pmcw.dnv) + return; + chp_id_init(&chpid); for (i = 0; i<8; i++) { mask = 0x80 >> i; diff --git a/drivers/s390/cio/device_ops.c b/drivers/s390/cio/device_ops.c index c1ba4a19368f..f2f7f8cba410 100644 --- a/drivers/s390/cio/device_ops.c +++ b/drivers/s390/cio/device_ops.c @@ -490,6 +490,8 @@ struct channel_path_desc_fmt0 *ccw_device_get_chp_desc(struct ccw_device *cdev, struct chp_id chpid; sch = to_subchannel(cdev->dev.parent); + if (!sch->schib.pmcw.dnv) + return NULL; chp_id_init(&chpid); chpid.id = sch->schib.pmcw.chpid[chp_idx]; return chp_get_chp_desc(chpid); @@ -510,6 +512,8 @@ u8 *ccw_device_get_util_str(struct ccw_device *cdev, int chp_idx) struct chp_id chpid; u8 *util_str; + if (!sch->schib.pmcw.dnv) + return NULL; chp_id_init(&chpid); chpid.id = sch->schib.pmcw.chpid[chp_idx]; chp = chpid_to_chp(chpid); @@ -662,6 +666,9 @@ int ccw_device_get_mdc(struct ccw_device *cdev, u8 mask) struct chp_id chpid; int mdc = 0, i; + if (!sch->schib.pmcw.dnv) + return 0; + /* Adjust requested path mask to excluded varied off paths. */ if (mask) mask &= sch->lpm; @@ -798,6 +805,8 @@ int ccw_device_get_chpid(struct ccw_device *cdev, int chp_idx, u8 *chpid) if ((chp_idx < 0) || (chp_idx > 7)) return -EINVAL; + if (!sch->schib.pmcw.dnv) + return -ENODEV; mask = 0x80 >> chp_idx; if (!(sch->schib.pmcw.pim & mask)) return -ENODEV; diff --git a/drivers/s390/cio/vfio_ccw_fsm.c b/drivers/s390/cio/vfio_ccw_fsm.c index 5fd94e9d5c61..9a000b0231d6 100644 --- a/drivers/s390/cio/vfio_ccw_fsm.c +++ b/drivers/s390/cio/vfio_ccw_fsm.c @@ -399,7 +399,7 @@ static void fsm_close(struct vfio_ccw_private *private, spin_lock_irq(&sch->lock); - if (!sch->schib.pmcw.ena) + if (!sch->schib.pmcw.dnv || !sch->schib.pmcw.ena) goto err_unlock; ret = cio_disable_subchannel(sch); From fc3ae66514ca5e87251f79044236e3e8babd24a3 Mon Sep 17 00:00:00 2001 From: David Howells Date: Mon, 14 Sep 2026 16:20:27 +0100 Subject: [PATCH 124/694] netfs: Fix netfs_read_gaps() to use separate sink folios Fix netfs_read_gaps() to use separate folios rather than re-using a single sink folio to discard the unwanted data so that cifs checksum checking sees all the data that was fetched. Fixes: 7f84a7b9892d ("netfs: Make netfs_read_folio() handle streaming-write pages") Reported-by: Frank Sorenson Closes: https://lore.kernel.org/r/a385053c-1c4a-4060-a3bb-befa007ddb33@redhat.com/ Signed-off-by: David Howells Link: https://patch.msgid.link/3228134.1789399227@warthog.procyon.org.uk Tested-by: Frank Sorenson Reviewed-by: Paulo Alcantara cc: Paulo Alcantara cc: Namjae Jeon cc: netfs@lists.linux.dev cc: linux-cifs@vger.kernel.org cc: linux-fsdevel@vger.kernel.org Signed-off-by: Christian Brauner (Amutable) --- fs/netfs/buffered_read.c | 34 +++++++++++++++++++--------------- 1 file changed, 19 insertions(+), 15 deletions(-) diff --git a/fs/netfs/buffered_read.c b/fs/netfs/buffered_read.c index 424df70a5c30..105194de6e13 100644 --- a/fs/netfs/buffered_read.c +++ b/fs/netfs/buffered_read.c @@ -482,15 +482,14 @@ static int netfs_read_gaps(struct file *file, struct folio *folio) struct netfs_group *group = netfs_folio_group(folio); struct netfs_folio *finfo = netfs_folio_info(folio); struct netfs_inode *ctx = netfs_inode(mapping->host); - struct folio *sink = NULL; - struct bio_vec *bvec; + struct bio_vec *bvec = NULL; unsigned int from = finfo->dirty_offset; unsigned int to = from + finfo->dirty_len; - unsigned int off = 0, i = 0; + unsigned int off = 0; size_t flen = folio_size(folio); size_t nr_bvec = flen / PAGE_SIZE + 2; size_t part; - int ret; + int ret, i = 0, sink_from = -1, sink_to = -1; _enter("%lx", folio->index); @@ -515,24 +514,23 @@ static int netfs_read_gaps(struct file *file, struct folio *folio) if (!bvec) goto discard; - sink = folio_alloc(GFP_KERNEL, 0); - if (!sink) { - kfree(bvec); - goto discard; - } - trace_netfs_folio(folio, netfs_folio_trace_read_gaps); - rreq->direct_bv = bvec; - rreq->direct_bv_count = nr_bvec; if (from > 0) { bvec_set_folio(&bvec[i++], folio, from, 0); off = from; } + sink_from = i; while (off < to) { + struct folio *sink = folio_alloc(GFP_KERNEL, 0); + + if (!sink) + goto discard; part = min_t(size_t, to - off, PAGE_SIZE); - bvec_set_folio(&bvec[i++], sink, part, 0); + bvec_set_folio(&bvec[i], sink, part, 0); off += part; + sink_to = i; + i++; } if (to < flen) bvec_set_folio(&bvec[i++], folio, flen - to, to); @@ -553,8 +551,10 @@ static int netfs_read_gaps(struct file *file, struct folio *folio) folio_mark_uptodate(folio); } - if (sink) - folio_put(sink); + if (sink_to >= 0) + for (; sink_from <= sink_to; sink_from++) + folio_put(bvec_folio(&bvec[sink_from])); + kfree(bvec); folio_unlock(folio); netfs_put_request(rreq, netfs_rreq_trace_put_return); return ret < 0 ? ret : 0; @@ -563,6 +563,10 @@ static int netfs_read_gaps(struct file *file, struct folio *folio) netfs_put_failed_request(rreq); alloc_error: folio_unlock(folio); + if (sink_to >= 0) + for (; sink_from <= sink_to; sink_from++) + folio_put(bvec_folio(&bvec[sink_from])); + kfree(bvec); return ret; } From e9d810279f84b30738f7790c0ed15f8dd5b9024a Mon Sep 17 00:00:00 2001 From: Wentao Liang Date: Wed, 16 Sep 2026 09:47:01 +0000 Subject: [PATCH 125/694] gpio: arizona: Fix runtime PM leak in arizona_gpio_direction_out() Switching a persistent GPIO line from input to output acquires a runtime PM reference on the parent device, but if the subsequent regmap_update_bits() fails the reference is never dropped and no later direction_in() can balance it since the direction was never changed. Drop the reference on the update failure path. Fixes: 27a49ed17e22 ("gpio: arizona: Add support for GPIOs that need to be maintained") Cc: stable@vger.kernel.org Signed-off-by: Wentao Liang Reviewed-by: Charles Keepax Link: https://patch.msgid.link/20260916094701.2007509-1-vulab@iscas.ac.cn Signed-off-by: Bartosz Golaszewski --- drivers/gpio/gpio-arizona.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/drivers/gpio/gpio-arizona.c b/drivers/gpio/gpio-arizona.c index a7e98d395d8e..88cf013f0d90 100644 --- a/drivers/gpio/gpio-arizona.c +++ b/drivers/gpio/gpio-arizona.c @@ -115,8 +115,12 @@ static int arizona_gpio_direction_out(struct gpio_chip *chip, if (value) value = ARIZONA_GPN_LVL; - return regmap_update_bits(arizona->regmap, ARIZONA_GPIO1_CTRL + offset, - ARIZONA_GPN_DIR | ARIZONA_GPN_LVL, value); + ret = regmap_update_bits(arizona->regmap, ARIZONA_GPIO1_CTRL + offset, + ARIZONA_GPN_DIR | ARIZONA_GPN_LVL, value); + if (ret < 0 && (val & ARIZONA_GPN_DIR) && persistent) + pm_runtime_put_autosuspend(chip->parent); + + return ret; } static int arizona_gpio_set(struct gpio_chip *chip, unsigned int offset, From 02af7eac17bc62a72335c1fc8c4af4471654f6cc Mon Sep 17 00:00:00 2001 From: Rosen Penev Date: Mon, 14 Sep 2026 17:44:06 -0700 Subject: [PATCH 126/694] gpio: mvebu: keep resume masks within the irqchip cache mvebu_gpio_resume() writes the edge/level mask registers saved at suspend time straight back to hardware, bypassing the irqchip's mask_cache_priv. genirq skips mask_irq() for a line it already considers masked, so restoring a bit in hardware that genirq thinks is still masked leaves that line unmasked behind genirq's back. An asserted level line then has nobody to ack it, and the moment interrupts are re-enabled the chained handler storms, hanging resume. AND the restored mask values with the matching irqchip mask cache so only lines genirq currently considers unmasked are unmasked again. Read the caches under gc->lock to keep them consistent with the mask/unmask handlers. Tested on Helios4 (armhf): 5 suspend cycles woken by magic packet, no hang; mvebu_gpio_resume() returns in 6 usecs. Assisted-by: LLM Signed-off-by: Rosen Penev Link: https://patch.msgid.link/20260915004406.115230-1-rosenp@gmail.com Signed-off-by: Bartosz Golaszewski --- drivers/gpio/gpio-mvebu.c | 33 +++++++++++++++++++++++++++------ 1 file changed, 27 insertions(+), 6 deletions(-) diff --git a/drivers/gpio/gpio-mvebu.c b/drivers/gpio/gpio-mvebu.c index 93b8a08b04b9..a2796240fe14 100644 --- a/drivers/gpio/gpio-mvebu.c +++ b/drivers/gpio/gpio-mvebu.c @@ -1034,6 +1034,8 @@ static int mvebu_gpio_suspend(struct platform_device *pdev, pm_message_t state) static int mvebu_gpio_resume(struct platform_device *pdev) { struct mvebu_gpio_chip *mvchip = platform_get_drvdata(pdev); + u32 edge_cache = ~0U, level_cache = ~0U; + unsigned long flags; int i; regmap_write(mvchip->regs, GPIO_OUT_OFF + mvchip->offset, @@ -1045,32 +1047,51 @@ static int mvebu_gpio_resume(struct platform_device *pdev) regmap_write(mvchip->regs, GPIO_IN_POL_OFF + mvchip->offset, mvchip->in_pol_reg); + /* + * genirq skips mask_irq() for a line it already considers masked, so + * unmasking one behind its back leaves an asserted level line that + * nobody masks. Restore only bits the irqchip cache still has set. + * + * Snapshot the caches under the raw spinlock, but release it before + * the regmap writes below: regmap_write() takes a sleepable lock on + * PREEMPT_RT. + */ + if (mvchip->domain) { + struct irq_chip_generic *gc; + + gc = irq_get_domain_generic_chip(mvchip->domain, 0); + raw_spin_lock_irqsave(&gc->lock, flags); + level_cache = gc->chip_types[0].mask_cache_priv; + edge_cache = gc->chip_types[1].mask_cache_priv; + raw_spin_unlock_irqrestore(&gc->lock, flags); + } + switch (mvchip->soc_variant) { case MVEBU_GPIO_SOC_VARIANT_ORION: case MVEBU_GPIO_SOC_VARIANT_A8K: regmap_write(mvchip->regs, GPIO_EDGE_MASK_OFF + mvchip->offset, - mvchip->edge_mask_regs[0]); + mvchip->edge_mask_regs[0] & edge_cache); regmap_write(mvchip->regs, GPIO_LEVEL_MASK_OFF + mvchip->offset, - mvchip->level_mask_regs[0]); + mvchip->level_mask_regs[0] & level_cache); break; case MVEBU_GPIO_SOC_VARIANT_MV78200: for (i = 0; i < 2; i++) { regmap_write(mvchip->regs, GPIO_EDGE_MASK_MV78200_OFF(i), - mvchip->edge_mask_regs[i]); + mvchip->edge_mask_regs[i] & edge_cache); regmap_write(mvchip->regs, GPIO_LEVEL_MASK_MV78200_OFF(i), - mvchip->level_mask_regs[i]); + mvchip->level_mask_regs[i] & level_cache); } break; case MVEBU_GPIO_SOC_VARIANT_ARMADAXP: for (i = 0; i < 4; i++) { regmap_write(mvchip->regs, GPIO_EDGE_MASK_ARMADAXP_OFF(i), - mvchip->edge_mask_regs[i]); + mvchip->edge_mask_regs[i] & edge_cache); regmap_write(mvchip->regs, GPIO_LEVEL_MASK_ARMADAXP_OFF(i), - mvchip->level_mask_regs[i]); + mvchip->level_mask_regs[i] & level_cache); } break; default: From c51e89c5b5db3e1927738fad7cd18b66dde68aed Mon Sep 17 00:00:00 2001 From: David Howells Date: Tue, 15 Sep 2026 17:21:59 +0100 Subject: [PATCH 127/694] netfs, afs: Fix symlink reading Fix the reading of symlinks from the cache in afs by making netfslib trim the amount read down to i_size. The problem is that afs sets the size of the iterator to the size of the buffer (PAGE_SIZE) so that the cache can round the read size up to the cache's DIO size. Note that this also impacts the reading of AFS mountpoints as they're just stored as symlinks with an odd file mode. Link: https://patch.msgid.link/3912795.1789489319@warthog.procyon.org.uk Fixes: c0410adf3da6 ("afs: Fix the locking used by afs_get_link()") Reviewed-by: Paulo Alcantara cc: Paulo Alcantara cc: Marc Dionne cc: linux-afs@lists.infradead.org cc: netfs@lists.linux.dev cc: linux-fsdevel@vger.kernel.org Signed-off-by: Christian Brauner (Amutable) --- fs/netfs/read_collect.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/fs/netfs/read_collect.c b/fs/netfs/read_collect.c index 5cf22087d243..a94197ef0181 100644 --- a/fs/netfs/read_collect.c +++ b/fs/netfs/read_collect.c @@ -435,6 +435,11 @@ static void netfs_rreq_assess_single(struct netfs_io_request *rreq) netfs_single_mark_inode_dirty(rreq->inode); } + /* To do DIO, the cache has to round the size up, so we need to undo + * the rounding. + */ + rreq->transferred = min(rreq->transferred, rreq->i_size); + if (rreq->iocb) { rreq->iocb->ki_pos += rreq->transferred; if (rreq->iocb->ki_complete) { From 7459c021874246c196f397686e100702f059b9e7 Mon Sep 17 00:00:00 2001 From: Julian Sun Date: Tue, 15 Sep 2026 12:49:12 +0800 Subject: [PATCH 128/694] fs: avoid repeated scans in evict_inodes() We observed hung tasks when users attempted to unmount a filesystem after its disk had been removed while still in use. During device removal, fs_bdev_mark_dead() calls evict_inodes() while holding s_umount. Each time evict_inodes() drops s_inode_list_lock to reschedule, it restarts the walk from the head of s_inodes. With many referenced inodes at the head of the list, these restarts repeatedly scan the same inodes without reclaiming them. This can keep s_umount held for a long time, blocking concurrent umount attempts and triggering hung-task reports. Keep the current inode, already marked I_FREEING, out of the disposal batch until s_inode_list_lock is reacquired. Resume the walk from this inode and dispose of it in a later batch or at the end of the walk. The zero-refcount and state checks under i_lock allow this walker to claim the inode by setting I_FREEING and removing it from the LRU. Other reclaimers skip the inode, leaving this walker responsible for eviction. Only evict() removes it from s_inodes, so keeping it out of the disposal batch ensures that it remains on the list while the lock is dropped. After reacquiring the lock, reading its current next pointer accounts for concurrent removal of following inodes. The existing inode lifetime rules prohibit acquiring a reference to an inode marked I_FREEING or I_WILL_FREE. __iget() requires its caller to hold i_lock and establish that taking a reference is valid. Inode lookup and igrab() check these flags under i_lock when acquiring a reference from zero. ihold() requires an existing reference, which would keep i_count nonzero and prevent this walker from claiming the inode. These rules already allow iput_final() and the inode shrinker to release i_lock after setting I_FREEING and before eviction completes. A temporary __iget() reference would also keep the inode on the list, but its release must preserve last-reference handling. Another user can acquire a reference, update lazy timestamps and drop its reference while the pin is held. If the pin becomes the last reference, dropping it with atomic_dec_and_test() and evicting directly bypasses iput()'s lazytime handling and can lose those timestamp updates. Releasing the pin with iput() preserves that handling, but does not guarantee eviction. fs_bdev_mark_dead() runs with SB_ACTIVE set, so iput() may retain the inode in cache, whereas evict_inodes() must evict eligible zero-reference inodes. The inode may also have been freed when iput() returns, so the walker cannot then use it to force eviction. Using I_FREEING preserves the existing eviction behavior without introducing an additional last-reference transition. The xfstests auto group passed on ext4 and XFS with known unrelated failures excluded. No new issues were observed, and the previously reproducible hung task no longer occurs with this patch. Fixes: ac05fbb40062 ("inode: don't softlockup when evicting inodes") Signed-off-by: Julian Sun Link: https://patch.msgid.link/20260915044912.3183440-1-sunjunchao@bytedance.com Reviewed-by: Jan Kara Signed-off-by: Christian Brauner (Amutable) --- fs/inode.c | 11 +++++------ 1 file changed, 5 insertions(+), 6 deletions(-) diff --git a/fs/inode.c b/fs/inode.c index ba7da39be4a3..a9d37be390a1 100644 --- a/fs/inode.c +++ b/fs/inode.c @@ -880,7 +880,6 @@ void evict_inodes(struct super_block *sb) struct inode *inode; LIST_HEAD(dispose); -again: spin_lock(&sb->s_inode_list_lock); list_for_each_entry(inode, &sb->s_inodes, i_sb_list) { if (icount_read_once(inode)) @@ -899,19 +898,19 @@ void evict_inodes(struct super_block *sb) inode_state_set(inode, I_FREEING); inode_lru_list_del(inode); spin_unlock(&inode->i_lock); - list_add(&inode->i_lru, &dispose); /* - * We can have a ton of inodes to evict at unmount time given - * enough memory, check to see if we need to go to sleep for a - * bit so we don't livelock. + * Keep this inode out of dispose so it stays on s_inodes while + * the list lock is dropped. I_FREEING prevents new references + * and leaves eviction to us, so we can resume the walk from it. */ if (need_resched()) { spin_unlock(&sb->s_inode_list_lock); cond_resched(); dispose_list(&dispose); - goto again; + spin_lock(&sb->s_inode_list_lock); } + list_add(&inode->i_lru, &dispose); } spin_unlock(&sb->s_inode_list_lock); From f6988c90671e83db79df1b7b9d6fdb0e5947fd84 Mon Sep 17 00:00:00 2001 From: "Patrick Lu (Anthropic)" Date: Fri, 11 Sep 2026 18:49:49 +0000 Subject: [PATCH 129/694] writeback: bound cleanup_offline_cgwb() rescans by rotating scanned inodes cleanup_offline_cgwb() prepares at most WB_MAX_INODES_PER_ISW inodes per call and is called again until the dying wb is drained, but every call walks wb->b_attached and then wb->b_dirty_time from the same end. Inodes already prepared (they stay on the list with I_WB_SWITCH set until the switch worker runs) and inodes that cannot be switched (I_FREEING, I_WILL_FREE, !SB_ACTIVE, DAX, already on the target wb) stay where they are, so each pass rescans a growing run of them under wb->list_lock and a full drain is quadratic in the number of inodes on the list. With ~17M inodes attached to one dying cgwb we saw this end in soft lockups, with CPUs reported stuck for 21-48s. Walk both lists from the oldest end and move every scanned inode to the newest end, so the next pass starts where the previous one stopped and the drain becomes linear. b_attached is unordered, so nobody sees the reorder there. b_dirty_time is ordered by dirtied_when, but the oldest unscanned inode stays at the end move_expired_inodes() picks from, sync takes the whole list regardless of order, and prepared inodes leave the list as soon as the switch work runs and get a new dirtied_time_when on the new wb anyway, so the only inodes left out of order are the ones that can never switch (DAX), and only on the dying wb. Fixes: c22d70a162d3 ("writeback, cgroup: release dying cgwbs by switching attached inodes") Cc: stable@vger.kernel.org Acked-by: Tejun Heo Acked-by: Roman Gushchin Signed-off-by: Patrick Lu (Anthropic) Link: https://patch.msgid.link/20260911-wb-cgwb-rotate-v2-1-a9ab253a1295@gmail.com Reviewed-by: Jan Kara Signed-off-by: Christian Brauner (Amutable) --- fs/fs-writeback.c | 25 ++++++++++++++++++++----- 1 file changed, 20 insertions(+), 5 deletions(-) diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c index e744f9f9d43f..ea3eb40bf828 100644 --- a/fs/fs-writeback.c +++ b/fs/fs-writeback.c @@ -727,19 +727,34 @@ static bool isw_prepare_wbs_switch(struct bdi_writeback *new_wb, struct inode_switch_wbs_context *isw, struct list_head *list, int *nr) { - struct inode *inode; + struct inode *inode, *tmp; + LIST_HEAD(scanned); + bool full = false; + + /* + * Walk from the oldest end and move scanned inodes to the newest + * end, so the next scan resumes at unscanned inodes instead of + * re-walking an ever-growing run of prepared and skipped ones. + * For b_dirty_time this keeps the oldest unscanned inode at the + * end move_expired_inodes() picks from; b_attached is unordered. + */ + list_for_each_entry_safe_reverse(inode, tmp, list, i_io_list) { + list_move(&inode->i_io_list, &scanned); - list_for_each_entry(inode, list, i_io_list) { if (!inode_prepare_wbs_switch(inode, new_wb)) continue; isw->inodes[*nr] = inode; (*nr)++; - if (*nr >= WB_MAX_INODES_PER_ISW - 1) - return true; + if (*nr >= WB_MAX_INODES_PER_ISW - 1) { + full = true; + break; + } } - return false; + list_splice(&scanned, list); + + return full; } /** From 1fca688e9443003e33cf30453e7a7560367656c9 Mon Sep 17 00:00:00 2001 From: shechenglong Date: Mon, 7 Sep 2026 11:51:47 +0800 Subject: [PATCH 130/694] drm/client: fix restore of partially initialized client I got a null-ptr-deref report when closing a DRM file descriptor: WARNING: drivers/gpu/drm/drm_atomic.c:2031 at __drm_atomic_helper_set_config+0x18e/0x1b0 [drm] Call Trace: drm_client_modeset_commit_atomic+0x16b/0x220 [drm] drm_client_modeset_commit_locked+0x56/0x160 [drm] drm_client_modeset_commit+0x21/0x40 [drm] __drm_fb_helper_restore_fbdev_mode_unlocked.part.0+0x7b/0x80 drm_fbdev_client_restore+0xe/0x20 [drm_client_lib] drm_client_dev_restore+0x9f/0xc0 [drm] drm_release+0xc5/0xe0 [drm] The warning is followed by a NULL pointer dereference: BUG: kernel NULL pointer dereference, address: 0000000000000008 RIP: __drm_fb_helper_restore_fbdev_mode_unlocked.part.0+0x41/0x80 [drm_kms_helper] Call Trace: drm_fbdev_client_restore+0xe/0x20 [drm_client_lib] drm_client_dev_restore+0x9f/0xc0 [drm] drm_release+0xc5/0xe0 [drm] __fput+0xdc/0x2b0 __x64_sys_close+0x39/0x80 do_syscall_64+0x8d/0x460 entry_SYSCALL_64_after_hwframe+0x76/0x7e drm_client_register() adds the DRM client to the device client list before invoking the initial hotplug callback. If the hotplug callback fails, the client remains registered. For the fbdev client, a failure during drm_fb_helper_initial_config() causes the partially initialized fbdev helper to be cleaned up. drm_fb_helper_fini() releases fb_helper->info and leaves it NULL. The fbdev client therefore remains registered even though there is no fully initialized framebuffer device. Later, when userspace closes the DRM file descriptor, drm_release() can invoke the restore callbacks of registered DRM clients: drm_release() drm_client_dev_restore() drm_fbdev_client_restore() drm_fb_helper_restore_fbdev_mode_unlocked() drm_fbdev_client_restore() currently restores the fbdev state unconditionally. For a partially initialized fbdev client this can submit an incomplete modeset state and subsequently access fbdev state which has not been initialized, resulting in the warning and NULL pointer dereference above. drm_fbdev_client_unregister() already uses fb_helper->info to distinguish a fully probed framebuffer device from a partially initialized client. Use the same condition in drm_fbdev_client_restore() and skip restore if no framebuffer device has been successfully initialized. Signed-off-by: shechenglong Reviewed-by: Thomas Zimmermann Fixes: 5d08c44e47b9 ("drm/fbdev: Add memory-agnostic fbdev client") Signed-off-by: Thomas Zimmermann Cc: # v6.13+ Link: https://patch.msgid.link/20260907035147.1339-1-shechenglong@xfusion.com --- drivers/gpu/drm/clients/drm_fbdev_client.c | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/drivers/gpu/drm/clients/drm_fbdev_client.c b/drivers/gpu/drm/clients/drm_fbdev_client.c index 91d196a397cf..1c16bc1084c4 100644 --- a/drivers/gpu/drm/clients/drm_fbdev_client.c +++ b/drivers/gpu/drm/clients/drm_fbdev_client.c @@ -42,6 +42,14 @@ static int drm_fbdev_client_restore(struct drm_client_dev *client, bool force) { struct drm_fb_helper *fb_helper = drm_fb_helper_from_client(client); + /* + * The client is registered before the initial fbdev probe. + * If probing failed, the client remains registered but there + * is no valid fbdev framebuffer to restore. + */ + if (!fb_helper->info || !fb_helper->fb) + return 0; + drm_fb_helper_restore_fbdev_mode_unlocked(fb_helper, force); return 0; From df5cdc2c832ca4e8a6d774596b9005558761a403 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Wed, 16 Sep 2026 12:00:24 -1000 Subject: [PATCH 131/694] sched_ext: Derive SCX_RQ_IN_WAKEUP from the core enqueue flags schedule_deferred_locked() skips scheduling a deferred action while SCX_RQ_IN_WAKEUP is set and relies on the task_woken_scx() call that follows a wakeup enqueue to run it. enqueue_task_scx() sets the flag from the merged enqueue flags, which include the flags stashed for a remote activation. move_remote_task_to_local_dsq() thus sets SCX_RQ_IN_WAKEUP on the destination rq when the moved task was woken up, although no task_woken_scx() follows that activation. An IMMED insert into a busy destination requests a local reenqueue during that enqueue. The request gets linked but not scheduled and stays pending until an unrelated wakeup or preemption on that CPU runs the deferred actions. The IMMED task sits behind the running task in the meantime. If nothing runs them before the scheduler is disabled, the request outlives the scheduler and points into its freed per-cpu area, which the next scheduler dereferences from run_deferred(). Test the core enqueue flags for the wakeup bit. Only the core's wakeup path is followed by task_woken_scx(). Fixes: 57ccf5ccdc56 ("sched_ext: Fix enqueue_task_scx() truncation of upper enqueue flags") Cc: stable@vger.kernel.org # v7.1+ Reported-by: Andrea Righi Link: https://lore.kernel.org/all/20260916145807.3250167-1-arighi@nvidia.com/ Signed-off-by: Tejun Heo Reviewed-by: Andrea Righi --- kernel/sched/ext/ext.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/kernel/sched/ext/ext.c b/kernel/sched/ext/ext.c index 725b890fcab9..e207ccd17164 100644 --- a/kernel/sched/ext/ext.c +++ b/kernel/sched/ext/ext.c @@ -2141,7 +2141,12 @@ static void enqueue_task_scx(struct rq *rq, struct task_struct *p, int core_enq_ int sticky_cpu = p->scx.sticky_cpu; u64 enq_flags = core_enq_flags | rq->scx.remote_activate_enq_flags; - if (enq_flags & ENQUEUE_WAKEUP) + /* + * SCX_RQ_IN_WAKEUP promises a task_woken_scx() call once this enqueue + * returns. Only the core's wakeup path delivers one. The flags stashed + * for a remote activation may carry the wakeup bit without it. + */ + if (core_enq_flags & ENQUEUE_WAKEUP) rq->scx.flags |= SCX_RQ_IN_WAKEUP; /* From 67b4411538c8341692548429d43256f25be99f7a Mon Sep 17 00:00:00 2001 From: Wentao Liang Date: Wed, 16 Sep 2026 18:00:36 +0000 Subject: [PATCH 132/694] drm/nouveau: Fix bridge reference leak in nv1a_ram_new() pci_get_domain_bus_and_slot() takes a reference to the PCI device, which is never released once the memory size has been read from its config space. Drop the reference before returning. Fixes: 2fa6d6cdaf283c05 ("drm/nouveau: deprecate pci_get_bus_and_slot()") Cc: stable@vger.kernel.org Signed-off-by: Wentao Liang Reviewed-by: Lyude Paul Signed-off-by: Lyude Paul Link: https://patch.msgid.link/20260916180036.2090118-1-vulab@iscas.ac.cn --- drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramnv1a.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramnv1a.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramnv1a.c index 18241c6ba5fa..4d52a158f320 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramnv1a.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramnv1a.c @@ -51,6 +51,8 @@ nv1a_ram_new(struct nvkm_fb *fb, struct nvkm_ram **pram) mib = ((mem >> 4) & 127) + 1; } + pci_dev_put(bridge); + return nvkm_ram_new_(&nv04_ram_func, fb, NVKM_RAM_TYPE_STOLEN, mib * 1024 * 1024, pram); } From cb86607ada73185f321baa7f9d94a08a3a40bbf5 Mon Sep 17 00:00:00 2001 From: fangqiurong Date: Thu, 17 Sep 2026 15:52:41 +0800 Subject: [PATCH 133/694] sched_ext: Don't run ops.dequeue() with a DSQ lock held ops.dequeue() is invoked with the source user DSQ's lock still held on the consume and move paths (scx_consume_dispatch_q(), move_task_between_dsqs()). A BPF scheduler which locks the source user DSQ from ops.dequeue() - e.g. by iterating it with bpf_iter_scx_dsq - self-deadlocks. ops.dequeue() can only call the "any" kfuncs and none of them can lock a builtin DSQ, so the global and bypass paths can't deadlock; however, all DSQ locks share one lockdep class, so iterating any user DSQ from ops.dequeue() on those paths trips the recursion check. Move the invocation after the DSQ unlock on all three paths. SCX_TASK_IN_CUSTODY is cleared under the lock serializing the transfer so that the callback is invoked exactly once. Fixes: ebf1ccff79c4 ("sched_ext: Fix ops.dequeue() semantics") Cc: stable@vger.kernel.org # v7.1+ Acked-by: Andrea Righi Signed-off-by: fangqiurong Signed-off-by: Tejun Heo --- kernel/sched/ext/ext.c | 44 ++++++++++++++++++------------------- kernel/sched/ext/internal.h | 3 +-- kernel/sched/ext/sub.c | 10 ++++----- 3 files changed, 28 insertions(+), 29 deletions(-) diff --git a/kernel/sched/ext/ext.c b/kernel/sched/ext/ext.c index e207ccd17164..f568fd9973f6 100644 --- a/kernel/sched/ext/ext.c +++ b/kernel/sched/ext/ext.c @@ -1499,27 +1499,22 @@ static inline bool task_scx_migrating(struct task_struct *p) return p->scx.sticky_cpu >= 0; } -/* - * Call ops.dequeue() if the task is in BPF custody and not migrating. - * Clears %SCX_TASK_IN_CUSTODY when the callback is invoked. - */ -static void call_task_dequeue(struct scx_sched *sch, struct rq *rq, - struct task_struct *p, u64 deq_flags) +/* Must be called under the lock serializing @p's custody transfers. */ +static bool task_leave_custody(struct task_struct *p) { if (!(p->scx.flags & SCX_TASK_IN_CUSTODY) || task_scx_migrating(p)) - return; - - if (SCX_HAS_OP(sch, dequeue)) - SCX_CALL_OP_TASK(sch, dequeue, rq, p, deq_flags); + return false; p->scx.flags &= ~SCX_TASK_IN_CUSTODY; + return true; } static void rq_owned_post_enq(struct scx_sched *sch, struct rq *rq, struct scx_dispatch_q *dsq, struct task_struct *p, u64 enq_flags) { - call_task_dequeue(sch, rq, p, 0); + if (task_leave_custody(p) && SCX_HAS_OP(sch, dequeue)) + SCX_CALL_OP_TASK(sch, dequeue, rq, p, 0); /* * Only local inserts get the wakeup treatment below. Rejects kick the @@ -1705,20 +1700,28 @@ static void scx_dispatch_enqueue(struct scx_sched *sch, struct rq *rq, if (is_rq_owned) { rq_owned_post_enq(sch, rq, dsq, p, enq_flags); } else { + bool call_dequeue = false; + /* * Global and bypass DSQs are terminal - the task leaves the - * scheduler's custody, so ops.dequeue() fires here. It can run + * scheduler's custody, so ops.dequeue() fires. It can run * without @p's rq lock (finish_dispatch() passes the dispatch * rq); that's safe because dequeue_task_scx() waits on * SCX_OPSS_DISPATCHING (see the ops_state note above) and so * can't race it. A non-terminal DSQ keeps the task in custody. + * The custody transfer happens under @dsq->lock so that later + * consumers see the flag clear; the callback runs after + * @dsq->lock is dropped because it may lock a DSQ itself. */ if (dsq->id == SCX_DSQ_GLOBAL || dsq->id == SCX_DSQ_BYPASS) - call_task_dequeue(sch, rq, p, 0); + call_dequeue = task_leave_custody(p); else p->scx.flags |= SCX_TASK_IN_CUSTODY; raw_spin_unlock(&dsq->lock); + + if (call_dequeue && SCX_HAS_OP(sch, dequeue)) + SCX_CALL_OP_TASK(sch, dequeue, rq, p, 0); } /* @@ -2215,7 +2218,7 @@ static void ops_dequeue(struct rq *rq, struct task_struct *p, u64 deq_flags) /* * A queued task must always be in BPF scheduler's custody. If * SCX_TASK_IN_CUSTODY is clear, finish_dispatch() on another - * CPU has already passed call_task_dequeue() (which clears the + * CPU has already passed task_leave_custody() (which clears the * flag), but has not yet written SCX_OPSS_NONE. That final * store does not require this rq's lock, so retrying with * cpu_relax() is bounded: we will observe NONE (or DISPATCHING, @@ -2263,7 +2266,8 @@ static void ops_dequeue(struct rq *rq, struct task_struct *p, u64 deq_flags) * NONE but the task may still have %SCX_TASK_IN_CUSTODY set until * it is enqueued on the destination. */ - call_task_dequeue(sch, rq, p, deq_flags); + if (task_leave_custody(p) && SCX_HAS_OP(sch, dequeue)) + SCX_CALL_OP_TASK(sch, dequeue, rq, p, deq_flags); } static bool dequeue_task_scx(struct rq *rq, struct task_struct *p, int core_deq_flags) @@ -2379,14 +2383,10 @@ static void wakeup_preempt_scx(struct rq *rq, struct task_struct *p, int wake_fl } void scx_move_local_task_to_local_dsq(struct scx_sched *sch, struct task_struct *p, - u64 enq_flags, struct scx_dispatch_q *src_dsq, - struct rq *dst_rq) + u64 enq_flags, struct rq *dst_rq) { struct scx_dispatch_q *dst_dsq = scx_resolve_local_dsq(sch, dst_rq, p, &enq_flags); - /* @p is on @dst_rq, an rq-owned @src_dsq is covered by the rq lock */ - if (!dsq_is_rq_owned(src_dsq)) - lockdep_assert_held(&src_dsq->lock); lockdep_assert_rq_held(dst_rq); WARN_ON_ONCE(p->scx.holding_cpu >= 0); @@ -2634,8 +2634,8 @@ static struct rq *move_task_between_dsqs(struct scx_sched *sch, /* @p is going from a non-local DSQ to a local DSQ */ if (src_rq == dst_rq) { scx_task_unlink_from_dsq(p, src_dsq); - scx_move_local_task_to_local_dsq(sch, p, enq_flags, src_dsq, dst_rq); raw_spin_unlock(&src_dsq->lock); + scx_move_local_task_to_local_dsq(sch, p, enq_flags, dst_rq); } else { raw_spin_unlock(&src_dsq->lock); move_remote_task_to_local_dsq(sch, p, enq_flags, src_rq, dst_rq); @@ -2685,8 +2685,8 @@ bool scx_consume_dispatch_q(struct scx_sched *sch, struct rq *rq, if (rq == task_rq) { scx_task_unlink_from_dsq(p, dsq); - scx_move_local_task_to_local_dsq(sch, p, enq_flags, dsq, rq); raw_spin_unlock(&dsq->lock); + scx_move_local_task_to_local_dsq(sch, p, enq_flags, rq); return true; } diff --git a/kernel/sched/ext/internal.h b/kernel/sched/ext/internal.h index 115c96fbf932..d150de10a5c9 100644 --- a/kernel/sched/ext/internal.h +++ b/kernel/sched/ext/internal.h @@ -2065,8 +2065,7 @@ void scx_dispatch_dequeue(struct rq *rq, struct task_struct *p); void scx_do_enqueue_task(struct rq *rq, struct task_struct *p, u64 enq_flags, int sticky_cpu); void scx_move_local_task_to_local_dsq(struct scx_sched *sch, struct task_struct *p, - u64 enq_flags, struct scx_dispatch_q *src_dsq, - struct rq *dst_rq); + u64 enq_flags, struct rq *dst_rq); bool scx_consume_dispatch_q(struct scx_sched *sch, struct rq *rq, struct scx_dispatch_q *dsq, u64 enq_flags); bool scx_consume_global_dsq(struct scx_sched *sch, struct rq *rq); diff --git a/kernel/sched/ext/sub.c b/kernel/sched/ext/sub.c index a30c965b175d..3036190b0662 100644 --- a/kernel/sched/ext/sub.c +++ b/kernel/sched/ext/sub.c @@ -555,8 +555,8 @@ static void scx_rescue_timerfn(struct timer_list *timer) scx.dsq_list.node); scx_task_unlink_from_dsq(p, &rq->scx.rescue.dsq); scx_rescue_admit(rq, p, slice); - scx_move_local_task_to_local_dsq(scx_task_sched(p), p, SCX_ENQ_IGNORE_CAPS, - &rq->scx.rescue.dsq, rq); + scx_move_local_task_to_local_dsq(scx_task_sched(p), p, + SCX_ENQ_IGNORE_CAPS, rq); if (sched_class_above(&ext_sched_class, rq->curr->sched_class)) resched_curr(rq); } else if (p->scx.dsq && rq->scx.rescue.budget > 2 * scx_rescue_quantum_ns) { @@ -572,7 +572,7 @@ static void scx_rescue_timerfn(struct timer_list *timer) scx_task_unlink_from_dsq(p, &rq->scx.local_dsq); scx_move_local_task_to_local_dsq(scx_task_sched(p), p, SCX_ENQ_HEAD | SCX_ENQ_PREEMPT | SCX_ENQ_IGNORE_CAPS, - &rq->scx.local_dsq, rq); + rq); } out_arm: scx_rescue_timer_arm(rq); @@ -596,8 +596,8 @@ void scx_rescue_flush(struct rq *rq) /* and flush out all pending ones */ list_for_each_entry_safe(p, n, &rq->scx.rescue.dsq.list, scx.dsq_list.node) { scx_task_unlink_from_dsq(p, &rq->scx.rescue.dsq); - scx_move_local_task_to_local_dsq(scx_task_sched(p), p, SCX_ENQ_IGNORE_CAPS, - &rq->scx.rescue.dsq, rq); + scx_move_local_task_to_local_dsq(scx_task_sched(p), p, + SCX_ENQ_IGNORE_CAPS, rq); } timer_delete(&rq->scx.rescue.timer); From 9ec7ba20c97d92fa59b351832aeeb934b3b16177 Mon Sep 17 00:00:00 2001 From: fangqiurong Date: Thu, 17 Sep 2026 15:52:42 +0800 Subject: [PATCH 134/694] selftests/sched_ext: Test that ops.dequeue() can iterate the consumed DSQ Add a scheduler whose ops.dequeue() iterates the source user DSQ with bpf_iter_scx_dsq. The iteration takes the DSQ's raw spinlock; on a kernel that runs ops.dequeue() while the consume path still holds that lock, the first task consumed self-deadlocks the CPU with IRQs disabled. The watchdog cannot recover from that state, so on an unfixed kernel this test wedges the system instead of failing cleanly. On a fixed kernel the scheduler runs clean and the test passes. Signed-off-by: fangqiurong Signed-off-by: Tejun Heo --- tools/testing/selftests/sched_ext/Makefile | 1 + .../selftests/sched_ext/dequeue_iter.bpf.c | 73 +++++++++++++++++ .../selftests/sched_ext/dequeue_iter.c | 79 +++++++++++++++++++ 3 files changed, 153 insertions(+) create mode 100644 tools/testing/selftests/sched_ext/dequeue_iter.bpf.c create mode 100644 tools/testing/selftests/sched_ext/dequeue_iter.c diff --git a/tools/testing/selftests/sched_ext/Makefile b/tools/testing/selftests/sched_ext/Makefile index 3cfe90e0f34f..49897727f535 100644 --- a/tools/testing/selftests/sched_ext/Makefile +++ b/tools/testing/selftests/sched_ext/Makefile @@ -164,6 +164,7 @@ all_test_bpfprogs := $(foreach prog,$(wildcard *.bpf.c),$(INCLUDE_DIR)/$(patsubs auto-test-targets := \ create_dsq \ dequeue \ + dequeue_iter \ enq_last_no_enq_fails \ ddsp_bogus_dsq_fail \ ddsp_vtimelocal_fail \ diff --git a/tools/testing/selftests/sched_ext/dequeue_iter.bpf.c b/tools/testing/selftests/sched_ext/dequeue_iter.bpf.c new file mode 100644 index 000000000000..76c2c71a90b6 --- /dev/null +++ b/tools/testing/selftests/sched_ext/dequeue_iter.bpf.c @@ -0,0 +1,73 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * ops.dequeue() of this scheduler iterates the user DSQ it consumes + * tasks from with bpf_iter_scx_dsq, which takes the DSQ lock. + * On a kernel that still runs ops.dequeue() with that lock held, the + * iteration self-deadlocks the CPU - this test wedges the system on + * unfixed kernels instead of failing cleanly. + * + * Copyright (c) 2026 fangqiurong + */ + +#include + +char _license[] SEC("license") = "GPL"; + +UEI_DEFINE(uei); + +#define TEST_DSQ_ID 1000 + +u64 dq_count; + +s32 BPF_STRUCT_OPS_SLEEPABLE(dequeue_iter_init) +{ + return scx_bpf_create_dsq(TEST_DSQ_ID, -1); +} + +s32 BPF_STRUCT_OPS(dequeue_iter_select_cpu, struct task_struct *p, + s32 prev_cpu, u64 wake_flags) +{ + return prev_cpu; +} + +void BPF_STRUCT_OPS(dequeue_iter_enqueue, struct task_struct *p, u64 enq_flags) +{ + scx_bpf_dsq_insert(p, TEST_DSQ_ID, SCX_SLICE_DFL, enq_flags); +} + +void BPF_STRUCT_OPS(dequeue_iter_dispatch, s32 cpu, struct task_struct *task) +{ + scx_bpf_dsq_move_to_local(TEST_DSQ_ID, 0); +} + +void BPF_STRUCT_OPS(dequeue_iter_dequeue, struct task_struct *p, u64 deq_flags) +{ + struct bpf_iter_scx_dsq it; + struct task_struct *t; + + if (!bpf_iter_scx_dsq_new(&it, TEST_DSQ_ID, 0)) { + while ((t = bpf_iter_scx_dsq_next(&it))) + ; + } + bpf_iter_scx_dsq_destroy(&it); + + __sync_fetch_and_add(&dq_count, 1); +} + +void BPF_STRUCT_OPS(dequeue_iter_exit, struct scx_exit_info *ei) +{ + UEI_RECORD(uei, ei); + scx_bpf_destroy_dsq(TEST_DSQ_ID); +} + +SEC(".struct_ops.link") +struct sched_ext_ops dequeue_iter_ops = { + .init = (void *)dequeue_iter_init, + .select_cpu = (void *)dequeue_iter_select_cpu, + .enqueue = (void *)dequeue_iter_enqueue, + .dispatch = (void *)dequeue_iter_dispatch, + .dequeue = (void *)dequeue_iter_dequeue, + .exit = (void *)dequeue_iter_exit, + .timeout_ms = 1000U, + .name = "dequeue_iter", +}; diff --git a/tools/testing/selftests/sched_ext/dequeue_iter.c b/tools/testing/selftests/sched_ext/dequeue_iter.c new file mode 100644 index 000000000000..f60711bf3b39 --- /dev/null +++ b/tools/testing/selftests/sched_ext/dequeue_iter.c @@ -0,0 +1,79 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Copyright (c) 2026 fangqiurong + */ +#include +#include +#include +#include +#include "dequeue_iter.bpf.skel.h" +#include "scx_test.h" + +#define DQ_TARGET 10 +#define DQ_DEADLINE_MS 3000 + +static unsigned long long now_ms(void) +{ + struct timespec ts; + + clock_gettime(CLOCK_MONOTONIC, &ts); + + return ts.tv_sec * 1000ULL + ts.tv_nsec / 1000000; +} + +static enum scx_test_status setup(void **ctx) +{ + struct dequeue_iter *skel; + + skel = dequeue_iter__open(); + SCX_FAIL_IF(!skel, "Failed to open"); + SCX_ENUM_INIT(skel); + SCX_FAIL_IF(dequeue_iter__load(skel), "Failed to load skel"); + + *ctx = skel; + + return SCX_TEST_PASS; +} + +static enum scx_test_status run(void *ctx) +{ + struct dequeue_iter *skel = ctx; + struct bpf_link *link; + unsigned long long end; + + link = bpf_map__attach_struct_ops(skel->maps.dequeue_iter_ops); + SCX_FAIL_IF(!link, "Failed to attach scheduler"); + + end = now_ms() + DQ_DEADLINE_MS; + while (skel->bss->dq_count < DQ_TARGET && !UEI_EXITED(skel, uei) && + now_ms() < end) + usleep(100); + + bpf_link__destroy(link); + + SCX_EQ(skel->data->uei.kind, EXIT_KIND(SCX_EXIT_UNREG)); + + if (skel->bss->dq_count < DQ_TARGET) { + SCX_ERR("ops.dequeue() fired only %llu times", + (unsigned long long)skel->bss->dq_count); + return SCX_TEST_FAIL; + } + + return SCX_TEST_PASS; +} + +static void cleanup(void *ctx) +{ + struct dequeue_iter *skel = ctx; + + dequeue_iter__destroy(skel); +} + +struct scx_test dequeue_iter = { + .name = "dequeue_iter", + .description = "Verify ops.dequeue() can iterate its source user DSQ", + .setup = setup, + .run = run, + .cleanup = cleanup, +}; +REGISTER_SCX_TEST(&dequeue_iter) From 40c2096961b4e8f48d1fc17406a90c5a36ac0a8d Mon Sep 17 00:00:00 2001 From: Kumar Kartikeya Dwivedi Date: Mon, 14 Sep 2026 15:19:20 +0200 Subject: [PATCH 135/694] bpf: Verify global subprogs in each sleepability context Global subprograms are verified independently with a fresh verifier root. do_check_common() currently seeds that root's in_sleepable state from the program, even though a global subprogram can also run from callbacks whose execution context differs from the program's main entry point. In particular, workqueue and task-work callbacks are sleepable even when the containing program is not. A global subprogram of that program is therefore verified as non-sleepable, making in_rcu_cs() true and allowing loads of RCU-protected kptrs to produce trusted MEM_RCU pointers. The same subprogram can then be called from a sleepable callback without a classic RCU reader. It can retain such a pointer while the object is freed and use it after free. The verifier's execution-context predicates are complementary. A state is sleepable only when in_sleepable is set and no RCU, preemption, IRQ, or lock region is active. Each condition which prevents sleeping also provides RCU protection, while in_rcu_cs() treats a non-sleepable state as implicitly protected. Use this relationship to represent a global subprogram caller with only the result of in_sleepable_context(). A protected sleepable caller is normalized to in_sleepable=false at the independent verification root. This both prevents sleepable operations and makes in_rcu_cs() true without copying caller-owned lock state. Track only the contexts in which each global subprogram is actually reached. Verify it once if all reachable calls use the same context, and twice only if both sleepable and non-sleepable calls reach it. Calls found while verifying globals or asynchronous callbacks mark further contexts for checking. Repeat the existing subprogram walk until all called contexts have been verified; unreachable global calls remain unchecked. Accumulate instruction counts over those verification passes. Preserve the total recorded before each pass, since path accounting has already added this pass's synchronous instructions and its root total must also include asynchronous subprograms. This makes an unprotected callback verify the global subprogram as sleepable, turning its RCU-protected kptr load into an untrusted pointer. Protected callers and global subprograms which do not depend on implicit RCU protection remain valid. Fixes: 81f1d7a583fa ("bpf: wq: add bpf_wq_set_callback_impl") Fixes: 38aa7003e369 ("bpf: task work scheduling kfuncs") Reported-by: Nicholas Carlini Suggested-by: Nicholas Carlini Signed-off-by: Kumar Kartikeya Dwivedi Acked-by: Eduard Zingerman Link: https://patch.msgid.link/20260914131923.2544250-2-memxor@gmail.com Signed-off-by: Eduard Zingerman --- include/linux/bpf.h | 5 +-- kernel/bpf/verifier.c | 71 ++++++++++++++++++++++--------------------- 2 files changed, 40 insertions(+), 36 deletions(-) diff --git a/include/linux/bpf.h b/include/linux/bpf.h index e57af902560c..1d2676782d70 100644 --- a/include/linux/bpf.h +++ b/include/linux/bpf.h @@ -1651,8 +1651,9 @@ static inline void bpf_trampoline_set_flags(struct bpf_trampoline *tr, u32 flags struct bpf_func_info_aux { u16 linkage; bool unreliable; - bool called : 1; - bool verified : 1; + /* Indexed by in_sleepable. */ + bool called[2]; + bool verified[2]; }; enum bpf_jit_poke_reason { diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index ddba53eaa333..5d7080c260d8 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -9931,6 +9931,7 @@ static int check_func_call(struct bpf_verifier_env *env, struct bpf_insn *insn, if (err == -EFAULT) return err; if (bpf_subprog_is_global(env, subprog)) { + struct bpf_func_info_aux *sub_aux = subprog_aux(env, subprog); const char *sub_name = bpf_subprog_name(env, subprog); const char *operation; bool returns_void; @@ -9962,11 +9963,10 @@ static int check_func_call(struct bpf_verifier_env *env, struct bpf_insn *insn, if (env->log.level & BPF_LOG_LEVEL) verbose(env, "Func#%d ('%s') is global and assumed valid.\n", subprog, sub_name); + sub_aux->called[in_sleepable_context(env)] = true; returns_void = subprog_returns_void(env, subprog); if (env->subprog_info[subprog].changes_pkt_data) clear_all_pkt_pointers(env); - /* mark global subprog for verifying after main prog */ - subprog_aux(env, subprog)->called = true; if (returns_void) bpf_diag_record_scrub(env, &caller->regs[BPF_REG_0], BPF_DIAG_MOD_CALLER_SAVED); else @@ -10784,11 +10784,7 @@ int bpf_get_helper_proto(struct bpf_verifier_env *env, int func_id, /* Check if we're in a sleepable context. */ static inline bool in_sleepable_context(struct bpf_verifier_env *env) { - return !env->cur_state->active_rcu_locks && - !env->cur_state->active_preempt_locks && - !env->cur_state->active_locks && - !env->cur_state->active_irq_id && - in_sleepable(env); + return !in_rcu_cs(env); } static const char *non_sleepable_context_description(struct bpf_verifier_env *env) @@ -19447,13 +19443,14 @@ static void free_states(struct bpf_verifier_env *env) } } -static int do_check_common(struct bpf_verifier_env *env, int subprog) +static int do_check_common(struct bpf_verifier_env *env, int subprog, bool is_sleepable) { bool pop_log = !(env->log.level & BPF_LOG_LEVEL2); struct bpf_subprog_info *sub = subprog_info(env, subprog); struct bpf_prog_aux *aux = env->prog->aux; struct bpf_verifier_state *state; struct bpf_reg_state *regs; + u32 old_insns_total = sub->insns_total; u32 insn_processed = env->insn_processed; int ret, i; @@ -19466,7 +19463,7 @@ static int do_check_common(struct bpf_verifier_env *env, int subprog) state->curframe = 0; state->speculative = false; state->branches = 1; - state->in_sleepable = env->prog->sleepable; + state->in_sleepable = is_sleepable; state->frame[0] = kzalloc_obj(struct bpf_func_state, GFP_KERNEL_ACCOUNT); if (!state->frame[0]) { kfree(state); @@ -19607,8 +19604,10 @@ static int do_check_common(struct bpf_verifier_env *env, int subprog) * not accounted as callees by account_current_path(). * Accumulate their total counts as total counts of the main or * global subprog hosting the async call. + * Start from the saved total of earlier contexts: adding to the current + * total would count this pass's synchronous paths twice. */ - env->subprog_info[subprog].insns_total = env->insn_processed - insn_processed; + sub->insns_total = old_insns_total + (env->insn_processed - insn_processed); return ret; } @@ -19636,14 +19635,19 @@ static int do_check_subprogs(struct bpf_verifier_env *env) { struct bpf_prog_aux *aux = env->prog->aux; struct bpf_func_info_aux *sub_aux; - int i, ret, new_cnt; + int context, i, ret, new_cnt; if (!aux->func_info) return 0; - /* exception callback is presumed to be always called */ - if (env->exception_callback_subprog) - subprog_aux(env, env->exception_callback_subprog)->called = true; + /* + * Callbacks cannot throw, so the exception callback always runs in the + * main program's context. It is presumed to be always called. + */ + if (env->exception_callback_subprog) { + sub_aux = subprog_aux(env, env->exception_callback_subprog); + sub_aux->called[env->prog->sleepable] = true; + } again: new_cnt = 0; @@ -19652,29 +19656,28 @@ static int do_check_subprogs(struct bpf_verifier_env *env) continue; sub_aux = subprog_aux(env, i); - if (!sub_aux->called || sub_aux->verified) - continue; + for (context = 0; context < ARRAY_SIZE(sub_aux->called); context++) { + if (!sub_aux->called[context] || sub_aux->verified[context]) + continue; - env->insn_idx = env->subprog_info[i].start; - WARN_ON_ONCE(env->insn_idx == 0); - ret = do_check_common(env, i); - if (ret) { - return ret; - } else if (env->log.level & BPF_LOG_LEVEL) { - verbose(env, "Func#%d ('%s') is safe for any args that match its prototype\n", - i, bpf_subprog_name(env, i)); + env->insn_idx = env->subprog_info[i].start; + WARN_ON_ONCE(env->insn_idx == 0); + ret = do_check_common(env, i, context); + if (ret) + return ret; + if (env->log.level & BPF_LOG_LEVEL) + verbose(env, "Func#%d ('%s') is safe for any args " + "that match its prototype\n", + i, bpf_subprog_name(env, i)); + + sub_aux->verified[context] = true; + new_cnt++; } - - /* We verified new global subprog, it might have called some - * more global subprogs that we haven't verified yet, so we - * need to do another pass over subprogs to verify those. - */ - sub_aux->verified = true; - new_cnt++; } - /* We can't loop forever as we verify at least one global subprog on - * each pass. + /* + * We can't loop forever as each pass verifies at least one new context, + * and there are only two contexts per global subprog. */ if (new_cnt) goto again; @@ -19687,7 +19690,7 @@ static int do_check_main(struct bpf_verifier_env *env) int ret; env->insn_idx = 0; - ret = do_check_common(env, 0); + ret = do_check_common(env, 0, env->prog->sleepable); if (!ret) env->prog->aux->stack_depth = env->subprog_info[0].stack_depth; return ret; From a452e729b7be317837bb2157d5d88aa3de9873e6 Mon Sep 17 00:00:00 2001 From: Kumar Kartikeya Dwivedi Date: Mon, 14 Sep 2026 15:19:21 +0200 Subject: [PATCH 136/694] selftests/bpf: Test global subprog callback contexts Exercise global subprogram verification from workqueue callbacks, which can run in a sleepable context even when the containing program is not sleepable. An unprotected callback must not let the global subprogram use implicit RCU protection inherited from the program. Add a negative case which loads an RCU-protected task kptr in a global subprogram reached from a workqueue callback. It fails on an unfixed kernel because the program is incorrectly accepted. Also cover a workqueue callback protected by an explicit RCU read-side critical section, where the same global subprogram remains valid. Call the same harmless global subprogram directly from the main program and from an unprotected callback. Mark it __weak __noinline so both calls survive optimization, and check that its instruction statistics account for both verification contexts. This also verifies that global calls from callbacks are not rejected wholesale. Workqueue callbacks return zero explicitly after the global call, as required by their contract. Signed-off-by: Kumar Kartikeya Dwivedi Acked-by: Eduard Zingerman Link: https://patch.msgid.link/20260914131923.2544250-3-memxor@gmail.com Signed-off-by: Eduard Zingerman --- .../bpf/progs/verifier_async_cb_context.c | 106 ++++++++++++++++++ 1 file changed, 106 insertions(+) diff --git a/tools/testing/selftests/bpf/progs/verifier_async_cb_context.c b/tools/testing/selftests/bpf/progs/verifier_async_cb_context.c index e0926767bbd3..1b653bfb63eb 100644 --- a/tools/testing/selftests/bpf/progs/verifier_async_cb_context.c +++ b/tools/testing/selftests/bpf/progs/verifier_async_cb_context.c @@ -9,6 +9,11 @@ char _license[] SEC("license") = "GPL"; +struct task_struct *bpf_task_acquire(struct task_struct *p) __ksym; +void bpf_task_release(struct task_struct *p) __ksym; +void bpf_rcu_read_lock(void) __ksym; +void bpf_rcu_read_unlock(void) __ksym; + /* Timer tests */ struct timer_elem { @@ -164,6 +169,7 @@ int syscall_btf_find_prog(void *ctx) struct wq_elem { struct bpf_wq w; + struct task_struct __kptr *task; }; struct { @@ -217,6 +223,106 @@ int wq_sleepable_prog(void *ctx) return 0; } +__noinline int wq_global_acquire(void) +{ + struct task_struct *task, *acquired; + struct wq_elem *val; + int key = 0; + + val = bpf_map_lookup_elem(&wq_map, &key); + if (!val) + return 0; + + task = val->task; + if (!task) + return 0; + + acquired = bpf_task_acquire(task); + if (acquired) + bpf_task_release(acquired); + return 0; +} + +static int wq_global_rcu_cb(void *map, int *key, void *value) +{ + wq_global_acquire(); + return 0; +} + +SEC("fentry/bpf_fentry_test1") +__failure __msg("R1 must be a rcu pointer") +int wq_global_rcu_prog(void *ctx) +{ + struct wq_elem *val; + int key = 0; + + val = bpf_map_lookup_elem(&wq_map, &key); + if (!val) + return 0; + + bpf_wq_init(&val->w, &wq_map, 0); + bpf_wq_set_callback(&val->w, wq_global_rcu_cb, 0); + return 0; +} + +static int wq_global_rcu_lock_cb(void *map, int *key, void *value) +{ + bpf_rcu_read_lock(); + wq_global_acquire(); + bpf_rcu_read_unlock(); + return 0; +} + +SEC("fentry/bpf_fentry_test1") +__success +int wq_global_rcu_lock_prog(void *ctx) +{ + struct wq_elem *val; + int key = 0; + + /* Verify the same global subprog in non-sleepable and protected contexts. */ + wq_global_acquire(); + + val = bpf_map_lookup_elem(&wq_map, &key); + if (!val) + return 0; + + bpf_wq_init(&val->w, &wq_map, 0); + bpf_wq_set_callback(&val->w, wq_global_rcu_lock_cb, 0); + return 0; +} + +__weak __noinline int wq_global_no_rcu(void) +{ + return 0; +} + +static int wq_global_no_rcu_cb(void *map, int *key, void *value) +{ + wq_global_no_rcu(); + return 0; +} + +SEC("fentry/bpf_fentry_test1") +__success __log_level(4) +__msg("subprog {{[0-9]+}} (wq_global_no_rcu) global insns_self 4 insns_total 4 stack 0") +int wq_global_no_rcu_prog(void *ctx) +{ + struct wq_elem *val; + int key = 0; + + /* Verify the same global in non-sleepable and unprotected contexts. */ + wq_global_no_rcu(); + + val = bpf_map_lookup_elem(&wq_map, &key); + if (!val) + return 0; + + bpf_wq_init(&val->w, &wq_map, 0); + bpf_wq_set_callback(&val->w, wq_global_no_rcu_cb, 0); + return 0; +} + /* Task work tests */ struct task_work_elem { From 70504de0bb627848667207bec7ccfd647deb8814 Mon Sep 17 00:00:00 2001 From: Zhiling Zou Date: Fri, 11 Sep 2026 00:18:24 +0800 Subject: [PATCH 137/694] xsk: Use a 32-bit compare in xsk_map_gen_lookup xsk_map_gen_lookup() loads a u32 key and compares it with max_entries using BPF_JMP_IMM. BPF immediates are sign-extended to 64 bits, so a max_entries value of 0x80000000 or higher becomes a threshold larger than every zero-extended 32-bit key. An out-of-range index then skips the bounds check and the generated lookup reads past xsk_map[]. Compare with BPF_JMP32_IMM so the check stays in 32-bit unsigned range. Fixes: e65650f291ee ("bpf: Implement map_gen_lookup() callback for XSKMAP") Reported-by: Vega Signed-off-by: Zhiling Zou Signed-off-by: Alexei Starovoitov Reviewed-by: Emil Tsalapatis Link: https://patch.msgid.link/7d2cb8e8dfaa9eb8fdff85156987a60960787dc3.1789056660.git.zhilinz@nebusec.ai Signed-off-by: Eduard Zingerman --- net/xdp/xskmap.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/xdp/xskmap.c b/net/xdp/xskmap.c index 3bff346308d0..bf00d6463c19 100644 --- a/net/xdp/xskmap.c +++ b/net/xdp/xskmap.c @@ -124,7 +124,7 @@ static int xsk_map_gen_lookup(struct bpf_map *map, struct bpf_insn *insn_buf) struct bpf_insn *insn = insn_buf; *insn++ = BPF_LDX_MEM(BPF_W, ret, index, 0); - *insn++ = BPF_JMP_IMM(BPF_JGE, ret, map->max_entries, 5); + *insn++ = BPF_JMP32_IMM(BPF_JGE, ret, map->max_entries, 5); *insn++ = BPF_ALU64_IMM(BPF_LSH, ret, ilog2(sizeof(struct xsk_sock *))); *insn++ = BPF_ALU64_IMM(BPF_ADD, mp, offsetof(struct xsk_map, xsk_map)); *insn++ = BPF_ALU64_REG(BPF_ADD, ret, mp); From 5ea72f7b7139b123713a7983448f910bc4514d9e Mon Sep 17 00:00:00 2001 From: Wentao Liang Date: Wed, 16 Sep 2026 18:02:02 +0000 Subject: [PATCH 138/694] drm/nouveau: Fix gem reference leak in validate_init() On the ttm_bo_reserve() failure and "vma not found" error paths, the loop breaks without adding the looked-up object to any validate list, so the reference taken by drm_gem_object_lookup() is never released; validate_fini() only walks the spliced lists. Drop the reference before breaking out on both paths. Fixes: 19ca10d82e33bcfe ("drm/nouveau/gem: lookup VMAs for buffers referenced by pushbuf ioctl") Cc: stable@vger.kernel.org Signed-off-by: Wentao Liang Reviewed-by: Lyude Paul Signed-off-by: Lyude Paul Link: https://patch.msgid.link/20260916180202.2090231-1-vulab@iscas.ac.cn --- drivers/gpu/drm/nouveau/nouveau_gem.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/gpu/drm/nouveau/nouveau_gem.c b/drivers/gpu/drm/nouveau/nouveau_gem.c index 0b7123b163e4..51188be57221 100644 --- a/drivers/gpu/drm/nouveau/nouveau_gem.c +++ b/drivers/gpu/drm/nouveau/nouveau_gem.c @@ -522,6 +522,7 @@ validate_init(struct nouveau_channel *chan, struct drm_file *file_priv, if (unlikely(ret)) { if (ret != -ERESTARTSYS) NV_PRINTK(err, cli, "fail reserve\n"); + drm_gem_object_put(gem); break; } } @@ -531,6 +532,7 @@ validate_init(struct nouveau_channel *chan, struct drm_file *file_priv, struct nouveau_vma *vma = nouveau_vma_find(nvbo, vmm); if (!vma) { NV_PRINTK(err, cli, "vma not found!\n"); + drm_gem_object_put(gem); ret = -EINVAL; break; } From 1e04611d3735543bd80a67d9d13dc13f503746fb Mon Sep 17 00:00:00 2001 From: Wentao Liang Date: Wed, 16 Sep 2026 18:03:42 +0000 Subject: [PATCH 139/694] drm/nouveau: Fix runtime PM leak in nouveau_connector_detect() If nvif_outp_edid_get() fails, nouveau_connector_detect() returns early without dropping the runtime PM reference taken at the start of the function, keeping the device powered on until the next successful detect. Balance the reference on the error path like the other exit paths do. Fixes: 0cd7e0718139 ("drm/nouveau/disp: add output method to fetch edid") Cc: stable@vger.kernel.org Signed-off-by: Wentao Liang Reviewed-by: Lyude Paul Signed-off-by: Lyude Paul Link: https://patch.msgid.link/20260916180342.2090360-1-vulab@iscas.ac.cn --- drivers/gpu/drm/nouveau/nouveau_connector.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/nouveau/nouveau_connector.c b/drivers/gpu/drm/nouveau/nouveau_connector.c index b0b0ad9a0c24..4cfc9c7c2ae0 100644 --- a/drivers/gpu/drm/nouveau/nouveau_connector.c +++ b/drivers/gpu/drm/nouveau/nouveau_connector.c @@ -600,8 +600,11 @@ nouveau_connector_detect(struct drm_connector *connector, bool force) new_edid = drm_get_edid(connector, nv_encoder->i2c); } else { ret = nvif_outp_edid_get(&nv_encoder->outp, (u8 **)&new_edid); - if (ret < 0) + if (ret < 0) { + pm_runtime_mark_last_busy(dev->dev); + pm_runtime_put_autosuspend(dev->dev); return connector_status_disconnected; + } } nouveau_connector_set_edid(nv_connector, new_edid); From 97077ac87afe9e91ec074ef0be64454e7ccbf344 Mon Sep 17 00:00:00 2001 From: Peiyang He Date: Wed, 16 Sep 2026 18:31:38 +0800 Subject: [PATCH 140/694] drm/nouveau: fix double-free in nvif_vmm_dtor On failure, nouveau_cli_init() calls nouveau_cli_fini() to tear the client down. Then, nouveau_drm_open() also enters into its cleanup path and calls nouveau_cli_fini() AGAIN. nouveau_cli_fini() calls nouveau_vmm_fini(): void nouveau_vmm_fini(struct nouveau_vmm *vmm) { nouveau_svmm_fini(&vmm->svmm); nvif_vmm_dtor(&vmm->vmm); vmm->cli = NULL; } Inside nvif_vmm_dtor(), vmm->page is freed unconditionally: void nvif_vmm_dtor(struct nvif_vmm *vmm) { kfree(vmm->page); nvif_object_dtor(&vmm->object); } vmm->page is never cleared after being freed, so the second call of nvif_vmm_dtor() will cause a double-free. Found by fuzzing the nouveau driver with a modified Syzkaller: BUG: KASAN: double-free in nvif_vmm_dtor+0x31/0x50 drivers/gpu/drm/nouveau/nvif/vmm.c:194 Free of addr ffff888010fcdc30 by task syz.0.173/2567 CPU: 1 UID: 0 PID: 2567 Comm: syz.0.173 Not tainted 7.2.0 #24 PREEMPT(lazy) Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.16.3-debian-1.16.3-2 04/01/2014 Call Trace: __dump_stack lib/dump_stack.c:94 [inline] dump_stack_lvl+0x95/0xe0 lib/dump_stack.c:120 print_address_description mm/kasan/report.c:378 [inline] print_report+0xcb/0x5a0 mm/kasan/report.c:482 kasan_report_invalid_free+0xaa/0xd0 mm/kasan/report.c:557 check_slab_allocation+0xe4/0x110 mm/kasan/common.c:235 kasan_slab_pre_free include/linux/kasan.h:199 [inline] slab_free_hook mm/slub.c:2622 [inline] slab_free mm/slub.c:6377 [inline] kfree+0x192/0x590 mm/slub.c:6692 nvif_vmm_dtor+0x31/0x50 drivers/gpu/drm/nouveau/nvif/vmm.c:194 nouveau_vmm_fini+0x16/0x50 drivers/gpu/drm/nouveau/nouveau_vmm.c:127 nouveau_cli_fini+0x10e/0x210 drivers/gpu/drm/nouveau/nouveau_drm.c:225 nouveau_drm_open+0x24e/0x740 drivers/gpu/drm/nouveau/nouveau_drm.c:1255 drm_file_alloc+0x5f2/0xad0 drivers/gpu/drm/drm_file.c:176 drm_open_helper+0x1d7/0x4a0 drivers/gpu/drm/drm_file.c:335 drm_open+0x190/0x3d0 drivers/gpu/drm/drm_file.c:388 drm_stub_open+0x1f2/0x460 drivers/gpu/drm/drm_drv.c:1211 chrdev_open+0x21c/0x660 fs/char_dev.c:411 do_dentry_open+0x59d/0x12b0 fs/open.c:947 vfs_open+0x82/0x390 fs/open.c:1052 do_open fs/namei.c:4700 [inline] path_openat+0x2345/0x3420 fs/namei.c:4863 do_file_open+0x207/0x460 fs/namei.c:4892 do_sys_openat2+0xd1/0x1d0 fs/open.c:1368 do_sys_open fs/open.c:1374 [inline] __do_sys_openat fs/open.c:1390 [inline] __se_sys_openat fs/open.c:1385 [inline] __x64_sys_openat+0x144/0x200 fs/open.c:1385 do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline] do_syscall_64+0x115/0x690 arch/x86/entry/syscall_64.c:94 entry_SYSCALL_64_after_hwframe+0x77/0x7f RIP: 0033:0x7fc6d687594d Code: ff c3 66 2e 0f 1f 84 00 00 00 00 00 90 f3 0f 1e fa 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 c7 c1 b0 ff ff ff f7 d8 64 89 01 48 RSP: 002b:00007fc6d5295008 EFLAGS: 00000246 ORIG_RAX: 0000000000000101 RAX: ffffffffffffffda RBX: 00007fc6d6b06180 RCX: 00007fc6d687594d RDX: 0000000000022501 RSI: 0000200000000000 RDI: ffffffffffffff9c RBP: 00007fc6d691c303 R08: 0000000000000000 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000000 R13: 00007fc6d6b06218 R14: 00007fc6d6b06180 R15: 00007ffd9451d760 Allocated by task 2567 on cpu 1 at 163.593900s: kasan_save_stack+0x24/0x50 mm/kasan/common.c:57 kasan_save_track+0x17/0x60 mm/kasan/common.c:78 poison_kmalloc_redzone mm/kasan/common.c:398 [inline] __kasan_kmalloc+0xaa/0xb0 mm/kasan/common.c:415 kasan_kmalloc include/linux/kasan.h:263 [inline] __do_kmalloc_node mm/slub.c:5334 [inline] __kmalloc_noprof+0x304/0x7c0 mm/slub.c:5359 _kmalloc_noprof include/linux/slab.h:992 [inline] nvif_vmm_ctor+0x3c0/0x7e0 drivers/gpu/drm/nouveau/nvif/vmm.c:237 nouveau_vmm_init+0x40/0x90 drivers/gpu/drm/nouveau/nouveau_vmm.c:134 nouveau_cli_init+0x7b9/0xe10 drivers/gpu/drm/nouveau/nouveau_drm.c:293 nouveau_drm_open+0x236/0x740 drivers/gpu/drm/nouveau/nouveau_drm.c:1243 drm_file_alloc+0x5f2/0xad0 drivers/gpu/drm/drm_file.c:176 drm_open_helper+0x1d7/0x4a0 drivers/gpu/drm/drm_file.c:335 drm_open+0x190/0x3d0 drivers/gpu/drm/drm_file.c:388 drm_stub_open+0x1f2/0x460 drivers/gpu/drm/drm_drv.c:1211 chrdev_open+0x21c/0x660 fs/char_dev.c:411 do_dentry_open+0x59d/0x12b0 fs/open.c:947 vfs_open+0x82/0x390 fs/open.c:1052 do_open fs/namei.c:4700 [inline] path_openat+0x2345/0x3420 fs/namei.c:4863 do_file_open+0x207/0x460 fs/namei.c:4892 do_sys_openat2+0xd1/0x1d0 fs/open.c:1368 do_sys_open fs/open.c:1374 [inline] __do_sys_openat fs/open.c:1390 [inline] __se_sys_openat fs/open.c:1385 [inline] __x64_sys_openat+0x144/0x200 fs/open.c:1385 do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline] do_syscall_64+0x115/0x690 arch/x86/entry/syscall_64.c:94 entry_SYSCALL_64_after_hwframe+0x77/0x7f Freed by task 2567 on cpu 1 at 163.601355s: kasan_save_stack+0x24/0x50 mm/kasan/common.c:57 kasan_save_track+0x17/0x60 mm/kasan/common.c:78 kasan_save_free_info+0x3b/0x60 mm/kasan/generic.c:584 poison_slab_object mm/kasan/common.c:253 [inline] __kasan_slab_free+0x61/0x80 mm/kasan/common.c:285 kasan_slab_free include/linux/kasan.h:235 [inline] slab_free_hook mm/slub.c:2677 [inline] slab_free mm/slub.c:6377 [inline] kfree+0x383/0x590 mm/slub.c:6692 nvif_vmm_dtor+0x31/0x50 drivers/gpu/drm/nouveau/nvif/vmm.c:194 nouveau_vmm_fini+0x16/0x50 drivers/gpu/drm/nouveau/nouveau_vmm.c:127 nouveau_cli_fini+0x10e/0x210 drivers/gpu/drm/nouveau/nouveau_drm.c:225 nouveau_cli_init+0x593/0xe10 drivers/gpu/drm/nouveau/nouveau_drm.c:324 nouveau_drm_open+0x236/0x740 drivers/gpu/drm/nouveau/nouveau_drm.c:1243 drm_file_alloc+0x5f2/0xad0 drivers/gpu/drm/drm_file.c:176 drm_open_helper+0x1d7/0x4a0 drivers/gpu/drm/drm_file.c:335 drm_open+0x190/0x3d0 drivers/gpu/drm/drm_file.c:388 drm_stub_open+0x1f2/0x460 drivers/gpu/drm/drm_drv.c:1211 chrdev_open+0x21c/0x660 fs/char_dev.c:411 do_dentry_open+0x59d/0x12b0 fs/open.c:947 vfs_open+0x82/0x390 fs/open.c:1052 do_open fs/namei.c:4700 [inline] path_openat+0x2345/0x3420 fs/namei.c:4863 do_file_open+0x207/0x460 fs/namei.c:4892 do_sys_openat2+0xd1/0x1d0 fs/open.c:1368 do_sys_open fs/open.c:1374 [inline] __do_sys_openat fs/open.c:1390 [inline] __se_sys_openat fs/open.c:1385 [inline] __x64_sys_openat+0x144/0x200 fs/open.c:1385 do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline] do_syscall_64+0x115/0x690 arch/x86/entry/syscall_64.c:94 entry_SYSCALL_64_after_hwframe+0x77/0x7f The buggy address belongs to the object at ffff888010fcdc30 which belongs to the cache kmalloc-16 of size 16 The buggy address is located 0 bytes inside of 16-byte region [ffff888010fcdc30, ffff888010fcdc40) The buggy address belongs to the physical page: page: refcount:0 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x10fcd flags: 0x100000000000000(node=0|zone=1) page_type: f5(slab) raw: 0100000000000000 ffff88800d441640 dead000000000100 dead000000000122 raw: 0000000000000000 0000000000550055 00000000f5000000 0000000000000000 page dumped because: kasan: bad access detected Memory state around the buggy address: ffff888010fcdb00: fc fc 00 04 fc fc fc fc fa fb fc fc fc fc fa fb ffff888010fcdb80: fc fc fc fc fa fb fc fc fc fc 00 07 fc fc fc fc >ffff888010fcdc00: fa fb fc fc fc fc fa fb fc fc fc fc fa fb fc fc ^ ffff888010fcdc80: fc fc fa fb fc fc fc fc 00 04 fc fc fc fc fa fb ffff888010fcdd00: fc fc fc fc 00 00 fc fc fc fc fa fb fc fc fc fc Fix by removing the redundant teardown in nouveau_drm_open(), since nouveau_cli_init() already does the cleanup work. Also clear vmm->page after its freeing. Cc: stable@vger.kernel.org Fixes: 20d8a88e557a ("drm/nouveau: tidy up the client init/fini interfaces") Signed-off-by: Peiyang He Assisted-by: LLM Reviewed-by: Lyude Paul Signed-off-by: Lyude Paul Link: https://patch.msgid.link/03BA723D9E5FF725+20260916103138.2651605-1-peiyang_he@smail.nju.edu.cn --- drivers/gpu/drm/nouveau/nouveau_drm.c | 4 +--- drivers/gpu/drm/nouveau/nvif/vmm.c | 1 + 2 files changed, 2 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/nouveau/nouveau_drm.c b/drivers/gpu/drm/nouveau/nouveau_drm.c index 4d1ad718e09b..b4dfaf70762a 100644 --- a/drivers/gpu/drm/nouveau/nouveau_drm.c +++ b/drivers/gpu/drm/nouveau/nouveau_drm.c @@ -1250,10 +1250,8 @@ nouveau_drm_open(struct drm_device *dev, struct drm_file *fpriv) mutex_unlock(&drm->clients_lock); done: - if (ret && cli) { - nouveau_cli_fini(cli); + if (ret && cli) kfree(cli); - } pm_runtime_mark_last_busy(dev->dev); pm_runtime_put_autosuspend(dev->dev); diff --git a/drivers/gpu/drm/nouveau/nvif/vmm.c b/drivers/gpu/drm/nouveau/nvif/vmm.c index 65c3e883b119..579af70766f2 100644 --- a/drivers/gpu/drm/nouveau/nvif/vmm.c +++ b/drivers/gpu/drm/nouveau/nvif/vmm.c @@ -192,6 +192,7 @@ void nvif_vmm_dtor(struct nvif_vmm *vmm) { kfree(vmm->page); + vmm->page = NULL; nvif_object_dtor(&vmm->object); } From 64ca4cdd1031206424e6455f0ae4fb0560d8f46a Mon Sep 17 00:00:00 2001 From: Junrui Luo Date: Mon, 17 Aug 2026 14:50:40 +0800 Subject: [PATCH 141/694] drm/nouveau/dmem: pin VRAM for the whole registered range Commit c32287471077 ("gpu/drm/nouveau: enable THP support for GPU memory migration") grew the device-private region that nouveau_dmem_chunk_alloc() registers from DMEM_CHUNK_SIZE to DMEM_CHUNK_SIZE * NR_CHUNKS, but left the VRAM buffer object backing that region at DMEM_CHUNK_SIZE. nouveau_dmem_page_addr() returns chunk->bo->offset plus the page's offset within the registered region, so every page past the first chunk resolves to VRAM outside the buffer object. Size the buffer object to the region it backs. Fixes: c32287471077 ("gpu/drm/nouveau: enable THP support for GPU memory migration") Reported-by: Yuhao Jiang Assisted-by: LLM Cc: stable@vger.kernel.org Signed-off-by: Junrui Luo Reviewed-by: Lyude Paul Signed-off-by: Lyude Paul Link: https://patch.msgid.link/20260817-nouveau-fixes-v1-1-f518d0c735f3@outlook.com --- drivers/gpu/drm/nouveau/nouveau_dmem.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/nouveau/nouveau_dmem.c b/drivers/gpu/drm/nouveau/nouveau_dmem.c index ad4570c50be7..e74d7bb975a8 100644 --- a/drivers/gpu/drm/nouveau/nouveau_dmem.c +++ b/drivers/gpu/drm/nouveau/nouveau_dmem.c @@ -339,8 +339,8 @@ nouveau_dmem_chunk_alloc(struct nouveau_drm *drm, struct page **ppage, chunk->pagemap.ops = &nouveau_dmem_pagemap_ops; chunk->pagemap.owner = drm->dev; - ret = nouveau_bo_new_pin(&drm->client, NOUVEAU_GEM_DOMAIN_VRAM, DMEM_CHUNK_SIZE, - &chunk->bo); + ret = nouveau_bo_new_pin(&drm->client, NOUVEAU_GEM_DOMAIN_VRAM, + DMEM_CHUNK_SIZE * NR_CHUNKS, &chunk->bo); if (ret) goto out_release; From cb4c7603678ccef4c52b38159f2aaad867586dbc Mon Sep 17 00:00:00 2001 From: Lyude Paul Date: Thu, 17 Sep 2026 14:54:25 -0400 Subject: [PATCH 142/694] drm/nouveau/gsp/r570: Add support for INTERNAL_GCX_ENTRY_PREREQUISITE OpenRM's runtime PM handling looks a bit different then nouveau's, one part in particular that differs from us: OpenRM actually consults GSP to ask whether the GPU should be allowed to enter Gc6 and/or GcOff before runtime suspending the GPU. In the event the card isn't ready, runtime suspend is simply delayed for a few seconds before retrying. Implement the command used for querying GSP about this, NV2080_CTRL_CMD_INTERNAL_GCX_ENTRY_PREREQUISITE, and check to ensure that the GPU is ready for runtime suspend in nouveau_pmops_runtime_suspend() using this query. If the GPU can't be runtime suspended, update the last busy counter of the device and then return -EBUSY from nouveau_pmops_runtime_suspend() - essentially delaying the runtime suspend process by whatever autosuspend_delay_ms is set to. Signed-off-by: Lyude Paul Reviewed-by: Dave Airlie Link: https://patch.msgid.link/20260917185916.1089621-2-lyude@redhat.com --- drivers/gpu/drm/nouveau/include/nvif/cl0080.h | 10 ++++++ drivers/gpu/drm/nouveau/include/nvif/device.h | 1 + .../gpu/drm/nouveau/include/nvkm/subdev/gsp.h | 2 ++ drivers/gpu/drm/nouveau/nouveau_drm.c | 13 +++++++ drivers/gpu/drm/nouveau/nvif/device.c | 13 +++++++ .../gpu/drm/nouveau/nvkm/engine/device/user.c | 35 +++++++++++++++++++ .../gpu/drm/nouveau/nvkm/subdev/gsp/base.c | 10 ++++++ .../gpu/drm/nouveau/nvkm/subdev/gsp/priv.h | 1 + .../drm/nouveau/nvkm/subdev/gsp/rm/r570/gsp.c | 28 +++++++++++++++ .../nvkm/subdev/gsp/rm/r570/nvrm/gsp.h | 7 ++++ .../gpu/drm/nouveau/nvkm/subdev/gsp/rm/rm.h | 2 ++ 11 files changed, 122 insertions(+) diff --git a/drivers/gpu/drm/nouveau/include/nvif/cl0080.h b/drivers/gpu/drm/nouveau/include/nvif/cl0080.h index ea8267e0d8da..9e639df1da46 100644 --- a/drivers/gpu/drm/nouveau/include/nvif/cl0080.h +++ b/drivers/gpu/drm/nouveau/include/nvif/cl0080.h @@ -4,6 +4,7 @@ #define NV_DEVICE_V0_INFO 0x00 #define NV_DEVICE_V0_TIME 0x01 +#define NV_DEVICE_V0_GCX_READY 0x02 struct nv_device_info_v0 { __u8 version; @@ -55,6 +56,15 @@ struct nv_device_time_v0 { __u64 time; }; +#define NV_DEVICE_GC6_READY BIT(0) +#define NV_DEVICE_GCOFF_READY BIT(1) + +struct nv_device_gcx_ready_v0 { + __u8 version; + __u8 pad01[6]; + __u8 ready; +}; + #define NV_DEVICE_INFO_UNIT (0xffffffffULL << 32) #define NV_DEVICE_INFO(n) ((n) | (0x00000000ULL << 32)) #define NV_DEVICE_HOST(n) ((n) | (0x00000001ULL << 32)) diff --git a/drivers/gpu/drm/nouveau/include/nvif/device.h b/drivers/gpu/drm/nouveau/include/nvif/device.h index 7877a2a79da9..ce2fadcb05d9 100644 --- a/drivers/gpu/drm/nouveau/include/nvif/device.h +++ b/drivers/gpu/drm/nouveau/include/nvif/device.h @@ -22,4 +22,5 @@ int nvif_device_ctor(struct nvif_client *, const char *name, struct nvif_device void nvif_device_dtor(struct nvif_device *); int nvif_device_map(struct nvif_device *); u64 nvif_device_time(struct nvif_device *); +int nvif_device_gcx_ready(struct nvif_device *); #endif diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/gsp.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/gsp.h index 64fed208e4cf..cd10c3705369 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/subdev/gsp.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/gsp.h @@ -495,6 +495,8 @@ nvkm_gsp_event_dtor(struct nvkm_gsp_event *event) int nvkm_gsp_intr_stall(struct nvkm_gsp *, enum nvkm_subdev_type, int); int nvkm_gsp_intr_nonstall(struct nvkm_gsp *, enum nvkm_subdev_type, int); +int nvkm_gsp_gcx_ready(struct nvkm_gsp *gsp); + int gv100_gsp_new(struct nvkm_device *, enum nvkm_subdev_type, int, struct nvkm_gsp **); int tu102_gsp_new(struct nvkm_device *, enum nvkm_subdev_type, int, struct nvkm_gsp **); int tu116_gsp_new(struct nvkm_device *, enum nvkm_subdev_type, int, struct nvkm_gsp **); diff --git a/drivers/gpu/drm/nouveau/nouveau_drm.c b/drivers/gpu/drm/nouveau/nouveau_drm.c index b4dfaf70762a..d0eb29583cb1 100644 --- a/drivers/gpu/drm/nouveau/nouveau_drm.c +++ b/drivers/gpu/drm/nouveau/nouveau_drm.c @@ -1148,6 +1148,7 @@ nouveau_pmops_runtime_suspend(struct device *dev) { struct pci_dev *pdev = to_pci_dev(dev); struct nouveau_drm *drm = pci_get_drvdata(pdev); + struct nvif_device *nvif = &drm->client.device; int ret; if (!nouveau_pmops_runtime()) { @@ -1155,6 +1156,18 @@ nouveau_pmops_runtime_suspend(struct device *dev) return -EBUSY; } + // Check if the GPU itself is ready for runtime suspend, otherwise mark as busy and check + // again in a bit. + ret = nvif_device_gcx_ready(nvif); + if (ret < 0) { + NV_ERROR(drm, "Failed to query GCX readiness (returned %d)\n", ret); + return -EBUSY; + } else if (!(ret & NV_DEVICE_GCOFF_READY)) { + NV_DEBUG(drm, "GPU isn't ready for suspend yet, delaying...\n"); + pm_runtime_mark_last_busy(dev); + return -EBUSY; + } + nouveau_switcheroo_optimus_dsm(); ret = nouveau_do_suspend(drm, true); pci_save_state(pdev); diff --git a/drivers/gpu/drm/nouveau/nvif/device.c b/drivers/gpu/drm/nouveau/nvif/device.c index 24880931039f..1be9fbe6cb70 100644 --- a/drivers/gpu/drm/nouveau/nvif/device.c +++ b/drivers/gpu/drm/nouveau/nvif/device.c @@ -38,6 +38,19 @@ nvif_device_time(struct nvif_device *device) return device->user.func->time(&device->user); } +int +nvif_device_gcx_ready(struct nvif_device *device) +{ + struct nv_device_gcx_ready_v0 args = {}; + int ret; + + ret = nvif_object_mthd(&device->object, NV_DEVICE_V0_GCX_READY, &args, sizeof(args)); + if (ret) + return ret; + + return args.ready; +} + int nvif_device_map(struct nvif_device *device) { diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/user.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/user.c index 23d11d8221cb..f78e6b9b4292 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/device/user.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/user.c @@ -27,6 +27,7 @@ #include #include +#include #include #include @@ -189,6 +190,38 @@ nvkm_udevice_time(struct nvkm_udevice *udev, void *data, u32 size) return ret; } +static int +nvkm_udevice_gcx_ready(struct nvkm_udevice *udev, void *data, u32 size) +{ + struct nvkm_object *object = &udev->object; + struct nvkm_device *device = udev->device; + struct nvkm_gsp *gsp = device->gsp; + union { + struct nv_device_gcx_ready_v0 v0; + } *args = data; + int ret = -ENOSYS; + + if (!gsp) { + args->v0.ready = NV_DEVICE_GC6_READY | NV_DEVICE_GCOFF_READY; + return 0; + } + + nvif_ioctl(object, "device gcx ready size %d\n", size); + ret = nvif_unpack(ret, &data, &size, args->v0, 0, 0, false); + if (!ret) { + nvif_ioctl(object, "device gcx ready vers %d\n", args->v0.version); + + ret = nvkm_gsp_gcx_ready(gsp); + if (ret < 0) + return ret; + + args->v0.ready = ret; + ret = 0; + } + + return ret; +} + static int nvkm_udevice_mthd(struct nvkm_object *object, u32 mthd, void *data, u32 size) { @@ -199,6 +232,8 @@ nvkm_udevice_mthd(struct nvkm_object *object, u32 mthd, void *data, u32 size) return nvkm_udevice_info(udev, data, size); case NV_DEVICE_V0_TIME: return nvkm_udevice_time(udev, data, size); + case NV_DEVICE_V0_GCX_READY: + return nvkm_udevice_gcx_ready(udev, data, size); default: break; } diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/base.c b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/base.c index 9ba1316831e7..e475d0e8fa7b 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/base.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/base.c @@ -20,6 +20,7 @@ * OTHER DEALINGS IN THE SOFTWARE. */ #include "priv.h" +#include int nvkm_gsp_intr_nonstall(struct nvkm_gsp *gsp, enum nvkm_subdev_type type, int inst) @@ -47,6 +48,15 @@ nvkm_gsp_intr_stall(struct nvkm_gsp *gsp, enum nvkm_subdev_type type, int inst) return -ENOENT; } +int +nvkm_gsp_gcx_ready(struct nvkm_gsp *gsp) +{ + if (!gsp->rm->api->gsp->gcx_ready) + return NV_DEVICE_GC6_READY | NV_DEVICE_GCOFF_READY; + + return gsp->rm->api->gsp->gcx_ready(gsp); +} + static int nvkm_gsp_fini(struct nvkm_subdev *subdev, enum nvkm_suspend_state suspend) { diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/priv.h b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/priv.h index 71b7203bef50..b07797813b04 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/priv.h +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/priv.h @@ -93,6 +93,7 @@ void r535_gsp_dtor(struct nvkm_gsp *); int r535_gsp_oneinit(struct nvkm_gsp *); int r535_gsp_init(struct nvkm_gsp *); int r535_gsp_fini(struct nvkm_gsp *, enum nvkm_suspend_state suspend); +int r535_gsp_gcx_ready(struct nvkm_gsp *gsp); int nvkm_gsp_new_(const struct nvkm_gsp_fwif *, struct nvkm_device *, enum nvkm_subdev_type, int, struct nvkm_gsp **); diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/gsp.c b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/gsp.c index b45781cd0dfd..89b801c1e609 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/gsp.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/gsp.c @@ -10,6 +10,7 @@ #include "nvrm/gsp.h" #include "nvrm/rpcfn.h" #include "nvrm/msgfn.h" +#include "nvif/cl0080.h" #include #include @@ -215,6 +216,32 @@ r570_gsp_set_rmargs(struct nvkm_gsp *gsp, bool resume) args->bDmemStack = 1; } +int +r570_gsp_gcx_ready(struct nvkm_gsp *gsp) +{ + NV2080_CTRL_INTERNAL_GCX_ENTRY_PREREQUISITE_PARAMS *ctrl; + int ret = 0; + + ctrl = nvkm_gsp_rm_ctrl_rd(&gsp->internal.device.subdevice, + NV2080_CTRL_CMD_INTERNAL_GCX_ENTRY_PREREQUISITE, + sizeof(*ctrl)); + if (IS_ERR(ctrl)) + return PTR_ERR(ctrl); + + if (ctrl->bIsGC6Satisfied) + ret |= NV_DEVICE_GC6_READY; + if (ctrl->bIsGCOFFSatisfied) + ret |= NV_DEVICE_GCOFF_READY; + + nvkm_debug(&gsp->subdev, + "GCX ready status: GC6=%s GCOFF=%s\n", + str_yes_no(ctrl->bIsGC6Satisfied), str_yes_no(ctrl->bIsGCOFFSatisfied)); + + nvkm_gsp_rm_ctrl_done(&gsp->internal.device.subdevice, ctrl); + return ret; +} + + const struct nvkm_rm_api_gsp r570_gsp = { .set_rmargs = r570_gsp_set_rmargs, @@ -223,4 +250,5 @@ r570_gsp = { .xlat_mc_engine_idx = r570_gsp_xlat_mc_engine_idx, .drop_post_nocat_record = r570_gsp_drop_post_nocat_record, .sr_data_size = r570_gsp_sr_data_size, + .gcx_ready = r570_gsp_gcx_ready, }; diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/nvrm/gsp.h b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/nvrm/gsp.h index c458569af9d7..2814629fddd2 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/nvrm/gsp.h +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/nvrm/gsp.h @@ -639,4 +639,11 @@ typedef struct GSP_FMC_BOOT_PARAMS } GSP_FMC_BOOT_PARAMS; #define GSP_FW_HEAP_PARAM_BASE_RM_SIZE_GH100 (14 << 20) // Hopper+ + +#define NV2080_CTRL_CMD_INTERNAL_GCX_ENTRY_PREREQUISITE (0x2080a7d7) + +typedef struct NV2080_CTRL_INTERNAL_GCX_ENTRY_PREREQUISITE_PARAMS { + NvBool bIsGC6Satisfied; + NvBool bIsGCOFFSatisfied; +} NV2080_CTRL_INTERNAL_GCX_ENTRY_PREREQUISITE_PARAMS; #endif diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/rm.h b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/rm.h index e9ac47d86b69..86970129ad96 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/rm.h +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/rm.h @@ -40,6 +40,7 @@ struct nvkm_rm_api { void (*drop_send_user_shared_data)(struct nvkm_gsp *); void (*drop_post_nocat_record)(struct nvkm_gsp *); u32 (*sr_data_size)(struct nvkm_gsp *); + int (*gcx_ready)(struct nvkm_gsp *gsp); } *gsp; const struct nvkm_rm_api_rpc { @@ -174,6 +175,7 @@ int r535_gr_chan_new(struct nvkm_gr *, struct nvkm_chan *, const struct nvkm_ocl int r535_gr_promote_ctx(struct r535_gr *, bool golden, struct nvkm_vmm *, struct nvkm_memory **pctxbuf_mem, struct nvkm_vma **pctxbuf_vma, struct nvkm_gsp_object *chan); +int r570_gsp_gcx_ready(struct nvkm_gsp *gsp); extern const struct nvkm_rm_api_engine r535_nvdec; extern const struct nvkm_rm_api_engine r535_nvenc; extern const struct nvkm_rm_api_engine r535_nvjpg; From 3217000f0b7e4f67e5b386d5b3491ec1b9b584b8 Mon Sep 17 00:00:00 2001 From: Lyude Paul Date: Thu, 17 Sep 2026 14:54:26 -0400 Subject: [PATCH 143/694] drm/nouveau/gsp/r535: Add support for MEMSYS_GET_STATIC_CONFIG This is a GSP structure describing various characteristics of the memory management system that GSP provides. Start by fetching it during driver load, but don't do anything with the information we get from it just yet. Signed-off-by: Lyude Paul Reviewed-by: Dave Airlie Link: https://patch.msgid.link/20260917185916.1089621-3-lyude@redhat.com --- .../gpu/drm/nouveau/include/nvkm/subdev/gsp.h | 4 ++ .../drm/nouveau/nvkm/subdev/gsp/rm/r535/gsp.c | 17 +++++++ .../nvkm/subdev/gsp/rm/r535/nvrm/gsp.h | 45 +++++++++++++++++++ .../drm/nouveau/nvkm/subdev/gsp/rm/r570/gsp.c | 8 ++++ .../gpu/drm/nouveau/nvkm/subdev/gsp/rm/rm.h | 1 + 5 files changed, 75 insertions(+) diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/gsp.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/gsp.h index cd10c3705369..ed5c6e0e68d3 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/subdev/gsp.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/gsp.h @@ -156,6 +156,10 @@ struct nvkm_gsp { struct sg_table fbsr; } sr; + struct { + bool use_raw_mode_comptagline_alloc; + } memsys; + struct { struct nvkm_gsp_mem mem; diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r535/gsp.c b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r535/gsp.c index 94925f1590ea..63aa30f94747 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r535/gsp.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r535/gsp.c @@ -1782,6 +1782,23 @@ r535_gsp_fini(struct nvkm_gsp *gsp, enum nvkm_suspend_state suspend) return 0; } +int +r535_gsp_get_static_memsys_info(struct nvkm_gsp *gsp) +{ + NV2080_CTRL_INTERNAL_MEMSYS_GET_STATIC_CONFIG_PARAMS *ctrl; + + ctrl = nvkm_gsp_rm_ctrl_rd(&gsp->internal.device.subdevice, + NV2080_CTRL_CMD_INTERNAL_MEMSYS_GET_STATIC_CONFIG, + sizeof(*ctrl)); + if (IS_ERR(ctrl)) + return PTR_ERR(ctrl); + + gsp->memsys.use_raw_mode_comptagline_alloc = ctrl->bUseRawModeComptaglineAllocation; + + nvkm_gsp_rm_ctrl_done(&gsp->internal.device.subdevice, ctrl); + return 0; +} + int r535_gsp_init(struct nvkm_gsp *gsp) { diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r535/nvrm/gsp.h b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r535/nvrm/gsp.h index b6683a5bf870..7b10b7548c57 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r535/nvrm/gsp.h +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r535/nvrm/gsp.h @@ -782,6 +782,51 @@ typedef struct NV2080_CTRL_INTERNAL_INTR_GET_KERNEL_TABLE_PARAMS { #define GSP_FW_HEAP_PARAM_CLIENT_ALLOC_SIZE ((48 << 10) * 2048) // Support 2048 channels +typedef struct NV2080_CTRL_INTERNAL_MEMSYS_GET_STATIC_CONFIG_PARAMS { + /*! Determines if RM should use 1 to 1 Comptagline allocation policy */ + NvBool bOneToOneComptagLineAllocation; + + /*! Determines if RM should use 1 to 4 Comptagline allocation policy */ + NvBool bUseOneToFourComptagLineAllocation; + + /*! Determines if RM should use raw Comptagline allocation policy */ + NvBool bUseRawModeComptaglineAllocation; + + /*! Has COMPBIT_BACKING_SIZE been overridden to zero (i.e. disabled)? */ + NvBool bDisableCompbitBacking; + + /*! Determine if we need to disable post L2 compression */ + NvBool bDisablePostL2Compression; + + /*! Is ECC DRAM feature supported? */ + NvBool bEnabledEccFBPA; + + NvBool bL2PreFill; + + /*! L2 cache size */ + NV_DECLARE_ALIGNED(NvU64 l2CacheSize, 8); + + /*! Indicate whether fpba is present or not */ + NvBool bFbpaPresent; + + /*! Size covered by one comptag */ + NvU32 comprPageSize; + + /*! log32(comprPageSize) */ + NvU32 comprPageShift; + + /*! RAM type */ + NvU32 ramType; + + /*! LTC count */ + NvU32 ltcCount; + + /*! LTS per LTC count */ + NvU32 ltsPerLtcCount; +} NV2080_CTRL_INTERNAL_MEMSYS_GET_STATIC_CONFIG_PARAMS; + +#define NV2080_CTRL_CMD_INTERNAL_MEMSYS_GET_STATIC_CONFIG (0x20800a1c) /* finn: Evaluated from "(FINN_NV20_SUBDEVICE_0_INTERNAL_INTERFACE_ID << 8) | NV2080_CTRL_INTERNAL_MEMSYS_GET_STATIC_CONFIG_PARAMS_MESSAGE_ID" */ + typedef union rpc_message_rpc_union_field_v03_00 { NvU32 spare; diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/gsp.c b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/gsp.c index 89b801c1e609..ea38a94211f4 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/gsp.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/gsp.c @@ -138,6 +138,14 @@ r570_gsp_get_static_info(struct nvkm_gsp *gsp) } } + ret = r535_gsp_get_static_memsys_info(gsp); + if (ret) { + nvkm_error(&gsp->subdev, "Retrieving static memsys info failed\n"); + return ret; + } + nvkm_debug(&gsp->subdev, "memsys: Use raw mode for comptag allocations? %s\n", + str_yes_no(gsp->memsys.use_raw_mode_comptagline_alloc)); + return 0; } diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/rm.h b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/rm.h index 86970129ad96..17480d4e527a 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/rm.h +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/rm.h @@ -176,6 +176,7 @@ int r535_gr_promote_ctx(struct r535_gr *, bool golden, struct nvkm_vmm *, struct nvkm_memory **pctxbuf_mem, struct nvkm_vma **pctxbuf_vma, struct nvkm_gsp_object *chan); int r570_gsp_gcx_ready(struct nvkm_gsp *gsp); +int r535_gsp_get_static_memsys_info(struct nvkm_gsp *gsp); extern const struct nvkm_rm_api_engine r535_nvdec; extern const struct nvkm_rm_api_engine r535_nvenc; extern const struct nvkm_rm_api_engine r535_nvjpg; From c7ef611a43bb3ab7e7738349a860418336d507db Mon Sep 17 00:00:00 2001 From: Lyude Paul Date: Thu, 17 Sep 2026 14:54:27 -0400 Subject: [PATCH 144/694] drm/nouveau/gsp/r570: Add comp mode workaround from issue #3172217 One of the things that OpenRM does right before initiating fbsr is apply a special workaround (nvidia issue #3172217) which temporarily disables raw compression mode on the GPU. It is later re-enabled after resuming with fbsr completes. Since we don't currently save the compbit backing with fbsr, this shouldn't currently make any functional difference in the suspend/resume process. But it will be required for implementing support for saving and restoring compbit backings from the GPU. Signed-off-by: Lyude Paul Reviewed-by: Dave Airlie Link: https://patch.msgid.link/20260917185916.1089621-4-lyude@redhat.com --- .../nouveau/nvkm/subdev/gsp/rm/r570/fbsr.c | 48 +++++++++++++++++++ .../nvkm/subdev/gsp/rm/r570/nvrm/fbsr.h | 6 +++ 2 files changed, 54 insertions(+) diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/fbsr.c b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/fbsr.c index af5aa5065c3d..f73d9b29e891 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/fbsr.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/fbsr.c @@ -26,6 +26,35 @@ r570_fbsr_suspend_channels(struct nvkm_gsp *gsp, bool suspend) return nvkm_gsp_rm_ctrl_wr(&gsp->internal.device.subdevice, ctrl); } +static int +r570_memsys_enable_raw_comp_mode(struct nvkm_gsp *gsp, bool enable) +{ + NV2080_CTRL_INTERNAL_MEMSYS_PROGRAM_RAW_COMPRESSION_MODE_PARAMS *ctrl; + int ret; + + ctrl = nvkm_gsp_rm_ctrl_get(&gsp->internal.device.subdevice, + NV2080_CTRL_CMD_INTERNAL_MEMSYS_PROGRAM_RAW_COMPRESSION_MODE, + sizeof(*ctrl)); + if (IS_ERR(ctrl)) + return PTR_ERR(ctrl); + + ctrl->bRawMode = enable; + + ret = nvkm_gsp_rm_ctrl_wr(&gsp->internal.device.subdevice, ctrl); + if (!ret) + nvkm_debug(&gsp->subdev, "memsys: Raw compression mode %s\n", + str_enabled_disabled(enable)); + + return ret; +} + +static bool +r570_need_raw_comp_war(struct nvkm_gsp *gsp, struct nvkm_device *device) +{ + return (device->card_type == GA100 || device->card_type == AD100) && + gsp->memsys.use_raw_mode_comptagline_alloc; +} + static void r570_fbsr_resume(struct nvkm_gsp *gsp) { @@ -33,6 +62,7 @@ r570_fbsr_resume(struct nvkm_gsp *gsp) struct nvkm_instmem *imem = device->imem; struct nvkm_instobj *iobj; struct nvkm_vmm *vmm; + int ret; /* Restore BAR2 page tables via BAR0 window, and re-enable BAR2. */ list_for_each_entry(iobj, &imem->boot, head) { @@ -54,6 +84,13 @@ r570_fbsr_resume(struct nvkm_gsp *gsp) vmm = nvkm_bar_bar1_vmm(device); vmm->func->flush(vmm, 0); + /* Re-enable raw mode if it was previously disabled */ + if (r570_need_raw_comp_war(gsp, device)) { + ret = r570_memsys_enable_raw_comp_mode(gsp, true); + if (ret) + nvkm_error(&gsp->subdev, "Failed to re-enable raw comp mode\n"); + } + /* Resume channel scheduling. */ r570_fbsr_suspend_channels(device->gsp, false); @@ -104,6 +141,17 @@ r570_fbsr_suspend(struct nvkm_gsp *gsp) /* Stop channel scheduling. */ r570_fbsr_suspend_channels(gsp, true); + /* Temporarily disable raw mode to prevent FBSR restore operations from corrupting + * compressed surfaces. Required for ampere and ada. + * + * Nvidia bug #3172217 + */ + if (r570_need_raw_comp_war(gsp, device)) { + ret = r570_memsys_enable_raw_comp_mode(gsp, false); + if (ret) + return ret; + } + /* Save BAR2 allocations to system memory. */ list_for_each_entry(iobj, &imem->list, head) { if (iobj->preserve) { diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/nvrm/fbsr.h b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/nvrm/fbsr.h index 8af432375f7a..9050a8274b27 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/nvrm/fbsr.h +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/nvrm/fbsr.h @@ -16,4 +16,10 @@ typedef struct NV2080_CTRL_INTERNAL_FBSR_INIT_PARAMS { NV_DECLARE_ALIGNED(NvU64 sysmemAddrOfSuspendResumeData, 8); } NV2080_CTRL_INTERNAL_FBSR_INIT_PARAMS; +#define NV2080_CTRL_CMD_INTERNAL_MEMSYS_PROGRAM_RAW_COMPRESSION_MODE (0x20800a6f) /* finn: Evaluated from "(FINN_NV20_SUBDEVICE_0_INTERNAL_INTERFACE_ID << 8) | NV2080_CTRL_INTERNAL_MEMSYS_PROGRAM_RAW_COMPRESSION_MODE_PARAMS_MESSAGE_ID" */ + +typedef struct NV2080_CTRL_INTERNAL_MEMSYS_PROGRAM_RAW_COMPRESSION_MODE_PARAMS { + NvBool bRawMode; +} NV2080_CTRL_INTERNAL_MEMSYS_PROGRAM_RAW_COMPRESSION_MODE_PARAMS; + #endif From 82f4394bbe223fda560153257e5cf21bdb12b6a3 Mon Sep 17 00:00:00 2001 From: Lyude Paul Date: Thu, 17 Sep 2026 14:54:28 -0400 Subject: [PATCH 145/694] drm/nouveau/gsp/r570: Start saving comptag backing stores One of the portions of OpenRM's fbsr process that we never implemented is the saving and restoring of comptag backing stores. This isn't strictly necessary for fbsr to work (as long as we don't specify bEnteringGcOff = 1), but implementing it brings us much closer to matching OpenRM's fbsr process - which means we can rely on things being well tested on Nvidia's side. Now that we have the required driver workarounds in place and fetch the required information from GSP's memsys on driver load, let's implement support for this by fetching the required space for the compbit backing stores and adding it to the amount of memory that we allocate for fbsr. With this, we should be able to safely enable bEnteringGcOff in fbsr. Signed-off-by: Lyude Paul Reviewed-by: Dave Airlie Link: https://patch.msgid.link/20260917185916.1089621-5-lyude@redhat.com --- .../nouveau/nvkm/subdev/gsp/rm/r570/fbsr.c | 27 ++++++++++++++++++- .../nvkm/subdev/gsp/rm/r570/nvrm/fbsr.h | 23 ++++++++++++++++ 2 files changed, 49 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/fbsr.c b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/fbsr.c index f73d9b29e891..a93a39ef7c69 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/fbsr.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/fbsr.c @@ -26,6 +26,23 @@ r570_fbsr_suspend_channels(struct nvkm_gsp *gsp, bool suspend) return nvkm_gsp_rm_ctrl_wr(&gsp->internal.device.subdevice, ctrl); } +static int +r570_fb_get_compbit_store_size(struct nvkm_gsp *gsp, u64 *size) +{ + NV0080_CTRL_FB_GET_COMPBIT_STORE_INFO_PARAMS *ctrl; + + ctrl = nvkm_gsp_rm_ctrl_rd(&gsp->internal.device.object, + NV0080_CTRL_CMD_FB_GET_COMPBIT_STORE_INFO, + sizeof(*ctrl)); + if (IS_ERR(ctrl)) + return PTR_ERR(ctrl); + + *size = ctrl->Size; + + nvkm_gsp_rm_ctrl_done(&gsp->internal.device.object, ctrl); + return 0; +} + static int r570_memsys_enable_raw_comp_mode(struct nvkm_gsp *gsp, bool enable) { @@ -135,7 +152,7 @@ r570_fbsr_suspend(struct nvkm_gsp *gsp) struct nvkm_device *device = subdev->device; struct nvkm_instmem *imem = device->imem; struct nvkm_instobj *iobj; - u64 size; + u64 size, compbit_store_size; int ret; /* Stop channel scheduling. */ @@ -152,6 +169,12 @@ r570_fbsr_suspend(struct nvkm_gsp *gsp) return ret; } + ret = r570_fb_get_compbit_store_size(gsp, &compbit_store_size); + if (ret < 0) + return ret; + nvkm_debug(&gsp->subdev, "fbsr: Compbit backing store size: 0x%llx bytes\n", + compbit_store_size); + /* Save BAR2 allocations to system memory. */ list_for_each_entry(iobj, &imem->list, head) { if (iobj->preserve) { @@ -174,6 +197,8 @@ r570_fbsr_suspend(struct nvkm_gsp *gsp) size = gsp->fb.heap.size; size += gsp->fb.rsvd_size; size += gsp->fb.bios.vga_workspace.size; + size += compbit_store_size; + nvkm_debug(subdev, "fbsr: size: 0x%llx bytes\n", size); ret = nvkm_gsp_sg(device, size, &gsp->sr.fbsr); diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/nvrm/fbsr.h b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/nvrm/fbsr.h index 9050a8274b27..cb3e448415b4 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/nvrm/fbsr.h +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/nvrm/fbsr.h @@ -16,6 +16,29 @@ typedef struct NV2080_CTRL_INTERNAL_FBSR_INIT_PARAMS { NV_DECLARE_ALIGNED(NvU64 sysmemAddrOfSuspendResumeData, 8); } NV2080_CTRL_INTERNAL_FBSR_INIT_PARAMS; +#define NV0080_CTRL_CMD_FB_GET_COMPBIT_STORE_INFO (0x801306) /* finn: Evaluated from "(FINN_NV01_DEVICE_0_FB_INTERFACE_ID << 8) | NV0080_CTRL_FB_GET_COMPBIT_STORE_INFO_PARAMS_MESSAGE_ID" */ + +typedef struct NV0080_CTRL_FB_GET_COMPBIT_STORE_INFO_PARAMS { + NV_DECLARE_ALIGNED(NvU64 Size, 8); + NV_DECLARE_ALIGNED(NvU64 Address, 8); + NvU32 AddressSpace; + NvU32 MaxCompbitLine; + NvU32 comptagsPerCacheLine; + NvU32 cacheLineSize; + NvU32 cacheLineSizePerSlice; + NvU32 cacheLineFetchAlignment; + NV_DECLARE_ALIGNED(NvU64 backingStoreBase, 8); + NvU32 gobsPerComptagPerSlice; + NvU32 backingStoreCbcBase; + NvU32 comptaglineAllocationPolicy; + NV_DECLARE_ALIGNED(NvU64 privRegionStartOffset, 8); + NvU32 cbcCoveragePerSlice; +} NV0080_CTRL_FB_GET_COMPBIT_STORE_INFO_PARAMS; + +#define NV0080_CTRL_CMD_FB_GET_COMPBIT_STORE_INFO_ADDRESS_SPACE_UNKNOWN 0 // ADDR_UNKNOWN +#define NV0080_CTRL_CMD_FB_GET_COMPBIT_STORE_INFO_ADDRESS_SPACE_SYSMEM 1 // ADDR_SYSMEM +#define NV0080_CTRL_CMD_FB_GET_COMPBIT_STORE_INFO_ADDRESS_SPACE_FBMEM 2 // ADDR_FBMEM + #define NV2080_CTRL_CMD_INTERNAL_MEMSYS_PROGRAM_RAW_COMPRESSION_MODE (0x20800a6f) /* finn: Evaluated from "(FINN_NV20_SUBDEVICE_0_INTERNAL_INTERFACE_ID << 8) | NV2080_CTRL_INTERNAL_MEMSYS_PROGRAM_RAW_COMPRESSION_MODE_PARAMS_MESSAGE_ID" */ typedef struct NV2080_CTRL_INTERNAL_MEMSYS_PROGRAM_RAW_COMPRESSION_MODE_PARAMS { From adb87c20081e5ed5b6eef1270fc08b28bd77e865 Mon Sep 17 00:00:00 2001 From: Lyude Paul Date: Thu, 17 Sep 2026 14:54:29 -0400 Subject: [PATCH 146/694] drm/nouveau/gsp/r570: Enable Gcoff in fbsr again Now that we're properly saving the compbit backing stores on fbsr init, we can start setting bEnteringGcOff = 1 again without things breaking, which brings us closer to following the exact same code-paths OpenRM does for fbsr. Signed-off-by: Lyude Paul Reviewed-by: Dave Airlie Link: https://patch.msgid.link/20260917185916.1089621-6-lyude@redhat.com --- drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/fbsr.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/fbsr.c b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/fbsr.c index a93a39ef7c69..469e7eed1d6f 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/fbsr.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/gsp/rm/r570/fbsr.c @@ -135,7 +135,7 @@ r570_fbsr_init(struct nvkm_gsp *gsp, struct sg_table *sgt, u64 size) ctrl->hClient = gsp->internal.client.object.handle; ctrl->hSysMem = memlist.handle; ctrl->sysmemAddrOfSuspendResumeData = gsp->sr.meta.addr; - ctrl->bEnteringGcoffState = 0; + ctrl->bEnteringGcoffState = 1; ret = nvkm_gsp_rm_ctrl_wr(&gsp->internal.device.subdevice, ctrl); if (ret) From 814f2da6995772c07a7c9c2cc6028b6e3f3ecc9c Mon Sep 17 00:00:00 2001 From: Abdurrahman Hussain Date: Mon, 31 Aug 2026 18:29:22 -0700 Subject: [PATCH 147/694] of/overlay: put property on deadprops only after changeset add succeeds add_changeset_property() links a new property of a not-yet-live target node into the node's deadprops list before handing it to of_changeset_add_property(). If that fails, the error path frees the property but leaves the freed pointer linked in deadprops, and of_node_release() frees it a second time when the aborted overlay's node is released. Record the changeset entry first and link the property into deadprops only on success. of_changeset_add_property() never looks at the node's property lists, so the order of the two steps is otherwise immaterial, and the error path frees a property that nothing references. Fixes: f96278810150 ("of: overlay: set node fields from properties when add new overlay node") Cc: stable@vger.kernel.org Assisted-by: Claude:claude-fable-5 [Claude Code] Signed-off-by: Abdurrahman Hussain Link: https://patch.msgid.link/20260831-nh-of-alias-overlay-v7-4-02754604805a@nexthop.ai Signed-off-by: Rob Herring (Arm) --- drivers/of/overlay.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/drivers/of/overlay.c b/drivers/of/overlay.c index 08d5351746be..51df2ad26f83 100644 --- a/drivers/of/overlay.c +++ b/drivers/of/overlay.c @@ -358,12 +358,13 @@ static int add_changeset_property(struct overlay_changeset *ovcs, return -ENOMEM; if (!prop) { - if (!target->in_livetree) { + ret = of_changeset_add_property(&ovcs->cset, target->np, + new_prop); + /* the detached node owns the property until the apply */ + if (!ret && !target->in_livetree) { new_prop->next = target->np->deadprops; target->np->deadprops = new_prop; } - ret = of_changeset_add_property(&ovcs->cset, target->np, - new_prop); } else { ret = of_changeset_update_property(&ovcs->cset, target->np, new_prop); From 44081d708b5b885487bac792db78324b7b4c267d Mon Sep 17 00:00:00 2001 From: Abdurrahman Hussain Date: Mon, 31 Aug 2026 18:29:23 -0700 Subject: [PATCH 148/694] of/overlay: only treat a positive changeset id as registered of_overlay_fdt_apply() stores the idr_alloc() return value in ovcs->id before checking it. On failure the stored id is negative, free_overlay_changeset()'s "if (ovcs->id)" check passes, idr_remove() is called with a negative id and list_del() runs on ovcs->ovcs_list, which is not initialized until after the id allocation. An allocation failure at that point dereferences NULL. Make free_overlay_changeset() treat only a strict-positive id as registered. The rest of the function already copes with a partially-initialized ovcs, so the error path stays a plain goto err_free_ovcs. Fixes: 61b4de4e0b38 ("of: overlay: minor restructuring") Cc: stable@vger.kernel.org Suggested-by: Geert Uytterhoeven Assisted-by: Claude:claude-fable-5 [Claude Code] Reviewed-by: Geert Uytterhoeven Signed-off-by: Abdurrahman Hussain Link: https://patch.msgid.link/20260831-nh-of-alias-overlay-v7-5-02754604805a@nexthop.ai Signed-off-by: Rob Herring (Arm) --- drivers/of/overlay.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/of/overlay.c b/drivers/of/overlay.c index 51df2ad26f83..0cc08567c300 100644 --- a/drivers/of/overlay.c +++ b/drivers/of/overlay.c @@ -864,7 +864,8 @@ static void free_overlay_changeset(struct overlay_changeset *ovcs) if (ovcs->cset.entries.next) of_changeset_destroy(&ovcs->cset); - if (ovcs->id) { + /* a failed idr_alloc() leaves its negative error in ovcs->id */ + if (ovcs->id > 0) { idr_remove(&ovcs_idr, ovcs->id); list_del(&ovcs->ovcs_list); ovcs->id = 0; From 1d62f0f2bc7edb12ed09f71203df0e47fa4efca4 Mon Sep 17 00:00:00 2001 From: Abdurrahman Hussain Date: Mon, 31 Aug 2026 18:29:24 -0700 Subject: [PATCH 149/694] of/overlay: don't leak fragment references when changeset init fails init_overlay_changeset() stores the number of initialized fragments in ovcs->count only on full success. When find_target() or the __symbols__ lookup fails partway through, the target and overlay references taken for the fragments initialized so far are never dropped: free_overlay_changeset() bounds its cleanup loop by ovcs->count, which is still 0. Store the running count on the error path so free_overlay_changeset() puts whatever was set up. The store is guarded by ovcs->fragments because on the allocation-failure path cnt still holds the counting pass total while there is no fragments array; everywhere else cnt only counts fully-initialized fragments. Reported-by: Sashiko AI Closes: https://lore.kernel.org/20260805204009.CF3891F000E9@smtp.kernel.org Fixes: 61b4de4e0b38 ("of: overlay: minor restructuring") Assisted-by: Claude:claude-opus-4-8 [Claude Code] Signed-off-by: Abdurrahman Hussain Link: https://patch.msgid.link/20260831-nh-of-alias-overlay-v7-6-02754604805a@nexthop.ai Signed-off-by: Rob Herring (Arm) --- drivers/of/overlay.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/drivers/of/overlay.c b/drivers/of/overlay.c index 0cc08567c300..276870291656 100644 --- a/drivers/of/overlay.c +++ b/drivers/of/overlay.c @@ -854,6 +854,10 @@ static int init_overlay_changeset(struct overlay_changeset *ovcs, err_out: pr_err("%s() failed, ret = %d\n", __func__, ret); + /* let free_overlay_changeset() put the fragments set up so far */ + if (ovcs->fragments) + ovcs->count = cnt; + return ret; } From 3359a372efb6d585c97019ee1b7f1874442bcebe Mon Sep 17 00:00:00 2001 From: Peiyang He Date: Mon, 7 Sep 2026 13:22:04 +0800 Subject: [PATCH 150/694] drm/nouveau/uvmm: fix UAF in nouveau_uvmm_sm when BO is in TTM_PL_SYSTEM nouveau_uvmm_sm() calls op_map(), which passes bo->resource through nouveau_mem() to nouveau_uvma_map(). nouveau_uvmm_vmm_map() then reads mem->mem.type. But this is only valid when bo->resource is backed by struct nouveau_mem, as is the case for VRAM and TT resources. If the BO is left in TTM_PL_SYSTEM, bo->resource is only a struct ttm_resource. Treating it as struct nouveau_mem makes the mem->mem.type read past the end of the resource, causing a KASAN: slab-use-after-free Read in nouveau_uvmm_sm report: BUG: KASAN: slab-use-after-free in nouveau_uvmm_vmm_map drivers/gpu/drm/nouveau/nouveau_uvmm.c:152 [inline] BUG: KASAN: slab-use-after-free in nouveau_uvma_map drivers/gpu/drm/nouveau/nouveau_uvmm.c:199 [inline] BUG: KASAN: slab-use-after-free in op_map drivers/gpu/drm/nouveau/nouveau_uvmm.c:849 [inline] BUG: KASAN: slab-use-after-free in nouveau_uvmm_sm.constprop.0+0x6ab/0x900 drivers/gpu/drm/nouveau/nouveau_uvmm.c:903 Read of size 1 at addr ffff888127d3e3a0 by task kworker/0:1/11 CPU: 0 UID: 0 PID: 11 Comm: kworker/0:1 Not tainted 7.2.0 #5 PREEMPT(lazy) Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.16.3-debian-1.16.3-2 04/01/2014 Workqueue: nouveau_sched_wq_2224 drm_sched_run_job_work Call Trace: __dump_stack lib/dump_stack.c:94 [inline] dump_stack_lvl+0x95/0xe0 lib/dump_stack.c:120 print_address_description mm/kasan/report.c:378 [inline] print_report+0xcb/0x5a0 mm/kasan/report.c:482 kasan_report+0xca/0x100 mm/kasan/report.c:595 nouveau_uvmm_vmm_map drivers/gpu/drm/nouveau/nouveau_uvmm.c:152 [inline] nouveau_uvma_map drivers/gpu/drm/nouveau/nouveau_uvmm.c:199 [inline] op_map drivers/gpu/drm/nouveau/nouveau_uvmm.c:849 [inline] nouveau_uvmm_sm.constprop.0+0x6ab/0x900 drivers/gpu/drm/nouveau/nouveau_uvmm.c:903 nouveau_uvmm_sm_unmap drivers/gpu/drm/nouveau/nouveau_uvmm.c:932 [inline] nouveau_uvmm_bind_job_run+0xd6/0x250 drivers/gpu/drm/nouveau/nouveau_uvmm.c:1532 nouveau_job_run drivers/gpu/drm/nouveau/nouveau_sched.c:350 [inline] nouveau_sched_run_job+0x62/0xd0 drivers/gpu/drm/nouveau/nouveau_sched.c:364 drm_sched_run_job_work+0x356/0xa10 drivers/gpu/drm/scheduler/sched_main.c:1061 process_one_work+0x8a5/0x1900 kernel/workqueue.c:3322 process_scheduled_works kernel/workqueue.c:3405 [inline] worker_thread+0x5dd/0xd80 kernel/workqueue.c:3486 kthread+0x31d/0x420 kernel/kthread.c:436 ret_from_fork+0x662/0x940 arch/x86/kernel/process.c:158 ret_from_fork_asm+0x1a/0x30 arch/x86/entry/entry_64.S:245 Allocated by task 2224 on cpu 0 at 66.550027s: kasan_save_stack+0x24/0x50 mm/kasan/common.c:57 kasan_save_track+0x17/0x60 mm/kasan/common.c:78 poison_kmalloc_redzone mm/kasan/common.c:398 [inline] __kasan_kmalloc+0xaa/0xb0 mm/kasan/common.c:415 kasan_kmalloc include/linux/kasan.h:263 [inline] __do_kmalloc_node mm/slub.c:5334 [inline] __kmalloc_noprof+0x304/0x7c0 mm/slub.c:5359 _kmalloc_noprof include/linux/slab.h:992 [inline] dma_resv_list_alloc+0x27/0x90 drivers/dma-buf/dma-resv.c:106 dma_resv_reserve_fences+0x60e/0xa30 drivers/dma-buf/dma-resv.c:205 ttm_bo_alloc_resource+0x12c/0xbd0 drivers/gpu/drm/ttm/ttm_bo.c:721 ttm_bo_validate+0x1bc/0x4a0 drivers/gpu/drm/ttm/ttm_bo.c:856 ttm_bo_init_reserved+0x2c3/0x570 drivers/gpu/drm/ttm/ttm_bo.c:970 nouveau_bo_init+0x159/0x2c0 drivers/gpu/drm/nouveau/nouveau_bo.c:359 nouveau_gem_new+0x234/0x5f0 drivers/gpu/drm/nouveau/nouveau_gem.c:272 nouveau_gem_ioctl_new+0x1eb/0x420 drivers/gpu/drm/nouveau/nouveau_gem.c:352 drm_ioctl_kernel+0x192/0x350 drivers/gpu/drm/drm_ioctl.c:817 drm_ioctl+0x4f8/0xb40 drivers/gpu/drm/drm_ioctl.c:914 nouveau_drm_ioctl+0xea/0x2c0 drivers/gpu/drm/nouveau/nouveau_drm.c:1338 vfs_ioctl fs/ioctl.c:51 [inline] __do_sys_ioctl fs/ioctl.c:597 [inline] __se_sys_ioctl fs/ioctl.c:583 [inline] __x64_sys_ioctl+0x180/0x1d0 fs/ioctl.c:583 do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline] do_syscall_64+0x115/0x690 arch/x86/entry/syscall_64.c:94 entry_SYSCALL_64_after_hwframe+0x77/0x7f Freed by task 2223 on cpu 0 at 66.554063s: kasan_save_stack+0x24/0x50 mm/kasan/common.c:57 kasan_save_track+0x17/0x60 mm/kasan/common.c:78 kasan_save_free_info+0x3b/0x60 mm/kasan/generic.c:584 poison_slab_object mm/kasan/common.c:253 [inline] __kasan_slab_free+0x61/0x80 mm/kasan/common.c:285 kasan_slab_free include/linux/kasan.h:235 [inline] slab_free_hook mm/slub.c:2677 [inline] __rcu_free_sheaf_prepare+0xb6/0x2e0 mm/slub.c:2928 rcu_free_sheaf+0x1b/0x120 mm/slub.c:5978 rcu_do_batch kernel/rcu/tree.c:2645 [inline] rcu_core+0x521/0x1490 kernel/rcu/tree.c:2897 handle_softirqs+0x1b1/0x8a0 kernel/softirq.c:622 __do_softirq kernel/softirq.c:656 [inline] invoke_softirq kernel/softirq.c:496 [inline] __irq_exit_rcu+0x137/0x1c0 kernel/softirq.c:735 irq_exit_rcu+0x9/0x20 kernel/softirq.c:752 instr_sysvec_apic_timer_interrupt arch/x86/kernel/apic/apic.c:1062 [inline] sysvec_apic_timer_interrupt+0x70/0x80 arch/x86/kernel/apic/apic.c:1062 asm_sysvec_apic_timer_interrupt+0x1a/0x20 arch/x86/include/asm/idtentry.h:674 The buggy address belongs to the object at ffff888127d3e380 which belongs to the cache kmalloc-96 of size 96 The buggy address is located 32 bytes inside of freed 96-byte region [ffff888127d3e380, ffff888127d3e3e0) The buggy address belongs to the physical page: page: refcount:0 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x127d3e flags: 0x200000000000000(node=0|zone=2) page_type: f5(slab) raw: 0200000000000000 ffff888100041280 dead000000000122 0000000000000000 raw: 0000000000000000 0000000000200020 00000000f5000000 0000000000000000 page dumped because: kasan: bad access detected Memory state around the buggy address: ffff888127d3e280: 00 00 00 00 00 00 00 00 00 fc fc fc fc fc fc fc ffff888127d3e300: 00 00 00 00 00 00 00 00 00 00 00 fc fc fc fc fc >ffff888127d3e380: fa fb fb fb fb fb fb fb fb fb fb fb fc fc fc fc ^ ffff888127d3e400: fa fb fb fb fb fb fb fb fb fb fb fb fc fc fc fc ffff888127d3e480: 00 00 00 00 00 00 00 00 00 00 fc fc fc fc fc fc Fix by resetting the placement to the BO's valid domains before calling nouveau_bo_validate(), matching the handling in nouveau_uvmm_bo_validate(), so map jobs do not run for SYSTEM resources; Reject BO that cannot reside in VRAM or GART; Also skip op_map() when the GPUVA has been invalidated, matching the handling in the unmap and remap paths. Found when fuzzing the nouveau driver with a modified Syzkaller. Fixes: b88baab82871 ("drm/nouveau: implement new VM_BIND uAPI") Cc: stable@vger.kernel.org Signed-off-by: Peiyang He Assisted-by: Codex:gpt-5.5 Reviewed-by: Lyude Paul Signed-off-by: Lyude Paul Link: https://patch.msgid.link/0D77BEC410CE0129+20260907052204.1431488-1-peiyang_he@smail.nju.edu.cn --- drivers/gpu/drm/nouveau/nouveau_uvmm.c | 13 +++++++++++-- 1 file changed, 11 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/nouveau/nouveau_uvmm.c b/drivers/gpu/drm/nouveau/nouveau_uvmm.c index fc125fd44a9b..2026fe6b48c6 100644 --- a/drivers/gpu/drm/nouveau/nouveau_uvmm.c +++ b/drivers/gpu/drm/nouveau/nouveau_uvmm.c @@ -846,6 +846,9 @@ op_map(struct nouveau_uvma *uvma) { struct nouveau_bo *nvbo = nouveau_gem_object(uvma->va.gem.obj); + if (drm_gpuva_invalidated(&uvma->va)) + return; + nouveau_uvma_map(uvma, nouveau_mem(nvbo->bo.resource)); } @@ -1232,6 +1235,7 @@ bind_lock_validate(struct nouveau_job *job, struct drm_exec *exec, drm_gpuva_for_each_op(va_op, op->ops) { struct drm_gem_object *obj = op_gem_obj(va_op); + struct nouveau_bo *nvbo; if (unlikely(!obj)) continue; @@ -1246,8 +1250,13 @@ bind_lock_validate(struct nouveau_job *job, struct drm_exec *exec, if (va_op->op == DRM_GPUVA_OP_UNMAP) continue; - ret = nouveau_bo_validate(nouveau_gem_object(obj), - true, false); + nvbo = nouveau_gem_object(obj); + if (!(nvbo->valid_domains & + (NOUVEAU_GEM_DOMAIN_VRAM | NOUVEAU_GEM_DOMAIN_GART))) + return -EINVAL; + + nouveau_bo_placement_set(nvbo, nvbo->valid_domains, 0); + ret = nouveau_bo_validate(nvbo, true, false); if (ret) return ret; } From 21c89ff1fc86ffa517f3a9ca1ba5c48e9e37c1ba Mon Sep 17 00:00:00 2001 From: Abdurrahman Hussain Date: Mon, 31 Aug 2026 18:29:25 -0700 Subject: [PATCH 151/694] of/overlay: don't create "//" paths for fragments targeting the root dup_and_fixup_symbol_prop() rewrites a symbol value by replacing its "/fragment/__overlay__" prefix with the fragment's target path. When the fragment targets the root node the target path is "/" and the result starts with "//", which __of_find_node_by_full_path() cannot resolve: symbols pointing into such fragments silently stop resolving. Drop the target path when it is the root and the tail is non-empty. A value naming the fragment root itself (empty tail) keeps the "/". Fixes: d1651b03c2df ("of: overlay: add overlay symbols to live device tree") Assisted-by: Claude:claude-fable-5 [Claude Code] Signed-off-by: Abdurrahman Hussain Link: https://patch.msgid.link/20260831-nh-of-alias-overlay-v7-7-02754604805a@nexthop.ai Signed-off-by: Rob Herring (Arm) --- drivers/of/overlay.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/of/overlay.c b/drivers/of/overlay.c index 276870291656..83c56dd9b5d8 100644 --- a/drivers/of/overlay.c +++ b/drivers/of/overlay.c @@ -256,6 +256,9 @@ static struct property *dup_and_fixup_symbol_prop( if (!target_path) return NULL; target_path_len = strlen(target_path); + /* a root target renders as "/"; drop it to avoid "//" results */ + if (target_path_len == 1 && target_path[0] == '/' && path_tail_len) + target_path_len = 0; new_prop = kzalloc_obj(*new_prop); if (!new_prop) From c0078f4d8c8fcd8edfe8c53ffe77d48565c1957e Mon Sep 17 00:00:00 2001 From: Wei Wang Date: Tue, 15 Sep 2026 20:14:12 +0000 Subject: [PATCH 152/694] mailmap: add entry for Wei Wang My Meta email address is no longer active. Map it to my current address so that git and get_maintainer.pl stop pointing at a dead address for my contributions. Signed-off-by: Wei Wang Link: https://patch.msgid.link/20260915201412.2201757-1-weiwan@google.com Signed-off-by: Jakub Kicinski --- .mailmap | 1 + 1 file changed, 1 insertion(+) diff --git a/.mailmap b/.mailmap index 1f5540bc33f1..526217c88495 100644 --- a/.mailmap +++ b/.mailmap @@ -975,6 +975,7 @@ Vladimir Davydov Vlastimil Babka WangYuli WangYuli +Wei Wang Weiwen Hu WeiXiong Liao Wen Gong From f7eae6d8d768fabd6b59779ca7da79e02c74e113 Mon Sep 17 00:00:00 2001 From: "Jonghyuk Kim(MalHyuk)" Date: Wed, 2 Sep 2026 10:27:16 +0900 Subject: [PATCH 153/694] drm/nouveau: RCU-free the scheduler-containing nouveau_sched struct nouveau_sched embeds a struct drm_gpu_scheduler (base). nouveau_sched_destroy() calls nouveau_sched_fini() (which does drm_sched_fini(&sched->base)) and then frees the object with plain kfree(sched). drm_sched_fence_get_timeline_name() returns fence->sched->name, and the scheduler fence keeps a .release callback so it is not ops-detached on signalling. A finished fence exported to userspace via drm_syncobj / sync_file therefore keeps pointing at &sched->base after nouveau_sched_destroy(), and a later get_timeline_name() -- reachable unprivileged through SYNC_IOC_FILE_INFO -- dereferences freed memory (KASAN slab-use-after-free read). Per the dma-fence lifetime contract the exporter must keep the data backing a signalled fence alive for an RCU grace period. Free the scheduler-containing object with kfree_rcu() instead of kfree(). Fixes: 5f03a507b29e ("drm/nouveau: implement 1:1 scheduler - entity relationship") Cc: stable@vger.kernel.org Signed-off-by: Jonghyuk Kim(MalHyuk) Reviewed-by: Lyude Paul Signed-off-by: Lyude Paul Link: https://patch.msgid.link/20260902012717.880724-1-malhyuk97@gmail.com --- drivers/gpu/drm/nouveau/nouveau_sched.c | 2 +- drivers/gpu/drm/nouveau/nouveau_sched.h | 1 + 2 files changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/nouveau/nouveau_sched.c b/drivers/gpu/drm/nouveau/nouveau_sched.c index 8b9f935afe09..b3f02c490ecb 100644 --- a/drivers/gpu/drm/nouveau/nouveau_sched.c +++ b/drivers/gpu/drm/nouveau/nouveau_sched.c @@ -517,7 +517,7 @@ nouveau_sched_destroy(struct nouveau_sched **psched) struct nouveau_sched *sched = *psched; nouveau_sched_fini(sched); - kfree(sched); + kfree_rcu(sched, rcu); *psched = NULL; } diff --git a/drivers/gpu/drm/nouveau/nouveau_sched.h b/drivers/gpu/drm/nouveau/nouveau_sched.h index 20cd1da8db73..51ce8dcf6285 100644 --- a/drivers/gpu/drm/nouveau/nouveau_sched.h +++ b/drivers/gpu/drm/nouveau/nouveau_sched.h @@ -98,6 +98,7 @@ void nouveau_job_free(struct nouveau_job *job); struct nouveau_sched { struct drm_gpu_scheduler base; + struct rcu_head rcu; struct drm_sched_entity entity; struct workqueue_struct *wq; struct mutex mutex; From 50e80e2bb5e2be8515205b9c496b9640ddefa434 Mon Sep 17 00:00:00 2001 From: Weiming Shi Date: Tue, 15 Sep 2026 01:02:07 +0800 Subject: [PATCH 154/694] bpf: Skip unsettled links in link iterator bpf_link_prime() inserts a link into link_idr before anon_inode_getfile() succeeds and before bpf_link_settle() publishes the ID in link->id. bpf_link_by_id() treats such an ID-zero link as unsettled, but the link iterator takes a reference without this check. If anon_inode_getfile() then fails, the creator removes the ID and frees its still-private link directly. The iterator is left with a dangling reference and its next bpf_link_put() accesses freed memory. Treat ID-zero entries as transient in bpf_link_get_curr_or_next(), just as bpf_link_by_id() does. BUG: KASAN: slab-use-after-free in bpf_link_put Write of size 8 by task exp/384 Call Trace: bpf_link_put kernel/bpf/syscall.c:3372 bpf_link_seq_next kernel/bpf/link_iter.c:33 bpf_seq_read kernel/bpf/bpf_iter.c:158 vfs_read fs/read_write.c:572 ksys_read fs/read_write.c:716 do_syscall_64 arch/x86/entry/syscall_64.c:84 entry_SYSCALL_64_after_hwframe arch/x86/entry/entry_64.S:121 Kernel panic - not syncing: KASAN: panic_on_warn set ... Fixes: 9f8836127308 ("bpf: Add bpf_link iterator") Reported-by: Xiang Mei Signed-off-by: Weiming Shi Signed-off-by: Andrii Nakryiko Link: https://lore.kernel.org/bpf/20260914170206.170723-2-bestswngs@gmail.com --- kernel/bpf/syscall.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c index 853b47f81384..244a939b9d2d 100644 --- a/kernel/bpf/syscall.c +++ b/kernel/bpf/syscall.c @@ -6042,7 +6042,10 @@ struct bpf_link *bpf_link_get_curr_or_next(u32 *id) again: link = idr_get_next(&link_idr, id); if (link) { - link = bpf_link_inc_not_zero(link); + if (link->id) + link = bpf_link_inc_not_zero(link); + else + link = ERR_PTR(-EAGAIN); if (IS_ERR(link)) { (*id)++; goto again; From fefd9480ec361969f1a836df46326a1801062c26 Mon Sep 17 00:00:00 2001 From: Guangshuo Li Date: Sat, 8 Aug 2026 21:41:37 +0800 Subject: [PATCH 155/694] drm/nouveau: fix autosuspend cleanup during teardown nouveau_drm_device_init() calls pm_runtime_use_autosuspend(), but nouveau_drm_device_fini() does not call the matching pm_runtime_dont_use_autosuspend(). If the autosuspend delay is set to a negative value while autosuspend is enabled, the runtime PM core increments usage_count to prevent runtime suspend. Without calling pm_runtime_dont_use_autosuspend() during teardown, this reference is not dropped and usage_count remains unbalanced. The documentation for pm_runtime_use_autosuspend() also notes that it is important to undo it with pm_runtime_dont_use_autosuspend() at driver exit time, unless runtime PM was initially enabled with devm_pm_runtime_enable(). Add the missing pm_runtime_dont_use_autosuspend() call to the common device teardown path. This issue was found by manual code inspection. Fixes: 5addcf0a5f0f ("nouveau: add runtime PM support (v0.9)") Cc: stable@vger.kernel.org Signed-off-by: Guangshuo Li Reviewed-by: Lyude Paul Signed-off-by: Lyude Paul Link: https://patch.msgid.link/20260808134137.2864847-1-lgs201920130244@gmail.com --- drivers/gpu/drm/nouveau/nouveau_drm.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/gpu/drm/nouveau/nouveau_drm.c b/drivers/gpu/drm/nouveau/nouveau_drm.c index d0eb29583cb1..2c7077a49888 100644 --- a/drivers/gpu/drm/nouveau/nouveau_drm.c +++ b/drivers/gpu/drm/nouveau/nouveau_drm.c @@ -585,6 +585,7 @@ nouveau_drm_device_fini(struct nouveau_drm *drm) if (nouveau_pmops_runtime()) { pm_runtime_get_sync(dev->dev); pm_runtime_forbid(dev->dev); + pm_runtime_dont_use_autosuspend(dev->dev); } nouveau_led_fini(dev); From 8e0b235bd918d06f54ba8fddd2c3ddc36ca59c15 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Bj=C3=B6rn=20T=C3=B6pel?= Date: Tue, 15 Sep 2026 12:49:15 +0200 Subject: [PATCH 156/694] eth: fbnic: Fix payload page pool error cleanup MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The payload page pool pointer contains an error pointer when its allocation fails. The cleanup path passes that error pointer to page_pool_destroy() instead of destroying the header page pool. This can dereference the error pointer and leave the header page pool allocated. Destroy the header page pool instead. Fixes: 8a11010fdd96 ("eth: fbnic: allocate unreadable page pool for the payloads") Reported-by: Sashiko Link: https://lore.kernel.org/netdev/178915061000.219967.7726187707862333281@kernel.org/ Signed-off-by: Björn Töpel Reviewed-by: Simon Horman Link: https://patch.msgid.link/20260915104917.3978113-1-bjorn@kernel.org Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/meta/fbnic/fbnic_txrx.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/ethernet/meta/fbnic/fbnic_txrx.c b/drivers/net/ethernet/meta/fbnic/fbnic_txrx.c index e7918d3f6aba..661dee1661af 100644 --- a/drivers/net/ethernet/meta/fbnic/fbnic_txrx.c +++ b/drivers/net/ethernet/meta/fbnic/fbnic_txrx.c @@ -1622,7 +1622,7 @@ fbnic_alloc_qt_page_pools(struct fbnic_net *fbn, struct fbnic_q_triad *qt, return 0; err_destroy_sub0: - page_pool_destroy(pp); + page_pool_destroy(qt->sub0.page_pool); return PTR_ERR(pp); } From 0a5f5d9e94dead312d32c366b917c64e552b72f7 Mon Sep 17 00:00:00 2001 From: Jamal Hadi Salim Date: Wed, 16 Sep 2026 06:01:14 -0400 Subject: [PATCH 157/694] net/sched: cls_u32: fix manual hash table handle IDR aliasing A u32 hash table created with an explicit handle ('tc filter add ... handle 801: u32 divisor N') keys its IDR entry on the raw handle, while the destroy paths free it under handle2id(handle). The two key domains disagree for handles in the 0x800..0xFFF htid range: handle2id() folds them back into the auto-allocated id space (1..0x7FF). A manual table therefore leaves its raw-keyed IDR entry unreachable on delete (a permanent leak), and its delete can drop the idr entry of an unrelated live auto table. A later auto allocation can then hand out a handle that aliases the live manual table; u32_lookup_ht() first-match routes lookups and TCA_U32_LINK for that htid to the wrong table. Key the divisor-path alloc on handle2id(handle) so allocation and removal share one key domain. A manual handle that maps onto an id already in use is rejected with -ENOSPC, and auto allocation skips ids held by live manual tables. Conditions to recreate: ip link add test0 type dummy tc qdisc add dev test0 clsact tc filter add dev test0 ingress protocol ip pref 1 \ handle 801: u32 divisor 16 tc filter add dev test0 ingress protocol ip pref 2 u32 divisor 16 tc -d filter show dev test0 ingress | grep 'fh 801:' # unpatched: two live tables with handle 0x80100000 (the pref 2 root # hnode is auto-allocated id 1); patched: the auto hnode takes id 2. Also tested with a poc with a live u32 table on the block, add/delete a manual table 'handle 901: u32 divisor 1' twice; unpatched, the re-add fails with -ENOSPC because the raw key leaked on the first delete. Fixes: 73af53d82076 ("net: sched: cls_u32: Fix u32's systematic failure to free IDR entries for hnodes.") Reported-by: Sashiko (gemini + nipa) Closes: https://sashiko.dev/#/patchset/20260822222049.114526-1-jhs@mojatatu.com Reviewed-by: Victor Nogueira Tested-by: hybris Signed-off-by: Jamal Hadi Salim Reviewed-by: Simon Horman Link: https://patch.msgid.link/QDISC-LQFE.v1.20260911041746.1@mojatatu.com Signed-off-by: Jakub Kicinski --- net/sched/cls_u32.c | 12 ++++++++++-- 1 file changed, 10 insertions(+), 2 deletions(-) diff --git a/net/sched/cls_u32.c b/net/sched/cls_u32.c index a3e65c8cf29e..76ce2d124079 100644 --- a/net/sched/cls_u32.c +++ b/net/sched/cls_u32.c @@ -1003,8 +1003,16 @@ static int u32_change(struct net *net, struct sk_buff *in_skb, return -ENOMEM; } } else { - err = idr_alloc_u32(&tp_c->handle_idr, ht, &handle, - handle, GFP_KERNEL); + /* The IDR is keyed on the mapped id, and that is + * what the destroy paths remove. Ask for it here, + * so a manual handle colliding with the + * auto-allocated id space is rejected (-ENOSPC) + * instead of aliasing a future auto id. + */ + u32 id = handle2id(handle); + + err = idr_alloc_u32(&tp_c->handle_idr, ht, &id, id, + GFP_KERNEL); if (err) { kfree(ht); return err; From 960ab631f3d8789586db71b9914b361059ddcb6e Mon Sep 17 00:00:00 2001 From: Jamal Hadi Salim Date: Wed, 16 Sep 2026 06:01:15 -0400 Subject: [PATCH 158/694] selftests/tc-testing: add u32 manual table handle IDR tests 35fc: create a manual table with handle 801:, then add an auto-allocated table. Before the fix, the auto allocation reuses id 1 and hands out the same handle 0x80100000, aliasing the manual table; the test requires the manual 801: handle to keep exactly one entry in the dump. a6e8: with a live u32 table keeping the tc_u_common alive, add and delete a manual table with handle 901:, then re-add it. Unpatched, the delete leaks the raw-keyed IDR entry and the re-add fails with -ENOSPC; the test requires the re-add to succeed. Reviewed-by: Victor Nogueira Tested-by: hybris Signed-off-by: Jamal Hadi Salim Reviewed-by: Simon Horman Link: https://patch.msgid.link/QDISC-LQFE.v1.20260911041746.2@mojatatu.com Signed-off-by: Jakub Kicinski --- .../tc-testing/tc-tests/filters/u32.json | 48 +++++++++++++++++++ 1 file changed, 48 insertions(+) diff --git a/tools/testing/selftests/tc-testing/tc-tests/filters/u32.json b/tools/testing/selftests/tc-testing/tc-tests/filters/u32.json index e2b03f2b5e89..edc5148a8d97 100644 --- a/tools/testing/selftests/tc-testing/tc-tests/filters/u32.json +++ b/tools/testing/selftests/tc-testing/tc-tests/filters/u32.json @@ -376,5 +376,53 @@ "teardown": [ "$TC qdisc del dev $DUMMY clsact" ] + }, + { + "id": "35fc", + "name": "u32 manual table then auto table: auto allocation must not alias a live manual handle", + "category": [ + "filter", + "u32" + ], + "plugins": { + "requires": "nsPlugin" + }, + "setup": [ + "$TC qdisc add dev $DEV1 ingress", + "$TC filter add dev $DEV1 ingress protocol ip pref 1 handle 801: u32 divisor 16" + ], + "cmdUnderTest": "$TC filter add dev $DEV1 ingress protocol ip pref 2 u32 divisor 16", + "expExitCode": "0", + "verifyCmd": "$TC -d filter show dev $DEV1 ingress", + "matchPattern": "fh 801:", + "matchCount": "1", + "teardown": [ + "$TC qdisc del dev $DEV1 ingress" + ] + }, + { + "id": "a6e8", + "name": "u32 manual table add/del does not leak its idr entry (re-adding the same handle succeeds)", + "category": [ + "filter", + "u32" + ], + "plugins": { + "requires": "nsPlugin" + }, + "setup": [ + "$TC qdisc add dev $DEV1 ingress", + "$TC filter add dev $DEV1 ingress protocol ip pref 1 u32 divisor 16", + "$TC filter add dev $DEV1 ingress protocol ip pref 5 handle 901: u32 divisor 1", + "$TC filter del dev $DEV1 ingress protocol ip pref 5 handle 901: u32" + ], + "cmdUnderTest": "$TC filter add dev $DEV1 ingress protocol ip pref 6 handle 901: u32 divisor 1", + "expExitCode": "0", + "verifyCmd": "$TC -d filter show dev $DEV1 ingress", + "matchPattern": "fh 901: ht divisor 1", + "matchCount": "1", + "teardown": [ + "$TC qdisc del dev $DEV1 ingress" + ] } ] From daf677c2c6449011ee695d55b48b5b2977a36f88 Mon Sep 17 00:00:00 2001 From: Kyle Hendry Date: Tue, 15 Sep 2026 10:39:20 -0700 Subject: [PATCH 159/694] net: pcs: rzn1-miic: Fix config array initialization Fix memset parameters to initialize the entire DT value array Fixes: f39e968dc168a7bd ("net: pcs: rzn1-miic: Move configuration data to SoC-specific struct") Reviewed-by: Geert Uytterhoeven Signed-off-by: Kyle Hendry Reviewed-by: Lad Prabhakar Link: https://patch.msgid.link/20260915-rzn1-miic-fix-array-v5-1-b7173fd5b97d@reliablecontrols.com Signed-off-by: Jakub Kicinski --- drivers/net/pcs/pcs-rzn1-miic.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/net/pcs/pcs-rzn1-miic.c b/drivers/net/pcs/pcs-rzn1-miic.c index 2b72fa98ddf1..cb74861e823c 100644 --- a/drivers/net/pcs/pcs-rzn1-miic.c +++ b/drivers/net/pcs/pcs-rzn1-miic.c @@ -683,7 +683,8 @@ static int miic_parse_dt(struct miic *miic, u32 *mode_cfg) if (!dt_val) return -ENOMEM; - memset(dt_val, MIIC_MODCTRL_CONF_NONE, sizeof(*dt_val)); + memset(dt_val, MIIC_MODCTRL_CONF_NONE, + sizeof(*dt_val) * miic->of_data->conf_conv_count); if (of_property_read_u32(np, "renesas,miic-switch-portin", &conf) == 0) dt_val[0] = conf; From 39c6580765dad6477fb2637f6f616e0d276aae65 Mon Sep 17 00:00:00 2001 From: Heyang Tan Date: Mon, 14 Sep 2026 10:05:21 +0800 Subject: [PATCH 160/694] octeontx2-af: use seq_file for rsrc_alloc debugfs The rsrc_alloc debugfs reader writes rows directly to userspace without respecting the caller's read count. It also uses the current row length as the userspace stride, which can corrupt output when rows have different widths. Use seq_file to handle userspace buffer sizes, offsets, and partial reads, and write output columns directly to the seq_file buffer. Fixes: 23205e6d06d4 ("octeontx2-af: Dump current resource provisioning status") Signed-off-by: Heyang Tan Reviewed-by: Ratheesh Kannoth Link: https://patch.msgid.link/20260914020521.146-1-thy15333007817@163.com Signed-off-by: Jakub Kicinski --- .../marvell/octeontx2/af/rvu_debugfs.c | 104 ++++++------------ 1 file changed, 33 insertions(+), 71 deletions(-) diff --git a/drivers/net/ethernet/marvell/octeontx2/af/rvu_debugfs.c b/drivers/net/ethernet/marvell/octeontx2/af/rvu_debugfs.c index 904374baae6f..2927633465d9 100644 --- a/drivers/net/ethernet/marvell/octeontx2/af/rvu_debugfs.c +++ b/drivers/net/ethernet/marvell/octeontx2/af/rvu_debugfs.c @@ -714,110 +714,72 @@ static int get_max_column_width(struct rvu *rvu) } /* Dumps current provisioning status of all RVU block LFs */ -static ssize_t rvu_dbg_rsrc_attach_status(struct file *filp, - char __user *buffer, - size_t count, loff_t *ppos) +static int rvu_dbg_rsrc_attach_status(struct seq_file *filp, void *unused) { - int index, off = 0, flag = 0, len = 0, i = 0; - struct rvu *rvu = filp->private_data; - int bytes_not_copied = 0; + struct rvu *rvu = filp->private; + int index, pf, vf, pcifunc; struct rvu_block block; - int pf, vf, pcifunc; - int buf_size = 2048; int lf_str_size; char *lfs; - char *buf; - /* don't allow partial reads */ - if (*ppos != 0) - return 0; - - buf = kzalloc(buf_size, GFP_KERNEL); - if (!buf) - return -ENOMEM; - - /* Get the maximum width of a column */ lf_str_size = get_max_column_width(rvu); + if (lf_str_size < 0) + return lf_str_size; lfs = kzalloc(lf_str_size, GFP_KERNEL); - if (!lfs) { - kfree(buf); + if (!lfs) return -ENOMEM; - } - off += scnprintf(&buf[off], buf_size - 1 - off, "%-*s", lf_str_size, - "pcifunc"); + + seq_printf(filp, "%-*s", lf_str_size, "pcifunc"); for (index = 0; index < BLK_COUNT; index++) - if (strlen(rvu->hw->block[index].name)) { - off += scnprintf(&buf[off], buf_size - 1 - off, - "%-*s", lf_str_size, - rvu->hw->block[index].name); - } + if (strlen(rvu->hw->block[index].name)) + seq_printf(filp, "%-*s", lf_str_size, + rvu->hw->block[index].name); - off += scnprintf(&buf[off], buf_size - 1 - off, "\n"); - bytes_not_copied = copy_to_user(buffer + (i * off), buf, off); - if (bytes_not_copied) - goto out; - - i++; - *ppos += off; + seq_putc(filp, '\n'); for (pf = 0; pf < rvu->hw->total_pfs; pf++) { for (vf = 0; vf <= rvu->hw->total_vfs; vf++) { - off = 0; - flag = 0; pcifunc = rvu_make_pcifunc(rvu->pdev, pf, vf); if (!pcifunc) continue; - if (vf) { - sprintf(lfs, "PF%d:VF%d", pf, vf - 1); - off = scnprintf(&buf[off], - buf_size - 1 - off, - "%-*s", lf_str_size, lfs); - } else { - sprintf(lfs, "PF%d", pf); - off = scnprintf(&buf[off], - buf_size - 1 - off, - "%-*s", lf_str_size, lfs); - } - for (index = 0; index < BLK_COUNT; index++) { block = rvu->hw->block[index]; if (!strlen(block.name)) continue; - len = 0; - lfs[len] = '\0'; + lfs[0] = '\0'; get_lf_str_list(&block, pcifunc, lfs); if (strlen(lfs)) - flag = 1; - - off += scnprintf(&buf[off], buf_size - 1 - off, - "%-*s", lf_str_size, lfs); + break; } - if (flag) { - off += scnprintf(&buf[off], - buf_size - 1 - off, "\n"); - bytes_not_copied = copy_to_user(buffer + - (i * off), - buf, off); - if (bytes_not_copied) - goto out; + if (index == BLK_COUNT) + continue; - i++; - *ppos += off; + if (vf) + sprintf(lfs, "PF%d:VF%d", pf, vf - 1); + else + sprintf(lfs, "PF%d", pf); + seq_printf(filp, "%-*s", lf_str_size, lfs); + + for (index = 0; index < BLK_COUNT; index++) { + block = rvu->hw->block[index]; + if (!strlen(block.name)) + continue; + + lfs[0] = '\0'; + get_lf_str_list(&block, pcifunc, lfs); + seq_printf(filp, "%-*s", lf_str_size, lfs); } + seq_putc(filp, '\n'); } } -out: kfree(lfs); - kfree(buf); - if (bytes_not_copied) - return -EFAULT; - return *ppos; + return 0; } -RVU_DEBUG_FOPS(rsrc_status, rsrc_attach_status, NULL); +RVU_DEBUG_SEQ_FOPS(rsrc_status, rsrc_attach_status, NULL); static int rvu_dbg_rvu_pf_cgx_map_display(struct seq_file *filp, void *unused) { From 1d653a183973f5283a3db5a38cd5e195eb152244 Mon Sep 17 00:00:00 2001 From: David Carlier Date: Thu, 17 Sep 2026 22:24:07 +0100 Subject: [PATCH 161/694] fprobe: Terminate the fgraph_data list when the reservation is not filled fprobe_fgraph_entry() reserves shadow stack space for every fprobe with an exit handler, but only fills it for those whose entry handler returns 0. fgraph_reserve_data() does not clear the area, so fprobe_return() parses the unused tail as headers left over from an earlier call, and an exit handler can run twice or despite its entry handler asking to skip it. Write a zero word after the last entry to terminate the walk. A zeroed slot does not decode to a NULL fprobe on the arches that encode the header into one unsigned long, since arch_decode_fprobe_header_fp() ORs in FPROBE_HEADER_MSB_PATTERN, so make read_fprobe_header() return NULL for a zeroed slot. Link: https://lore.kernel.org/all/20260917212407.384468-1-devnexen@gmail.com/ Fixes: e0a384434ae1 ("tracing: fprobe: do not zero out unused fgraph_data") Cc: stable@vger.kernel.org Suggested-by: Masami Hiramatsu (Google) Signed-off-by: David Carlier Signed-off-by: Masami Hiramatsu (Google) --- kernel/trace/fprobe.c | 15 +++++++++++++++ 1 file changed, 15 insertions(+) diff --git a/kernel/trace/fprobe.c b/kernel/trace/fprobe.c index 1e9b00997ff2..9f2d98181779 100644 --- a/kernel/trace/fprobe.c +++ b/kernel/trace/fprobe.c @@ -171,6 +171,11 @@ static inline bool write_fprobe_header(unsigned long *stack, static inline void read_fprobe_header(unsigned long *stack, struct fprobe **fp, unsigned int *size_words) { + if (!*stack) { + *fp = NULL; + *size_words = 0; + return; + } *fp = arch_decode_fprobe_header_fp(*stack); *size_words = arch_decode_fprobe_header_size(*stack); } @@ -203,6 +208,12 @@ static inline void read_fprobe_header(unsigned long *stack, { struct __fprobe_header *fph = (struct __fprobe_header *)stack; + if (!*stack) { + *fp = NULL; + *size_words = 0; + return; + } + *fp = fph->fp; *size_words = fph->size_words; } @@ -635,6 +646,10 @@ static int fprobe_fgraph_entry(struct ftrace_graph_ent *trace, struct fgraph_ops } } + /* Terminate the list, fgraph_reserve_data() does not clear it. */ + if (used && used < reserved_words) + fgraph_data[used] = 0; + /* If any exit_handler is set, data must be used. */ return used != 0; } From d09e8f64653c93da5793c16be19330968f2a32e6 Mon Sep 17 00:00:00 2001 From: Shay Drory Date: Tue, 15 Sep 2026 14:34:57 +0300 Subject: [PATCH 162/694] net/mlx5: devcom, Base component size on linked devices mlx5_devcom_comp_get_size() returns the component's kref count. That kref is bumped in mlx5_devcom_register_component() under comp_list_lock, before the comp_dev is linked onto comp_dev_list_head under comp->sem. The event broadcast (mlx5_devcom_locked_send_event()) walks that list. Hence, a caller can read the expected size, but send_event won't be sent to all peers. In the SD group registration path, this lets a member broadcast its role-election event over an incomplete list, electing a primary that never completes the group, is never marked ready, and leaves the group with a stale primary. Track the number of linked comp_devs in a dedicated counter, maintained under comp->sem together with the list add/remove, and return it from mlx5_devcom_comp_get_size(). Fixes: 9bb1ac80738a ("net/mlx5: devcom, Add component size getter") Signed-off-by: Shay Drory Reviewed-by: Akiva Goldberger Signed-off-by: Tariq Toukan Link: https://patch.msgid.link/20260915113459.3934760-2-tariqt@nvidia.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/mellanox/mlx5/core/lib/devcom.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lib/devcom.c b/drivers/net/ethernet/mellanox/mlx5/core/lib/devcom.c index 64f92427602d..75855481522b 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/lib/devcom.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/lib/devcom.c @@ -37,6 +37,7 @@ struct mlx5_devcom_comp { struct mlx5_devcom_key key; mlx5_devcom_event_handler_t handler; struct kref ref; + int nr_devs; bool ready; struct rw_semaphore sem; struct lock_class_key lock_key; @@ -170,6 +171,7 @@ devcom_alloc_comp_dev(struct mlx5_devcom_dev *devc, down_write(&comp->sem); list_add_tail(&devcom->list, &comp->comp_dev_list_head); + WRITE_ONCE(comp->nr_devs, comp->nr_devs + 1); up_write(&comp->sem); return devcom; @@ -182,6 +184,7 @@ devcom_free_comp_dev(struct mlx5_devcom_comp_dev *devcom) down_write(&comp->sem); list_del(&devcom->list); + WRITE_ONCE(comp->nr_devs, comp->nr_devs - 1); up_write(&comp->sem); kref_put(&devcom->devc->ref, mlx5_devcom_dev_release); @@ -284,7 +287,7 @@ int mlx5_devcom_comp_get_size(struct mlx5_devcom_comp_dev *devcom) { struct mlx5_devcom_comp *comp = devcom->comp; - return kref_read(&comp->ref); + return READ_ONCE(comp->nr_devs); } int mlx5_devcom_locked_send_event(struct mlx5_devcom_comp_dev *devcom, From e1e29ada2b938b13ba689a06a8bd8604564da2b3 Mon Sep 17 00:00:00 2001 From: Shay Drory Date: Tue, 15 Sep 2026 14:34:58 +0300 Subject: [PATCH 163/694] net/mlx5: SD, unload reps on shared FDB create error path mlx5_lag_shared_fdb_create() sets sd_fdb_active on every group member before reloading the representors, so mlx5_lag_is_active() is already true and the guard in mlx5_esw_offloads_rep_load() does not skip the VF/SF reps. If the reload then fails, the error path clears sd_fdb_active and destroys the shared FDB, leaving the reps loaded while SD LAG is inactive - the state cited commit was written to prevent. Unload the reps in the error path as well. Fixes: 68c2dd59a6c7 ("net/mlx5: E-Switch, Tie rep load/unload to SD LAG state") Signed-off-by: Shay Drory Reviewed-by: Akiva Goldberger Signed-off-by: Tariq Toukan Link: https://patch.msgid.link/20260915113459.3934760-3-tariqt@nvidia.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/mellanox/mlx5/core/lag/shared_fdb.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lag/shared_fdb.c b/drivers/net/ethernet/mellanox/mlx5/core/lag/shared_fdb.c index 6b4ad3c53f2f..424040918fa3 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/lag/shared_fdb.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/lag/shared_fdb.c @@ -270,6 +270,7 @@ int mlx5_lag_shared_fdb_create(struct mlx5_lag *ldev, pf->sd_fdb_active = false; } mlx5_lag_destroy_single_fdb_filter(ldev, group_id); + mlx5_lag_unload_reps_from_locked(ldev, filter); } err_add_devices: mlx5_lag_add_devices_filter(ldev, filter); From bae23d1ae62092c7f0ec6d5f7e1be5d164822638 Mon Sep 17 00:00:00 2001 From: Shay Drory Date: Tue, 15 Sep 2026 14:34:59 +0300 Subject: [PATCH 164/694] net/mlx5: LAG, reload IB reps of LAG master before the rest In a shared-FDB LAG the master device creates the bond IB device; the other LAG members do not create their own, they populate a port inside the master's IB device. mlx5_lag_reload_ib_reps_unlocked() reloaded the members' IB reps in iteration order, with no guarantee the master is reloaded first. When a non-master member is reloaded before the master, it tries to populate its port in an IB device that has not been recreated yet. Hence, reload the master's IB reps first, then every other member. Fixes: 2b204cdb1206 ("net/mlx5: LAG, use xa_alloc to manage LAG device indices") Signed-off-by: Shay Drory Reviewed-by: Akiva Goldberger Signed-off-by: Tariq Toukan Link: https://patch.msgid.link/20260915113459.3934760-4-tariqt@nvidia.com Signed-off-by: Jakub Kicinski --- .../net/ethernet/mellanox/mlx5/core/lag/lag.c | 44 ++++++++++++++----- 1 file changed, 32 insertions(+), 12 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lag/lag.c b/drivers/net/ethernet/mellanox/mlx5/core/lag/lag.c index c655f6e32e9b..dd14cdc378de 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/lag/lag.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/lag/lag.c @@ -1266,25 +1266,45 @@ void mlx5_lag_remove_devices(struct mlx5_lag *ldev) mlx5_lag_remove_devices_filter(ldev, MLX5_LAG_FILTER_PORTS); } +static int mlx5_lag_reload_ib_reps_idx(struct mlx5_lag *ldev, int idx, + u32 flags) +{ + struct lag_func *pf = mlx5_lag_pf(ldev, idx); + struct mlx5_eswitch *esw; + int ret; + + if (pf->dev->priv.flags & flags) + return 0; + + esw = pf->dev->priv.eswitch; + mlx5_esw_reps_block(esw); + ret = mlx5_eswitch_reload_ib_reps(esw); + mlx5_esw_reps_unblock(esw); + + return ret; +} + static int mlx5_lag_reload_ib_reps_unlocked(struct mlx5_lag *ldev, u32 flags, u32 filter, bool cont_on_fail) { - struct lag_func *pf; + int master_idx = mlx5_lag_get_dev_index_by_seq_filter(ldev, MLX5_LAG_P1, + filter); int ret; int i; - mlx5_lag_for_each(i, 0, ldev, filter) { - pf = mlx5_lag_pf(ldev, i); - if (!(pf->dev->priv.flags & flags)) { - struct mlx5_eswitch *esw; + if (master_idx < 0) + return -EINVAL; - esw = pf->dev->priv.eswitch; - mlx5_esw_reps_block(esw); - ret = mlx5_eswitch_reload_ib_reps(esw); - mlx5_esw_reps_unblock(esw); - if (ret && !cont_on_fail) - return ret; - } + ret = mlx5_lag_reload_ib_reps_idx(ldev, master_idx, flags); + if (ret && !cont_on_fail) + return ret; + + mlx5_lag_for_each(i, 0, ldev, filter) { + if (i == master_idx) + continue; + ret = mlx5_lag_reload_ib_reps_idx(ldev, i, flags); + if (ret && !cont_on_fail) + return ret; } return 0; From 261b61d3735b042ae25634f795c4540be0fc140c Mon Sep 17 00:00:00 2001 From: Kumar Kartikeya Dwivedi Date: Fri, 18 Sep 2026 01:32:09 +0200 Subject: [PATCH 165/694] bpf: Make post-verification instruction rewrites killable After do_check() returns, the verifier runs several instruction rewrite passes. Some of them patch or remove one instruction at a time. Each operation moves the remaining instruction and auxiliary-data arrays and adjusts all branch offsets, making the overall work quadratic in the program length. A privileged loader can submit 131072 unconditional jumps by zero followed by a valid return. Verification finishes quickly, but bpf_opt_remove_nops() then spends a long time removing each jump separately. Since this post-verification work neither checks for signals nor reschedules, a pending SIGKILL cannot terminate the task until the rewrite finishes. Make bpf_patch_insn_data() and verifier_remove_insns() common cancellation and rescheduling points. These helpers run from BPF_PROG_LOAD process context, and bpf_patch_insn_data() can already sleep while reallocating auxiliary data. Report interrupted constant blinding as -EINTR and propagate it through both JIT paths, including kernels that permit interpreter fallback. Other blinding failures retain the existing fallback behavior. This does not reduce the quadratic cost of the rewrite passes, but it makes the work preemptible and allows a killed loader to be torn down promptly. Fixes: 52875a04f4b2 ("bpf: verifier: remove dead code") Reported-by: Nicholas Carlini Suggested-by: Nicholas Carlini Signed-off-by: Kumar Kartikeya Dwivedi Acked-by: Eduard Zingerman Link: https://patch.msgid.link/20260917233222.2542500-2-memxor@gmail.com Signed-off-by: Eduard Zingerman --- kernel/bpf/core.c | 21 +++++++++++++++++---- kernel/bpf/fixups.c | 24 ++++++++++++++++++++++-- 2 files changed, 39 insertions(+), 6 deletions(-) diff --git a/kernel/bpf/core.c b/kernel/bpf/core.c index 8b294dfc1ad4..2e3bf8113ae9 100644 --- a/kernel/bpf/core.c +++ b/kernel/bpf/core.c @@ -19,6 +19,7 @@ #include #include +#include #include #include #include @@ -1619,6 +1620,8 @@ struct bpf_prog *bpf_jit_blind_constants(struct bpf_verifier_env *env, struct bp * fix it up here on error. */ bpf_jit_prog_release_other(prog, clone); + if (env && fatal_signal_pending(current)) + return ERR_PTR(-EINTR); return IS_ERR(tmp) ? tmp : ERR_PTR(-ENOMEM); } @@ -2636,11 +2639,14 @@ static struct bpf_prog *bpf_prog_jit_compile(struct bpf_verifier_env *env, struc orig_prog = prog; prog = bpf_jit_blind_constants(env, prog); /* - * If blinding was requested and we failed during blinding, we must fall - * back to the interpreter. + * Fall back to the interpreter after blinding failures, except when + * the loader was killed. */ - if (IS_ERR(prog)) + if (IS_ERR(prog)) { + if (PTR_ERR(prog) == -EINTR) + return prog; goto out_restore; + } prog = bpf_int_jit_compile(env, prog); if (prog->jited) { @@ -2659,6 +2665,8 @@ static struct bpf_prog *bpf_prog_jit_compile(struct bpf_verifier_env *env, struc struct bpf_prog *__bpf_prog_select_runtime(struct bpf_verifier_env *env, struct bpf_prog *fp, int *err) { + struct bpf_prog *jit_prog; + /* In case of BPF to BPF calls, verifier did all the prep * work with regards to JITing, etc. */ @@ -2681,7 +2689,12 @@ struct bpf_prog *__bpf_prog_select_runtime(struct bpf_verifier_env *env, struct if (*err) return fp; - fp = bpf_prog_jit_compile(env, fp); + jit_prog = bpf_prog_jit_compile(env, fp); + if (IS_ERR(jit_prog)) { + *err = PTR_ERR(jit_prog); + return fp; + } + fp = jit_prog; bpf_prog_jit_attempt_done(fp); if (!fp->jited && jit_needed) { *err = -ENOTSUPP; diff --git a/kernel/bpf/fixups.c b/kernel/bpf/fixups.c index 52d3cec33672..d6f83521fc78 100644 --- a/kernel/bpf/fixups.c +++ b/kernel/bpf/fixups.c @@ -8,6 +8,7 @@ #include #include #include +#include #include #include "disasm.h" @@ -306,12 +307,28 @@ static void adjust_poke_descs(struct bpf_prog *prog, u32 off, u32 len) } } +/* + * Some post-verification instruction rewriting passes require an + * O(prog->len) operation per instruction. Keep their shared primitives + * killable and preemptible. + */ +static bool bpf_rewrite_must_abort(void) +{ + if (fatal_signal_pending(current)) + return true; + cond_resched(); + return false; +} + struct bpf_prog *bpf_patch_insn_data(struct bpf_verifier_env *env, u32 off, const struct bpf_insn *patch, u32 len) { struct bpf_prog *new_prog; struct bpf_insn_aux_data *new_data = NULL; + if (bpf_rewrite_must_abort()) + return NULL; + if (len > 1) { new_data = vrealloc(env->insn_aux_data, array_size(env->prog->len + len - 1, @@ -523,6 +540,9 @@ static int verifier_remove_insns(struct bpf_verifier_env *env, u32 off, u32 cnt) unsigned int orig_prog_len = env->prog->len; int err; + if (bpf_rewrite_must_abort()) + return -EINTR; + if (bpf_prog_is_offloaded(env->prog->aux)) bpf_prog_offload_remove_insns(env, off, cnt); @@ -1356,7 +1376,7 @@ int bpf_jit_subprogs(struct bpf_verifier_env *env) } prog = bpf_jit_blind_constants(env, prog); if (IS_ERR(prog)) { - err = -ENOMEM; + err = PTR_ERR(prog); prog = orig_prog; goto out_restore; } @@ -1433,7 +1453,7 @@ int bpf_fixup_call_args(struct bpf_verifier_env *env) err = bpf_jit_subprogs(env); if (err == 0) return 0; - if (err == -EFAULT) + if (err == -EFAULT || err == -EINTR) return err; } #ifndef CONFIG_BPF_JIT_ALWAYS_ON From fd16449a9b3b31a8f18944c2f0e29e4e218ca2cf Mon Sep 17 00:00:00 2001 From: Kumar Kartikeya Dwivedi Date: Fri, 18 Sep 2026 01:32:10 +0200 Subject: [PATCH 166/694] bpf: Preserve packet pointer class displacement in regsafe() regsafe() maps packet pointer IDs between states and checks that each current register range is a subset of the corresponding explored register range. It does not, however, preserve the displacement between registers that share a packet pointer ID. This is unsound because packet range is shared by ID. A bounds check on one class member updates every member, and a later access can consume the range through another member. Commit 022ac0750883 ("bpf: use reg->var_off instead of reg->off for pointers") folded the fixed pointer offset into r64 and removed the old off equality check, so two individually narrower registers can prune even when their displacement has changed. The explored path can then license an out-of-bounds packet access on the pruned path. Require matching range bases for packet pointers with an ID. Together with the existing ID mapping, this preserves the displacement between members of each packet-pointer class without adding per-ID state. Packet pointers without an ID remain unaffected. Fixes: 022ac0750883 ("bpf: use reg->var_off instead of reg->off for pointers") Reported-by: Nicholas Carlini Suggested-by: Nicholas Carlini Signed-off-by: Kumar Kartikeya Dwivedi Acked-by: Eduard Zingerman Link: https://patch.msgid.link/20260917233222.2542500-3-memxor@gmail.com Signed-off-by: Eduard Zingerman --- kernel/bpf/states.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/kernel/bpf/states.c b/kernel/bpf/states.c index 66fb11b6c6a7..6c88ad95b63b 100644 --- a/kernel/bpf/states.c +++ b/kernel/bpf/states.c @@ -635,6 +635,9 @@ static bool regsafe(struct bpf_verifier_env *env, struct bpf_reg_state *rold, /* id relations must be preserved */ if (!check_ids(rold->id, rcur->id, idmap)) return false; + /* Preserve displacements between pointers sharing an ID. */ + if (rold->id && rold->r64.base != rcur->r64.base) + return false; /* new val must satisfy old val knowledge */ return range_within(rold, rcur) && tnum_in(rold->var_off, rcur->var_off); From 2059d9af54f0d66deb237aa75d2ed28648df05a1 Mon Sep 17 00:00:00 2001 From: Kumar Kartikeya Dwivedi Date: Fri, 18 Sep 2026 01:32:11 +0200 Subject: [PATCH 167/694] selftests/bpf: Test packet pointer class displacement pruning Add two paths whose packet pointer ranges are individually compatible at a join but whose members have different relative displacements. The first path proves an eight-byte access through one member. On the second path, the same guard only proves that the access starts before data_end. An affected verifier prunes the second path and accepts the program. With packet pointer class displacement preserved, it explores that path and rejects the out-of-bounds access. Read the unknown offset and branch selector directly from XDP context fields, and force state checkpoints so the pruning attempt does not depend on the verifier checkpoint heuristics. Signed-off-by: Kumar Kartikeya Dwivedi Link: https://patch.msgid.link/20260917233222.2542500-4-memxor@gmail.com Signed-off-by: Eduard Zingerman --- .../progs/verifier_xdp_direct_packet_access.c | 35 +++++++++++++++++++ 1 file changed, 35 insertions(+) diff --git a/tools/testing/selftests/bpf/progs/verifier_xdp_direct_packet_access.c b/tools/testing/selftests/bpf/progs/verifier_xdp_direct_packet_access.c index 0b86d95a4133..9866bc154194 100644 --- a/tools/testing/selftests/bpf/progs/verifier_xdp_direct_packet_access.c +++ b/tools/testing/selftests/bpf/progs/verifier_xdp_direct_packet_access.c @@ -1719,4 +1719,39 @@ l0_%=: r0 = 0; \ : __clobber_all); } +SEC("xdp") +__description("XDP pkt regsafe preserves packet pointer class displacement") +__failure __msg("R2 min value is outside of the allowed memory range") +__flag(BPF_F_ANY_ALIGNMENT) __flag(BPF_F_TEST_STATE_FREQ) +__naked void pkt_regsafe_class_displacement(void) +{ + asm volatile (" \ + r8 = *(u32 *)(r1 + %[xdp_md_data_end]); \ + r9 = *(u32 *)(r1 + %[xdp_md_data]); \ + r4 = *(u32 *)(r1 + %[xdp_md_rx_queue_index]); \ + r4 &= 15; \ + r0 = *(u32 *)(r1 + %[xdp_md_ingress_ifindex]); \ + if r0 != 0 goto l0_%=; \ + r2 = r9; \ + r2 += r4; \ + r3 = r2; \ + r3 += 8; \ + goto l1_%=; \ +l0_%=: r4 &= 3; \ + r4 += 8; \ + r2 = r9; \ + r2 += r4; \ + r3 = r2; \ +l1_%=: if r3 > r8 goto l2_%=; \ + r0 = *(u64 *)(r2 + 0); \ +l2_%=: r0 = 0; \ + exit; \ +" : + : __imm_const(xdp_md_data, offsetof(struct xdp_md, data)), + __imm_const(xdp_md_data_end, offsetof(struct xdp_md, data_end)), + __imm_const(xdp_md_rx_queue_index, offsetof(struct xdp_md, rx_queue_index)), + __imm_const(xdp_md_ingress_ifindex, offsetof(struct xdp_md, ingress_ifindex)) + : __clobber_all); +} + char _license[] SEC("license") = "GPL"; From c26e97721b172163042b98572fead234797f2c3c Mon Sep 17 00:00:00 2001 From: Kumar Kartikeya Dwivedi Date: Fri, 18 Sep 2026 01:32:12 +0200 Subject: [PATCH 168/694] bpf: Apply CO-RE relocations before subprogram validation check_subprogs() verifies that each subprogram ends in an exit or an unconditional jump before in-kernel CO-RE relocations are applied. An unresolved relocation can then replace that terminal instruction with an invalid helper call. The resulting fall-through into another subprogram breaks the CFG invariant used by postorder and stack liveness analysis, which can write past their per-subprogram arrays. Apply CO-RE relocations immediately after preparing the program BTF, before subprogram discovery and validation. Keep func_info and line_info validation after subprogram discovery because those records depend on the complete subprogram layout. Reject an ldimm64 first slot at the end of the instruction stream before CO-RE can inspect its missing second slot. check_subprogs() previously rejected this form before relocation processing because it is not a valid subprogram terminator. Moving CO-RE ahead of check_subprogs() removes that implicit protection, so perform an explicit check before applying relocations. Include core_relo_cnt when deciding whether to prepare program BTF. A load that supplied only CO-RE relocation metadata previously skipped both BTF setup and relocation processing. Fixes: fbd94c7afcf9 ("bpf: Pass a set of bpf_core_relo-s to prog_load command.") Suggested-by: Andrii Nakryiko Suggested-by: Alexei Starovoitov Suggested-by: Eduard Zingerman Signed-off-by: Kumar Kartikeya Dwivedi Acked-by: Eduard Zingerman Link: https://patch.msgid.link/20260917233222.2542500-5-memxor@gmail.com Signed-off-by: Eduard Zingerman --- include/linux/bpf_verifier.h | 2 ++ kernel/bpf/check_btf.c | 12 ++++-------- kernel/bpf/verifier.c | 12 +++++++++++- 3 files changed, 17 insertions(+), 9 deletions(-) diff --git a/include/linux/bpf_verifier.h b/include/linux/bpf_verifier.h index 36b65797877d..bba5a727c651 100644 --- a/include/linux/bpf_verifier.h +++ b/include/linux/bpf_verifier.h @@ -1197,6 +1197,8 @@ static inline void bpf_trampoline_unpack_key(u64 key, u32 *obj_id, u32 *btf_id) int bpf_prepare_btf_info(struct bpf_verifier_env *env, const union bpf_attr *attr, bpfptr_t uattr); +int bpf_check_core_relo(struct bpf_verifier_env *env, + const union bpf_attr *attr, bpfptr_t uattr); int bpf_check_btf_info(struct bpf_verifier_env *env, const union bpf_attr *attr, bpfptr_t uattr); diff --git a/kernel/bpf/check_btf.c b/kernel/bpf/check_btf.c index 0e8b3ccc7a5b..4c1ed842f661 100644 --- a/kernel/bpf/check_btf.c +++ b/kernel/bpf/check_btf.c @@ -338,9 +338,9 @@ static int check_btf_line(struct bpf_verifier_env *env, #define MIN_CORE_RELO_SIZE sizeof(struct bpf_core_relo) #define MAX_CORE_RELO_SIZE MAX_FUNCINFO_REC_SIZE -static int check_core_relo(struct bpf_verifier_env *env, - const union bpf_attr *attr, - bpfptr_t uattr) +int bpf_check_core_relo(struct bpf_verifier_env *env, + const union bpf_attr *attr, + bpfptr_t uattr) { u32 i, nr_core_relo, ncopy, expected_size, rec_size; struct bpf_core_relo core_relo = {}; @@ -414,7 +414,7 @@ int bpf_prepare_btf_info(struct bpf_verifier_env *env, struct btf *btf; int err; - if (!attr->func_info_cnt && !attr->line_info_cnt) { + if (!attr->func_info_cnt && !attr->line_info_cnt && !attr->core_relo_cnt) { if (check_abnormal_return(env)) return -EINVAL; return 0; @@ -455,9 +455,5 @@ int bpf_check_btf_info(struct bpf_verifier_env *env, if (err) return err; - err = check_core_relo(env, attr, uattr); - if (err) - return err; - return 0; } diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 5d7080c260d8..33161dc64568 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -21199,6 +21199,11 @@ int bpf_check(struct bpf_prog **prog, union bpf_attr *attr, bpfptr_t uattr, ret = bpf_diag_init(env); if (ret) goto err_prep; + if (env->prog->insnsi[env->prog->len - 1].code == (BPF_LD | BPF_IMM | BPF_DW)) { + verbose(env, "invalid bpf_ld_imm64 insn\n"); + ret = -EINVAL; + goto err_prep; + } if (env->signature) { ret = bpf_prog_calc_tag(env->prog); if (ret < 0) @@ -21274,6 +21279,11 @@ int bpf_check(struct bpf_prog **prog, union bpf_attr *attr, bpfptr_t uattr, if (ret < 0) goto skip_full_check; + /* Apply CO-RE before validating the program's instruction layout. */ + ret = bpf_check_core_relo(env, attr, uattr); + if (ret < 0) + goto skip_full_check; + /* Discover all subprograms before validating their layout and BTF. */ ret = add_subprogs(env); if (ret < 0) @@ -21283,7 +21293,7 @@ int bpf_check(struct bpf_prog **prog, union bpf_attr *attr, bpfptr_t uattr, if (ret < 0) goto skip_full_check; - /* Validate BTF against the complete subprogram layout and apply CO-RE. */ + /* Validate BTF against the complete subprogram layout. */ ret = bpf_check_btf_info(env, attr, uattr); if (ret < 0) goto skip_full_check; From 968ee7c06b62f1ac4597c351a45c2219a678a458 Mon Sep 17 00:00:00 2001 From: Kumar Kartikeya Dwivedi Date: Fri, 18 Sep 2026 01:32:13 +0200 Subject: [PATCH 169/694] selftests/bpf: Test early in-kernel CO-RE relocation Add a raw program load with CO-RE relocation metadata but no func_info or line_info. Place the relocation in dead code and require the poisoning log, proving that the kernel processes standalone CO-RE metadata instead of silently skipping it. Also give a subprogram a relocatable immediate as its terminal instruction. Require the relocation's poisoning log before check_subprogs() rejects the resulting fall-through. With the old ordering, check_subprogs() rejects the original terminal instruction before CO-RE can emit the substitution log, so the test continues to distinguish the ordering after relocation target validation is tightened. Submit a trailing ldimm64 first slot with CO-RE metadata and require the early structural diagnostic. This exercises the check that protects relocation processing instead of the later regular instruction validation. Load the standalone instruction stream without relocation metadata first to ensure that CO-RE processing causes its poisoning diagnostic. Encode the fixed BTF metadata directly with the selftest BTF helpers. Suggested-by: Eduard Zingerman Signed-off-by: Kumar Kartikeya Dwivedi Acked-by: Eduard Zingerman Link: https://patch.msgid.link/20260917233222.2542500-6-memxor@gmail.com Signed-off-by: Eduard Zingerman --- .../selftests/bpf/prog_tests/core_reloc_raw.c | 133 ++++++++++++++++++ 1 file changed, 133 insertions(+) diff --git a/tools/testing/selftests/bpf/prog_tests/core_reloc_raw.c b/tools/testing/selftests/bpf/prog_tests/core_reloc_raw.c index a18d3680fb16..bb19e49dd87d 100644 --- a/tools/testing/selftests/bpf/prog_tests/core_reloc_raw.c +++ b/tools/testing/selftests/bpf/prog_tests/core_reloc_raw.c @@ -14,6 +14,138 @@ static char log[16 * 1024]; +static int load_core_relo_insns(int btf_fd, struct bpf_insn *insns, int insn_cnt, + struct bpf_func_info *funcs, int func_cnt, + int enum_id, int access_str_off, int insn_idx, + bool relocate) +{ + struct bpf_core_relo relo = { + .insn_off = insn_idx * sizeof(struct bpf_insn), + .type_id = enum_id, + .access_str_off = access_str_off, + .kind = BPF_CORE_ENUMVAL_VALUE, + }; + union bpf_attr attr = { + .prog_type = BPF_PROG_TYPE_SOCKET_FILTER, + .insn_cnt = insn_cnt, + .insns = (__u64)insns, + .license = (__u64)"GPL", + .log_buf = (__u64)log, + .log_size = sizeof(log), + .log_level = 2, + .prog_btf_fd = btf_fd, + .func_info_rec_size = sizeof(struct bpf_func_info), + .func_info = (__u64)funcs, + .func_info_cnt = func_cnt, + }; + + if (relocate) { + attr.core_relo_cnt = 1; + attr.core_relos = (__u64)&relo; + attr.core_relo_rec_size = sizeof(relo); + } + memset(log, 0, sizeof(log)); + return sys_bpf_prog_load(&attr, sizeof(attr), 1); +} + +static void test_early_core_relo(void) +{ + struct test_btf { + struct btf_header hdr; + __u32 types[18]; + char strings[64]; + } raw_btf = { + .hdr = { + .magic = BTF_MAGIC, + .version = BTF_VERSION, + .hdr_len = sizeof(struct btf_header), + .type_off = 0, + .type_len = sizeof(raw_btf.types), + .str_off = offsetof(struct test_btf, strings) - + offsetof(struct test_btf, types), + .str_len = sizeof(raw_btf.strings), + }, + .types = { + BTF_TYPE_INT_ENC(1, BTF_INT_SIGNED, 0, 32, 4), /* [1] int */ + BTF_FUNC_PROTO_ENC(1, 0), /* [2] int (*)(void) */ + BTF_FUNC_ENC(5, 2), /* [3] main_fn */ + BTF_FUNC_ENC(13, 2), /* [4] sub_fn */ + BTF_TYPE_ENC(20, BTF_INFO_ENC(BTF_KIND_ENUM, 0, 1), 4), /* [5] enum */ + BTF_ENUM_ENC(45, 0), /* value = 0 */ + }, + .strings = "\0int\0main_fn\0sub_fn\0core_relo_poison_missing\0value\0" "0", + }; + struct bpf_func_info funcs[] = { + { .insn_off = 0, .type_id = 3 }, + { .insn_off = 3, .type_id = 4 }, + }; + struct bpf_insn core_only[] = { + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_JMP_IMM(BPF_JEQ, BPF_REG_0, 0, 1), + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_EXIT_INSN(), + }; + struct bpf_insn subprog[] = { + BPF_CALL_REL(2), + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_EXIT_INSN(), + }; + struct bpf_insn truncated_ldimm64[] = { + BPF_RAW_INSN(BPF_LD | BPF_IMM | BPF_DW, 0, 0, 0, 0), + }; + int access_str_off = 51; /* offset of "0" */ + int enum_id = 5; + int btf_fd, prog_fd = -1; + + btf_fd = bpf_btf_load(&raw_btf, sizeof(raw_btf), NULL); + if (!ASSERT_GE(btf_fd, 0, "btf_load")) + goto cleanup; + + if (test__start_subtest("without_func_info")) { + prog_fd = load_core_relo_insns(btf_fd, core_only, ARRAY_SIZE(core_only), NULL, 0, + enum_id, access_str_off, 2, false); + if (!ASSERT_GE(prog_fd, 0, "control_load")) + goto cleanup; + close(prog_fd); + prog_fd = load_core_relo_insns(btf_fd, core_only, ARRAY_SIZE(core_only), NULL, 0, + enum_id, access_str_off, 2, true); + if (!ASSERT_GE(prog_fd, 0, "poisoned_load")) + goto cleanup; + ASSERT_HAS_SUBSTR(log, "substituting insn #2", "poison_log"); + close(prog_fd); + prog_fd = -1; + } + + if (test__start_subtest("before_subprog_validation")) { + prog_fd = load_core_relo_insns(btf_fd, subprog, ARRAY_SIZE(subprog), funcs, 2, + enum_id, access_str_off, 2, true); + if (!ASSERT_LT(prog_fd, 0, "poisoned_load")) + goto cleanup; + ASSERT_HAS_SUBSTR(log, "substituting insn #2", "poison_log"); + ASSERT_HAS_SUBSTR(log, "last insn is not an exit or jmp", "poisoned_load_log"); + } + + if (test__start_subtest("truncated_ldimm64")) { + prog_fd = load_core_relo_insns(btf_fd, truncated_ldimm64, + ARRAY_SIZE(truncated_ldimm64), NULL, 0, + enum_id, access_str_off, 0, true); + if (!ASSERT_LT(prog_fd, 0, "truncated_load")) + goto cleanup; + ASSERT_HAS_SUBSTR(log, "invalid bpf_ld_imm64 insn", "truncated_load_log"); + } + +cleanup: + if (env.verbosity > VERBOSE_NORMAL && log[0]) { + printf("-------- program load log start --------\n"); + printf("%s", log); + printf("-------- program load log end ----------\n"); + } + close(prog_fd); + close(btf_fd); +} + /* Check that verifier rejects BPF program containing relocation * pointing to non-existent BTF type. */ @@ -120,6 +252,7 @@ static void test_bad_local_id(void) void test_core_reloc_raw(void) { + test_early_core_relo(); if (test__start_subtest("bad_local_id")) test_bad_local_id(); } From 394ae398337c5f87e567f6cd63b937fc2b2f6ddc Mon Sep 17 00:00:00 2001 From: Kumar Kartikeya Dwivedi Date: Fri, 18 Sep 2026 01:32:14 +0200 Subject: [PATCH 170/694] bpf: Restrict CO-RE poisoning to relocatable instructions CO-RE relocation records can name any instruction offset. When a relocation cannot be resolved, bpf_core_patch_insn() currently poisons its target before checking whether that instruction is a valid relocation target. Malformed metadata can therefore replace jumps, calls, exits, register-source arithmetic, or non-immediate loads instead of failing at the relocation step. Handle poisoning only after the instruction has passed the same class and operand-form checks used for a resolved relocation. Route invalid forms through the existing diagnostic and return a hard error. Keep poisoning supported instructions, including both halves of a plain ldimm64, so an unresolved relocation in dead code remains valid. Extend bpf_core_poison_insn() to poison both halves of ldimm64, and return its status directly from each validated instruction case. This avoids routing the success path through a common label and leaves the helper free to report errors. The shared relocation code applies this restriction to both libbpf and in-kernel CO-RE. Fixes: d7a252708dbc ("libbpf: Improve handling of failed CO-RE relocations") Reported-by: Nicholas Carlini Suggested-by: Nicholas Carlini Signed-off-by: Kumar Kartikeya Dwivedi Acked-by: Eduard Zingerman Link: https://patch.msgid.link/20260917233222.2542500-7-memxor@gmail.com Signed-off-by: Eduard Zingerman --- tools/lib/bpf/relo_core.c | 58 +++++++++++++++++++++------------------ 1 file changed, 31 insertions(+), 27 deletions(-) diff --git a/tools/lib/bpf/relo_core.c b/tools/lib/bpf/relo_core.c index 8ad2715721cf..2672623a4198 100644 --- a/tools/lib/bpf/relo_core.c +++ b/tools/lib/bpf/relo_core.c @@ -980,23 +980,30 @@ static int bpf_core_calc_relo(const char *prog_name, } /* - * Turn instruction for which CO_RE relocation failed into invalid one with + * Turn instruction for which CO-RE relocation failed into invalid one with * distinct signature. */ -static void bpf_core_poison_insn(const char *prog_name, int relo_idx, - int insn_idx, struct bpf_insn *insn) +static int bpf_core_poison_insn(const char *prog_name, int relo_idx, + struct bpf_insn *insn, int insn_idx) { - pr_debug("prog '%s': relo #%d: substituting insn #%d w/ invalid insn\n", - prog_name, relo_idx, insn_idx); - insn->code = BPF_JMP | BPF_CALL; - insn->dst_reg = 0; - insn->src_reg = 0; - insn->off = 0; - /* if this instruction is reachable (not a dead code), - * verifier will complain with the following message: - * invalid func unknown#195896080 - */ - insn->imm = 195896080; /* => 0xbad2310 => "bad relo" */ + int insn_cnt = is_ldimm64_insn(insn) ? 2 : 1; + int i; + + for (i = 0; i < insn_cnt; i++) { + pr_debug("prog '%s': relo #%d: substituting insn #%d w/ invalid insn\n", + prog_name, relo_idx, insn_idx + i); + insn[i].code = BPF_JMP | BPF_CALL; + insn[i].dst_reg = 0; + insn[i].src_reg = 0; + insn[i].off = 0; + /* + * If this instruction is reachable (not dead code), the verifier + * will complain with "invalid func unknown#195896080". + */ + insn[i].imm = 195896080; /* => 0xbad2310 => "bad relo" */ + } + + return 0; } static int insn_bpf_size_to_bytes(struct bpf_insn *insn) @@ -1047,17 +1054,6 @@ int bpf_core_patch_insn(const char *prog_name, struct bpf_insn *insn, class = BPF_CLASS(insn->code); - if (res->poison) { -poison: - /* poison second part of ldimm64 to avoid confusing error from - * verifier about "unknown opcode 00" - */ - if (is_ldimm64_insn(insn)) - bpf_core_poison_insn(prog_name, relo_idx, insn_idx + 1, insn + 1); - bpf_core_poison_insn(prog_name, relo_idx, insn_idx, insn); - return 0; - } - orig_val = res->orig_val; new_val = res->new_val; @@ -1065,7 +1061,9 @@ int bpf_core_patch_insn(const char *prog_name, struct bpf_insn *insn, case BPF_ALU: case BPF_ALU64: if (BPF_SRC(insn->code) != BPF_K) - return -EINVAL; + goto bad_insn; + if (res->poison) + return bpf_core_poison_insn(prog_name, relo_idx, insn, insn_idx); if (res->validate && insn->imm != orig_val) { pr_warn("prog '%s': relo #%d: unexpected insn #%d (ALU/ALU64) value: got %d, exp %llu -> %llu\n", prog_name, relo_idx, @@ -1082,6 +1080,8 @@ int bpf_core_patch_insn(const char *prog_name, struct bpf_insn *insn, case BPF_LDX: case BPF_ST: case BPF_STX: + if (res->poison) + return bpf_core_poison_insn(prog_name, relo_idx, insn, insn_idx); if (res->validate && insn->off != orig_val) { pr_warn("prog '%s': relo #%d: unexpected insn #%d (LDX/ST/STX) value: got %d, exp %llu -> %llu\n", prog_name, relo_idx, insn_idx, insn->off, (unsigned long long)orig_val, @@ -1097,7 +1097,7 @@ int bpf_core_patch_insn(const char *prog_name, struct bpf_insn *insn, pr_warn("prog '%s': relo #%d: insn #%d (LDX/ST/STX) accesses field incorrectly. " "Make sure you are accessing pointers, unsigned integers, or fields of matching type and size.\n", prog_name, relo_idx, insn_idx); - goto poison; + return bpf_core_poison_insn(prog_name, relo_idx, insn, insn_idx); } orig_val = insn->off; @@ -1140,6 +1140,9 @@ int bpf_core_patch_insn(const char *prog_name, struct bpf_insn *insn, return -EINVAL; } + if (res->poison) + return bpf_core_poison_insn(prog_name, relo_idx, insn, insn_idx); + imm = (__u32)insn[0].imm | ((__u64)insn[1].imm << 32); if (res->validate && imm != orig_val) { pr_warn("prog '%s': relo #%d: unexpected insn #%d (LDIMM64) value: got %llu, exp %llu -> %llu\n", @@ -1157,6 +1160,7 @@ int bpf_core_patch_insn(const char *prog_name, struct bpf_insn *insn, break; } default: +bad_insn: pr_warn("prog '%s': relo #%d: trying to relocate unrecognized insn #%d, code:0x%x, src:0x%x, dst:0x%x, off:0x%x, imm:0x%x\n", prog_name, relo_idx, insn_idx, insn->code, (unsigned)insn->src_reg, (unsigned)insn->dst_reg, (unsigned)insn->off, (unsigned)insn->imm); From 3440505aca926e726a26c0fd30356454334322bd Mon Sep 17 00:00:00 2001 From: Kumar Kartikeya Dwivedi Date: Fri, 18 Sep 2026 01:32:15 +0200 Subject: [PATCH 171/694] selftests/bpf: Test CO-RE instruction poisoning restrictions Add raw CO-RE relocations that fail to resolve their target enum value. Place each supported and unsupported instruction form in dead code. Unsupported targets must fail relocation with a diagnostic even when they are unreachable. Supported ALU immediates, memory accesses, and ldimm64 instructions must still be poisoned and removed as dead code, allowing the program to load. Check that both halves of ldimm64 are poisoned. Load every instruction stream without relocations first to ensure that rejection is caused by the relocation rather than the original program. Signed-off-by: Kumar Kartikeya Dwivedi Acked-by: Eduard Zingerman Link: https://patch.msgid.link/20260917233222.2542500-8-memxor@gmail.com Signed-off-by: Eduard Zingerman --- .../selftests/bpf/prog_tests/core_reloc_raw.c | 61 ++++++++++++++++++- 1 file changed, 60 insertions(+), 1 deletion(-) diff --git a/tools/testing/selftests/bpf/prog_tests/core_reloc_raw.c b/tools/testing/selftests/bpf/prog_tests/core_reloc_raw.c index bb19e49dd87d..51f42b02a267 100644 --- a/tools/testing/selftests/bpf/prog_tests/core_reloc_raw.c +++ b/tools/testing/selftests/bpf/prog_tests/core_reloc_raw.c @@ -50,6 +50,28 @@ static int load_core_relo_insns(int btf_fd, struct bpf_insn *insns, int insn_cnt static void test_early_core_relo(void) { + static const char unrecognized[] = "trying to relocate unrecognized insn #2"; + static const struct { + const char *name; + struct bpf_insn insns[2]; + const char *err_msg; + } tests[] = { + { "poison_exit", { BPF_EXIT_INSN() }, unrecognized }, + { "poison_ja", { BPF_JMP_A(1) }, unrecognized }, + { "poison_jmp", { BPF_JMP_IMM(BPF_JEQ, BPF_REG_0, 0, 1) }, unrecognized }, + { "poison_jmp32", { BPF_JMP32_IMM(BPF_JEQ, BPF_REG_0, 0, 1) }, unrecognized }, + { "poison_call", { BPF_EMIT_CALL(BPF_FUNC_get_prandom_u32) }, unrecognized }, + { "poison_alu_reg", { BPF_MOV32_REG(BPF_REG_0, BPF_REG_1) }, unrecognized }, + { "poison_alu64_reg", { BPF_MOV64_REG(BPF_REG_0, BPF_REG_1) }, unrecognized }, + { "poison_ld_abs", { BPF_LD_ABS(BPF_W, 0) }, + "insn #2 (LDIMM64) has unexpected form" }, + { "poison_alu_imm", { BPF_MOV32_IMM(BPF_REG_0, 0) } }, + { "poison_alu64_imm", { BPF_MOV64_IMM(BPF_REG_0, 0) } }, + { "poison_ldx", { BPF_LDX_MEM(BPF_W, BPF_REG_0, BPF_REG_1, 0) } }, + { "poison_st", { BPF_ST_MEM(BPF_W, BPF_REG_10, -4, 0) } }, + { "poison_stx", { BPF_STX_MEM(BPF_W, BPF_REG_10, BPF_REG_0, -4) } }, + { "poison_ldimm64", { BPF_LD_IMM64(BPF_REG_0, 0) } }, + }; struct test_btf { struct btf_header hdr; __u32 types[18]; @@ -97,7 +119,7 @@ static void test_early_core_relo(void) }; int access_str_off = 51; /* offset of "0" */ int enum_id = 5; - int btf_fd, prog_fd = -1; + int btf_fd, prog_fd = -1, i; btf_fd = bpf_btf_load(&raw_btf, sizeof(raw_btf), NULL); if (!ASSERT_GE(btf_fd, 0, "btf_load")) @@ -136,6 +158,43 @@ static void test_early_core_relo(void) ASSERT_HAS_SUBSTR(log, "invalid bpf_ld_imm64 insn", "truncated_load_log"); } + for (i = 0; i < ARRAY_SIZE(tests); i++) { + struct bpf_insn insns[] = { + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_JMP_IMM(BPF_JEQ, BPF_REG_0, 0, 1), + tests[i].insns[0], + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_EXIT_INSN(), + }; + bool is_ldimm64 = insns[2].code == (BPF_LD | BPF_DW | BPF_IMM); + + if (!test__start_subtest(tests[i].name)) + continue; + if (is_ldimm64) { + insns[1].off = 2; + insns[3] = tests[i].insns[1]; + } + prog_fd = load_core_relo_insns(btf_fd, insns, ARRAY_SIZE(insns), funcs, 1, + enum_id, access_str_off, 2, false); + if (!ASSERT_GE(prog_fd, 0, "control_load")) + goto cleanup; + close(prog_fd); + prog_fd = load_core_relo_insns(btf_fd, insns, ARRAY_SIZE(insns), funcs, 1, + enum_id, access_str_off, 2, true); + if (!tests[i].err_msg) { + ASSERT_GE(prog_fd, 0, "dead_poison_load"); + ASSERT_HAS_SUBSTR(log, "substituting insn #2", "poison_log"); + if (is_ldimm64) + ASSERT_HAS_SUBSTR(log, "substituting insn #3", "poison_ldimm64_log"); + } else { + ASSERT_LT(prog_fd, 0, "invalid_poison_load"); + ASSERT_HAS_SUBSTR(log, tests[i].err_msg, "invalid_poison_log"); + ASSERT_NULL(strstr(log, "substituting insn"), "invalid_poison_substitution"); + } + close(prog_fd); + prog_fd = -1; + } + cleanup: if (env.verbosity > VERBOSE_NORMAL && log[0]) { printf("-------- program load log start --------\n"); From 71919742c83c32afcccf06a7a28e28f3f55f21a2 Mon Sep 17 00:00:00 2001 From: Kumar Kartikeya Dwivedi Date: Fri, 18 Sep 2026 01:32:16 +0200 Subject: [PATCH 172/694] bpf: Assign lock identity to callback map values A nested bpf_for_each_map_elem() callback can unlock a different element of the same map: static long inner(void *map, int *key, struct value *v, struct value **outer_value) { bpf_spin_lock(&v->lock); bpf_spin_unlock(&(*outer_value)->lock); return 0; } static long outer(void *map, int *key, struct value *v, void *ctx) { bpf_for_each_map_elem(map, inner, &v, 0); return 0; } Both callback values currently have ID zero and the same map_ptr. process_spin_lock() compares those two fields, so it accepts the unlock even though the two callbacks can receive different map elements. Assign a fresh ID to every callback map value in the for-each, timer/workqueue, and task-work constructors. Copies of one callback argument retain its ID, so locking and unlocking through that argument continues to work. Distinct callbacks also get distinct IDs for single-element arrays, including inner arrays sharing inner_map_meta. Preserve map_uid for every inner-map lookup and compare it through check_ids() during state pruning. This preserves relationships between maps, keys, and values while allowing equivalent states with different lookup IDs to match. It avoids field-specific rules for when an inner map needs an identity. Move map_uid out of the metadata union and next to the other IDs, so register comparisons can use the existing memcmp() ranges and remap the IDs separately. Clear it when resetting a register or converting a map lookup result to a socket pointer. Shrink frameno to u8, which is enough for MAX_CALL_FRAMES, to make room without growing bpf_reg_state. Fixes: d0d78c1df9b1 ("bpf: Allow locking bpf_spin_lock global variables") Reported-by: Nicholas Carlini Suggested-by: Nicholas Carlini Signed-off-by: Kumar Kartikeya Dwivedi Acked-by: Eduard Zingerman Link: https://patch.msgid.link/20260917233222.2542500-9-memxor@gmail.com Signed-off-by: Eduard Zingerman --- include/linux/bpf_verifier.h | 25 +++++++++++++------------ kernel/bpf/states.c | 6 ++++-- kernel/bpf/verifier.c | 13 +++++++++---- 3 files changed, 26 insertions(+), 18 deletions(-) diff --git a/include/linux/bpf_verifier.h b/include/linux/bpf_verifier.h index bba5a727c651..a7202b44ab10 100644 --- a/include/linux/bpf_verifier.h +++ b/include/linux/bpf_verifier.h @@ -46,17 +46,11 @@ struct bpf_reg_state { /* valid when type == PTR_TO_PACKET */ int range; - /* valid when type == CONST_PTR_TO_MAP | PTR_TO_MAP_VALUE | - * PTR_TO_MAP_VALUE_OR_NULL + /* + * For CONST_PTR_TO_MAP, PTR_TO_MAP_KEY, PTR_TO_MAP_VALUE and + * PTR_TO_INSN. */ - struct { - struct bpf_map *map_ptr; - /* To distinguish map lookups from outer map - * the map_uid is non-zero for registers - * pointing to inner maps. - */ - u32 map_uid; - }; + struct bpf_map *map_ptr; /* for PTR_TO_BTF_ID */ struct { @@ -155,13 +149,20 @@ struct bpf_reg_state { * gets parent_id set to the dynptr's id. */ u32 parent_id; - /* Inside the callee two registers can be both PTR_TO_STACK like + /* + * Distinguishes inner-map lookups and their keys and values. Zero for + * other registers. Kept outside the metadata union for ID remapping + * during state comparisons. + */ + u32 map_uid; + /* + * Inside the callee two registers can be both PTR_TO_STACK like * R1=fp-8 and R2=fp-8, but one of them points to this function stack * while another to the caller's stack. To differentiate them 'frameno' * is used which is an index in bpf_verifier_state->frame[] array * pointing to bpf_func_state. */ - u32 frameno; + u8 frameno; /* if (!precise && SCALAR_VALUE) min/max/tnum don't affect safety */ bool precise; }; diff --git a/kernel/bpf/states.c b/kernel/bpf/states.c index 6c88ad95b63b..e4ec007f7fa6 100644 --- a/kernel/bpf/states.c +++ b/kernel/bpf/states.c @@ -491,7 +491,8 @@ static bool regs_exact(const struct bpf_reg_state *rold, { return memcmp(rold, rcur, offsetof(struct bpf_reg_state, id)) == 0 && check_ids(rold->id, rcur->id, idmap) && - check_ids(rold->parent_id, rcur->parent_id, idmap); + check_ids(rold->parent_id, rcur->parent_id, idmap) && + check_ids(rold->map_uid, rcur->map_uid, idmap); } enum exact_level { @@ -616,7 +617,8 @@ static bool regsafe(struct bpf_verifier_env *env, struct bpf_reg_state *rold, range_within(rold, rcur) && tnum_in(rold->var_off, rcur->var_off) && check_ids(rold->id, rcur->id, idmap) && - check_ids(rold->parent_id, rcur->parent_id, idmap); + check_ids(rold->parent_id, rcur->parent_id, idmap) && + check_ids(rold->map_uid, rcur->map_uid, idmap); case PTR_TO_PACKET_META: case PTR_TO_PACKET: /* We must have at least as much range as the old ptr diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 33161dc64568..02be326f235c 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -1864,6 +1864,7 @@ static void __mark_reg_known(struct bpf_reg_state *reg, u64 imm) offsetof(struct bpf_reg_state, var_off) - sizeof(reg->type)); reg->id = 0; reg->parent_id = 0; + reg->map_uid = 0; ___mark_reg_known(reg, imm); } @@ -1925,17 +1926,18 @@ static void refine_map_lookup_value(struct bpf_reg_state *reg) if (map->inner_map_meta) { reg->type = CONST_PTR_TO_MAP | maybe_null; reg->map_ptr = map->inner_map_meta; - /* transfer reg's id which is unique for every map_lookup_elem + /* + * transfer reg's id which is unique for every map_lookup_elem * as UID of the inner map. */ - if (btf_record_has_field(map->inner_map_meta->record, - BPF_TIMER | BPF_WORKQUEUE | BPF_TASK_WORK)) - reg->map_uid = reg->id; + reg->map_uid = reg->id; } else if (map->map_type == BPF_MAP_TYPE_XSKMAP) { reg->type = PTR_TO_XDP_SOCK | maybe_null; + reg->map_uid = 0; } else if (map->map_type == BPF_MAP_TYPE_SOCKMAP || map->map_type == BPF_MAP_TYPE_SOCKHASH) { reg->type = PTR_TO_SOCKET | maybe_null; + reg->map_uid = 0; } } @@ -10048,6 +10050,7 @@ int map_set_for_each_callback_args(struct bpf_verifier_env *env, __mark_reg_known_zero(&callee->regs[BPF_REG_3]); callee->regs[BPF_REG_3].map_ptr = caller->regs[BPF_REG_1].map_ptr; callee->regs[BPF_REG_3].map_uid = caller->regs[BPF_REG_1].map_uid; + callee->regs[BPF_REG_3].id = ++env->id_gen; /* pointer to stack or null */ callee->regs[BPF_REG_4] = caller->regs[BPF_REG_3]; @@ -10144,6 +10147,7 @@ static int set_timer_callback_state(struct bpf_verifier_env *env, __mark_reg_known_zero(&callee->regs[BPF_REG_3]); callee->regs[BPF_REG_3].map_ptr = map_ptr; callee->regs[BPF_REG_3].map_uid = map_uid; + callee->regs[BPF_REG_3].id = ++env->id_gen; /* unused */ bpf_mark_reg_not_init(env, &callee->regs[BPF_REG_4]); @@ -10262,6 +10266,7 @@ static int set_task_work_schedule_callback_state(struct bpf_verifier_env *env, __mark_reg_known_zero(&callee->regs[BPF_REG_3]); callee->regs[BPF_REG_3].map_ptr = map_ptr; callee->regs[BPF_REG_3].map_uid = map_uid; + callee->regs[BPF_REG_3].id = ++env->id_gen; /* unused */ bpf_mark_reg_not_init(env, &callee->regs[BPF_REG_4]); From 04ae4ffc57a6b7a8215779f0e9c536cacda9f761 Mon Sep 17 00:00:00 2001 From: Kumar Kartikeya Dwivedi Date: Fri, 18 Sep 2026 01:32:17 +0200 Subject: [PATCH 173/694] selftests/bpf: Check callback map value lock identity Add a verifier test which retains a map value from an outer callback and then acquires a lock through an inner callback value before attempting to release the outer callback value. Both values can denote different elements, so the verifier must reject the mismatched unlock. Also exercise callbacks reached through two inner-map lookups. The lookup results share inner_map_meta but may refer to different one-element arrays, so their callback values must retain distinct lock identities. Extend the existing spin_lock failure table and reuse its array and inner-map fixtures to keep these cases alongside the other lock identity tests. Update the nested callback reference-leak expectation for the extra callback value ID. Signed-off-by: Kumar Kartikeya Dwivedi Link: https://patch.msgid.link/20260917233222.2542500-10-memxor@gmail.com Signed-off-by: Eduard Zingerman --- .../selftests/bpf/prog_tests/cb_refs.c | 2 +- .../selftests/bpf/prog_tests/spin_lock.c | 2 + .../selftests/bpf/progs/test_spin_lock_fail.c | 67 ++++++++++++++++++- 3 files changed, 68 insertions(+), 3 deletions(-) diff --git a/tools/testing/selftests/bpf/prog_tests/cb_refs.c b/tools/testing/selftests/bpf/prog_tests/cb_refs.c index 78566b817fd7..490e15e7126d 100644 --- a/tools/testing/selftests/bpf/prog_tests/cb_refs.c +++ b/tools/testing/selftests/bpf/prog_tests/cb_refs.c @@ -13,7 +13,7 @@ struct { } cb_refs_tests[] = { { "underflow_prog", "release kfunc bpf_kfunc_call_test_release expects referenced PTR_TO_BTF_ID passed to R1" }, { "leak_prog", "Possibly NULL pointer passed to helper R2" }, - { "nested_cb", "Unreleased reference id=4 alloc_insn=2" }, /* alloc_insn=2{4,5} */ + { "nested_cb", "Unreleased reference id=5 alloc_insn=2" }, /* alloc_insn=2{4,5} */ { "non_cb_transfer_ref", "Unreleased reference id=4 alloc_insn=1" }, /* alloc_insn=1{1,2} */ }; diff --git a/tools/testing/selftests/bpf/prog_tests/spin_lock.c b/tools/testing/selftests/bpf/prog_tests/spin_lock.c index 5c3579438427..e368370262c8 100644 --- a/tools/testing/selftests/bpf/prog_tests/spin_lock.c +++ b/tools/testing/selftests/bpf/prog_tests/spin_lock.c @@ -54,6 +54,8 @@ static struct { { "lock_global_sleepable_helper_subprog", "global function calls are not allowed while holding a lock" }, { "lock_global_sleepable_kfunc_subprog", "global function calls are not allowed while holding a lock" }, { "lock_global_sleepable_subprog_indirect", "global function calls are not allowed while holding a lock" }, + { "callback_value_lock_identity", "bpf_spin_unlock of different lock" }, + { "callback_inner_map_value_lock_identity", "bpf_spin_unlock of different lock" }, }; static int match_regex(const char *pattern, const char *string) diff --git a/tools/testing/selftests/bpf/progs/test_spin_lock_fail.c b/tools/testing/selftests/bpf/progs/test_spin_lock_fail.c index f678ee6bd7ea..55282f20fa32 100644 --- a/tools/testing/selftests/bpf/progs/test_spin_lock_fail.c +++ b/tools/testing/selftests/bpf/progs/test_spin_lock_fail.c @@ -14,17 +14,18 @@ struct array_map { __type(key, int); __type(value, struct foo); __uint(max_entries, 1); -} array_map SEC(".maps"); +} array_map SEC(".maps"), array_map_b SEC(".maps"); struct { __uint(type, BPF_MAP_TYPE_ARRAY_OF_MAPS); - __uint(max_entries, 1); + __uint(max_entries, 2); __type(key, int); __type(value, int); __array(values, struct array_map); } map_of_maps SEC(".maps") = { .values = { [0] = &array_map, + [1] = &array_map_b, }, }; @@ -314,4 +315,66 @@ int lock_global_sleepable_subprog_indirect(struct __sk_buff *ctx) return ret; } +struct { + __uint(type, BPF_MAP_TYPE_ARRAY); + __uint(max_entries, 2); + __type(key, int); + __type(value, struct foo); +} callback_array_map SEC(".maps"); + +struct callback_ctx { + struct foo *value; +}; + +static long lock_different_value(struct bpf_map *map, int *key, + struct foo *value, struct callback_ctx *ctx) +{ + bpf_spin_lock(&value->lock); + bpf_spin_unlock(&ctx->value->lock); + return 0; +} + +static long nest_lock_different_value(struct bpf_map *map, int *key, + struct foo *value, void *data) +{ + struct callback_ctx ctx = { .value = value }; + + bpf_for_each_map_elem(&callback_array_map, lock_different_value, &ctx, 0); + return 0; +} + +SEC("?tc") +int callback_value_lock_identity(void *ctx) +{ + bpf_for_each_map_elem(&callback_array_map, nest_lock_different_value, NULL, 0); + return 0; +} + +static long nest_lock_different_inner_value(struct bpf_map *map, int *key, + struct foo *value, void *data) +{ + struct callback_ctx ctx = { .value = value }; + int inner_key = 1; + void *inner_map; + + inner_map = bpf_map_lookup_elem(&map_of_maps, &inner_key); + if (!inner_map) + return 0; + bpf_for_each_map_elem(inner_map, lock_different_value, &ctx, 0); + return 0; +} + +SEC("?tc") +int callback_inner_map_value_lock_identity(void *ctx) +{ + int inner_key = 0; + void *inner_map; + + inner_map = bpf_map_lookup_elem(&map_of_maps, &inner_key); + if (!inner_map) + return 0; + bpf_for_each_map_elem(inner_map, nest_lock_different_inner_value, NULL, 0); + return 0; +} + char _license[] SEC("license") = "GPL"; From b4e875d397da451fb4e9c573ff4b86db53caba05 Mon Sep 17 00:00:00 2001 From: Kumar Kartikeya Dwivedi Date: Fri, 18 Sep 2026 01:32:18 +0200 Subject: [PATCH 174/694] libbpf: Reject truncated ldimm64 CO-RE relocations CO-RE relocation of an ldimm64 instruction operates on two instruction slots. A malformed BPF ELF can end a function after the first slot and attach a CO-RE relocation to it. libbpf allocates the instruction array according to the function symbol size, so the shared relocation code would then access beyond the allocation. Reject a terminal ldimm64 in libbpf's relocation loop, where the program length is available, before resolving or applying the relocation. Both resolved and unresolved relocations validate the absent second slot, and unresolved relocation poisoning would additionally write past the array. The in-kernel caller is protected by the verifier's early instruction-stream check before it applies CO-RE relocations. Fixes: eacaaed784e2 ("libbpf: Implement enum value-based CO-RE relocations") Reported-by: Sashiko Signed-off-by: Kumar Kartikeya Dwivedi Link: https://lore.kernel.org/20260914140852.03DA21F0089B@smtp.kernel.org Link: https://patch.msgid.link/20260917233222.2542500-11-memxor@gmail.com Signed-off-by: Eduard Zingerman --- tools/lib/bpf/libbpf.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c index b749c01742ee..bfa64ae6c94d 100644 --- a/tools/lib/bpf/libbpf.c +++ b/tools/lib/bpf/libbpf.c @@ -6206,6 +6206,13 @@ bpf_object__relocate_core(struct bpf_object *obj, const char *targ_btf_path) return -EINVAL; insn = &prog->insns[insn_idx]; + if (is_ldimm64_insn(insn) && (size_t)insn_idx + 1 >= prog->insns_cnt) { + pr_warn("prog '%s': relo #%d: insn #%d (LDIMM64) is truncated\n", + prog->name, i, insn_idx); + err = -EINVAL; + goto out; + } + err = record_relo_core(prog, rec, insn_idx); if (err) { pr_warn("prog '%s': relo #%d: failed to record relocation: %s\n", From 2ec28c09b320ba241bea8a70ee5cb9ccf4a099e8 Mon Sep 17 00:00:00 2001 From: Aldo Ariel Panzardo Date: Tue, 15 Sep 2026 14:30:50 -0300 Subject: [PATCH 175/694] vsock: ignore empty child namespace mode writes __vsock_net_mode_string() returns success without updating new_mode when the transfer length is zero. Its caller then reads the uninitialized enum and may permanently store a stack-derived value in the write-once child mode. Return before calling __vsock_net_mode_string() when *lenp is zero so that the helper is never invoked with nothing to parse and new_mode is never read uninitialized. This also prevents an empty write from locking the current mode. Fixes: eafb64f40ca4 ("vsock: add netns to vsock core") Cc: stable@vger.kernel.org Reviewed-by: Luigi Leonardi Signed-off-by: Aldo Ariel Panzardo Reviewed-by: Stefano Garzarella Reviewed-by: Bobby Eshleman Link: https://patch.msgid.link/20260915173050.3176344-1-qwe.aldo@gmail.com Signed-off-by: Jakub Kicinski --- net/vmw_vsock/af_vsock.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/net/vmw_vsock/af_vsock.c b/net/vmw_vsock/af_vsock.c index f840498b58af..9b71479a2b29 100644 --- a/net/vmw_vsock/af_vsock.c +++ b/net/vmw_vsock/af_vsock.c @@ -2889,6 +2889,9 @@ static int vsock_net_child_mode_string(const struct ctl_table *table, int write, net = container_of(table->data, struct net, vsock.child_ns_mode); + if (!*lenp) + return 0; + ret = __vsock_net_mode_string(table, write, buffer, lenp, ppos, vsock_net_child_mode(net), &new_mode); if (ret) From 651010592bdce7005c1179498327e51bfc4fe1a5 Mon Sep 17 00:00:00 2001 From: Zhang Yunfei Date: Fri, 11 Sep 2026 17:11:23 +0800 Subject: [PATCH 176/694] net: txgbe: fix FDIR filter restore for VF rules txgbe_fdir_filter_restore() reprograms every filter from txgbe->fdir_filter_list after a reset. It extracts the ring part of filter->action with ethtool_get_flow_spec_ring() and maps it onto a PF rx ring, silently dropping the VF part of the cookie that txgbe_add_ethtool_fdir_entry() stores there (input->action = fsp->ring_cookie). For a rule directed at a VF, restore therefore reprograms the filter to the PF queue with the same ring index: after any down/up or txgbe_reinit_locked(), traffic matching the rule is steered to the PF instead of the VF. Handle VF rules the same way txgbe_add_ethtool_fdir_entry() does: validate vf against wx->num_vfs and ring against wx->num_rx_queues_per_pool, and map the ring onto the absolute queue index ((vf - 1) * wx->num_rx_queues_per_pool) + ring. Fixes: 7a91722e0dd4 ("net: txgbe: Support the FDIR rules assigned to VFs") Cc: stable@vger.kernel.org Signed-off-by: Zhang Yunfei Link: https://patch.msgid.link/20260911091123.798931-1-zhangyunfei1@kylinos.cn Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/wangxun/txgbe/txgbe_fdir.c | 13 +++++++++++-- 1 file changed, 11 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/wangxun/txgbe/txgbe_fdir.c b/drivers/net/ethernet/wangxun/txgbe/txgbe_fdir.c index a84010828551..59a47532618c 100644 --- a/drivers/net/ethernet/wangxun/txgbe/txgbe_fdir.c +++ b/drivers/net/ethernet/wangxun/txgbe/txgbe_fdir.c @@ -591,15 +591,24 @@ static void txgbe_fdir_filter_restore(struct wx *wx) queue = TXGBE_RDB_FDIR_DROP_QUEUE; } else { u32 ring = ethtool_get_flow_spec_ring(filter->action); + u8 vf = ethtool_get_flow_spec_ring_vf(filter->action); - if (ring >= wx->num_rx_queues) { + if (!vf && ring >= wx->num_rx_queues) { wx_err(wx, "FDIR restore failed, ring:%u\n", ring); continue; + } else if (vf && (vf > wx->num_vfs || + ring >= wx->num_rx_queues_per_pool)) { + wx_err(wx, "FDIR restore failed, vf:%u, ring:%u\n", + vf, ring); + continue; } /* Map the ring onto the absolute queue index */ - queue = wx->rx_ring[ring]->reg_idx; + if (!vf) + queue = wx->rx_ring[ring]->reg_idx; + else + queue = ((vf - 1) * wx->num_rx_queues_per_pool) + ring; } ret = txgbe_fdir_write_perfect_filter(wx, From 46bc52d13594848023e681860df8700c8db14354 Mon Sep 17 00:00:00 2001 From: Linkui Xiao Date: Wed, 16 Sep 2026 20:53:16 +0800 Subject: [PATCH 177/694] ipv4: fib: fix data-race and stale genid check around nh->nh_saddr fib_select_multipath() compares nexthop_nh->nh_saddr against the flow source address with no lock held, while fib_info_update_nhc_saddr() stores a new value from another CPU as soon as the preferred source address of the egress device changes. Commit 195374d89368 ("ipv4: fib: annotate races around nh->nh_saddr_genid and nh->nh_saddr") added WRITE_ONCE() on the store side and READ_ONCE() in fib_result_prefsrc() after syzbot reported BUG: KCSAN: data-race in fib_select_path / fib_select_path but it only covered that reader. fib_select_multipath(), reached from fib_select_path(), is a second lockless reader of nh->nh_saddr and was left bare. Moreover, nh_saddr is only meaningful when nh_saddr_genid matches dev_addr_genid, as established by commit 436c3b66ec98 ("ipv4: Invalidate nexthop cache nh_saddr more correctly."). fib_select_multipath() skips that validation, so it can score a nexthop using a stale source address and skew the ECMP selection. Annotate both reads with READ_ONCE() and refresh the cached source address via fib_info_update_nhc_saddr() when the genid does not match, mirroring fib_result_prefsrc(). Fixes: 32607a332cfe ("ipv4: prefer multipath nexthop that matches source address") Signed-off-by: Linkui Xiao Reviewed-by: Ido Schimmel Reviewed-by: Eric Dumazet Link: https://patch.msgid.link/20260916125316.988044-1-xiaolinkui@126.com Signed-off-by: Jakub Kicinski --- net/ipv4/fib_semantics.c | 13 ++++++++++++- 1 file changed, 12 insertions(+), 1 deletion(-) diff --git a/net/ipv4/fib_semantics.c b/net/ipv4/fib_semantics.c index 7a362f2e2c2b..5c9021ea3a79 100644 --- a/net/ipv4/fib_semantics.c +++ b/net/ipv4/fib_semantics.c @@ -2176,6 +2176,15 @@ static bool fib_good_nh(const struct fib_nh *nh) return !!(state & NUD_VALID); } +static __be32 fib_nh_saddr(struct net *net, const struct fib_info *fi, + struct fib_nh *nh, int genid) +{ + if (READ_ONCE(nh->nh_saddr_genid) == genid) + return READ_ONCE(nh->nh_saddr); + + return fib_info_update_nhc_saddr(net, &nh->nh_common, fi->fib_scope); +} + void fib_select_multipath(struct fib_result *res, int hash, const struct flowi4 *fl4) { @@ -2184,6 +2193,7 @@ void fib_select_multipath(struct fib_result *res, int hash, bool use_neigh; int score = -1; __be32 saddr; + int genid; if (unlikely(res->fi->nh)) { nexthop_path_fib_result(res, hash); @@ -2192,6 +2202,7 @@ void fib_select_multipath(struct fib_result *res, int hash, use_neigh = READ_ONCE(net->ipv4.sysctl_fib_multipath_use_neigh); saddr = fl4 ? fl4->saddr : 0; + genid = saddr ? atomic_read(&net->ipv4.dev_addr_genid) : 0; change_nexthops(fi) { int nh_upper_bound, nh_score = 0; @@ -2204,7 +2215,7 @@ void fib_select_multipath(struct fib_result *res, int hash, (use_neigh && !fib_good_nh(nexthop_nh))) continue; - if (saddr && nexthop_nh->nh_saddr == saddr) + if (saddr && fib_nh_saddr(net, fi, nexthop_nh, genid) == saddr) nh_score += 2; if (hash <= nh_upper_bound) nh_score++; From 5179241521401ef364294128bb43cdbce7252457 Mon Sep 17 00:00:00 2001 From: Christian Brauner Date: Thu, 17 Sep 2026 13:20:25 +0200 Subject: [PATCH 178/694] fs: don't create the private nullfs mount under namespace_sem init_mount_tree() mounts the mutable rootfs on top of nullfs via LOCK_MOUNT_EXACT(). That declares a pinned mountpoint with a cleanup attribute in the scope of the whole function so the nullfs root inode lock and namespace_sem are only dropped when init_mount_tree() returns. This became a problem when the private nullfs instance for kthreads was added. kern_mount() allocates a new superblock and alloc_super() takes the new s_umount with SINGLE_DEPTH_NESTING and then shrinker_mutex via shrinker_alloc(). Doing that with namespace_sem held teaches lockdep the dependency namespace_sem -> s_umount/1 -> shrinker_mutex With CONFIG_SHRINKER_DEBUG shrinker_debugfs_rename() takes the debugfs directory inode lock under shrinker_mutex every time a block device is mounted and lock_mount_exact() takes namespace_sem under the inode lock of the mountpoint for every mount. So mounting anything on debugfs, e.g. the tracefs automount on /sys/kernel/debug/tracing, closes the cycle: WARNING: possible circular locking dependency detected 7.3.0-rc3+ #17 Not tainted ------------------------------------------------------ rasdaemon/4449 is trying to acquire lock: (namespace_sem){++++}-{4:4}, at: lock_mount_exact+0x4c/0x308 but task is already holding lock: (&sb->s_type->i_mutex_key#17){++++}-{4:4}, at: lock_mount_exact+0x3c/0x308 which lock already depends on the new lock. ... Chain exists of: namespace_sem --> shrinker_mutex --> &sb->s_type->i_mutex_key#17 This can't actually deadlock. init_mount_tree() runs single-threaded during early boot before any other task exists and nothing allocates a superblock under namespace_sem after that. But lockdep can't know that and disables itself for the rest of the boot. Move mounting the rootfs on top of nullfs into a helper so the locks are dropped when it returns. Fixes: 32750c77e811 ("fs: start all kthreads in nullfs") Reported-by: Zenghui Yu Closes: https://lore.kernel.org/15174353-3f4a-a1ca-5bd1-ea2a4c77828e@huawei.com Link: https://patch.msgid.link/20260917-atemtechnik-bleichen-befassen-9a57db01baf0@brauner Signed-off-by: Christian Brauner (Amutable) --- fs/namespace.c | 25 ++++++++++++++++--------- 1 file changed, 16 insertions(+), 9 deletions(-) diff --git a/fs/namespace.c b/fs/namespace.c index ae5dc64f8b45..5e41021eaa63 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -6184,6 +6184,21 @@ struct mnt_namespace init_mnt_ns = { .poll = __WAIT_QUEUE_HEAD_INITIALIZER(init_mnt_ns.poll), }; +static void __init mount_rootfs_on_nullfs(struct vfsmount *mnt, + struct vfsmount *nullfs_mnt) +{ + struct path root = { + .mnt = nullfs_mnt, + .dentry = nullfs_mnt->mnt_root, + }; + + LOCK_MOUNT_EXACT(mp, &root); + if (unlikely(IS_ERR(mp.parent))) + panic("VFS: Failed to mount rootfs on nullfs"); + scoped_guard(mount_writer) + attach_mnt(real_mount(mnt), mp.parent, mp.mp); +} + static void __init init_mount_tree(void) { struct vfsmount *mnt, *nullfs_mnt; @@ -6215,15 +6230,7 @@ static void __init init_mount_tree(void) mnt_root = real_mount(nullfs_mnt); init_mnt_ns.root = mnt_root; - /* Mount mutable rootfs on top of nullfs. */ - root.mnt = nullfs_mnt; - root.dentry = nullfs_mnt->mnt_root; - - LOCK_MOUNT_EXACT(mp, &root); - if (unlikely(IS_ERR(mp.parent))) - panic("VFS: Failed to mount rootfs on nullfs"); - scoped_guard(mount_writer) - attach_mnt(real_mount(mnt), mp.parent, mp.mp); + mount_rootfs_on_nullfs(mnt, nullfs_mnt); pr_info("VFS: Finished mounting rootfs on nullfs\n"); From d54a489c8c4b627775445d54a6cbd32f209bda8f Mon Sep 17 00:00:00 2001 From: Frank Wunderlich Date: Thu, 17 Sep 2026 17:37:11 +0200 Subject: [PATCH 179/694] gpiolib: use of_node_name if line-name is missing Until v7.0, GPIO hogs inherited the DT node name when no line-name property was specified. This was implemented as a fallback in of_parse_own_gpio(). Commit d1d564ec4992 ("gpio: move hogs into GPIO core") moved hog parsing into the GPIO core and removed this fallback. Consequently, GPIO hogs without a line-name property are now displayed with a ? in /sys/kernel/debug/gpio. Restore the old fallback. Fixes: d1d564ec4992 ("gpio: move hogs into GPIO core") Signed-off-by: Frank Wunderlich Reviewed-by: Andy Shevchenko Link: https://patch.msgid.link/20260917153712.134367-1-linux@fw-web.de Signed-off-by: Bartosz Golaszewski --- drivers/gpio/gpiolib.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/drivers/gpio/gpiolib.c b/drivers/gpio/gpiolib.c index ef8ccaf17c9c..28f7265c5d56 100644 --- a/drivers/gpio/gpiolib.c +++ b/drivers/gpio/gpiolib.c @@ -1029,6 +1029,13 @@ int gpiochip_add_hog(struct gpio_chip *gc, struct fwnode_handle *fwnode) ret = of_gpiochip_get_lflags(gc, &gpiospec, &lflags); if (ret) return ret; + + /* + * If no line-name property is present, fall back to the OF + * node name as in the previous implementation. + */ + if (!name) + name = to_of_node(fwnode)->name; } else { /* * GPIO_ACTIVE_LOW is currently the only lookup flag From d644b23afe1ef509c9961a6d84a093c2587edf02 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?J=C3=A9r=C3=A9my=20Jean?= Date: Tue, 18 Aug 2026 20:00:15 +0000 Subject: [PATCH 180/694] netfilter: flowtable: publish HW_DEAD after worker is done MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit flow_offload_work_del() sets NF_FLOW_HW_DEAD before the work handler clears NF_FLOW_HW_PENDING. Once a flow is both HW_DYING and HW_DEAD, a concurrent garbage collection pass can remove it and schedule it for RCU freeing. The offload worker holds neither an RCU read lock nor a reference to the flow. If it is preempted after publishing HW_DEAD, the RCU callback can free the flow before the worker resumes and clears HW_PENDING, resulting in a use-after-free. Move HW_DEAD publication to the common worker epilogue after the pending bit is cleared, making it the final flow access by destroy work. Order all preceding flow accesses before publishing the bit that allows garbage collection to free the object. Fixes: 2c8897953f3b ("netfilter: flowtable: Add pending bit for offload work") Assisted-by: Codex:gpt-5 Signed-off-by: Jérémy Jean Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nf_flow_table_offload.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/net/netfilter/nf_flow_table_offload.c b/net/netfilter/nf_flow_table_offload.c index 801a3dd9ceea..6757fd89c1f1 100644 --- a/net/netfilter/nf_flow_table_offload.c +++ b/net/netfilter/nf_flow_table_offload.c @@ -995,7 +995,6 @@ static void flow_offload_work_del(struct flow_offload_work *offload) flow_offload_tuple_del(offload, FLOW_OFFLOAD_DIR_ORIGINAL); if (test_bit(NF_FLOW_HW_BIDIRECTIONAL, &offload->flow->flags)) flow_offload_tuple_del(offload, FLOW_OFFLOAD_DIR_REPLY); - set_bit(NF_FLOW_HW_DEAD, &offload->flow->flags); } static void flow_offload_tuple_stats(struct flow_offload_work *offload, @@ -1059,6 +1058,12 @@ static void flow_offload_work_handler(struct work_struct *work) } clear_bit(NF_FLOW_HW_PENDING, &offload->flow->flags); + if (offload->cmd == FLOW_CLS_DESTROY) { + /* Publish after the worker's last flow access. */ + smp_mb__before_atomic(); + set_bit(NF_FLOW_HW_DEAD, &offload->flow->flags); + } + kfree(offload); } From 9461613afc59acef44a0071b0dd5075f6e993ffe Mon Sep 17 00:00:00 2001 From: Florian Westphal Date: Thu, 3 Sep 2026 02:41:46 +0200 Subject: [PATCH 181/694] netfilter: nfnetlink_queue: hold nfnl mutex in event notifier We must serialize the release notifier and the config netlink function. A concurrent thread can issue close() which can call the release function while unrelated socket processes UNBIND request for same portid: Oops: general protection fault, [..] RIP: 0010:__instance_destroy+0x60/0x210 [nfnetlink_queue] Call Trace: nfqnl_recv_config+0x9b0/0xdc0 [nfnetlink_queue] nfnetlink_rcv_msg+0x7c2/0xeb0 ? __pfx_nfnetlink_rcv_msg+0x10/0x10 After this, parallel UNBIND and URELEASE events are impossible. This change isn't nice, but its the shortest fix given instances are not refcounted and the nfnetlink config callback drops the rcu read lock early due to need for sleeping allocations. Fixes: 7af4cc3fa158 ("[NETFILTER]: Add "nfnetlink_queue" netfilter queue handler over nfnetlink") Signed-off-by: Florian Westphal Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nfnetlink_queue.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/net/netfilter/nfnetlink_queue.c b/net/netfilter/nfnetlink_queue.c index c727668b0c5b..a3bc00280051 100644 --- a/net/netfilter/nfnetlink_queue.c +++ b/net/netfilter/nfnetlink_queue.c @@ -1593,6 +1593,7 @@ nfqnl_rcv_nl_event(struct notifier_block *this, if (event == NETLINK_URELEASE && n->protocol == NETLINK_NETFILTER) { int i; + nfnl_lock(NFNL_SUBSYS_QUEUE); /* destroy all instances for this portid */ spin_lock(&q->instances_lock); for (i = 0; i < INSTANCE_BUCKETS; i++) { @@ -1606,6 +1607,7 @@ nfqnl_rcv_nl_event(struct notifier_block *this, } } spin_unlock(&q->instances_lock); + nfnl_unlock(NFNL_SUBSYS_QUEUE); } return NOTIFY_DONE; } @@ -1925,9 +1927,9 @@ static int nfqnl_recv_config(struct sk_buff *skb, const struct nfnl_info *info, /* Lookup queue under RCU. After peer_portid check (or for new queue * in BIND case), the queue is owned by the socket sending this message. - * A socket cannot simultaneously send a message and close, so while - * processing this CONFIG message, nfqnl_rcv_nl_event() (triggered by - * socket close) cannot destroy this queue. Safe to use without RCU. + * nfqnl_rcv_nl_event() will block on the nfnl subsys mutex that is + * held by the caller, so the queue cannot be destroyed in parallel, + * even after we drop the RCU read lock. */ rcu_read_lock(); queue = instance_lookup(q, queue_num); From 1b9b5323725e458906c7620a3bc10398b51ad954 Mon Sep 17 00:00:00 2001 From: Weiming Shi Date: Sun, 6 Sep 2026 16:44:10 +0800 Subject: [PATCH 182/694] netfilter: ip6t_rpfilter: reject routes without inet6_dev ip6_route_lookup() can return an error-free route whose rt6i_idev is NULL. Lowering an external nexthop device's MTU below IPV6_MIN_MTU tears down its inet6_dev while fib6_ifdown() leaves routes using nexthop objects in the FIB. An unprivileged user can construct this state with rtnetlink in a private user and network namespace, then trigger a NULL dereference through an IPv6 rpfilter lookup: Oops: general protection fault, probably for non-canonical address 0xdffffc0000000000 KASAN: null-ptr-deref in range [0x0000000000000000-0x0000000000000007] RIP: rpfilter_mt (net/ipv6/netfilter/ip6t_rpfilter.c:75) Call Trace: ip6t_do_table (net/ipv6/netfilter/ip6_tables.c:316) nf_hook_slow (net/netfilter/core.c:619) ipv6_rcv (net/ipv6/ip6_input.c:351) __netif_receive_skb_one_core (net/core/dev.c:6216) process_backlog (net/core/dev.c:6680) __napi_poll (net/core/dev.c:7739) net_rx_action (net/core/dev.c:7959) handle_softirqs (kernel/softirq.c:622) do_softirq.part.0 (kernel/softirq.c:523) __local_bh_enable_ip (kernel/softirq.c:450) __dev_queue_xmit (net/core/dev.c:4913) packet_sendmsg (net/packet/af_packet.c:3139) __sys_sendto (net/socket.c:2252) __x64_sys_sendto (net/socket.c:2259) do_syscall_64 (arch/x86/entry/syscall_64.c:94) entry_SYSCALL_64_after_hwframe (arch/x86/entry/entry_64.S:121) Kernel panic - not syncing: Fatal exception in interrupt Reject routes without an inet6_dev immediately after lookup. Such routes are not eligible for reverse-path filtering, and the check protects all later rt6i_idev dereferences. Fixes: e26f9a480fb6 ("netfilter: add ipv6 reverse path filter match") Reported-by: co+459f67f4d8af8ce6@bugs.sh Closes: https://lore.kernel.org/all/VtWUkE8QzJt5CroTj2V2v3ZQ0gwbXZ7nq7I3@bugs.sh/ Suggested-by: Florian Westphal Assisted-by: Claude:gpt-5 Cc: stable@vger.kernel.org Signed-off-by: Weiming Shi Signed-off-by: Pablo Neira Ayuso --- net/ipv6/netfilter/ip6t_rpfilter.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/ipv6/netfilter/ip6t_rpfilter.c b/net/ipv6/netfilter/ip6t_rpfilter.c index 67c87a88cde4..b5def30c3127 100644 --- a/net/ipv6/netfilter/ip6t_rpfilter.c +++ b/net/ipv6/netfilter/ip6t_rpfilter.c @@ -61,7 +61,7 @@ static bool rpfilter_lookup_reverse6(struct net *net, const struct sk_buff *skb, fl6.flowi6_oif = dev->ifindex; rt = (void *)ip6_route_lookup(net, &fl6, skb, lookup_flags); - if (rt->dst.error) + if (rt->dst.error || !rt->rt6i_idev) goto out; if (rt->rt6i_flags & (RTF_REJECT|RTF_ANYCAST)) From 82313c169eddc02b1bf5ba6b427803e272d3ec42 Mon Sep 17 00:00:00 2001 From: Luxiao Xu Date: Sun, 6 Sep 2026 21:29:55 +0800 Subject: [PATCH 183/694] netfilter: ip6t_rt: fix zero-address non-strict match out-of-bounds read rt_mt6_check() permits rules to be configured with rtinfo->addrnr == 0 even when address matching (IP6T_RT_FST_MASK) is requested. In the IP6T_RT_FST_NSTRICT path, rt_mt6() evaluates packet routing addresses against rtinfo->addrs[i] and terminates backwards at the bottom of the loop: if (ipv6_addr_equal(ap, &rtinfo->addrs[i])) { i++; } if (i == rtinfo->addrnr) break; When addrnr is 0, if the first packet address matches rtinfo->addrs[0], i is incremented to 1. Because i is now strictly greater than addrnr (0), the loop termination condition (i == rtinfo->addrnr) is bypassed and will never be satisfied. If a crafted IPv6 packet contains matching routing addresses, i will advance past IP6T_RT_HOPS (16). The subsequent call to ipv6_addr_equal() reads beyond struct ip6t_rt, triggering UBSAN/KASAN out-of-bounds warnings or kernel panics. Fix this by: 1. Rejecting rules in rt_mt6_check() where IP6T_RT_FST_MASK is set but rtinfo->addrnr is zero. 2. In rt_mt6(), moving the termination condition (i < rtinfo->addrnr) into the for-loop header condition and removing the backwards break at the end of the loop body. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Cc: stable@vger.kernel.org Reported-by: Vega Suggested-by: Florian Westphal Assisted-by: LLM Signed-off-by: Luxiao Xu Signed-off-by: Ren Wei Signed-off-by: Pablo Neira Ayuso --- net/ipv6/netfilter/ip6t_rt.c | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-) diff --git a/net/ipv6/netfilter/ip6t_rt.c b/net/ipv6/netfilter/ip6t_rt.c index 8051425213dd..9880faf3cc7d 100644 --- a/net/ipv6/netfilter/ip6t_rt.c +++ b/net/ipv6/netfilter/ip6t_rt.c @@ -96,7 +96,8 @@ static bool rt_mt6(const struct sk_buff *skb, struct xt_action_param *par) unsigned int i = 0; for (temp = 0; - temp < (unsigned int)((hdrlen - 8) / 16); + temp < (unsigned int)((hdrlen - 8) / 16) && + i < rtinfo->addrnr; temp++) { ap = skb_header_pointer(skb, ptr @@ -112,8 +113,6 @@ static bool rt_mt6(const struct sk_buff *skb, struct xt_action_param *par) if (ipv6_addr_equal(ap, &rtinfo->addrs[i])) i++; - if (i == rtinfo->addrnr) - break; } if (i == rtinfo->addrnr) return ret; @@ -162,6 +161,12 @@ static int rt_mt6_check(const struct xt_mtchk_param *par) pr_info_ratelimited("too many addresses specified\n"); return -EINVAL; } + + if ((rtinfo->flags & IP6T_RT_FST_MASK) && !rtinfo->addrnr) { + pr_info_ratelimited("address list match requested but addrnr is 0\n"); + return -EINVAL; + } + if ((rtinfo->flags & (IP6T_RT_RES | IP6T_RT_FST_MASK)) && (!(rtinfo->flags & IP6T_RT_TYP) || (rtinfo->rt_type != 0) || From a311a898172743558b82f6035ef2aa8c310a4223 Mon Sep 17 00:00:00 2001 From: Karl Mehltretter Date: Thu, 10 Sep 2026 22:02:28 +0200 Subject: [PATCH 184/694] netfilter: nft_synproxy: use the family-aware checksum helper nft_synproxy_do_eval() verifies the TCP checksum before it switches on skb->protocol. It uses nf_ip_checksum(), which constructs an IPv4 pseudo header and relies on the IPv4 header checksum when folding the whole skb. Neither operation is valid for an IPv6 packet. A correctly checksummed IPv6 segment can therefore fail verification when it reaches the hook as CHECKSUM_NONE or, at NF_INET_LOCAL_IN, CHECKSUM_COMPLETE. nft_synproxy_do_eval() returns NF_DROP before nft_synproxy_eval_v6() can send a SYN-ACK. nft_synproxy_validate() deliberately admits NFPROTO_IPV6 and NFPROTO_INET, and the xtables counterpart ip6t_SYNPROXY.c already calls nf_ip6_checksum(). Use nf_checksum() with nft_pf() so the checksum helper dispatches to the packet family's implementation. Fixes: ad49d86e07a4 ("netfilter: nf_tables: Add synproxy support") Assisted-by: LLM Signed-off-by: Karl Mehltretter Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nft_synproxy.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/net/netfilter/nft_synproxy.c b/net/netfilter/nft_synproxy.c index 9ed288c9d168..554a96a000f4 100644 --- a/net/netfilter/nft_synproxy.c +++ b/net/netfilter/nft_synproxy.c @@ -118,7 +118,8 @@ static void nft_synproxy_do_eval(const struct nft_synproxy *priv, return; } - if (nf_ip_checksum(skb, nft_hook(pkt), thoff, IPPROTO_TCP)) { + if (nf_checksum(skb, nft_hook(pkt), thoff, IPPROTO_TCP, + nft_pf(pkt))) { regs->verdict.code = NF_DROP; return; } From e290145564886d6a3038810c621f738c1fe9fa51 Mon Sep 17 00:00:00 2001 From: Julian Anastasov Date: Fri, 11 Sep 2026 14:43:15 +0300 Subject: [PATCH 185/694] ipvs: revalidate ihl before icmp_send While the outer IP header is already pulled into the skb head, we must be careful and revalidate the embedded headers after reading them from the skb frags to prevent possible out-of-bounds access. One such place reported by Sashiko is ip_vs_in_icmp() where local process can change the ihl field and after pskb_may_pull() we can see larger value. Even if icmp_send() has checks to prevent out-of-bounds access, play safe and add check to drop the packet if the ihl field is changed. As the outer headers are pulled, make sure the transport header is updated too, it was used before commit 7fcc2fe39fed ("net: icmp: avoid invalid transport header access in icmp_send tracepoint") Fixes: f2edb9f7706d ("ipvs: implement passive PMTUD for IPIP packets") Link: https://sashiko.dev/#/patchset/20260806105211.34622-1-ja%40ssi.bg Signed-off-by: Julian Anastasov Signed-off-by: Pablo Neira Ayuso --- net/netfilter/ipvs/ip_vs_core.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/net/netfilter/ipvs/ip_vs_core.c b/net/netfilter/ipvs/ip_vs_core.c index ba0957798bad..fd503f0efb57 100644 --- a/net/netfilter/ipvs/ip_vs_core.c +++ b/net/netfilter/ipvs/ip_vs_core.c @@ -1960,6 +1960,12 @@ ip_vs_in_icmp(struct netns_ipvs *ipvs, struct sk_buff *skb, int *related, /* Ensure the IP header is present in headroom */ if (!pskb_may_pull(skb, hlen_orig)) goto ignore_tunnel; + skb_set_transport_header(skb, hlen_orig); + /* Before now we may used ihl from skb frag, revalidate it after + * copying it into skb head to prevent out-of-bounds access + */ + if (ip_hdr(skb)->ihl * 4 != hlen_orig) + goto ignore_tunnel; IP_VS_DBG(12, "Sending ICMP for %pI4->%pI4: t=%u, c=%u, i=%u\n", &ip_hdr(skb)->saddr, &ip_hdr(skb)->daddr, type, code, ntohl(info)); From 207d591c353201f3bd3e0c89bb7d44a849c8fd59 Mon Sep 17 00:00:00 2001 From: Naman Gulati Date: Sat, 12 Sep 2026 01:10:51 +0000 Subject: [PATCH 186/694] netfilter: ctnetlink: fix suspicious RCU usage in expect_iter_name expect_iter_name() is invoked by nf_ct_expect_iterate_net() under spin_lock_bh(&nf_conntrack_expect_lock). It does not hold rcu_read_lock(). When accessing exp->helper with rcu_dereference() in syzbot's report, lockdep warns: ============================= WARNING: suspicious RCU usage syzkaller #0 Not tainted ----------------------------- net/netfilter/nf_conntrack_netlink.c:3393 suspicious rcu_dereference_check() usage! locks held by syz-executor381/5628: 2, last CPU#1: #0: ffffffff9aee42a0 (nfnl_subsys_ctnetlink_exp){+.+.}-{4:4}, at: nfnetlink_rcv_msg+0xa69/0x12b0 #1: ffffffff8ea74d58 (nf_conntrack_expect_lock){+...}-{3:3}, at: nf_ct_expect_iterate_net+0x38/0x180 Call Trace: dump_stack_lvl+0xe8/0x150 lockdep_rcu_suspicious+0x140/0x1d0 expect_iter_name+0xfb/0x100 nf_ct_expect_iterate_net+0xf2/0x180 ctnetlink_del_expect+0x45d/0x640 nfnetlink_rcv_msg+0xcc2/0x12b0 netlink_rcv_skb+0x226/0x4a0 nfnetlink_rcv+0x2b9/0x28c0 netlink_unicast+0x7bd/0x940 netlink_sendmsg+0x813/0xb40 ____sys_sendmsg+0x54e/0x850 ___sys_sendmsg+0x2a5/0x360 __sys_sendmsg+0x2a5/0x360 do_syscall_64+0x166/0x520 entry_SYSCALL_64_after_hwframe+0x77/0x7f Use rcu_dereference_protected() with lockdep_is_held() on nf_conntrack_expect_lock instead, similar to expect_iter_me() in nf_conntrack_helper.c. Fixes: f01794106042 ("netfilter: nf_conntrack_expect: use expect->helper") Reported-by: syzbot+4bd730aede2791e40bdf@syzkaller.appspotmail.com Closes: https://lore.kernel.org/netdev/6aa4a377.f81106d8.2ab401.0024.GAE@google.com/T/#u Signed-off-by: Naman Gulati Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nf_conntrack_netlink.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/net/netfilter/nf_conntrack_netlink.c b/net/netfilter/nf_conntrack_netlink.c index 579ada063b1b..4e5d7c701436 100644 --- a/net/netfilter/nf_conntrack_netlink.c +++ b/net/netfilter/nf_conntrack_netlink.c @@ -3392,7 +3392,8 @@ static bool expect_iter_name(struct nf_conntrack_expect *exp, void *data) struct nf_conntrack_helper *helper; const char *name = data; - helper = rcu_dereference(exp->helper); + helper = rcu_dereference_protected(exp->helper, + lockdep_is_held(&nf_conntrack_expect_lock)); if (!helper) return false; From 4f948b5949d2e418b4506752e7c514fe91bd408b Mon Sep 17 00:00:00 2001 From: Chris Mason Date: Fri, 18 Sep 2026 11:19:59 +0200 Subject: [PATCH 187/694] binfmt_misc: fix OOB read in bpf_binprm_select_interp() bpf_binprm_select_interp() checks the name its load program passes with strnlen(name, name__sz) and then hands the same buffer to binfmt_misc_find_interp(), which compares it with an unbounded strcmp(). The buffer can be a BPF map value that another CPU rewrites between the two reads. If the terminating NUL is overwritten in that window, strcmp() reads past the name__sz bytes the verifier checked. That is an out-of-bounds read of up to 31 bytes of whatever follows the checked name__sz bytes. The verifier checks the name and name__sz pair with BPF_READ | BPF_WRITE, so a writable array map value is an accepted argument. bpf(BPF_MAP_UPDATE_ELEM) on an array map copies the new value over the old one in place and takes no lock. The NUL that strnlen() finds can be overwritten before strcmp() reads the buffer again: CPU0 CPU1 bpf_binprm_select_interp() strnlen(name, name__sz) finds the NUL inside name__sz bpf(BPF_MAP_UPDATE_ELEM) array_map_update_elem() copy_map_value() overwrites the NUL binfmt_misc_find_interp() strcmp(interp->name, name) reads past name__sz strnlen() proves that a NUL lies inside name__sz only at the moment it runs. The map update on CPU1 takes no lock, so it can store over the NUL right after. The lookup on CPU0 then walks the live buffer again, once per bound interpreter: fs/binfmt_misc.c:binfmt_misc_find_interp list_for_each_entry(interp, interps, list) if (!strcmp(interp->name, name)) return interp; strcmp() stops at the first mismatch or at the end of interp->name. bm_entry_add_interp() caps a bound name at BINFMT_MISC_INTERP_NAME_MAX (32) bytes, so strcmp() reads at most 33 bytes of name. The smallest name__sz the kfunc accepts is 2, which leaves up to 31 bytes read beyond the checked extent. The handler's own load program has to pass a writable map value, and something has to store into it while the kfunc runs. The window between strnlen() and strcmp() is short, but with a BPF_F_MMAPABLE array the store is a plain user space write into the mapped value, so a loop can hit it without a single bpf() call. Copy the name into a stack buffer of BINFMT_MISC_INTERP_NAME_MAX + 1 bytes, terminate it, and look up the copy. The memcpy() length is below name__sz, so the copy stays inside the extent the verifier checked, and the BPF buffer is not read again afterwards. Return -ENOENT first for a name longer than BINFMT_MISC_INTERP_NAME_MAX. bm_entry_add_interp() rejects a longer name, and the only other binding site attaches the empty name. No entry can bind such a name, so that lookup already ended in -ENOENT and no result changes. Check the first byte of the copy and return -EINVAL if it is NUL, as the existing "!len" test does for an empty name. Only an 'F' entry binds the empty name and a 'B' entry cannot carry 'F', so without that check a racing store of NUL to byte 0 would look up a name no entry binds and end in -ENOENT rather than -EINVAL. A NUL stored further into the name only shortens it to another name the program could have passed anyway. binfmt_misc_find_interp() itself is left alone: entry_attach_interpreter() calls it with a kernel string, and this kfunc now calls it with a private copy. Fixes: 6ec7c96bee30 ("binfmt_misc: let a 'B' entry bind its interpreters") Signed-off-by: Chris Mason Link: https://patch.msgid.link/20260918-work-binfmt_misc-fixes-v1-1-647b24bc1c46@kernel.org Signed-off-by: Christian Brauner (Amutable) --- fs/binfmt_misc_bpf.c | 15 ++++++++++++++- 1 file changed, 14 insertions(+), 1 deletion(-) diff --git a/fs/binfmt_misc_bpf.c b/fs/binfmt_misc_bpf.c index 91576ff05911..ce1bc78e8511 100644 --- a/fs/binfmt_misc_bpf.c +++ b/fs/binfmt_misc_bpf.c @@ -176,6 +176,7 @@ __bpf_kfunc int bpf_binprm_select_interp(struct linux_binprm *bprm, const char *name, size_t name__sz) { const struct binfmt_misc_interp *interp; + char buf[BINFMT_MISC_INTERP_NAME_MAX + 1]; size_t len; char *path; @@ -184,8 +185,20 @@ __bpf_kfunc int bpf_binprm_select_interp(struct linux_binprm *bprm, len = strnlen(name, name__sz); if (len == name__sz || !len) return -EINVAL; + /* No entry binds a longer name, so it cannot be found. */ + if (len > BINFMT_MISC_INTERP_NAME_MAX) + return -ENOENT; - interp = binfmt_misc_find_interp(bprm->bpf_interps, name); + /* + * The program may pass memory that is written to while this runs, + * so look the name up in a private copy and check that instead. + */ + memcpy(buf, name, len); + buf[len] = '\0'; + if (!buf[0]) + return -EINVAL; + + interp = binfmt_misc_find_interp(bprm->bpf_interps, buf); if (!interp) return -ENOENT; From 1970fc4ecb52dabce2e57f9be721964b936a052d Mon Sep 17 00:00:00 2001 From: Chris Mason Date: Fri, 18 Sep 2026 11:20:00 +0200 Subject: [PATCH 188/694] binfmt_misc: fix racy checks in bpf set_interp kfuncs bpf_binprm_set_interp() tests path[0] != '/' on the buffer its load program passes and then reads the same buffer again to copy it with kmemdup_nul(). The buffer can be a BPF map value that another CPU rewrites between the two reads. If byte 0 is overwritten in that window, the kfunc stages a relative or empty interpreter path. The staged path is not checked again, so open_exec() resolves a relative path against the working directory of the task doing the exec. bpf_binprm_set_interp_arg() has the same pattern for its "!len" test and can stage an empty argument, which the interpreter then receives as an empty argv entry. The verifier checks the path and path__sz pair with BPF_READ | BPF_WRITE, so a writable array map value is an accepted argument. bpf(BPF_MAP_UPDATE_ELEM) on an array map copies the new value over the old one in place and takes no lock. Both kfuncs are KF_SLEEPABLE and allocate with GFP_KERNEL between the test and the copy, so the task can sleep inside the window: load program bpf(BPF_MAP_UPDATE_ELEM) bpf_binprm_set_interp() strnlen(path, path__sz) path[0] != '/' is false kmemdup_nul(path, len, GFP_KERNEL) allocation may sleep array_map_update_elem() copy_map_value() rewrites byte 0 copy reads path again bm_bpf_stage_selection() The test in the load program's column proves what byte 0 held only at the moment the test ran. The map update takes no lock, so it can store to byte 0 right after. kmemdup_nul() then copies the rewritten bytes, and bm_bpf_stage_selection() publishes them as bprm->bpf_interp. The staged path is not checked again on its way to open_exec(): load_misc_binary() entry_select_interpreter() returns bprm->bpf_interp unchanged build_interp_argv() copy_string_kernel() copies it as argv[0] bprm_change_interp() kstrdup() entry_open_interpreter() open_exec() unless a bound file is staged or the entry is an 'F' entry None of these functions tests the first byte, and load_misc_binary() hands the pointer to nothing else. In bpf_binprm_set_interp_arg(), strnlen() finds a non-zero len, a NUL is then stored to byte 0, and build_interp_argv() later copies the empty bprm->bpf_interp_arg with copy_string_kernel(). The handler's own load program has to pass a writable map value, and something has to store into it while the kfunc runs. The allocation can sleep inside the window, and with a BPF_F_MMAPABLE array the store is a plain user space write into the mapped value, so a loop can hit it without a single bpf() call. Check the private copy in both kfuncs, so that the string that gets staged is the string that was checked. bpf_binprm_select_interp() already looks its name up in a private copy for the same reason. The remaining tests work on path__sz, arg__sz or the local len, and the copy length is len, so the copy stays inside the extent the verifier checked. Results of bpf_binprm_set_interp() with the check on the copy: - A NUL stored to byte 0 fails interp[0] != '/' and gets -EINVAL. - For len == 0, kmemdup_nul() returns an empty string, so an empty path still gets -EINVAL. - A NUL stored further into the string only shortens it to another absolute path, or another non-empty argument, that the program could have passed anyway. - A path that both lacks the leading '/' and is PATH_MAX or longer now gets -ENAMETOOLONG instead of -EINVAL. - A path that is empty or lacks the leading '/' is now rejected after the copy rather than before it, so such a call makes an allocation and returns -ENOMEM instead of -EINVAL if that allocation fails. bpf_binprm_set_interp_arg() still rejects an empty argument before allocating, so its results are unchanged apart from the raced case fixed here. Both new checks run before the previously staged string is freed or replaced. A failing call frees only its own allocation and leaves the earlier selection in place, as the -ENOMEM path already does. Fixes: b4bfe2f6b011 ("binfmt_misc: add binfmt_misc_ops bpf struct_ops") Signed-off-by: Chris Mason Link: https://patch.msgid.link/20260918-work-binfmt_misc-fixes-v1-2-647b24bc1c46@kernel.org Signed-off-by: Christian Brauner (Amutable) --- fs/binfmt_misc_bpf.c | 20 ++++++++++++++++++-- 1 file changed, 18 insertions(+), 2 deletions(-) diff --git a/fs/binfmt_misc_bpf.c b/fs/binfmt_misc_bpf.c index ce1bc78e8511..a3e26e8a4027 100644 --- a/fs/binfmt_misc_bpf.c +++ b/fs/binfmt_misc_bpf.c @@ -141,8 +141,6 @@ __bpf_kfunc int bpf_binprm_set_interp(struct linux_binprm *bprm, len = strnlen(path, path__sz); if (len == path__sz) return -EINVAL; - if (path[0] != '/') - return -EINVAL; if (len >= PATH_MAX) return -ENAMETOOLONG; @@ -150,6 +148,15 @@ __bpf_kfunc int bpf_binprm_set_interp(struct linux_binprm *bprm, if (!interp) return -ENOMEM; + /* + * The program may pass memory that is written to while this runs, + * so check the private copy and not the buffer it was made from. + */ + if (interp[0] != '/') { + kfree(interp); + return -EINVAL; + } + bm_bpf_stage_selection(bprm, interp, NULL); return 0; } @@ -241,6 +248,15 @@ __bpf_kfunc int bpf_binprm_set_interp_arg(struct linux_binprm *bprm, if (!val) return -ENOMEM; + /* + * The program may pass memory that is written to while this runs, + * so check the private copy and not the buffer it was made from. + */ + if (!val[0]) { + kfree(val); + return -EINVAL; + } + kfree(bprm->bpf_interp_arg); bprm->bpf_interp_arg = val; return 0; From 70194dc37670bd08e44b471389861cc01bd3a3c9 Mon Sep 17 00:00:00 2001 From: Aohan Mei Date: Mon, 14 Sep 2026 19:51:47 +0800 Subject: [PATCH 189/694] netfilter: nf_tables: skip expired catchall elements on insert and delete nft_setelem_catchall_insert() looks up duplicates with nft_set_elem_active() only, while nft_set_catchall_lookup() and the dump path additionally skip expired elements. Once a catchall element with a timeout expires, this predicate drift makes it invisible to userspace dumps, yet it still blocks re-insertion: with NLM_F_EXCL the request fails with -EEXIST, and without it the request reports success but silently inserts nothing. The stale entry only goes away when the (user-tunable) gc interval elapses, so the catchall rule may silently stop matching for an arbitrarily long time after its first expiration. The delete path shows the same drift: nft_setelem_catchall_deactivate() picks the first active-next entry in the catchall list, so with an expired entry still pending GC it retires the stale entry instead of the fresh one, and it deactivates an element that userspace no longer sees instead of failing with -ENOENT. Align both walks with the lookup and dump predicates: only an element that is active and not expired counts as a duplicate or delete candidate, using the per-netns timestamp taken at transaction start, in line with the set backend .insert/.deactivate and catchall GC sync paths. Reported-by: TencentOS Corvus AI Cc: stable@vger.kernel.org Fixes: aaa31047a6d2 ("netfilter: nftables: add catch-all set element support") Assisted-by: CodeBuddy:Kimi-K3 Signed-off-by: Aohan Mei Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nf_tables_api.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c index c0b754a2d45b..b59628e6240c 100644 --- a/net/netfilter/nf_tables_api.c +++ b/net/netfilter/nf_tables_api.c @@ -6995,11 +6995,14 @@ static int nft_setelem_catchall_insert(const struct net *net, { struct nft_set_elem_catchall *catchall; u8 genmask = nft_genmask_next(net); + u64 tstamp = nft_net_tstamp(net); struct nft_set_ext *ext; list_for_each_entry(catchall, &set->catchall_list, list) { ext = nft_set_elem_ext(set, catchall->elem); - if (nft_set_elem_active(ext, genmask)) { + if (nft_set_elem_active(ext, genmask) && + !__nft_set_elem_expired(ext, tstamp) && + !nft_set_elem_is_dead(ext)) { *priv = catchall->elem; return -EEXIST; } @@ -7092,11 +7095,14 @@ static int nft_setelem_catchall_deactivate(const struct net *net, struct nft_set_elem *elem) { struct nft_set_elem_catchall *catchall; + u64 tstamp = nft_net_tstamp(net); struct nft_set_ext *ext; list_for_each_entry(catchall, &set->catchall_list, list) { ext = nft_set_elem_ext(set, catchall->elem); - if (!nft_is_active_next(net, ext)) + if (!nft_is_active_next(net, ext) || + __nft_set_elem_expired(ext, tstamp) || + nft_set_elem_is_dead(ext)) continue; kfree(elem->priv); From 406aa2b186d3f13a35bc1ad6aff4274917851bc4 Mon Sep 17 00:00:00 2001 From: Robin Murphy Date: Tue, 15 Sep 2026 13:24:22 +0100 Subject: [PATCH 190/694] perf/arm-cmn: Fix multi-filter encoding The current special-case for EVICT_STATE_SEL filtering effectively assigns the "filter" and "filter2" controls in the opposite order from how the CMN S3 r2 TRM states "Filtering is programmed in pmu_hbt_lbt_sel and pmu_evict_state_sel". On reflection, not only does this seem unnecessarily non-obvious to users, but it's also likely to be a problem for scaling to a full multi-filter abstraction in future. There is a logical order to filters based on their bitfield positions in the pmu_event_sel register, which the TRM descriptions allude to, and the cmn_filter_select enum already (almost) follows, so let's fix the UABI to follow suit while it's still unreleased. Fixes: 09178f536bb9 ("perf/arm-cmn: Plumb in new filter types") Signed-off-by: Robin Murphy Reviewed-by: Ilkka Koskinen Reviewed-by: Leo Yan Signed-off-by: Will Deacon --- drivers/perf/arm-cmn.c | 58 +++++++++++++++++++++++++----------------- 1 file changed, 34 insertions(+), 24 deletions(-) diff --git a/drivers/perf/arm-cmn.c b/drivers/perf/arm-cmn.c index 33ee2be9b386..5378fba916cf 100644 --- a/drivers/perf/arm-cmn.c +++ b/drivers/perf/arm-cmn.c @@ -305,9 +305,9 @@ enum cmn_filter_select { SEL_NONE, SEL_OCCUP1_ID, SEL_CLASS_OCCUP_ID, - SEL_CBUSY_SNTHROTTLE_SEL, SEL_HBT_LBT_SEL, SEL_SN_HOME_SEL, + SEL_CBUSY_SNTHROTTLE_SEL, SEL_SNP_VC_SEL, SEL_ENHANCED_HBT_LBT_SEL, SEL_EVICT_STATE_SEL, @@ -978,10 +978,14 @@ static umode_t arm_cmn_event_attr_is_visible(struct kobject *kobj, _CMN_EVENT_HNS(_model, _name##_all, _event, _sel, 0), \ _CMN_EVENT_HNS(_model, _name##_hbt, _event, _sel, 1), \ _CMN_EVENT_HNS(_model, _name##_lbt, _event, _sel, 2) -#define _CMN_EVENT_HNS_HBT2(_model, _name, _event, _fsel1, f1) \ - _CMN_EVENT_HNS(_model, _name##_all, _event, _fsel1, f1, SEL_HBT_LBT_SEL, 0), \ - _CMN_EVENT_HNS(_model, _name##_hbt, _event, _fsel1, f1, SEL_HBT_LBT_SEL, 1), \ - _CMN_EVENT_HNS(_model, _name##_lbt, _event, _fsel1, f1, SEL_HBT_LBT_SEL, 2) +#define _CMN_EVENT_HNS_EVICT(_model, _name, _event, _fsel1, f1) \ + _CMN_EVENT_HNS(_model, _name##_all, _event, _fsel1, f1, SEL_EVICT_STATE_SEL, 0), \ + _CMN_EVENT_HNS(_model, _name##_eu, _event, _fsel1, f1, SEL_EVICT_STATE_SEL, 1), \ + _CMN_EVENT_HNS(_model, _name##_en, _event, _fsel1, f1, SEL_EVICT_STATE_SEL, 2), \ + _CMN_EVENT_HNS(_model, _name##_su, _event, _fsel1, f1, SEL_EVICT_STATE_SEL, 3), \ + _CMN_EVENT_HNS(_model, _name##_sn, _event, _fsel1, f1, SEL_EVICT_STATE_SEL, 4), \ + _CMN_EVENT_HNS(_model, _name##_mu, _event, _fsel1, f1, SEL_EVICT_STATE_SEL, 5), \ + _CMN_EVENT_HNS(_model, _name##_mn, _event, _fsel1, f1, SEL_EVICT_STATE_SEL, 6) #define CMN_EVENT_HNS_OCC(_model, _name, _event) \ CMN_EVENT_HN_OCC(_model, hns_##_name, CMN_TYPE_HNS, _event), \ @@ -1014,13 +1018,9 @@ static umode_t arm_cmn_event_attr_is_visible(struct kobject *kobj, _CMN_EVENT_HNS(CMNS3R2, _name##_ccglcn, _event, SEL_ENHANCED_HBT_LBT_SEL, 5), \ _CMN_EVENT_HNS(CMNS3R2, _name##_ccgrn, _event, SEL_ENHANCED_HBT_LBT_SEL, 6) #define CMN_EVENT_HNS_EVICT(_model, _name, _event) \ - _CMN_EVENT_HNS_HBT2(_model, _name##_all, _event, SEL_EVICT_STATE_SEL, 0), \ - _CMN_EVENT_HNS_HBT2(_model, _name##_eu, _event, SEL_EVICT_STATE_SEL, 1), \ - _CMN_EVENT_HNS_HBT2(_model, _name##_en, _event, SEL_EVICT_STATE_SEL, 2), \ - _CMN_EVENT_HNS_HBT2(_model, _name##_su, _event, SEL_EVICT_STATE_SEL, 3), \ - _CMN_EVENT_HNS_HBT2(_model, _name##_sn, _event, SEL_EVICT_STATE_SEL, 4), \ - _CMN_EVENT_HNS_HBT2(_model, _name##_mu, _event, SEL_EVICT_STATE_SEL, 5), \ - _CMN_EVENT_HNS_HBT2(_model, _name##_mn, _event, SEL_EVICT_STATE_SEL, 6) + _CMN_EVENT_HNS_EVICT(_model, _name##_all, _event, SEL_HBT_LBT_SEL, 0), \ + _CMN_EVENT_HNS_EVICT(_model, _name##_hbt, _event, SEL_HBT_LBT_SEL, 1), \ + _CMN_EVENT_HNS_EVICT(_model, _name##_lbt, _event, SEL_HBT_LBT_SEL, 2) #define CMN_EVENT_HNSR0_HBT(_name, _event) \ _CMN_EVENT_HNS_HBT(CMN700 | CMNS3R01, _name, _event, SEL_HBT_LBT_SEL) @@ -1860,10 +1860,12 @@ static void arm_cmn_val_add_event(struct arm_cmn *cmn, struct arm_cmn_val *val, val->dtm_count[dtm]++; + if (sel == SEL_EVICT_STATE_SEL) { + val->filter[dtm][sel] = CMN_EVENT_FILTER2(event) + 1; + sel = SEL_HBT_LBT_SEL; + } if (sel) val->filter[dtm][sel] = CMN_EVENT_FILTER(event) + 1; - if (sel == SEL_EVICT_STATE_SEL) - val->filter[dtm][SEL_HBT_LBT_SEL] = CMN_EVENT_FILTER2(event) + 1; if (type != CMN_TYPE_WP) continue; @@ -1914,14 +1916,17 @@ static int arm_cmn_validate_group(struct arm_cmn *cmn, struct perf_event *event) if (val->dtm_count[dtm] == CMN_DTM_NUM_COUNTERS) goto done; + if (sel == SEL_EVICT_STATE_SEL) { + if (val->filter[dtm][sel] && + val->filter[dtm][sel] != CMN_EVENT_FILTER2(event) + 1) + goto done; + sel = SEL_HBT_LBT_SEL; + } + if (sel && val->filter[dtm][sel] && val->filter[dtm][sel] != CMN_EVENT_FILTER(event) + 1) goto done; - if (sel == SEL_EVICT_STATE_SEL && val->filter[dtm][SEL_HBT_LBT_SEL] && - val->filter[dtm][SEL_HBT_LBT_SEL] != CMN_EVENT_FILTER2(event) + 1) - goto done; - if (type != CMN_TYPE_WP) continue; @@ -1950,7 +1955,7 @@ static enum cmn_filter_select arm_cmn_event_filter(const struct arm_cmn *cmn, for (int i = 0; i < ARRAY_SIZE(arm_cmn_event_attrs) - 1; i++) { e = container_of(arm_cmn_event_attrs[i], typeof(*e), attr.attr); if (e->model & model && e->type == type && e->eventid == eventid) - return e->filter[0].sel; + return e->filter[1].sel ?: e->filter[0].sel; } return SEL_NONE; } @@ -2081,16 +2086,21 @@ static void arm_cmn_event_clear(struct arm_cmn *cmn, struct perf_event *event, static int arm_cmn_set_event_filter(struct arm_cmn_node *dn, struct perf_event *event) { enum cmn_filter_select fsel = to_cmn_hw(event)->filter_sel; + bool evict_state = fsel == SEL_EVICT_STATE_SEL; int ret = 0; + if (evict_state) { + ret = arm_cmn_set_event_sel_hi(dn, fsel, CMN_EVENT_FILTER2(event)); + if (ret) + return ret; + fsel = SEL_HBT_LBT_SEL; + } if (fsel) ret = arm_cmn_set_event_sel_hi(dn, fsel, CMN_EVENT_FILTER(event)); - if (fsel == SEL_EVICT_STATE_SEL && !ret) { - ret = arm_cmn_set_event_sel_hi(dn, SEL_HBT_LBT_SEL, CMN_EVENT_FILTER2(event)); - if (ret) - dn->filter[fsel].count--; - } + if (ret && evict_state) + dn->filter[SEL_EVICT_STATE_SEL].count--; + return ret; } From bb756b11ad63832ebee58caf9e8f9381eaecff9f Mon Sep 17 00:00:00 2001 From: Zeng Heng Date: Fri, 11 Sep 2026 09:58:59 +0800 Subject: [PATCH 191/694] arm64: io: Reject non-user protection in ioremap_prot() Mapping a stack-top page via /dev/mem with PROT_NONE and then reading that process's /proc//cmdline triggers a spurious WARN in ioremap_prot() through generic_access_phys(): WARNING: ./arch/arm64/include/asm/io.h:275 at generic_access_phys Call trace: generic_access_phys+0x1c8/0x228 (P) __access_remote_vm+0x2b4/0x398 access_remote_vm+0x14/0x30 get_mm_cmdline+0xf8/0x2a0 proc_pid_cmdline_read+0x68/0x120 generic_access_phys() passes the protection derived from the user PTE to ioremap_prot(). On arm64, a PROT_NONE mapping is represented by a present-invalid PTE, so pte_present() still returns true and the protection reaches ioremap_prot(). A PROT_NONE mapping does not have PTE_USER, causing the existing WARN_ON_ONCE() in ioremap_prot() to fire even though this is a valid user mapping. Execute-only mappings have the same issue and must not be readable through this path either. ioremap_prot() should therefore reject protection values without PTE_USER without warning. This makes the access fail cleanly for PROT_NONE and execute-only mappings while retaining the existing user-protection contract. Fixes: 8f098037139b ("arm64: io: Extract user memory type in ioremap_prot()") Signed-off-by: Zeng Heng Reviewed-by: Catalin Marinas Signed-off-by: Will Deacon --- arch/arm64/include/asm/io.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/arch/arm64/include/asm/io.h b/arch/arm64/include/asm/io.h index 49a7002661a9..9aa0bb08ab60 100644 --- a/arch/arm64/include/asm/io.h +++ b/arch/arm64/include/asm/io.h @@ -280,7 +280,8 @@ static inline void __iomem *ioremap_prot(phys_addr_t phys, size_t size, pgprot_t prot; ptval_t user_prot_val = pgprot_val(user_prot); - if (WARN_ON_ONCE(!(user_prot_val & PTE_USER))) + /* Reject PROT_NONE and exec-only */ + if (!(user_prot_val & PTE_USER)) return NULL; prot = __pgprot_modify(PAGE_KERNEL, PTE_ATTRINDX_MASK, From baaa9b126b875b40ffd9356da52c7c871917e5bb Mon Sep 17 00:00:00 2001 From: Yureka Lilian Date: Fri, 11 Sep 2026 19:02:10 +0200 Subject: [PATCH 192/694] arm64: Add override for WFxT Add an override for WFxT support within ID_AA64ISAR2_EL1 to allow it to be disabled using the new arm64.nowfxt command line parameter. This accompanies the idle=nop param introduced in a previous patch series [1] in dealing with misbehaving WFI and WFIT instructions on Apple Silicon SoCs, and eases debugging of other quirky WFxT implementations. Link[1]: https://lore.kernel.org/all/20260804-arm64-idle-param-v3-1-d10f8159062b@cyberchaos.dev/ Suggested-by: Will Deacon Signed-off-by: Yureka Lilian Signed-off-by: Will Deacon --- Documentation/admin-guide/kernel-parameters.txt | 3 +++ arch/arm64/kernel/pi/idreg-override.c | 2 ++ 2 files changed, 5 insertions(+) diff --git a/Documentation/admin-guide/kernel-parameters.txt b/Documentation/admin-guide/kernel-parameters.txt index 68647ff4bdd2..0fe9e36ce61a 100644 --- a/Documentation/admin-guide/kernel-parameters.txt +++ b/Documentation/admin-guide/kernel-parameters.txt @@ -589,6 +589,9 @@ Kernel parameters arm64.nosve [ARM64] Unconditionally disable Scalable Vector Extension support + arm64.nowfxt [ARM64] Unconditionally disable Wait For Event with + Timeout and Wait For Interrupt with Timeout support + ataflop= [HW,M68k] atarimouse= [HW,MOUSE] Atari Mouse diff --git a/arch/arm64/kernel/pi/idreg-override.c b/arch/arm64/kernel/pi/idreg-override.c index 274bc72ba98d..acaf56234973 100644 --- a/arch/arm64/kernel/pi/idreg-override.c +++ b/arch/arm64/kernel/pi/idreg-override.c @@ -186,6 +186,7 @@ static const struct ftr_set_desc isar2 __prel64_initconst = { .name = "id_aa64isar2", .override = &id_aa64isar2_override, .fields = { + FIELD("wfxt", ID_AA64ISAR2_EL1_WFxT_SHIFT, NULL), FIELD("gpa3", ID_AA64ISAR2_EL1_GPA3_SHIFT, NULL), FIELD("apa3", ID_AA64ISAR2_EL1_APA3_SHIFT, NULL), FIELD("mops", ID_AA64ISAR2_EL1_MOPS_SHIFT, NULL), @@ -254,6 +255,7 @@ static const struct { "id_aa64isar1.api=0 id_aa64isar1.apa=0 " "id_aa64isar2.gpa3=0 id_aa64isar2.apa3=0" }, { "arm64.nomops", "id_aa64isar2.mops=0" }, + { "arm64.nowfxt", "id_aa64isar2.wfxt=0" }, { "arm64.nomte", "id_aa64pfr1.mte=0" }, { "nokaslr", "arm64_sw.nokaslr=1" }, { "rodata=off", "arm64_sw.rodataoff=1" }, From b7403afb7a5f85073243df10238b3483958ad69e Mon Sep 17 00:00:00 2001 From: David Carlier Date: Sun, 6 Sep 2026 13:14:16 +0100 Subject: [PATCH 193/694] arm64: errata: match the target implementation CPU's own MIDR __is_affected_midr_range() is handed the MIDR and REVIDR of one target implementation CPU, but tests the erratum's range with is_midr_in_range(), which re-scans all of target_impl_cpus[] and ignores the @midr argument. The range test is thus constant across the per-CPU loop in is_affected_midr_range() and only answers "is any target CPU in range". Since just the fixed_revs REVIDR check uses the iteration's own registers, an out-of-range target CPU can decide whether a MIDR_FIXED() exemption applies. A VM then enables a workaround whose only in-range CPU is fixed silicon, e.g. erratum 2658417 on a Cortex-A510 r1p1 with REVIDR_EL1[25] set. Factor the range test into __is_midr_in_range(), which takes an explicit MIDR, and use it in __is_affected_midr_range(). Fixes: 86edf6bdcf05 ("smccc/kvm_guest: Enable errata based on implementation CPUs") Cc: stable@vger.kernel.org Assisted-by: Claude:claude-opus-5 Signed-off-by: David Carlier Signed-off-by: Will Deacon --- arch/arm64/kernel/cpu_errata.c | 15 +++++++++------ 1 file changed, 9 insertions(+), 6 deletions(-) diff --git a/arch/arm64/kernel/cpu_errata.c b/arch/arm64/kernel/cpu_errata.c index b33dccfafaf8..8ec47d89b45b 100644 --- a/arch/arm64/kernel/cpu_errata.c +++ b/arch/arm64/kernel/cpu_errata.c @@ -28,18 +28,21 @@ bool cpu_errata_set_target_impl(u64 num, void *impl_cpus) return true; } +static inline bool __is_midr_in_range(u32 midr, struct midr_range const *range) +{ + return midr_is_cpu_model_range(midr, range->model, + range->rv_min, range->rv_max); +} + static inline bool is_midr_in_range(struct midr_range const *range) { int i; if (!target_impl_cpu_num) - return midr_is_cpu_model_range(read_cpuid_id(), range->model, - range->rv_min, range->rv_max); + return __is_midr_in_range(read_cpuid_id(), range); for (i = 0; i < target_impl_cpu_num; i++) { - if (midr_is_cpu_model_range(target_impl_cpus[i].midr, - range->model, - range->rv_min, range->rv_max)) + if (__is_midr_in_range(target_impl_cpus[i].midr, range)) return true; } return false; @@ -59,7 +62,7 @@ __is_affected_midr_range(const struct arm64_cpu_capabilities *entry, u32 midr, u32 revidr) { const struct arm64_midr_revidr *fix; - if (!is_midr_in_range(&entry->midr_range)) + if (!__is_midr_in_range(midr, &entry->midr_range)) return false; midr &= MIDR_REVISION_MASK | MIDR_VARIANT_MASK; From 4485a01f4df1c9683d8ffe3e4ade6c33c9572d3c Mon Sep 17 00:00:00 2001 From: Sean Young Date: Sun, 13 Sep 2026 15:12:27 +0100 Subject: [PATCH 194/694] parisc: unwind: Replace open-coded binary search with bsearch() There is a bug in the binary search where hi can underflow. If addr is less than the first entry, then "hi = mid - 1" will underflow to ULONG_MAX. Then we have an out-of-bounds read. Replace the open-coded binary search with bsearch(). Issue found by an LLM. Signed-off-by: Sean Young Signed-off-by: Helge Deller --- arch/parisc/kernel/unwind.c | 33 +++++++++++++++------------------ 1 file changed, 15 insertions(+), 18 deletions(-) diff --git a/arch/parisc/kernel/unwind.c b/arch/parisc/kernel/unwind.c index 32103a270a8e..fab9ae22191a 100644 --- a/arch/parisc/kernel/unwind.c +++ b/arch/parisc/kernel/unwind.c @@ -14,6 +14,7 @@ #include #include #include +#include #include #include @@ -49,27 +50,23 @@ static DEFINE_SPINLOCK(unwind_lock); static struct unwind_table kernel_unwind_table __ro_after_init; static LIST_HEAD(unwind_tables); +static int cmp_unwind_entry(const void *key, const void *elt) +{ + unsigned long addr = (unsigned long)key; + const struct unwind_table_entry *e = elt; + + if (addr < e->region_start) + return -1; + if (addr > e->region_end) + return 1; + return 0; +} + static inline const struct unwind_table_entry * find_unwind_entry_in_table(const struct unwind_table *table, unsigned long addr) { - const struct unwind_table_entry *e = NULL; - unsigned long lo, hi, mid; - - lo = 0; - hi = table->length - 1; - - while (lo <= hi) { - mid = (hi - lo) / 2 + lo; - e = &table->table[mid]; - if (addr < e->region_start) - hi = mid - 1; - else if (addr > e->region_end) - lo = mid + 1; - else - return e; - } - - return NULL; + return bsearch((void *)addr, table->table, table->length, + sizeof(*table->table), cmp_unwind_entry); } static const struct unwind_table_entry * From cb917b1e1c23f6f9b73802cd576b48bd606458c0 Mon Sep 17 00:00:00 2001 From: Ethan Nelson-Moore Date: Thu, 17 Sep 2026 20:04:14 -0700 Subject: [PATCH 195/694] parisc: remove unused header The last use of arch/parisc/include/asm/compat_ucontext.h was removed in commit d313d4e72df3 ("parisc: remove unused compat_rt_sigframe.h header") but it was left behind. Remove it. Signed-off-by: Ethan Nelson-Moore Signed-off-by: Helge Deller --- arch/parisc/include/asm/compat_ucontext.h | 18 ------------------ 1 file changed, 18 deletions(-) delete mode 100644 arch/parisc/include/asm/compat_ucontext.h diff --git a/arch/parisc/include/asm/compat_ucontext.h b/arch/parisc/include/asm/compat_ucontext.h deleted file mode 100644 index c606f1bc891d..000000000000 --- a/arch/parisc/include/asm/compat_ucontext.h +++ /dev/null @@ -1,18 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef _ASM_PARISC_COMPAT_UCONTEXT_H -#define _ASM_PARISC_COMPAT_UCONTEXT_H - -#include - -/* 32-bit ucontext as seen from an 64-bit kernel */ -struct compat_ucontext { - compat_uint_t uc_flags; - compat_uptr_t uc_link; - compat_stack_t uc_stack; /* struct compat_sigaltstack (12 bytes)*/ - /* FIXME: Pad out to get uc_mcontext to start at an 8-byte aligned boundary */ - compat_uint_t pad[1]; - struct compat_sigcontext uc_mcontext; - compat_sigset_t uc_sigmask; /* mask last for extensibility */ -}; - -#endif /* !_ASM_PARISC_COMPAT_UCONTEXT_H */ From 289e99e7a263c6fd6a5d07d6d8f2156b70f3a9d5 Mon Sep 17 00:00:00 2001 From: Zhenghui Hao Date: Fri, 18 Sep 2026 13:46:49 +0800 Subject: [PATCH 196/694] parisc: parse early parameters in setup_arch() parisc is one of the few architectures that does not call parse_early_param() from setup_arch(). That was mostly harmless until commit d49004c5f0c1 ("arch, mm: consolidate initialization of nodes, zones and memory map") moved the consumer of several hugetlb command line parameters into mm_core_init_early(), which runs before the generic parse_early_param() call in start_kernel(). As a result hugepages=, hugepagesz=, default_hugepagesz=, hugetlb_cma= and hugetlb_free_vmemmap= are recorded after they have already been consumed and are silently dropped on parisc. Call parse_early_param() from setup_arch(), after the command line has been set up and the memory inventory has been taken. jump_label_init() must be called first because early parameter handlers may enable or disable static keys. Both functions are safe to call more than once: the generic calls in start_kernel() remain in place and turn into no-ops. Suggested-by: Mike Rapoport (Microsoft) Fixes: d49004c5f0c1 ("arch, mm: consolidate initialization of nodes, zones and memory map") Cc: Signed-off-by: Zhenghui Hao Tested-by: Helge Deller Signed-off-by: Helge Deller --- arch/parisc/kernel/setup.c | 12 ++++++++++++ 1 file changed, 12 insertions(+) diff --git a/arch/parisc/kernel/setup.c b/arch/parisc/kernel/setup.c index d3e17a7a8901..4d3015a411d6 100644 --- a/arch/parisc/kernel/setup.c +++ b/arch/parisc/kernel/setup.c @@ -132,6 +132,18 @@ void __init setup_arch(char **cmdline_p) parisc_cache_init(); paging_init(); + /* + * Parse early parameters before mm_core_init_early() runs. + * Several early_param() handlers only record data that is consumed + * from there - for example hugepages=, hugepagesz=, + * default_hugepagesz=, hugetlb_cma= and hugetlb_free_vmemmap= - so + * the generic parse_early_param() call in start_kernel() is too late + * for them. jump_label_init() must come first, since early param + * handlers may enable or disable static keys. + */ + jump_label_init(); + parse_early_param(); + #ifdef CONFIG_PA11 dma_ops_init(); #endif From b0be01f133aa36d2fd12d71c916cb8a47826b4ed Mon Sep 17 00:00:00 2001 From: Shawn Guo Date: Sun, 30 Aug 2026 11:02:01 +0800 Subject: [PATCH 197/694] pinctrl: qcom: nord: Split QUP1 SE2/SE3 into lane-pair functions QUP1 SE2 and SE3 pack all four of their lanes pair-wise onto only two pins each: lanes 0/1 (I2C SDA/SCL) at mux value 2 and lanes 2/3 (UART TX/RX) at mux value 1, on gpio127/gpio128 and gpio129/gpio130 respectively. Both mux values were named "qup1_se2" (respectively "qup1_se3"), so the two distinct lane pairs became indistinguishable. msm_pinmux_set_mux() stops at the first entry matching the requested function, which means mux value 1 was always selected and the I2C lanes could never be muxed out. In practice i2c9 and i2c10 got the UART lanes and did not work, while uart9 and uart10 happened to be muxed correctly. Give each lane pair its own function, following the _01/_23 naming already used for the same hardware arrangement by the shikra, eliza, hawi and maili TLMM drivers. Both functions still cover the full pin pair, so a single pinctrl state per protocol remains sufficient. Drop gpio129/gpio130 from the SE2 group list, since those pins belong to SE3 and were never reachable through the SE2 function. Also rename QUP1 SE2/SE3 functions in the binding doc accordingly. While at it, add missing "gpio", "qup3_se0_mira" and "qup3_se0_mirb" to the binding function enum to get the list complete. Fixes: c24dd0826f06 ("pinctrl: qcom: add the TLMM driver for the Nord platforms") Cc: stable@vger.kernel.org Assisted-by: LLM Signed-off-by: Shawn Guo Reviewed-by: Konrad Dybcio Link: https://patch.msgid.link/20260830030201.135637-1-shengchao.guo@oss.qualcomm.com Signed-off-by: Bartosz Golaszewski Signed-off-by: Linus Walleij --- .../bindings/pinctrl/qcom,nord-tlmm.yaml | 7 ++-- drivers/pinctrl/qcom/pinctrl-nord.c | 34 +++++++++++++------ 2 files changed, 27 insertions(+), 14 deletions(-) diff --git a/Documentation/devicetree/bindings/pinctrl/qcom,nord-tlmm.yaml b/Documentation/devicetree/bindings/pinctrl/qcom,nord-tlmm.yaml index 4bb511719f31..56758a85ead8 100644 --- a/Documentation/devicetree/bindings/pinctrl/qcom,nord-tlmm.yaml +++ b/Documentation/devicetree/bindings/pinctrl/qcom,nord-tlmm.yaml @@ -67,7 +67,7 @@ $defs: Specify the alternative function to be configured for the specified pins. - enum: [ aoss_cti, atest_char, atest_usb20, atest_usb21, + enum: [ gpio, aoss_cti, atest_char, atest_usb20, atest_usb21, aud_intfc0_clk, aud_intfc0_data, aud_intfc0_ws, aud_intfc10_clk, aud_intfc10_data, aud_intfc10_ws, aud_intfc1_clk, aud_intfc1_data, aud_intfc1_ws, @@ -98,9 +98,10 @@ $defs: pcie3_clk_req_n, phase_flag, pll_bist_sync, pll_clk_aux, prng_rosc0, prng_rosc1, pwrbrk_i_n, qdss, qdss_cti, qspi, qup0_se0, qup0_se1, qup0_se2, qup0_se3, qup0_se4, qup0_se5, - qup1_se0, qup1_se1, qup1_se3, qup1_se2, qup1_se4, qup1_se5, + qup1_se0, qup1_se1, qup1_se2_01, qup1_se2_23, qup1_se3_01, + qup1_se3_23, qup1_se4, qup1_se5, qup1_se6, qup2_se0, qup2_se1, qup2_se2, qup2_se3, qup2_se4, - qup2_se5, qup2_se6, + qup2_se5, qup2_se6, qup3_se0_mira, qup3_se0_mirb, sailss_ospi, sdc4_clk, sdc4_cmd, sdc4_data, smb_alert, smb_alert_n, smb_clk, smb_dat, tb_trig_sdc4, tmess_prng0, tmess_prng1, tsc_timer, tsense_pwm, usb0_hs, diff --git a/drivers/pinctrl/qcom/pinctrl-nord.c b/drivers/pinctrl/qcom/pinctrl-nord.c index 7c21306e77ff..7f37f8e819ba 100644 --- a/drivers/pinctrl/qcom/pinctrl-nord.c +++ b/drivers/pinctrl/qcom/pinctrl-nord.c @@ -570,8 +570,10 @@ enum nord_functions { msm_mux_qup0_se5, msm_mux_qup1_se0, msm_mux_qup1_se1, - msm_mux_qup1_se2, - msm_mux_qup1_se3, + msm_mux_qup1_se2_01, + msm_mux_qup1_se2_23, + msm_mux_qup1_se3_01, + msm_mux_qup1_se3_23, msm_mux_qup1_se4, msm_mux_qup1_se5, msm_mux_qup1_se6, @@ -1152,11 +1154,19 @@ static const char *const qup1_se1_groups[] = { "gpio123", "gpio124", "gpio125", "gpio126", }; -static const char *const qup1_se2_groups[] = { - "gpio127", "gpio128", "gpio129", "gpio130", +static const char *const qup1_se2_01_groups[] = { + "gpio127", "gpio128", }; -static const char *const qup1_se3_groups[] = { +static const char *const qup1_se2_23_groups[] = { + "gpio127", "gpio128", +}; + +static const char *const qup1_se3_01_groups[] = { + "gpio129", "gpio130", +}; + +static const char *const qup1_se3_23_groups[] = { "gpio129", "gpio130", }; @@ -1428,8 +1438,10 @@ static const struct pinfunction nord_functions[] = { MSM_PIN_FUNCTION(qup0_se5), MSM_PIN_FUNCTION(qup1_se0), MSM_PIN_FUNCTION(qup1_se1), - MSM_PIN_FUNCTION(qup1_se2), - MSM_PIN_FUNCTION(qup1_se3), + MSM_PIN_FUNCTION(qup1_se2_01), + MSM_PIN_FUNCTION(qup1_se2_23), + MSM_PIN_FUNCTION(qup1_se3_01), + MSM_PIN_FUNCTION(qup1_se3_23), MSM_PIN_FUNCTION(qup1_se4), MSM_PIN_FUNCTION(qup1_se5), MSM_PIN_FUNCTION(qup1_se6), @@ -1633,13 +1645,13 @@ static const struct msm_pingroup nord_groups[] = { _, _, _, _, _, _, _), [126] = PINGROUP(126, qup1_se1, qup1_se0, ccu_i2c_scl, mdp1_vsync_out, _, atest_usb20, ddr_pxi, _, _, _, _), - [127] = PINGROUP(127, qup1_se2, qup1_se2, _, atest_usb21, ddr_pxi, + [127] = PINGROUP(127, qup1_se2_23, qup1_se2_01, _, atest_usb21, ddr_pxi, _, _, _, _, _, _), - [128] = PINGROUP(128, qup1_se2, qup1_se2, _, atest_usb20, ddr_pxi, + [128] = PINGROUP(128, qup1_se2_23, qup1_se2_01, _, atest_usb20, ddr_pxi, _, _, _, _, _, _), - [129] = PINGROUP(129, qup1_se3, qup1_se3, ccu_i2c_sda, mdp1_vsync_out, + [129] = PINGROUP(129, qup1_se3_23, qup1_se3_01, ccu_i2c_sda, mdp1_vsync_out, _, atest_usb21, ddr_pxi, _, _, _, _), - [130] = PINGROUP(130, qup1_se3, qup1_se3, ccu_i2c_scl, mdp1_vsync_out, + [130] = PINGROUP(130, qup1_se3_23, qup1_se3_01, ccu_i2c_scl, mdp1_vsync_out, _, atest_usb20, ddr_pxi, _, _, _, _), [131] = PINGROUP(131, qup1_se4, qup1_se6, ccu_i2c_sda, mdp1_vsync_out, _, atest_usb21, ddr_pxi, _, _, _, _), From 447dcff90557748a5ac384aaf5d70b2c7e3b961d Mon Sep 17 00:00:00 2001 From: "hpp.iscas" Date: Sat, 5 Sep 2026 21:43:40 +0800 Subject: [PATCH 198/694] pinctrl: qcom: ipq5210: Publish the OF module alias The IPQ5210 TLMM platform driver can be a module and matches through ipq5210_tlmm_of_match. This table is not published for OF modalias matching. Publish the existing table, preserving arch_initcall ordering and the shared MSM pinctrl probe. Fixes: a549fe22376f ("pinctrl: qcom: Introduce IPQ5210 TLMM driver") Signed-off-by: hpp.iscas Reviewed-by: Konrad Dybcio Link: https://patch.msgid.link/20260905134340.67477-1-hppiscas@163.com Signed-off-by: Bartosz Golaszewski Signed-off-by: Linus Walleij --- drivers/pinctrl/qcom/pinctrl-ipq5210.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/pinctrl/qcom/pinctrl-ipq5210.c b/drivers/pinctrl/qcom/pinctrl-ipq5210.c index 827a4ad07a6b..d5b4eb3e7343 100644 --- a/drivers/pinctrl/qcom/pinctrl-ipq5210.c +++ b/drivers/pinctrl/qcom/pinctrl-ipq5210.c @@ -867,6 +867,7 @@ static const struct of_device_id ipq5210_tlmm_of_match[] = { { .compatible = "qcom,ipq5210-tlmm", }, { }, }; +MODULE_DEVICE_TABLE(of, ipq5210_tlmm_of_match); static int ipq5210_tlmm_probe(struct platform_device *pdev) { From aff09d9e37e02dc60bde79035ac15b136d602259 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Fri, 18 Sep 2026 15:16:17 +0200 Subject: [PATCH 199/694] drm/nouveau/clk: fix list cursor use after loop in nvkm_clk_ustate_update If list_for_each_entry() exits without hitting a break then "pstate" is not a valid pstate pointer. Introduce a "found" variable instead. The check is reachable from userspace: nvkm_clk_ustate_update() takes the pstate id straight from the 'pstate' debugfs file, so requesting an id that is not in clk->states - or any id at all when the perf tables are broken and the list is empty - makes the pstate->pstate != req test dereference the list head cast to a struct nvkm_pstate, which is an out-of-bounds read. Fixes: 7c8565220697 ("drm/nouveau/clk: implement power state and engine clock control in core") Signed-off-by: Dan Carpenter [Francesco: rebased on drm-misc-next, expanded the commit message] Signed-off-by: Francesco Magazzu Reviewed-by: Lyude Paul Signed-off-by: Lyude Paul Link: https://patch.msgid.link/20260918131620.405133-2-postadelmaga@gmail.com --- drivers/gpu/drm/nouveau/nvkm/subdev/clk/base.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/clk/base.c b/drivers/gpu/drm/nouveau/nvkm/subdev/clk/base.c index 572e63846315..5da82db71dde 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/clk/base.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/clk/base.c @@ -473,6 +473,7 @@ static int nvkm_clk_ustate_update(struct nvkm_clk *clk, int req) { struct nvkm_pstate *pstate; + bool found = false; int i = 0; if (!clk->allow_reclock) @@ -480,12 +481,14 @@ nvkm_clk_ustate_update(struct nvkm_clk *clk, int req) if (req != -1 && req != -2) { list_for_each_entry(pstate, &clk->states, head) { - if (pstate->pstate == req) + if (pstate->pstate == req) { + found = true; break; + } i++; } - if (pstate->pstate != req) + if (!found) return -EINVAL; req = i; } From 866dbd17c3d5f599b9d93ea9bd5b3d6859ff8350 Mon Sep 17 00:00:00 2001 From: Francesco Magazzu Date: Fri, 18 Sep 2026 15:16:18 +0200 Subject: [PATCH 200/694] drm/nouveau/clk: don't use the pstate cursor after the loop nvkm_pstate_prog() walks clk->states looking for the entry at index 'pstatei' and then keeps using the list_for_each_entry cursor after the loop. This is not triggerable today: every caller clamps the index against clk->state_nr before calling, so the loop always breaks on a real entry. It is safe by virtue of what the callers happen to do, not by anything the function itself checks. Should a caller ever pass an index that is not on the list, the cursor would point at the list head rather than at a pstate, and the pstate->base.domain[] and pstate->fanspeed accesses that follow would read past it. Rather than leave that trap in place for the next caller, track whether the entry was found and return -EINVAL if it was not. No functional change. Signed-off-by: Francesco Magazzu Reviewed-by: Lyude Paul Signed-off-by: Lyude Paul Link: https://patch.msgid.link/20260918131620.405133-3-postadelmaga@gmail.com --- drivers/gpu/drm/nouveau/nvkm/subdev/clk/base.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/clk/base.c b/drivers/gpu/drm/nouveau/nvkm/subdev/clk/base.c index 5da82db71dde..a43246ae681f 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/clk/base.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/clk/base.c @@ -270,13 +270,19 @@ nvkm_pstate_prog(struct nvkm_clk *clk, int pstatei) struct nvkm_fb *fb = subdev->device->fb; struct nvkm_pci *pci = subdev->device->pci; struct nvkm_pstate *pstate; + bool found = false; int ret, idx = 0; list_for_each_entry(pstate, &clk->states, head) { - if (idx++ == pstatei) + if (idx++ == pstatei) { + found = true; break; + } } + if (!found) + return -EINVAL; + nvkm_debug(subdev, "setting performance state %d\n", pstatei); clk->pstate = pstatei; From 7ca7b8b5f2cc89ee843c2eab3272aac5aafb7b73 Mon Sep 17 00:00:00 2001 From: Francesco Magazzu Date: Fri, 18 Sep 2026 15:16:19 +0200 Subject: [PATCH 201/694] drm/nouveau/device: don't use the pstate cursor after the loop nvkm_control_mthd_pstate_attr() looks up the pstate at the index supplied by userspace by walking clk->states, and then keeps using the list_for_each_entry cursor after the loop. This is not triggerable today: the function already rejects args->v0.state >= clk->state_nr before the loop, and clk->state_nr is kept in sync with the number of entries on clk->states, so the lookup always breaks on a real entry. Should the loop ever run to completion, the cursor would point at the list head rather than at a pstate, and the pstate->base.domain[] read and the walk of pstate->list that follow would read past it. Rather than leave that trap in place, track whether the entry was found and return -EINVAL if it was not, like the other lookup failures in this function. No functional change. Signed-off-by: Francesco Magazzu Reviewed-by: Lyude Paul Signed-off-by: Lyude Paul Link: https://patch.msgid.link/20260918131620.405133-4-postadelmaga@gmail.com --- drivers/gpu/drm/nouveau/nvkm/engine/device/ctrl.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/ctrl.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/ctrl.c index f2e9a06263ce..28702741a88b 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/device/ctrl.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/ctrl.c @@ -74,6 +74,7 @@ nvkm_control_mthd_pstate_attr(struct nvkm_control *ctrl, void *data, u32 size) const struct nvkm_domain *domain; struct nvkm_pstate *pstate; struct nvkm_cstate *cstate; + bool found = false; int i = 0, j = -1; u32 lo, hi; int ret = -ENOSYS; @@ -104,10 +105,15 @@ nvkm_control_mthd_pstate_attr(struct nvkm_control *ctrl, void *data, u32 size) if (args->v0.state != NVIF_CONTROL_PSTATE_ATTR_V0_STATE_CURRENT) { list_for_each_entry(pstate, &clk->states, head) { - if (i++ == args->v0.state) + if (i++ == args->v0.state) { + found = true; break; + } } + if (!found) + return -EINVAL; + lo = pstate->base.domain[domain->name]; hi = lo; list_for_each_entry(cstate, &pstate->list, head) { From e5cccdafc855cd5f96f4b51d38114a0360b075d7 Mon Sep 17 00:00:00 2001 From: Francesco Magazzu Date: Fri, 18 Sep 2026 15:16:20 +0200 Subject: [PATCH 202/694] drm/nouveau/clk: don't clobber reclock status when restoring volt/fan nvkm_cstate_prog() reuses 'ret' for the voltage and fan-speed restore calls it makes after reprogramming the clocks. Those calls almost always succeed, so the status of the reclock itself is overwritten and the function reports success even when clk->func->calc() or clk->func->prog() failed. The converse is also true: a successful reclock is reported as an error if the final restore call fails, even though that failure is only logged and otherwise ignored. The only consumer of the return value is the error message in nvkm_pstate_work(), so in practice a failing reclock is simply never reported. Nothing else changes, but a function that returns success on failure is a trap for the next caller. Keep the calc/prog status in 'ret' and use a separate local for the restore calls. Fixes: 3eca809b3c05 ("drm/nouveau/clk: cosmetic changes") Signed-off-by: Francesco Magazzu Reviewed-by: Lyude Paul Signed-off-by: Lyude Paul Link: https://patch.msgid.link/20260918131620.405133-5-postadelmaga@gmail.com --- drivers/gpu/drm/nouveau/nvkm/subdev/clk/base.c | 16 +++++++++------- 1 file changed, 9 insertions(+), 7 deletions(-) diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/clk/base.c b/drivers/gpu/drm/nouveau/nvkm/subdev/clk/base.c index a43246ae681f..1cb83edc78dc 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/clk/base.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/clk/base.c @@ -199,16 +199,18 @@ nvkm_cstate_prog(struct nvkm_clk *clk, struct nvkm_pstate *pstate, int cstatei) } if (volt) { - ret = nvkm_volt_set_id(volt, cstate->voltage, - pstate->base.voltage, clk->temp, -1); - if (ret && ret != -ENODEV) - nvkm_error(subdev, "failed to lower voltage: %d\n", ret); + int err = nvkm_volt_set_id(volt, cstate->voltage, + pstate->base.voltage, clk->temp, -1); + + if (err && err != -ENODEV) + nvkm_error(subdev, "failed to lower voltage: %d\n", err); } if (therm) { - ret = nvkm_therm_cstate(therm, pstate->fanspeed, -1); - if (ret && ret != -ENODEV) - nvkm_error(subdev, "failed to lower fan speed: %d\n", ret); + int err = nvkm_therm_cstate(therm, pstate->fanspeed, -1); + + if (err && err != -ENODEV) + nvkm_error(subdev, "failed to lower fan speed: %d\n", err); } return ret; From 6a6870d3077faa501ca97760057ddca22b68418d Mon Sep 17 00:00:00 2001 From: Peiyang He Date: Fri, 18 Sep 2026 10:53:12 +0800 Subject: [PATCH 203/694] drm/nouveau: don't bump pin count on failed re-pin in nouveau_bo_pin_locked() nouveau_bo_pin_locked() checks whether an already pinned BO is in a memory domain compatible with a new pin request. When the domains are incompatible, it sets -EBUSY but still calls ttm_bo_pin() before returning. Callers treat a failed nouveau_bo_pin() as not having acquired a new pin, so the extra pin count is never decreased by a matching unpin. This triggers the warning in ttm_bo_release(): WARN_ON_ONCE(bo->pin_count); Found when fuzzing the nouveau driver with a modified Syzkaller: WARNING: drivers/gpu/drm/ttm/ttm_bo.c:256 at ttm_bo_release+0x827/0x9e0 drivers/gpu/drm/ttm/ttm_bo.c:256, CPU#1: syz.3.24/2212 Modules linked in: CPU: 1 UID: 0 PID: 2212 Comm: syz.3.24 Not tainted 7.2.0 #24 PREEMPT(lazy) nouveau 0000:01:00.0: gsp:msg fn:103 len:0x40/0x20 res:0x19 resp:0x19 Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.16.3-debian-1.16.3-2 04/01/2014 RIP: 0010:ttm_bo_release+0x827/0x9e0 drivers/gpu/drm/ttm/ttm_bo.c:256 Code: 02 00 0f 85 51 01 00 00 48 8b 7b 08 e8 d2 20 01 00 e9 80 fd ff ff e8 d8 15 c0 fe 90 0f 0b 90 e9 e1 f8 ff ff e8 ca 15 c0 fe 90 <0f> 0b 90 e9 a4 f8 ff ff e8 bc 15 c0 fe be 03 00 00 00 4c 89 e7 e8 msg: 00000000: 05 00 d0 c1 04 00 f0 f1 01 30 00 00 2d 90 00 00 .........0..-... RSP: 0018:ffffc9000f5cf710 EFLAGS: 00010293 RAX: 0000000000000000 RBX: ffff888018e5d2a8 RCX: ffffffff82bb1b36 RDX: ffff888017b68000 RSI: 0000000000000004 RDI: ffff888018e5d2a8 msg: 00000010: 19 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................ RBP: ffff88801261c720 R08: 0000000000000001 R09: ffffed10031cba55 R10: ffff888018e5d2ab R11: 00000000000000f3 R12: ffff888018e5d290 R13: ffff888018e5d2d4 R14: ffff88801b219c18 R15: dffffc0000000000 FS: 0000000000000000(0000) GS:ffff8880e0f6f000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000001b31223ffc CR3: 0000000028e00005 CR4: 0000000000770ef0 PKRU: 80000000 Call Trace: kref_put include/linux/kref.h:65 [inline] ttm_bo_put drivers/gpu/drm/ttm/ttm_bo.c:325 [inline] ttm_bo_fini+0x55/0x80 drivers/gpu/drm/ttm/ttm_bo.c:330 nouveau_gem_object_del+0xb2/0x1b0 drivers/gpu/drm/nouveau/nouveau_gem.c:90 drm_gem_object_free+0x5f/0x90 drivers/gpu/drm/drm_gem.c:1165 kref_put include/linux/kref.h:65 [inline] __drm_gem_object_put include/drm/drm_gem.h:562 [inline] drm_gem_object_put include/drm/drm_gem.h:575 [inline] nouveau_abi16_chan_fini.constprop.0+0x44f/0x5a0 drivers/gpu/drm/nouveau/nouveau_abi16.c:195 nouveau 0000:01:00.0: syz.2.23[2209]: Unknown handle 0x00000000 nouveau_abi16_fini+0x1d0/0x340 drivers/gpu/drm/nouveau/nouveau_abi16.c:225 nouveau_drm_postclose+0x18b/0x3e0 drivers/gpu/drm/nouveau/nouveau_drm.c:1284 nouveau 0000:01:00.0: syz.2.23[2209]: validate_init drm_file_free.part.0+0x6d6/0xb60 drivers/gpu/drm/drm_file.c:267 drm_file_free drivers/gpu/drm/drm_file.c:237 [inline] drm_close_helper.isra.0+0x11a/0x160 drivers/gpu/drm/drm_file.c:290 drm_release+0x1ab/0x330 drivers/gpu/drm/drm_file.c:438 __fput+0x39c/0xa60 fs/file_table.c:512 nouveau 0000:01:00.0: syz.2.23[2209]: validate: -2 task_work_run+0x15a/0x230 kernel/task_work.c:233 exit_task_work include/linux/task_work.h:40 [inline] do_exit+0x82b/0x25a0 kernel/exit.c:1009 do_group_exit+0xc2/0x280 kernel/exit.c:1152 get_signal+0x1d6e/0x1f30 kernel/signal.c:3046 arch_do_signal_or_restart+0x7d/0x6e0 arch/x86/kernel/signal.c:337 __exit_to_user_mode_loop kernel/entry/common.c:66 [inline] exit_to_user_mode_loop+0xdf/0x440 kernel/entry/common.c:101 __exit_to_user_mode_prepare include/linux/irq-entry-common.h:207 [inline] syscall_exit_to_user_mode_prepare include/linux/irq-entry-common.h:230 [inline] syscall_exit_to_user_mode include/linux/entry-common.h:318 [inline] do_syscall_64+0x4f8/0x690 arch/x86/entry/syscall_64.c:100 entry_SYSCALL_64_after_hwframe+0x77/0x7f RIP: 0033:0x7f12bac8594d Code: Unable to access opcode bytes at 0x7f12bac85923. RSP: 002b:00007f12b96e70d8 EFLAGS: 00000246 ORIG_RAX: 00000000000000ca RAX: 0000000000000001 RBX: 00007f12baf15fa8 RCX: 00007f12bac8594d RDX: 00000000000f4240 RSI: 0000000000000081 RDI: 00007f12baf15fac RBP: 00007f12baf15fa0 R08: 00007f12baee8000 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000000 R13: 00007f12baf16038 R14: 0000000000000006 R15: 00007ffe2ed394b0 irq event stamp: 47867 hardirqs last enabled at (47883): [] __up_console_sem+0x66/0x70 kernel/printk/printk.c:347 hardirqs last disabled at (47892): [] __up_console_sem+0x4b/0x70 kernel/printk/printk.c:345 softirqs last enabled at (47880): [] __do_softirq kernel/softirq.c:656 [inline] softirqs last enabled at (47880): [] invoke_softirq kernel/softirq.c:496 [inline] softirqs last enabled at (47880): [] __irq_exit_rcu+0x137/0x1c0 kernel/softirq.c:735 softirqs last disabled at (47875): [] __do_softirq kernel/softirq.c:656 [inline] softirqs last disabled at (47875): [] invoke_softirq kernel/softirq.c:496 [inline] softirqs last disabled at (47875): [] __irq_exit_rcu+0x137/0x1c0 kernel/softirq.c:735 Fix by calling ttm_bo_pin() only when the existing placement is compatible with the new pin request. This matches the correct behavior in other DRM drivers such as amdgpu_bo_pin() in amdgpu. Cc: stable@vger.kernel.org Fixes: ad76b3f7c7a0 ("drm/nouveau: teach nouveau_bo_pin() how to force a contig vram allocation") Signed-off-by: Peiyang He Assisted-by: LLM Reviewed-by: Lyude Paul Signed-off-by: Lyude Paul Link: https://patch.msgid.link/EACEF2F4E098413F+20260918025312.2814889-1-peiyang_he@smail.nju.edu.cn --- drivers/gpu/drm/nouveau/nouveau_bo.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/nouveau/nouveau_bo.c b/drivers/gpu/drm/nouveau/nouveau_bo.c index 0e8de6d4b36f..6dcb92575eb4 100644 --- a/drivers/gpu/drm/nouveau/nouveau_bo.c +++ b/drivers/gpu/drm/nouveau/nouveau_bo.c @@ -578,8 +578,9 @@ int nouveau_bo_pin_locked(struct nouveau_bo *nvbo, uint32_t domain, bool contig) "0x%08x vs 0x%08x\n", bo, bo->resource->mem_type, domain); ret = -EBUSY; + } else { + ttm_bo_pin(&nvbo->bo); } - ttm_bo_pin(&nvbo->bo); goto out; } From 1717fcc5be575d4768279148ae9465a8b13d4339 Mon Sep 17 00:00:00 2001 From: Giuseppe Ranieri Date: Thu, 17 Sep 2026 21:51:14 +0000 Subject: [PATCH 204/694] drm/nouveau/disp: don't reject HDMI config on cards without SCDC nv50_hdmi_enable() passes the sink's SCDC capability from its EDID straight through to nvif_outp_hdmi(). On pre-Maxwell-2 cards there is no hdmi->scdc callback, so nvkm_uoutp_mthd_hdmi() rejects the whole configuration with -EINVAL, and nv50_hdmi_enable() returns before hdmi->ctrl() runs and before the AVI and VSI infoframes are sent. The result on such a card driving an SCDC-capable HDMI 2.0 sink is that HDMI audio silently stops working. Video is unaffected, and nothing is logged, which makes the failure hard to attribute. SCDC is optional, and the hdmi->scdc() call further down is already guarded against a missing callback. Requesting it on a card that cannot do it need not invalidate the rest of the HDMI configuration, so drop that term from the condition and let the existing guard skip SCDC alone. Fixes: 6c6abab20b99 ("drm/nouveau/disp: add output hdmi config method") Signed-off-by: Giuseppe Ranieri Co-Authored-By: Tano Dzhinski Signed-off-by: Tano Dzhinski Tested-by: Tano Dzhinski Reviewed-by: Lyude Paul Signed-off-by: Lyude Paul Link: https://patch.msgid.link/20260917215114.1136715-1-tano.dzhinski@gmail.com --- drivers/gpu/drm/nouveau/nvkm/engine/disp/uoutp.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/uoutp.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/uoutp.c index 377d0e0cef84..9887b3898505 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/uoutp.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/uoutp.c @@ -253,8 +253,7 @@ nvkm_uoutp_mthd_hdmi(struct nvkm_outp *outp, void *argv, u32 argc) if (!ior->func->hdmi || args->v0.max_ac_packet > 0x1f || - args->v0.rekey > 0x7f || - (args->v0.scdc && !ior->func->hdmi->scdc)) + args->v0.rekey > 0x7f) return -EINVAL; if (!args->v0.enable) { From d58384c22739848efe14b34e9586e4f1242f33c0 Mon Sep 17 00:00:00 2001 From: Liz Fong-Jones Date: Fri, 18 Sep 2026 03:56:33 +0000 Subject: [PATCH 205/694] PCI: Fix BAR resize for devices on a root bus MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit pci_do_resource_release_and_resize() releases device BARs that share a bridge window with the BAR being resized, but when the device sits directly on a root bus (pdev->bus->self == NULL) it then skips resource assignment entirely and returns success, leaving the BARs it just released unassigned (IORESOURCE_UNSET). Skipping pbus_reassign_bridge_resources() is correct in that case -- there is no bridge window to adjust -- but the device BARs still have to be reassigned. Before the BAR release was consolidated into the PCI core, this case worked for amdgpu because the driver released the BARs itself and then called pci_assign_unassigned_bus_resources() unconditionally after the resize, which assigns unassigned device BARs also on a root bus. Commit db92e3fef53e ("drm/amdgpu: Remove driver side BAR release before resize") removed that call, so nothing assigns the released BARs anymore. This breaks amdgpu completely on the SolidRun HoneyComb LX2K (NXP LX2160A, arm64, ACPI), where ACPI doesn't expose the Root Port so the GPU endpoint appears directly on a "root bus" of its segment: amdgpu 0004:01:00.0: BAR 0 [mem 0xa400000000-0xa40fffffff 64bit pref]: releasing amdgpu 0004:01:00.0: BAR 2 [mem 0xa410000000-0xa4101fffff 64bit pref]: releasing amdgpu 0004:01:00.0: sw_init of IP block failed -19 amdgpu 0004:01:00.0: amdgpu_device_ip_init failed amdgpu 0004:01:00.0: Fatal error during GPU init No error is logged because the resize path reports success; amdgpu then finds BAR 0 IORESOURCE_UNSET and bails out with -ENODEV. When there is no upstream bridge, call pci_bus_assign_resources() on the root bus to place the BARs released above, using the same alignment-sorted algorithm as normal enumeration instead of a manual per-BAR loop. This also walks the rest of the hierarchy under the root bus, as pci_assign_unassigned_bus_resources() used to for amdgpu before commit db92e3fef53e ("drm/amdgpu: Remove driver side BAR release before resize") removed that call -- the core-side fix that commit asked for ("such a problem should be fixed inside pci_resize_resource() instead"). pci_bus_assign_resources() returns void, so failure is detected by checking whether the released BARs are still assigned afterward; if not, roll back as in the bridged case. This is stricter than the bridged path -- it fails on any unplaced resource, not just required ones -- since a root bus typically has one shared window, and failing loudly seemed better than leaving something silently unassigned. The root bus path also had a locking bug that any fix here necessarily touches: the old "goto out" jumped to up_read(&pci_bus_sem) without a matching down_read() (as does the "goto restore" taken when pci_dev_res_add_to_list() fails in the release loop). Take pci_bus_sem before the BAR release loop so every path through the function holds it exactly once. Fixes: 337b1b566db0 ("PCI: Fix restoring BARs on BAR resize rollback path") Link: https://bugs.launchpad.net/ubuntu/+source/linux-hwe-7.0/+bug/2159596 Suggested-by: Ilpo Järvinen Assisted-by: Claude:claude-fable-5 checkpatch Assisted-by: Claude:claude-sonnet-5 Signed-off-by: Liz Fong-Jones [bhelgaas: commit log] Signed-off-by: Bjorn Helgaas Reviewed-by: Ilpo Järvinen Cc: stable@vger.kernel.org Link: https://patch.msgid.link/20260918035633.566823-1-lizf@honeycomb.io --- drivers/pci/setup-bus.c | 23 +++++++++++++++++------ 1 file changed, 17 insertions(+), 6 deletions(-) diff --git a/drivers/pci/setup-bus.c b/drivers/pci/setup-bus.c index e8c94aa1d3c1..ed16ef7c26fa 100644 --- a/drivers/pci/setup-bus.c +++ b/drivers/pci/setup-bus.c @@ -2380,6 +2380,7 @@ int pci_do_resource_release_and_resize(struct pci_dev *pdev, int resno, int size struct resource *res = pci_resource_n(pdev, resno); struct pci_dev_resource *dev_res; struct pci_bus *bus = pdev->bus; + struct pci_dev *bridge = pci_upstream_bridge(pdev); struct resource *b_win, *r; LIST_HEAD(saved); unsigned int i; @@ -2397,6 +2398,8 @@ int pci_do_resource_release_and_resize(struct pci_dev *pdev, int resno, int size if (ret) return ret; + down_read(&pci_bus_sem); + pci_dev_for_each_resource(pdev, r, i) { if (i >= PCI_BRIDGE_RESOURCES) break; @@ -2415,13 +2418,21 @@ int pci_do_resource_release_and_resize(struct pci_dev *pdev, int resno, int size pci_resize_resource_set_size(pdev, resno, size); - if (!bus->self) - goto out; + if (bridge) { + ret = pbus_reassign_bridge_resources(bus, res, &saved); + if (ret) + goto restore; + } else { + /* No bridge window to adjust; let the core reassign the bus. */ + pci_bus_assign_resources(bus); - down_read(&pci_bus_sem); - ret = pbus_reassign_bridge_resources(bus, res, &saved); - if (ret) - goto restore; + list_for_each_entry(dev_res, &saved, list) { + if (!resource_assigned(dev_res->res)) { + ret = -ENOSPC; + goto restore; + } + } + } out: up_read(&pci_bus_sem); From 8805840aad73df7146778be243a196d48b4f6430 Mon Sep 17 00:00:00 2001 From: Angel J Date: Fri, 18 Sep 2026 14:55:40 -0500 Subject: [PATCH 206/694] PCI: of_property: Omit bus properties without a subordinate bus A bridge (a device with a Type 1 header) may not have a secondary bus allocated (pdev->subordinate), e.g., if there are no available bus numbers or the bridge secondary/subordinate bus numbers are not writable. The dynamic OF helpers of_pci_prop_bus_range() and of_pci_prop_intr_map() dereference pdev->subordinate without checking it. When CONFIG_PCI_DYNAMIC_OF_NODES is enabled, this can cause a NULL pointer dereference and early boot hang. Generate 'bus-range' and 'interrupt-map' properties only when a subordinate bus exists. Keep the node and its remaining properties for bridges without one. The problem was latent since 407d1a51921e ("PCI: Create device tree node for bridge"), but wasn't reachable until 1f340724419e ("PCI: of: Create device tree PCI host bridge node"), which appeared in v6.15. Before 1f340724419e, of_pci_make_dev_node() returned early because the parent OF node was missing. Fixes: 407d1a51921e ("PCI: Create device tree node for bridge") Signed-off-by: Angel J [bhelgaas: move pdev->subordinate test to callees, commit log] Signed-off-by: Bjorn Helgaas Cc: stable@vger.kernel.org # v6.6+ Link: https://patch.msgid.link/20260918195540.GA1187209@bhelgaas --- drivers/pci/of_property.c | 11 +++++++++-- 1 file changed, 9 insertions(+), 2 deletions(-) diff --git a/drivers/pci/of_property.c b/drivers/pci/of_property.c index 75a358f73e69..1500740cc55d 100644 --- a/drivers/pci/of_property.c +++ b/drivers/pci/of_property.c @@ -95,9 +95,13 @@ static int of_pci_prop_bus_range(struct pci_dev *pdev, struct of_changeset *ocs, struct device_node *np) { - u32 bus_range[] = { pdev->subordinate->busn_res.start, - pdev->subordinate->busn_res.end }; + u32 bus_range[2]; + if (!pdev->subordinate) + return 0; + + bus_range[0] = pdev->subordinate->busn_res.start; + bus_range[1] = pdev->subordinate->busn_res.end; return of_changeset_add_prop_u32_array(ocs, np, "bus-range", bus_range, ARRAY_SIZE(bus_range)); } @@ -220,6 +224,9 @@ static int of_pci_prop_intr_map(struct pci_dev *pdev, struct of_changeset *ocs, int ret; u8 pin; + if (!pdev->subordinate) + return 0; + pnode = pci_device_to_OF_node(pdev->bus->self); if (!pnode) pnode = pci_bus_to_OF_node(pdev->bus); From 99cc2a62e07a44a22254d7beca9ef1f8ad886d0d Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Sun, 13 Sep 2026 04:42:33 +0000 Subject: [PATCH 207/694] tipc: reject invalid and unexpected GRP_ACK_MSG to prevent bc_ackers underflow Commit 48a5fe38772b ("tipc: fix bc_ackers underflow on duplicate GRP_ACK_MSG") rejected duplicate/stale ACKs in tipc_group_proto_rcv() by returning early when less_eq(acked, m->bc_acked). However, that check remains incomplete in two ways: 1. When grp->bc_ackers is zero (e.g. on a quiet group, when replicast ACKs were not requested, or after all expected members have already acknowledged), an unexpected GRP_ACK_MSG with acked > m->bc_acked passes less_eq() and unconditionally decrements grp->bc_ackers. Because bc_ackers is a u16, this wraps to 65535, causing tipc_group_bc_cong() to permanently report congestion and blocking all future group broadcasts on the socket. 2. During an active broadcast round (grp->bc_ackers > 0), the sender transmits packet S and advances grp->bc_snd_nxt to S + 1. Receivers increment their expected counter to S + 1 upon consuming packet S, so the only valid ACK value for the current round is strictly acked == grp->bc_snd_nxt. However, tipc_group_update_bc_members() initializes each member's m->bc_acked to prev = grp->bc_snd_nxt - 1 (S - 1 before increment). This leaves a 2-sequence gap (S - 1 to S + 1) in sequence space. An incoming ACK is therefore neither rejected as duplicate nor prevented from decrementing grp->bc_ackers if an unexpected or stale value (such as S) is received. A member sending acked = S followed by acked = S + 1 could decrement grp->bc_ackers twice in the same round, prematurely clearing bc_ackers or underflowing it. Fix this by: - Dropping GRP_ACK_MSG immediately if grp->bc_ackers is zero. - Requiring acked == grp->bc_snd_nxt and rejecting duplicates where m->bc_acked == acked. Because replicast broadcast rounds are strictly sequential, only grp->bc_snd_nxt can be acknowledged, and each member can acknowledge at most once per round. Note that a related pre-existing issue in tipc_group_delete_member() (where grp->bc_ackers decrementing to zero upon member departure does not restore *grp->open or trigger a socket wakeup) will be addressed in a separate patch. Fixes: 48a5fe38772b ("tipc: fix bc_ackers underflow on duplicate GRP_ACK_MSG") Fixes: 2f487712b893 ("tipc: guarantee that group broadcast doesn't bypass group unicast") Reported-by: James Burton Cc: stable@vger.kernel.org Signed-off-by: Eric Dumazet Link: https://patch.msgid.link/20260913044233.193927-1-edumazet@google.com Signed-off-by: Jakub Kicinski --- net/tipc/group.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/net/tipc/group.c b/net/tipc/group.c index 14e6732624e2..74f6d3dac078 100644 --- a/net/tipc/group.c +++ b/net/tipc/group.c @@ -797,10 +797,10 @@ void tipc_group_proto_rcv(struct tipc_group *grp, bool *usr_wakeup, tipc_group_open(m, usr_wakeup); return; case GRP_ACK_MSG: - if (!m) + if (!m || !grp->bc_ackers) return; acked = msg_grp_bc_acked(hdr); - if (less_eq(acked, m->bc_acked)) + if (acked != grp->bc_snd_nxt || m->bc_acked == acked) return; m->bc_acked = acked; if (--grp->bc_ackers) From 47abe7a5c4eb53269aca3506446f851572a059a3 Mon Sep 17 00:00:00 2001 From: Nguyen Ngoc Thang Date: Tue, 15 Sep 2026 22:08:16 +0700 Subject: [PATCH 208/694] net/sched: act_ct: don't WARN on benign flow_offload_alloc() failure flow_offload_alloc() returns NULL when the conntrack entry is dying (e.g. raced with a conntrack flush) or when the GFP_ATOMIC allocation fails; both are expected under load and neither is a kernel bug. This path runs from softirq on every committed packet, so with panic_on_warn=1 an unprivileged user can panic the box just by racing a conntrack flush against a `tc ... action ct commit` classifier. Reproduced with a custom repro under QEMU: a small, fixed set of UDP flows through `tc filter ... action ct commit` on lo, raced against threads flooding bare ctnetlink CT_DELETE (flush) requests. Hits WARNING: net/sched/act_ct.c:437 (tcf_ct_flow_table_add(), inlined into tcf_ct_act() in this build) within ~15s on the unpatched kernel; same setup is clean on the patched kernel. The fix itself is behavior-preserving: both branches already did `goto err_alloc` before and after, only the WARN is removed. Fixes: 64ff70b80fd4 ("net/sched: act_ct: Offload established connections to flow table") Reported-by: syzbot+6cc37aba98dac721c415@syzkaller.appspotmail.com Closes: https://syzkaller.appspot.com/bug?extid=6cc37aba98dac721c415 Signed-off-by: Nguyen Ngoc Thang Reviewed-by: Simon Horman Link: https://patch.msgid.link/20260915150816.36487-1-ngocthang2710.1999@gmail.com Signed-off-by: Jakub Kicinski --- net/sched/act_ct.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/net/sched/act_ct.c b/net/sched/act_ct.c index 9080cb386c16..55f3521edb4c 100644 --- a/net/sched/act_ct.c +++ b/net/sched/act_ct.c @@ -432,11 +432,10 @@ static void tcf_ct_flow_table_add(struct tcf_ct_flow_table *ct_ft, if (test_and_set_bit(IPS_OFFLOAD_BIT, &ct->status)) return; + /* NULL if ct is dying (raced flush) or the atomic alloc failed. */ entry = flow_offload_alloc(ct); - if (!entry) { - WARN_ON_ONCE(1); + if (!entry) goto err_alloc; - } if (tcp) { ct->proto.tcp.seen[0].flags |= IP_CT_TCP_FLAG_BE_LIBERAL; From 2566866fc30965d915d0b52b5c3323b362619f0e Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?J=C3=A9r=C3=A9my=20Jean?= Date: Tue, 15 Sep 2026 12:48:07 +0000 Subject: [PATCH 209/694] net: gue: reject invalid REMCSUM offsets MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The REMCSUM option carries an absolute checksum start and checksum field offset. gue_remcsum() passes them to skb_remcsum_process(), whose partial path stores offset - start in the u16 skb->csum_offset variable. If offset is less than start, this underflows. A forwarded packet can retain CHECKSUM_PARTIAL and reach a NETIF_F_HW_CSUM driver which trusts the metadata, leading skb_copy_and_csum_dev() to write two bytes about 64 KiB beyond the destination buffer. Reject reversed tuples in validate_gue_flags(), after the existing length validation, so all GUE parsers enforce the ordering in one place. Fixes: fe881ef11cf0 ("gue: Use checksum partial with remote checksum offload") Signed-off-by: Jérémy Jean Link: https://patch.msgid.link/20260915124806.2852293-2-Jeremy.Jean@oss.cyber.gouv.fr Signed-off-by: Jakub Kicinski --- include/net/gue.h | 19 +++++++++++++++---- 1 file changed, 15 insertions(+), 4 deletions(-) diff --git a/include/net/gue.h b/include/net/gue.h index caefd6da8693..d377155fd0b3 100644 --- a/include/net/gue.h +++ b/include/net/gue.h @@ -84,8 +84,9 @@ static inline size_t guehdr_priv_flags_len(__be32 flags) } /* Validate standard and private flags. Returns non-zero (meaning invalid) - * if there is an unknown standard or private flags, or the options length for - * the flags exceeds the options length specific in hlen of the GUE header. + * if there is an unknown standard or private flags, if the options length for + * the flags exceeds the options length specified in hlen of the GUE header, or + * if a private option contains invalid data. */ static inline int validate_gue_flags(struct guehdr *guehdr, size_t optlen) { @@ -103,8 +104,8 @@ static inline int validate_gue_flags(struct guehdr *guehdr, size_t optlen) /* Private flags are last four bytes accounted in * guehdr_flags_len */ - __be32 pflags = *(__be32 *)((void *)&guehdr[1] + - len - GUE_LEN_PRIV); + void *data = (void *)&guehdr[1] + len; + __be32 pflags = *(__be32 *)(data - GUE_LEN_PRIV); if (pflags & ~GUE_PFLAGS_ALL) return 1; @@ -112,6 +113,16 @@ static inline int validate_gue_flags(struct guehdr *guehdr, size_t optlen) len += guehdr_priv_flags_len(pflags); if (len > optlen) return 1; + + if (pflags & GUE_PFLAG_REMCSUM) { + __be16 *pd = data; + + /* The field offset pd[1] must not be less + * than the start pd[0]. + */ + if (ntohs(pd[1]) < ntohs(pd[0])) + return 1; + } } return 0; From 310d1ac61a4d5a2ca8356a3a48d263acf54503ce Mon Sep 17 00:00:00 2001 From: Lorenzo Bianconi Date: Wed, 16 Sep 2026 15:30:13 +0200 Subject: [PATCH 210/694] net: ethernet: mtk_eth_soc: unregister net_devices in case of probe failure If register_netdev() fails for one of the MTK_MAX_DEVS devices in mtk_probe(), the error path jumps to err_deinit_ppe, skipping mtk_unreg_dev(). The previously registered net_devices are then freed by mtk_free_dev() while still in NETREG_REGISTERED state, hitting the BUG_ON(dev->reg_state != NETREG_UNREGISTERED). Route the register_netdev() failure to err_unreg_netdev so the net_devices registered so far are properly unregistered before being freed. Fixes: 8a8a9e89f801 ("net: ethernet: mediatek: cleanup error path inside mtk_hw_init") Signed-off-by: Lorenzo Bianconi Link: https://patch.msgid.link/20260916-mtk_eth_soc-netdev-fix-v1-1-5dac50eb65b1@oss.qualcomm.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/mediatek/mtk_eth_soc.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/mediatek/mtk_eth_soc.c b/drivers/net/ethernet/mediatek/mtk_eth_soc.c index fd7a49ae88d0..2ea5dfe85539 100644 --- a/drivers/net/ethernet/mediatek/mtk_eth_soc.c +++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.c @@ -4509,6 +4509,10 @@ static int mtk_unreg_dev(struct mtk_eth *eth) mac = netdev_priv(eth->netdev[i]); if (MTK_HAS_CAPS(eth->soc->caps, MTK_QDMA)) unregister_netdevice_notifier(&mac->device_notifier); + + if (eth->netdev[i]->reg_state != NETREG_REGISTERED) + continue; + unregister_netdev(eth->netdev[i]); } @@ -5344,7 +5348,7 @@ static int mtk_probe(struct platform_device *pdev) err = register_netdev(eth->netdev[i]); if (err) { dev_err(eth->dev, "error bringing up device\n"); - goto err_deinit_ppe; + goto err_unreg_netdev; } else netif_info(eth, probe, eth->netdev[i], "mediatek frame engine at 0x%08lx, irq %d\n", From 24fedc7a569bce181728f0dd616504bef9f1513b Mon Sep 17 00:00:00 2001 From: Andy Moreton Date: Wed, 16 Sep 2026 13:56:41 +0100 Subject: [PATCH 211/694] sfc: add X4D PF support X4D is an X4 controller instance as an IP block in an SoC. It has the same feature set as X4. Signed-off-by: Andy Moreton Reviewed-by: Pieter Jansen van Vuuren Reviewed-by: Alejandro Lucero Reviewed-by: Simon Horman Link: https://patch.msgid.link/20260916125641.12238-1-alucerop@amd.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/sfc/efx.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/drivers/net/ethernet/sfc/efx.c b/drivers/net/ethernet/sfc/efx.c index 3806cd3dd7f4..953f1d90b9b1 100644 --- a/drivers/net/ethernet/sfc/efx.c +++ b/drivers/net/ethernet/sfc/efx.c @@ -904,6 +904,10 @@ static const struct pci_device_id efx_pci_table[] = { .driver_data = (unsigned long)&efx_x4_nic_type}, {PCI_DEVICE(PCI_VENDOR_ID_SOLARFLARE, 0x2c03), /* X4 PF (FF only) */ .driver_data = (unsigned long)&efx_x4_nic_type}, + {PCI_DEVICE(PCI_VENDOR_ID_SOLARFLARE, 0x8c03), /* X4D PF (FF/LL) */ + .driver_data = (unsigned long)&efx_x4_nic_type}, + {PCI_DEVICE(PCI_VENDOR_ID_SOLARFLARE, 0xac03), /* X4D PF (FF only) */ + .driver_data = (unsigned long)&efx_x4_nic_type}, {0} /* end of list */ }; From ee319bd3a0e976af5087cbe59ebc50a66f31d202 Mon Sep 17 00:00:00 2001 From: Norbert Szetei Date: Wed, 16 Sep 2026 21:57:53 +0200 Subject: [PATCH 212/694] ipv6: do not let ipv6_find_hdr() return an offset past the packet end ipv6_find_hdr() walks the extension header chain, skipping each header by the length that header itself declares. ipv6_optlen() returns up to 2048, and the skip is never checked against skb->len, so the offset stored in *offset can point past the end of the packet. openvswitch installs that offset as the transport header, and update_ipv6_checksum() then reads and writes the transport checksum field out of bounds: BUG: KASAN: slab-use-after-free in inet_proto_csum_replace16+0x445/0x470 Read of size 2 at addr ffff88810b754b06 by task ovs_ipv6_oob/629 CPU: 4 UID: 1000 PID: 629 Comm: ovs_ipv6_oob Tainted: G N 7.3.0-rc3+ #348 Call Trace: inet_proto_csum_replace16+0x445/0x470 set_ipv6_addr+0x3dd/0x460 do_execute_actions+0x6a3d/0x7c40 ovs_execute_actions+0xfd/0x480 ovs_packet_cmd_execute+0xc38/0xf20 genl_rcv_msg+0x59e/0x870 netlink_rcv_skb+0x18b/0x450 genl_rcv+0x2d/0x40 netlink_unicast+0x6bc/0xa20 The buggy address belongs to the object at ffff88810b754980 which belongs to the cache skbuff_small_head of size 704 The buggy address is located 390 bytes inside of freed 704-byte region [ffff88810b754980, ffff88810b754c40) Other callers use that offset too, so bound it here rather than in one caller. Reject a header whose declared length does not fit in the packet. ipv6_find_hdr() already fails with -EBADMSG on a malformed chain, so this adds no new failure mode. Fixes: f8f626754ebe ("ipv6: Move ipv6_find_hdr() out of Netfilter code.") Suggested-by: Ilya Maximets Suggested-by: Eric Dumazet Cc: stable@vger.kernel.org Signed-off-by: Norbert Szetei Reviewed-by: Ido Schimmel Reviewed-by: Ilya Maximets Link: https://patch.msgid.link/8F80BA1A-DDFD-432D-9075-242A3435FEB5@doyensec.com Signed-off-by: Jakub Kicinski --- net/ipv6/exthdrs_core.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/net/ipv6/exthdrs_core.c b/net/ipv6/exthdrs_core.c index 9d06d487e8b1..4a9748338cf4 100644 --- a/net/ipv6/exthdrs_core.c +++ b/net/ipv6/exthdrs_core.c @@ -278,6 +278,9 @@ int ipv6_find_hdr(const struct sk_buff *skb, unsigned int *offset, hdrlen = ipv6_optlen(hp); if (!found) { + if (skb->len - start < hdrlen) + return -EBADMSG; + nexthdr = hp->nexthdr; start += hdrlen; } From dd47bcf279f1083f09bf5266890b26263361022b Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Wed, 16 Sep 2026 23:09:24 +0000 Subject: [PATCH 213/694] ip6_gre: Call ip6erspan_tunnel_unlink_md() in ip6erspan_changelink(). The cited commit accidentally added ip6gre_tunnel_unlink_md() in ip6erspan_changelink(). Let's correct it to ip6erspan_tunnel_unlink_md(). Fixes: b80d0b93b991 ("net: ip6_gre: fix tunnel metadata device sharing.") Signed-off-by: Kuniyuki Iwashima Reviewed-by: Xuanqiang Luo Reviewed-by: Ido Schimmel Link: https://patch.msgid.link/20260916230927.378957-1-kuniyu@google.com Signed-off-by: Jakub Kicinski --- net/ipv6/ip6_gre.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/ipv6/ip6_gre.c b/net/ipv6/ip6_gre.c index 8ebda0b6a78b..e61cb10b50dc 100644 --- a/net/ipv6/ip6_gre.c +++ b/net/ipv6/ip6_gre.c @@ -2279,7 +2279,7 @@ static int ip6erspan_changelink(struct net_device *dev, struct nlattr *tb[], return PTR_ERR(t); ip6erspan_set_version(data, &p); - ip6gre_tunnel_unlink_md(ign, t); + ip6erspan_tunnel_unlink_md(ign, t); ip6gre_tunnel_unlink(ign, t); ip6erspan_tnl_change(t, &p, !tb[IFLA_MTU]); ip6erspan_tunnel_link_md(ign, t); From 95c4d54ed02283e9a09e8cd7360e384daa67a741 Mon Sep 17 00:00:00 2001 From: Abhishek Ojha Date: Wed, 16 Sep 2026 19:19:28 -0400 Subject: [PATCH 214/694] net: phy: micrel: Advance register data pointer in write loop lanphy_write_reg_data() does not advance the data pointer while iterating over the register table. As a result, it writes the first entry num times and leaves the remaining errata registers unconfigured. Single-entry tables are unaffected, but tables with multiple entries leave every entry after the first unapplied. Advance the data pointer after each successful write so every table entry is applied in order. Fixes: c8732e933925 ("net: phy: micrel: lan8842 errata") Cc: stable@vger.kernel.org Signed-off-by: Abhishek Ojha Reviewed-by: Andrew Lunn Link: https://patch.msgid.link/20260916231928.1336305-1-abhishek.ojha@savoirfairelinux.com Signed-off-by: Jakub Kicinski --- drivers/net/phy/micrel.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/net/phy/micrel.c b/drivers/net/phy/micrel.c index ae830781824b..5c8461db7b4b 100644 --- a/drivers/net/phy/micrel.c +++ b/drivers/net/phy/micrel.c @@ -6344,6 +6344,7 @@ static int lanphy_write_reg_data(struct phy_device *phydev, data->val); if (ret) break; + data++; } return ret; From 1b82958f3f035df5ccaab5430a2302f08a5d5351 Mon Sep 17 00:00:00 2001 From: Alexander Duyck Date: Mon, 14 Sep 2026 14:09:57 -0700 Subject: [PATCH 215/694] net: ethtool: keep rtnl_lock for the ioctl self test An offline self test that brings the interface down and back up with netif_close() / netif_open() requires rtnl_lock for both. Since the ethtool IOCTL path became rtnl-optional for ops-locked drivers, the ETHTOOL_TEST ioctl runs holding only the netdev instance lock, so on an ops-locked driver the self test now tears the device down without rtnl_lock. With lockdep this reproduces deterministically on every offline self test on such a driver; note the sole lock held is the instance lock, not rtnl: WARNING: suspicious RCU usage net/core/netpoll.c:207 suspicious rcu_dereference_protected() usage! 1 lock held by ethtool/107: #0: (&dev->lock){+.+.}, at: dev_ethtool Call Trace: netpoll_poll_disable __dev_close_many netif_close_many netif_close fbnic_self_test dev_ethtool_locked dev_ethtool dev_ioctl sock_ioctl __x64_sys_ioctl Without lockdep the same condition trips ASSERT_RTNL() in __dev_close_many() / __dev_open(); that check only samples the global rtnl state, so it can be masked by a concurrent rtnl holder, but the device is still being reconfigured without the lock it requires. The ethtool self_test is a legacy ioctl-only command, so an ETHTOOL_TEST case is only needed on the ioctl path. Add an opt-in bit for drivers whose self test needs rtnl_lock and set it on the ops-locked drivers whose offline self test tears the interface down and up: - fbnic (ops-locked via queue_mgmt_ops): fbnic_self_test() offline path uses netif_close() / netif_open(). - bnxt (ops-locked via queue_mgmt_ops): bnxt_self_test() offline path goes through bnxt_close_nic() / bnxt_half_open_nic() / bnxt_half_close_nic() / bnxt_open_nic(), which close and reopen the device. Fixes: f994752b1127 ("net: ethtool: optionally skip rtnl_lock on IOCTL path") Signed-off-by: Alexander Duyck Reviewed-by: Simon Horman Link: https://patch.msgid.link/178942019771.7700.338431553546884773.stgit@ahduyck-xeon-server.home.arpa Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c | 3 ++- drivers/net/ethernet/meta/fbnic/fbnic_ethtool.c | 3 ++- include/linux/ethtool.h | 2 ++ net/ethtool/common.h | 2 ++ 4 files changed, 8 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c index 62bc9cae613c..622e89587e5d 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c @@ -5733,7 +5733,8 @@ const struct ethtool_ops bnxt_ethtool_ops = { .op_needs_rtnl = ETHTOOL_OP_NEEDS_RTNL_SCHANNELS | ETHTOOL_OP_NEEDS_RTNL_SRINGPARAM | ETHTOOL_OP_NEEDS_RTNL_SCOALESCE | - ETHTOOL_OP_NEEDS_RTNL_RSS, + ETHTOOL_OP_NEEDS_RTNL_RSS | + ETHTOOL_OP_NEEDS_RTNL_TEST, .supported_coalesce_params = ETHTOOL_COALESCE_USECS | ETHTOOL_COALESCE_MAX_FRAMES | ETHTOOL_COALESCE_USECS_IRQ | diff --git a/drivers/net/ethernet/meta/fbnic/fbnic_ethtool.c b/drivers/net/ethernet/meta/fbnic/fbnic_ethtool.c index 0e47088ec44b..423f179c9d47 100644 --- a/drivers/net/ethernet/meta/fbnic/fbnic_ethtool.c +++ b/drivers/net/ethernet/meta/fbnic/fbnic_ethtool.c @@ -2025,7 +2025,8 @@ static const struct ethtool_ops fbnic_ethtool_ops = { ETHTOOL_OP_NEEDS_RTNL_SPAUSEPARAM | ETHTOOL_OP_NEEDS_RTNL_SCHANNELS | ETHTOOL_OP_NEEDS_RTNL_SRINGPARAM | - ETHTOOL_OP_NEEDS_RTNL_GLINK, + ETHTOOL_OP_NEEDS_RTNL_GLINK | + ETHTOOL_OP_NEEDS_RTNL_TEST, .get_drvinfo = fbnic_get_drvinfo, .get_regs_len = fbnic_get_regs_len, .get_regs = fbnic_get_regs, diff --git a/include/linux/ethtool.h b/include/linux/ethtool.h index 253600c0eccd..c4c9ce038611 100644 --- a/include/linux/ethtool.h +++ b/include/linux/ethtool.h @@ -944,6 +944,7 @@ struct kernel_ethtool_ts_info { #define ETHTOOL_OP_NEEDS_RTNL_SPAUSEPARAM BIT(6) #define ETHTOOL_OP_NEEDS_RTNL_RSS BIT(7) #define ETHTOOL_OP_NEEDS_RTNL_GLINK BIT(8) +#define ETHTOOL_OP_NEEDS_RTNL_TEST BIT(9) /** * struct ethtool_ops - optional netdev operations @@ -981,6 +982,7 @@ struct kernel_ethtool_ts_info { * - netdev_update_features() * - netif_set_real_num_tx_queues() * - ethtool_op_get_link() (syncs link watch under rtnl_lock) + * - netif_open() / netif_close() (used by @self_test) * * @get_drvinfo: Report driver/device information. Modern drivers no * longer have to implement this callback. Most fields are diff --git a/net/ethtool/common.h b/net/ethtool/common.h index 4e5356e26f40..ae32e7fdb563 100644 --- a/net/ethtool/common.h +++ b/net/ethtool/common.h @@ -163,6 +163,8 @@ ethtool_ioctl_needs_rtnl(const struct net_device *dev, u32 ethcmd) return ops->op_needs_rtnl & ETHTOOL_OP_NEEDS_RTNL_RSS; case ETHTOOL_GLINK: return ops->op_needs_rtnl & ETHTOOL_OP_NEEDS_RTNL_GLINK; + case ETHTOOL_TEST: + return ops->op_needs_rtnl & ETHTOOL_OP_NEEDS_RTNL_TEST; } return false; } From 1f4c73064a50f53d596c6f1d06d2d700f43c4b32 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Bj=C3=B6rn=20T=C3=B6pel?= Date: Mon, 14 Sep 2026 14:10:04 -0700 Subject: [PATCH 216/694] eth: fbnic: Handle maximum standalone channels MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Standalone channels use one NAPI vector for each Tx and Rx queue. fbnic's allocation path excludes FBNIC_MAX_TXQS from that layout. A 64-Tx/64-Rx configuration therefore records 128 vectors but allocates only 64, leaving NULL entries that resource setup dereferences. Include the maximum vector count in standalone allocation. Fixes: bc6107771bb4 ("eth: fbnic: Allocate a netdevice and napi vectors with queues") Signed-off-by: Björn Töpel Reviewed-by: Simon Horman Link: https://patch.msgid.link/178942020457.7700.13129750616387075931.stgit@ahduyck-xeon-server.home.arpa Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/meta/fbnic/fbnic_txrx.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/ethernet/meta/fbnic/fbnic_txrx.c b/drivers/net/ethernet/meta/fbnic/fbnic_txrx.c index 661dee1661af..a30aa4450848 100644 --- a/drivers/net/ethernet/meta/fbnic/fbnic_txrx.c +++ b/drivers/net/ethernet/meta/fbnic/fbnic_txrx.c @@ -1791,7 +1791,7 @@ int fbnic_alloc_napi_vectors(struct fbnic_net *fbn) int err; /* Allocate 1 Tx queue per napi vector */ - if (num_napi < FBNIC_MAX_TXQS && num_napi == num_tx + num_rx) { + if (num_napi <= FBNIC_MAX_TXQS && num_napi == num_tx + num_rx) { while (num_tx) { err = fbnic_alloc_napi_vector(fbd, fbn, num_napi, v_idx, From b5d9e9d4d0c13bc8b60d8d97e7a07fb25fea639e Mon Sep 17 00:00:00 2001 From: Alexander Duyck Date: Mon, 14 Sep 2026 14:10:11 -0700 Subject: [PATCH 217/694] eth: fbnic: use the Rx queue napi pointer to find the napi vector The queue management ndos pick the napi vector for an Rx queue with: nv = fbn->napi[idx % fbn->num_napi]; The issue is this is only correct in the cases where there are no standalone Tx vectors. In those cases we were allocating the Tx vectors first and then the Rx so the queues would be pointing to Tx NAPI vectors instead of the Rx ones. The mapping the ndos want is already recorded. fbnic_set_netif_napi() publishes it with netif_queue_set_napi(), which stores the napi pointer in netdev_rx_queue.napi, and fbnic_reset_netif_napi() clears it again. Both run under the netdev instance lock that the queue management ndos also hold, so the pointer can be read directly. Use it and drop the divide. The pointer is NULL exactly while the datapath is down, so fbnic_queue_mem_alloc() can reject that case rather than reaching into freed state: netdev_rx_queue_restart() calls it before it tests netif_running(), and fbnic_pm_suspend() leaves netif_running() true across a PCIe recovery that never completes, so a queue restart can arrive after fbnic_stop() has freed the rings and the vectors. fbnic_stop() clears the association in fbnic_reset_netif_queues() before fbnic_free_napi_vectors(), so the NULL is always published first. fbnic_queue_start() and fbnic_queue_stop() need no check of their own, as netdev_rx_queue_reconfig() only reaches them once fbnic_queue_mem_alloc() has succeeded under the same instance lock. Fixes: da43127a8edc ("eth: fbnic: support queue ops / zero-copy Rx") Signed-off-by: Alexander Duyck Reviewed-by: Simon Horman Link: https://patch.msgid.link/178942021136.7700.4391219358260544104.stgit@ahduyck-xeon-server.home.arpa Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/meta/fbnic/fbnic_txrx.c | 26 +++++++++++++++++--- 1 file changed, 23 insertions(+), 3 deletions(-) diff --git a/drivers/net/ethernet/meta/fbnic/fbnic_txrx.c b/drivers/net/ethernet/meta/fbnic/fbnic_txrx.c index a30aa4450848..10caacffee0f 100644 --- a/drivers/net/ethernet/meta/fbnic/fbnic_txrx.c +++ b/drivers/net/ethernet/meta/fbnic/fbnic_txrx.c @@ -7,6 +7,7 @@ #include #include #include +#include #include #include #include @@ -2853,6 +2854,17 @@ void fbnic_napi_depletion_check(struct net_device *netdev) fbnic_wrfl(fbd); } +/* Returns the napi vector servicing an Rx queue, or NULL if the datapath + * is torn down. The association is published by fbnic_set_netif_napi() + * and cleared by fbnic_reset_netif_napi(), both under the instance lock. + */ +static struct fbnic_napi_vector *fbnic_rxq_nv(struct net_device *dev, int idx) +{ + struct napi_struct *napi = __netif_get_rx_queue(dev, idx)->napi; + + return napi ? container_of(napi, struct fbnic_napi_vector, napi) : NULL; +} + static int fbnic_queue_mem_alloc(struct net_device *dev, struct netdev_queue_config *qcfg, void *qmem, int idx) @@ -2865,8 +2877,16 @@ static int fbnic_queue_mem_alloc(struct net_device *dev, if (!netif_running(dev)) return fbnic_alloc_qt_page_pools(fbn, qt, idx); + /* A failed PCIe recovery or resume can leave the datapath torn down + * while netif_running() is still true. This ndo runs before + * netdev_rx_queue_restart() checks netif_running(), so bail out + * rather than touching rings and vectors that are already freed. + */ + nv = fbnic_rxq_nv(dev, idx); + if (!nv) + return -ENETDOWN; + real = container_of(fbn->rx[idx], struct fbnic_q_triad, cmpl); - nv = fbn->napi[idx % fbn->num_napi]; fbnic_ring_init(&qt->sub0, real->sub0.doorbell, real->sub0.q_idx, real->sub0.flags); @@ -2917,7 +2937,7 @@ static int fbnic_queue_start(struct net_device *dev, struct fbnic_q_triad *real; real = container_of(fbn->rx[idx], struct fbnic_q_triad, cmpl); - nv = fbn->napi[idx % fbn->num_napi]; + nv = fbnic_rxq_nv(dev, idx); fbnic_aggregate_ring_bdq_counters(fbn, &real->sub0); fbnic_aggregate_ring_bdq_counters(fbn, &real->sub1); @@ -2939,7 +2959,7 @@ static int fbnic_queue_stop(struct net_device *dev, void *qmem, int idx) int err; real = container_of(fbn->rx[idx], struct fbnic_q_triad, cmpl); - nv = fbn->napi[idx % fbn->num_napi]; + nv = fbnic_rxq_nv(dev, idx); fbnic_dbg_nv_exit(nv); napi_disable_locked(&nv->napi); From 4bcc4a92c603fe7f062cea22e20da2e0ad6b12c3 Mon Sep 17 00:00:00 2001 From: Alexander Duyck Date: Mon, 14 Sep 2026 14:10:18 -0700 Subject: [PATCH 218/694] eth: fbnic: reset num_napi when the napi vectors are freed fbn->num_napi is the count of live napi vectors, each of which owns an IRQ. The PM path had freed them without clearing the count. fbnic_pm_suspend() tears the datapath down via ndo_stop() and frees the IRQs, but leaves netif_running() true so resume knows to re-open. Resume rebuilds the datapath in __fbnic_pm_resume() and fbnic_reset_queues() sets num_napi and __fbnic_open() re-allocates the vectors. When the datapath is torn down but never rebuilt, num_napi is left pointing at freed vectors under 2 different scenarios: - a PCIe error recovery that fails (fbnic_err_slot_reset() -> __fbnic_pm_resume() returns an error -> PCI_ERS_RESULT_DISCONNECT), so .resume never runs; or - an __fbnic_open() that fails partway on resume and unwinds, freeing the vectors after fbnic_reset_queues() has already set num_napi. The netdev is then running with num_napi > 0 but napi[] freed, and the eventual remove/unbind close re-enters fbnic_down() -> fbnic_dbg_down() and dereferences the freed vectors: BUG: kernel NULL pointer dereference, address: 0000000000000210 RIP: fbnic_dbg_down+0x28 Clear num_napi when the vectors are freed: in the suspend teardown (a good resume re-establishes it before __fbnic_open()) and on the resume open failure. A redundant ndo_stop() then walks an empty napi[]. The normal ndo_stop() down/up cycle is untouched and keeps num_napi for the next ndo_open(). Fixes: bc6107771bb4 ("eth: fbnic: Allocate a netdevice and napi vectors with queues") Signed-off-by: Alexander Duyck Reviewed-by: Simon Horman Link: https://patch.msgid.link/178942021809.7700.10804028989308077839.stgit@ahduyck-xeon-server.home.arpa Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/meta/fbnic/fbnic_pci.c | 18 ++++++++++++++---- 1 file changed, 14 insertions(+), 4 deletions(-) diff --git a/drivers/net/ethernet/meta/fbnic/fbnic_pci.c b/drivers/net/ethernet/meta/fbnic/fbnic_pci.c index 8b9bc9e8ea56..c6698e3002a1 100644 --- a/drivers/net/ethernet/meta/fbnic/fbnic_pci.c +++ b/drivers/net/ethernet/meta/fbnic/fbnic_pci.c @@ -434,6 +434,7 @@ static int fbnic_pm_suspend(struct device *dev) { struct fbnic_dev *fbd = dev_get_drvdata(dev); struct net_device *netdev = fbd->netdev; + struct fbnic_net *fbn; if (fbnic_init_failure(fbd)) goto null_uc_addr; @@ -441,11 +442,16 @@ static int fbnic_pm_suspend(struct device *dev) rtnl_lock(); netdev_lock(netdev); + fbn = netdev_priv(netdev); + netif_device_detach(netdev); if (netif_running(netdev)) netdev->netdev_ops->ndo_stop(netdev); + /* The IRQs are about to be freed, so drop the napi vector count */ + fbn->num_napi = 0; + netdev_unlock(netdev); rtnl_unlock(); @@ -508,16 +514,20 @@ static int __fbnic_pm_resume(struct device *dev) if (fbnic_init_failure(fbd)) return 0; + rtnl_lock(); + netdev_lock(netdev); + fbn = netdev_priv(netdev); /* Reset the queues if needed */ fbnic_reset_queues(fbn, fbn->num_tx_queues, fbn->num_rx_queues); - rtnl_lock(); - netdev_lock(netdev); - - if (netif_running(netdev)) + if (netif_running(netdev)) { err = __fbnic_open(fbn); + /* On failure the vectors are freed, so drop the count */ + if (err) + fbn->num_napi = 0; + } netdev_unlock(netdev); rtnl_unlock(); From 8947f13e436a4ff5eed9f8f019b2865a07af4bb2 Mon Sep 17 00:00:00 2001 From: Alexander Duyck Date: Mon, 14 Sep 2026 14:10:25 -0700 Subject: [PATCH 219/694] eth: fbnic: Set AW_FLUSH_MODE alongside AW_FLUSH when flushing the mailbox When tearing down the FW mailbox Rx ring, fbnic_mbx_reset_desc_ring() writes AW_CFG with FLUSH set and everything else, BME included, cleared. Clearing BME halts the device's writes to the host but leaves the staged requests parked in the PUL write pipeline rather than draining them, so on the write path FLUSH alone never terminates the outstanding requests and the flush the firmware waits on never completes. Add the FLUSH_MODE definition and set both bits so the staged writes drain out of the pipeline on their own. BME stays cleared, so nothing lands on the host; it is restored later in fbnic_mbx_init_desc_ring() when the ring is rebuilt, once the outstanding writes are gone. The read path is unaffected. AR_CFG has no equivalent mode bit and AR_FLUSH terminates the outstanding reads by itself, so it is left as is. Both writes remain plain stores rather than read-modify-writes. That is deliberate: the matching write in fbnic_mbx_init_desc_ring() restores BME and the TLP attributes, and clears both flush bits as a side effect. Fixes: 3b12f00ddd08 ("fbnic: Gate AXI read/write enabling on FW mailbox") Signed-off-by: Alexander Duyck Reviewed-by: Simon Horman Link: https://patch.msgid.link/178942022583.7700.11050671998277309744.stgit@ahduyck-xeon-server.home.arpa Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/meta/fbnic/fbnic_csr.h | 1 + drivers/net/ethernet/meta/fbnic/fbnic_fw.c | 9 ++++++++- 2 files changed, 9 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/meta/fbnic/fbnic_csr.h b/drivers/net/ethernet/meta/fbnic/fbnic_csr.h index 64b958df7774..14af30e189d6 100644 --- a/drivers/net/ethernet/meta/fbnic/fbnic_csr.h +++ b/drivers/net/ethernet/meta/fbnic/fbnic_csr.h @@ -974,6 +974,7 @@ enum { /* PUL User Registers */ #define FBNIC_CSR_START_PUL_USER 0x31000 /* CSR section delimiter */ #define FBNIC_PUL_OB_TLP_HDR_AW_CFG 0x3103d /* 0xc40f4 */ +#define FBNIC_PUL_OB_TLP_HDR_AW_CFG_FLUSH_MODE CSR_BIT(20) #define FBNIC_PUL_OB_TLP_HDR_AW_CFG_FLUSH CSR_BIT(19) #define FBNIC_PUL_OB_TLP_HDR_AW_CFG_BME CSR_BIT(18) #define FBNIC_PUL_OB_TLP_HDR_AW_CFG_RDE_ATTR CSR_GENMASK(17, 15) diff --git a/drivers/net/ethernet/meta/fbnic/fbnic_fw.c b/drivers/net/ethernet/meta/fbnic/fbnic_fw.c index 283d25fae79e..59aa879798b9 100644 --- a/drivers/net/ethernet/meta/fbnic/fbnic_fw.c +++ b/drivers/net/ethernet/meta/fbnic/fbnic_fw.c @@ -60,8 +60,15 @@ static void fbnic_mbx_reset_desc_ring(struct fbnic_dev *fbd, int mbx_idx) */ switch (mbx_idx) { case FBNIC_IPC_MBX_RX_IDX: + /* Clearing BME blocks the device from writing to the host + * but leaves the requests parked in the write pipeline. The + * write path only clears outstanding requests when both FLUSH + * and FLUSH_MODE are set; FLUSH_MODE lets them drain without + * landing on the host. + */ wr32(fbd, FBNIC_PUL_OB_TLP_HDR_AW_CFG, - FBNIC_PUL_OB_TLP_HDR_AW_CFG_FLUSH); + FBNIC_PUL_OB_TLP_HDR_AW_CFG_FLUSH | + FBNIC_PUL_OB_TLP_HDR_AW_CFG_FLUSH_MODE); break; case FBNIC_IPC_MBX_TX_IDX: wr32(fbd, FBNIC_PUL_OB_TLP_HDR_AR_CFG, From 1b97a269a5bdde20d4e69511f27649c9cb82b7c7 Mon Sep 17 00:00:00 2001 From: Alexander Duyck Date: Mon, 14 Sep 2026 14:10:33 -0700 Subject: [PATCH 220/694] eth: fbnic: Handle FW mailbox completions flagged with an error The firmware can complete a mailbox descriptor while also setting FW_ERR to indicate it could not process the request, for example on a mailbox DMA error. The completion carries no valid data. The driver did not check FW_ERR. On the Rx mailbox it would sync and parse the stale page as a normal message, and on the Tx mailbox it silently freed the request. If the initial capabilities exchange in fbnic_mbx_poll_tx_ready() hit FW_ERR -- on the Tx request or on the Rx response descriptor -- no response was parsed and the poll spun until it timed out even though the ring was healthy. Check FW_ERR on both mailboxes. Count it per-mailbox in fbnic_fw_mbx.resp_error, which is also shown in debugfs, warn (rate limited, since the bit is firmware controlled), and drop the Rx page instead of parsing it. In fbnic_mbx_poll_tx_ready() re-issue the capabilities request when either the Tx or the Rx resp_error counter advances, so a FW_ERR on the request or on its response triggers a retry rather than a timeout. A valid capabilities response is honored before the retry check, so a response parsed in the same poll as an unrelated FW_ERR is not discarded. The counters are mailbox-wide rather than keyed to the capabilities request; that is sufficient here because the exchange runs during bring-up before any other mailbox traffic, and any spurious retry is bounded by the existing 10s timeout. Fixes: da3cde08209e ("eth: fbnic: Add FW communication mechanism") Signed-off-by: Alexander Duyck Reviewed-by: Simon Horman Link: https://patch.msgid.link/178942023343.7700.9423398932961964439.stgit@ahduyck-xeon-server.home.arpa Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/meta/fbnic/fbnic_csr.h | 4 ++ .../net/ethernet/meta/fbnic/fbnic_debugfs.c | 4 +- drivers/net/ethernet/meta/fbnic/fbnic_fw.c | 38 ++++++++++++++++++- drivers/net/ethernet/meta/fbnic/fbnic_fw.h | 1 + 4 files changed, 44 insertions(+), 3 deletions(-) diff --git a/drivers/net/ethernet/meta/fbnic/fbnic_csr.h b/drivers/net/ethernet/meta/fbnic/fbnic_csr.h index 14af30e189d6..baba3471bf5a 100644 --- a/drivers/net/ethernet/meta/fbnic/fbnic_csr.h +++ b/drivers/net/ethernet/meta/fbnic/fbnic_csr.h @@ -1216,6 +1216,10 @@ enum { #define FBNIC_IPC_MBX_DESC_LEN_MASK DESC_GENMASK(63, 48) #define FBNIC_IPC_MBX_DESC_EOM DESC_BIT(46) #define FBNIC_IPC_MBX_DESC_ADDR_MASK DESC_GENMASK(45, 3) +/* Set with FW_CMPL when the FW completed a descriptor without successfully + * processing it (e.g. a mailbox DMA error); the completion has no valid data. + */ +#define FBNIC_IPC_MBX_DESC_FW_ERR DESC_BIT(2) #define FBNIC_IPC_MBX_DESC_FW_CMPL DESC_BIT(1) #define FBNIC_IPC_MBX_DESC_HOST_CMPL DESC_BIT(0) diff --git a/drivers/net/ethernet/meta/fbnic/fbnic_debugfs.c b/drivers/net/ethernet/meta/fbnic/fbnic_debugfs.c index 3c4563c8f403..6edfa0aa69f1 100644 --- a/drivers/net/ethernet/meta/fbnic/fbnic_debugfs.c +++ b/drivers/net/ethernet/meta/fbnic/fbnic_debugfs.c @@ -539,8 +539,8 @@ static void fbnic_dbg_fw_mbx_display(struct seq_file *s, /* Generate header */ seq_puts(s, mbx_idx == FBNIC_IPC_MBX_RX_IDX ? "Rx\n" : "Tx\n"); - seq_printf(s, "Rdy: %d Head: %d Tail: %d\n", - mbx->ready, mbx->head, mbx->tail); + seq_printf(s, "Rdy: %d Head: %d Tail: %d resp_error: %llu\n", + mbx->ready, mbx->head, mbx->tail, mbx->resp_error); snprintf(hdr, sizeof(hdr), "%3s %-4s %s %-12s %s %-3s %-16s\n", "Idx", "Len", "E", "Addr", "F", "H", "Raw"); diff --git a/drivers/net/ethernet/meta/fbnic/fbnic_fw.c b/drivers/net/ethernet/meta/fbnic/fbnic_fw.c index 59aa879798b9..6d7eb8479edf 100644 --- a/drivers/net/ethernet/meta/fbnic/fbnic_fw.c +++ b/drivers/net/ethernet/meta/fbnic/fbnic_fw.c @@ -292,6 +292,12 @@ static void fbnic_mbx_process_tx_msgs(struct fbnic_dev *fbd) if (!(desc & FBNIC_IPC_MBX_DESC_FW_CMPL)) break; + if (desc & FBNIC_IPC_MBX_DESC_FW_ERR) { + tx_mbx->resp_error++; + dev_warn_ratelimited(fbd->dev, + "FW completed a Tx mailbox request with an error\n"); + } + fbnic_mbx_unmap_and_free_msg(fbd, FBNIC_IPC_MBX_TX_IDX, head); head++; @@ -1673,6 +1679,13 @@ static void fbnic_mbx_process_rx_msgs(struct fbnic_dev *fbd) if (!(desc & FBNIC_IPC_MBX_DESC_FW_CMPL)) break; + if (desc & FBNIC_IPC_MBX_DESC_FW_ERR) { + rx_mbx->resp_error++; + dev_warn_ratelimited(fbd->dev, + "FW reported an error on an Rx mailbox message; dropping\n"); + goto next_page; + } + dma_sync_single_for_cpu(fbd->dev, rx_mbx->buf_info[head].addr, FBNIC_RX_PAGE_SIZE, DMA_FROM_DEVICE); @@ -1740,7 +1753,9 @@ void fbnic_mbx_poll(struct fbnic_dev *fbd) int fbnic_mbx_poll_tx_ready(struct fbnic_dev *fbd) { struct fbnic_fw_mbx *tx_mbx = &fbd->mbx[FBNIC_IPC_MBX_TX_IDX]; + struct fbnic_fw_mbx *rx_mbx = &fbd->mbx[FBNIC_IPC_MBX_RX_IDX]; unsigned long timeout = jiffies + 10 * HZ + 1; + u64 tx_resp_error, rx_resp_error; int err, i; do { @@ -1771,6 +1786,9 @@ int fbnic_mbx_poll_tx_ready(struct fbnic_dev *fbd) * mgmt.version once we get the actual version from the firmware * in the capabilities request message. */ +send_cap_req: + tx_resp_error = tx_mbx->resp_error; + rx_resp_error = rx_mbx->resp_error; err = fbnic_fw_xmit_simple_msg(fbd, FBNIC_TLV_MSG_ID_HOST_CAP_REQ); if (err) goto clean_mbx; @@ -1788,9 +1806,27 @@ int fbnic_mbx_poll_tx_ready(struct fbnic_dev *fbd) msleep(20); fbnic_mbx_poll(fbd); + /* A valid capabilities response ends the poll. Check it + * before the FW_ERR retry below so a response parsed in the + * same poll as an unrelated FW_ERR is not discarded. + */ + if (fbd->fw_cap.running.mgmt.version >= MIN_FW_VER_CODE) + break; + /* set err, but wait till mgmt.version check to report it */ - if (!time_is_after_jiffies(timeout)) + if (!time_is_after_jiffies(timeout)) { err = -ETIMEDOUT; + continue; + } + + /* The FW can flag our capabilities request (Tx) or its + * response (Rx) with FW_ERR, in which case it produced no + * usable response. The ring is not wedged, so re-issue the + * request instead of spinning until the timeout. + */ + if (tx_mbx->resp_error != tx_resp_error || + rx_mbx->resp_error != rx_resp_error) + goto send_cap_req; } return 0; diff --git a/drivers/net/ethernet/meta/fbnic/fbnic_fw.h b/drivers/net/ethernet/meta/fbnic/fbnic_fw.h index d84723e4cfa3..5f9969247e30 100644 --- a/drivers/net/ethernet/meta/fbnic/fbnic_fw.h +++ b/drivers/net/ethernet/meta/fbnic/fbnic_fw.h @@ -13,6 +13,7 @@ struct fbnic_tlv_msg; struct fbnic_fw_mbx { u8 ready, head, tail; + u64 resp_error; struct { struct fbnic_tlv_msg *msg; dma_addr_t addr; From d2c31b837406395e576afeb25958c98e9938f3f6 Mon Sep 17 00:00:00 2001 From: Yiqi Sun Date: Tue, 15 Sep 2026 17:50:17 +0800 Subject: [PATCH 221/694] sctp: avoid livelock while updating retransmit path sctp_assoc_update_retran_path() can loop forever when every remaining transport, including retran_path, is SCTP_UNCONFIRMED: the state check runs before the wraparound test, so the loop cannot observe that it has completed a full pass. Fix this by considering a transport only when it is not UNCONFIRMED, then checking whether the walk has returned to retran_path. This makes the full-pass termination independent of the transport state while preserving the existing fallback selection semantics. Also restore the NULL guard around the retran_path assignment. In the all-UNCONFIRMED case there is no eligible replacement transport, and installing NULL would leave later retransmit-path users and the debug print with a NULL path. Fixes: 4c47af4d5eb2 ("net: sctp: rework multihoming retransmission path selection to rfc4960") Signed-off-by: Yiqi Sun Acked-by: Xin Long Link: https://patch.msgid.link/20260915095017.942213-1-sunyiqixm@gmail.com Signed-off-by: Jakub Kicinski --- net/sctp/associola.c | 15 ++++++++------- 1 file changed, 8 insertions(+), 7 deletions(-) diff --git a/net/sctp/associola.c b/net/sctp/associola.c index c0512c827d0f..4521be3bd85a 100644 --- a/net/sctp/associola.c +++ b/net/sctp/associola.c @@ -1289,18 +1289,19 @@ void sctp_assoc_update_retran_path(struct sctp_association *asoc) /* Manually skip the head element. */ if (&trans->transports == &asoc->peer.transport_addr_list) continue; - if (trans->state == SCTP_UNCONFIRMED) - continue; - trans_next = sctp_trans_elect_best(trans, trans_next); - /* Active is good enough for immediate return. */ - if (trans_next->state == SCTP_ACTIVE) - break; + if (trans->state != SCTP_UNCONFIRMED) { + trans_next = sctp_trans_elect_best(trans, trans_next); + /* Active is good enough for immediate return. */ + if (trans_next->state == SCTP_ACTIVE) + break; + } /* We've reached the end, time to update path. */ if (trans == asoc->peer.retran_path) break; } - asoc->peer.retran_path = trans_next; + if (trans_next) + asoc->peer.retran_path = trans_next; pr_debug("%s: association:%p updated new path to addr:%pISpc\n", __func__, asoc, &asoc->peer.retran_path->ipaddr.sa); From 4581c3d2adc3c73a019bc38db64ca11f28bbd7fd Mon Sep 17 00:00:00 2001 From: Ralf Lici Date: Thu, 17 Sep 2026 14:27:23 +0200 Subject: [PATCH 222/694] net/mlx5e: advertise MACsec offload only when supported Commit 339ccec8d43d ("net/mlx5: Enable MACsec offload feature for VLAN interface") added NETIF_F_HW_MACSEC unconditionally to vlan_features so that VLAN devices could inherit MACsec offload support. mlx5e_build_nic_netdev subsequently copies vlan_features into hw_features and features. As a result, all mlx5e NIC netdevices advertise MACsec hardware offload, even when the firmware does not support it and the driver does not install macsec_ops. Set the MACsec feature bits in mlx5e_macsec_build_netdev, after device capabilities have been validated. This preserves MACsec-over-VLAN support and the ethtool feature control on capable devices, without advertising either on unsupported hardware. Fixes: 339ccec8d43d ("net/mlx5: Enable MACsec offload feature for VLAN interface") Cc: stable@vger.kernel.org Reviewed-by: Tariq Toukan Signed-off-by: Ralf Lici Link: https://patch.msgid.link/20260917122724.654639-1-ralf@mandelbit.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/mellanox/mlx5/core/en_accel/macsec.c | 2 ++ drivers/net/ethernet/mellanox/mlx5/core/en_main.c | 1 - 2 files changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/macsec.c b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/macsec.c index daff53ba7d09..38a3415acf7a 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/macsec.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/macsec.c @@ -1724,6 +1724,8 @@ void mlx5e_macsec_build_netdev(struct mlx5e_priv *priv) mlx5_core_dbg(priv->mdev, "mlx5e: MACsec acceleration enabled\n"); netdev->macsec_ops = &macsec_offload_ops; netdev->features |= NETIF_F_HW_MACSEC; + netdev->hw_features |= NETIF_F_HW_MACSEC; + netdev->vlan_features |= NETIF_F_HW_MACSEC; netif_keep_dst(netdev); } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c index fc110a7d16e8..4c6060d54fbe 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c @@ -5851,7 +5851,6 @@ static void mlx5e_build_nic_netdev(struct net_device *netdev) netdev->vlan_features |= NETIF_F_SG; netdev->vlan_features |= NETIF_F_HW_CSUM; - netdev->vlan_features |= NETIF_F_HW_MACSEC; netdev->vlan_features |= NETIF_F_GRO; netdev->vlan_features |= NETIF_F_TSO; netdev->vlan_features |= NETIF_F_TSO6; From 6c096bb08de97cdca051fecddad22cac6a1fd275 Mon Sep 17 00:00:00 2001 From: Quentin Armitage Date: Tue, 15 Sep 2026 22:33:21 +0100 Subject: [PATCH 223/694] net: allow IFLA_INET_CONF messages when NLA_F_NESTED unset Commit fa8fca88714c ("ipv4: validate IPV4_DEVCONF attributes properly") added validation of IFLA_INET_CONF attributes, and in the process changed the call of nla_for_each_nested() to nla_parse_nested(). A side effect of this change is that the IFLA_INET_CONF option is now tested for NLA_F_NESTED being set, and fails if it is not. Prior to the commit there was no check of NLA_F_NESTED. Change nla_parse_nested() to nla_parse(). This restores the previous functionality of not checking NLA_F_NESTED, thereby allowing code that (incorrectly) doesn't set NLA_F_NESTED to continue to work. This issue was identified because keepalived started logging errors when it was configuring macvlans that it created. Fixes: fa8fca88714c ("ipv4: validate IPV4_DEVCONF attributes properly") Signed-off-by: Quentin Armitage Reviewed-by: Ido Schimmel Link: https://patch.msgid.link/20260915213320.1527029-2-quentin@armitage.org.uk Signed-off-by: Jakub Kicinski --- net/ipv4/devinet.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/net/ipv4/devinet.c b/net/ipv4/devinet.c index a90be57c63be..5b6b11c943e4 100644 --- a/net/ipv4/devinet.c +++ b/net/ipv4/devinet.c @@ -2117,9 +2117,10 @@ static int inet_validate_link_af(const struct net_device *dev, return err; if (tb[IFLA_INET_CONF]) { - err = nla_parse_nested(nested_tb, IPV4_DEVCONF_MAX, - tb[IFLA_INET_CONF], inet_devconf_policy, - extack); + err = nla_parse(nested_tb, IPV4_DEVCONF_MAX, + nla_data(tb[IFLA_INET_CONF]), + nla_len(tb[IFLA_INET_CONF]), + inet_devconf_policy, extack); if (err < 0) return err; From 8901cee9316d53a5a97398f026c2d59a3043159d Mon Sep 17 00:00:00 2001 From: Kumar Kartikeya Dwivedi Date: Sat, 19 Sep 2026 03:42:10 +0200 Subject: [PATCH 224/694] bpf: Compare stack frames in regs_exact() regs_exact() compares the register state up to id, followed by the ID mappings, but does not compare frameno. The PTR_TO_STACK case in regsafe() checks frameno separately, which is bypassed when exact comparison is requested. Consequently, infinite-loop detection can treat pointers to different stack frames as the same pointer and reject a finite loop. For example, initialize fp-8 to zero in the caller and to one in the callee, then pass the caller's fp-8 to the callee as r1: loop: r0 = *(u64 *)(r1 + 0); if r0 != 0 goto done; r1 = r10; r1 += -8; goto loop; done: exit; The loop terminates after reading the callee's slot on its second iteration. At the loop header, however, the only relevant difference is r1's frameno, so exact comparison incorrectly reports an infinite loop. The same problem occurs when the pointer is spilled to the stack. Move frameno into the type-specific metadata union, ahead of id, so the existing prefix comparison in regs_exact() covers it. Ordinary stack pointers do not use another union member. Iterator and IRQ stack-slot states use their dedicated union views and do not need a frame lookup. This also keeps bpf_reg_state at 80 bytes. Since frameno now shares storage with other pointer metadata, it is only meaningful for PTR_TO_STACK registers. Return NULL from bpf_func() for other register types. process_iter_arg(), get_constant_map_key() and is_dynptr_reg_valid_init() look up the frame before checking the register type and would otherwise index frame[] with a byte of the register's map or BTF pointer. They dereference the frame only after their type check. Move the states_maybe_looping() boundary from frameno to precise after the field relocation. Its prefix comparison continues to cover the complete value state and now includes frameno. Continue to ignore precise. Precision marks control whether pruning may ignore scalar ranges; they do not change the represented values, and exact comparison already compares those ranges unconditionally. Marks can also change through backtracking while an ancestor state is still being explored. Fixes: d5b892fd607a ("bpf: make infinite loop detection in is_state_visited() exact") Reported-by: Eduard Zingerman Signed-off-by: Kumar Kartikeya Dwivedi Signed-off-by: Alexei Starovoitov Acked-by: Eduard Zingerman Link: https://patch.msgid.link/20260919014213.1840880-2-memxor@gmail.com --- include/linux/bpf_verifier.h | 23 +++++++++++++++-------- kernel/bpf/states.c | 7 ++----- 2 files changed, 17 insertions(+), 13 deletions(-) diff --git a/include/linux/bpf_verifier.h b/include/linux/bpf_verifier.h index a7202b44ab10..6fe8e5dc57aa 100644 --- a/include/linux/bpf_verifier.h +++ b/include/linux/bpf_verifier.h @@ -45,6 +45,14 @@ struct bpf_reg_state { union { /* valid when type == PTR_TO_PACKET */ int range; + /* + * Valid when type == PTR_TO_STACK. Inside the callee two registers + * can be both PTR_TO_STACK like R1=fp-8 and R2=fp-8, but one of them + * points to this function stack while another to the caller's stack. + * To differentiate them 'frameno' is used which is an index in + * bpf_verifier_state->frame[] array pointing to bpf_func_state. + */ + u8 frameno; /* * For CONST_PTR_TO_MAP, PTR_TO_MAP_KEY, PTR_TO_MAP_VALUE and @@ -155,14 +163,6 @@ struct bpf_reg_state { * during state comparisons. */ u32 map_uid; - /* - * Inside the callee two registers can be both PTR_TO_STACK like - * R1=fp-8 and R2=fp-8, but one of them points to this function stack - * while another to the caller's stack. To differentiate them 'frameno' - * is used which is an index in bpf_verifier_state->frame[] array - * pointing to bpf_func_state. - */ - u8 frameno; /* if (!precise && SCALAR_VALUE) min/max/tnum don't affect safety */ bool precise; }; @@ -1239,11 +1239,18 @@ static inline int bpf_get_spi(s32 off) return (-off - 1) / BPF_REG_SIZE; } +/* + * Return the function state a stack pointer register refers to. frameno + * shares storage with other pointer metadata, so return NULL for any + * other register type instead of indexing frame[] with aliased bytes. + */ static inline struct bpf_func_state *bpf_func(struct bpf_verifier_env *env, const struct bpf_reg_state *reg) { struct bpf_verifier_state *cur = env->cur_state; + if (reg->type != PTR_TO_STACK) + return NULL; return cur->frame[reg->frameno]; } diff --git a/kernel/bpf/states.c b/kernel/bpf/states.c index e4ec007f7fa6..012b82513a3b 100644 --- a/kernel/bpf/states.c +++ b/kernel/bpf/states.c @@ -644,10 +644,7 @@ static bool regsafe(struct bpf_verifier_env *env, struct bpf_reg_state *rold, return range_within(rold, rcur) && tnum_in(rold->var_off, rcur->var_off); case PTR_TO_STACK: - /* two stack pointers are equal only if they're pointing to - * the same stack frame, since fp-8 in foo != fp-8 in bar - */ - return regs_exact(rold, rcur, idmap) && rold->frameno == rcur->frameno; + return regs_exact(rold, rcur, idmap); case PTR_TO_ARENA: return true; case PTR_TO_INSN: @@ -1126,7 +1123,7 @@ static bool states_maybe_looping(struct bpf_verifier_state *old, fcur = cur->frame[fr]; for (i = 0; i < MAX_BPF_REG; i++) if (memcmp(&fold->regs[i], &fcur->regs[i], - offsetof(struct bpf_reg_state, frameno))) + offsetof(struct bpf_reg_state, precise))) return false; return true; } From 070587848985efcfcd45104c5266ba76a1165f55 Mon Sep 17 00:00:00 2001 From: Kumar Kartikeya Dwivedi Date: Sat, 19 Sep 2026 03:42:11 +0200 Subject: [PATCH 225/694] selftests/bpf: Cover frame changes in bounded loops Add a finite loop whose progress is represented only by changing the frame number of a stack pointer. The loop first reads zero from the caller's stack, switches to the same offset in the callee's stack, and exits after reading one on its next iteration. Force frequent checkpoints so the test exercises infinite-loop detection, and check that the program returns one when run. Without the frameno comparison in regs_exact(), the program is rejected with an "infinite loop detected" diagnostic instead of loading successfully. Signed-off-by: Kumar Kartikeya Dwivedi Signed-off-by: Alexei Starovoitov Tested-by: Eduard Zingerman Link: https://patch.msgid.link/20260919014213.1840880-3-memxor@gmail.com --- .../selftests/bpf/progs/verifier_loops1.c | 36 +++++++++++++++++++ 1 file changed, 36 insertions(+) diff --git a/tools/testing/selftests/bpf/progs/verifier_loops1.c b/tools/testing/selftests/bpf/progs/verifier_loops1.c index d248ce877f14..48a966cda199 100644 --- a/tools/testing/selftests/bpf/progs/verifier_loops1.c +++ b/tools/testing/selftests/bpf/progs/verifier_loops1.c @@ -303,4 +303,40 @@ __naked void maybe_exit_scc_bug1(void) ::: __clobber_all); } +/* + * The loop reads zero from the caller's stack on its first iteration and + * one from the callee's stack on its second iteration. At the loop header, + * only the frame number of the pointer in r1 changes. + */ +static __naked __noinline __used +void loop_stack_frames_reg(void) +{ + asm volatile ( + "*(u64 *)(r10 - 8) = 1;" +"1:" + "r0 = *(u64 *)(r1 + 0);" + "if r0 != 0 goto 2f;" + "r1 = r10;" + "r1 += -8;" + "goto 1b;" +"2:" + "exit;" + ::: __clobber_all); +} + +SEC("xdp") +__description("bounded loop changing stack frame in a register") +__success __retval(1) +__flag(BPF_F_TEST_STATE_FREQ) +__naked void bounded_loop_stack_frames_reg(void) +{ + asm volatile ( + "*(u64 *)(r10 - 8) = 0;" + "r1 = r10;" + "r1 += -8;" + "call loop_stack_frames_reg;" + "exit;" + ::: __clobber_all); +} + char _license[] SEC("license") = "GPL"; From bfc888f04588f591851e95c974954cfca58e6c19 Mon Sep 17 00:00:00 2001 From: Kumar Kartikeya Dwivedi Date: Mon, 14 Sep 2026 15:24:42 +0200 Subject: [PATCH 226/694] bpf: Bound ownership depth through local kptrs and graph roots Program-allocated objects can own other local objects through referenced kptrs. bpf_obj_free_fields() follows those pointers through __bpf_obj_drop_impl() synchronously, before the object storage is freed through RCU. A self-referential local kptr type therefore permits arbitrarily deep object chains, and dropping the head can exhaust the kernel stack. Long acyclic type chains have the same problem. btf_check_and_fixup_fields() still assumes referenced kptrs only point to kernel types and checks ownership through list and rbtree roots only. Its existing rule is sufficient for graph-only cycles: the target of each graph edge must contain a node, so every type in a cycle has both a root and a node. The rule rejects such a type owning another root, breaking every cycle. It also limits graph-only chains to three types, or two if the first type contains a node, and conservatively rejects longer acyclic chains. The missing local-kptr edges, rather than a missed graph-only cycle, are the bug introduced by support for bpf_kptr_xchg() into local kptrs. Replace that restriction with one bounded ownership walk covering graph roots and local referenced kptrs. Run it after all BTF records have been fixed up, reject cycles and paths deeper than eight record-bearing types, and cache each type's suffix depth while checking it against the remaining budget. This also permits the longer acyclic graph-only layouts rejected by the old rule; update their existing BTF tests accordingly. Keep the bound independent of MAX_CALL_FRAMES because recursive destruction can run below a BPF call chain. A plain local pointee without special-field metadata adds only a final non-recursing drop. Non-owning kptrs and kernel-BTF kptrs do not recurse through local records and remain outside the walk. Include local percpu-kptr edges too, although allocation of percpu objects with special fields is currently forbidden, so that relaxing that restriction cannot bypass the ownership bound. btf_check_and_fixup_fields() continues to initialize graph_root.value_rec, including for separately allocated map records. The ownership relationships belong to immutable program BTF and only need validation at BTF load time. Fixes: b0966c724584 ("bpf: Support bpf_kptr_xchg into local kptr") Reported-by: Nicholas Carlini Suggested-by: Nicholas Carlini Signed-off-by: Kumar Kartikeya Dwivedi Signed-off-by: Alexei Starovoitov Link: https://patch.msgid.link/20260914132444.2564218-2-memxor@gmail.com --- kernel/bpf/btf.c | 138 +++++++++++------- .../selftests/bpf/prog_tests/linked_list.c | 4 +- 2 files changed, 90 insertions(+), 52 deletions(-) diff --git a/kernel/bpf/btf.c b/kernel/bpf/btf.c index d67a169ba2bb..d870bc5e50bc 100644 --- a/kernel/bpf/btf.c +++ b/kernel/bpf/btf.c @@ -4268,13 +4268,10 @@ int btf_check_and_fixup_fields(const struct btf *btf, struct btf_record *rec) { int i; - /* There are three types that signify ownership of some other type: - * kptr_ref, bpf_list_head, bpf_rb_root. - * kptr_ref only supports storing kernel types, which can't store - * references to program allocated local types. - * - * Hence we only need to ensure that bpf_{list_head,rb_root} ownership - * does not form cycles. + /* + * Check fields which require the complete BTF and initialize runtime + * metadata. Ownership relationships are validated after every record has + * been fixed up. */ if (IS_ERR_OR_NULL(rec) || !(rec->field_mask & (BPF_GRAPH_ROOT | BPF_UPTR))) return 0; @@ -4305,53 +4302,90 @@ int btf_check_and_fixup_fields(const struct btf *btf, struct btf_record *rec) if (!meta) return -EFAULT; rec->fields[i].graph_root.value_rec = meta->record; - - /* We need to set value_rec for all root types, but no need - * to check ownership cycle for a type unless it's also a - * node type. - */ - if (!(rec->field_mask & BPF_GRAPH_NODE)) - continue; - - /* We need to ensure ownership acyclicity among all types. The - * proper way to do it would be to topologically sort all BTF - * IDs based on the ownership edges, since there can be multiple - * bpf_{list_head,rb_node} in a type. Instead, we use the - * following resaoning: - * - * - A type can only be owned by another type in user BTF if it - * has a bpf_{list,rb}_node. Let's call these node types. - * - A type can only _own_ another type in user BTF if it has a - * bpf_{list_head,rb_root}. Let's call these root types. - * - * We ensure that if a type is both a root and node, its - * element types cannot be root types. - * - * To ensure acyclicity: - * - * When A is an root type but not a node, its ownership - * chain can be: - * A -> B -> C - * Where: - * - A is an root, e.g. has bpf_rb_root. - * - B is both a root and node, e.g. has bpf_rb_node and - * bpf_list_head. - * - C is only an root, e.g. has bpf_list_node - * - * When A is both a root and node, some other type already - * owns it in the BTF domain, hence it can not own - * another root type through any of the ownership edges. - * A -> B - * Where: - * - A is both an root and node. - * - B is only an node. - */ - if (meta->record->field_mask & BPF_GRAPH_ROOT) - return -ELOOP; } return 0; } +static int btf_owned_type_idx(const struct btf *btf, struct btf_struct_metas *tab, + const struct btf_field *field) +{ + struct btf_struct_meta *meta; + u32 btf_id; + + if (field->type & BPF_GRAPH_ROOT) { + btf_id = field->graph_root.value_btf_id; + } else if (field->type == BPF_KPTR_REF || field->type == BPF_KPTR_PERCPU) { + if (btf_is_kernel(field->kptr.btf)) + return -ENOENT; + btf_id = field->kptr.btf_id; + } else { + return -ENOENT; + } + + meta = btf_find_struct_meta(btf, btf_id); + if (!meta) + return field->type & BPF_GRAPH_ROOT ? -EFAULT : -ENOENT; + return meta - tab->types; +} + +/* + * Each ownership edge adds kernel frames through bpf_obj_free_fields() and + * __bpf_obj_drop_impl(). Keep the bound deliberately small because object + * destruction can itself run below a BPF call chain. A final pointee without + * special fields is not present in the struct metadata table and adds only a + * non-recursing drop. + */ +#define BTF_MAX_OWNERSHIP_DEPTH 8 + +static int btf_ownership_depth(const struct btf *btf, + struct btf_struct_metas *tab, u8 *depth, + int idx, int depth_left) +{ + const struct btf_record *rec = tab->types[idx].record; + int i, ret, max_depth = 0; + + if (!depth_left) + return -ELOOP; + if (depth[idx]) + goto done; + + for (i = 0; i < rec->cnt; i++) { + ret = btf_owned_type_idx(btf, tab, &rec->fields[i]); + if (ret == -ENOENT) + continue; + if (ret < 0) + return ret; + ret = btf_ownership_depth(btf, tab, depth, ret, depth_left - 1); + if (ret < 0) + return ret; + max_depth = max(max_depth, ret); + } + depth[idx] = max_depth + 1; +done: + return depth[idx] > depth_left ? -ELOOP : depth[idx]; +} + +static int btf_check_ownership_depth(const struct btf *btf, + struct btf_struct_metas *tab) +{ + u8 *depth; + int i, ret = 0; + + depth = kvcalloc(tab->cnt, sizeof(*depth), GFP_KERNEL | __GFP_NOWARN); + if (!depth) + return -ENOMEM; + + for (i = 0; i < tab->cnt; i++) { + ret = btf_ownership_depth(btf, tab, depth, i, + BTF_MAX_OWNERSHIP_DEPTH); + if (ret < 0) + break; + ret = 0; + } + kvfree(depth); + return ret; +} + static void __btf_struct_show(const struct btf *btf, const struct btf_type *t, u32 type_id, void *data, u8 bits_offset, struct btf_show *show) @@ -6044,6 +6078,10 @@ static struct btf *btf_parse(const union bpf_attr *attr, bpfptr_t uattr, if (err < 0) goto errout_meta; } + + err = btf_check_ownership_depth(btf, struct_meta_tab); + if (err < 0) + goto errout_meta; } err = bpf_log_attr_finalize(attr_log, &env->log); diff --git a/tools/testing/selftests/bpf/prog_tests/linked_list.c b/tools/testing/selftests/bpf/prog_tests/linked_list.c index c3d133c6a00d..52fabbee3dd5 100644 --- a/tools/testing/selftests/bpf/prog_tests/linked_list.c +++ b/tools/testing/selftests/bpf/prog_tests/linked_list.c @@ -714,7 +714,7 @@ static void test_btf(void) break; err = btf__load_into_kernel(btf); - ASSERT_EQ(err, -ELOOP, "check btf"); + ASSERT_EQ(err, 0, "check btf"); btf__free(btf); break; } @@ -773,7 +773,7 @@ static void test_btf(void) break; err = btf__load_into_kernel(btf); - ASSERT_EQ(err, -ELOOP, "check btf"); + ASSERT_EQ(err, 0, "check btf"); btf__free(btf); break; } From 0288ed67482b6e370eb3fa6b07720df435907970 Mon Sep 17 00:00:00 2001 From: Kumar Kartikeya Dwivedi Date: Mon, 14 Sep 2026 15:24:43 +0200 Subject: [PATCH 227/694] selftests/bpf: Check local object ownership depth Build raw program BTF records to exercise local object ownership without constructing a runtime chain deep enough to threaten the kernel stack. Cover referenced-kptr and percpu-kptr self-cycles, a two-type cycle, and a cycle mixing a graph root with a referenced kptr. The existing graph-only check accepts these local-kptr cycles and over-limit kptr chains; the fix rejects them with -ELOOP. Pin the eight-record depth boundary with a terminal plain object. Exercise both parent-first and child-first BTF orders, and a shared suffix reached first through a shorter path. These cases require cached suffix depths to be checked against the remaining depth budget on each path. Check list and rbtree chains of three, four, eight, and nine types. This covers the old graph-only depth boundary and the new explicit bound. The existing linked-list BTF tests still reject pure graph cycles and now accept the longer acyclic layouts previously rejected by the conservative rule. Although bpf_percpu_obj_new() currently rejects types with special fields, require the percpu cycle to fail at BTF load so future support cannot bypass the ownership bound. Keep a positive control for a non-owning kptr, which remains outside the ownership graph. Signed-off-by: Kumar Kartikeya Dwivedi Signed-off-by: Alexei Starovoitov Link: https://patch.msgid.link/20260914132444.2564218-3-memxor@gmail.com --- .../bpf/prog_tests/local_kptr_ownership.c | 296 ++++++++++++++++++ 1 file changed, 296 insertions(+) create mode 100644 tools/testing/selftests/bpf/prog_tests/local_kptr_ownership.c diff --git a/tools/testing/selftests/bpf/prog_tests/local_kptr_ownership.c b/tools/testing/selftests/bpf/prog_tests/local_kptr_ownership.c new file mode 100644 index 000000000000..a487aa68f2ee --- /dev/null +++ b/tools/testing/selftests/bpf/prog_tests/local_kptr_ownership.c @@ -0,0 +1,296 @@ +// SPDX-License-Identifier: GPL-2.0 +/* Copyright (c) 2026 Meta Platforms, Inc. and affiliates. */ + +#include +#include +#include + +#define SPIN_LOCK 2 +#define LIST_HEAD 3 +#define LIST_NODE 4 +/* Keep in sync with BTF_MAX_OWNERSHIP_DEPTH. */ +#define MAX_OWNERSHIP_DEPTH 8 + +static struct btf *init_btf(void) +{ + struct btf *btf; + int id; + + btf = btf__new_empty(); + if (!ASSERT_OK_PTR(btf, "btf__new_empty")) + return NULL; + id = btf__add_int(btf, "int", 4, BTF_INT_SIGNED); + if (!ASSERT_EQ(id, 1, "btf__add_int")) + goto err_out; + id = btf__add_struct(btf, "bpf_spin_lock", 4); + if (!ASSERT_EQ(id, SPIN_LOCK, "btf__add_struct bpf_spin_lock")) + goto err_out; + id = btf__add_struct(btf, "bpf_list_head", 16); + if (!ASSERT_EQ(id, LIST_HEAD, "btf__add_struct bpf_list_head")) + goto err_out; + id = btf__add_struct(btf, "bpf_list_node", 24); + if (!ASSERT_EQ(id, LIST_NODE, "btf__add_struct bpf_list_node")) + goto err_out; + return btf; + +err_out: + btf__free(btf); + return NULL; +} + +static int add_local_kptr(struct btf *btf, int pointee_id, const char *tag) +{ + int id; + + id = btf__add_type_tag(btf, tag, pointee_id); + if (!ASSERT_GT(id, 0, "btf__add_type_tag")) + return id; + id = btf__add_ptr(btf, id); + ASSERT_GT(id, 0, "btf__add_ptr"); + return id; +} + +static void test_self_cycle(const char *tag, int expected_err) +{ + struct btf *btf; + int id, err; + + btf = init_btf(); + if (!ASSERT_OK_PTR(btf, "init_btf")) + return; + id = add_local_kptr(btf, 7, tag); + if (id <= 0) + goto out; + id = btf__add_struct(btf, "self_cycle", 8); + if (!ASSERT_EQ(id, 7, "btf__add_struct self_cycle")) + goto out; + err = btf__add_field(btf, "next", 6, 0, 0); + if (!ASSERT_OK(err, "btf__add_field self_cycle::next")) + goto out; + + err = btf__load_into_kernel(btf); + ASSERT_EQ(err, expected_err, "check btf"); +out: + btf__free(btf); +} + +static void test_aba_cycle(void) +{ + struct btf *btf; + int id, err; + + btf = init_btf(); + if (!ASSERT_OK_PTR(btf, "init_btf")) + return; + id = add_local_kptr(btf, 10, "kptr"); + if (id <= 0) + goto out; + id = add_local_kptr(btf, 9, "kptr"); + if (id <= 0) + goto out; + id = btf__add_struct(btf, "cycle_a", 8); + if (!ASSERT_EQ(id, 9, "btf__add_struct cycle_a")) + goto out; + err = btf__add_field(btf, "b", 6, 0, 0); + if (!ASSERT_OK(err, "btf__add_field cycle_a::b")) + goto out; + id = btf__add_struct(btf, "cycle_b", 8); + if (!ASSERT_EQ(id, 10, "btf__add_struct cycle_b")) + goto out; + err = btf__add_field(btf, "a", 8, 0, 0); + if (!ASSERT_OK(err, "btf__add_field cycle_b::a")) + goto out; + + err = btf__load_into_kernel(btf); + ASSERT_EQ(err, -ELOOP, "check btf"); +out: + btf__free(btf); +} + +static void test_mixed_cycle(void) +{ + struct btf *btf; + int id, err; + + btf = init_btf(); + if (!ASSERT_OK_PTR(btf, "init_btf")) + return; + id = add_local_kptr(btf, 7, "kptr"); + if (id <= 0) + goto out; + id = btf__add_struct(btf, "mixed_owner", 20); + if (!ASSERT_EQ(id, 7, "btf__add_struct mixed_owner")) + goto out; + err = btf__add_field(btf, "root", LIST_HEAD, 0, 0); + if (!ASSERT_OK(err, "btf__add_field mixed_owner::root")) + goto out; + err = btf__add_field(btf, "lock", SPIN_LOCK, 128, 0); + if (!ASSERT_OK(err, "btf__add_field mixed_owner::lock")) + goto out; + id = btf__add_decl_tag(btf, "contains:mixed_node:node", 7, 0); + if (!ASSERT_EQ(id, 8, "btf__add_decl_tag mixed_owner")) + goto out; + id = btf__add_struct(btf, "mixed_node", 32); + if (!ASSERT_EQ(id, 9, "btf__add_struct mixed_node")) + goto out; + err = btf__add_field(btf, "node", LIST_NODE, 0, 0); + if (!ASSERT_OK(err, "btf__add_field mixed_node::node")) + goto out; + err = btf__add_field(btf, "owner", 6, 192, 0); + if (!ASSERT_OK(err, "btf__add_field mixed_node::owner")) + goto out; + + err = btf__load_into_kernel(btf); + ASSERT_EQ(err, -ELOOP, "check btf"); +out: + btf__free(btf); +} + +static void test_acyclic_depth(int depth, bool child_first, bool shared_suffix, int expected_err) +{ + int ptr_id[MAX_OWNERSHIP_DEPTH + 1]; + int first_struct_id; + struct btf *btf; + int id, err, i, n, pointee_id; + + btf = init_btf(); + if (!ASSERT_OK_PTR(btf, "init_btf")) + return; + first_struct_id = 5 + 2 * depth; + for (i = 0; i < depth; i++) { + if (i == depth - 1) + pointee_id = first_struct_id + depth; + else + pointee_id = first_struct_id + (child_first ? depth - 2 - i : i + 1); + ptr_id[i] = add_local_kptr(btf, pointee_id, "kptr"); + if (ptr_id[i] <= 0) + goto out; + } + for (n = 0; n < depth; n++) { + char name[32]; + int offset = 0; + + i = child_first ? depth - 1 - n : n; + snprintf(name, sizeof(name), "owner_%d", i); + id = btf__add_struct(btf, name, shared_suffix && !i ? 16 : 8); + if (!ASSERT_EQ(id, first_struct_id + n, "btf__add_struct owner")) + goto out; + if (shared_suffix && !i) { + /* + * Visit the shared suffix through the shorter path before + * reaching it again with less remaining depth. + */ + err = btf__add_field(btf, "suffix", ptr_id[1], 0, 0); + if (!ASSERT_OK(err, "btf__add_field owner::suffix")) + goto out; + offset = 64; + } + err = btf__add_field(btf, "next", ptr_id[i], offset, 0); + if (!ASSERT_OK(err, "btf__add_field owner::next")) + goto out; + } + id = btf__add_struct(btf, "plain_leaf", 4); + if (!ASSERT_EQ(id, first_struct_id + depth, "btf__add_struct plain_leaf")) + goto out; + + err = btf__load_into_kernel(btf); + ASSERT_EQ(err, expected_err, "check btf"); +out: + btf__free(btf); +} + +static void test_graph_depth(bool rbtree, int depth, int expected_err) +{ + int root_type = LIST_HEAD, node_type = LIST_NODE, node_size = 24; + int id, err, i, lock_off, root_off, size; + struct btf *btf; + + btf = init_btf(); + if (!ASSERT_OK_PTR(btf, "init_btf")) + return; + if (rbtree) { + root_type = btf__add_struct(btf, "bpf_rb_root", 16); + if (!ASSERT_GT(root_type, 0, "btf__add_struct bpf_rb_root")) + goto out; + node_type = btf__add_struct(btf, "bpf_rb_node", 32); + if (!ASSERT_GT(node_type, 0, "btf__add_struct bpf_rb_node")) + goto out; + node_size = 32; + } + + for (i = 0; i < depth; i++) { + char name[32], tag[64]; + + lock_off = i ? node_size : 0; + root_off = lock_off + 8; + size = i == depth - 1 ? node_size : root_off + 16; + snprintf(name, sizeof(name), "graph_owner_%d", i); + id = btf__add_struct(btf, name, size); + if (!ASSERT_GT(id, 0, "btf__add_struct graph_owner")) + goto out; + if (i) { + err = btf__add_field(btf, "node", node_type, 0, 0); + if (!ASSERT_OK(err, "btf__add_field graph_owner::node")) + goto out; + } + if (i == depth - 1) + continue; + err = btf__add_field(btf, "lock", SPIN_LOCK, lock_off * 8, 0); + if (!ASSERT_OK(err, "btf__add_field graph_owner::lock")) + goto out; + err = btf__add_field(btf, "root", root_type, root_off * 8, 0); + if (!ASSERT_OK(err, "btf__add_field graph_owner::root")) + goto out; + snprintf(tag, sizeof(tag), "contains:graph_owner_%d:node", i + 1); + err = btf__add_decl_tag(btf, tag, id, i ? 2 : 1); + if (!ASSERT_GT(err, 0, "btf__add_decl_tag graph_owner")) + goto out; + } + + err = btf__load_into_kernel(btf); + ASSERT_EQ(err, expected_err, "check btf"); +out: + btf__free(btf); +} + +void test_local_kptr_ownership(void) +{ + if (test__start_subtest("self_cycle")) + test_self_cycle("kptr", -ELOOP); + if (test__start_subtest("untrusted_self_cycle")) + test_self_cycle("kptr_untrusted", 0); + if (test__start_subtest("percpu_self_cycle")) + test_self_cycle("percpu_kptr", -ELOOP); + if (test__start_subtest("ABA_cycle")) + test_aba_cycle(); + if (test__start_subtest("mixed_graph_root_cycle")) + test_mixed_cycle(); + if (test__start_subtest("max_acyclic")) + test_acyclic_depth(MAX_OWNERSHIP_DEPTH, false, false, 0); + if (test__start_subtest("too_deep_acyclic")) + test_acyclic_depth(MAX_OWNERSHIP_DEPTH + 1, false, false, -ELOOP); + if (test__start_subtest("max_acyclic_child_first")) + test_acyclic_depth(MAX_OWNERSHIP_DEPTH, true, false, 0); + if (test__start_subtest("too_deep_acyclic_child_first")) + test_acyclic_depth(MAX_OWNERSHIP_DEPTH + 1, true, false, -ELOOP); + if (test__start_subtest("max_acyclic_shared_suffix")) + test_acyclic_depth(MAX_OWNERSHIP_DEPTH, false, true, 0); + if (test__start_subtest("too_deep_acyclic_shared_suffix")) + test_acyclic_depth(MAX_OWNERSHIP_DEPTH + 1, false, true, -ELOOP); + if (test__start_subtest("list_three_types")) + test_graph_depth(false, 3, 0); + if (test__start_subtest("list_four_types")) + test_graph_depth(false, 4, 0); + if (test__start_subtest("list_max_depth")) + test_graph_depth(false, MAX_OWNERSHIP_DEPTH, 0); + if (test__start_subtest("list_too_deep")) + test_graph_depth(false, MAX_OWNERSHIP_DEPTH + 1, -ELOOP); + if (test__start_subtest("rbtree_three_types")) + test_graph_depth(true, 3, 0); + if (test__start_subtest("rbtree_four_types")) + test_graph_depth(true, 4, 0); + if (test__start_subtest("rbtree_max_depth")) + test_graph_depth(true, MAX_OWNERSHIP_DEPTH, 0); + if (test__start_subtest("rbtree_too_deep")) + test_graph_depth(true, MAX_OWNERSHIP_DEPTH + 1, -ELOOP); +} From 3bd46666cfe51e2bd333fb46a0234694ca594230 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Fri, 18 Sep 2026 14:28:37 -1000 Subject: [PATCH 228/694] sched_ext: Pass the initial cmask to cid-form ops.enable() The cid-form API has an obvious hole. A task's cid mask is only visible through ops.set_cmask(), which fires on affinity changes and class switches but not when a task enters a scheduler through fork, sub-sched enable or re-home, and there is no p->cpus_ptr equivalent to fall back on. Schedulers work around it by seeding the mask in ops.init_task() from p->cpus_ptr cid by cid, which is subtly wrong: on sub-sched enable and re-home, an affinity change between init_task() and enable() is delivered to the sched the task is still on, and nothing corrects the new sched's copy afterwards. Fix it by adding struct scx_enable_args to cid-form ops.enable() carrying the task's cmask, built in the per-cpu scratch under the rq lock as the task enters the scheduler, and calling set_cmask() with the same mask right after enable(), ahead of set_weight(). A scheduler can then track affinity in set_cmask() alone, and scx_qmap drops its init_task() seed. set_cmask() no longer fires for a cid-form task before it is enabled, and the class-switch republish in switching_to_scx() is limited to the cpu form. This changes the cid-form ops.enable() signature, which is fine as the cid-form API is still considered unreleased. An args struct rather than a bare cmask argument leaves room for more initial state without another signature change, and the cmask travels as a plain arena address because BTF can't type arena struct members yet. v2: The cmask travels as a u64 arena address, cmask_arena_addr, instead of a kernel-typed pointer, with the typing limitation and the planned typed alias documented (Sashiko review). v3: The initial set_cmask() is delivered before set_weight() so that the mask is in place when weight-dependent state is derived (Andrea Righi). Selftest added. Signed-off-by: Tejun Heo Reviewed-by: Andrea Righi --- kernel/sched/ext/ext.c | 92 +++++++++++++++++++++++----------- kernel/sched/ext/internal.h | 56 ++++++++++++++++++--- tools/sched_ext/scx_qmap.bpf.c | 3 -- 3 files changed, 110 insertions(+), 41 deletions(-) diff --git a/kernel/sched/ext/ext.c b/kernel/sched/ext/ext.c index f568fd9973f6..ad391a8cbd05 100644 --- a/kernel/sched/ext/ext.c +++ b/kernel/sched/ext/ext.c @@ -447,37 +447,45 @@ static void switch_rq_lock(struct rq *from, struct rq *to) DEFINE_STATIC_KEY_FALSE(__scx_is_cid_type); /** - * scx_call_op_set_cpumask - invoke ops.set_cpumask / ops_cid.set_cmask for @task + * scx_fill_cmask_scratch - Build this cpu's arena cmask from @cpumask + * @sch: scx_sched whose scratch to fill + * @cpumask: cpus to translate into cids + * + * The scratch lives in BPF-writable arena memory and its header can't be + * trusted, so it is rewritten from kernel geometry rather than read. Caller + * must hold an rq lock so this cpu is the sole kernel writer for as long as the + * returned address is in use. + */ +static struct scx_cmask *scx_fill_cmask_scratch(struct scx_sched *sch, + const struct cpumask *cpumask) +{ + struct scx_cmask *kern_va = *this_cpu_ptr(sch->set_cmask_scratch); + struct scx_cmask_ref ref; + + scx_cmask_ref_init_kern(sch, kern_va, 0, num_possible_cpus(), &ref); + scx_cmask_ref_from_cpumask(&ref, cpumask); + return kern_va; +} + +/** + * scx_call_op_set_cpumask - Invoke the set_cpumask or set_cmask op for @task * @sch: scx_sched being invoked * @rq: rq to update as the currently-locked rq, or NULL * @task: task whose affinity is changing * @cpumask: new cpumask * - * For cid-form schedulers, translate @cpumask to a cmask via the per-cpu - * scratch in cid.c and dispatch through the ops_cid union view. Caller - * must hold @rq's rq lock so this_cpu_ptr is stable across the call. + * For cid-form schedulers, translate @cpumask to a cmask in the per-cpu scratch + * and dispatch through the ops_cid union view. Caller must hold @rq's rq lock. */ static inline void scx_call_op_set_cpumask(struct scx_sched *sch, struct rq *rq, struct task_struct *task, const struct cpumask *cpumask) { - if (scx_is_cid_type()) { - struct scx_cmask *kern_va = *this_cpu_ptr(sch->set_cmask_scratch); - struct scx_cmask_ref ref; - - /* - * Build the per-cpu arena cmask from kernel geometry via @ref, - * never reading its BPF-writable header. set_cmask()'s __arena - * argument takes the kernel address and the struct_ops - * trampoline rebases it into BPF's arena pointer form. The rq - * lock makes this cpu the sole kernel writer. - */ - scx_cmask_ref_init_kern(sch, kern_va, 0, num_possible_cpus(), &ref); - scx_cmask_ref_from_cpumask(&ref, cpumask); - SCX_CALL_CID_OP_TASK(sch, set_cmask, rq, task, kern_va); - } else { + if (scx_is_cid_type()) + SCX_CALL_CID_OP_TASK(sch, set_cmask, rq, task, + scx_fill_cmask_scratch(sch, cpumask)); + else SCX_CALL_OP_TASK(sch, set_cpumask, rq, task, cpumask); - } } enum scx_dsq_iter_flags { @@ -3634,8 +3642,12 @@ static void set_cpus_allowed_scx(struct task_struct *p, * * Fine-grained memory write control is enforced by BPF making the const * designation pointless. Cast it away when calling the operation. + * + * The cid form receives the initial mask when the task is enabled and + * hears about changes only afterwards, see struct scx_enable_args. */ - if (SCX_HAS_OP(sch, set_cpumask)) + if (SCX_HAS_OP(sch, set_cpumask) && + (!scx_is_cid_type() || scx_get_task_state(p) == SCX_TASK_ENABLED)) scx_call_op_set_cpumask(sch, task_rq(p), p, (struct cpumask *)p->cpus_ptr); } @@ -3944,8 +3956,27 @@ static void __scx_enable_task(struct scx_sched *sch, struct task_struct *p) p->scx.weight = sched_weight_to_cgroup(weight); - if (SCX_HAS_OP(sch, enable)) - SCX_CALL_OP_TASK(sch, enable, rq, p); + if (SCX_HAS_OP(sch, enable)) { + if (scx_is_cid_type()) { + struct scx_cmask *cmask = scx_fill_cmask_scratch(sch, p->cpus_ptr); + struct scx_enable_args args = { + .cmask_arena_addr = scx_kaddr_to_arena(sch, cmask), + }; + + SCX_CALL_CID_OP_TASK(sch, enable, rq, p, &args); + } else { + SCX_CALL_OP_TASK(sch, enable, rq, p); + } + } + + /* + * The initial mask also goes out through set_cmask() so a scheduler can + * track affinity there alone, and before set_weight() so that the mask + * is in place when weight-dependent state is derived, see struct + * scx_enable_args. + */ + if (scx_is_cid_type() && SCX_HAS_OP(sch, set_cpumask)) + scx_call_op_set_cpumask(sch, rq, p, p->cpus_ptr); if (SCX_HAS_OP(sch, set_weight)) SCX_CALL_OP_TASK(sch, set_weight, rq, p, p->scx.weight); @@ -4288,9 +4319,10 @@ static void switching_to_scx(struct rq *rq, struct task_struct *p) /* * set_cpus_allowed_scx() is not called while @p is associated with a - * different scheduler class. Keep the BPF scheduler up-to-date. + * different scheduler class. Keep the BPF scheduler up-to-date. The cid + * form gets its mask from scx_enable_task(). */ - if (SCX_HAS_OP(sch, set_cpumask)) + if (!scx_is_cid_type() && SCX_HAS_OP(sch, set_cpumask)) scx_call_op_set_cpumask(sch, rq, p, (struct cpumask *)p->cpus_ptr); } @@ -8374,10 +8406,11 @@ static struct bpf_struct_ops bpf_sched_ext_ops = { /* * cid-form cfi stubs. Stubs whose signatures match the cpu-form (param types * identical, only param names differ across structs) are reused. Some need - * fresh stubs, set_cmask due to an argument type difference and the sub-sched - * notifiers because no cpu-form stub exists to reuse. + * fresh stubs, set_cmask and enable due to argument differences and the + * sub-sched notifiers because no cpu-form stub exists to reuse. */ static void sched_ext_ops_cid__set_cmask(struct task_struct *p, const struct scx_cmask *cmask__arena) {} +static void sched_ext_ops_cid__enable(struct task_struct *p, struct scx_enable_args *args) {} static void sched_ext_ops__sub_caps_updated(const struct scx_cmask *cmask__arena, u64 caps) {} static void sched_ext_ops__sub_ecaps_updated(s32 cid, u64 before, u64 after) {} @@ -8398,7 +8431,7 @@ static struct sched_ext_ops_cid __bpf_ops_sched_ext_ops_cid = { .update_idle = sched_ext_ops__update_idle, .init_task = sched_ext_ops__init_task, .exit_task = sched_ext_ops__exit_task, - .enable = sched_ext_ops__enable, + .enable = sched_ext_ops_cid__enable, .disable = sched_ext_ops__disable, #ifdef CONFIG_EXT_GROUP_SCHED .cpuctl_init = sched_ext_ops__cgroup_init, @@ -10421,8 +10454,7 @@ __bpf_kfunc const void *scx_bpf_online_cmask(const struct bpf_prog_aux *aux) if (unlikely(!online)) return NULL; - /* BPF rebases by the low 32 bits, like __arena callback args */ - return (void *)((unsigned long)online - sch->arena_kern_base); + return (void *)scx_kaddr_to_arena(sch, online); } /** diff --git a/kernel/sched/ext/internal.h b/kernel/sched/ext/internal.h index d150de10a5c9..1df8f583b0ec 100644 --- a/kernel/sched/ext/internal.h +++ b/kernel/sched/ext/internal.h @@ -250,6 +250,31 @@ struct scx_exit_task_args { bool cancelled; }; +/** + * struct scx_enable_args - Argument container for cid-form ops.enable() + * @cmask_arena_addr: BPF arena address of the cmask of cids the task may run on + * + * @cmask_arena_addr is the task's affinity as it enters the scheduler. + * set_cmask() delivers the same mask right after enable(), before set_weight() + * and the first enqueue, then every affinity change afterwards, and is never + * called before enable(). A scheduler may therefore track affinity in + * set_cmask() alone. + * + * The kernel builds the mask in the scheduler arena from its own geometry, so + * the header is valid regardless of what the scheduler last wrote there. The + * memory is per-cpu scratch reused once the callback returns: copy the bits + * out, don't keep the address. The set_cmask() argument follows the same rules. + * + * The address is a plain value rather than a typed pointer because BTF can't + * mark a struct member as an arena pointer yet and a pointer member would reach + * the program typed as a kernel pointer. Cast it to struct scx_cmask __arena * + * before use. Once arena members can be typed, a typed alias will join this + * field in an anonymous union at the same offset. + */ +struct scx_enable_args { + u64 cmask_arena_addr; +}; + /* argument container for ops.cgroup_init() */ struct scx_cgroup_init_args { /* the weight of the cgroup [1..10000] */ @@ -1037,6 +1062,7 @@ struct sched_ext_ops { * - dispatch -> dispatch (cpu arg is now cid) * - update_idle -> update_idle (cpu arg is now cid) * - set_cpumask -> set_cmask (cmask instead of cpumask) + * - enable -> enable (takes struct scx_enable_args) * - cpu_online -> cid_online * - cpu_offline -> cid_offline * - dump_cpu -> dump_cid @@ -1070,7 +1096,7 @@ struct sched_ext_ops_cid { struct scx_init_task_args *args); void (*exit_task)(struct task_struct *p, struct scx_exit_task_args *args); - void (*enable)(struct task_struct *p); + void (*enable)(struct task_struct *p, struct scx_enable_args *args); void (*disable)(struct task_struct *p); void (*dump)(struct scx_dump_ctx *ctx); void (*dump_cid)(struct scx_dump_ctx *ctx, s32 cid, bool idle); @@ -1533,7 +1559,8 @@ struct scx_sched { * by BUILD_BUG_ON in scx_init()). The anonymous union lets the kernel * access either view of the same storage without function-pointer * casts: use .ops for cpu-form and shared fields, .ops_cid for the - * cid-renamed callbacks (set_cmask, select_cid, cid_online, ...). + * callbacks whose cid-form signature differs (set_cmask, enable, + * select_cid, cid_online, ...). */ union { struct sched_ext_ops ops; @@ -1556,9 +1583,9 @@ struct scx_sched { uintptr_t arena_kern_base; /* - * Per-CPU arena cmask used by scx_call_op_set_cpumask() to hand a cmask - * to ops_cid.set_cmask(). The kernel writes through the stored kern_va - * and passes it to the callback's __arena argument. + * Per-CPU arena cmask the kernel fills from a task's cpumask and hands + * to ops_cid.enable() and ops_cid.set_cmask(). The stored pointers are + * the kernel addresses. */ struct scx_cmask * __percpu *set_cmask_scratch; struct scx_cmask *online_cmask; @@ -1669,6 +1696,19 @@ static inline void *scx_arena_to_kaddr(struct scx_sched *sch, const void *bpf_pt return (void *)(sch->arena_kern_base + (u32)(uintptr_t)bpf_ptr); } +/** + * scx_kaddr_to_arena - Translate a kernel arena address to the BPF form + * @sch: scheduler whose arena hosts @kaddr + * @kaddr: kernel address inside @sch's arena + * + * __arena callback arguments need no translation. Addresses handed to BPF any + * other way, such as struct fields and kfunc return values, go through this. + */ +static inline uintptr_t scx_kaddr_to_arena(struct scx_sched *sch, const void *kaddr) +{ + return (uintptr_t)kaddr - sch->arena_kern_base; +} + enum scx_wake_flags { /* expose select WF_* flags as enums */ SCX_WAKE_FORK = WF_FORK, @@ -2302,9 +2342,9 @@ do { \ } while (0) /* - * Dispatch a task op through the cid-form ops_cid table. Only set_cmask() needs - * this: it takes an arena cmask address instead of a cpumask, so it cannot be - * invoked via its cpu-form set_cpumask() slot. + * Dispatch a task op through the cid-form ops_cid table, for the ops whose + * cid-form signature differs from the cpu-form slot: set_cmask() takes an arena + * cmask instead of a cpumask and enable() takes scx_enable_args. */ #define SCX_CALL_CID_OP_TASK(sch, op, locked_rq, task, args...) \ __SCX_CALL_OP_TASK(sch, ops_cid, op, locked_rq, task, ##args) diff --git a/tools/sched_ext/scx_qmap.bpf.c b/tools/sched_ext/scx_qmap.bpf.c index 67b7c01cae55..2f3653199842 100644 --- a/tools/sched_ext/scx_qmap.bpf.c +++ b/tools/sched_ext/scx_qmap.bpf.c @@ -961,9 +961,6 @@ s32 BPF_STRUCT_OPS_SLEEPABLE(qmap_init_task, struct task_struct *p, taskc->highpri = false; taskc->core_sched_seq = 0; cmask_init(&taskc->cpus_allowed, 0, scx_bpf_nr_cids()); - bpf_rcu_read_lock(); - cmask_from_cpumask(&taskc->cpus_allowed, p->cpus_ptr); - bpf_rcu_read_unlock(); v = bpf_task_storage_get(&task_ctx_stor, p, NULL, BPF_LOCAL_STORAGE_GET_F_CREATE); From d781d1b78acf547bafeb1592e8ba4020dce515c6 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Sat, 19 Sep 2026 04:08:20 -1000 Subject: [PATCH 229/694] selftests/sched_ext: Check the cmask cid-form ops.enable() receives cid-form ops.enable() now hands the task's cmask to the scheduler and set_cmask() repeats it right after. Add a cid-form selftest that checks both against p->cpus_ptr, that they match each other, that the initial set_cmask() lands before set_weight() and before the task first becomes runnable, and that set_cmask() never precedes enable(), across class-switch enables, fork-path enables and live affinity changes. v2: Mismatch details returned through a caller-local struct instead of globals, alloc_words validated in the header check, loop bounded by nr_cids directly (Andrea Righi). Signed-off-by: Tejun Heo --- tools/testing/selftests/sched_ext/Makefile | 1 + .../selftests/sched_ext/enable_cmask.bpf.c | 217 ++++++++++++++++++ .../selftests/sched_ext/enable_cmask.c | 138 +++++++++++ 3 files changed, 356 insertions(+) create mode 100644 tools/testing/selftests/sched_ext/enable_cmask.bpf.c create mode 100644 tools/testing/selftests/sched_ext/enable_cmask.c diff --git a/tools/testing/selftests/sched_ext/Makefile b/tools/testing/selftests/sched_ext/Makefile index 49897727f535..4e06d0baaeec 100644 --- a/tools/testing/selftests/sched_ext/Makefile +++ b/tools/testing/selftests/sched_ext/Makefile @@ -169,6 +169,7 @@ auto-test-targets := \ ddsp_bogus_dsq_fail \ ddsp_vtimelocal_fail \ dsp_local_on \ + enable_cmask \ enq_select_cpu \ exit \ hotplug \ diff --git a/tools/testing/selftests/sched_ext/enable_cmask.bpf.c b/tools/testing/selftests/sched_ext/enable_cmask.bpf.c new file mode 100644 index 000000000000..0068f3c7ab3c --- /dev/null +++ b/tools/testing/selftests/sched_ext/enable_cmask.bpf.c @@ -0,0 +1,217 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * A cid-form scheduler checking the cmask cid-form ops.enable() receives: the + * header, every cid bit against p->cpus_ptr, and that set_cmask() follows with + * the same mask before set_weight() and before the task first becomes runnable, + * and never runs before enable(). + * + * Copyright (c) 2026 Tejun Heo + */ +#include + +char _license[] SEC("license") = "GPL"; + +struct { + __uint(type, BPF_MAP_TYPE_ARENA); + __uint(map_flags, BPF_F_MMAPABLE); + __uint(max_entries, 1 << 16); +} arena SEC(".maps"); + +struct task_ctx { + u64 enable_fp; /* fingerprint of the mask enable() received */ + bool enabled; + bool pending; /* enable() ran, the initial set_cmask() hasn't */ +}; + +struct { + __uint(type, BPF_MAP_TYPE_TASK_STORAGE); + __uint(map_flags, BPF_F_NO_PREALLOC); + __type(key, int); + __type(value, struct task_ctx); +} task_ctx_stor SEC(".maps"); + +/* details of a cid bit mismatch, filled by check_mask() */ +struct mask_mismatch { + s32 cid; + bool want; + bool got; +}; + +u64 nr_enable, nr_initial_set_cmask, nr_set_cmask, nr_set_weight; + +UEI_DEFINE(uei); + +static struct task_ctx *lookup_task_ctx(struct task_struct *p) +{ + struct task_ctx *tctx; + + tctx = bpf_task_storage_get(&task_ctx_stor, p, 0, 0); + if (!tctx) + scx_bpf_error("task_ctx lookup failed for %s[%d]", p->comm, p->pid); + return tctx; +} + +/* + * Verify @m's header and every cid bit against @p's cpumask and fingerprint the + * bits into @fp. Return 0 on success, -EINVAL on a bad header, -ENOENT on a cid + * without a cpu and -EIO on a bit mismatch with the details in @mm. + */ +static int check_mask(struct task_struct *p, const struct scx_cmask __arena *m, u64 *fp, + struct mask_mismatch *mm) +{ + u32 nr_cids = scx_bpf_nr_cids(); + u64 h = 0; + s32 cid; + + if (m->base || m->nr_cids != nr_cids || m->alloc_words != CMASK_NR_WORDS(nr_cids)) + return -EINVAL; + + bpf_for(cid, 0, nr_cids) { + bool want, got; + s32 cpu; + + cpu = scx_bpf_cid_to_cpu(cid); + if (cpu < 0) + return -ENOENT; + want = bpf_cpumask_test_cpu(cpu, p->cpus_ptr); + got = cmask_test(cid, m); + if (want != got) { + mm->cid = cid; + mm->want = want; + mm->got = got; + return -EIO; + } + h = h * 31 + got; + } + + *fp = h; + return 0; +} + +s32 BPF_STRUCT_OPS_SLEEPABLE(enable_cmask_init_task, struct task_struct *p, + struct scx_init_task_args *args) +{ + if (!bpf_task_storage_get(&task_ctx_stor, p, 0, BPF_LOCAL_STORAGE_GET_F_CREATE)) + return -ENOMEM; + return 0; +} + +void BPF_STRUCT_OPS(enable_cmask_enable, struct task_struct *p, struct scx_enable_args *args) +{ + struct scx_cmask __arena *m = (struct scx_cmask __arena *)args->cmask_arena_addr; + struct mask_mismatch mm = {}; + struct task_ctx *tctx; + int ret; + + asm volatile("" :: "r"(&arena)); + tctx = lookup_task_ctx(p); + if (!tctx) + return; + + __sync_fetch_and_add(&nr_enable, 1); + if (tctx->enabled || tctx->pending) { + scx_bpf_error("enable: %s[%d] enabled twice", p->comm, p->pid); + return; + } + + ret = check_mask(p, m, &tctx->enable_fp, &mm); + if (ret) { + scx_bpf_error("enable: %s[%d] cmask check failed %d cid=%d want=%d got=%d", + p->comm, p->pid, ret, mm.cid, mm.want, mm.got); + return; + } + tctx->enabled = true; + tctx->pending = true; +} + +void BPF_STRUCT_OPS(enable_cmask_set_cmask, struct task_struct *p, + struct scx_cmask __arena *m) +{ + struct mask_mismatch mm = {}; + struct task_ctx *tctx; + u64 fp; + int ret; + + asm volatile("" :: "r"(&arena)); + tctx = lookup_task_ctx(p); + if (!tctx) + return; + + __sync_fetch_and_add(&nr_set_cmask, 1); + if (!tctx->enabled) { + scx_bpf_error("set_cmask: %s[%d] not enabled", p->comm, p->pid); + return; + } + + ret = check_mask(p, m, &fp, &mm); + if (ret) { + scx_bpf_error("set_cmask: %s[%d] cmask check failed %d cid=%d want=%d got=%d", + p->comm, p->pid, ret, mm.cid, mm.want, mm.got); + return; + } + + if (tctx->pending) { + if (fp != tctx->enable_fp) { + scx_bpf_error("set_cmask: %s[%d] initial mask differs from enable()", + p->comm, p->pid); + return; + } + tctx->pending = false; + __sync_fetch_and_add(&nr_initial_set_cmask, 1); + } +} + +void BPF_STRUCT_OPS(enable_cmask_set_weight, struct task_struct *p, u32 weight) +{ + struct task_ctx *tctx; + + tctx = lookup_task_ctx(p); + if (!tctx) + return; + + __sync_fetch_and_add(&nr_set_weight, 1); + if (tctx->pending) + scx_bpf_error("set_weight: %s[%d] before the initial set_cmask()", p->comm, + p->pid); +} + +void BPF_STRUCT_OPS(enable_cmask_runnable, struct task_struct *p, u64 enq_flags) +{ + struct task_ctx *tctx; + + tctx = lookup_task_ctx(p); + if (!tctx) + return; + + if (tctx->pending) + scx_bpf_error("runnable: %s[%d] before the initial set_cmask()", p->comm, + p->pid); +} + +void BPF_STRUCT_OPS(enable_cmask_disable, struct task_struct *p) +{ + struct task_ctx *tctx; + + tctx = lookup_task_ctx(p); + if (!tctx) + return; + + tctx->enabled = false; + tctx->pending = false; +} + +void BPF_STRUCT_OPS(enable_cmask_exit, struct scx_exit_info *ei) +{ + UEI_RECORD(uei, ei); +} + +SCX_OPS_CID_DEFINE(enable_cmask_ops, + .init_task = (void *)enable_cmask_init_task, + .enable = (void *)enable_cmask_enable, + .set_cmask = (void *)enable_cmask_set_cmask, + .set_weight = (void *)enable_cmask_set_weight, + .runnable = (void *)enable_cmask_runnable, + .disable = (void *)enable_cmask_disable, + .exit = (void *)enable_cmask_exit, + .flags = SCX_OPS_SWITCH_PARTIAL, + .name = "enable_cmask"); diff --git a/tools/testing/selftests/sched_ext/enable_cmask.c b/tools/testing/selftests/sched_ext/enable_cmask.c new file mode 100644 index 000000000000..556bcad4431d --- /dev/null +++ b/tools/testing/selftests/sched_ext/enable_cmask.c @@ -0,0 +1,138 @@ +// SPDX-License-Identifier: GPL-2.0 +/* Copyright (c) 2026 Tejun Heo */ +#define _GNU_SOURCE +#include +#include +#include +#include +#include +#include +#include +#include +#include "enable_cmask.bpf.skel.h" +#include "scx_test.h" + +#define SCHED_EXT 7 +#define NR_CHILDREN 8 +#define MAX_CPUS 1024 + +static int cpus[MAX_CPUS]; +static int nr_cpus; + +static void spin_ms(int ms) +{ + struct timespec start, now; + + clock_gettime(CLOCK_MONOTONIC, &start); + do { + clock_gettime(CLOCK_MONOTONIC, &now); + } while ((now.tv_sec - start.tv_sec) * 1000 + + (now.tv_nsec - start.tv_nsec) / 1000000 < ms); +} + +static int pin(pid_t pid, int idx) +{ + cpu_set_t set; + + CPU_ZERO(&set); + CPU_SET(cpus[idx % nr_cpus], &set); + return sched_setaffinity(pid, sizeof(set), &set); +} + +/* + * Pin, switch to SCHED_EXT for a class-switch enable, fork a grandchild that + * inherits the policy for a fork-path enable, then change affinity a few times + * while running for set_cmask() on live tasks. + */ +static int child(int idx) +{ + struct sched_param param = {}; + int i, status; + pid_t pid; + + if (pin(0, idx) || sched_setscheduler(0, SCHED_EXT, ¶m)) + return 1; + + pid = fork(); + if (pid < 0) + return 1; + if (!pid) { + spin_ms(20); + return 0; + } + + for (i = 1; i <= 4; i++) { + if (pin(0, idx + i)) + return 1; + spin_ms(5); + } + + return waitpid(pid, &status, 0) == pid && !status ? 0 : 1; +} + +static enum scx_test_status run(void *ctx) +{ + struct enable_cmask *skel; + struct bpf_link *link; + pid_t pids[NR_CHILDREN]; + cpu_set_t set; + int i, status, failed = 0; + + if (!__COMPAT_struct_has_field("scx_enable_args", "cmask_arena_addr")) + return SCX_TEST_SKIP; + + SCX_FAIL_IF(sched_getaffinity(0, sizeof(set), &set), "Failed to read affinity"); + for (i = 0; i < MAX_CPUS && i < CPU_SETSIZE; i++) + if (CPU_ISSET(i, &set)) + cpus[nr_cpus++] = i; + if (nr_cpus < 2) + return SCX_TEST_SKIP; + + skel = enable_cmask__open(); + SCX_FAIL_IF(!skel, "Failed to open"); + SCX_ENUM_INIT(skel); + SCX_FAIL_IF(enable_cmask__load(skel), "Failed to load skel"); + + link = bpf_map__attach_struct_ops(skel->maps.enable_cmask_ops); + SCX_FAIL_IF(!link, "Failed to attach struct_ops"); + + for (i = 0; i < NR_CHILDREN; i++) { + pids[i] = fork(); + SCX_FAIL_IF(pids[i] < 0, "Failed to fork"); + if (!pids[i]) + exit(child(i)); + } + + /* affinity changes from the outside race with the children's own */ + for (i = 0; i < NR_CHILDREN; i++) + pin(pids[i], i + NR_CHILDREN); + + for (i = 0; i < NR_CHILDREN; i++) { + if (waitpid(pids[i], &status, 0) != pids[i] || status) + failed++; + } + + bpf_link__destroy(link); + + SCX_EQ(skel->data->uei.kind, EXIT_KIND(SCX_EXIT_UNREG)); + SCX_EQ(failed, 0); + SCX_GE(skel->bss->nr_enable, 2 * NR_CHILDREN); + SCX_EQ(skel->bss->nr_initial_set_cmask, skel->bss->nr_enable); + SCX_GT(skel->bss->nr_set_cmask, skel->bss->nr_initial_set_cmask); + SCX_GE(skel->bss->nr_set_weight, skel->bss->nr_enable); + printf("enable=%lu initial_set_cmask=%lu set_cmask=%lu set_weight=%lu\n", + (unsigned long)skel->bss->nr_enable, + (unsigned long)skel->bss->nr_initial_set_cmask, + (unsigned long)skel->bss->nr_set_cmask, + (unsigned long)skel->bss->nr_set_weight); + + enable_cmask__destroy(skel); + return SCX_TEST_PASS; +} + +struct scx_test enable_cmask = { + .name = "enable_cmask", + .description = "Check the cid-form ops.enable() cmask and the set_cmask() after it", + .run = run, +}; +REGISTER_SCX_TEST(&enable_cmask) From 6b1bca1b1ab77f60a62087337bfe6e2f0efb9e6d Mon Sep 17 00:00:00 2001 From: Karl Mehltretter Date: Mon, 10 Aug 2026 02:56:16 +0200 Subject: [PATCH 230/694] KVM: arm64: Fix AArch32 DBGBXVR handling The consolidation of the breakpoint and watchpoint register accessors switched DBGBXVR from trap_bvr() to trap_dbg_wb_reg(). The latter selects backing storage with demux_wb_reg(), which only handles Op2 values 4 through 7. Since DBGBXVR uses Op2 1, an AArch32 guest access hits KVM_BUG_ON() and marks the VM dead. DBGBXVR aliases DBGBVR_EL1[63:32], and its AA32(HI) descriptor already selects the upper half. Map Op2 1 to dbg_bvr[] alongside Op2 4, restoring the pre-regression behavior. Fixes: 3ce9f3357e9e ("KVM: arm64: Fold DBGxVR/DBGxCR accessors into common set") Cc: stable@vger.kernel.org Assisted-by: Claude:claude-opus-5 Signed-off-by: Karl Mehltretter Reviewed-by: Marc Zyngier Link: https://patch.msgid.link/20260810005616.13227-1-kmehltretter@gmail.com Signed-off-by: Oliver Upton --- arch/arm64/kvm/sys_regs.c | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/arm64/kvm/sys_regs.c b/arch/arm64/kvm/sys_regs.c index 44aae52c473d..a2f4e769a428 100644 --- a/arch/arm64/kvm/sys_regs.c +++ b/arch/arm64/kvm/sys_regs.c @@ -926,6 +926,7 @@ static u64 *demux_wb_reg(struct kvm_vcpu *vcpu, const struct sys_reg_desc *rd) struct kvm_guest_debug_arch *dbg = &vcpu->arch.vcpu_debug_state; switch (rd->Op2) { + case 0b001: case 0b100: return &dbg->dbg_bvr[rd->CRm]; case 0b101: From c82b797abe668d0b668601a93ba2c0b071a63574 Mon Sep 17 00:00:00 2001 From: Weiming Shi Date: Mon, 14 Sep 2026 14:51:23 +0800 Subject: [PATCH 231/694] net/sched: reject IDR error pointers when deleting actions tcf_action_delete() drops the reference held by its lookup before calling tcf_idr_delete_index() with the saved action index. An unlocked classifier can remove that action and reserve the same IDR slot with ERR_PTR(-EBUSY) in between. tcf_idr_delete_index() only checks the lookup result for NULL. It therefore treats the reservation as a tc_action and dereferences tcfa_bindcnt. A hardware execution breakpoint was used to schedule the interleaving without changing the kernel source. KASAN reported this decoded trace: BUG: KASAN: null-ptr-deref in tca_action_gd+0x5b9/0x1010 Read of size 4 at addr 0000000000000010 by task poc/150 Oops: general protection fault, probably for non-canonical address 0xdffffc0000000002 RIP: tca_action_gd+0x5c0/0x1010: arch_atomic_read at arch/x86/include/asm/atomic.h:23 raw_atomic_read at include/linux/atomic/atomic-arch-fallback.h:457 atomic_read at include/linux/atomic/atomic-instrumented.h:33 tcf_idr_delete_index at net/sched/act_api.c:766 tcf_action_delete at net/sched/act_api.c:1859 tcf_del_notify at net/sched/act_api.c:2014 tca_action_gd at net/sched/act_api.c:2064 R13: 0000000000000010 R15: fffffffffffffff0 Kernel panic - not syncing: Fatal exception R15 contains ERR_PTR(-EBUSY), and adding the tcfa_bindcnt offset produces the address in R13. With the guard applied, the same reproducer returned -ENOENT without a KASAN report or panic. Treat error pointers as absent and return -ENOENT. Fixes: 0190c1d452a9 ("net: sched: atomically check-allocate action") Cc: stable@vger.kernel.org Reported-by: Xiang Mei Signed-off-by: Weiming Shi Link: https://patch.msgid.link/20260914065123.4109709-2-bestswngs@gmail.com Signed-off-by: Jakub Kicinski --- net/sched/act_api.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/sched/act_api.c b/net/sched/act_api.c index 3f653721c45f..e45a63be397c 100644 --- a/net/sched/act_api.c +++ b/net/sched/act_api.c @@ -758,7 +758,7 @@ static int tcf_idr_delete_index(struct tcf_idrinfo *idrinfo, u32 index) mutex_lock(&idrinfo->lock); p = idr_find(&idrinfo->action_idr, index); - if (!p) { + if (IS_ERR_OR_NULL(p)) { mutex_unlock(&idrinfo->lock); return -ENOENT; } From 9d565b6b72fe3f41fd43636e143072848105189f Mon Sep 17 00:00:00 2001 From: Aamir Ahmed Date: Tue, 15 Sep 2026 00:06:58 +0100 Subject: [PATCH 232/694] net: usb: catc: bound the RX packet length in catc_rx_done() catc_rx_done() walks a multi-packet URB, reading a two-byte length from each packet header. Its bound, pkt_len > urb->actual_length, ignores the header offset and compares against the whole transfer rather than the bytes left from pkt_start, so a crafted packet header makes skb_copy_to_linear_data() read past the buffer. A length below ETH_HLEN is also accepted, including zero, and eth_type_trans() then reads a MAC header from the uninitialised tailroom of a shorter skb. The is_f5u011 branch takes its length straight from the transfer, so a zero-length URB reaches the same path. Track the bytes remaining from the current packet, and reject a header that does not fit, a length past what is left, and a length below an Ethernet header. A transfer shorter than an Ethernet header, including a zero-length one, previously became a runt skb passed to netif_rx() and counted as received; it is now counted in rx_length_errors and ends the walk. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Signed-off-by: Aamir Ahmed Reviewed-by: Simon Horman Link: https://patch.msgid.link/AS8P251MB00015FD7716F38C345619B56C8BB2@AS8P251MB0001.EURP251.PROD.OUTLOOK.COM Signed-off-by: Jakub Kicinski --- drivers/net/usb/catc.c | 15 ++++++++++++--- 1 file changed, 12 insertions(+), 3 deletions(-) diff --git a/drivers/net/usb/catc.c b/drivers/net/usb/catc.c index 96e82f94edcf..39b678f175dd 100644 --- a/drivers/net/usb/catc.c +++ b/drivers/net/usb/catc.c @@ -233,17 +233,26 @@ static void catc_rx_done(struct urb *urb) } do { - if(!catc->is_f5u011) { - pkt_len = le16_to_cpup((__le16*)pkt_start); - if (pkt_len > urb->actual_length) { + int remaining = urb->actual_length - + (pkt_start - (u8 *)urb->transfer_buffer); + + if (!catc->is_f5u011) { + if (remaining < pkt_offset) { catc->netdev->stats.rx_length_errors++; catc->netdev->stats.rx_errors++; break; } + pkt_len = le16_to_cpup((__le16 *)pkt_start); } else { pkt_len = urb->actual_length; } + if (pkt_len < ETH_HLEN || pkt_len + pkt_offset > remaining) { + catc->netdev->stats.rx_length_errors++; + catc->netdev->stats.rx_errors++; + break; + } + if (!(skb = dev_alloc_skb(pkt_len))) return; From ab888242fce4f16f6c4d4c6ec53939ad36aa3b3a Mon Sep 17 00:00:00 2001 From: Xiang Mei Date: Tue, 15 Sep 2026 01:31:52 -0700 Subject: [PATCH 233/694] vlan: require the MAC header to be present in __vlan_insert_inner_tag() __vlan_insert_inner_tag() only guarantees head room via skb_cow_head(), never that mac_len bytes of MAC header are present. Its ETH_HLEN wrappers - __vlan_insert_tag() under skb_vlan_push(), and vlan_insert_tag() under validate_xmit_vlan() on the generic transmit path - therefore rewrite the first 16 bytes at skb->data: a 12-byte memmove plus two 2-byte stores at +12 and +14. No caller supplies the bound, while the pop helpers use skb_ensure_writable()/pskb_may_pull(). An IFF_TUN device has hard_header_len == 0, so packet_snd() accepts a one-byte AF_PACKET/SOCK_RAW frame. The first vlan push only sets a hwaccel tag; the next - clsact "action vlan push" or bpf_skb_vlan_push() - enters the helper with skb->len still 1. The head comes from skbuff_small_head without __GFP_ZERO, so each push drags bytes from beyond skb->tail into the frame. After three the one-byte send leaves as 13 bytes carrying 11 bytes of uninitialised slab: 0000: 5a b3 62 12 80 88 ff ff 00 b3 62 12 81 `------------------------------' only 0x5a was sent; the rest is slab, here the top 56 bits of a linear-map address Require the MAC header the helper rewrites to be present, so such a frame is dropped rather than transmitted. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Cc: stable@vger.kernel.org Reported-by: co+0ea1ac045375cf05@bugs.sh Signed-off-by: Xiang Mei Reviewed-by: Simon Horman Link: https://patch.msgid.link/20260915083152.705309-1-xmei5@asu.edu Signed-off-by: Jakub Kicinski --- include/linux/if_vlan.h | 3 +++ 1 file changed, 3 insertions(+) diff --git a/include/linux/if_vlan.h b/include/linux/if_vlan.h index 20cc16ea4e5a..4846032bf4ff 100644 --- a/include/linux/if_vlan.h +++ b/include/linux/if_vlan.h @@ -365,6 +365,9 @@ static inline int __vlan_insert_inner_tag(struct sk_buff *skb, const u8 meta_len = mac_len > ETH_TLEN ? skb_metadata_len(skb) : 0; struct vlan_ethhdr *veth; + if (unlikely(!pskb_may_pull(skb, mac_len))) + return -EINVAL; + if (skb_cow_head(skb, meta_len + VLAN_HLEN) < 0) return -ENOMEM; From a11212910cf09b2fe8db9afa41ef60c4f81879c5 Mon Sep 17 00:00:00 2001 From: Yuqi Xu Date: Sat, 19 Sep 2026 16:45:02 +0800 Subject: [PATCH 234/694] bpf: Check params size before reading reserved fields bpf_crypto_ctx_create() is a kfunc whose second argument is declared with the __sz annotation, so the verifier only guarantees that params__sz bytes of params are valid. The function nevertheless reads params->reserved[0] and params->reserved[1] (offsets 14 and 15) before comparing params__sz against the size of struct bpf_crypto_params, so a BPF program can pass a shorter buffer and have the kernel read past the region that was validated for it. Move the size check in front of the reserved field reads. Fixes: 3e1c6f35409f ("bpf: make common crypto API for TC/XDP programs") Reported-by: Vega Signed-off-by: Yuqi Xu Signed-off-by: Alexei Starovoitov Reviewed-by: Ren Wei Link: https://patch.msgid.link/4f3ab4b03e79017e215521743996555439bf0bb3.1789802413.git.xuyuqiabc@gmail.com --- kernel/bpf/crypto.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/kernel/bpf/crypto.c b/kernel/bpf/crypto.c index 51f89cecefb4..3f3fe2450fc6 100644 --- a/kernel/bpf/crypto.c +++ b/kernel/bpf/crypto.c @@ -149,8 +149,9 @@ bpf_crypto_ctx_create(const struct bpf_crypto_params *params, u32 params__sz, const struct bpf_crypto_type *type; struct bpf_crypto_ctx *ctx; - if (!params || params->reserved[0] || params->reserved[1] || - params__sz != sizeof(struct bpf_crypto_params)) { + if (!params || + params__sz != sizeof(struct bpf_crypto_params) || + params->reserved[0] || params->reserved[1]) { *err = -EINVAL; return NULL; } From 8a60ade2277e1f0e0d0578d565354e52292fa46d Mon Sep 17 00:00:00 2001 From: Jamal Hadi Salim Date: Thu, 17 Sep 2026 06:57:32 -0400 Subject: [PATCH 235/694] net/sched: sch_hfsc: bound the classify inner-filter walk with a drift budget MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit hfsc_classify() applies the "filter may only point downwards" level check only when the filter result carries no bound class. A filter created with a flowid gets res.class set once at bind time, so the check never runs for it during classification. hfsc_adjust_levels() can later raise a class's level without revalidating existing bindings, leaving two binds that were each legal at bind time pointing at each other; the classify walk then bounces between two interior classes forever with the qdisc lock held and BH disabled — a soft lockup from a single packet. The stuck walk trips the watchdog: watchdog: BUG: soft lockup - CPU#3 stuck for 13s! [ping:444] RIP: 0010:u32_classify+0x542/0x17f0 ... tcf_classify+0x66/0xa0 hfsc_enqueue+0x166/0xdf0 Bound the traversal with a budget of non-descending hops, the only way a configured walk can move without descending the class tree once levels drift after bind time. The budget is cumulative over the whole walk and is deliberately not reset on a descending hop: a chain that alternates a descent with a lateral hop would return the budget every lap and never trip. Descending hops never decrement it, so legitimately deep trees are unaffected and a terminating lateral chain still classifies normally. Drop the packet with a rate-limited warning once the budget is exhausted, mirroring the merged HTB fix. This is a follow-up to commit 729c4896ab82 ("net/sched: sch_htb: limit htb_classify inner-class filter hops"), which bounded the same classify loop on the HTB side but left the HFSC walk unbounded. Conditions to recreate the bug: - CONFIG_NET_SCHED, CONFIG_NET_SCH_HFSC, CONFIG_NET_CLS_U32, CONFIG_LOCKUP_DETECTOR. - Build a cycle with two legal-at-bind-time flowid binds and a level drift: class X 1:1 (child of root) with leaf child 1:10; class Y 1:2 (sibling of X) with children 1:20 and 1:200; root u32 filter flowid 1:1; filter on X flowid 1:2 (legal when Y is a leaf); after Y's level rises to 2, filter on Y flowid 1:1 (legal then). Send one packet (ping on the device). Unfixed kernel: classify spins with the qdisc lock held; with softlockup_panic=1 it panics. - Reachable from unprivileged user via unshare -Urn (CAP_NET_ADMIN). Fixes: a2f79227138c ("net_sched: sch_hfsc: fix classification loops") Reported-by: Sashiko (gemini + nipa) Closes: https://lore.kernel.org/netdev/QDISC-CTUU.v2.20260913192614@mojatatu.com/ Link: https://sashiko.dev/#/patchset/QDISC-CTUU.v2.20260913192614@mojatatu.com Link: https://netdev-ai.bots.linux.dev/sashiko/#/patchset/QDISC-CTUU.v2.20260913192614%40mojatatu.com Reviewed-by: Victor Nogueira Tested-by: hybris Signed-off-by: Jamal Hadi Salim Link: https://patch.msgid.link/QDISC-CTUU.v3.20260916184908@mojatatu.com Signed-off-by: Jakub Kicinski --- net/sched/sch_hfsc.c | 22 ++++++++++++++++++++++ 1 file changed, 22 insertions(+) diff --git a/net/sched/sch_hfsc.c b/net/sched/sch_hfsc.c index e87f5021a199..284490fd6ca9 100644 --- a/net/sched/sch_hfsc.c +++ b/net/sched/sch_hfsc.c @@ -386,6 +386,15 @@ cftree_update(struct hfsc_class *cl) #define SM_MASK ((1ULL << SM_SHIFT) - 1) #define ISM_MASK ((1ULL << ISM_SHIFT) - 1) +/* + * Cap on the non-descending hops a classify walk may take before its + * filter chain is treated as misconfigured. A flowid binding that was + * legal at bind time can become lateral once hfsc_adjust_levels() + * raises a class level; a few such hops are legitimate, an unbounded + * run means the chain cycles. + */ +#define HFSC_CLASSIFY_MAX_DRIFT 8 + static inline u64 seg_x2y(u64 x, u64 sm) { @@ -1133,6 +1142,7 @@ hfsc_classify(struct sk_buff *skb, struct Qdisc *sch, int *qerr) struct hfsc_class *head, *cl; struct tcf_result res; struct tcf_proto *tcf; + unsigned int drift; int result; if (TC_H_MAJ(skb->priority ^ sch->handle) == 0 && @@ -1142,6 +1152,7 @@ hfsc_classify(struct sk_buff *skb, struct Qdisc *sch, int *qerr) *qerr = NET_XMIT_SUCCESS | __NET_XMIT_BYPASS; head = &q->root; + drift = HFSC_CLASSIFY_MAX_DRIFT; tcf = rcu_dereference_bh(q->root.filter_list); while (tcf && (result = tcf_classify_qdisc(skb, tcf, &res, false)) >= 0) { #ifdef CONFIG_NET_CLS_ACT @@ -1167,6 +1178,17 @@ hfsc_classify(struct sk_buff *skb, struct Qdisc *sch, int *qerr) if (cl->level == 0) return cl; /* hit leaf class */ + /* + * flowid binds skip the level check above (res.class is set + * at bind time and levels drift after), so a walk can follow + * lateral hops without descending; a bounded number of them + * is legal, more means the chain cycles. + */ + if (cl->level >= head->level && drift-- == 0) { + pr_warn_ratelimited("hfsc: classify hop budget exhausted, dropping packet\n"); + return NULL; + } + /* apply inner filter chain */ tcf = rcu_dereference_bh(cl->filter_list); head = cl; From 1e24c4f2ee44be0eee94092b5d13cbdb4bdf0d60 Mon Sep 17 00:00:00 2001 From: Jamal Hadi Salim Date: Thu, 17 Sep 2026 06:57:33 -0400 Subject: [PATCH 236/694] selftests: tc-testing: add a lateral-drift hfsc classify-walk test The classify-loop fix bounds a walk's non-descending hops, so the guard must not misfire on a legal walk that reaches its leaf through a level-drift lateral chain. Add a case that builds exactly that chain and asserts traffic still reaches the chain's own leaf. A lateral hop can only exist because a bind was legal when it was made and a later class add raised the target's level, so the setup binds each hop while the target is still a leaf and only then deepens it: bind 1:1 -> 1:2 while 1:2 is a leaf, add 1:20 under 1:2, add 1:3 and bind 1:2 -> 1:3 while 1:3 is a leaf, then add 1:30 and 1:31 under 1:3 and bind 1:3 -> 1:31. The walk root -> 1:1 -> 1:2 -> 1:3 -> 1:31 then takes two lateral hops and must reach leaf 1:31. The default class is 1:30, distinct from the asserted leaf, and the verify pattern is anchored to the 1:31 stats line, so neither a fall-through to the default nor a nonzero count on another class can satisfy the check. On the patched kernel the test passes; with the bound forced to zero the walk falls to the default and 1:31 stays idle, so the test fails. Reviewed-by: Victor Nogueira Tested-by: hybris Signed-off-by: Jamal Hadi Salim Link: https://patch.msgid.link/QDISC-CTUU.v3.20260916184908@mojatatu.com.2 Signed-off-by: Jakub Kicinski --- .../tc-testing/tc-tests/qdiscs/hfsc.json | 34 +++++++++++++++++++ 1 file changed, 34 insertions(+) diff --git a/tools/testing/selftests/tc-testing/tc-tests/qdiscs/hfsc.json b/tools/testing/selftests/tc-testing/tc-tests/qdiscs/hfsc.json index c98c339424d4..4f6bbb8b57f9 100644 --- a/tools/testing/selftests/tc-testing/tc-tests/qdiscs/hfsc.json +++ b/tools/testing/selftests/tc-testing/tc-tests/qdiscs/hfsc.json @@ -169,5 +169,39 @@ "teardown": [ "$TC qdisc del dev $DUMMY handle 1: root" ] + }, + { + "id": "8c39", + "name": "HFSC classify walk still reaches leaf after lateral drift", + "category": [ + "qdisc", + "hfsc" + ], + "plugins": { + "requires": "nsPlugin" + }, + "setup": [ + "ip link set lo up", + "$TC qdisc add dev lo handle 1: root hfsc default 30", + "$TC class add dev lo parent 1: classid 1:1 hfsc rt m2 100kbit", + "$TC class add dev lo parent 1:1 classid 1:10 hfsc rt m2 50kbit", + "$TC class add dev lo parent 1: classid 1:2 hfsc rt m2 100kbit", + "$TC filter add dev lo parent 1: protocol ip prio 1 u32 match u8 0 0 at 0 flowid 1:1", + "$TC filter add dev lo parent 1:1 protocol ip prio 1 u32 match u8 0 0 at 0 flowid 1:2", + "$TC class add dev lo parent 1:2 classid 1:20 hfsc rt m2 10kbit", + "$TC class add dev lo parent 1: classid 1:3 hfsc rt m2 100kbit", + "$TC filter add dev lo parent 1:2 protocol ip prio 1 u32 match u8 0 0 at 0 flowid 1:3", + "$TC class add dev lo parent 1:3 classid 1:30 hfsc rt m2 10kbit", + "$TC class add dev lo parent 1:3 classid 1:31 hfsc rt m2 100kbit", + "$TC filter add dev lo parent 1:3 protocol ip prio 1 u32 match u8 0 0 at 0 flowid 1:31" + ], + "cmdUnderTest": "ping -n -c 10 -W 1 127.0.0.1", + "expExitCode": "0", + "verifyCmd": "$TC -s class show dev lo", + "matchPattern": "class hfsc 1:31 parent 1:3 rt[^\\n]*\\n Sent [0-9]+ bytes [1-9][0-9]* pkt", + "matchCount": "1", + "teardown": [ + "$TC qdisc del dev lo handle 1: root" + ] } ] From 8f6f8a48399f82f4a83f7b9f25b9707e0062a4f9 Mon Sep 17 00:00:00 2001 From: Adarsh Das Date: Sat, 19 Sep 2026 16:27:32 +0530 Subject: [PATCH 237/694] smb: client: delete compound mids on send failure before unlock When sending a compound request fails, smb_send_rqst() kicks off a reconnect. compound_send_recv() still has those mids on pending_mid_q, but it unlocks the server without removing them first. During reconnect, cifs_abort_connection() walks pending_mid_q and runs each mid callback. With no response yet, those callbacks return credits and drop in_flight. Then compound_send_recv()'s send-error path returns the same credits again. in_flight ends up decremented twice and smb2_add_credits() WARNs. syzbot hits this during SMB2_negotiate when the socket send fails. cifs_call_async() already calls delete_mid() before unlock on send failure. Do the same for compound chains and set cancelled_mid[] so the out: path does not delete them again. Fixes: ee258d79159a ("CIFS: Move credit processing to mid callbacks for SMB3") Reported-by: syzbot+eeb58d2197d88720a228@syzkaller.appspotmail.com Closes: https://lore.kernel.org/r/6a727d22.40259c87.584f4.04ce.GAE@google.com Tested-by: syzbot+eeb58d2197d88720a228@syzkaller.appspotmail.com Assisted-by: LLM Cc: stable@vger.kernel.org Signed-off-by: Adarsh Das Signed-off-by: Paulo Alcantara --- fs/smb/client/transport.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/fs/smb/client/transport.c b/fs/smb/client/transport.c index e266859818a4..7df5b3447aea 100644 --- a/fs/smb/client/transport.c +++ b/fs/smb/client/transport.c @@ -965,6 +965,10 @@ compound_send_recv(const unsigned int xid, struct cifs_ses *ses, if (rc < 0) { revert_current_mid(server, num_rqst); server->sequence_number -= 2; + for (i = 0; i < num_rqst; i++) { + delete_mid(server, mid[i]); + cancelled_mid[i] = true; + } } cifs_server_unlock(server); From c4e941bb7654bcbdfb0b6f3341dc2acfdf235c8d Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Tue, 15 Sep 2026 22:10:30 +0200 Subject: [PATCH 238/694] landlock: Work around gcc-16 -Wuninitialized warning MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit gcc has a bug with -ftrivial-auto-var-init=pattern that produces a warning for correct code that uses sparse bitfields: security/landlock/fs.c: In function 'is_access_to_paths_allowed.isra': security/landlock/fs.c:767:28: error: '_layer_masks_child1' is used uninitialized [-Werror=uninitialized] 767 | struct layer_masks _layer_masks_child1, _layer_masks_child2; | ^~~~~~~~~~~~~~~~~~~ security/landlock/fs.c:767:28: note: '_layer_masks_child1' declared here 767 | struct layer_masks _layer_masks_child1, _layer_masks_child2; | ^~~~~~~~~~~~~~~~~~~ security/landlock/fs.c: In function 'hook_unix_find': security/landlock/fs.c:1649:28: error: 'layer_masks' is used uninitialized [-Werror=uninitialized] 1649 | struct layer_masks layer_masks; | ^~~~~~~~~~~ security/landlock/fs.c:1649:28: note: 'layer_masks' declared here 1649 | struct layer_masks layer_masks; | ^~~~~~~~~~~ To work around this, change the definition of struct layer_mask to use an explictit padding field. Link: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=110743 Link: https://lore.kernel.org/all/20260619082133.3504146-1-arnd@kernel.org/ Fixes: a260c0055665 ("landlock: Add a place for flags to layer rules") Signed-off-by: Arnd Bergmann Link: https://patch.msgid.link/20260915201036.3527935-1-arnd@kernel.org [mic: Use BITS_PER_TYPE(), fix kdoc warnings, fix commit message according to v2 changes] Cc: stable@vger.kernel.org Signed-off-by: Mickaël Salaün --- security/landlock/access.h | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/security/landlock/access.h b/security/landlock/access.h index bbbb41f41147..f843835851d0 100644 --- a/security/landlock/access.h +++ b/security/landlock/access.h @@ -61,6 +61,10 @@ union access_masks_all { static_assert(sizeof(typeof_member(union access_masks_all, masks)) == sizeof(typeof_member(union access_masks_all, all))); +#define _LANDLOCK_LAYER_MASK_PADDING \ + (BITS_PER_TYPE(access_mask_t) - LANDLOCK_NUM_ACCESS_MAX - \ + IS_ENABLED(CONFIG_SECURITY_LANDLOCK_LOG)) + /** * struct layer_mask - The access rights and rule flags for a layer. * @@ -81,6 +85,10 @@ struct layer_mask { */ access_mask_t quiet : 1; #endif /* CONFIG_SECURITY_LANDLOCK_LOG */ + /** + * @__pad: Padding for the compiler's bitfield initialization. + */ + access_mask_t __pad : _LANDLOCK_LAYER_MASK_PADDING; } __packed __aligned(sizeof(access_mask_t)); /* From f71ecaece401cef287cdca12aad785fec809cb9e Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Micka=C3=ABl=20Sala=C3=BCn?= Date: Fri, 18 Sep 2026 20:50:24 +0200 Subject: [PATCH 239/694] landlock: Fix tracepoint fixed-width type names MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The new Landlock tracepoints use UAPI-prefixed __u32 and __u64 names for callback arguments and record fields, including internal IDs that are not Landlock UAPI values. Typed BPF consumers see callback typedef names through BTF. Use the kernel u32 and u64 aliases before release so the tracepoint contract does not present internal values as Landlock UAPI types. This changes BTF-visible typedef spelling but not integer widths, calling conventions, tracefs formats, or record layouts. Cc: Günther Noack Cc: Steven Rostedt Link: https://patch.msgid.link/20260918185036.608651-2-mic@digikod.net Signed-off-by: Mickaël Salaün --- include/trace/events/landlock.h | 68 ++++++++++++++++----------------- 1 file changed, 34 insertions(+), 34 deletions(-) diff --git a/include/trace/events/landlock.h b/include/trace/events/landlock.h index d05253afaf59..4984f80923ed 100644 --- a/include/trace/events/landlock.h +++ b/include/trace/events/landlock.h @@ -243,8 +243,8 @@ static inline const char *__trace_landlock_print_layers( * Field encoding * ~~~~~~~~~~~~~~ * - * Fields that mirror the Landlock UAPI use the same C types and endianness - * (e.g. network ports are __u64 in host endianness, like + * Fields that mirror the Landlock UAPI preserve their widths and endianness + * (e.g. network ports are u64 in host endianness, like * landlock_net_port_attr.port). Per-event details, such as where a value * is byte-swapped, live in the field's own kdoc. * @@ -319,8 +319,8 @@ TRACE_EVENT(landlock_create_ruleset, TP_ARGS(ruleset), TP_STRUCT__entry( - __field( __u64, ruleset_id ) - __field( __u32, ruleset_version ) + __field( u64, ruleset_id ) + __field( u32, ruleset_version ) __field( access_mask_t, handled_fs ) __field( access_mask_t, handled_net ) __field( access_mask_t, scoped ) @@ -359,8 +359,8 @@ TRACE_EVENT(landlock_free_ruleset, TP_ARGS(ruleset), TP_STRUCT__entry( - __field( __u64, ruleset_id ) - __field( __u32, ruleset_version ) + __field( u64, ruleset_id ) + __field( u32, ruleset_version ) ), TP_fast_assign( @@ -396,8 +396,8 @@ TRACE_EVENT(landlock_add_rule_fs, TP_ARGS(ruleset, access_rights, path, pathname), TP_STRUCT__entry( - __field( __u64, ruleset_id ) - __field( __u32, ruleset_version ) + __field( u64, ruleset_id ) + __field( u32, ruleset_version ) __field( access_mask_t, access_rights ) __field( dev_t, dev ) __field( ino_t, ino ) @@ -443,15 +443,15 @@ TRACE_EVENT(landlock_add_rule_fs, TRACE_EVENT(landlock_add_rule_net, TP_PROTO(const struct landlock_ruleset *ruleset, - access_mask_t access_rights, __u64 port), + access_mask_t access_rights, u64 port), TP_ARGS(ruleset, access_rights, port), TP_STRUCT__entry( - __field( __u64, ruleset_id ) - __field( __u32, ruleset_version ) + __field( u64, ruleset_id ) + __field( u32, ruleset_version ) __field( access_mask_t, access_rights ) - __field( __u64, port ) + __field( u64, port ) ), TP_fast_assign( @@ -495,10 +495,10 @@ TRACE_EVENT(landlock_create_domain, TP_ARGS(domain, ruleset), TP_STRUCT__entry( - __field( __u64, domain_id ) - __field( __u64, parent_id ) - __field( __u64, ruleset_id ) - __field( __u32, ruleset_version ) + __field( u64, domain_id ) + __field( u64, parent_id ) + __field( u64, ruleset_id ) + __field( u32, ruleset_version ) ), TP_fast_assign( @@ -557,7 +557,7 @@ TRACE_EVENT(landlock_enforce_domain, TP_ARGS(domain, complete, process_wide, no_new_privs), TP_STRUCT__entry( - __field( __u64, domain_id ) + __field( u64, domain_id ) __field( bool, complete ) __field( bool, process_wide ) __field( bool, no_new_privs ) @@ -595,8 +595,8 @@ TRACE_EVENT(landlock_free_domain, TP_ARGS(hierarchy), TP_STRUCT__entry( - __field( __u64, domain_id ) - __field( __u64, denials ) + __field( u64, domain_id ) + __field( u64, denials ) ), TP_fast_assign( @@ -631,7 +631,7 @@ TRACE_EVENT(landlock_check_rule_fs, TP_ARGS(domain, rule, access_request, dentry), TP_STRUCT__entry( - __field( __u64, domain_id ) + __field( u64, domain_id ) __field( access_mask_t, access_request ) __field( dev_t, dev ) __field( ino_t, ino ) @@ -675,14 +675,14 @@ TRACE_EVENT(landlock_check_rule_net, TP_PROTO(const struct landlock_domain *domain, const struct landlock_rule *rule, - access_mask_t access_request, __u64 port), + access_mask_t access_request, u64 port), TP_ARGS(domain, rule, access_request, port), TP_STRUCT__entry( - __field( __u64, domain_id ) + __field( u64, domain_id ) __field( access_mask_t, access_request ) - __field( __u64, port ) + __field( u64, port ) __dynamic_array(access_mask_t, grants, domain->num_layers) ), @@ -729,7 +729,7 @@ TRACE_EVENT(landlock_deny_access_fs, TP_ARGS(hierarchy, same_exec, logged, blockers, path, pathname), TP_STRUCT__entry( - __field( __u64, domain_id ) + __field( u64, domain_id ) __field( bool, same_exec ) __field( bool, logged ) __field( access_mask_t, blockers ) @@ -791,17 +791,17 @@ TRACE_EVENT(landlock_deny_access_net, TP_PROTO(const struct landlock_hierarchy *hierarchy, bool same_exec, bool logged, access_mask_t blockers, const struct sock *sk, - __u64 sport, __u64 dport), + u64 sport, u64 dport), TP_ARGS(hierarchy, same_exec, logged, blockers, sk, sport, dport), TP_STRUCT__entry( - __field( __u64, domain_id ) + __field( u64, domain_id ) __field( bool, same_exec ) __field( bool, logged ) __field( access_mask_t, blockers ) - __field( __u64, sport ) - __field( __u64, dport ) + __field( u64, sport ) + __field( u64, dport ) ), TP_fast_assign( @@ -842,10 +842,10 @@ TRACE_EVENT(landlock_deny_ptrace, TP_ARGS(hierarchy, same_exec, logged, tracee_domain_id, tracee), TP_STRUCT__entry( - __field( __u64, domain_id ) + __field( u64, domain_id ) __field( bool, same_exec ) __field( bool, logged ) - __field( __u64, tracee_domain_id) + __field( u64, tracee_domain_id) __field( pid_t, tracee_pid ) __string( tracee_comm, tracee->comm ) ), @@ -891,10 +891,10 @@ TRACE_EVENT(landlock_deny_scope_signal, TP_ARGS(hierarchy, same_exec, logged, target_domain_id, target), TP_STRUCT__entry( - __field( __u64, domain_id ) + __field( u64, domain_id ) __field( bool, same_exec ) __field( bool, logged ) - __field( __u64, target_domain_id) + __field( u64, target_domain_id) __field( pid_t, target_pid ) __string( target_comm, target->comm ) ), @@ -940,10 +940,10 @@ TRACE_EVENT(landlock_deny_scope_abstract_unix_socket, TP_ARGS(hierarchy, same_exec, logged, peer_domain_id, peer), TP_STRUCT__entry( - __field( __u64, domain_id ) + __field( u64, domain_id ) __field( bool, same_exec ) __field( bool, logged ) - __field( __u64, peer_domain_id ) + __field( u64, peer_domain_id ) __field( pid_t, peer_pid ) /* * Abstract socket names are untrusted binary data from From 0de33ca344fbf983d380d78db6eeb6fe312d5a5e Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Micka=C3=ABl=20Sala=C3=BCn?= Date: Fri, 18 Sep 2026 20:50:25 +0200 Subject: [PATCH 240/694] landlock: Fix filesystem denial blocker reporting MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Filesystem topology denials are rendered with an empty blockers value because their blocker is identified by the request type instead of an access mask. Introduce the private struct landlock_blockers to carry the request type and final missing access mask to filesystem and network denial tracepoints. Copy both members into named trace-record fields, then use the type to print change_topology for topology denials while preserving symbolic access masks for ordinary denials. The request type lets typed BPF consumers distinguish topology denials from access denials. Keeping the native access mask in a pointer-reached field also lets CO-RE adjust existing programs' load width if access_mask_t grows. Cc: Günther Noack Cc: Steven Rostedt Fixes: 01ce260f5ccf ("landlock: Add landlock_deny_access_fs and landlock_deny_access_net") Link: https://patch.msgid.link/20260918185036.608651-3-mic@digikod.net Signed-off-by: Mickaël Salaün --- include/trace/events/landlock.h | 51 ++++++++++++++++++++++++--------- security/landlock/log.h | 5 ++++ security/landlock/trace.c | 19 +++++++++--- 3 files changed, 58 insertions(+), 17 deletions(-) diff --git a/include/trace/events/landlock.h b/include/trace/events/landlock.h index 4984f80923ed..4e304cab1dab 100644 --- a/include/trace/events/landlock.h +++ b/include/trace/events/landlock.h @@ -17,7 +17,9 @@ #include #include +enum landlock_request_type; struct dentry; +struct landlock_blockers; struct landlock_domain; struct landlock_hierarchy; struct landlock_rule; @@ -26,6 +28,10 @@ struct path; struct sock; struct task_struct; +TRACE_DEFINE_ENUM(LANDLOCK_REQUEST_FS_CHANGE_TOPOLOGY); +TRACE_DEFINE_ENUM(LANDLOCK_REQUEST_FS_ACCESS); +TRACE_DEFINE_ENUM(LANDLOCK_REQUEST_NET_ACCESS); + #ifdef CREATE_TRACE_POINTS /* About 6 KiB, leaving about 2 KiB for sibling helpers and fixed fields. */ @@ -182,6 +188,9 @@ static inline const char *__trace_landlock_print_layers( /* Maps a shared _LANDLOCK_*_NAMES entry to a __print_flags() pair. */ #define _LANDLOCK_NAME_ENTRY(mask, name) { mask, name } +#define _LANDLOCK_FS_BLOCKER_TYPE_NAMES \ + { LANDLOCK_REQUEST_FS_CHANGE_TOPOLOGY, "change_topology" } + /** * DOC: Landlock trace events * @@ -281,6 +290,13 @@ static inline const char *__trace_landlock_print_layers( * the two parties without kernel-internal state. The ID is a scalar * snapshot, not a live domain pointer that could dangle: an optional * relational referent is a scalar (0 sentinel), not a nullable pointer. + * + * Blocker fields + * ~~~~~~~~~~~~~~ + * + * The filesystem and network blocker arguments identify the request type + * and carry its final missing access subset when applicable. The type + * determines how to interpret the access value. */ /* @@ -712,8 +728,7 @@ TRACE_EVENT(landlock_check_rule_net, * domain field. * @same_exec: Whether the current task entered the denying domain itself. * @logged: The domain's audit-logging decision for this denial. - * @blockers: Access mask that was blocked (zero for a mount-topology - * change, whose only blocker is the operation itself). + * @blockers: Request type and final missing access subset (never NULL). * @path: Filesystem path that was denied (never NULL). * @pathname: Resolved path string (never NULL; an error placeholder on * resolution failure). @@ -723,8 +738,8 @@ TRACE_EVENT(landlock_check_rule_net, TRACE_EVENT(landlock_deny_access_fs, TP_PROTO(const struct landlock_hierarchy *hierarchy, bool same_exec, - bool logged, access_mask_t blockers, const struct path *path, - const char *pathname), + bool logged, const struct landlock_blockers *blockers, + const struct path *path, const char *pathname), TP_ARGS(hierarchy, same_exec, logged, blockers, path, pathname), @@ -732,7 +747,8 @@ TRACE_EVENT(landlock_deny_access_fs, __field( u64, domain_id ) __field( bool, same_exec ) __field( bool, logged ) - __field( access_mask_t, blockers ) + __field( enum landlock_request_type, blockers_type ) + __field( access_mask_t, blockers_access ) __field( dev_t, dev ) __field( ino_t, ino ) __string( pathname, pathname ) @@ -744,7 +760,8 @@ TRACE_EVENT(landlock_deny_access_fs, __entry->domain_id = hierarchy->id; __entry->same_exec = same_exec; __entry->logged = logged; - __entry->blockers = blockers; + __entry->blockers_type = blockers->type; + __entry->blockers_access = blockers->access; __entry->dev = path->dentry->d_sb->s_dev; /* * A negative dentry has no backing inode, so mirror the @@ -756,7 +773,10 @@ TRACE_EVENT(landlock_deny_access_fs, TP_printk("domain=%llx same_exec=%d logged=%d blockers=%s dev=%u:%u ino=%lu path=%s", __entry->domain_id, __entry->same_exec, __entry->logged, - __print_flags(__entry->blockers, "|", _LANDLOCK_ACCESS_FS_NAMES), + __entry->blockers_type == LANDLOCK_REQUEST_FS_ACCESS ? + __print_flags(__entry->blockers_access, "|", _LANDLOCK_ACCESS_FS_NAMES) : + __print_symbolic(__entry->blockers_type, + _LANDLOCK_FS_BLOCKER_TYPE_NAMES), MAJOR(__entry->dev), MINOR(__entry->dev), __entry->ino, __trace_print_untrusted_str(p, __get_str(pathname), __get_dynamic_array_len(pathname) - 1)) @@ -769,7 +789,7 @@ TRACE_EVENT(landlock_deny_access_fs, * domain field. * @same_exec: Whether the current task entered the denying domain itself. * @logged: The domain's audit-logging decision for this denial. - * @blockers: Access mask that was blocked. + * @blockers: Request type and final missing access subset (never NULL). * @sk: Socket object (never NULL), read without a socket lock, so its * fields are a best-effort snapshot. The denied endpoint is not * available: the hook runs before :manpage:`bind(2)` / @@ -790,8 +810,8 @@ TRACE_EVENT(landlock_deny_access_fs, TRACE_EVENT(landlock_deny_access_net, TP_PROTO(const struct landlock_hierarchy *hierarchy, bool same_exec, - bool logged, access_mask_t blockers, const struct sock *sk, - u64 sport, u64 dport), + bool logged, const struct landlock_blockers *blockers, + const struct sock *sk, u64 sport, u64 dport), TP_ARGS(hierarchy, same_exec, logged, blockers, sk, sport, dport), @@ -799,7 +819,8 @@ TRACE_EVENT(landlock_deny_access_net, __field( u64, domain_id ) __field( bool, same_exec ) __field( bool, logged ) - __field( access_mask_t, blockers ) + __field( enum landlock_request_type, blockers_type ) + __field( access_mask_t, blockers_access ) __field( u64, sport ) __field( u64, dport ) ), @@ -808,14 +829,17 @@ TRACE_EVENT(landlock_deny_access_net, __entry->domain_id = hierarchy->id; __entry->same_exec = same_exec; __entry->logged = logged; - __entry->blockers = blockers; + __entry->blockers_type = blockers->type; + __entry->blockers_access = blockers->access; __entry->sport = sport; __entry->dport = dport; ), TP_printk("domain=%llx same_exec=%d logged=%d blockers=%s sport=%llu dport=%llu", __entry->domain_id, __entry->same_exec, __entry->logged, - __print_flags(__entry->blockers, "|", _LANDLOCK_ACCESS_NET_NAMES), + __entry->blockers_type == LANDLOCK_REQUEST_NET_ACCESS ? + __print_flags(__entry->blockers_access, "|", _LANDLOCK_ACCESS_NET_NAMES) : + "unknown", __entry->sport, __entry->dport) ); @@ -991,6 +1015,7 @@ TRACE_EVENT(landlock_deny_scope_abstract_unix_socket, __get_dynamic_array_len(sun_path) - 1)) ); +#undef _LANDLOCK_FS_BLOCKER_TYPE_NAMES #undef _LANDLOCK_NAME_ENTRY #endif /* _TRACE_LANDLOCK_H */ diff --git a/security/landlock/log.h b/security/landlock/log.h index e0a6e44f3ddd..04f3e241e765 100644 --- a/security/landlock/log.h +++ b/security/landlock/log.h @@ -25,6 +25,11 @@ enum landlock_request_type { LANDLOCK_REQUEST_SCOPE_SIGNAL, }; +struct landlock_blockers { + access_mask_t access; + enum landlock_request_type type; +}; + /* * We should be careful to only use a variable of this type for * landlock_log_denial(). This way, the compiler can remove it entirely if diff --git a/security/landlock/trace.c b/security/landlock/trace.c index 8c21e5de6f0d..58276cc32d3f 100644 --- a/security/landlock/trace.c +++ b/security/landlock/trace.c @@ -61,7 +61,7 @@ void landlock_trace_free_domain(const struct landlock_hierarchy *const hierarchy * * @request: Detail of the user space request. * @youngest_denied: The youngest hierarchy node that denied the access. - * @missing: The set of denied access rights. + * @missing: The final missing access subset, when applicable. * @same_exec: Whether the current task is the same executable that called * landlock_restrict_self() for the denying domain, as computed * by landlock_log_denial(). @@ -83,6 +83,10 @@ void landlock_trace_denial( case LANDLOCK_REQUEST_FS_ACCESS: case LANDLOCK_REQUEST_FS_CHANGE_TOPOLOGY: if (trace_landlock_deny_access_fs_enabled()) { + const struct landlock_blockers blockers = { + .access = missing, + .type = request->type, + }; char *buf __free(__putname) = __getname(); struct path dentry_path; const char *pathname; @@ -147,16 +151,23 @@ void landlock_trace_denial( trace_landlock_deny_access_fs(youngest_denied, same_exec, logged, - missing, path, pathname); + &blockers, path, + pathname); } break; case LANDLOCK_REQUEST_NET_ACCESS: - if (trace_landlock_deny_access_net_enabled()) + if (trace_landlock_deny_access_net_enabled()) { + const struct landlock_blockers blockers = { + .access = missing, + .type = request->type, + }; + trace_landlock_deny_access_net( - youngest_denied, same_exec, logged, missing, + youngest_denied, same_exec, logged, &blockers, request->audit.u.net->sk, ntohs(request->audit.u.net->sport), ntohs(request->audit.u.net->dport)); + } break; case LANDLOCK_REQUEST_PTRACE: if (trace_landlock_deny_ptrace_enabled()) From 1a985d3890ed8caa428390a5682e957503f14060 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Micka=C3=ABl=20Sala=C3=BCn?= Date: Fri, 18 Sep 2026 20:50:26 +0200 Subject: [PATCH 241/694] landlock: Fix rule tracepoint context MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Name each event after the identity it reports. Add-rule events describe UAPI rule insertion, so rename them after LANDLOCK_RULE_PATH_BENEATH and LANDLOCK_RULE_NET_PORT. Check-rule events describe matches in internal rule trees, so rename them after LANDLOCK_KEY_INODE and LANDLOCK_KEY_NET_PORT. This remains accurate if multiple UAPI rule types share one lookup and stored rule. Keep denial event names based on filesystem and network families because they describe final access decisions. Use u64 for growable access masks passed by value to add-rule and check-rule typed BTF callbacks. CO-RE can relocate pointer-reached fields, but it cannot widen a scalar callback slot declared by a BPF program. Keep native access_mask_t for internal state and trace records. For add-rule callbacks, report the normalized per-call contribution passed to landlock_insert_rule() and expose the complete validated flags value. Put the ruleset and flags first as a common invocation prefix. This distinguishes duplicate and effective-zero additions without recovering arguments from saved syscall registers. Cc: Günther Noack Cc: Steven Rostedt Fixes: 63747c94774d ("landlock: Add landlock_add_rule_fs and landlock_add_rule_net tracepoints") Fixes: 3f1f106e4c14 ("landlock: Add tracepoints for rule checking") Link: https://patch.msgid.link/20260918185036.608651-4-mic@digikod.net Signed-off-by: Mickaël Salaün --- Documentation/trace/events-landlock.rst | 22 +++---- include/trace/events/landlock.h | 62 +++++++++++-------- security/landlock/fs.c | 10 +-- security/landlock/net.c | 5 +- tools/testing/selftests/landlock/fs_test.c | 28 ++++----- tools/testing/selftests/landlock/net_test.c | 11 ++-- tools/testing/selftests/landlock/trace.h | 56 ++++++++--------- .../selftests/landlock/trace_fs_test.c | 54 +++++++++------- tools/testing/selftests/landlock/trace_test.c | 51 ++++++++------- 9 files changed, 162 insertions(+), 137 deletions(-) diff --git a/Documentation/trace/events-landlock.rst b/Documentation/trace/events-landlock.rst index af9267cca47d..9bb81a5c676e 100644 --- a/Documentation/trace/events-landlock.rst +++ b/Documentation/trace/events-landlock.rst @@ -6,7 +6,7 @@ Landlock Trace Events ===================== :Author: Mickaël Salaün -:Date: August 2026 +:Date: September 2026 Landlock emits trace events for sandbox lifecycle operations and access denials. These events can be consumed by ftrace (for human-readable @@ -31,8 +31,8 @@ Landlock trace events are organized in four categories: **Syscall events** are emitted during Landlock system calls: - ``landlock_create_ruleset``: a new ruleset is created -- ``landlock_add_rule_fs``: a filesystem rule is added to a ruleset -- ``landlock_add_rule_net``: a network port rule is added to a ruleset +- ``landlock_add_rule_path_beneath``: a filesystem rule is added to a ruleset +- ``landlock_add_rule_net_port``: a network port rule is added to a ruleset - ``landlock_create_domain``: a new domain is created from a ruleset - ``landlock_enforce_domain``: a domain is enforced on a thread @@ -47,8 +47,8 @@ Landlock trace events are organized in four categories: **Rule evaluation events** are emitted during rule matching: -- ``landlock_check_rule_fs``: a filesystem rule is evaluated -- ``landlock_check_rule_net``: a network port rule is evaluated +- ``landlock_check_rule_inode``: an inode-keyed rule is evaluated +- ``landlock_check_rule_net_port``: a network-port-keyed rule is evaluated **Lifecycle events**: @@ -189,7 +189,7 @@ rather than the caller's, so correlate those to the syscall by domain ID. Interpreting check_rule events ============================== -The ``check_rule_fs`` and ``check_rule_net`` events expose the per-layer +The ``check_rule_inode`` and ``check_rule_net_port`` events expose the per-layer rule evaluation, which is useful for understanding *why* a specific access is allowed or denied. @@ -244,22 +244,22 @@ check. For example, a program sandboxed with read and execute access to the whole filesystem reads ``/etc/passwd``; both the ``execve()`` and the -read match the rule covering ``/`` (inode 2), so ``check_rule_fs`` fires +read match the rule covering ``/`` (inode 2), so ``check_rule_inode`` fires with the requested rights intersected against what that rule grants. The ``access_request=`` mask includes ``truncate`` because the file-open hook evaluates that optional right alongside the required access, but the rule does not grant it, so ``truncate`` never appears in ``grants=``:: - cat-127 [...] landlock_check_rule_fs: domain=1e40cb56f access_request=execute|read_file|truncate dev=0:17 ino=2 grants={execute|read_file} - cat-127 [...] landlock_check_rule_fs: domain=1e40cb56f access_request=read_file|truncate dev=0:17 ino=2 grants={read_file} + cat-127 [...] landlock_check_rule_inode: domain=1e40cb56f access_request=execute|read_file|truncate dev=0:17 ino=2 grants={execute|read_file} + cat-127 [...] landlock_check_rule_inode: domain=1e40cb56f access_request=read_file|truncate dev=0:17 ino=2 grants={read_file} The ``[...]`` replaces the ftrace CPU, flags, and timestamp columns. A single ``grants=`` group means the enforcing domain has one layer. With two nested sandboxes that each grant the same rights, the rule spans both layers, so ``grants=`` has one group per layer:: - cat-128 [...] landlock_check_rule_fs: domain=184788b52 access_request=execute|read_file|truncate dev=0:17 ino=2 grants={execute|read_file,execute|read_file} - cat-128 [...] landlock_check_rule_fs: domain=184788b52 access_request=read_file|truncate dev=0:17 ino=2 grants={read_file,read_file} + cat-128 [...] landlock_check_rule_inode: domain=184788b52 access_request=execute|read_file|truncate dev=0:17 ino=2 grants={execute|read_file,execute|read_file} + cat-128 [...] landlock_check_rule_inode: domain=184788b52 access_request=read_file|truncate dev=0:17 ino=2 grants={read_file,read_file} eBPF access =========== diff --git a/include/trace/events/landlock.h b/include/trace/events/landlock.h index 4e304cab1dab..b7e6a6638822 100644 --- a/include/trace/events/landlock.h +++ b/include/trace/events/landlock.h @@ -28,6 +28,8 @@ struct path; struct sock; struct task_struct; +static_assert(sizeof(access_mask_t) <= sizeof(u64)); + TRACE_DEFINE_ENUM(LANDLOCK_REQUEST_FS_CHANGE_TOPOLOGY); TRACE_DEFINE_ENUM(LANDLOCK_REQUEST_FS_ACCESS); TRACE_DEFINE_ENUM(LANDLOCK_REQUEST_NET_ACCESS); @@ -114,7 +116,7 @@ __trace_print_untrusted_str(struct trace_seq *p, const char *src, size_t len) * Fills the dense per-domain-layer array layers (one access mask per layer, * indexed by level - 1) from rule's sparse layer stack, keeping only the * requested rights (access_request). Layers with no matching rule entry get - * a zero mask. Shared by the check_rule_fs and check_rule_net events. + * a zero mask. Shared by the check_rule_inode and check_rule_net_port events. * * rule->layers is sorted by ascending level, with levels in the domain's * [1, num_layers] range (see landlock_merge_ruleset()), so every entry maps @@ -389,12 +391,14 @@ TRACE_EVENT(landlock_free_ruleset, ); /** - * landlock_add_rule_fs - Filesystem rule added to a ruleset + * landlock_add_rule_path_beneath - Path-beneath rule added to a ruleset * * @ruleset: Source ruleset (never NULL). - * @access_rights: Effective access mask stored in the rule, not the raw - * sys_landlock_add_rule() argument (unhandled rights - * added). + * @flags: Complete validated landlock_add_rule_flags value supplied by this + * successful call, not the rule's accumulated quiet state. + * @access_rights: Canonical per-call access mask passed to + * landlock_insert_rule() after normalization, not the raw + * sys_landlock_add_rule() argument or accumulated rule. * @path: Filesystem path for the rule (never NULL). * @pathname: Resolved absolute path string (never NULL; error placeholder * on resolution failure). @@ -403,13 +407,13 @@ TRACE_EVENT(landlock_free_ruleset, * the reported ruleset is a stable snapshot that no concurrent writer can * change. */ -TRACE_EVENT(landlock_add_rule_fs, +TRACE_EVENT(landlock_add_rule_path_beneath, - TP_PROTO(const struct landlock_ruleset *ruleset, - access_mask_t access_rights, const struct path *path, + TP_PROTO(const struct landlock_ruleset *ruleset, u32 flags, + u64 access_rights, const struct path *path, const char *pathname), - TP_ARGS(ruleset, access_rights, path, pathname), + TP_ARGS(ruleset, flags, access_rights, path, pathname), TP_STRUCT__entry( __field( u64, ruleset_id ) @@ -443,25 +447,27 @@ TRACE_EVENT(landlock_add_rule_fs, ); /** - * landlock_add_rule_net - Network port rule added to a ruleset + * landlock_add_rule_net_port - Network-port rule added to a ruleset * * @ruleset: Source ruleset (never NULL). - * @access_rights: Effective access mask stored in the rule, not the raw - * sys_landlock_add_rule() argument (unhandled rights - * added). - * @port: Network port, the landlock_net_port_attr.port UAPI value - * forwarded directly. + * @flags: Complete validated landlock_add_rule_flags value supplied by this + * successful call, not the rule's accumulated quiet state. + * @access_rights: Canonical per-call access mask passed to + * landlock_insert_rule() after normalization, not the raw + * sys_landlock_add_rule() argument or accumulated rule. + * @port: Network port in host endianness, forwarded directly from + * &landlock_net_port_attr.port. * * Emitted by sys_landlock_add_rule() under the modified ruleset's lock, so * the reported ruleset is a stable snapshot that no concurrent writer can * change. */ -TRACE_EVENT(landlock_add_rule_net, +TRACE_EVENT(landlock_add_rule_net_port, - TP_PROTO(const struct landlock_ruleset *ruleset, - access_mask_t access_rights, u64 port), + TP_PROTO(const struct landlock_ruleset *ruleset, u32 flags, + u64 access_rights, u64 port), - TP_ARGS(ruleset, access_rights, port), + TP_ARGS(ruleset, flags, access_rights, port), TP_STRUCT__entry( __field( u64, ruleset_id ) @@ -625,7 +631,7 @@ TRACE_EVENT(landlock_free_domain, ); /** - * landlock_check_rule_fs - Filesystem rule evaluated during access check + * landlock_check_rule_inode - Inode rule evaluated during access check * * @domain: Enforcing domain (never NULL). * @rule: Matching rule with per-layer access masks (never NULL). @@ -638,11 +644,11 @@ TRACE_EVENT(landlock_free_domain, * domain layer. See Documentation/trace/events-landlock.rst for how to * interpret it. */ -TRACE_EVENT(landlock_check_rule_fs, +TRACE_EVENT(landlock_check_rule_inode, TP_PROTO(const struct landlock_domain *domain, const struct landlock_rule *rule, - access_mask_t access_request, const struct dentry *dentry), + u64 access_request, const struct dentry *dentry), TP_ARGS(domain, rule, access_request, dentry), @@ -664,7 +670,8 @@ TRACE_EVENT(landlock_check_rule_fs, __trace_landlock_fill_layers(__get_dynamic_array(grants), __get_dynamic_array_len(grants) / sizeof(access_mask_t), - rule, access_request); + rule, + (access_mask_t)access_request); ), TP_printk("domain=%llx access_request=%s dev=%u:%u ino=%lu grants=%s", @@ -675,7 +682,7 @@ TRACE_EVENT(landlock_check_rule_fs, ); /** - * landlock_check_rule_net - Network port rule evaluated during access check + * landlock_check_rule_net_port - Network port rule evaluated * * @domain: Enforcing domain (never NULL). * @rule: Matching rule with per-layer access masks (never NULL). @@ -687,11 +694,11 @@ TRACE_EVENT(landlock_check_rule_fs, * layer. See Documentation/trace/events-landlock.rst for how to * interpret it. */ -TRACE_EVENT(landlock_check_rule_net, +TRACE_EVENT(landlock_check_rule_net_port, TP_PROTO(const struct landlock_domain *domain, const struct landlock_rule *rule, - access_mask_t access_request, u64 port), + u64 access_request, u64 port), TP_ARGS(domain, rule, access_request, port), @@ -711,7 +718,8 @@ TRACE_EVENT(landlock_check_rule_net, __trace_landlock_fill_layers(__get_dynamic_array(grants), __get_dynamic_array_len(grants) / sizeof(access_mask_t), - rule, access_request); + rule, + (access_mask_t)access_request); ), TP_printk("domain=%llx access_request=%s port=%llu grants=%s", diff --git a/security/landlock/fs.c b/security/landlock/fs.c index 330a1871bf94..cab43892ec2f 100644 --- a/security/landlock/fs.c +++ b/security/landlock/fs.c @@ -356,14 +356,14 @@ int landlock_append_fs_rule(struct landlock_ruleset *const ruleset, * held for BTF consistency (enforced by lockdep_assert_held in * TP_fast_assign). */ - if (!err && trace_landlock_add_rule_fs_enabled()) { + if (!err && trace_landlock_add_rule_path_beneath_enabled()) { char *buffer __free(__putname) = __getname(); const char *pathname = buffer ? resolve_path_for_trace(path, buffer) : ""; - trace_landlock_add_rule_fs(ruleset, access_rights, path, - pathname); + trace_landlock_add_rule_path_beneath( + ruleset, flags, access_rights, path, pathname); } mutex_unlock(&ruleset->lock); @@ -423,8 +423,8 @@ static bool unmask_layers_fs(const struct landlock_domain *const domain, ret = landlock_unmask_layers(domain, id, masks, &rule); if (rule) - trace_landlock_check_rule_fs(domain, rule, access_request, - dentry); + trace_landlock_check_rule_inode(domain, rule, access_request, + dentry); return ret; } diff --git a/security/landlock/net.c b/security/landlock/net.c index 8f2aaac54b33..5552c60388f8 100644 --- a/security/landlock/net.c +++ b/security/landlock/net.c @@ -47,7 +47,8 @@ int landlock_append_net_rule(struct landlock_ruleset *const ruleset, * TP_fast_assign). */ if (!err) - trace_landlock_add_rule_net(ruleset, access_rights, port); + trace_landlock_add_rule_net_port(ruleset, flags, access_rights, + port); mutex_unlock(&ruleset->lock); return err; @@ -63,7 +64,7 @@ static bool unmask_layers_net(const struct landlock_domain *const domain, ret = landlock_unmask_layers(domain, id, masks, &rule); if (rule) - trace_landlock_check_rule_net( + trace_landlock_check_rule_net_port( domain, rule, access_request, ntohs((__force __be16)id.key.data)); return ret; diff --git a/tools/testing/selftests/landlock/fs_test.c b/tools/testing/selftests/landlock/fs_test.c index 18dbdb99aeba..6e979cef884d 100644 --- a/tools/testing/selftests/landlock/fs_test.c +++ b/tools/testing/selftests/landlock/fs_test.c @@ -10493,9 +10493,9 @@ FIXTURE_TEARDOWN_PARENT(trace_layout1) } /* - * Verifies that check_rule_fs events include correct field values: domain, dev, - * ino, access_request, and grants. All values are verified against stat() of - * the rule path on a deterministic tmpfs layout. + * Verifies that check_rule_inode events include correct field values: domain, + * dev, ino, access_request, and grants. All values are verified against stat() + * of the rule path on a deterministic tmpfs layout. */ TEST_F(trace_layout1, check_rule_fs_fields) { @@ -10529,7 +10529,7 @@ TEST_F(trace_layout1, check_rule_fs_fields) EXPECT_EQ(1, tracefs_count_matches(buf, REGEX_CHECK_RULE_FS(TRACE_TASK))) { - TH_LOG("Expected 1 check_rule_fs event\n%s", buf); + TH_LOG("Expected 1 check_rule_inode event\n%s", buf); } ASSERT_EQ(0, tracefs_extract_field(buf, REGEX_CHECK_RULE_FS(TRACE_TASK), @@ -10570,8 +10570,8 @@ TEST_F(trace_layout1, check_rule_fs_fields) } /* - * Verifies check_rule_fs behavior with multiple rules. With rules at s1d1 and - * s1d2 (a child of s1d1), accessing s1d2 produces only 1 event because the + * Verifies check_rule_inode behavior with multiple rules. With rules at s1d1 + * and s1d2 (a child of s1d1), accessing s1d2 produces only 1 event because the * pathwalk short-circuits after the first rule fully unmasks the single layer. */ TEST_F(trace_layout1, check_rule_fs_multiple_rules) @@ -10643,14 +10643,14 @@ TEST_F(trace_layout1, check_rule_fs_multiple_rules) ASSERT_NE(NULL, buf); /* - * Only 1 check_rule_fs event: the rule on dir_s1d2 fully unmasked the - * single layer, so the pathwalk short-circuits before reaching the + * Only one check_rule_inode event: the rule on dir_s1d2 fully unmasks + * the single layer, so the pathwalk short-circuits before reaching the * dir_s1d1 rule. */ count = tracefs_count_matches(buf, REGEX_CHECK_RULE_FS(TRACE_TASK)); EXPECT_EQ(1, count) { - TH_LOG("Expected 1 check_rule_fs event, got %d\n%s", count, + TH_LOG("Expected 1 check_rule_inode event, got %d\n%s", count, buf); } @@ -10777,7 +10777,7 @@ TEST_F(trace_layout1, check_rule_fs_optional_access) count = tracefs_count_matches(buf, REGEX_CHECK_RULE_FS(TRACE_TASK)); EXPECT_EQ(1, count) { - TH_LOG("Expected 1 check_rule_fs event, got %d\n%s", count, + TH_LOG("Expected 1 check_rule_inode event, got %d\n%s", count, buf); } @@ -10796,7 +10796,7 @@ TEST_F(trace_layout1, check_rule_fs_optional_access) } /* - * Verifies that check_rule_fs fires for a rule that matches the inode even when + * Verifies that check_rule_inode fires for a rule matching the inode even when * it grants none of the requested rights, so the grants set is empty. Landlock * cannot know a rule ignores the request before reading it, so the event is * still emitted (grants={}), which lets a tracer see that the rule matched. @@ -10884,7 +10884,7 @@ TEST_F(trace_layout1, check_rule_fs_empty_grant) count = tracefs_count_matches(buf, REGEX_CHECK_RULE_FS(TRACE_TASK)); EXPECT_EQ(2, count) { - TH_LOG("Expected 2 check_rule_fs events, got %d\n%s", count, + TH_LOG("Expected 2 check_rule_inode events, got %d\n%s", count, buf); } @@ -10894,7 +10894,7 @@ TEST_F(trace_layout1, check_rule_fs_empty_grant) tracefs_count_matches( buf, TRACE_PREFIX( - TRACE_TASK) "landlock_check_rule_fs: domain=[0-9a-f]\\+ " + TRACE_TASK) "landlock_check_rule_inode: domain=[0-9a-f]\\+ " "access_request=read_dir " "dev=[0-9]\\+:[0-9]\\+ ino=[0-9]\\+ " "grants={}$")) @@ -10908,7 +10908,7 @@ TEST_F(trace_layout1, check_rule_fs_empty_grant) tracefs_count_matches( buf, TRACE_PREFIX( - TRACE_TASK) "landlock_check_rule_fs: domain=[0-9a-f]\\+ " + TRACE_TASK) "landlock_check_rule_inode: domain=[0-9a-f]\\+ " "access_request=read_dir " "dev=[0-9]\\+:[0-9]\\+ ino=[0-9]\\+ " "grants={read_dir}$")) diff --git a/tools/testing/selftests/landlock/net_test.c b/tools/testing/selftests/landlock/net_test.c index a18761e0fd82..4fb705e1596d 100644 --- a/tools/testing/selftests/landlock/net_test.c +++ b/tools/testing/selftests/landlock/net_test.c @@ -3712,7 +3712,7 @@ TEST_F(trace_net_connect, deny_access_net) free(buf); } -/* Field verification for the check_rule_net event on an allowed access. */ +/* Field verification for the check_rule_net_port event on an allowed access. */ /* clang-format off */ FIXTURE(trace_net_check_rule) { @@ -3757,10 +3757,11 @@ FIXTURE_TEARDOWN(trace_net_check_rule) /* * Verifies that an allowed bind matching a net-port rule emits exactly one - * landlock_check_rule_net event with the enforcing domain, the requested + * landlock_check_rule_net_port event with the enforcing domain, the requested * access, the checked port (host endianness), and the per-layer grants. The - * whole event is anchored to exact values so a revert of the check_rule_net - * emit (or a byte-order or field-plumbing regression) fails the test. + * whole event is anchored to exact values so removing the check_rule_net_port + * emission or introducing a byte-order or field-plumbing regression fails the + * test. */ TEST_F(trace_net_check_rule, check_rule_net_fields) { @@ -3832,7 +3833,7 @@ TEST_F(trace_net_check_rule, check_rule_net_fields) EXPECT_EQ(1, tracefs_count_matches(buf, REGEX_CHECK_RULE_NET(TRACE_TASK))) { - TH_LOG("Expected 1 check_rule_net event\n%s", buf); + TH_LOG("Expected 1 check_rule_net_port event\n%s", buf); } ASSERT_EQ(0, diff --git a/tools/testing/selftests/landlock/trace.h b/tools/testing/selftests/landlock/trace.h index ba0c5e92001f..e6873853376c 100644 --- a/tools/testing/selftests/landlock/trace.h +++ b/tools/testing/selftests/landlock/trace.h @@ -27,14 +27,14 @@ TRACEFS_LANDLOCK_DIR "/landlock_create_domain/enable" #define TRACEFS_ENFORCE_DOMAIN_ENABLE \ TRACEFS_LANDLOCK_DIR "/landlock_enforce_domain/enable" -#define TRACEFS_ADD_RULE_FS_ENABLE \ - TRACEFS_LANDLOCK_DIR "/landlock_add_rule_fs/enable" -#define TRACEFS_ADD_RULE_NET_ENABLE \ - TRACEFS_LANDLOCK_DIR "/landlock_add_rule_net/enable" +#define TRACEFS_ADD_RULE_PATH_BENEATH_ENABLE \ + TRACEFS_LANDLOCK_DIR "/landlock_add_rule_path_beneath/enable" +#define TRACEFS_ADD_RULE_NET_PORT_ENABLE \ + TRACEFS_LANDLOCK_DIR "/landlock_add_rule_net_port/enable" #define TRACEFS_CHECK_RULE_FS_ENABLE \ - TRACEFS_LANDLOCK_DIR "/landlock_check_rule_fs/enable" + TRACEFS_LANDLOCK_DIR "/landlock_check_rule_inode/enable" #define TRACEFS_CHECK_RULE_NET_ENABLE \ - TRACEFS_LANDLOCK_DIR "/landlock_check_rule_net/enable" + TRACEFS_LANDLOCK_DIR "/landlock_check_rule_net_port/enable" #define TRACEFS_DENY_ACCESS_FS_ENABLE \ TRACEFS_LANDLOCK_DIR "/landlock_deny_access_fs/enable" #define TRACEFS_DENY_ACCESS_NET_ENABLE \ @@ -79,18 +79,18 @@ */ #define KWORKER_TASK "kworker/[0-9]\\+:[0-9]\\+" -#define REGEX_ADD_RULE_FS(task) \ - TRACE_PREFIX(task) \ - "landlock_add_rule_fs: " \ - "ruleset=[0-9a-f]\\+\\.[0-9]\\+ " \ - "access_rights=[a-z_|]* " \ - "dev=[0-9]\\+:[0-9]\\+ " \ - "ino=[0-9]\\+ " \ +#define REGEX_ADD_RULE_PATH_BENEATH(task) \ + TRACE_PREFIX(task) \ + "landlock_add_rule_path_beneath: " \ + "ruleset=[0-9a-f]\\+\\.[0-9]\\+ " \ + "access_rights=[a-z_|]* " \ + "dev=[0-9]\\+:[0-9]\\+ " \ + "ino=[0-9]\\+ " \ "path=[^ ]\\+$" -#define REGEX_ADD_RULE_NET(task) \ +#define REGEX_ADD_RULE_NET_PORT(task) \ TRACE_PREFIX(task) \ - "landlock_add_rule_net: " \ + "landlock_add_rule_net_port: " \ "ruleset=[0-9a-f]\\+\\.[0-9]\\+ " \ "access_rights=[a-z_|]* " \ "port=[0-9]\\+$" @@ -110,21 +110,21 @@ "parent=[0-9a-f]\\+ " \ "ruleset=[0-9a-f]\\+\\.[0-9]\\+$" -#define REGEX_CHECK_RULE_FS(task) \ - TRACE_PREFIX(task) \ - "landlock_check_rule_fs: " \ - "domain=[0-9a-f]\\+ " \ - "access_request=[a-z_|]* " \ - "dev=[0-9]\\+:[0-9]\\+ " \ - "ino=[0-9]\\+ " \ +#define REGEX_CHECK_RULE_FS(task) \ + TRACE_PREFIX(task) \ + "landlock_check_rule_inode: " \ + "domain=[0-9a-f]\\+ " \ + "access_request=[a-z_|]* " \ + "dev=[0-9]\\+:[0-9]\\+ " \ + "ino=[0-9]\\+ " \ "grants={[a-z_|,]*}$" -#define REGEX_CHECK_RULE_NET(task) \ - TRACE_PREFIX(task) \ - "landlock_check_rule_net: " \ - "domain=[0-9a-f]\\+ " \ - "access_request=[a-z_|]* " \ - "port=[0-9]\\+ " \ +#define REGEX_CHECK_RULE_NET(task) \ + TRACE_PREFIX(task) \ + "landlock_check_rule_net_port: " \ + "domain=[0-9a-f]\\+ " \ + "access_request=[a-z_|]* " \ + "port=[0-9]\\+ " \ "grants={[a-z_|,]*}$" #define REGEX_DENY_ACCESS_FS(task) \ diff --git a/tools/testing/selftests/landlock/trace_fs_test.c b/tools/testing/selftests/landlock/trace_fs_test.c index 4543a25c1f55..6666d4746cb1 100644 --- a/tools/testing/selftests/landlock/trace_fs_test.c +++ b/tools/testing/selftests/landlock/trace_fs_test.c @@ -121,7 +121,8 @@ FIXTURE_SETUP(trace_fs) } self->tracefs_ok = 1; - ASSERT_EQ(0, tracefs_enable_event(TRACEFS_ADD_RULE_FS_ENABLE, true)); + ASSERT_EQ(0, tracefs_enable_event(TRACEFS_ADD_RULE_PATH_BENEATH_ENABLE, + true)); ASSERT_EQ(0, tracefs_enable_event(TRACEFS_CHECK_RULE_FS_ENABLE, true)); ASSERT_EQ(0, tracefs_enable_event(TRACEFS_DENY_ACCESS_FS_ENABLE, true)); ASSERT_EQ(0, tracefs_clear()); @@ -134,7 +135,7 @@ FIXTURE_TEARDOWN(trace_fs) return; set_cap(_metadata, CAP_SYS_ADMIN); - tracefs_enable_event(TRACEFS_ADD_RULE_FS_ENABLE, false); + tracefs_enable_event(TRACEFS_ADD_RULE_PATH_BENEATH_ENABLE, false); tracefs_enable_event(TRACEFS_CHECK_RULE_FS_ENABLE, false); tracefs_enable_event(TRACEFS_DENY_ACCESS_FS_ENABLE, false); tracefs_fixture_teardown(); @@ -183,11 +184,11 @@ TEST_F(trace_fs, unsandboxed) } /* - * Verifies that adding a filesystem rule emits a landlock_add_rule_fs trace - * event with the expected path and field values: ruleset ID is non-zero, - * access_rights is non-zero, and path matches. + * Verifies that adding a filesystem rule emits a landlock_add_rule_path_beneath + * event with the expected path and field values: the ruleset ID and + * access_rights are non-zero, and the path matches. */ -TEST_F(trace_fs, add_rule_fs) +TEST_F(trace_fs, add_rule_path_beneath) { struct landlock_ruleset_attr ruleset_attr = { .handled_access_fs = LANDLOCK_ACCESS_FS_READ_FILE | @@ -215,28 +216,30 @@ TEST_F(trace_fs, add_rule_fs) buf = tracefs_read_buf(); ASSERT_NE(NULL, buf); - count = tracefs_count_matches(buf, REGEX_ADD_RULE_FS(TRACE_TASK)); + count = tracefs_count_matches(buf, + REGEX_ADD_RULE_PATH_BENEATH(TRACE_TASK)); EXPECT_EQ(1, count) { - TH_LOG("Expected 1 add_rule_fs event, got %d\n%s", count, buf); + TH_LOG("Expected 1 add_rule_path_beneath event, got %d\n%s", + count, buf); } /* Ruleset ID should be non-zero. */ - ASSERT_EQ(0, tracefs_extract_field(buf, REGEX_ADD_RULE_FS(TRACE_TASK), - "ruleset", field_buf, - sizeof(field_buf))); + ASSERT_EQ(0, tracefs_extract_field( + buf, REGEX_ADD_RULE_PATH_BENEATH(TRACE_TASK), + "ruleset", field_buf, sizeof(field_buf))); EXPECT_STRNE("0", field_buf); /* Access rights should be non-zero. */ - ASSERT_EQ(0, tracefs_extract_field(buf, REGEX_ADD_RULE_FS(TRACE_TASK), - "access_rights", field_buf, - sizeof(field_buf))); + ASSERT_EQ(0, tracefs_extract_field( + buf, REGEX_ADD_RULE_PATH_BENEATH(TRACE_TASK), + "access_rights", field_buf, sizeof(field_buf))); EXPECT_STRNE("", field_buf); /* Path should be /usr. */ - ASSERT_EQ(0, - tracefs_extract_field(buf, REGEX_ADD_RULE_FS(TRACE_TASK), - "path", field_buf, sizeof(field_buf))); + ASSERT_EQ(0, tracefs_extract_field( + buf, REGEX_ADD_RULE_PATH_BENEATH(TRACE_TASK), + "path", field_buf, sizeof(field_buf))); EXPECT_STREQ("/usr", field_buf); free(buf); @@ -246,7 +249,7 @@ TEST_F(trace_fs, add_rule_fs) * Verifies that a path whose escaping exceeds the trace scratch sequence does * not corrupt a sibling symbolic field. */ -TEST_F(trace_fs, add_rule_fs_escaped_path_overflow) +TEST_F(trace_fs, add_rule_path_beneath_escaped_path_overflow) { static const char access_prefix[] = "execute|write_file|read_file|"; static const char access_suffix[] = "|ioctl_dev|resolve_unix"; @@ -277,10 +280,12 @@ TEST_F(trace_fs, add_rule_fs_escaped_path_overflow) buf = tracefs_read_buf(); ASSERT_NE(NULL, buf); - count = tracefs_count_matches(buf, REGEX_ADD_RULE_FS(TRACE_TASK)); + count = tracefs_count_matches(buf, + REGEX_ADD_RULE_PATH_BENEATH(TRACE_TASK)); EXPECT_EQ(1, count) { - TH_LOG("Expected 1 add_rule_fs event, got %d\n%s", count, buf); + TH_LOG("Expected 1 add_rule_path_beneath event, got %d\n%s", + count, buf); } /* @@ -288,9 +293,9 @@ TEST_F(trace_fs, add_rule_fs_escaped_path_overflow) * field also catches scratch-sequence poisoning when the compiler * evaluates the overflowing path first, as GCC currently does. */ - ASSERT_EQ(0, tracefs_extract_field(buf, REGEX_ADD_RULE_FS(TRACE_TASK), - "access_rights", field_buf, - sizeof(field_buf))); + ASSERT_EQ(0, tracefs_extract_field( + buf, REGEX_ADD_RULE_PATH_BENEATH(TRACE_TASK), + "access_rights", field_buf, sizeof(field_buf))); EXPECT_EQ(0, strncmp(field_buf, access_prefix, sizeof(access_prefix) - 1)); EXPECT_EQ(NULL, strstr(field_buf, "|refer|")); @@ -298,7 +303,8 @@ TEST_F(trace_fs, add_rule_fs_escaped_path_overflow) ASSERT_LE(sizeof(access_suffix) - 1, field_len); EXPECT_STREQ(access_suffix, field_buf + field_len - (sizeof(access_suffix) - 1)); - expect_truncated_path(_metadata, buf, REGEX_ADD_RULE_FS(TRACE_TASK)); + expect_truncated_path(_metadata, buf, + REGEX_ADD_RULE_PATH_BENEATH(TRACE_TASK)); free(buf); } diff --git a/tools/testing/selftests/landlock/trace_test.c b/tools/testing/selftests/landlock/trace_test.c index afdaf8511b3a..f9b293a9dd56 100644 --- a/tools/testing/selftests/landlock/trace_test.c +++ b/tools/testing/selftests/landlock/trace_test.c @@ -49,8 +49,10 @@ FIXTURE_SETUP(trace) ASSERT_EQ(0, tracefs_enable_event(TRACEFS_CREATE_RULESET_ENABLE, true)); ASSERT_EQ(0, tracefs_enable_event(TRACEFS_CREATE_DOMAIN_ENABLE, true)); ASSERT_EQ(0, tracefs_enable_event(TRACEFS_ENFORCE_DOMAIN_ENABLE, true)); - ASSERT_EQ(0, tracefs_enable_event(TRACEFS_ADD_RULE_FS_ENABLE, true)); - ASSERT_EQ(0, tracefs_enable_event(TRACEFS_ADD_RULE_NET_ENABLE, true)); + ASSERT_EQ(0, tracefs_enable_event(TRACEFS_ADD_RULE_PATH_BENEATH_ENABLE, + true)); + ASSERT_EQ(0, + tracefs_enable_event(TRACEFS_ADD_RULE_NET_PORT_ENABLE, true)); ASSERT_EQ(0, tracefs_enable_event(TRACEFS_CHECK_RULE_FS_ENABLE, true)); ASSERT_EQ(0, tracefs_enable_event(TRACEFS_CHECK_RULE_NET_ENABLE, true)); ASSERT_EQ(0, tracefs_enable_event(TRACEFS_DENY_ACCESS_FS_ENABLE, true)); @@ -72,8 +74,8 @@ FIXTURE_TEARDOWN(trace) tracefs_enable_event(TRACEFS_CREATE_RULESET_ENABLE, false); tracefs_enable_event(TRACEFS_CREATE_DOMAIN_ENABLE, false); tracefs_enable_event(TRACEFS_ENFORCE_DOMAIN_ENABLE, false); - tracefs_enable_event(TRACEFS_ADD_RULE_FS_ENABLE, false); - tracefs_enable_event(TRACEFS_ADD_RULE_NET_ENABLE, false); + tracefs_enable_event(TRACEFS_ADD_RULE_PATH_BENEATH_ENABLE, false); + tracefs_enable_event(TRACEFS_ADD_RULE_NET_PORT_ENABLE, false); tracefs_enable_event(TRACEFS_CHECK_RULE_FS_ENABLE, false); tracefs_enable_event(TRACEFS_CHECK_RULE_NET_ENABLE, false); tracefs_enable_event(TRACEFS_DENY_ACCESS_FS_ENABLE, false); @@ -103,8 +105,10 @@ TEST_F(trace, no_trace_when_disabled) ASSERT_EQ(0, tracefs_enable_event(TRACEFS_CREATE_DOMAIN_ENABLE, false)); ASSERT_EQ(0, tracefs_enable_event(TRACEFS_ENFORCE_DOMAIN_ENABLE, false)); - ASSERT_EQ(0, tracefs_enable_event(TRACEFS_ADD_RULE_FS_ENABLE, false)); - ASSERT_EQ(0, tracefs_enable_event(TRACEFS_ADD_RULE_NET_ENABLE, false)); + ASSERT_EQ(0, tracefs_enable_event(TRACEFS_ADD_RULE_PATH_BENEATH_ENABLE, + false)); + ASSERT_EQ(0, tracefs_enable_event(TRACEFS_ADD_RULE_NET_PORT_ENABLE, + false)); ASSERT_EQ(0, tracefs_enable_event(TRACEFS_CHECK_RULE_FS_ENABLE, false)); ASSERT_EQ(0, tracefs_enable_event(TRACEFS_CHECK_RULE_NET_ENABLE, false)); @@ -265,10 +269,11 @@ TEST_F(trace, ruleset_version) ASSERT_NE(0, !!dot); EXPECT_STREQ("0", dot + 1); - /* Verify 2 add_rule_fs events were emitted. */ - EXPECT_EQ(2, tracefs_count_matches(buf, REGEX_ADD_RULE_FS(TRACE_TASK))) + /* Verify two add_rule_path_beneath events were emitted. */ + EXPECT_EQ(2, tracefs_count_matches( + buf, REGEX_ADD_RULE_PATH_BENEATH(TRACE_TASK))) { - TH_LOG("Expected 2 add_rule_fs events\n%s", buf); + TH_LOG("Expected 2 add_rule_path_beneath events\n%s", buf); } /* @@ -373,7 +378,7 @@ TEST_F(trace, create_domain) tracefs_count_matches(buf, REGEX_CHECK_RULE_FS(TRACE_TASK)); ASSERT_LE(1, check_count) { - TH_LOG("Expected check_rule_fs events\n%s", buf); + TH_LOG("Expected check_rule_inode events\n%s", buf); } EXPECT_EQ(0, tracefs_extract_field(buf, REGEX_CHECK_RULE_FS(TRACE_TASK), @@ -508,9 +513,11 @@ TEST_F(trace, add_rule_invalid_fd) buf = tracefs_read_buf(); ASSERT_NE(NULL, buf); - EXPECT_EQ(0, tracefs_count_matches(buf, REGEX_ADD_RULE_FS(TRACE_TASK))) + EXPECT_EQ(0, tracefs_count_matches( + buf, REGEX_ADD_RULE_PATH_BENEATH(TRACE_TASK))) { - TH_LOG("No add_rule_fs event expected on invalid fd\n%s", buf); + TH_LOG("No add_rule_path_beneath event expected on invalid fd\n%s", + buf); } free(buf); @@ -902,10 +909,10 @@ TEST_F(trace, non_audit_visible_denial_counting) } /* - * Verifies that landlock_add_rule_net emits a trace event with the correct port - * and allowed access mask fields. + * Verifies that landlock_add_rule_net_port emits a trace event with the correct + * port and allowed access mask fields. */ -TEST_F(trace, add_rule_net_fields) +TEST_F(trace, add_rule_net_port_fields) { struct landlock_ruleset_attr ruleset_attr = { .handled_access_net = LANDLOCK_ACCESS_NET_BIND_TCP, @@ -931,9 +938,10 @@ TEST_F(trace, add_rule_net_fields) buf = tracefs_read_buf(); ASSERT_NE(NULL, buf); - EXPECT_EQ(1, tracefs_count_matches(buf, REGEX_ADD_RULE_NET(TRACE_TASK))) + EXPECT_EQ(1, tracefs_count_matches(buf, + REGEX_ADD_RULE_NET_PORT(TRACE_TASK))) { - TH_LOG("Expected 1 add_rule_net event\n%s", buf); + TH_LOG("Expected 1 add_rule_net_port event\n%s", buf); } /* @@ -941,7 +949,8 @@ TEST_F(trace, add_rule_net_fields) * (landlock_net_port_attr.port). On little-endian, htons(8080) is * 36895, so this comparison catches byte-order bugs. */ - EXPECT_EQ(0, tracefs_extract_field(buf, REGEX_ADD_RULE_NET(TRACE_TASK), + EXPECT_EQ(0, tracefs_extract_field(buf, + REGEX_ADD_RULE_NET_PORT(TRACE_TASK), "port", field, sizeof(field))); EXPECT_STREQ("8080", field); /* @@ -950,9 +959,9 @@ TEST_F(trace, add_rule_net_fields) * net access bits are unhandled because the ruleset only handles * BIND_TCP). */ - EXPECT_EQ(0, - tracefs_extract_field(buf, REGEX_ADD_RULE_NET(TRACE_TASK), - "access_rights", field, sizeof(field))); + EXPECT_EQ(0, tracefs_extract_field( + buf, REGEX_ADD_RULE_NET_PORT(TRACE_TASK), + "access_rights", field, sizeof(field))); EXPECT_STREQ("bind_tcp|connect_tcp|bind_udp|connect_send_udp", field); free(buf); From 98b04ab00f0e738d69bd718228db55483a911b7a Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Micka=C3=ABl=20Sala=C3=BCn?= Date: Fri, 18 Sep 2026 20:50:27 +0200 Subject: [PATCH 242/694] landlock: Fix network denial trace context MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Network denial events report source and destination ports reconstructed from audit data. Their zero values are ambiguous, and neither identifies the complete endpoint that Landlock checked. Carry the checked sockaddr and its signed length in a private trace-only context. For an enabled event, validate the length and copy only the initialized prefix into zeroed local storage. This prevents a typed BPF program from reading uninitialized bytes while exposing the socket family, socket, address, and length. Replace the source and destination trace-record fields with one signed port derived from the checked address. A value of -1 means that no port was checked, zero is a valid port, and positive values use host endianness. Bind blockers select the bind address; connect and send blockers select the destination. Cc: Günther Noack Cc: Steven Rostedt Fixes: 01ce260f5ccf ("landlock: Add landlock_deny_access_fs and landlock_deny_access_net") Link: https://patch.msgid.link/20260918185036.608651-5-mic@digikod.net Signed-off-by: Mickaël Salaün --- include/trace/events/landlock.h | 73 +++++++++++++-------- security/landlock/log.h | 32 +++++++-- security/landlock/net.c | 37 ++++++++--- security/landlock/trace.c | 20 +++++- tools/testing/selftests/landlock/net_test.c | 49 +++++--------- tools/testing/selftests/landlock/trace.h | 3 +- 6 files changed, 135 insertions(+), 79 deletions(-) diff --git a/include/trace/events/landlock.h b/include/trace/events/landlock.h index b7e6a6638822..a982a7cfa881 100644 --- a/include/trace/events/landlock.h +++ b/include/trace/events/landlock.h @@ -10,7 +10,10 @@ #if !defined(_TRACE_LANDLOCK_H) || defined(TRACE_HEADER_MULTI_READ) #define _TRACE_LANDLOCK_H +#include +#include #include +#include #include #include #include @@ -790,6 +793,11 @@ TRACE_EVENT(landlock_deny_access_fs, __get_dynamic_array_len(pathname) - 1)) ); +static_assert(offsetof(struct sockaddr_in, sin_port) == + offsetof(struct sockaddr_in6, sin6_port)); +static_assert(sizeof_field(struct sockaddr_in, sin_port) == + sizeof_field(struct sockaddr_in6, sin6_port)); + /** * landlock_deny_access_net - Network access denied * @@ -798,30 +806,31 @@ TRACE_EVENT(landlock_deny_access_fs, * @same_exec: Whether the current task entered the denying domain itself. * @logged: The domain's audit-logging decision for this denial. * @blockers: Request type and final missing access subset (never NULL). - * @sk: Socket object (never NULL), read without a socket lock, so its - * fields are a best-effort snapshot. The denied endpoint is not - * available: the hook runs before :manpage:`bind(2)` / - * :manpage:`connect(2)` sets the socket addresses. - * @sport: Source port in host endianness, set for bind denials (zero for - * an autobind/ephemeral port); zero for connect and send denials. - * @dport: Destination port in host endianness, set for connect and send - * denials; zero for bind denials, and also zero for a UDP send to - * an AF_UNSPEC address on an IPv6 socket (indistinguishable from a - * real destination port 0). The bind-vs-connect direction is - * given by @blockers, not by which port is set. + * @sk: Socket object (never NULL), read without a socket lock, so its fields + * are a best-effort snapshot. + * @socket_family: Socket-family snapshot used by the verdict. + * @address: Authoritative address checked by the verdict (never NULL). + * The producer copies @addrlen bytes from the checked address and + * zeroes the remaining storage before emission. The + * &sockaddr_in.sin_port or &sockaddr_in6.sin6_port member, when + * present, remains in network endianness. + * @addrlen: Validated signed length of @address. * - * Emitted when a Landlock domain denies a network operation. - * - * The port fields are converted from the socket's network byte order to - * host endianness before emitting. + * Emitted when a Landlock domain denies a network operation. The blocker + * identifies whether the address is a bind or connect/send policy object. + * The flattened port field is converted from the checked address to host + * endianness, or is -1 when no port was checked. Zero is a valid checked + * port. */ TRACE_EVENT(landlock_deny_access_net, TP_PROTO(const struct landlock_hierarchy *hierarchy, bool same_exec, bool logged, const struct landlock_blockers *blockers, - const struct sock *sk, u64 sport, u64 dport), + const struct sock *sk, u16 socket_family, + const struct sockaddr_storage *address, int addrlen), - TP_ARGS(hierarchy, same_exec, logged, blockers, sk, sport, dport), + TP_ARGS(hierarchy, same_exec, logged, blockers, sk, socket_family, + address, addrlen), TP_STRUCT__entry( __field( u64, domain_id ) @@ -829,26 +838,36 @@ TRACE_EVENT(landlock_deny_access_net, __field( bool, logged ) __field( enum landlock_request_type, blockers_type ) __field( access_mask_t, blockers_access ) - __field( u64, sport ) - __field( u64, dport ) + __field( s64, port ) ), TP_fast_assign( + const struct sockaddr *const addr = + (const struct sockaddr *)address; + const bool has_port = + addrlen >= (int)offsetofend(struct sockaddr_in, sin_port) && + (addr->sa_family == AF_INET || + addr->sa_family == AF_INET6 || + (addr->sa_family == AF_UNSPEC && + socket_family == AF_INET)); + __entry->domain_id = hierarchy->id; __entry->same_exec = same_exec; __entry->logged = logged; __entry->blockers_type = blockers->type; __entry->blockers_access = blockers->access; - __entry->sport = sport; - __entry->dport = dport; + __entry->port = + has_port ? + ntohs(((const struct sockaddr_in *)addr)->sin_port) : + -1; ), - TP_printk("domain=%llx same_exec=%d logged=%d blockers=%s sport=%llu dport=%llu", - __entry->domain_id, __entry->same_exec, __entry->logged, - __entry->blockers_type == LANDLOCK_REQUEST_NET_ACCESS ? - __print_flags(__entry->blockers_access, "|", _LANDLOCK_ACCESS_NET_NAMES) : - "unknown", - __entry->sport, __entry->dport) + TP_printk("domain=%llx same_exec=%d logged=%d blockers=%s port=%lld", + __entry->domain_id, __entry->same_exec, __entry->logged, + __entry->blockers_type == LANDLOCK_REQUEST_NET_ACCESS ? + __print_flags(__entry->blockers_access, "|", _LANDLOCK_ACCESS_NET_NAMES) : + "unknown", + __entry->port) ); /** diff --git a/security/landlock/log.h b/security/landlock/log.h index 04f3e241e765..4587c2b1566d 100644 --- a/security/landlock/log.h +++ b/security/landlock/log.h @@ -15,6 +15,7 @@ struct landlock_cred_security; struct landlock_hierarchy; +struct sockaddr; enum landlock_request_type { LANDLOCK_REQUEST_PTRACE = 1, @@ -30,6 +31,16 @@ struct landlock_blockers { enum landlock_request_type type; }; +#ifdef CONFIG_TRACEPOINTS + +struct landlock_net_trace { + const struct sockaddr *address; + int addrlen; + u16 socket_family; +}; + +#endif /* CONFIG_TRACEPOINTS */ + /* * We should be careful to only use a variable of this type for * landlock_log_denial(). This way, the compiler can remove it entirely if @@ -57,13 +68,20 @@ struct landlock_request { deny_masks_t deny_masks; optional_access_t quiet_optional_accesses; - /* - * Other-party domain ID for a relational (scope/ptrace) denial, or 0 if - * that party is unsandboxed. An ID, not a pointer: the other task can - * replace its credential and free the domain it referenced. Trace path - * only; audit ignores it. - */ - u64 other_domain_id; + union { + /* + * Other-party domain ID for a relational (scope/ptrace) denial, + * or 0 if that party is unsandboxed. Store an ID, not a + * pointer: the other task can replace its credential and free + * the domain it referenced. Trace-only; audit ignores it. + */ + u64 other_domain_id; + +#ifdef CONFIG_TRACEPOINTS + /* Synchronous context for a network denial. */ + const struct landlock_net_trace *trace_net; +#endif /* CONFIG_TRACEPOINTS */ + }; }; #ifdef CONFIG_SECURITY_LANDLOCK_LOG diff --git a/security/landlock/net.c b/security/landlock/net.c index 5552c60388f8..6fe0dbde3b78 100644 --- a/security/landlock/net.c +++ b/security/landlock/net.c @@ -93,7 +93,7 @@ static int current_check_access_socket(struct socket *const sock, return 0; /* Checks for minimal header length to safely read sa_family. */ - if (addrlen < offsetofend(typeof(*address), sa_family)) + if (addrlen < (int)offsetofend(typeof(*address), sa_family)) return -EINVAL; /* @@ -145,6 +145,17 @@ static int current_check_access_socket(struct socket *const sock, .audit.u.net = &audit_net, .access = access_request, .layer_masks = &layer_masks, +#ifdef CONFIG_TRACEPOINTS + .trace_net = + &(struct landlock_net_trace){ + .address = + address, + .addrlen = + addrlen, + .socket_family = + sock_family, + }, +#endif /* CONFIG_TRACEPOINTS */ }); return -EACCES; } @@ -276,14 +287,22 @@ static int current_check_access_socket(struct socket *const sock, audit_net.family = address->sa_family; audit_net.sk = sock->sk; - landlock_log_denial(subject, - &(struct landlock_request){ - .type = LANDLOCK_REQUEST_NET_ACCESS, - .audit.type = LSM_AUDIT_DATA_NET, - .audit.u.net = &audit_net, - .access = access_request, - .layer_masks = &layer_masks, - }); + landlock_log_denial( + subject, &(struct landlock_request){ + .type = LANDLOCK_REQUEST_NET_ACCESS, + .audit.type = LSM_AUDIT_DATA_NET, + .audit.u.net = &audit_net, + .access = access_request, + .layer_masks = &layer_masks, +#ifdef CONFIG_TRACEPOINTS + .trace_net = + &(struct landlock_net_trace){ + .address = address, + .addrlen = addrlen, + .socket_family = sock_family, + }, +#endif /* CONFIG_TRACEPOINTS */ + }); return -EACCES; } diff --git a/security/landlock/trace.c b/security/landlock/trace.c index 58276cc32d3f..9be86638f905 100644 --- a/security/landlock/trace.c +++ b/security/landlock/trace.c @@ -12,6 +12,7 @@ #include #include #include +#include #include #include "access.h" @@ -157,16 +158,31 @@ void landlock_trace_denial( break; case LANDLOCK_REQUEST_NET_ACCESS: if (trace_landlock_deny_access_net_enabled()) { + const struct landlock_net_trace *const trace_net = + request->trace_net; const struct landlock_blockers blockers = { .access = missing, .type = request->type, }; + struct sockaddr_storage address = {}; + if (WARN_ON_ONCE(!trace_net || !trace_net->address)) + return; + + if (WARN_ON_ONCE( + trace_net->addrlen < + (int)offsetofend(struct sockaddr, + sa_family) || + trace_net->addrlen > (int)sizeof(address))) + return; + + memcpy(&address, trace_net->address, + trace_net->addrlen); trace_landlock_deny_access_net( youngest_denied, same_exec, logged, &blockers, request->audit.u.net->sk, - ntohs(request->audit.u.net->sport), - ntohs(request->audit.u.net->dport)); + trace_net->socket_family, &address, + trace_net->addrlen); } break; case LANDLOCK_REQUEST_PTRACE: diff --git a/tools/testing/selftests/landlock/net_test.c b/tools/testing/selftests/landlock/net_test.c index 4fb705e1596d..28942438e154 100644 --- a/tools/testing/selftests/landlock/net_test.c +++ b/tools/testing/selftests/landlock/net_test.c @@ -3482,8 +3482,8 @@ TEST_F(trace_net, deny_access_net_bind) /* * Anchors the denial fields shared by every deny_access_net event so a field - * test proves more than sport/dport: the denying domain, the same-exec bit, the - * audit-logging verdict, and the blocked access all stay populated. + * test proves more than the checked endpoint: the denying domain, the same-exec + * bit, the audit-logging verdict, and the blocked access all stay populated. */ static void expect_net_deny_common_fields(struct __test_metadata *const _metadata, @@ -3569,38 +3569,38 @@ FIXTURE_VARIANT(trace_net_connect) { bool deny_connect; }; +/* Denied connect() to the next IPv4 port. */ /* clang-format off */ - -/* Denied connect(): sport=0, dport=. */ FIXTURE_VARIANT_ADD(trace_net_connect, connect_denied) { + /* clang-format on */ .handled = LANDLOCK_ACCESS_NET_CONNECT_TCP, .bind_base_first = false, .deny_connect = true, }; -/* Denied bind(): sport=, dport=0. */ +/* Denied bind() to the next IPv4 port. */ +/* clang-format off */ FIXTURE_VARIANT_ADD(trace_net_connect, bind_fields) { + /* clang-format on */ .handled = LANDLOCK_ACCESS_NET_BIND_TCP, .bind_base_first = false, .deny_connect = false, }; -/* Denied connect() after an allowed bind(): the connect fields (sport=0). */ +/* Denied connect() after an allowed bind() uses the checked destination. */ +/* clang-format off */ FIXTURE_VARIANT_ADD(trace_net_connect, connect_after_bind) { - .handled = LANDLOCK_ACCESS_NET_BIND_TCP | LANDLOCK_ACCESS_NET_CONNECT_TCP, + /* clang-format on */ + .handled = LANDLOCK_ACCESS_NET_BIND_TCP | + LANDLOCK_ACCESS_NET_CONNECT_TCP, .bind_base_first = true, .deny_connect = true, }; -/* clang-format on */ - /* - * A denied TCP bind(2) or connect(2) emits one deny_access_net event. The port - * is reported in the field matching the denied operation, in host endianness - * (the UAPI landlock_net_port_attr.port convention): a connect denial reports - * sport=0 dport=, a bind denial reports sport= dport=0, so a - * byte-order or field-swap bug is caught. A prior allowed bind - * (connect_after_bind) does not change the connect denial's fields. + * A denied TCP bind(2) or connect(2) emits one deny_access_net event with the + * checked IPv4 port in host endianness (the UAPI landlock_net_port_attr.port + * convention). A prior allowed bind does not change a connect denial's port. */ TEST_F(trace_net_connect, deny_access_net) { @@ -3693,21 +3693,13 @@ TEST_F(trace_net_connect, deny_access_net) expect_net_deny_common_fields(_metadata, buf); - /* - * The denied operation's port field carries the port; the other is 0. - */ snprintf(expected, sizeof(expected), "%llu", (unsigned long long)(sock_port_start + 1)); ASSERT_EQ(0, tracefs_extract_field(buf, REGEX_DENY_ACCESS_NET(TRACE_TASK), - "sport", field, sizeof(field))); - EXPECT_STREQ(variant->deny_connect ? "0" : expected, field); - - ASSERT_EQ(0, - tracefs_extract_field(buf, REGEX_DENY_ACCESS_NET(TRACE_TASK), - "dport", field, sizeof(field))); - EXPECT_STREQ(variant->deny_connect ? expected : "0", field); + "port", field, sizeof(field))); + EXPECT_STREQ(expected, field); free(buf); } @@ -3867,11 +3859,4 @@ TEST_F(trace_net_check_rule, check_rule_net_fields) free(buf); } -/* - * IPv6 network trace tests are intentionally elided. IPv6 hook dispatch uses - * the same current_check_access_socket() code path as IPv4, validated by the - * audit tests in this file. The trace events use the same blockers/sport/dport - * fields regardless of address family. - */ - TEST_HARNESS_MAIN diff --git a/tools/testing/selftests/landlock/trace.h b/tools/testing/selftests/landlock/trace.h index e6873853376c..2ec863362173 100644 --- a/tools/testing/selftests/landlock/trace.h +++ b/tools/testing/selftests/landlock/trace.h @@ -145,8 +145,7 @@ "same_exec=[01] " \ "logged=[01] " \ "blockers=[a-z_|]* " \ - "sport=[0-9]\\+ " \ - "dport=[0-9]\\+$" + "port=-\\?[0-9]\\+$" #define REGEX_DENY_PTRACE(task) \ TRACE_PREFIX(task) \ From 7ad69ac63315506e2091bf46d5c96c49f6ce439e Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Micka=C3=ABl=20Sala=C3=BCn?= Date: Fri, 18 Sep 2026 20:50:28 +0200 Subject: [PATCH 243/694] landlock: Report the actual ptrace tracer MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The ptrace denial callback identifies only the tracee. Current is the tracer during hook_ptrace_access_check(), but it is the tracee during PTRACE_TRACEME, where the parent is the actual tracer. A consumer therefore cannot infer both parties from the existing arguments. Append the actual tracer task to the typed-BPF callback: current for hook_ptrace_access_check() and parent for hook_ptrace_traceme(). Carry it with the tracee domain ID in a private ptrace context. Both hooks keep the selected tasks alive through synchronous dispatch, so no extra task reference is needed. Keep the tracefs record unchanged. The new context is available only to typed BPF, while same_exec continues to describe the tracer that owns the denying policy. Cc: Günther Noack Cc: Steven Rostedt Fixes: bb91730f16c0 ("landlock: Add tracepoints for ptrace and scope denials") Link: https://patch.msgid.link/20260918185036.608651-6-mic@digikod.net Signed-off-by: Mickaël Salaün --- include/trace/events/landlock.h | 11 ++++++++--- security/landlock/log.h | 16 ++++++++++++---- security/landlock/task.c | 28 +++++++++++++++++----------- security/landlock/trace.c | 20 ++++++++++++++------ 4 files changed, 51 insertions(+), 24 deletions(-) diff --git a/include/trace/events/landlock.h b/include/trace/events/landlock.h index a982a7cfa881..8c6ebf958d66 100644 --- a/include/trace/events/landlock.h +++ b/include/trace/events/landlock.h @@ -295,6 +295,8 @@ static inline const char *__trace_landlock_print_layers( * the two parties without kernel-internal state. The ID is a scalar * snapshot, not a live domain pointer that could dangle: an optional * relational referent is a scalar (0 sentinel), not a nullable pointer. + * For ptrace, same_exec instead describes the tracer, even for + * PTRACE_TRACEME, and may differ from the current task. * * Blocker fields * ~~~~~~~~~~~~~~ @@ -875,12 +877,14 @@ TRACE_EVENT(landlock_deny_access_net, * * @hierarchy: Denying domain's hierarchy node (never NULL); its id is the * domain field. - * @same_exec: Whether the current task entered the denying domain itself. + * @same_exec: Whether the tracer entered the denying domain itself. * @logged: The domain's audit-logging decision for this denial. * @tracee_domain_id: The tracee's Landlock domain ID, or 0 if the tracee * is unsandboxed. * @tracee: The target task ptrace acted on (never NULL). tracee_pid is * the init-namespace TGID (like audit's opid). + * @tracer: The tracer or proposed tracer (never NULL); for PTRACE_TRACEME + * this is the parent, not the syscall caller. * * Emitted when a Landlock domain denies a ptrace operation. */ @@ -888,9 +892,10 @@ TRACE_EVENT(landlock_deny_ptrace, TP_PROTO(const struct landlock_hierarchy *hierarchy, bool same_exec, bool logged, u64 tracee_domain_id, - const struct task_struct *tracee), + const struct task_struct *tracee, + const struct task_struct *tracer), - TP_ARGS(hierarchy, same_exec, logged, tracee_domain_id, tracee), + TP_ARGS(hierarchy, same_exec, logged, tracee_domain_id, tracee, tracer), TP_STRUCT__entry( __field( u64, domain_id ) diff --git a/security/landlock/log.h b/security/landlock/log.h index 4587c2b1566d..821df6f711c6 100644 --- a/security/landlock/log.h +++ b/security/landlock/log.h @@ -16,6 +16,7 @@ struct landlock_cred_security; struct landlock_hierarchy; struct sockaddr; +struct task_struct; enum landlock_request_type { LANDLOCK_REQUEST_PTRACE = 1, @@ -39,6 +40,11 @@ struct landlock_net_trace { u16 socket_family; }; +struct landlock_ptrace_trace { + u64 tracee_domain_id; + const struct task_struct *tracer; +}; + #endif /* CONFIG_TRACEPOINTS */ /* @@ -70,16 +76,18 @@ struct landlock_request { union { /* - * Other-party domain ID for a relational (scope/ptrace) denial, - * or 0 if that party is unsandboxed. Store an ID, not a - * pointer: the other task can replace its credential and free - * the domain it referenced. Trace-only; audit ignores it. + * Other-party domain ID for a scope denial, or 0 if that party + * is unsandboxed. Store an ID, not a pointer: the other task + * can replace its credential and free the domain it referenced. + * Audit ignores this trace-only field. */ u64 other_domain_id; #ifdef CONFIG_TRACEPOINTS /* Synchronous context for a network denial. */ const struct landlock_net_trace *trace_net; + /* Consumed only by the synchronous trace dispatcher. */ + const struct landlock_ptrace_trace *trace_ptrace; #endif /* CONFIG_TRACEPOINTS */ }; }; diff --git a/security/landlock/task.c b/security/landlock/task.c index 4491ce31ae04..445f6b921a87 100644 --- a/security/landlock/task.c +++ b/security/landlock/task.c @@ -88,7 +88,9 @@ static int hook_ptrace_access_check(struct task_struct *const child, const unsigned int mode) { const struct landlock_cred_security *parent_subject; +#ifdef CONFIG_TRACEPOINTS u64 tracee_domain_id = 0; +#endif /* CONFIG_TRACEPOINTS */ int err; /* Quick return for non-landlocked tasks. */ @@ -100,10 +102,10 @@ static int hook_ptrace_access_check(struct task_struct *const child, const struct landlock_domain *const child_dom = landlock_get_task_domain(child); err = domain_ptrace(parent_subject->domain, child_dom); -#ifdef CONFIG_SECURITY_LANDLOCK_LOG +#ifdef CONFIG_TRACEPOINTS if (child_dom) tracee_domain_id = child_dom->hierarchy->id; -#endif /* CONFIG_SECURITY_LANDLOCK_LOG */ +#endif /* CONFIG_TRACEPOINTS */ } if (!err) @@ -121,7 +123,12 @@ static int hook_ptrace_access_check(struct task_struct *const child, .u.tsk = child, }, .layer_plus_one = parent_subject->domain->num_layers, - .other_domain_id = tracee_domain_id, +#ifdef CONFIG_TRACEPOINTS + .trace_ptrace = &(struct landlock_ptrace_trace) { + .tracee_domain_id = tracee_domain_id, + .tracer = current, + }, +#endif /* CONFIG_TRACEPOINTS */ }); return err; @@ -142,7 +149,6 @@ static int hook_ptrace_traceme(struct task_struct *const parent) { const struct landlock_cred_security *parent_subject; const struct landlock_domain *child_dom; - u64 tracee_domain_id = 0; int err; child_dom = landlock_get_current_domain(); @@ -154,12 +160,6 @@ static int hook_ptrace_traceme(struct task_struct *const parent) if (!err) return 0; -#ifdef CONFIG_SECURITY_LANDLOCK_LOG - /* The tracee is the current task; its domain is stable here. */ - if (child_dom) - tracee_domain_id = child_dom->hierarchy->id; -#endif /* CONFIG_SECURITY_LANDLOCK_LOG */ - /* * For the ptrace_traceme case, we log the domain which is the cause of * the denial, which means the parent domain instead of the current @@ -174,7 +174,13 @@ static int hook_ptrace_traceme(struct task_struct *const parent) .u.tsk = current, }, .layer_plus_one = parent_subject->domain->num_layers, - .other_domain_id = tracee_domain_id, +#ifdef CONFIG_TRACEPOINTS + .trace_ptrace = &(struct landlock_ptrace_trace) { + /* The current task's domain is stable here. */ + .tracee_domain_id = child_dom ? child_dom->hierarchy->id : 0, + .tracer = parent, + }, +#endif /* CONFIG_TRACEPOINTS */ }); return err; } diff --git a/security/landlock/trace.c b/security/landlock/trace.c index 9be86638f905..43091c052f77 100644 --- a/security/landlock/trace.c +++ b/security/landlock/trace.c @@ -63,7 +63,7 @@ void landlock_trace_free_domain(const struct landlock_hierarchy *const hierarchy * @request: Detail of the user space request. * @youngest_denied: The youngest hierarchy node that denied the access. * @missing: The final missing access subset, when applicable. - * @same_exec: Whether the current task is the same executable that called + * @same_exec: Whether the policy subject is the same executable that called * landlock_restrict_self() for the denying domain, as computed * by landlock_log_denial(). * @logged: Whether the domain's policy selects this denial for logging, as @@ -186,11 +186,19 @@ void landlock_trace_denial( } break; case LANDLOCK_REQUEST_PTRACE: - if (trace_landlock_deny_ptrace_enabled()) - trace_landlock_deny_ptrace(youngest_denied, same_exec, - logged, - request->other_domain_id, - request->audit.u.tsk); + if (trace_landlock_deny_ptrace_enabled()) { + const struct landlock_ptrace_trace *const trace_ptrace = + request->trace_ptrace; + + if (WARN_ON_ONCE(!trace_ptrace || + !trace_ptrace->tracer)) + return; + + trace_landlock_deny_ptrace( + youngest_denied, same_exec, logged, + trace_ptrace->tracee_domain_id, + request->audit.u.tsk, trace_ptrace->tracer); + } break; case LANDLOCK_REQUEST_SCOPE_SIGNAL: if (trace_landlock_deny_scope_signal_enabled()) From 0889db596a25ecde210e66fa0db70bc6a6d91f6c Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Micka=C3=ABl=20Sala=C3=BCn?= Date: Fri, 18 Sep 2026 20:50:29 +0200 Subject: [PATCH 244/694] landlock: Report the effective signal number MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The signal-scope denial callback identifies its target but not the effective signal. This loses permission-probe signal zero and makes the file-owner hook's zero sentinel ambiguous. Append an int signal argument to the typed-BPF callback. Preserve sig, including zero, in hook_task_kill(). In hook_file_send_sigiotask(), translate signum zero to SIGIO at the producer, where its meaning is known. Carry the effective signal and target domain ID in a private, stack-backed context consumed synchronously. This requires no allocation or task reference in the interrupt-capable file-owner path. Gate this context and the remaining scope-only domain IDs with CONFIG_TRACEPOINTS. Keep the tracefs record and audit output unchanged. Cc: Günther Noack Cc: Steven Rostedt Fixes: bb91730f16c0 ("landlock: Add tracepoints for ptrace and scope denials") Link: https://patch.msgid.link/20260918185036.608651-7-mic@digikod.net Signed-off-by: Mickaël Salaün --- include/trace/events/landlock.h | 8 +++-- security/landlock/log.h | 21 ++++++++---- security/landlock/task.c | 59 +++++++++++++++++++++++---------- security/landlock/trace.c | 12 +++++-- 4 files changed, 71 insertions(+), 29 deletions(-) diff --git a/include/trace/events/landlock.h b/include/trace/events/landlock.h index 8c6ebf958d66..5da0f12ed2c3 100644 --- a/include/trace/events/landlock.h +++ b/include/trace/events/landlock.h @@ -928,12 +928,14 @@ TRACE_EVENT(landlock_deny_ptrace, * * @hierarchy: Denying domain's hierarchy node (never NULL); its id is the * domain field. - * @same_exec: Whether the current task entered the denying domain itself. + * @same_exec: Whether the policy subject entered the denying domain itself. * @logged: The domain's audit-logging decision for this denial. * @target_domain_id: The target's Landlock domain ID, or 0 if the target * is unsandboxed. * @target: The task the signal was aimed at (never NULL). target_pid is * the init-namespace TGID (like audit's opid). + * @signal: The signal selected by the denied check. Zero is a permission + * probe, not an absent value. * * Emitted when a Landlock domain denies signal delivery to a scoped-out * target. @@ -942,9 +944,9 @@ TRACE_EVENT(landlock_deny_scope_signal, TP_PROTO(const struct landlock_hierarchy *hierarchy, bool same_exec, bool logged, u64 target_domain_id, - const struct task_struct *target), + const struct task_struct *target, int signal), - TP_ARGS(hierarchy, same_exec, logged, target_domain_id, target), + TP_ARGS(hierarchy, same_exec, logged, target_domain_id, target, signal), TP_STRUCT__entry( __field( u64, domain_id ) diff --git a/security/landlock/log.h b/security/landlock/log.h index 821df6f711c6..faa30e26e42a 100644 --- a/security/landlock/log.h +++ b/security/landlock/log.h @@ -45,6 +45,11 @@ struct landlock_ptrace_trace { const struct task_struct *tracer; }; +struct landlock_signal_trace { + u64 target_domain_id; + int signal; +}; + #endif /* CONFIG_TRACEPOINTS */ /* @@ -74,22 +79,24 @@ struct landlock_request { deny_masks_t deny_masks; optional_access_t quiet_optional_accesses; +#ifdef CONFIG_TRACEPOINTS union { /* - * Other-party domain ID for a scope denial, or 0 if that party - * is unsandboxed. Store an ID, not a pointer: the other task - * can replace its credential and free the domain it referenced. - * Audit ignores this trace-only field. + * Other-party domain ID for an abstract UNIX socket scope + * denial, or 0 if that party is unsandboxed. Store an ID, not + * a pointer: the other task can replace its credential and free + * the domain it referenced. */ u64 other_domain_id; -#ifdef CONFIG_TRACEPOINTS /* Synchronous context for a network denial. */ const struct landlock_net_trace *trace_net; - /* Consumed only by the synchronous trace dispatcher. */ + /* Synchronous context for a ptrace denial. */ const struct landlock_ptrace_trace *trace_ptrace; -#endif /* CONFIG_TRACEPOINTS */ + /* Synchronous context for a signal denial. */ + const struct landlock_signal_trace *trace_signal; }; +#endif /* CONFIG_TRACEPOINTS */ }; #ifdef CONFIG_SECURITY_LANDLOCK_LOG diff --git a/security/landlock/task.c b/security/landlock/task.c index 445f6b921a87..d9eae86fc552 100644 --- a/security/landlock/task.c +++ b/security/landlock/task.c @@ -256,8 +256,7 @@ static bool domain_is_scoped(const struct landlock_domain *const client, } static bool sock_is_scoped(struct sock *const other, - const struct landlock_domain *const domain, - u64 *const peer_domain_id) + const struct landlock_domain *const domain) { const struct landlock_domain *dom_other; @@ -275,13 +274,23 @@ static bool sock_is_scoped(struct sock *const other, return false; dom_other = landlock_cred(other->sk_socket->file->f_cred)->domain; -#ifdef CONFIG_SECURITY_LANDLOCK_LOG - *peer_domain_id = dom_other ? dom_other->hierarchy->id : 0; -#endif /* CONFIG_SECURITY_LANDLOCK_LOG */ return domain_is_scoped(domain, dom_other, LANDLOCK_SCOPE_ABSTRACT_UNIX_SOCKET); } +#ifdef CONFIG_TRACEPOINTS + +static u64 get_socket_domain_id(const struct sock *const other) +{ + const struct landlock_domain *domain; + + lockdep_assert_held(&unix_sk(other)->lock); + domain = landlock_cred(other->sk_socket->file->f_cred)->domain; + return domain ? domain->hierarchy->id : 0; +} + +#endif /* CONFIG_TRACEPOINTS */ + static bool is_abstract_socket(struct sock *const sock) { struct unix_address *addr = unix_sk(sock)->addr; @@ -305,7 +314,6 @@ static int hook_unix_stream_connect(struct sock *const sock, struct sock *const newsk) { size_t handle_layer; - u64 peer_domain_id = 0; const struct landlock_cred_security *const subject = landlock_get_applicable_subject(current_cred(), unix_scope, &handle_layer); @@ -317,7 +325,7 @@ static int hook_unix_stream_connect(struct sock *const sock, if (!is_abstract_socket(other)) return 0; - if (!sock_is_scoped(other, subject->domain, &peer_domain_id)) + if (!sock_is_scoped(other, subject->domain)) return 0; landlock_log_denial(subject, &(struct landlock_request) { @@ -329,7 +337,9 @@ static int hook_unix_stream_connect(struct sock *const sock, }, }, .layer_plus_one = handle_layer + 1, - .other_domain_id = peer_domain_id, +#ifdef CONFIG_TRACEPOINTS + .other_domain_id = get_socket_domain_id(other), +#endif /* CONFIG_TRACEPOINTS */ }); return -EPERM; } @@ -338,7 +348,6 @@ static int hook_unix_may_send(struct socket *const sock, struct socket *const other) { size_t handle_layer; - u64 peer_domain_id = 0; const struct landlock_cred_security *const subject = landlock_get_applicable_subject(current_cred(), unix_scope, &handle_layer); @@ -356,7 +365,7 @@ static int hook_unix_may_send(struct socket *const sock, if (!is_abstract_socket(other->sk)) return 0; - if (!sock_is_scoped(other->sk, subject->domain, &peer_domain_id)) + if (!sock_is_scoped(other->sk, subject->domain)) return 0; landlock_log_denial(subject, &(struct landlock_request) { @@ -368,7 +377,9 @@ static int hook_unix_may_send(struct socket *const sock, }, }, .layer_plus_one = handle_layer + 1, - .other_domain_id = peer_domain_id, +#ifdef CONFIG_TRACEPOINTS + .other_domain_id = get_socket_domain_id(other->sk), +#endif /* CONFIG_TRACEPOINTS */ }); return -EPERM; } @@ -383,7 +394,9 @@ static int hook_task_kill(struct task_struct *const p, { bool is_scoped; size_t handle_layer; +#ifdef CONFIG_TRACEPOINTS u64 target_domain_id = 0; +#endif /* CONFIG_TRACEPOINTS */ const struct landlock_cred_security *subject; if (!cred) { @@ -415,10 +428,10 @@ static int hook_task_kill(struct task_struct *const p, is_scoped = domain_is_scoped(subject->domain, other, signal_scope.scope); -#ifdef CONFIG_SECURITY_LANDLOCK_LOG +#ifdef CONFIG_TRACEPOINTS if (other) target_domain_id = other->hierarchy->id; -#endif /* CONFIG_SECURITY_LANDLOCK_LOG */ +#endif /* CONFIG_TRACEPOINTS */ } if (!is_scoped) @@ -431,7 +444,12 @@ static int hook_task_kill(struct task_struct *const p, .u.tsk = p, }, .layer_plus_one = handle_layer + 1, - .other_domain_id = target_domain_id, +#ifdef CONFIG_TRACEPOINTS + .trace_signal = &(struct landlock_signal_trace) { + .target_domain_id = target_domain_id, + .signal = sig, + }, +#endif /* CONFIG_TRACEPOINTS */ }); return -EPERM; } @@ -441,7 +459,9 @@ static int hook_file_send_sigiotask(struct task_struct *tsk, { const struct landlock_cred_security *subject; bool is_scoped = false; +#ifdef CONFIG_TRACEPOINTS u64 target_domain_id = 0; +#endif /* CONFIG_TRACEPOINTS */ /* Lock already held by send_sigio() and send_sigurg(). */ lockdep_assert_held(&fown->lock); @@ -474,10 +494,10 @@ static int hook_file_send_sigiotask(struct task_struct *tsk, is_scoped = domain_is_scoped(subject->domain, other, signal_scope.scope); -#ifdef CONFIG_SECURITY_LANDLOCK_LOG +#ifdef CONFIG_TRACEPOINTS if (other) target_domain_id = other->hierarchy->id; -#endif /* CONFIG_SECURITY_LANDLOCK_LOG */ +#endif /* CONFIG_TRACEPOINTS */ } if (!is_scoped) @@ -492,7 +512,12 @@ static int hook_file_send_sigiotask(struct task_struct *tsk, #ifdef CONFIG_SECURITY_LANDLOCK_LOG .layer_plus_one = landlock_file(fown->file)->fown_layer + 1, #endif /* CONFIG_SECURITY_LANDLOCK_LOG */ - .other_domain_id = target_domain_id, +#ifdef CONFIG_TRACEPOINTS + .trace_signal = &(struct landlock_signal_trace) { + .target_domain_id = target_domain_id, + .signal = signum ? signum : SIGIO, + }, +#endif /* CONFIG_TRACEPOINTS */ }); return -EPERM; } diff --git a/security/landlock/trace.c b/security/landlock/trace.c index 43091c052f77..225dbf37bab0 100644 --- a/security/landlock/trace.c +++ b/security/landlock/trace.c @@ -201,10 +201,18 @@ void landlock_trace_denial( } break; case LANDLOCK_REQUEST_SCOPE_SIGNAL: - if (trace_landlock_deny_scope_signal_enabled()) + if (trace_landlock_deny_scope_signal_enabled()) { + const struct landlock_signal_trace *const trace_signal = + request->trace_signal; + + if (WARN_ON_ONCE(!trace_signal)) + return; + trace_landlock_deny_scope_signal( youngest_denied, same_exec, logged, - request->other_domain_id, request->audit.u.tsk); + trace_signal->target_domain_id, + request->audit.u.tsk, trace_signal->signal); + } break; case LANDLOCK_REQUEST_SCOPE_ABSTRACT_UNIX_SOCKET: if (trace_landlock_deny_scope_abstract_unix_socket_enabled()) From c6dea91d846f192eb6ddbd44f5fd873035b8b285 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Micka=C3=ABl=20Sala=C3=BCn?= Date: Fri, 18 Sep 2026 20:50:30 +0200 Subject: [PATCH 245/694] selftests/landlock: Test filesystem denial blockers MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Filesystem denial traces identify the policy change needed to allow a request, so require exact blocker values rather than merely nonempty output. Pin a READ_DIR denial to exactly one event with blockers=read_dir. Pin a REFER-only mount denial to EPERM and exactly one event with blockers=change_topology. The mount child retains CAP_SYS_ADMIN so Landlock is the only expected source of EPERM. This prevents a later capability failure from masking a Landlock regression; the trace-collecting parent remains unsandboxed. Cc: Günther Noack Cc: Steven Rostedt Link: https://patch.msgid.link/20260918185036.608651-8-mic@digikod.net Signed-off-by: Mickaël Salaün --- .../selftests/landlock/trace_fs_test.c | 76 ++++++++++++++++++- 1 file changed, 73 insertions(+), 3 deletions(-) diff --git a/tools/testing/selftests/landlock/trace_fs_test.c b/tools/testing/selftests/landlock/trace_fs_test.c index 6666d4746cb1..64014ade3a0e 100644 --- a/tools/testing/selftests/landlock/trace_fs_test.c +++ b/tools/testing/selftests/landlock/trace_fs_test.c @@ -548,7 +548,8 @@ TEST_F(trace_fs, check_rule_nested) */ TEST_F(trace_fs, deny_access_fs_denied) { - char *buf; + const char *const event_regex = REGEX_DENY_ACCESS_FS(TRACE_TASK); + char *buf, blockers[64]; int count; ASSERT_EQ(0, tracefs_clear_buf()); @@ -564,8 +565,77 @@ TEST_F(trace_fs, deny_access_fs_denied) buf = tracefs_read_buf(); ASSERT_NE(NULL, buf); - count = tracefs_count_matches(buf, REGEX_DENY_ACCESS_FS(TRACE_TASK)); - EXPECT_LE(1, count); + count = tracefs_count_matches(buf, event_regex); + EXPECT_EQ(1, count) + { + TH_LOG("Expected 1 access denial, got %d\n%s", count, buf); + } + ASSERT_EQ(0, tracefs_extract_field(buf, event_regex, "blockers", + blockers, sizeof(blockers))); + EXPECT_STREQ("read_dir", blockers); + + free(buf); +} + +/* + * Verifies that a denied mount reports the singleton topology blocker rather + * than an empty access mask. + */ +TEST_F(trace_fs, deny_change_topology) +{ + const char *const event_regex = REGEX_DENY_ACCESS_FS(TRACE_TASK); + const struct landlock_ruleset_attr ruleset_attr = { + .handled_access_fs = LANDLOCK_ACCESS_FS_REFER, + }; + char *buf, blockers[64]; + int count, ruleset_fd, status; + pid_t pid; + + ruleset_fd = + landlock_create_ruleset(&ruleset_attr, sizeof(ruleset_attr), 0); + ASSERT_LE(0, ruleset_fd); + ASSERT_EQ(0, tracefs_clear_buf()); + + /* Ensure that Landlock is the only expected mount denial. */ + set_cap(_metadata, CAP_SYS_ADMIN); + pid = fork(); + ASSERT_LE(0, pid); + if (pid == 0) { + if (prctl(PR_SET_NO_NEW_PRIVS, 1, 0, 0, 0)) { + close(ruleset_fd); + _exit(1); + } + if (landlock_restrict_self(ruleset_fd, 0)) { + close(ruleset_fd); + _exit(2); + } + close(ruleset_fd); + + if (mount(NULL, "/", NULL, MS_PRIVATE | MS_REC, NULL) != -1) + _exit(3); + + if (errno != EPERM) + _exit(4); + + _exit(0); + } + close(ruleset_fd); + clear_cap(_metadata, CAP_SYS_ADMIN); + + ASSERT_EQ(pid, waitpid(pid, &status, 0)); + ASSERT_TRUE(WIFEXITED(status)); + EXPECT_EQ(0, WEXITSTATUS(status)); + + buf = tracefs_read_buf(); + ASSERT_NE(NULL, buf); + count = tracefs_count_matches(buf, event_regex); + EXPECT_EQ(1, count) + { + TH_LOG("Expected 1 topology denial, got %d\n%s", count, buf); + } + ASSERT_EQ(0, tracefs_extract_field(buf, event_regex, "blockers", + blockers, sizeof(blockers))); + EXPECT_STREQ("change_topology", blockers); free(buf); } From c8dcb17205a689e96e2b8e46f22575a39b5b6320 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Micka=C3=ABl=20Sala=C3=BCn?= Date: Fri, 18 Sep 2026 20:50:31 +0200 Subject: [PATCH 246/694] selftests/landlock: Test network denial context MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Network denial events now report the port from their one authoritative validated address through one signed field. Verify this value directly without inferring a source or destination role. Exercise a nonzero IPv4 TCP bind, an explicit-zero IPv4 UDP bind, a synthetic-zero IPv6 UDP autobind, and a family-only AF_UNSPEC UDP send. Require exactly one event with the exact policy blocker for each shape. The value -1 distinguishes a family-only address with no validated port from the two valid port-zero cases. Test coverage for security/landlock is 91.6% of 2625 lines according to LLVM 22. Cc: Günther Noack Cc: Steven Rostedt Link: https://patch.msgid.link/20260918185036.608651-9-mic@digikod.net [mic: Add test coverage] Signed-off-by: Mickaël Salaün --- tools/testing/selftests/landlock/net_test.c | 257 +++++++++++--------- 1 file changed, 146 insertions(+), 111 deletions(-) diff --git a/tools/testing/selftests/landlock/net_test.c b/tools/testing/selftests/landlock/net_test.c index 28942438e154..16afbfdf06bb 100644 --- a/tools/testing/selftests/landlock/net_test.c +++ b/tools/testing/selftests/landlock/net_test.c @@ -3481,13 +3481,14 @@ TEST_F(trace_net, deny_access_net_bind) } /* - * Anchors the denial fields shared by every deny_access_net event so a field - * test proves more than the checked endpoint: the denying domain, the same-exec - * bit, the audit-logging verdict, and the blocked access all stay populated. + * Anchors the denial fields shared by every deny_access_net event so a port + * test also proves the denying domain, execution status, logging verdict, and + * exact blocked access. */ static void expect_net_deny_common_fields(struct __test_metadata *const _metadata, - const char *const buf) + const char *const buf, + const char *const expected_blockers) { char field[64]; @@ -3511,18 +3512,21 @@ expect_net_deny_common_fields(struct __test_metadata *const _metadata, ASSERT_EQ(0, tracefs_extract_field(buf, REGEX_DENY_ACCESS_NET(TRACE_TASK), "blockers", field, sizeof(field))); - EXPECT_STRNE("", field); + EXPECT_STREQ(expected_blockers, field); } -/* Connect and field-check tests use a separate fixture without variants. */ +enum trace_net_operation { + TRACE_NET_BIND, + TRACE_NET_SEND, +}; /* clang-format off */ -FIXTURE(trace_net_connect) { +FIXTURE(trace_net_address) { /* clang-format on */ int tracefs_ok; }; -FIXTURE_SETUP(trace_net_connect) +FIXTURE_SETUP(trace_net_address) { int ret; @@ -3547,7 +3551,7 @@ FIXTURE_SETUP(trace_net_connect) clear_cap(_metadata, CAP_SYS_ADMIN); } -FIXTURE_TEARDOWN(trace_net_connect) +FIXTURE_TEARDOWN(trace_net_address) { if (!self->tracefs_ok) return; @@ -3559,146 +3563,177 @@ FIXTURE_TEARDOWN(trace_net_connect) } /* clang-format off */ -FIXTURE_VARIANT(trace_net_connect) { +FIXTURE_VARIANT(trace_net_address) { /* clang-format on */ - /* handled_access_net, also the access allowed on the base port. */ - __u64 handled; - /* Bind the allowed base port before the denied operation. */ - bool bind_base_first; - /* Denied operation on the next port: connect (true) or bind (false). */ - bool deny_connect; + int socket_family; + int socket_type; + enum trace_net_operation operation; + int address_family; + socklen_t addrlen; + __u64 handled_access; + const char *expected_blockers; + bool address_port_zero; + bool expected_address_port; + int expected_port; }; -/* Denied connect() to the next IPv4 port. */ /* clang-format off */ -FIXTURE_VARIANT_ADD(trace_net_connect, connect_denied) { +FIXTURE_VARIANT_ADD(trace_net_address, ipv4_tcp_bind) { /* clang-format on */ - .handled = LANDLOCK_ACCESS_NET_CONNECT_TCP, - .bind_base_first = false, - .deny_connect = true, + .socket_family = AF_INET, + .socket_type = SOCK_STREAM, + .operation = TRACE_NET_BIND, + .address_family = AF_INET, + .addrlen = sizeof(struct sockaddr_in), + .handled_access = LANDLOCK_ACCESS_NET_BIND_TCP, + .expected_blockers = "bind_tcp", + .expected_address_port = true, }; -/* Denied bind() to the next IPv4 port. */ +/* Explicit bind(0) has a checked zero port. */ /* clang-format off */ -FIXTURE_VARIANT_ADD(trace_net_connect, bind_fields) { +FIXTURE_VARIANT_ADD(trace_net_address, ipv4_udp_bind_zero) { /* clang-format on */ - .handled = LANDLOCK_ACCESS_NET_BIND_TCP, - .bind_base_first = false, - .deny_connect = false, + .socket_family = AF_INET, + .socket_type = SOCK_DGRAM, + .operation = TRACE_NET_BIND, + .address_family = AF_INET, + .addrlen = sizeof(struct sockaddr_in), + .handled_access = LANDLOCK_ACCESS_NET_BIND_UDP, + .expected_blockers = "bind_udp", + .address_port_zero = true, + .expected_port = 0, }; -/* Denied connect() after an allowed bind() uses the checked destination. */ +/* A UDP send can deny its synthetic unspecified bind endpoint. */ /* clang-format off */ -FIXTURE_VARIANT_ADD(trace_net_connect, connect_after_bind) { +FIXTURE_VARIANT_ADD(trace_net_address, ipv6_udp_autobind) { /* clang-format on */ - .handled = LANDLOCK_ACCESS_NET_BIND_TCP | - LANDLOCK_ACCESS_NET_CONNECT_TCP, - .bind_base_first = true, - .deny_connect = true, + .socket_family = AF_INET6, + .socket_type = SOCK_DGRAM, + .operation = TRACE_NET_SEND, + .address_family = AF_INET6, + .addrlen = sizeof(struct sockaddr_in6), + .handled_access = LANDLOCK_ACCESS_NET_BIND_UDP, + .expected_blockers = "bind_udp", + .expected_port = 0, }; -/* - * A denied TCP bind(2) or connect(2) emits one deny_access_net event with the - * checked IPv4 port in host endianness (the UAPI landlock_net_port_attr.port - * convention). A prior allowed bind does not change a connect denial's port. - */ -TEST_F(trace_net_connect, deny_access_net) +/* A family-only address has no checked port. */ +/* clang-format off */ +FIXTURE_VARIANT_ADD(trace_net_address, ipv6_unspec_udp_send_min) { + /* clang-format on */ + .socket_family = AF_INET6, + .socket_type = SOCK_DGRAM, + .operation = TRACE_NET_SEND, + .address_family = AF_UNSPEC, + .addrlen = sizeof(sa_family_t), + .handled_access = LANDLOCK_ACCESS_NET_CONNECT_SEND_UDP, + .expected_blockers = "connect_send_udp", + .expected_port = -1, +}; + +static void set_trace_net_address(struct sockaddr_storage *const storage, + const int socket_family, + const int address_family, + const unsigned short port) { - pid_t child; - int status; - char *buf; + memset(storage, 0, sizeof(*storage)); + + if (socket_family == AF_INET) { + struct sockaddr_in *const addr4 = (struct sockaddr_in *)storage; + + addr4->sin_family = address_family; + addr4->sin_port = htons(port); + addr4->sin_addr.s_addr = htonl(INADDR_LOOPBACK); + } else { + struct sockaddr_in6 *const addr6 = + (struct sockaddr_in6 *)storage; + + addr6->sin6_family = address_family; + addr6->sin6_port = htons(port); + addr6->sin6_addr = in6addr_loopback; + } +} + +/* Verifies the actionable signed port for representative checked shapes. */ +TEST_F(trace_net_address, deny_access_net) +{ + const char *const event_regex = REGEX_DENY_ACCESS_NET(TRACE_TASK); + const unsigned short address_port = + variant->address_port_zero ? 0 : sock_port_start + 1; + const int expected_port = variant->expected_address_port ? + address_port : + variant->expected_port; + const struct landlock_ruleset_attr ruleset_attr = { + .handled_access_net = variant->handled_access, + }; + struct sockaddr_storage address; char field[64], expected[16]; + char *buf; + int count, ret, ruleset_fd, socket_fd, status; + pid_t child; if (!self->tracefs_ok) SKIP(return, "tracefs not available"); + set_trace_net_address(&address, variant->socket_family, + variant->address_family, address_port); + socket_fd = socket(variant->socket_family, + variant->socket_type | SOCK_CLOEXEC, 0); + ASSERT_LE(0, socket_fd); + ruleset_fd = + landlock_create_ruleset(&ruleset_attr, sizeof(ruleset_attr), 0); + ASSERT_LE(0, ruleset_fd); + ASSERT_EQ(0, tracefs_clear_buf()); + child = fork(); ASSERT_LE(0, child); - if (child == 0) { - struct landlock_ruleset_attr ruleset_attr = { - .handled_access_net = variant->handled, - }; - struct landlock_net_port_attr port_attr = { - .allowed_access = variant->handled, - .port = sock_port_start, - }; - struct sockaddr_in addr = { - .sin_family = AF_INET, - .sin_addr.s_addr = htonl(INADDR_LOOPBACK), - }; - int ruleset_fd, sock_fd, optval = 1, ret; - - ruleset_fd = landlock_create_ruleset(&ruleset_attr, - sizeof(ruleset_attr), 0); - if (ruleset_fd < 0) + if (prctl(PR_SET_NO_NEW_PRIVS, 1, 0, 0, 0)) _exit(1); - if (landlock_add_rule(ruleset_fd, LANDLOCK_RULE_NET_PORT, - &port_attr, 0)) { - close(ruleset_fd); - _exit(1); - } - prctl(PR_SET_NO_NEW_PRIVS, 1, 0, 0, 0); - if (landlock_restrict_self(ruleset_fd, 0)) { - close(ruleset_fd); - _exit(1); - } + if (landlock_restrict_self(ruleset_fd, 0)) + _exit(2); close(ruleset_fd); - sock_fd = socket(AF_INET, SOCK_STREAM | SOCK_CLOEXEC, 0); - if (sock_fd < 0) - _exit(1); - - /* Bind the allowed base port first (succeeds, no event). */ - if (variant->bind_base_first) { - setsockopt(sock_fd, SOL_SOCKET, SO_REUSEADDR, &optval, - sizeof(optval)); - addr.sin_port = htons(sock_port_start); - if (bind(sock_fd, (struct sockaddr *)&addr, - sizeof(addr))) { - close(sock_fd); - _exit(1); - } - } - - /* Denied operation on the next port. */ - addr.sin_port = htons(sock_port_start + 1); - if (variant->deny_connect) - ret = connect(sock_fd, (struct sockaddr *)&addr, - sizeof(addr)); - else - ret = bind(sock_fd, (struct sockaddr *)&addr, - sizeof(addr)); - if (ret == 0) { - close(sock_fd); - _exit(2); - } - if (errno != EACCES) { - close(sock_fd); + switch (variant->operation) { + case TRACE_NET_BIND: + ret = bind(socket_fd, (const struct sockaddr *)&address, + variant->addrlen); + break; + case TRACE_NET_SEND: + ret = sendto(socket_fd, "A", 1, MSG_NOSIGNAL, + (const struct sockaddr *)&address, + variant->addrlen); + break; + default: _exit(3); } - close(sock_fd); + if (ret >= 0 || errno != EACCES) + _exit(4); + close(socket_fd); + _exit(0); } + close(ruleset_fd); + close(socket_fd); ASSERT_EQ(child, waitpid(child, &status, 0)); ASSERT_TRUE(WIFEXITED(status)); - EXPECT_EQ(0, WEXITSTATUS(status)); + ASSERT_EQ(0, WEXITSTATUS(status)); buf = tracefs_read_buf(); ASSERT_NE(NULL, buf); + count = tracefs_count_matches(buf, event_regex); + if (count != 1) + TH_LOG("Expected 1 denial event, got %d\n%s", count, buf); + ASSERT_EQ(1, count); + expect_net_deny_common_fields(_metadata, buf, + variant->expected_blockers); - EXPECT_EQ(1, tracefs_count_matches(buf, - REGEX_DENY_ACCESS_NET(TRACE_TASK))); - - expect_net_deny_common_fields(_metadata, buf); - - snprintf(expected, sizeof(expected), "%llu", - (unsigned long long)(sock_port_start + 1)); - - ASSERT_EQ(0, - tracefs_extract_field(buf, REGEX_DENY_ACCESS_NET(TRACE_TASK), - "port", field, sizeof(field))); + ASSERT_EQ(0, tracefs_extract_field(buf, event_regex, "port", field, + sizeof(field))); + snprintf(expected, sizeof(expected), "%d", expected_port); EXPECT_STREQ(expected, field); free(buf); From 3fa5aa398edf94653926ad5e68b89f69490481b5 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Micka=C3=ABl=20Sala=C3=BCn?= Date: Fri, 18 Sep 2026 20:50:32 +0200 Subject: [PATCH 247/694] landlock: Fix tracepoint contract documentation MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The tracepoint documentation claims that denial and lifecycle events expose every input needed to reproduce a verdict. Instead document how denial, ruleset, and domain events identify the denying policy, checked operation and object, and reason for denial. Direct consumers to generic tracepoints for additional operational context. State the reconstruction limits: IDs are boot-local, rule checks have no request ID, and exported records may be lost or cross-CPU reordered. Also replace the incorrect BPF_RAW_TRACEPOINT guidance with libbpf SEC("tp_btf/...") attachment and refer consumers to the event prototypes for callback argument layouts. Cc: Günther Noack Cc: Steven Rostedt Link: https://patch.msgid.link/20260918185036.608651-10-mic@digikod.net Signed-off-by: Mickaël Salaün --- Documentation/trace/events-landlock.rst | 30 ++++++++++++--------- include/trace/events/landlock.h | 36 ++++++++++++------------- 2 files changed, 36 insertions(+), 30 deletions(-) diff --git a/Documentation/trace/events-landlock.rst b/Documentation/trace/events-landlock.rst index 9bb81a5c676e..304a8d06273e 100644 --- a/Documentation/trace/events-landlock.rst +++ b/Documentation/trace/events-landlock.rst @@ -141,10 +141,10 @@ in some field formats: treated as untrusted input and escaped in the trace text output so it cannot inject field separators or control characters. -- **Other party's domain**: A scope or ptrace denial compares the - subject's denying domain (``domain=``, always the enforcing domain and - never the current task) with the other party's domain, so these - tracepoints also report the other party's domain as a scalar ID: +- **Other party's domain**: A scope or ptrace denial compares the subject's + denying domain (``domain=``), which is the enforcing domain and not + necessarily the current task's domain, with the other party's domain. + These tracepoints also report the other party's domain as a scalar ID: ``tracee_domain=`` (ptrace), ``target_domain=`` (signal), and ``peer_domain=`` (abstract unix socket). It is ``0`` when the other party is unsandboxed, and otherwise a domain ID that a consumer resolves @@ -231,9 +231,10 @@ contribution, not the final decision: through another matching rule, or the right is denied and appears in the ``blockers=`` field of the corresponding ``deny_access`` event. -To reconstruct the decision for an object, aggregate the ``grants=`` -groups of all ``check_rule`` events emitted for that object during the -check. +For an access check that a consumer can delimit, aggregate the ``grants=`` +groups of all matching ``check_rule`` events. These events do not carry a +request ID; use their execution context and generic tracepoints to separate +concurrent or successive checks of the same object. .. note:: @@ -264,9 +265,11 @@ layers, so ``grants=`` has one group per layer:: eBPF access =========== -eBPF programs attached via ``BPF_RAW_TRACEPOINT`` can access the -tracepoint arguments directly through BTF. The arguments include both -standard kernel objects and Landlock-internal objects: +BTF-enabled raw tracepoint programs attached through libbpf +``SEC("tp_btf/...")`` sections receive typed callback arguments. The +event prototypes in `Event reference`_ document their argument layouts. +The arguments include both standard kernel objects and Landlock-internal +objects: - Standard kernel objects (``struct task_struct``, ``struct sock``, ``struct path``, ``struct dentry``) can be used with existing BPF @@ -277,7 +280,7 @@ standard kernel objects and Landlock-internal objects: Internal struct layouts may change between kernel versions; use CO-RE for field relocation. -A stateful eBPF program observes the full event stream and maintains +A stateful eBPF program attached before sandbox construction can maintain per-domain state in BPF maps: 1. On ``landlock_create_domain``: record the domain ID and parent (the @@ -293,7 +296,10 @@ per-domain state in BPF maps: final statistics. This approach requires no kernel modification and no Landlock-specific -BPF helpers. The Landlock IDs serve as correlation keys across events. +BPF helpers. Landlock IDs serve as correlation keys within one boot. +Records exported through tracing or BPF buffers can be lost, and records +from different CPUs are not globally ordered, so consumers must detect and +reconcile incomplete state. Audit filtering equivalence =========================== diff --git a/include/trace/events/landlock.h b/include/trace/events/landlock.h index 5da0f12ed2c3..523ba5ea9870 100644 --- a/include/trace/events/landlock.h +++ b/include/trace/events/landlock.h @@ -206,22 +206,22 @@ static inline const char *__trace_landlock_print_layers( * Decision context * ~~~~~~~~~~~~~~~~ * - * A denial event, together with the lifecycle events, exposes the full - * set of inputs the verdict consumed, so a consumer that tracked domain - * creation (landlock_create_ruleset, landlock_create_domain) can verify - * or reproduce the Landlock decision rather than merely observe it - * happened. In who/what/why terms: who is the denying domain (the domain - * field, always the subject that enforced the policy, never the current - * task), what is the operation and its object, and why is every other - * input the verdict weighed. + * A denial event identifies the domain whose policy denied the request, the + * Landlock operation and policy object that were checked, and the blocker or + * domain relationship responsible for the denial. When tracing starts with + * sandbox construction, ruleset and domain events provide the policy history + * needed to interpret these identifiers. The denying domain is the subject + * that enforced the policy, not necessarily current. Generic tracepoints can + * provide additional operational context. * * Lifecycle consistency * ~~~~~~~~~~~~~~~~~~~~~~ * - * Lifecycle events are balanced: a creation event always has a matching - * deallocation event and vice versa, so an eBPF program can model object - * lifetimes from the trace stream without reconciliation logic. A creation - * event fires while the object is still private to the calling thread + * Lifecycle emission is balanced: a creation event always has a matching + * deallocation event and vice versa. A consumer that observes an object's + * complete lifetime can model it from this pair; one that attaches late or + * loses exported records must reconcile incomplete state. A creation event + * fires while the object is still private to the calling thread * (landlock_create_ruleset fires before the ruleset's file descriptor is * installed, so it cannot race a concurrent :manpage:`close(2)`); if fd * installation later fails and the ruleset is freed, free_ruleset still @@ -295,7 +295,8 @@ static inline const char *__trace_landlock_print_layers( * the two parties without kernel-internal state. The ID is a scalar * snapshot, not a live domain pointer that could dangle: an optional * relational referent is a scalar (0 sentinel), not a nullable pointer. - * For ptrace, same_exec instead describes the tracer, even for + * Nonzero IDs are unique within one boot. For ptrace, same_exec instead + * describes the tracer, even for * PTRACE_TRACEME, and may differ from the current task. * * Blocker fields @@ -1031,11 +1032,10 @@ TRACE_EVENT(landlock_deny_scope_abstract_unix_socket, __entry->logged = logged; __entry->peer_domain_id = peer_domain_id; /* - * Best-effort (0 for a datagram peer). sk_peer_pid is - * canonically guarded by sk->sk_peer_lock, but the target - * peer's peercred is set once and not updated concurrently in - * these hooks, so this READ_ONCE() is safe; sun_path is the - * reliable identifier. + * Best-effort (0 for a datagram peer). The caller holds the + * peer's AF_UNIX state lock, serializing published peercred + * updates. The peer socket keeps a reference to sk_peer_pid + * through pid_nr(); sun_path is the reliable identifier. */ peer_pid = READ_ONCE(peer->sk_peer_pid); __entry->peer_pid = peer_pid ? pid_nr(peer_pid) : 0; From b74aad23d99b279bb34d135795f39a6d8ecdc075 Mon Sep 17 00:00:00 2001 From: Peiyang He Date: Tue, 8 Sep 2026 20:13:23 +0800 Subject: [PATCH 248/694] drm/virtio: fix memory leak of fence event on execbuffer failure virtio_gpu_execbuffer_ioctl() reserves a DRM event with drm_event_reserve_init() when VIRTGPU_EXECBUF_RING_IDX selects a ring that userspace has enabled polling for. virtio_gpu_init_submit() does this before the BO handles, the command buffer, the syncobj arrays and the in-fence are processed, so every later error path runs with the event already pending, including plain argument validation failures such as an invalid bo_handle or an in-syncobj that carries no fence. On those paths, virtio_gpu_cleanup_submit() drops the out-fence without cancelling the event. The fence is freed without ever having been emitted, taking the only driver-side pointer to the event with it. Closing the DRM file does not help. drm_events_release() unlinks pending events but deliberately leaves the freeing to the driver's later drm_send_event(), which never runs for an orphaned event, so the allocation is leaked permanently. Found when fuzzing the virtio driver with Syzkaller: BUG: memory leak unreferenced object 0xffff88802c176e80 (size 96): comm "syz.1.367", pid 10561, jiffies 4294960122 hex dump (first 32 bytes): 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................ c8 6e 17 2c 80 88 ff ff 00 00 00 00 00 00 00 00 .n.,............ backtrace (crc e1973c6b): kmemleak_alloc_recursive include/linux/kmemleak.h:44 [inline] slab_post_alloc_hook mm/slub.c:4597 [inline] slab_alloc_node mm/slub.c:4917 [inline] __kmalloc_cache_noprof+0x49d/0x6f0 mm/slub.c:5485 _kmalloc_noprof include/linux/slab.h:988 [inline] _kzalloc_noprof include/linux/slab.h:1309 [inline] virtio_gpu_fence_event_create drivers/gpu/drm/virtio/virtgpu_submit.c:282 [inline] virtio_gpu_init_submit drivers/gpu/drm/virtio/virtgpu_submit.c:398 [inline] virtio_gpu_execbuffer_ioctl+0xbbf/0x1aa0 drivers/gpu/drm/virtio/virtgpu_submit.c:505 drm_ioctl_kernel+0x1f4/0x3e0 drivers/gpu/drm/drm_ioctl.c:817 drm_ioctl+0x5f4/0xc70 drivers/gpu/drm/drm_ioctl.c:914 vfs_ioctl fs/ioctl.c:51 [inline] __do_sys_ioctl fs/ioctl.c:597 [inline] __se_sys_ioctl fs/ioctl.c:583 [inline] __x64_sys_ioctl+0x18e/0x210 fs/ioctl.c:583 do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline] do_syscall_64+0x116/0x800 arch/x86/entry/syscall_64.c:94 entry_SYSCALL_64_after_hwframe+0x77/0x7f Fix by cancelling and freeing the DRM event on the execbuffer error path before dropping the fence. Clear the fence's event pointer after cancellation so it does not retain a dangling pointer. Fixes: cd7f5ca33585 ("drm/virtio: implement context init: add virtio_gpu_fence_event") Cc: stable@vger.kernel.org Signed-off-by: Peiyang He Assisted-by: Codex:gpt-5.6-luna Signed-off-by: Dmitry Osipenko Link: https://patch.msgid.link/D320EAB5680C1411+20260908121323.2405044-1-peiyang_he@smail.nju.edu.cn --- drivers/gpu/drm/virtio/virtgpu_submit.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/drivers/gpu/drm/virtio/virtgpu_submit.c b/drivers/gpu/drm/virtio/virtgpu_submit.c index 32cb1e4aa425..734b1e976a75 100644 --- a/drivers/gpu/drm/virtio/virtgpu_submit.c +++ b/drivers/gpu/drm/virtio/virtgpu_submit.c @@ -538,6 +538,10 @@ int virtio_gpu_execbuffer_ioctl(struct drm_device *dev, void *data, virtio_gpu_process_post_deps(&submit); virtio_gpu_complete_submit(&submit); cleanup: + if (ret && submit.out_fence && submit.out_fence->e) { + drm_event_cancel_free(dev, &submit.out_fence->e->base); + submit.out_fence->e = NULL; + } virtio_gpu_cleanup_submit(&submit); return ret; From 36570ef2244cc4d7563b1f0157bc0f032498638c Mon Sep 17 00:00:00 2001 From: Junrui Luo Date: Tue, 15 Sep 2026 15:39:10 +0800 Subject: [PATCH 249/694] drm/virtio: fix object leak when drm_gem_handle_create() fails virtio_gpu_gem_create() owns the reference taken by virtio_gpu_object_create(). On the drm_gem_handle_create() error path it calls drm_gem_object_release() instead of dropping that reference. drm_gem_object_release() is the inverse of drm_gem_object_init() and does not touch the reference count or call obj->funcs->free(), so it is only correct as the last step of a destructor, as in virtio_gpu_cleanup_object(). Using it here leaves the bo at refcount 1 with no remaining reference, so virtio_gpu_free_object() never runs and the shmem pages, sg table and virtio_gpu_object are leaked. Since virtio_gpu_object_create() has already set bo->created, VIRTIO_GPU_CMD_RESOURCE_UNREF is not queued either, leaking the host-side resource and the resource id. drm_gem_handle_create_tail() drops the handle reference on all of its internal error paths, so the caller only has to drop its own. Use drm_gem_object_put(), matching the success path below. Fixes: dc5698e80cf7 ("Add virtio gpu driver.") Reported-by: Yuhao Jiang Assisted-by: Claude:claude-opus-5 Signed-off-by: Junrui Luo Signed-off-by: Dmitry Osipenko Link: https://patch.msgid.link/20260915-fixes-v2-1-a0d799e4db66@outlook.com --- drivers/gpu/drm/virtio/virtgpu_gem.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/virtio/virtgpu_gem.c b/drivers/gpu/drm/virtio/virtgpu_gem.c index 66c3f6f74e9c..d2f0b8a3f172 100644 --- a/drivers/gpu/drm/virtio/virtgpu_gem.c +++ b/drivers/gpu/drm/virtio/virtgpu_gem.c @@ -45,7 +45,7 @@ static int virtio_gpu_gem_create(struct drm_file *file, ret = drm_gem_handle_create(file, &obj->base.base, &handle); if (ret) { - drm_gem_object_release(&obj->base.base); + drm_gem_object_put(&obj->base.base); return ret; } From 477bc3068fc3777b9d8ffd79e265b0dfdf2d3a6b Mon Sep 17 00:00:00 2001 From: Junrui Luo Date: Tue, 15 Sep 2026 15:39:11 +0800 Subject: [PATCH 250/694] drm/virtio: fix object leak in virtio_gpu_resource_create_ioctl() virtio_gpu_resource_create_ioctl() calls drm_gem_object_release() on the drm_gem_handle_create() error path instead of dropping the reference it owns, so obj->funcs->free() never runs and the virtio_gpu_object, its pages and sg table, the resource id and the host-side resource are leaked. Use drm_gem_object_put() instead. Fixes: 62fb7a5e1096 ("virtio-gpu: add 3d/virgl support") Assisted-by: Claude:claude-opus-5 Signed-off-by: Junrui Luo Signed-off-by: Dmitry Osipenko Link: https://patch.msgid.link/20260915-fixes-v2-2-a0d799e4db66@outlook.com --- drivers/gpu/drm/virtio/virtgpu_ioctl.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/virtio/virtgpu_ioctl.c b/drivers/gpu/drm/virtio/virtgpu_ioctl.c index 3d8e4ccdb7c1..d16f07abb266 100644 --- a/drivers/gpu/drm/virtio/virtgpu_ioctl.c +++ b/drivers/gpu/drm/virtio/virtgpu_ioctl.c @@ -185,7 +185,7 @@ static int virtio_gpu_resource_create_ioctl(struct drm_device *dev, void *data, ret = drm_gem_handle_create(file, obj, &handle); if (ret) { - drm_gem_object_release(obj); + drm_gem_object_put(obj); return ret; } From 24b6d5c7641412c9ebef0d4c8b888d49a0e6b880 Mon Sep 17 00:00:00 2001 From: Junrui Luo Date: Tue, 15 Sep 2026 15:39:12 +0800 Subject: [PATCH 251/694] drm/virtio: fix object leaks in virtio_gpu_resource_create_blob_ioctl() virtio_gpu_resource_create_blob_ioctl() calls drm_gem_object_release() on both the virtio_gpu_resource_assign_uuid() and drm_gem_handle_create() error paths instead of dropping the reference it owns, so obj->funcs->free() never runs and the virtio_gpu_object, the resource id and the host-side resource are leaked. Use drm_gem_object_put() instead. Fixes: 897b4d1acaf5 ("drm/virtio: implement blob resources: resource create blob ioctl") Assisted-by: Claude:claude-opus-5 Signed-off-by: Junrui Luo Signed-off-by: Dmitry Osipenko Link: https://patch.msgid.link/20260915-fixes-v2-3-a0d799e4db66@outlook.com --- drivers/gpu/drm/virtio/virtgpu_ioctl.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/virtio/virtgpu_ioctl.c b/drivers/gpu/drm/virtio/virtgpu_ioctl.c index d16f07abb266..fcdb07a37972 100644 --- a/drivers/gpu/drm/virtio/virtgpu_ioctl.c +++ b/drivers/gpu/drm/virtio/virtgpu_ioctl.c @@ -557,14 +557,14 @@ static int virtio_gpu_resource_create_blob_ioctl(struct drm_device *dev, if (params.blob_flags & VIRTGPU_BLOB_FLAG_USE_CROSS_DEVICE) { ret = virtio_gpu_resource_assign_uuid(vgdev, bo); if (ret) { - drm_gem_object_release(obj); + drm_gem_object_put(obj); return ret; } } ret = drm_gem_handle_create(file, obj, &handle); if (ret) { - drm_gem_object_release(obj); + drm_gem_object_put(obj); return ret; } From 036d28db1818af2f9d80db771f5405da84d7732d Mon Sep 17 00:00:00 2001 From: Junrui Luo Date: Tue, 15 Sep 2026 15:39:13 +0800 Subject: [PATCH 252/694] drm/virtio: release the GEM object on virtio_gpu_vram_create() errors virtio_gpu_vram_create() frees the object with a bare kfree(vram) on both error paths after drm_gem_private_object_init() has run, and on the second one after drm_gem_create_mmap_offset() has linked obj->vma_node into the device's VMA offset manager. The freed object stays in that interval tree, so a later lookup or insertion walks freed memory, and the dma_resv and gpuva lock are never destroyed. Call drm_gem_object_release() before kfree() on both paths. Fixes: 16845c5d5409 ("drm/virtio: implement blob resources: implement vram object") Assisted-by: Claude:claude-opus-5 Signed-off-by: Junrui Luo Signed-off-by: Dmitry Osipenko Link: https://patch.msgid.link/20260915-fixes-v2-4-a0d799e4db66@outlook.com --- drivers/gpu/drm/virtio/virtgpu_vram.c | 17 +++++++++-------- 1 file changed, 9 insertions(+), 8 deletions(-) diff --git a/drivers/gpu/drm/virtio/virtgpu_vram.c b/drivers/gpu/drm/virtio/virtgpu_vram.c index 5b4a3ab81cd5..01241ce4d07c 100644 --- a/drivers/gpu/drm/virtio/virtgpu_vram.c +++ b/drivers/gpu/drm/virtio/virtgpu_vram.c @@ -215,16 +215,12 @@ int virtio_gpu_vram_create(struct virtio_gpu_device *vgdev, /* Create fake offset */ ret = drm_gem_create_mmap_offset(obj); - if (ret) { - kfree(vram); - return ret; - } + if (ret) + goto err_release_obj; ret = virtio_gpu_resource_id_get(vgdev, &vram->base.hw_res_handle); - if (ret) { - kfree(vram); - return ret; - } + if (ret) + goto err_release_obj; virtio_gpu_cmd_resource_create_blob(vgdev, &vram->base, params, NULL, 0); @@ -240,6 +236,11 @@ int virtio_gpu_vram_create(struct virtio_gpu_device *vgdev, *bo_ptr = &vram->base; return 0; + +err_release_obj: + drm_gem_object_release(obj); + kfree(vram); + return ret; } void virtio_gpu_vram_map_deferred(struct virtio_gpu_object_vram *vram) From 1e3b08de63274d0b009e99ef51cd6a9c0c6bf08c Mon Sep 17 00:00:00 2001 From: Dmitry Osipenko Date: Fri, 11 Sep 2026 17:42:03 +0300 Subject: [PATCH 253/694] Revert "drm/virtio: Allow importing prime buffers when 3D is enabled" Guest userspace may import udmabuf to vrend. Vrend doesn't support guest blobs, and thus, further 3d operations with the imported blob are failing. Typical scenario of the problem shown with mouse cursor RGBA image imported into virtio-gpu, which previously was rejected by virtio-gpu driver. Revert enabling guest blobs importing into vrend to fix the regression. Link: https://gitlab.freedesktop.org/virgl/virglrenderer/-/work_items/674 Fixes: df4dc947c46b ("drm/virtio: Allow importing prime buffers when 3D is enabled") Signed-off-by: Dmitry Osipenko Reviewed-by: Val Packett Link: https://patch.msgid.link/20260911144204.2089401-1-dmitry.osipenko@collabora.com --- drivers/gpu/drm/virtio/virtgpu_prime.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/virtio/virtgpu_prime.c b/drivers/gpu/drm/virtio/virtgpu_prime.c index 149e6bcb5878..ebf471044d06 100644 --- a/drivers/gpu/drm/virtio/virtgpu_prime.c +++ b/drivers/gpu/drm/virtio/virtgpu_prime.c @@ -349,7 +349,7 @@ struct drm_gem_object *virtgpu_gem_prime_import(struct drm_device *dev, } } - if (!vgdev->has_resource_blob) + if (!vgdev->has_resource_blob || vgdev->has_virgl_3d) return drm_gem_prime_import(dev, buf); bo = kzalloc_obj(*bo); From 846b3c64fe3e77d9db20a7e3e62dbbb637c773e1 Mon Sep 17 00:00:00 2001 From: Peiyang He Date: Wed, 9 Sep 2026 17:11:14 +0800 Subject: [PATCH 254/694] drm/virtio: fix NULL pointer dereference on fence allocation failure virtio_gpu_fence_alloc() can fail due to memory pressure and return NULL, but its caller like virtio_gpu_init_submit() never checks it. Later, virtio_gpu_init_submit() passes the NULL fence to virtio_gpu_fence_event_create(), which unconditionally dereferences it. Found when fuzzing the virtio driver with Syzkaller: Oops: general protection fault, probably for non-canonical address 0xdffffc0000000012: 0000 [#1] SMP KASAN NOPTI KASAN: null-ptr-deref in range [0x0000000000000090-0x0000000000000097] CPU: 1 UID: 0 PID: 9991 Comm: syz.0.121 Not tainted 7.2.0 #4 PREEMPT(full) Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS rel-1.17.0-0-gb52ca86e094d-prebuilt.qemu.org 04/01/2014 RIP: 0010:virtio_gpu_fence_event_create drivers/gpu/drm/virtio/virtgpu_submit.c:295 [inline] RIP: 0010:virtio_gpu_init_submit drivers/gpu/drm/virtio/virtgpu_submit.c:398 [inline] RIP: 0010:virtio_gpu_execbuffer_ioctl+0xc78/0x1aa0 drivers/gpu/drm/virtio/virtgpu_submit.c:505 Code: 85 ed 0f 85 21 09 00 00 e8 05 5a c9 fb 48 8b 44 24 10 48 8d b8 90 00 00 00 48 b8 00 00 00 00 00 fc ff df 48 89 fa 48 c1 ea 03 <80> 3c 02 00 0f 85 9a 0d 00 00 48 8b 44 24 10 4c 89 b0 90 00 00 00 RSP: 0018:ffffc900039dfad0 EFLAGS: 00010216 RAX: dffffc0000000000 RBX: ffffc900039dfdd8 RCX: ffffffff85f6fd3d RDX: 0000000000000012 RSI: ffffffff85f6fd4b RDI: 0000000000000090 RBP: 0000000000000000 R0virtio_gpu_virgl_process_cmd: ctrl 0x102, error 0x1203 R10: 0000000000000000 R11: 0000000000000000 R12: ffff8880132c4000 R13: 0000000000000000 R14: ffff888073b6c700 R15: 000000000000003b FS: 00007fab480b96c0(0000) GS:ffff8880eb6e9000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 00007effbf5e55a8 CR3: 0000000048d19000 CR4: 0000000000350ef0 Call Trace: drm_ioctl_kernel+0x1f4/0x3e0 drivers/gpu/drm/drm_ioctl.c:817 drm_ioctl+0x5f4/0xc70 drivers/gpu/drm/drm_ioctl.c:914 vfs_ioctl fs/ioctl.c:51 [inline] __do_sys_ioctl fs/ioctl.c:597 [inline] __se_sys_ioctl fs/ioctl.c:583 [inline] __x64_sys_ioctl+0x18e/0x210 fs/ioctl.c:583 do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline] do_syscall_64+0x116/0x800 arch/x86/entry/syscall_64.c:94 entry_SYSCALL_64_after_hwframe+0x77/0x7f RIP: 0033:0x7fab471a82bd Code: ff c3 66 2e 0f 1f 84 00 00 00 00 00 90 f3 0f 1e fa 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 c7 c1 b0 ff ff ff f7 d8 64 89 01 48 RSP: 002b:00007fab480b9018 EFLAGS: 00000246 ORIG_RAX: 0000000000000010 RAX: ffffffffffffffda RBX: 00007fab47435fa0 RCX: 00007fab471a82bd RDX: 00002000000000c0 RSI: 00000000c0406442 RDI: 0000000000000003 RBP: 00007fab480b9080 R08: 0000000000000000 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000001 R13: 00007fab47436038 R14: 00007fab47435fa0 R15: 00007ffe85ab0740 Modules linked in: ---[ end trace 0000000000000000 ]--- RIP: 0010:virtio_gpu_fence_event_create drivers/gpu/drm/virtio/virtgpu_submit.c:295 [inline] RIP: 0010:virtio_gpu_init_submit drivers/gpu/drm/virtio/virtgpu_submit.c:398 [inline] RIP: 0010:virtio_gpu_execbuffer_ioctl+0xc78/0x1aa0 drivers/gpu/drm/virtio/virtgpu_submit.c:505 Code: 85 ed 0f 85 21 09 00 00 e8 05 5a c9 fb 48 8b 44 24 10 48 8d b8 90 00 00 00 48 b8 00 00 00 00 00 fc ff df 48 89 fa 48 c1 ea 03 <80> 3c 02 00 0f 85 9a 0d 00 00 48 8b 44 24 10 4c 89 b0 90 00 00 00 RSP: 0018:ffffc900039dfad0 EFLAGS: 00010216 RAX: dffffc0000000000 RBX: ffffc900039dfdd8 RCX: ffffffff85f6fd3d RDX: 0000000000000012 RSI: ffffffff85f6fd4b RDI: 0000000000000090 RBP: 0000000000000000 R08: 0000000000000005 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000000 R12: ffff8880132c4000 R13: 0000000000000000 R14: ffff888073b6c700 R15: 000000000000003b FS: 00007fab480b96c0(0000) GS:ffff888098ae9000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 00007f24c3759000 CR3: 0000000048d19000 CR4: 0000000000350ef0 ---------------- Code disassembly (best guess): 0: 85 ed test %ebp,%ebp 2: 0f 85 21 09 00 00 jne 0x929 8: e8 05 5a c9 fb call 0xfbc95a12 d: 48 8b 44 24 10 mov 0x10(%rsp),%rax 12: 48 8d b8 90 00 00 00 lea 0x90(%rax),%rdi 19: 48 b8 00 00 00 00 00 movabs $0xdffffc0000000000,%rax 20: fc ff df 23: 48 89 fa mov %rdi,%rdx 26: 48 c1 ea 03 shr $0x3,%rdx * 2a: 80 3c 02 00 cmpb $0x0,(%rdx,%rax,1) <-- trapping instruction 2e: 0f 85 9a 0d 00 00 jne 0xdce 34: 48 8b 44 24 10 mov 0x10(%rsp),%rax 39: 4c 89 b0 90 00 00 00 mov %r14,0x90(%rax) Fix by checking virtio_gpu_fence_alloc() in virtio_gpu_init_submit() and returning -ENOMEM before any later code can dereference the NULL fence. Fixes: 70d1ace56db6 ("drm/virtio: Conditionally allocate virtio_gpu_fence") Cc: stable@vger.kernel.org Signed-off-by: Peiyang He Assisted-by: Codex:gpt-5.5 Signed-off-by: Dmitry Osipenko Link: https://patch.msgid.link/00EFE4BA92889B14+20260909091114.2622550-1-peiyang_he@smail.nju.edu.cn --- drivers/gpu/drm/virtio/virtgpu_submit.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/virtio/virtgpu_submit.c b/drivers/gpu/drm/virtio/virtgpu_submit.c index 734b1e976a75..3d35326dd904 100644 --- a/drivers/gpu/drm/virtio/virtgpu_submit.c +++ b/drivers/gpu/drm/virtio/virtgpu_submit.c @@ -389,10 +389,13 @@ static int virtio_gpu_init_submit(struct virtio_gpu_submit *submit, if ((exbuf->flags & VIRTGPU_EXECBUF_FENCE_FD_OUT) || exbuf->num_out_syncobjs || exbuf->num_bo_handles || - drm_fence_event) + drm_fence_event) { out_fence = virtio_gpu_fence_alloc(vgdev, fence_ctx, ring_idx); - else + if (!out_fence) + return -ENOMEM; + } else { out_fence = NULL; + } if (drm_fence_event) { err = virtio_gpu_fence_event_create(dev, file, out_fence, ring_idx); From 6947b78df4d25bd1e86b26870b614ea54c625b1e Mon Sep 17 00:00:00 2001 From: Shixiong Ou Date: Fri, 28 Aug 2026 17:01:23 +0800 Subject: [PATCH 255/694] drm/virtio: Add pixel blend mode property to cursor plane The cursor plane exposes a format with an alpha channel (DRM_FORMAT_ARGB8888) without a pixel blend mode property. Since commit 860e748bddcc ("drm: ensure blend mode supported if pixel format with alpha exposed") this triggers a warning during drm_mode_config_validate(): [ 0.649020] ------------[ cut here ]------------ [ 0.649040] [PLANE:36:plane-1] pixel format with alpha exposed but blend mode not setup [ 0.649081] WARNING: drivers/gpu/drm/drm_mode_config.c:872 at drm_mode_config_validate ...... [ 0.649761] Call trace: [ 0.649764] drm_mode_config_validate+0x398/0x558 [drm] (P) [ 0.649912] drm_dev_register+0x1cc/0x2a0 [drm] [ 0.650058] virtio_gpu_probe+0xd4/0x1c0 [virtio_gpu] [ 0.650088] virtio_dev_probe+0x1c8/0x310 ...... [ 0.650261] ---[ end trace 0000000000000000 ]--- Create the property with the only supported blend mode, DRM_MODE_BLEND_PREMULTI, which is also the property's default and matches what userspace had to assume before the property existed. Fixes: 860e748bddcc ("drm: ensure blend mode supported if pixel format with alpha exposed") Reported-by: Ye Liu Signed-off-by: Shixiong Ou Signed-off-by: Dmitry Osipenko Link: https://patch.msgid.link/20260828090123.784944-1-oushixiong1025@163.com --- drivers/gpu/drm/virtio/virtgpu_plane.c | 13 +++++++++++++ 1 file changed, 13 insertions(+) diff --git a/drivers/gpu/drm/virtio/virtgpu_plane.c b/drivers/gpu/drm/virtio/virtgpu_plane.c index 640815af4098..b422eba42a5f 100644 --- a/drivers/gpu/drm/virtio/virtgpu_plane.c +++ b/drivers/gpu/drm/virtio/virtgpu_plane.c @@ -589,6 +589,7 @@ struct drm_plane *virtio_gpu_plane_init(struct virtio_gpu_device *vgdev, struct drm_plane *plane; const uint32_t *formats; int nformats; + int ret; if (type == DRM_PLANE_TYPE_CURSOR) { formats = virtio_gpu_cursor_formats; @@ -614,5 +615,17 @@ struct drm_plane *virtio_gpu_plane_init(struct virtio_gpu_device *vgdev, drm_plane_create_blend_mode_property(plane, BIT(DRM_MODE_BLEND_PREMULTI)); + if (type == DRM_PLANE_TYPE_CURSOR) { + /* + * The cursor plane exposes a format with an alpha channel, + * which requires a blend mode property. The host blends + * premultiplied alpha, matching the property's default. + */ + ret = drm_plane_create_blend_mode_property(plane, + BIT(DRM_MODE_BLEND_PREMULTI)); + if (ret) + return ERR_PTR(ret); + } + return plane; } From 598c1c3e895590f845e04455d5580ea28ffde666 Mon Sep 17 00:00:00 2001 From: Benjamin Leggett Date: Fri, 14 Aug 2026 18:20:11 -0400 Subject: [PATCH 256/694] drm/virtio: sync shmem backing on guest-bound transfers virtio_gpu_cmd_transfer_to_host_{2d,3d}() sync the shmem backing for the device before the transfer, but nothing syncs for the CPU when a transfer runs the other way. That breaks two ways. Where the DMA layer bounces, the device writes into the bounce buffer while the guest keeps reading the original pages. Where DMA is not coherent, the device writes memory while the CPU keeps stale cache lines, because nothing reaches arch_sync_dma_for_cpu(). Either way DRM_IOCTL_VIRTGPU_TRANSFER_FROM_HOST hands back stale data. Sashiko originally found this in https://lore.kernel.org/dri-devel/20260806231002.27B4D1F000E9@smtp.kernel.org but the suggestion there to fix this with dma_sync_sgtable_for_cpu() isn't a sufficient fix, for two reasons. - The transfer is asynchronous. virtio_gpu_cmd_transfer_from_host_3d() only queues the command, so a sync there would run before the device had written anything. It belongs on completion, and ahead of any fence signalling. A waiter woken by the fence would otherwise race the sync and read the backing pages regardless. It needs its own pass over the reclaim list rather than a step inside the existing one, because virtio_gpu_fence_event_process() also signals every earlier fence in the same context, so any entry in that loop may signal an earlier entry's fence. - The transfer is also partial, carrying an offset, a level and a box. Where the mapping bounces, a sync for the CPU copies the whole mapping back, so unless the mapping is primed first the regions the device did not write come back holding whatever the bounce buffer contained, discarding data the guest owned. So the fix: Prime the mapping before queueing, tag the vbuffer, and sync for the CPU on completion before the fence is signalled. A second transfer must not snapshot the mapping while an earlier one is still in flight, or the snapshot would predate whatever the CPU wrote once the earlier fence signalled and the later sync would discard it. To mitigate this, wait for outstanding fences under the reservation before priming. Neither sync copies anything unless the mapping genuinely bounces: swiotlb_sync_single_for_cpu() and its Xen counterpart look the address up in the bounce pool and return early when it is absent. On a platform with non-coherent DMA they still perform the necessary cache maintenance. The range cannot be narrowed to the box, since for a non-blob resource virtio_gpu_transfer_from_host_ioctl() rejects a caller-supplied stride and layer_stride, leaving the layout to the host and the guest with no way to work out which bytes the device writes. A host3d guest blob does carry both, so its extent could be bounded, but the sync is left whole there too rather than special-cased: priming makes the untouched regions round-trip unchanged either way. Behaviour changes worth noting: - TRANSFER_FROM_HOST can now block, where before it returned as soon as the command was queued. Repeated readbacks of one resource serialise, and a readback can wait behind an earlier queued command that touched it, since virtio_gpu_array_add_fence() tags uploads, execbufs and plane flushes alike with DMA_RESV_USAGE_WRITE. -ERESTARTSYS was already possible here via dma_resv_lock_interruptible(). - A CPU write racing an in-flight transfer to the same resource is now lost, where before it survived and the transfer was lost instead. Priming captures the pages as of queueing, so a write landing before completion is overwritten by the sync. - TRANSFER_TO_HOST can also block now, but only while a guest-bound transfer on the same resource is outstanding, which happens only for callers that issue both without waiting. - Where a batch of completions contains a guest-bound transfer, the sync pass delays fence signalling for the whole batch. Only bounced pages are copied and the swiotlb pool bounds it. A batch with no such transfer is unaffected. Tested under QEMU on x86 with swiotlb=force and virtio-vga-gl iommu_platform=on, which forces both preconditions required to hit the original bug. Fixes: a3b815f09bb8 ("drm/virtio: add iommu support.") Reported-by: Sashiko AI review Closes: https://lore.kernel.org/dri-devel/20260806231002.27B4D1F000E9@smtp.kernel.org/ Signed-off-by: Benjamin Leggett Signed-off-by: Dmitry Osipenko Link: https://patch.msgid.link/20260814-virtgpu-from-host-sync-v4-1-64dd736b1779@edera.io --- drivers/gpu/drm/virtio/virtgpu_drv.h | 5 +++ drivers/gpu/drm/virtio/virtgpu_ioctl.c | 43 ++++++++++++++++++++++++ drivers/gpu/drm/virtio/virtgpu_vq.c | 46 ++++++++++++++++++++++++++ 3 files changed, 94 insertions(+) diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.h b/drivers/gpu/drm/virtio/virtgpu_drv.h index 9df4c7117341..f3bbbe4468bf 100644 --- a/drivers/gpu/drm/virtio/virtgpu_drv.h +++ b/drivers/gpu/drm/virtio/virtgpu_drv.h @@ -114,6 +114,8 @@ struct virtio_gpu_object { bool dumb; bool created; bool attached; + /* a guest-bound transfer is queued and its mapping not yet synced */ + bool from_host_pending; bool host3d_blob, guest_blob; uint32_t blob_mem, blob_flags; @@ -196,6 +198,9 @@ struct virtio_gpu_vbuffer { struct list_head list; uint32_t seqno; + + /* guest-bound transfer whose shmem backing needs a CPU sync */ + bool sync_for_cpu; }; struct virtio_gpu_output { diff --git a/drivers/gpu/drm/virtio/virtgpu_ioctl.c b/drivers/gpu/drm/virtio/virtgpu_ioctl.c index fcdb07a37972..81e70a12b356 100644 --- a/drivers/gpu/drm/virtio/virtgpu_ioctl.c +++ b/drivers/gpu/drm/virtio/virtgpu_ioctl.c @@ -261,6 +261,27 @@ static int virtio_gpu_transfer_from_host_ioctl(struct drm_device *dev, if (ret != 0) goto err_put_free; + if (virtio_gpu_is_shmem(bo) && virtio_gpu_use_dma_api(vgdev->vdev)) { + /* + * The sync on completion restores the whole mapping, so an + * earlier transfer has to be done before this one snapshots it. + * Otherwise the snapshot predates anything the CPU wrote once + * that transfer's fence signalled, and the later sync would + * discard it. Nothing can add a fence behind our back here, + * since doing so takes the reservation we already hold. + * This writes the pages, so it waits as a writer does. READ + * usage covers existing readers. + */ + long wait = dma_resv_wait_timeout(objs->objs[0]->resv, + DMA_RESV_USAGE_READ, true, + MAX_SCHEDULE_TIMEOUT); + + if (wait < 0) { + ret = wait; + goto err_unlock; + } + } + fence = virtio_gpu_fence_alloc(vgdev, vgdev->fence_drv.context, 0); if (!fence) { ret = -ENOMEM; @@ -320,6 +341,28 @@ static int virtio_gpu_transfer_to_host_ioctl(struct drm_device *dev, void *data, if (ret != 0) goto err_put_free; + /* + * A transfer the other way may have queued without yet syncing + * its mapping. Pushing the guest pages into it now would + * discard what the device wrote there, so wait for that sync: + * it runs before the fence it belongs to is signalled. The + * flag is only set under this reservation, so it cannot appear + * behind our back, and the acquire pairs with the release in + * that sync, so finding it clear means the pages it wrote are + * visible here too. + */ + if (smp_load_acquire(&bo->from_host_pending)) { + long wait = dma_resv_wait_timeout(objs->objs[0]->resv, + DMA_RESV_USAGE_WRITE, + true, + MAX_SCHEDULE_TIMEOUT); + + if (wait < 0) { + ret = wait; + goto err_unlock; + } + } + ret = -ENOMEM; fence = virtio_gpu_fence_alloc(vgdev, vgdev->fence_drv.context, 0); diff --git a/drivers/gpu/drm/virtio/virtgpu_vq.c b/drivers/gpu/drm/virtio/virtgpu_vq.c index c02c03c10d92..d99fb9e326e8 100644 --- a/drivers/gpu/drm/virtio/virtgpu_vq.c +++ b/drivers/gpu/drm/virtio/virtgpu_vq.c @@ -256,6 +256,33 @@ void virtio_gpu_dequeue_ctrl_func(struct work_struct *work) } while (!virtqueue_enable_cb(vgdev->ctrlq.vq)); spin_unlock(&vgdev->ctrlq.qlock); + /* + * Sync guest-bound transfers before signalling anything, so that a + * waiter cannot read the backing pages while what the device wrote is + * still in a bounce buffer. This cannot be folded into the loop below: + * virtio_gpu_fence_event_process() also signals every earlier fence in + * the same context, so any entry there may signal this entry's fence. + */ + list_for_each_entry(entry, &reclaim_list, list) { + if (entry->sync_for_cpu) { + struct virtio_gpu_object *bo = + gem_to_virtio_gpu_obj(entry->objs->objs[0]); + + dma_sync_sgtable_for_cpu(vgdev->vdev->dev.parent, + bo->base.sgt, DMA_FROM_DEVICE); + /* + * Release, so a transfer the other way that skips its + * wait on the strength of this cannot go on to read + * the backing pages before the sync above is visible. + * Nothing orders the two otherwise: where the mapping + * bounces on a coherent device the sync is a plain + * copy, and dma_direct_sync_sg_for_cpu() emits its + * barrier only for the non-coherent case. + */ + smp_store_release(&bo->from_host_pending, false); + } + } + list_for_each_entry(entry, &reclaim_list, list) { resp = (struct virtio_gpu_ctrl_hdr *)entry->resp_buf; @@ -1278,12 +1305,31 @@ void virtio_gpu_cmd_transfer_from_host_3d(struct virtio_gpu_device *vgdev, struct virtio_gpu_object *bo = gem_to_virtio_gpu_obj(objs->objs[0]); struct virtio_gpu_transfer_host_3d *cmd_p; struct virtio_gpu_vbuffer *vbuf; + bool use_dma_api = virtio_gpu_use_dma_api(vgdev->vdev); cmd_p = virtio_gpu_alloc_cmd(vgdev, &vbuf, sizeof(*cmd_p)); memset(cmd_p, 0, sizeof(*cmd_p)); vbuf->objs = objs; + if (virtio_gpu_is_shmem(bo) && use_dma_api) { + /* + * The device writes only the requested box, so prime the + * mapping with the current contents: otherwise the sync on + * completion would hand back whatever a bounce buffer held for + * the regions the device does not touch. + */ + dma_sync_sgtable_for_device(vgdev->vdev->dev.parent, + bo->base.sgt, DMA_TO_DEVICE); + vbuf->sync_for_cpu = true; + /* + * Set under the reservation the caller holds, so a transfer + * the other way cannot miss it and push the guest pages into + * the mapping while the device still owns it. + */ + WRITE_ONCE(bo->from_host_pending, true); + } + cmd_p->hdr.type = cpu_to_le32(VIRTIO_GPU_CMD_TRANSFER_FROM_HOST_3D); cmd_p->hdr.ctx_id = cpu_to_le32(ctx_id); cmd_p->resource_id = cpu_to_le32(bo->hw_res_handle); From ae2c5bf969573708cd5b6bb6393631255d83c3fe Mon Sep 17 00:00:00 2001 From: Prathamesh Shete Date: Wed, 16 Sep 2026 06:51:32 +0000 Subject: [PATCH 257/694] pinctrl: tegra238: Fix register bank for AON pin groups The AON pin controller has a single register region and therefore, the bank defined in the tegra238_functions[] and tegra238_aon_groups[] for the AON pin groups must be 0. However, commit 25cac7292d49 ("pinctrl: tegra: Add Tegra238 pinmux driver") incorrectly specified the bank for these pins as 1 and not 0. This means that in the tegra_pinctrl_probe() function we use an invalid index when accessing the pmx->regs[] array which causes an incorrect address to be used for accessing the pinmux registers. Fix this by correcting the bank for the AON pin groups. Fixes: 25cac7292d49 ("pinctrl: tegra: Add Tegra238 pinmux driver") Signed-off-by: Prathamesh Shete Signed-off-by: Linus Walleij --- drivers/pinctrl/tegra/pinctrl-tegra238.c | 204 +++++++++++------------ 1 file changed, 102 insertions(+), 102 deletions(-) diff --git a/drivers/pinctrl/tegra/pinctrl-tegra238.c b/drivers/pinctrl/tegra/pinctrl-tegra238.c index ec482365f14f..40aba285944e 100644 --- a/drivers/pinctrl/tegra/pinctrl-tegra238.c +++ b/drivers/pinctrl/tegra/pinctrl-tegra238.c @@ -1744,57 +1744,57 @@ static const char * const tegra238_functions[] = { #define drive_sdmmc1_dat0_pu2 DRV_PINGROUP_ENTRY_Y(0x8034, 28, 2, 30, 2, -1, -1, -1, -1, 0) #define drive_ufs0_rst_n_pv1 DRV_PINGROUP_ENTRY_Y(0x11004, 12, 5, 24, 5, -1, -1, -1, -1, 0) #define drive_ufs0_ref_clk_pv0 DRV_PINGROUP_ENTRY_Y(0x1100c, 12, 5, 24, 5, -1, -1, -1, -1, 0) -#define drive_batt_oc_paa4 DRV_PINGROUP_ENTRY_Y(0x1024, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_bootv_ctl_n_paa0 DRV_PINGROUP_ENTRY_Y(0x102c, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_vcomp_alert_paa2 DRV_PINGROUP_ENTRY_Y(0x105c, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_hdmi_cec_pbb0 DRV_PINGROUP_ENTRY_Y(0x1064, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_touch_clk_pdd3 DRV_PINGROUP_ENTRY_Y(0x106c, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_uart3_rx_pcc6 DRV_PINGROUP_ENTRY_Y(0x1074, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_uart3_tx_pcc5 DRV_PINGROUP_ENTRY_Y(0x107c, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_gen8_i2c_sda_pdd2 DRV_PINGROUP_ENTRY_Y(0x1084, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_gen8_i2c_scl_pdd1 DRV_PINGROUP_ENTRY_Y(0x108c, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_spi2_mosi_pcc2 DRV_PINGROUP_ENTRY_Y(0x1094, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_gen2_i2c_scl_pcc7 DRV_PINGROUP_ENTRY_Y(0x109c, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_spi2_cs0_pcc3 DRV_PINGROUP_ENTRY_Y(0x10a4, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_gen2_i2c_sda_pdd0 DRV_PINGROUP_ENTRY_Y(0x10ac, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_spi2_sck_pcc0 DRV_PINGROUP_ENTRY_Y(0x10b4, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_spi2_miso_pcc1 DRV_PINGROUP_ENTRY_Y(0x10bc, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio49_pee2 DRV_PINGROUP_ENTRY_Y(0x10c4, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio50_pee4 DRV_PINGROUP_ENTRY_Y(0x10cc, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio82_pee3 DRV_PINGROUP_ENTRY_Y(0x10d4, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio71_pff2 DRV_PINGROUP_ENTRY_Y(0x10dc, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio76_pff7 DRV_PINGROUP_ENTRY_Y(0x10e4, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio74_pff5 DRV_PINGROUP_ENTRY_Y(0x10ec, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio00_paa1 DRV_PINGROUP_ENTRY_Y(0x10f4, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio19_pdd6 DRV_PINGROUP_ENTRY_Y(0x10fc, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio86_phh3 DRV_PINGROUP_ENTRY_Y(0x1104, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio72_pff3 DRV_PINGROUP_ENTRY_Y(0x110c, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio77_pgg0 DRV_PINGROUP_ENTRY_Y(0x1114, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio80_pff6 DRV_PINGROUP_ENTRY_Y(0x111c, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio84_pgg1 DRV_PINGROUP_ENTRY_Y(0x1124, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio83_pee5 DRV_PINGROUP_ENTRY_Y(0x112c, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio73_pff4 DRV_PINGROUP_ENTRY_Y(0x1134, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio70_pff1 DRV_PINGROUP_ENTRY_Y(0x113c, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio04_paa5 DRV_PINGROUP_ENTRY_Y(0x1144, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio85_pgg6 DRV_PINGROUP_ENTRY_Y(0x114c, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio69_pff0 DRV_PINGROUP_ENTRY_Y(0x1154, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio25_paa6 DRV_PINGROUP_ENTRY_Y(0x115c, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_soc_gpio26_paa7 DRV_PINGROUP_ENTRY_Y(0x1164, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_uart5_tx_pgg7 DRV_PINGROUP_ENTRY_Y(0x116c, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_uart5_rx_phh0 DRV_PINGROUP_ENTRY_Y(0x1174, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_uart2_tx_pgg2 DRV_PINGROUP_ENTRY_Y(0x117c, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_uart2_rx_pgg3 DRV_PINGROUP_ENTRY_Y(0x1184, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_uart2_cts_pgg5 DRV_PINGROUP_ENTRY_Y(0x118c, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_uart2_rts_pgg4 DRV_PINGROUP_ENTRY_Y(0x1194, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_uart5_cts_phh2 DRV_PINGROUP_ENTRY_Y(0x119c, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_uart5_rts_phh1 DRV_PINGROUP_ENTRY_Y(0x11a4, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_pwm7_pee1 DRV_PINGROUP_ENTRY_Y(0x11ac, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_pwm2_pdd7 DRV_PINGROUP_ENTRY_Y(0x11b4, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_pwm3_pee0 DRV_PINGROUP_ENTRY_Y(0x11bc, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_pwm1_paa3 DRV_PINGROUP_ENTRY_Y(0x11c4, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_spi2_cs1_pcc4 DRV_PINGROUP_ENTRY_Y(0x11cc, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_dmic1_clk_pdd4 DRV_PINGROUP_ENTRY_Y(0x11d4, 12, 5, 20, 5, -1, -1, -1, -1, 1) -#define drive_dmic1_dat_pdd5 DRV_PINGROUP_ENTRY_Y(0x11dc, 12, 5, 20, 5, -1, -1, -1, -1, 1) +#define drive_batt_oc_paa4 DRV_PINGROUP_ENTRY_Y(0x1024, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_bootv_ctl_n_paa0 DRV_PINGROUP_ENTRY_Y(0x102c, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_vcomp_alert_paa2 DRV_PINGROUP_ENTRY_Y(0x105c, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_hdmi_cec_pbb0 DRV_PINGROUP_ENTRY_Y(0x1064, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_touch_clk_pdd3 DRV_PINGROUP_ENTRY_Y(0x106c, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_uart3_rx_pcc6 DRV_PINGROUP_ENTRY_Y(0x1074, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_uart3_tx_pcc5 DRV_PINGROUP_ENTRY_Y(0x107c, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_gen8_i2c_sda_pdd2 DRV_PINGROUP_ENTRY_Y(0x1084, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_gen8_i2c_scl_pdd1 DRV_PINGROUP_ENTRY_Y(0x108c, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_spi2_mosi_pcc2 DRV_PINGROUP_ENTRY_Y(0x1094, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_gen2_i2c_scl_pcc7 DRV_PINGROUP_ENTRY_Y(0x109c, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_spi2_cs0_pcc3 DRV_PINGROUP_ENTRY_Y(0x10a4, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_gen2_i2c_sda_pdd0 DRV_PINGROUP_ENTRY_Y(0x10ac, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_spi2_sck_pcc0 DRV_PINGROUP_ENTRY_Y(0x10b4, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_spi2_miso_pcc1 DRV_PINGROUP_ENTRY_Y(0x10bc, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio49_pee2 DRV_PINGROUP_ENTRY_Y(0x10c4, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio50_pee4 DRV_PINGROUP_ENTRY_Y(0x10cc, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio82_pee3 DRV_PINGROUP_ENTRY_Y(0x10d4, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio71_pff2 DRV_PINGROUP_ENTRY_Y(0x10dc, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio76_pff7 DRV_PINGROUP_ENTRY_Y(0x10e4, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio74_pff5 DRV_PINGROUP_ENTRY_Y(0x10ec, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio00_paa1 DRV_PINGROUP_ENTRY_Y(0x10f4, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio19_pdd6 DRV_PINGROUP_ENTRY_Y(0x10fc, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio86_phh3 DRV_PINGROUP_ENTRY_Y(0x1104, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio72_pff3 DRV_PINGROUP_ENTRY_Y(0x110c, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio77_pgg0 DRV_PINGROUP_ENTRY_Y(0x1114, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio80_pff6 DRV_PINGROUP_ENTRY_Y(0x111c, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio84_pgg1 DRV_PINGROUP_ENTRY_Y(0x1124, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio83_pee5 DRV_PINGROUP_ENTRY_Y(0x112c, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio73_pff4 DRV_PINGROUP_ENTRY_Y(0x1134, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio70_pff1 DRV_PINGROUP_ENTRY_Y(0x113c, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio04_paa5 DRV_PINGROUP_ENTRY_Y(0x1144, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio85_pgg6 DRV_PINGROUP_ENTRY_Y(0x114c, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio69_pff0 DRV_PINGROUP_ENTRY_Y(0x1154, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio25_paa6 DRV_PINGROUP_ENTRY_Y(0x115c, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_soc_gpio26_paa7 DRV_PINGROUP_ENTRY_Y(0x1164, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_uart5_tx_pgg7 DRV_PINGROUP_ENTRY_Y(0x116c, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_uart5_rx_phh0 DRV_PINGROUP_ENTRY_Y(0x1174, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_uart2_tx_pgg2 DRV_PINGROUP_ENTRY_Y(0x117c, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_uart2_rx_pgg3 DRV_PINGROUP_ENTRY_Y(0x1184, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_uart2_cts_pgg5 DRV_PINGROUP_ENTRY_Y(0x118c, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_uart2_rts_pgg4 DRV_PINGROUP_ENTRY_Y(0x1194, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_uart5_cts_phh2 DRV_PINGROUP_ENTRY_Y(0x119c, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_uart5_rts_phh1 DRV_PINGROUP_ENTRY_Y(0x11a4, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_pwm7_pee1 DRV_PINGROUP_ENTRY_Y(0x11ac, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_pwm2_pdd7 DRV_PINGROUP_ENTRY_Y(0x11b4, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_pwm3_pee0 DRV_PINGROUP_ENTRY_Y(0x11bc, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_pwm1_paa3 DRV_PINGROUP_ENTRY_Y(0x11c4, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_spi2_cs1_pcc4 DRV_PINGROUP_ENTRY_Y(0x11cc, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_dmic1_clk_pdd4 DRV_PINGROUP_ENTRY_Y(0x11d4, 12, 5, 20, 5, -1, -1, -1, -1, 0) +#define drive_dmic1_dat_pdd5 DRV_PINGROUP_ENTRY_Y(0x11dc, 12, 5, 20, 5, -1, -1, -1, -1, 0) #define drive_sdmmc1_comp DRV_PINGROUP_ENTRY_N @@ -1961,57 +1961,57 @@ static const struct tegra_pingroup tegra238_groups[] = { }; static const struct tegra_pingroup tegra238_aon_groups[] = { - PINGROUP(bootv_ctl_n_paa0, RSVD0, RSVD1, RSVD2, RSVD3, 0x1028, 1, Y, -1, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio00_paa1, RSVD0, RSVD1, RSVD2, RSVD3, 0x10f0, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(vcomp_alert_paa2, SOC_THERM_OC1, RSVD1, RSVD2, RSVD3, 0x1058, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(pwm1_paa3, GP_PWM1, RSVD1, RSVD2, RSVD3, 0x11c0, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(batt_oc_paa4, SOC_THERM_OC2, RSVD1, RSVD2, RSVD3, 0x1020, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio04_paa5, RSVD0, RSVD1, RSVD2, RSVD3, 0x1140, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio25_paa6, RSVD0, RSVD1, RSVD2, RSVD3, 0x1158, 1, Y, -1, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio26_paa7, RSVD0, SOC_THERM_OC3, RSVD2, RSVD3, 0x1160, 1, Y, -1, 7, 6, 8, -1, 10, 12), - PINGROUP(hdmi_cec_pbb0, HDMI_CEC, RSVD1, RSVD2, RSVD3, 0x1060, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(spi2_sck_pcc0, SPI2_SCK, RSVD1, RSVD2, RSVD3, 0x10b0, 1, Y, -1, 7, 6, 8, -1, 10, 12), - PINGROUP(spi2_miso_pcc1, SPI2_DIN, RSVD1, RSVD2, RSVD3, 0x10b8, 1, Y, -1, 7, 6, 8, -1, 10, 12), - PINGROUP(spi2_mosi_pcc2, SPI2_DOUT, RSVD1, RSVD2, RSVD3, 0x1090, 1, Y, -1, 7, 6, 8, -1, 10, 12), - PINGROUP(spi2_cs0_pcc3, SPI2_CS0, RSVD1, RSVD2, RSVD3, 0x10a0, 1, Y, -1, 7, 6, 8, -1, 10, 12), - PINGROUP(spi2_cs1_pcc4, SPI2_CS1, RSVD1, RSVD2, RSVD3, 0x11c8, 1, Y, -1, 7, 6, 8, -1, 10, 12), - PINGROUP(uart3_tx_pcc5, UARTC_TXD, RSVD1, RSVD2, RSVD3, 0x1078, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(uart3_rx_pcc6, UARTC_RXD, RSVD1, RSVD2, RSVD3, 0x1070, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(gen2_i2c_scl_pcc7, I2C2_CLK, RSVD1, RSVD2, RSVD3, 0x1098, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(gen2_i2c_sda_pdd0, I2C2_DAT, RSVD1, RSVD2, RSVD3, 0x10a8, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(gen8_i2c_scl_pdd1, I2C8_CLK, RSVD1, RSVD2, RSVD3, 0x1088, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(gen8_i2c_sda_pdd2, I2C8_DAT, RSVD1, RSVD2, RSVD3, 0x1080, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(touch_clk_pdd3, GP_PWM4, TOUCH_CLK, RSVD2, RSVD3, 0x1068, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(dmic1_clk_pdd4, DMIC1_CLK, RSVD1, DMIC5_CLK, RSVD3, 0x11d0, 1, Y, -1, 7, 6, 8, -1, 10, 12), - PINGROUP(dmic1_dat_pdd5, DMIC1_DAT, RSVD1, DMIC5_DAT, RSVD3, 0x11d8, 1, Y, -1, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio19_pdd6, RSVD0, WDT_RESET_OUTB, RSVD2, RSVD3, 0x10f8, 1, Y, -1, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio49_pee2, RSVD0, RSVD1, RSVD2, RSVD3, 0x10c0, 1, Y, -1, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio50_pee4, RSVD0, RSVD1, RSVD2, RSVD3, 0x10c8, 1, Y, -1, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio82_pee3, RSVD0, RSVD1, RSVD2, RSVD3, 0x10d0, 1, Y, -1, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio71_pff2, PPC_MODE_1, RSVD1, RSVD2, RSVD3, 0x10d8, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio76_pff7, RSVD0, RSVD1, TSC_EDGE_OUT0, TSC_EDGE_OUT0A, 0x10e0, 1, Y, -1, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio74_pff5, PPC_READY, PPC_I2C_DAT, RSVD2, RSVD3, 0x10e8, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio86_phh3, RSVD0, SPI5_CS1, TSC_EDGE_OUT3, TSC_EDGE_OUT0D, 0x1100, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio72_pff3, PPC_MODE_2, RSVD1, RSVD2, RSVD3, 0x1108, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio77_pgg0, RSVD0, RSVD1, TSC_EDGE_OUT1, TSC_EDGE_OUT0B, 0x1110, 1, Y, -1, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio80_pff6, RSVD0, PPC_RST_N, RSVD2, RSVD3, 0x1118, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio84_pgg1, RSVD0, RSVD1, TSC_EDGE_OUT2, TSC_EDGE_OUT0C, 0x1120, 1, Y, -1, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio83_pee5, RSVD0, RSVD1, RSVD2, RSVD3, 0x1128, 1, Y, -1, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio73_pff4, PPC_CC, PPC_I2C_CLK, RSVD2, RSVD3, 0x1130, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio70_pff1, PPC_MODE_0, RSVD1, RSVD2, RSVD3, 0x1138, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio85_pgg6, RSVD0, SPI4_CS1, RSVD2, RSVD3, 0x1148, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(soc_gpio69_pff0, PPC_INT_N, RSVD1, RSVD2, RSVD3, 0x1150, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(uart5_tx_pgg7, UARTE_TXD, SPI5_SCK, RSVD2, RSVD3, 0x1168, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(uart5_rx_phh0, UARTE_RXD, SPI5_MISO, RSVD2, RSVD3, 0x1170, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(uart2_tx_pgg2, UARTB_TXD, SPI4_SCK, RSVD2, RSVD3, 0x1178, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(uart2_rx_pgg3, UARTB_RXD, SPI4_MISO, RSVD2, RSVD3, 0x1180, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(uart2_cts_pgg5, UARTB_CTS, SPI4_CS0, RSVD2, RSVD3, 0x1188, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(uart2_rts_pgg4, UARTB_RTS, SPI4_MOSI, RSVD2, RSVD3, 0x1190, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(uart5_cts_phh2, UARTE_CTS, SPI5_CS0, RSVD2, RSVD3, 0x1198, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(uart5_rts_phh1, UARTE_RTS, SPI5_MOSI, RSVD2, RSVD3, 0x11a0, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(pwm2_pdd7, GP_PWM2, LED_BLINK, RSVD2, RSVD3, 0x11b0, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(pwm3_pee0, GP_PWM3, RSVD1, RSVD2, RSVD3, 0x11b8, 1, Y, 5, 7, 6, 8, -1, 10, 12), - PINGROUP(pwm7_pee1, GP_PWM7, RSVD1, RSVD2, RSVD3, 0x11a8, 1, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(bootv_ctl_n_paa0, RSVD0, RSVD1, RSVD2, RSVD3, 0x1028, 0, Y, -1, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio00_paa1, RSVD0, RSVD1, RSVD2, RSVD3, 0x10f0, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(vcomp_alert_paa2, SOC_THERM_OC1, RSVD1, RSVD2, RSVD3, 0x1058, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(pwm1_paa3, GP_PWM1, RSVD1, RSVD2, RSVD3, 0x11c0, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(batt_oc_paa4, SOC_THERM_OC2, RSVD1, RSVD2, RSVD3, 0x1020, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio04_paa5, RSVD0, RSVD1, RSVD2, RSVD3, 0x1140, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio25_paa6, RSVD0, RSVD1, RSVD2, RSVD3, 0x1158, 0, Y, -1, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio26_paa7, RSVD0, SOC_THERM_OC3, RSVD2, RSVD3, 0x1160, 0, Y, -1, 7, 6, 8, -1, 10, 12), + PINGROUP(hdmi_cec_pbb0, HDMI_CEC, RSVD1, RSVD2, RSVD3, 0x1060, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(spi2_sck_pcc0, SPI2_SCK, RSVD1, RSVD2, RSVD3, 0x10b0, 0, Y, -1, 7, 6, 8, -1, 10, 12), + PINGROUP(spi2_miso_pcc1, SPI2_DIN, RSVD1, RSVD2, RSVD3, 0x10b8, 0, Y, -1, 7, 6, 8, -1, 10, 12), + PINGROUP(spi2_mosi_pcc2, SPI2_DOUT, RSVD1, RSVD2, RSVD3, 0x1090, 0, Y, -1, 7, 6, 8, -1, 10, 12), + PINGROUP(spi2_cs0_pcc3, SPI2_CS0, RSVD1, RSVD2, RSVD3, 0x10a0, 0, Y, -1, 7, 6, 8, -1, 10, 12), + PINGROUP(spi2_cs1_pcc4, SPI2_CS1, RSVD1, RSVD2, RSVD3, 0x11c8, 0, Y, -1, 7, 6, 8, -1, 10, 12), + PINGROUP(uart3_tx_pcc5, UARTC_TXD, RSVD1, RSVD2, RSVD3, 0x1078, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(uart3_rx_pcc6, UARTC_RXD, RSVD1, RSVD2, RSVD3, 0x1070, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(gen2_i2c_scl_pcc7, I2C2_CLK, RSVD1, RSVD2, RSVD3, 0x1098, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(gen2_i2c_sda_pdd0, I2C2_DAT, RSVD1, RSVD2, RSVD3, 0x10a8, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(gen8_i2c_scl_pdd1, I2C8_CLK, RSVD1, RSVD2, RSVD3, 0x1088, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(gen8_i2c_sda_pdd2, I2C8_DAT, RSVD1, RSVD2, RSVD3, 0x1080, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(touch_clk_pdd3, GP_PWM4, TOUCH_CLK, RSVD2, RSVD3, 0x1068, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(dmic1_clk_pdd4, DMIC1_CLK, RSVD1, DMIC5_CLK, RSVD3, 0x11d0, 0, Y, -1, 7, 6, 8, -1, 10, 12), + PINGROUP(dmic1_dat_pdd5, DMIC1_DAT, RSVD1, DMIC5_DAT, RSVD3, 0x11d8, 0, Y, -1, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio19_pdd6, RSVD0, WDT_RESET_OUTB, RSVD2, RSVD3, 0x10f8, 0, Y, -1, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio49_pee2, RSVD0, RSVD1, RSVD2, RSVD3, 0x10c0, 0, Y, -1, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio50_pee4, RSVD0, RSVD1, RSVD2, RSVD3, 0x10c8, 0, Y, -1, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio82_pee3, RSVD0, RSVD1, RSVD2, RSVD3, 0x10d0, 0, Y, -1, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio71_pff2, PPC_MODE_1, RSVD1, RSVD2, RSVD3, 0x10d8, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio76_pff7, RSVD0, RSVD1, TSC_EDGE_OUT0, TSC_EDGE_OUT0A, 0x10e0, 0, Y, -1, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio74_pff5, PPC_READY, PPC_I2C_DAT, RSVD2, RSVD3, 0x10e8, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio86_phh3, RSVD0, SPI5_CS1, TSC_EDGE_OUT3, TSC_EDGE_OUT0D, 0x1100, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio72_pff3, PPC_MODE_2, RSVD1, RSVD2, RSVD3, 0x1108, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio77_pgg0, RSVD0, RSVD1, TSC_EDGE_OUT1, TSC_EDGE_OUT0B, 0x1110, 0, Y, -1, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio80_pff6, RSVD0, PPC_RST_N, RSVD2, RSVD3, 0x1118, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio84_pgg1, RSVD0, RSVD1, TSC_EDGE_OUT2, TSC_EDGE_OUT0C, 0x1120, 0, Y, -1, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio83_pee5, RSVD0, RSVD1, RSVD2, RSVD3, 0x1128, 0, Y, -1, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio73_pff4, PPC_CC, PPC_I2C_CLK, RSVD2, RSVD3, 0x1130, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio70_pff1, PPC_MODE_0, RSVD1, RSVD2, RSVD3, 0x1138, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio85_pgg6, RSVD0, SPI4_CS1, RSVD2, RSVD3, 0x1148, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(soc_gpio69_pff0, PPC_INT_N, RSVD1, RSVD2, RSVD3, 0x1150, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(uart5_tx_pgg7, UARTE_TXD, SPI5_SCK, RSVD2, RSVD3, 0x1168, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(uart5_rx_phh0, UARTE_RXD, SPI5_MISO, RSVD2, RSVD3, 0x1170, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(uart2_tx_pgg2, UARTB_TXD, SPI4_SCK, RSVD2, RSVD3, 0x1178, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(uart2_rx_pgg3, UARTB_RXD, SPI4_MISO, RSVD2, RSVD3, 0x1180, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(uart2_cts_pgg5, UARTB_CTS, SPI4_CS0, RSVD2, RSVD3, 0x1188, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(uart2_rts_pgg4, UARTB_RTS, SPI4_MOSI, RSVD2, RSVD3, 0x1190, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(uart5_cts_phh2, UARTE_CTS, SPI5_CS0, RSVD2, RSVD3, 0x1198, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(uart5_rts_phh1, UARTE_RTS, SPI5_MOSI, RSVD2, RSVD3, 0x11a0, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(pwm2_pdd7, GP_PWM2, LED_BLINK, RSVD2, RSVD3, 0x11b0, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(pwm3_pee0, GP_PWM3, RSVD1, RSVD2, RSVD3, 0x11b8, 0, Y, 5, 7, 6, 8, -1, 10, 12), + PINGROUP(pwm7_pee1, GP_PWM7, RSVD1, RSVD2, RSVD3, 0x11a8, 0, Y, 5, 7, 6, 8, -1, 10, 12), }; static const struct tegra_pinctrl_soc_data tegra238_pinctrl_aon = { From ada667890773e033d2f40dc94176e3beb930b516 Mon Sep 17 00:00:00 2001 From: Li Youhong Date: Fri, 4 Sep 2026 09:49:58 +0800 Subject: [PATCH 258/694] drm/bridge: samsung-dsim: fix TE GPIO lifetime for host attach When the Exynos DSI driver was generalized into samsung-dsim, the TE GPIO acquisition was switched from gpiod_get_optional() to devm_gpiod_get_optional() while keeping the matching gpiod_put() calls. That combination is wrong for a managed descriptor. However, dropping the puts and keeping the managed get is also wrong: samsung_dsim_register_te_irq() runs from the DSI host attach callback, and host detach/reattach can happen without destroying the device that owns the managed action. A second attach would then request the GPIO again without having released it. Switch back to a non-managed gpiod_get_optional() and keep the explicit gpiod_put() on the request_irq() error path and in samsung_dsim_unregister_te_irq(). Fixes: e7447128ca4a ("drm: bridge: Generalize Exynos-DSI driver into a Samsung DSIM bridge") Suggested-by: Luca Ceresoli Signed-off-by: Li Youhong Reviewed-by: Luca Ceresoli Tested-by: Luca Ceresoli Link: https://patch.msgid.link/20260904014958.1572918-1-dayou5941@163.com [Luca: remove unnecessary comment] Signed-off-by: Luca Ceresoli --- drivers/gpu/drm/bridge/samsung-dsim.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/bridge/samsung-dsim.c b/drivers/gpu/drm/bridge/samsung-dsim.c index e2fc69fc51b6..4694241f4d22 100644 --- a/drivers/gpu/drm/bridge/samsung-dsim.c +++ b/drivers/gpu/drm/bridge/samsung-dsim.c @@ -1862,7 +1862,7 @@ static int samsung_dsim_register_te_irq(struct samsung_dsim *dsi, struct device int te_gpio_irq; int ret; - dsi->te_gpio = devm_gpiod_get_optional(dev, "te", GPIOD_IN); + dsi->te_gpio = gpiod_get_optional(dev, "te", GPIOD_IN); if (!dsi->te_gpio) return 0; else if (IS_ERR(dsi->te_gpio)) From 7a6d08ee0f0e30023d18779bb314db8fd9a3b6d4 Mon Sep 17 00:00:00 2001 From: Ralf Lici Date: Fri, 28 Aug 2026 16:50:22 +0200 Subject: [PATCH 259/694] ovpn: preserve IPv6 scope id for netlink peer endpoints ovpn accepts OVPN_A_PEER_REMOTE_IPV6_SCOPE_ID and reports bind->remote.in6.sin6_scope_id in peer dumps, but the netlink endpoint parser never copied the attribute into the sockaddr_in6 used to create or update the peer bind. As a result, an IPv6 link-local remote endpoint configured through netlink loses its interface scope, unlike on the peer float path where ipv6_iface_scope_id populates the field. The UDPv6 output path then builds a flow with flowi6_oif set to zero and route lookup can fail or select the wrong interface. Copy the scope id when parsing non-v4-mapped IPv6 remote endpoints. The existing precheck already rejects the scope-id attribute for IPv4 and v4-mapped IPv6 remotes. Fixes: 1d36a36f6d53 ("ovpn: implement peer add/get/dump/delete via netlink") Signed-off-by: Ralf Lici Signed-off-by: Antonio Quartulli --- drivers/net/ovpn/netlink.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/drivers/net/ovpn/netlink.c b/drivers/net/ovpn/netlink.c index 4dad85294198..2ba762082acc 100644 --- a/drivers/net/ovpn/netlink.c +++ b/drivers/net/ovpn/netlink.c @@ -100,6 +100,8 @@ static bool ovpn_nl_attr_sockaddr_remote(struct nlattr **attrs, struct sockaddr_in6 *sin6; struct sockaddr_in *sin; struct in6_addr *in6; + struct nlattr *scope; + u32 scope_id = 0; __be16 port = 0; __be32 *in; @@ -114,6 +116,9 @@ static bool ovpn_nl_attr_sockaddr_remote(struct nlattr **attrs, } else if (attrs[OVPN_A_PEER_REMOTE_IPV6]) { ss->ss_family = AF_INET6; in6 = nla_data(attrs[OVPN_A_PEER_REMOTE_IPV6]); + scope = attrs[OVPN_A_PEER_REMOTE_IPV6_SCOPE_ID]; + if (scope) + scope_id = nla_get_u32(scope); } else { return false; } @@ -126,6 +131,7 @@ static bool ovpn_nl_attr_sockaddr_remote(struct nlattr **attrs, if (!ipv6_addr_v4mapped(in6)) { sin6 = (struct sockaddr_in6 *)ss; sin6->sin6_port = port; + sin6->sin6_scope_id = scope_id; memcpy(&sin6->sin6_addr, in6, sizeof(*in6)); break; } From 77393b4d72dfeb764b2af2b848acc659f6fcfd0a Mon Sep 17 00:00:00 2001 From: Ralf Lici Date: Fri, 28 Aug 2026 16:50:23 +0200 Subject: [PATCH 260/694] ovpn: skip UDP source validation for unspecified addresses ovpn validates the cached local UDP source address before reusing or refreshing a peer dst cache. This is only meaningful when a concrete source address is selected. For IPv6, calling ipv6_chk_addr with :: checks whether the unspecified address itself is configured on the host. A peer may legitimately have bind->local.ipv6 set to :: when no local endpoint was configured or after a stale learned address was cleared. In that case the source should be left unspecified and selected by ip6_dst_lookup_flow(). For IPv4, inet_confirm_addr(..., local = 0, ...) asks for local address autoselection rather than validating a chosen source. Skip the precheck there as well and let ip_route_output_flow select or reject the source. Only validate non-zero/non-any source addresses. Fixes: 08857b5ec5d9 ("ovpn: implement basic TX path (UDP)") Signed-off-by: Ralf Lici Signed-off-by: Antonio Quartulli --- drivers/net/ovpn/udp.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/drivers/net/ovpn/udp.c b/drivers/net/ovpn/udp.c index 7f69e8890b5b..df4750dabd1e 100644 --- a/drivers/net/ovpn/udp.c +++ b/drivers/net/ovpn/udp.c @@ -161,8 +161,8 @@ static int ovpn_udp4_output(struct ovpn_peer *peer, struct ovpn_bind *bind, if (rt) goto transmit; - if (unlikely(!inet_confirm_addr(sock_net(sk), NULL, 0, fl.saddr, - RT_SCOPE_HOST))) { + if (fl.saddr && unlikely(!inet_confirm_addr(sock_net(sk), NULL, 0, + fl.saddr, RT_SCOPE_HOST))) { /* we may end up here when the cached address is not usable * anymore. In this case we reset address/cache and perform a * new look up @@ -238,7 +238,8 @@ static int ovpn_udp6_output(struct ovpn_peer *peer, struct ovpn_bind *bind, if (dst) goto transmit; - if (unlikely(!ipv6_chk_addr(sock_net(sk), &fl.saddr, NULL, 0))) { + if (!ipv6_addr_any(&fl.saddr) && + unlikely(!ipv6_chk_addr(sock_net(sk), &fl.saddr, NULL, 0))) { /* we may end up here when the cached address is not usable * anymore. In this case we reset address/cache and perform a * new look up From 7c66b7a4ae80a9309e6dc1d24b7b6b897e6348eb Mon Sep 17 00:00:00 2001 From: Ralf Lici Date: Fri, 28 Aug 2026 16:50:24 +0200 Subject: [PATCH 261/694] ovpn: track UDP socket route key for peer dst cache ovpn stores the route used to transmit UDP packets in a per-peer dst cache. A cached dst is only valid for the route lookup inputs used when it was resolved. Some of those inputs are mutable while userspace still owns the UDP socket. In particular, changes to the socket mark or UDP source port do not invalidate ovpn's peer dst cache, so ovpn can keep using a route selected with an old socket route key. Replace the cached mark with a route key containing the socket-owned lookup inputs currently used by ovpn, and reset the peer dst cache when the key changes. Before storing a newly looked-up dst, recheck the route key under the peer lock so a dst resolved for stale socket state is not published. Fixes: 08857b5ec5d9 ("ovpn: implement basic TX path (UDP)") Signed-off-by: Ralf Lici Signed-off-by: Antonio Quartulli --- drivers/net/ovpn/peer.c | 1 + drivers/net/ovpn/peer.h | 19 ++++++++- drivers/net/ovpn/udp.c | 90 +++++++++++++++++++++++++++++++++++------ 3 files changed, 95 insertions(+), 15 deletions(-) diff --git a/drivers/net/ovpn/peer.c b/drivers/net/ovpn/peer.c index c95656ca7c35..b400783c2efa 100644 --- a/drivers/net/ovpn/peer.c +++ b/drivers/net/ovpn/peer.c @@ -113,6 +113,7 @@ struct ovpn_peer *ovpn_peer_new(struct ovpn_priv *ovpn, u32 id) RCU_INIT_POINTER(peer->bind, NULL); ovpn_crypto_state_init(&peer->crypto); spin_lock_init(&peer->lock); + seqcount_spinlock_init(&peer->route_key_seq, &peer->lock); kref_init(&peer->refcount); ovpn_peer_stats_init(&peer->vpn_stats); ovpn_peer_stats_init(&peer->link_stats); diff --git a/drivers/net/ovpn/peer.h b/drivers/net/ovpn/peer.h index dfa5c0037e02..063535699ecd 100644 --- a/drivers/net/ovpn/peer.h +++ b/drivers/net/ovpn/peer.h @@ -10,6 +10,7 @@ #ifndef _NET_OVPN_OVPNPEER_H_ #define _NET_OVPN_OVPNPEER_H_ +#include #include #include @@ -17,6 +18,16 @@ #include "socket.h" #include "stats.h" +/** + * struct ovpn_route_key - route key used for the peer dst cache + * @mark: fwmark used for route lookup + * @sport: UDP source port used for route lookup + */ +struct ovpn_route_key { + u32 mark; + __be16 sport; +}; + /** * struct ovpn_peer - the main remote peer object * @ovpn: main openvpn instance this peer belongs to @@ -45,6 +56,8 @@ * @tcp.sk_cb.ops: pointer to the original prot_ops object (TCP only) * @crypto: the crypto configuration (ciphers, keys, etc..) * @dst_cache: cache for dst_entry used to send to peer + * @route_key: route key matching the current dst cache contents + * @route_key_seq: seqcount protecting lockless route_key reads * @bind: remote peer binding * @keepalive_interval: seconds after which a new keepalive should be sent * @keepalive_xmit_exp: future timestamp when next keepalive should be sent @@ -55,7 +68,7 @@ * @vpn_stats: per-peer in-VPN TX/RX stats * @link_stats: per-peer link/transport TX/RX stats * @delete_reason: why peer was deleted (i.e. timeout, transport error, ..) - * @lock: protects binding to peer (bind) and keepalive* fields + * @lock: protects binding to peer (bind), route_key and keepalive* fields * @refcount: reference counter * @rcu: used to free peer in an RCU safe way * @release_entry: entry for the socket release list @@ -99,6 +112,8 @@ struct ovpn_peer { } tcp; struct ovpn_crypto_state crypto; struct dst_cache dst_cache; + struct ovpn_route_key route_key; + seqcount_spinlock_t route_key_seq; struct ovpn_bind __rcu *bind; unsigned long keepalive_interval; unsigned long keepalive_xmit_exp; @@ -109,7 +124,7 @@ struct ovpn_peer { struct ovpn_peer_stats vpn_stats; struct ovpn_peer_stats link_stats; enum ovpn_del_peer_reason delete_reason; - spinlock_t lock; /* protects bind and keepalive* */ + spinlock_t lock; /* protects bind, route_key and keepalive* */ struct kref refcount; struct rcu_head rcu; struct llist_node release_entry; diff --git a/drivers/net/ovpn/udp.c b/drivers/net/ovpn/udp.c index df4750dabd1e..c6d591cb7ff4 100644 --- a/drivers/net/ovpn/udp.c +++ b/drivers/net/ovpn/udp.c @@ -131,6 +131,48 @@ static int ovpn_udp_encap_recv(struct sock *sk, struct sk_buff *skb) return 0; } +static bool ovpn_route_key_equal(const struct ovpn_route_key *a, + const struct ovpn_route_key *b) +{ + return a->mark == b->mark && a->sport == b->sport; +} + +/** + * ovpn_dst_cache_check_key - reset peer dst cache after key changes + * @peer: the peer owning the dst cache + * @cache: the cache that might need to be reset + * @key: the route key for the packet being transmitted + * + * Reset the peer dst cache if it was populated for a different route key. + */ +static void ovpn_dst_cache_check_key(struct ovpn_peer *peer, + struct dst_cache *cache, + const struct ovpn_route_key *key) +{ + struct ovpn_route_key old_key; + unsigned int seq; + + /* snapshot the saved key before deciding whether the cache matches */ + do { + seq = read_seqcount_begin(&peer->route_key_seq); + old_key = peer->route_key; + } while (read_seqcount_retry(&peer->route_key_seq, seq)); + + /* nothing changed: the current cache can be reused */ + if (likely(ovpn_route_key_equal(&old_key, key))) + return; + + /* recheck under lock because another path may have updated the key */ + spin_lock_bh(&peer->lock); + if (!ovpn_route_key_equal(&peer->route_key, key)) { + write_seqcount_begin(&peer->route_key_seq); + peer->route_key = *key; + dst_cache_reset(cache); + write_seqcount_end(&peer->route_key_seq); + } + spin_unlock_bh(&peer->lock); +} + /** * ovpn_udp4_output - send IPv4 packet over udp socket * @peer: the destination peer @@ -138,21 +180,23 @@ static int ovpn_udp_encap_recv(struct sock *sk, struct sk_buff *skb) * @cache: dst cache * @sk: the socket to send the packet over * @skb: the packet to send + * @key: the route key snapshot used for cache validation and flow lookup * * Return: 0 on success or a negative error code otherwise */ static int ovpn_udp4_output(struct ovpn_peer *peer, struct ovpn_bind *bind, struct dst_cache *cache, struct sock *sk, - struct sk_buff *skb) + struct sk_buff *skb, + const struct ovpn_route_key *key) { struct rtable *rt; struct flowi4 fl = { .saddr = bind->local.ipv4.s_addr, .daddr = bind->remote.in4.sin_addr.s_addr, - .fl4_sport = inet_sk(sk)->inet_sport, + .fl4_sport = key->sport, .fl4_dport = bind->remote.in4.sin_port, .flowi4_proto = sk->sk_protocol, - .flowi4_mark = sk->sk_mark, + .flowi4_mark = key->mark, }; int ret; @@ -193,7 +237,12 @@ static int ovpn_udp4_output(struct ovpn_peer *peer, struct ovpn_bind *bind, ret); goto err; } - dst_cache_set_ip4(cache, &rt->dst, fl.saddr); + + /* avoid storing a stale cache */ + spin_lock_bh(&peer->lock); + if (likely(ovpn_route_key_equal(key, &peer->route_key))) + dst_cache_set_ip4(cache, &rt->dst, fl.saddr); + spin_unlock_bh(&peer->lock); transmit: udp_tunnel_xmit_skb(rt, sk, skb, fl.saddr, fl.daddr, 0, @@ -213,12 +262,14 @@ static int ovpn_udp4_output(struct ovpn_peer *peer, struct ovpn_bind *bind, * @cache: dst cache * @sk: the socket to send the packet over * @skb: the packet to send + * @key: the route key snapshot used for cache validation and flow lookup * * Return: 0 on success or a negative error code otherwise */ static int ovpn_udp6_output(struct ovpn_peer *peer, struct ovpn_bind *bind, struct dst_cache *cache, struct sock *sk, - struct sk_buff *skb) + struct sk_buff *skb, + const struct ovpn_route_key *key) { struct dst_entry *dst; int ret; @@ -226,10 +277,10 @@ static int ovpn_udp6_output(struct ovpn_peer *peer, struct ovpn_bind *bind, struct flowi6 fl = { .saddr = bind->local.ipv6, .daddr = bind->remote.in6.sin6_addr, - .fl6_sport = inet_sk(sk)->inet_sport, + .fl6_sport = key->sport, .fl6_dport = bind->remote.in6.sin6_port, .flowi6_proto = sk->sk_protocol, - .flowi6_mark = sk->sk_mark, + .flowi6_mark = key->mark, .flowi6_oif = bind->remote.in6.sin6_scope_id, }; @@ -259,7 +310,12 @@ static int ovpn_udp6_output(struct ovpn_peer *peer, struct ovpn_bind *bind, &bind->remote.in6, ret); goto err; } - dst_cache_set_ip6(cache, dst, &fl.saddr); + + /* avoid storing a stale cache */ + spin_lock_bh(&peer->lock); + if (likely(ovpn_route_key_equal(key, &peer->route_key))) + dst_cache_set_ip6(cache, dst, &fl.saddr); + spin_unlock_bh(&peer->lock); transmit: /* user IPv6 packets may be larger than the transport interface @@ -288,6 +344,7 @@ static int ovpn_udp6_output(struct ovpn_peer *peer, struct ovpn_bind *bind, * @cache: dst cache * @sk: the socket to send the packet over * @skb: the packet to send + * @key: route key snapshot used for cache validation and flow lookup * * rcu_read_lock should be held on entry. * On return, the skb is consumed. @@ -295,7 +352,8 @@ static int ovpn_udp6_output(struct ovpn_peer *peer, struct ovpn_bind *bind, * Return: 0 on success or a negative error code otherwise */ static int ovpn_udp_output(struct ovpn_peer *peer, struct dst_cache *cache, - struct sock *sk, struct sk_buff *skb) + struct sock *sk, struct sk_buff *skb, + struct ovpn_route_key *key) { struct ovpn_bind *bind; int ret; @@ -315,11 +373,11 @@ static int ovpn_udp_output(struct ovpn_peer *peer, struct dst_cache *cache, switch (bind->remote.in4.sin_family) { case AF_INET: - ret = ovpn_udp4_output(peer, bind, cache, sk, skb); + ret = ovpn_udp4_output(peer, bind, cache, sk, skb, key); break; #if IS_ENABLED(CONFIG_IPV6) case AF_INET6: - ret = ovpn_udp6_output(peer, bind, cache, sk, skb); + ret = ovpn_udp6_output(peer, bind, cache, sk, skb, key); break; #endif default: @@ -341,15 +399,21 @@ static int ovpn_udp_output(struct ovpn_peer *peer, struct dst_cache *cache, void ovpn_udp_send_skb(struct ovpn_peer *peer, struct sock *sk, struct sk_buff *skb) { + struct ovpn_route_key key = { + .mark = READ_ONCE(sk->sk_mark), + .sport = READ_ONCE(inet_sk(sk)->inet_sport), + }; int ret; skb->dev = peer->ovpn->dev; - skb->mark = READ_ONCE(sk->sk_mark); + skb->mark = key.mark; /* no checksum performed at this layer */ skb->ip_summed = CHECKSUM_NONE; + ovpn_dst_cache_check_key(peer, &peer->dst_cache, &key); + /* crypto layer -> transport (UDP) */ - ret = ovpn_udp_output(peer, &peer->dst_cache, sk, skb); + ret = ovpn_udp_output(peer, &peer->dst_cache, sk, skb, &key); if (unlikely(ret < 0)) kfree_skb(skb); } From fa603710bdb9aea33c0d9cc2c05ed24d84f58753 Mon Sep 17 00:00:00 2001 From: Ralf Lici Date: Fri, 28 Aug 2026 16:50:25 +0200 Subject: [PATCH 262/694] ovpn: validate peer state before caching UDP dst UDP route lookup runs without peer->lock while the bind is protected by RCU. The route key is snapshotted separately. Either can change while the lookup is in progress. The TX path currently checks only the route key before publishing the looked-up dst. If the bind changes but the route key does not, a dst resolved from the old endpoint can be installed in the cache after the bind replacement. Compare both the bind pointer and the route key under peer->lock before updating the cache. The RCU read-side critical section keeps the old bind alive throughout the lookup, so pointer identity is sufficient to detect a replacement. Fixes: f0281c1d3732 ("ovpn: add support for updating local or remote UDP endpoint") Signed-off-by: Ralf Lici Signed-off-by: Antonio Quartulli --- drivers/net/ovpn/udp.c | 33 +++++++++++++++++++++++++++++++-- 1 file changed, 31 insertions(+), 2 deletions(-) diff --git a/drivers/net/ovpn/udp.c b/drivers/net/ovpn/udp.c index c6d591cb7ff4..eeef4a7229f5 100644 --- a/drivers/net/ovpn/udp.c +++ b/drivers/net/ovpn/udp.c @@ -173,6 +173,35 @@ static void ovpn_dst_cache_check_key(struct ovpn_peer *peer, spin_unlock_bh(&peer->lock); } +/** + * ovpn_dst_cache_current - check whether a route lookup matches peer state + * @peer: the peer owning the bind and dst cache + * @bind: the RCU bind used for the route lookup + * @key: the route key used for the route lookup + * + * Check that @bind is still the current peer bind and that @key still matches + * the peer route key. The caller must hold @peer->lock. The TX path keeps + * @bind inside an RCU read-side critical section, so pointer identity is enough + * to detect whether the bind was replaced while the route lookup was running. + * + * Return: true if the lookup result still matches the current peer state and + * may update the dst cache. + */ +static bool ovpn_dst_cache_current(const struct ovpn_peer *peer, + const struct ovpn_bind *bind, + const struct ovpn_route_key *key) +{ + const struct ovpn_bind *curr_bind; + + lockdep_assert_held(&peer->lock); + + curr_bind = rcu_dereference_protected(peer->bind, + lockdep_is_held(&peer->lock)); + + return curr_bind == bind && + ovpn_route_key_equal(key, &peer->route_key); +} + /** * ovpn_udp4_output - send IPv4 packet over udp socket * @peer: the destination peer @@ -240,7 +269,7 @@ static int ovpn_udp4_output(struct ovpn_peer *peer, struct ovpn_bind *bind, /* avoid storing a stale cache */ spin_lock_bh(&peer->lock); - if (likely(ovpn_route_key_equal(key, &peer->route_key))) + if (likely(ovpn_dst_cache_current(peer, bind, key))) dst_cache_set_ip4(cache, &rt->dst, fl.saddr); spin_unlock_bh(&peer->lock); @@ -313,7 +342,7 @@ static int ovpn_udp6_output(struct ovpn_peer *peer, struct ovpn_bind *bind, /* avoid storing a stale cache */ spin_lock_bh(&peer->lock); - if (likely(ovpn_route_key_equal(key, &peer->route_key))) + if (likely(ovpn_dst_cache_current(peer, bind, key))) dst_cache_set_ip6(cache, dst, &fl.saddr); spin_unlock_bh(&peer->lock); From aea934a221ec6a867221e5b765f65f1857befd53 Mon Sep 17 00:00:00 2001 From: Ralf Lici Date: Fri, 28 Aug 2026 16:50:26 +0200 Subject: [PATCH 263/694] ovpn: replace bind when learning local endpoint struct ovpn_bind is published through peer->bind with RCU, but local endpoint learning updates bind->local in place under peer->lock. UDP TX reads the field without that lock. In particular, a concurrent IPv6 update can therefore result in a torn address read. Use ovpn_peer_reset_sockaddr to publish a replacement bind when learning a new local endpoint, just as a remote endpoint change does. Preserve the current remote address and reset the dst cache only after the new bind has been published successfully. Track remote endpoint changes separately so that float notification and transport-address rehashing remain limited to actual peer floats. Fixes: f0281c1d3732 ("ovpn: add support for updating local or remote UDP endpoint") Signed-off-by: Ralf Lici Signed-off-by: Antonio Quartulli --- drivers/net/ovpn/peer.c | 43 ++++++++++++++++++++++------------------- 1 file changed, 23 insertions(+), 20 deletions(-) diff --git a/drivers/net/ovpn/peer.c b/drivers/net/ovpn/peer.c index b400783c2efa..430c6cd48db8 100644 --- a/drivers/net/ovpn/peer.c +++ b/drivers/net/ovpn/peer.c @@ -200,13 +200,12 @@ static void __ovpn_peer_hash_transp_addr(struct ovpn_peer *peer, */ void ovpn_peer_endpoints_update(struct ovpn_peer *peer, struct sk_buff *skb) { + const void *local_ip = NULL; struct sockaddr_storage ss; struct sockaddr_in6 *sa6; - bool reset_cache = false; struct sockaddr_in *sa; struct ovpn_bind *bind; - const void *local_ip; - size_t salen = 0; + bool floated = false; spin_lock_bh(&peer->lock); bind = rcu_dereference_protected(peer->bind, @@ -233,8 +232,7 @@ void ovpn_peer_endpoints_update(struct ovpn_peer *peer, struct sk_buff *skb) .sin_addr.s_addr = ip_hdr(skb)->saddr, .sin_port = udp_hdr(skb)->source, }; - salen = sizeof(*sa); - reset_cache = true; + floated = true; break; } @@ -246,10 +244,12 @@ void ovpn_peer_endpoints_update(struct ovpn_peer *peer, struct sk_buff *skb) netdev_name(peer->ovpn->dev), peer->id, &bind->local.ipv4.s_addr, &ip_hdr(skb)->daddr); - bind->local.ipv4.s_addr = ip_hdr(skb)->daddr; - reset_cache = true; + local_ip = &ip_hdr(skb)->daddr; + memcpy(&ss, &bind->remote, sizeof(struct sockaddr_in)); + break; } - break; + /* nothing changed */ + goto unlock; case htons(ETH_P_IPV6): /* float check */ if (unlikely(!ovpn_bind_skb_src_match(bind, skb))) { @@ -271,8 +271,7 @@ void ovpn_peer_endpoints_update(struct ovpn_peer *peer, struct sk_buff *skb) ipv6_iface_scope_id(&ipv6_hdr(skb)->saddr, skb->skb_iif), }; - salen = sizeof(*sa6); - reset_cache = true; + floated = true; break; } @@ -285,26 +284,30 @@ void ovpn_peer_endpoints_update(struct ovpn_peer *peer, struct sk_buff *skb) netdev_name(peer->ovpn->dev), peer->id, &bind->local.ipv6, &ipv6_hdr(skb)->daddr); - bind->local.ipv6 = ipv6_hdr(skb)->daddr; - reset_cache = true; + local_ip = &ipv6_hdr(skb)->daddr; + memcpy(&ss, &bind->remote, sizeof(struct sockaddr_in6)); + break; } - break; + /* nothing changed */ + goto unlock; default: goto unlock; } - if (unlikely(reset_cache)) - dst_cache_reset(&peer->dst_cache); - - /* if the peer did not float, we can bail out now */ - if (likely(!salen)) - goto unlock; - if (unlikely(ovpn_peer_reset_sockaddr(peer, (struct sockaddr_storage *)&ss, local_ip) < 0)) goto unlock; + /* reset the cache only after a successful bind update to avoid useless + * cache misses on concurrent TX + */ + dst_cache_reset(&peer->dst_cache); + + /* if only the local address changed, bail out now */ + if (!floated) + goto unlock; + net_dbg_ratelimited("%s: peer %d floated to %pIScp", netdev_name(peer->ovpn->dev), peer->id, &ss); From 7d8104988f423572df1f3347ce578037b1043f34 Mon Sep 17 00:00:00 2001 From: Ralf Lici Date: Fri, 28 Aug 2026 16:50:27 +0200 Subject: [PATCH 264/694] ovpn: replace bind when clearing stale local source The UDP output fallback clears bind->local in place when the remembered source address is no longer usable. The bind is RCU-published and read locklessly by concurrent TX, so an IPv6 reader can observe a torn address. Retry the route lookup with source address autoselection without modifying the bind. After a successful lookup, revalidate the bind and route key under peer->lock, reset the dst cache, and best-effort publish a replacement bind with a wildcard local address. Do not cache the resolved dst when clearing the local source. Replacing the source invalidates all per-CPU cache entries, while dst_cache_set_ip4 and dst_cache_set_ip6 update only the current CPU slot. The current packet can still use the resolved route; if bind allocation fails, a later cache miss retries the repair. Fixes: 08857b5ec5d9 ("ovpn: implement basic TX path (UDP)") Signed-off-by: Ralf Lici Signed-off-by: Antonio Quartulli --- drivers/net/ovpn/udp.c | 81 +++++++++++++++++++++++++++++------------- 1 file changed, 56 insertions(+), 25 deletions(-) diff --git a/drivers/net/ovpn/udp.c b/drivers/net/ovpn/udp.c index eeef4a7229f5..055cdb1bee13 100644 --- a/drivers/net/ovpn/udp.c +++ b/drivers/net/ovpn/udp.c @@ -185,7 +185,7 @@ static void ovpn_dst_cache_check_key(struct ovpn_peer *peer, * to detect whether the bind was replaced while the route lookup was running. * * Return: true if the lookup result still matches the current peer state and - * may update the dst cache. + * may update the dst cache or replace the bind. */ static bool ovpn_dst_cache_current(const struct ovpn_peer *peer, const struct ovpn_bind *bind, @@ -218,6 +218,9 @@ static int ovpn_udp4_output(struct ovpn_peer *peer, struct ovpn_bind *bind, struct sk_buff *skb, const struct ovpn_route_key *key) { + struct sockaddr_storage remote; + struct in_addr local = {}; + bool reset_local = false; struct rtable *rt; struct flowi4 fl = { .saddr = bind->local.ipv4.s_addr, @@ -236,24 +239,17 @@ static int ovpn_udp4_output(struct ovpn_peer *peer, struct ovpn_bind *bind, if (fl.saddr && unlikely(!inet_confirm_addr(sock_net(sk), NULL, 0, fl.saddr, RT_SCOPE_HOST))) { - /* we may end up here when the cached address is not usable - * anymore. In this case we reset address/cache and perform a - * new look up + /* The learned local address is not usable anymore. + * Retry with source address autoselection. */ fl.saddr = 0; - spin_lock_bh(&peer->lock); - bind->local.ipv4.s_addr = 0; - spin_unlock_bh(&peer->lock); - dst_cache_reset(cache); + reset_local = true; } rt = ip_route_output_flow(sock_net(sk), &fl, sk); if (IS_ERR(rt) && PTR_ERR(rt) == -EINVAL) { fl.saddr = 0; - spin_lock_bh(&peer->lock); - bind->local.ipv4.s_addr = 0; - spin_unlock_bh(&peer->lock); - dst_cache_reset(cache); + reset_local = true; rt = ip_route_output_flow(sock_net(sk), &fl, sk); } @@ -267,10 +263,28 @@ static int ovpn_udp4_output(struct ovpn_peer *peer, struct ovpn_bind *bind, goto err; } - /* avoid storing a stale cache */ + /* avoid storing a stale cache or local address */ spin_lock_bh(&peer->lock); - if (likely(ovpn_dst_cache_current(peer, bind, key))) - dst_cache_set_ip4(cache, &rt->dst, fl.saddr); + if (likely(ovpn_dst_cache_current(peer, bind, key))) { + if (!reset_local) { + dst_cache_set_ip4(cache, &rt->dst, fl.saddr); + spin_unlock_bh(&peer->lock); + goto transmit; + } + + /* invalidate per-CPU dst entries that may still carry + * the stale source + */ + dst_cache_reset(cache); + + /* preserve the current remote */ + memcpy(&remote, &bind->remote, sizeof(struct sockaddr_in)); + /* The current packet already has a valid wildcard-source route. + * If replacing the bind fails, leave the stale local in place; + * a later cache miss will retry the repair. + */ + ovpn_peer_reset_sockaddr(peer, &remote, &local); + } spin_unlock_bh(&peer->lock); transmit: @@ -300,6 +314,9 @@ static int ovpn_udp6_output(struct ovpn_peer *peer, struct ovpn_bind *bind, struct sk_buff *skb, const struct ovpn_route_key *key) { + struct in6_addr local = in6addr_any; + struct sockaddr_storage remote; + bool reset_local = false; struct dst_entry *dst; int ret; @@ -320,15 +337,11 @@ static int ovpn_udp6_output(struct ovpn_peer *peer, struct ovpn_bind *bind, if (!ipv6_addr_any(&fl.saddr) && unlikely(!ipv6_chk_addr(sock_net(sk), &fl.saddr, NULL, 0))) { - /* we may end up here when the cached address is not usable - * anymore. In this case we reset address/cache and perform a - * new look up + /* The learned local address is not usable anymore. + * Retry with source address autoselection. */ fl.saddr = in6addr_any; - spin_lock_bh(&peer->lock); - bind->local.ipv6 = in6addr_any; - spin_unlock_bh(&peer->lock); - dst_cache_reset(cache); + reset_local = true; } dst = ip6_dst_lookup_flow(sock_net(sk), sk, &fl, NULL); @@ -340,10 +353,28 @@ static int ovpn_udp6_output(struct ovpn_peer *peer, struct ovpn_bind *bind, goto err; } - /* avoid storing a stale cache */ + /* avoid storing a stale cache or local address */ spin_lock_bh(&peer->lock); - if (likely(ovpn_dst_cache_current(peer, bind, key))) - dst_cache_set_ip6(cache, dst, &fl.saddr); + if (likely(ovpn_dst_cache_current(peer, bind, key))) { + if (!reset_local) { + dst_cache_set_ip6(cache, dst, &fl.saddr); + spin_unlock_bh(&peer->lock); + goto transmit; + } + + /* invalidate per-CPU dst entries that may still carry + * the stale source + */ + dst_cache_reset(cache); + + /* preserve the current remote */ + memcpy(&remote, &bind->remote, sizeof(struct sockaddr_in6)); + /* The current packet already has a valid wildcard-source route. + * If replacing the bind fails, leave the stale local in place; + * a later cache miss will retry the repair. + */ + ovpn_peer_reset_sockaddr(peer, &remote, &local); + } spin_unlock_bh(&peer->lock); transmit: From b43beccb3713fafada57814b0a652f4a876eb75f Mon Sep 17 00:00:00 2001 From: Ralf Lici Date: Fri, 28 Aug 2026 15:00:06 +0200 Subject: [PATCH 265/694] ovpn: always unhash old VPN addresses before rehashing ovpn_peer_hash_vpn_ip updates the per-peer VPN address hash entries after userspace changes a peer VPN address. The current code removes an old hash entry only when the new address for that family is not the unspecified address. When an address is cleared to 0.0.0.0 or ::, its hash node therefore remains linked in the bucket selected by the old address. The address comparison performed during lookup prevents the old address from matching, but the table retains a stale entry until the peer is removed or another address is configured for that family. Always remove both old VPN address hash entries before conditionally adding the currently configured addresses back. This ensures that a cleared address leaves its hash node unhashed. Fixes: 1d36a36f6d53 ("ovpn: implement peer add/get/dump/delete via netlink") Signed-off-by: Ralf Lici Signed-off-by: Antonio Quartulli --- drivers/net/ovpn/peer.c | 10 ++++------ 1 file changed, 4 insertions(+), 6 deletions(-) diff --git a/drivers/net/ovpn/peer.c b/drivers/net/ovpn/peer.c index 430c6cd48db8..bbd9e17fb0bf 100644 --- a/drivers/net/ovpn/peer.c +++ b/drivers/net/ovpn/peer.c @@ -994,10 +994,11 @@ void ovpn_peer_hash_vpn_ip(struct ovpn_peer *peer) if (hlist_unhashed(&peer->hash_entry_id)) return; - if (peer->vpn_addrs.ipv4.s_addr != htonl(INADDR_ANY)) { - /* remove potential old hashing */ - hlist_nulls_del_init_rcu(&peer->hash_entry_addr4); + /* remove potential old hashing */ + hlist_nulls_del_init_rcu(&peer->hash_entry_addr4); + hlist_nulls_del_init_rcu(&peer->hash_entry_addr6); + if (peer->vpn_addrs.ipv4.s_addr != htonl(INADDR_ANY)) { nhead = ovpn_get_hash_head(peer->ovpn->peers->by_vpn_addr4, &peer->vpn_addrs.ipv4, sizeof(peer->vpn_addrs.ipv4)); @@ -1005,9 +1006,6 @@ void ovpn_peer_hash_vpn_ip(struct ovpn_peer *peer) } if (!ipv6_addr_any(&peer->vpn_addrs.ipv6)) { - /* remove potential old hashing */ - hlist_nulls_del_init_rcu(&peer->hash_entry_addr6); - nhead = ovpn_get_hash_head(peer->ovpn->peers->by_vpn_addr6, &peer->vpn_addrs.ipv6, sizeof(peer->vpn_addrs.ipv6)); From d25e885b31a0f2808d936f95c9a558a8a792669b Mon Sep 17 00:00:00 2001 From: Ralf Lici Date: Fri, 28 Aug 2026 15:00:07 +0200 Subject: [PATCH 266/694] ovpn: reject duplicate peer VPN addresses In MP mode, ovpn uses the peer VPN addresses as lookup keys for selecting the peer that should receive an outgoing tunnel packet. However, the netlink peer configuration path does not currently reject duplicate VPN addresses. If two peers are configured with the same VPN address, both can be inserted in the VPN address hash table and lookups return whichever peer is found first. This makes peer selection ambiguous and dependent on hash insertion order. Reject peer creation or update when the resulting VPN address is already assigned to another peer. Ignore unspecified addresses because those are not inserted in the VPN address hash tables. This changes such configurations from being accepted to being rejected, but they have never worked reliably because peer selection is ambiguous. Fixes: 1d36a36f6d53 ("ovpn: implement peer add/get/dump/delete via netlink") Signed-off-by: Ralf Lici Signed-off-by: Antonio Quartulli --- drivers/net/ovpn/netlink.c | 37 ++++++++++++++++----- drivers/net/ovpn/peer.c | 67 +++++++++++++++++++++++++++++++++++++- drivers/net/ovpn/peer.h | 6 ++++ 3 files changed, 101 insertions(+), 9 deletions(-) diff --git a/drivers/net/ovpn/netlink.c b/drivers/net/ovpn/netlink.c index 2ba762082acc..e23f7d1f49e0 100644 --- a/drivers/net/ovpn/netlink.c +++ b/drivers/net/ovpn/netlink.c @@ -480,8 +480,10 @@ int ovpn_nl_peer_new_doit(struct sk_buff *skb, struct genl_info *info) int ovpn_nl_peer_set_doit(struct sk_buff *skb, struct genl_info *info) { - struct nlattr *attrs[OVPN_A_PEER_MAX + 1]; struct ovpn_priv *ovpn = info->user_ptr[0]; + struct nlattr *attrs[OVPN_A_PEER_MAX + 1]; + struct in6_addr vpn_addr6; + struct in_addr vpn_addr4; struct ovpn_socket *sock; struct ovpn_peer *peer; u32 peer_id; @@ -528,28 +530,47 @@ int ovpn_nl_peer_set_doit(struct sk_buff *skb, struct genl_info *info) rcu_read_unlock(); spin_lock_bh(&ovpn->lock); - ret = ovpn_nl_peer_modify(peer, info, attrs); - if (ret < 0) { - spin_unlock_bh(&ovpn->lock); - ovpn_peer_put(peer); - return ret; + + /* reject peer with conflicting VPN address */ + if (attrs[OVPN_A_PEER_VPN_IPV4]) { + vpn_addr4.s_addr = nla_get_in_addr(attrs[OVPN_A_PEER_VPN_IPV4]); + if (ovpn_peer_vpn_addr_conflict4(ovpn, peer, &vpn_addr4)) + goto addr_conflict; } + if (attrs[OVPN_A_PEER_VPN_IPV6]) { + vpn_addr6 = nla_get_in6_addr(attrs[OVPN_A_PEER_VPN_IPV6]); + if (ovpn_peer_vpn_addr_conflict6(ovpn, peer, &vpn_addr6)) + goto addr_conflict; + } + + ret = ovpn_nl_peer_modify(peer, info, attrs); + if (ret < 0) + goto unlock; /* ret == 1 means that VPN IPv4/6 has been modified and rehashing * is required */ - if (ret > 0) + if (ret > 0) { ovpn_peer_hash_vpn_ip(peer); + ret = 0; + } /* if the remote endpoint was updated, the by_transp_addr hash bucket * also needs to be refreshed, otherwise incoming packets from the new * remote address would fail the lockless lookup */ if (attrs[OVPN_A_PEER_REMOTE_IPV4] || attrs[OVPN_A_PEER_REMOTE_IPV6]) ovpn_peer_hash_transp_addr(peer); + +unlock: spin_unlock_bh(&ovpn->lock); ovpn_peer_put(peer); - return 0; + return ret; +addr_conflict: + NL_SET_ERR_MSG_FMT_MOD(info->extack, + "VPN IP is already assigned to another peer"); + ret = -EADDRINUSE; + goto unlock; } static int ovpn_nl_send_peer(struct sk_buff *skb, const struct genl_info *info, diff --git a/drivers/net/ovpn/peer.c b/drivers/net/ovpn/peer.c index bbd9e17fb0bf..2067825bb5b6 100644 --- a/drivers/net/ovpn/peer.c +++ b/drivers/net/ovpn/peer.c @@ -488,7 +488,7 @@ static struct ovpn_peer *ovpn_peer_get_by_vpn_addr4(struct ovpn_priv *ovpn, * Return: the peer if found or NULL otherwise */ static struct ovpn_peer *ovpn_peer_get_by_vpn_addr6(struct ovpn_priv *ovpn, - struct in6_addr *addr) + const struct in6_addr *addr) { struct hlist_nulls_head *nhead; struct hlist_nulls_node *ntmp; @@ -513,6 +513,64 @@ static struct ovpn_peer *ovpn_peer_get_by_vpn_addr6(struct ovpn_priv *ovpn, return NULL; } +/** + * ovpn_peer_vpn_addr_conflict4 - check if the VPN v4 address is already in use + * @ovpn: the openvpn instance to search + * @peer: peer being added or updated, or NULL + * @addr: VPN IPv4 address to check + * + * Check whether @addr is already assigned to another peer. @peer is ignored + * when found, allowing peer updates that keep an existing address. + * Unspecified addresses are ignored. + * + * Note: the caller must hold @ovpn->lock. + * + * Return: true on conflict, false otherwise. + */ +bool ovpn_peer_vpn_addr_conflict4(struct ovpn_priv *ovpn, + const struct ovpn_peer *peer, + const struct in_addr *addr) +{ + struct ovpn_peer *tmp = NULL; + + lockdep_assert_held(&ovpn->lock); + + /* we don't hash INADDR_ANY, no conflict in that case */ + if (addr->s_addr != htonl(INADDR_ANY)) + tmp = ovpn_peer_get_by_vpn_addr4(ovpn, addr->s_addr); + + return tmp && tmp != peer; +} + +/** + * ovpn_peer_vpn_addr_conflict6 - check if the VPN v6 address is already in use + * @ovpn: the openvpn instance to search + * @peer: peer being added or updated, or NULL + * @addr: VPN IPv6 address to check + * + * Check whether @addr is already assigned to another peer. @peer is ignored + * when found, allowing peer updates that keep an existing address. + * Unspecified addresses are ignored. + * + * Note: the caller must hold @ovpn->lock. + * + * Return: true on conflict, false otherwise. + */ +bool ovpn_peer_vpn_addr_conflict6(struct ovpn_priv *ovpn, + const struct ovpn_peer *peer, + const struct in6_addr *addr) +{ + struct ovpn_peer *tmp = NULL; + + lockdep_assert_held(&ovpn->lock); + + /* we don't hash ::, no conflict in that case */ + if (!ipv6_addr_any(addr)) + tmp = ovpn_peer_get_by_vpn_addr6(ovpn, addr); + + return tmp && tmp != peer; +} + /** * ovpn_peer_transp_match - check if sockaddr and peer binding match * @peer: the peer to get the binding from @@ -1040,6 +1098,13 @@ static int ovpn_peer_add_mp(struct ovpn_priv *ovpn, struct ovpn_peer *peer) goto out; } + /* reject peer with conflicting VPN address */ + if (ovpn_peer_vpn_addr_conflict4(ovpn, NULL, &peer->vpn_addrs.ipv4) || + ovpn_peer_vpn_addr_conflict6(ovpn, NULL, &peer->vpn_addrs.ipv6)) { + ret = -EADDRINUSE; + goto out; + } + bind = rcu_dereference_protected(peer->bind, true); /* peers connected via TCP have bind == NULL */ if (bind) { diff --git a/drivers/net/ovpn/peer.h b/drivers/net/ovpn/peer.h index 063535699ecd..1879bfb76992 100644 --- a/drivers/net/ovpn/peer.h +++ b/drivers/net/ovpn/peer.h @@ -164,6 +164,12 @@ struct ovpn_peer *ovpn_peer_get_by_transp_addr(struct ovpn_priv *ovpn, struct ovpn_peer *ovpn_peer_get_by_id(struct ovpn_priv *ovpn, u32 peer_id); struct ovpn_peer *ovpn_peer_get_by_dst(struct ovpn_priv *ovpn, struct sk_buff *skb); +bool ovpn_peer_vpn_addr_conflict4(struct ovpn_priv *ovpn, + const struct ovpn_peer *peer, + const struct in_addr *addr); +bool ovpn_peer_vpn_addr_conflict6(struct ovpn_priv *ovpn, + const struct ovpn_peer *peer, + const struct in6_addr *addr); void ovpn_peer_hash_vpn_ip(struct ovpn_peer *peer); void ovpn_peer_hash_transp_addr(struct ovpn_peer *peer); bool ovpn_peer_check_by_src(struct ovpn_priv *ovpn, struct sk_buff *skb, From 025af3a0a892514f9f27f186338ba3d44365547a Mon Sep 17 00:00:00 2001 From: Ralf Lici Date: Fri, 28 Aug 2026 15:00:08 +0200 Subject: [PATCH 267/694] ovpn: reject multipeer peers without VPN addresses In MP mode, ovpn uses the peer VPN addresses to select the peer for outgoing tunnel packets. Peer creation currently requires a VPN IPv4 or IPv6 attribute, but it only checks for the presence of the attribute and not for a usable address value. This allows userspace to create an MP peer with only unspecified VPN addresses, or to update an existing peer so that both VPN address families become unspecified. Such a peer cannot be selected through the VPN address hash tables. Reject MP peer creation or update when the resulting peer would not have at least one VPN address configured. This changes such configurations from being accepted to being rejected, but they have never been usable because the peer cannot be selected through the VPN address hash tables. Fixes: 1d36a36f6d53 ("ovpn: implement peer add/get/dump/delete via netlink") Signed-off-by: Ralf Lici Signed-off-by: Antonio Quartulli --- drivers/net/ovpn/netlink.c | 36 ++++++++++++++++++++++++++++++------ 1 file changed, 30 insertions(+), 6 deletions(-) diff --git a/drivers/net/ovpn/netlink.c b/drivers/net/ovpn/netlink.c index e23f7d1f49e0..e9e0f75e0443 100644 --- a/drivers/net/ovpn/netlink.c +++ b/drivers/net/ovpn/netlink.c @@ -352,8 +352,10 @@ static int ovpn_nl_peer_modify(struct ovpn_peer *peer, struct genl_info *info, int ovpn_nl_peer_new_doit(struct sk_buff *skb, struct genl_info *info) { - struct nlattr *attrs[OVPN_A_PEER_MAX + 1]; + struct in_addr vpn_addr4 = { .s_addr = htonl(INADDR_ANY) }; + struct in6_addr vpn_addr6 = IN6ADDR_ANY_INIT; struct ovpn_priv *ovpn = info->user_ptr[0]; + struct nlattr *attrs[OVPN_A_PEER_MAX + 1]; struct ovpn_socket *ovpn_sock; struct socket *sock = NULL; struct ovpn_peer *peer; @@ -377,11 +379,20 @@ int ovpn_nl_peer_new_doit(struct sk_buff *skb, struct genl_info *info) return -EINVAL; /* in MP mode VPN IPs are required for selecting the right peer */ - if (ovpn->mode == OVPN_MODE_MP && !attrs[OVPN_A_PEER_VPN_IPV4] && - !attrs[OVPN_A_PEER_VPN_IPV6]) { - NL_SET_ERR_MSG_FMT_MOD(info->extack, - "VPN IP must be provided in MP mode"); - return -EINVAL; + if (ovpn->mode == OVPN_MODE_MP) { + if (attrs[OVPN_A_PEER_VPN_IPV4]) + vpn_addr4.s_addr = + nla_get_in_addr(attrs[OVPN_A_PEER_VPN_IPV4]); + if (attrs[OVPN_A_PEER_VPN_IPV6]) + vpn_addr6 = + nla_get_in6_addr(attrs[OVPN_A_PEER_VPN_IPV6]); + + if (vpn_addr4.s_addr == htonl(INADDR_ANY) && + ipv6_addr_any(&vpn_addr6)) { + NL_SET_ERR_MSG_FMT_MOD(info->extack, + "at least one VPN IP must be configured in MP mode"); + return -EINVAL; + } } peer_id = nla_get_u32(attrs[OVPN_A_PEER_ID]); @@ -531,6 +542,9 @@ int ovpn_nl_peer_set_doit(struct sk_buff *skb, struct genl_info *info) spin_lock_bh(&ovpn->lock); + vpn_addr4 = peer->vpn_addrs.ipv4; + vpn_addr6 = peer->vpn_addrs.ipv6; + /* reject peer with conflicting VPN address */ if (attrs[OVPN_A_PEER_VPN_IPV4]) { vpn_addr4.s_addr = nla_get_in_addr(attrs[OVPN_A_PEER_VPN_IPV4]); @@ -543,6 +557,16 @@ int ovpn_nl_peer_set_doit(struct sk_buff *skb, struct genl_info *info) goto addr_conflict; } + /* in MP mode VPN IPs are required for selecting the right peer */ + if (ovpn->mode == OVPN_MODE_MP && + vpn_addr4.s_addr == htonl(INADDR_ANY) && + ipv6_addr_any(&vpn_addr6)) { + NL_SET_ERR_MSG_FMT_MOD(info->extack, + "at least one VPN IP must be configured in MP mode"); + ret = -EINVAL; + goto unlock; + } + ret = ovpn_nl_peer_modify(peer, info, attrs); if (ret < 0) goto unlock; From 5940f3407b78062442cb01f541ef6eed709fc380 Mon Sep 17 00:00:00 2001 From: Ralf Lici Date: Fri, 28 Aug 2026 15:00:09 +0200 Subject: [PATCH 268/694] ovpn: reject invalid peer VPN addresses In MP mode, ovpn uses peer VPN addresses as lookup keys for selecting the peer that should receive outgoing tunnel packets. The netlink configuration path currently accepts address values that cannot sensibly identify a VPN peer, such as multicast, broadcast or loopback addresses. Reject invalid peer VPN addresses when creating or updating an MP peer. Keep accepting the unspecified address as the internal unset value, provided that at least one VPN address family remains configured. Fixes: 1d36a36f6d53 ("ovpn: implement peer add/get/dump/delete via netlink") Signed-off-by: Ralf Lici Signed-off-by: Antonio Quartulli --- drivers/net/ovpn/netlink.c | 55 +++++++++++++++++++++++++++++--------- 1 file changed, 42 insertions(+), 13 deletions(-) diff --git a/drivers/net/ovpn/netlink.c b/drivers/net/ovpn/netlink.c index e9e0f75e0443..5432bc2eb8e8 100644 --- a/drivers/net/ovpn/netlink.c +++ b/drivers/net/ovpn/netlink.c @@ -185,6 +185,39 @@ static sa_family_t ovpn_nl_family_get(struct nlattr *addr4, return AF_UNSPEC; } +static int ovpn_nl_peer_check_vpn_addrs(const struct in_addr *addr4, + const struct in6_addr *addr6, + struct genl_info *info) +{ + int addr6_type; + + if (addr4->s_addr == htonl(INADDR_ANY) && ipv6_addr_any(addr6)) { + NL_SET_ERR_MSG_MOD(info->extack, + "at least one VPN IP must be configured in MP mode"); + return -EINVAL; + } + + if (ipv4_is_multicast(addr4->s_addr) || ipv4_is_lbcast(addr4->s_addr) || + ipv4_is_loopback(addr4->s_addr)) { + NL_SET_ERR_MSG_MOD(info->extack, + "VPN IPv4 address must be valid unicast or any"); + return -EADDRNOTAVAIL; + } + + if (!ipv6_addr_any(addr6)) { + addr6_type = ipv6_addr_type(addr6); + + if (!(addr6_type & IPV6_ADDR_UNICAST) || + (addr6_type & (IPV6_ADDR_LOOPBACK | IPV6_ADDR_COMPATv4))) { + NL_SET_ERR_MSG_MOD(info->extack, + "VPN IPv6 address must be valid unicast or any"); + return -EADDRNOTAVAIL; + } + } + + return 0; +} + static int ovpn_nl_peer_precheck(struct ovpn_priv *ovpn, struct genl_info *info, struct nlattr **attrs) @@ -387,12 +420,10 @@ int ovpn_nl_peer_new_doit(struct sk_buff *skb, struct genl_info *info) vpn_addr6 = nla_get_in6_addr(attrs[OVPN_A_PEER_VPN_IPV6]); - if (vpn_addr4.s_addr == htonl(INADDR_ANY) && - ipv6_addr_any(&vpn_addr6)) { - NL_SET_ERR_MSG_FMT_MOD(info->extack, - "at least one VPN IP must be configured in MP mode"); - return -EINVAL; - } + ret = ovpn_nl_peer_check_vpn_addrs(&vpn_addr4, &vpn_addr6, + info); + if (ret < 0) + return ret; } peer_id = nla_get_u32(attrs[OVPN_A_PEER_ID]); @@ -558,13 +589,11 @@ int ovpn_nl_peer_set_doit(struct sk_buff *skb, struct genl_info *info) } /* in MP mode VPN IPs are required for selecting the right peer */ - if (ovpn->mode == OVPN_MODE_MP && - vpn_addr4.s_addr == htonl(INADDR_ANY) && - ipv6_addr_any(&vpn_addr6)) { - NL_SET_ERR_MSG_FMT_MOD(info->extack, - "at least one VPN IP must be configured in MP mode"); - ret = -EINVAL; - goto unlock; + if (ovpn->mode == OVPN_MODE_MP) { + ret = ovpn_nl_peer_check_vpn_addrs(&vpn_addr4, &vpn_addr6, + info); + if (ret < 0) + goto unlock; } ret = ovpn_nl_peer_modify(peer, info, attrs); From 006208026819d5e9e5ec07b3e73d95960a059327 Mon Sep 17 00:00:00 2001 From: Ralf Lici Date: Fri, 28 Aug 2026 15:00:10 +0200 Subject: [PATCH 269/694] selftests: ovpn: validate peer VPN addresses Exercise peer VPN address validation through both peer creation and update. Check missing, unspecified, duplicate, multicast, broadcast, loopback, IPv4-compatible and IPv4-mapped addresses. Temporarily configure a peer with both address families to verify that either family can be cleared while the other remains configured, then restore the original addresses before running the existing traffic tests. Extend ovpn-cli peer updates with an optional VPN address and preserve peer creation errors so the negative tests can observe rejected netlink requests. Signed-off-by: Ralf Lici Signed-off-by: Antonio Quartulli --- tools/testing/selftests/net/ovpn/common.sh | 13 ++++ tools/testing/selftests/net/ovpn/ovpn-cli.c | 54 ++++++++++----- tools/testing/selftests/net/ovpn/test.sh | 75 ++++++++++++++++++++- 3 files changed, 123 insertions(+), 19 deletions(-) diff --git a/tools/testing/selftests/net/ovpn/common.sh b/tools/testing/selftests/net/ovpn/common.sh index 2d844eb3aa6e..5e9c81e885e6 100644 --- a/tools/testing/selftests/net/ovpn/common.sh +++ b/tools/testing/selftests/net/ovpn/common.sh @@ -136,6 +136,19 @@ ovpn_create_ns() { ip netns add "ovpn_peer${1}" } +ovpn_peer_vpn_addr() { + local peer="$1" + local file + + if [ "${OVPN_PROTO}" == "UDP" ]; then + file="${OVPN_UDP_PEERS_FILE}" + else + file="${OVPN_TCP_PEERS_FILE}" + fi + + awk -v peer="${peer}" '$1 == peer {print $NF; exit}' "${file}" +} + ovpn_setup_ns() { local peer="ovpn_peer${1}" local server_ns="ovpn_peer0" diff --git a/tools/testing/selftests/net/ovpn/ovpn-cli.c b/tools/testing/selftests/net/ovpn/ovpn-cli.c index f4effa7580c0..3b612a8a18fe 100644 --- a/tools/testing/selftests/net/ovpn/ovpn-cli.c +++ b/tools/testing/selftests/net/ovpn/ovpn-cli.c @@ -650,6 +650,26 @@ static int ovpn_connect(struct ovpn_ctx *ovpn) return ret; } +static int ovpn_nl_put_vpn_addr(struct nl_msg *msg, + const struct ovpn_ctx *ovpn) +{ + if (!ovpn->peer_ip_set) + return 0; + + switch (ovpn->peer_ip.in4.sin_family) { + case AF_INET: + return nla_put_u32(msg, OVPN_A_PEER_VPN_IPV4, + ovpn->peer_ip.in4.sin_addr.s_addr); + case AF_INET6: + return nla_put(msg, OVPN_A_PEER_VPN_IPV6, + sizeof(struct in6_addr), + &ovpn->peer_ip.in6.sin6_addr); + default: + fprintf(stderr, "Invalid family for peer address\n"); + return -EAFNOSUPPORT; + } +} + static int ovpn_new_peer(struct ovpn_ctx *ovpn, bool is_tcp) { struct nlattr *attr; @@ -691,22 +711,9 @@ static int ovpn_new_peer(struct ovpn_ctx *ovpn, bool is_tcp) } } - if (ovpn->peer_ip_set) { - switch (ovpn->peer_ip.in4.sin_family) { - case AF_INET: - NLA_PUT_U32(ctx->nl_msg, OVPN_A_PEER_VPN_IPV4, - ovpn->peer_ip.in4.sin_addr.s_addr); - break; - case AF_INET6: - NLA_PUT(ctx->nl_msg, OVPN_A_PEER_VPN_IPV6, - sizeof(struct in6_addr), - &ovpn->peer_ip.in6.sin6_addr); - break; - default: - fprintf(stderr, "Invalid family for peer address\n"); - goto nla_put_failure; - } - } + ret = ovpn_nl_put_vpn_addr(ctx->nl_msg, ovpn); + if (ret) + goto nla_put_failure; nla_nest_end(ctx->nl_msg, attr); @@ -732,6 +739,10 @@ static int ovpn_set_peer(struct ovpn_ctx *ovpn) ovpn->keepalive_interval); NLA_PUT_U32(ctx->nl_msg, OVPN_A_PEER_KEEPALIVE_TIMEOUT, ovpn->keepalive_timeout); + + ret = ovpn_nl_put_vpn_addr(ctx->nl_msg, ovpn); + if (ret) + goto nla_put_failure; nla_nest_end(ctx->nl_msg, attr); ret = ovpn_nl_msg_send(ctx, NULL); @@ -1730,13 +1741,14 @@ static void usage(const char *cmd) fprintf(stderr, "\tmark: socket FW mark value\n"); fprintf(stderr, - "* set_peer : set peer attributes\n"); + "* set_peer [vpnaddr]: set peer attributes\n"); fprintf(stderr, "\tiface: ovpn interface name\n"); fprintf(stderr, "\tpeer_id: peer ID of the peer to modify\n"); fprintf(stderr, "\tkeepalive_interval: interval for sending ping messages\n"); fprintf(stderr, "\tkeepalive_timeout: time after which a peer is timed out\n"); + fprintf(stderr, "\tvpnaddr: peer VPN IP\n"); fprintf(stderr, "* del_peer : delete peer\n"); fprintf(stderr, "\tiface: ovpn interface name\n"); @@ -2090,6 +2102,8 @@ static int ovpn_run_cmd(struct ovpn_ctx *ovpn) return ret; ret = ovpn_new_peer(ovpn, false); + if (ret < 0) + return ret; ovpn_waitbg(); break; case CMD_NEW_MULTI_PEER: @@ -2331,6 +2345,12 @@ static int ovpn_parse_cmd_args(struct ovpn_ctx *ovpn, int argc, char *argv[]) "keepalive interval value out of range\n"); return -1; } + + if (argc > 6) { + ret = ovpn_parse_remote(ovpn, NULL, NULL, argv[6]); + if (ret < 0) + return -1; + } break; case CMD_DEL_PEER: if (argc < 4) diff --git a/tools/testing/selftests/net/ovpn/test.sh b/tools/testing/selftests/net/ovpn/test.sh index 9b5610837032..392109d5e14e 100755 --- a/tools/testing/selftests/net/ovpn/test.sh +++ b/tools/testing/selftests/net/ovpn/test.sh @@ -56,6 +56,76 @@ ovpn_prepare_network() { done } +ovpn_new_test_peer() { + local peer_id="$1" + + shift + ip netns exec ovpn_peer0 "${OVPN_CLI}" new_peer tun0 \ + "${peer_id}" none 65000 10.10.1.2 1 "$@" +} + +ovpn_set_peer_vpn_addr() { + ip netns exec ovpn_peer0 "${OVPN_CLI}" set_peer tun0 \ + "$1" 60 120 "$2" +} + +ovpn_run_vpn_addr_validation() { + local addr + local peer1_addr4 + local test_peer_id=$((OVPN_NUM_PEERS + 1)) + local test_peer_addr6="2001:db8::2" + # Do not include 0.0.0.0 or :: here. They are invalid on creation, but + # clear one address family on update and are valid if the other remains. + local -a invalid_addrs=( + "127.0.0.1" + "224.0.0.1" + "255.255.255.255" + "::1" + "::192.0.2.1" + "::ffff:192.0.2.1" + "ff02::1" + ) + + peer1_addr4=$(ovpn_peer_vpn_addr 1) + + ovpn_cmd_fail "reject peer without VPN address" \ + ovpn_new_test_peer "${test_peer_id}" + + for addr in "0.0.0.0" "::" "${invalid_addrs[@]}"; do + ovpn_cmd_fail "reject new peer VPN address ${addr}" \ + ovpn_new_test_peer "${test_peer_id}" "${addr}" + done + + ovpn_cmd_fail "reject duplicate IPv4 address on peer creation" \ + ovpn_new_test_peer "${test_peer_id}" "${peer1_addr4}" + ovpn_cmd_fail "reject clearing the last peer VPN address" \ + ovpn_set_peer_vpn_addr 1 0.0.0.0 + + for addr in "${invalid_addrs[@]}"; do + ovpn_cmd_fail "reject updated peer VPN address ${addr}" \ + ovpn_set_peer_vpn_addr 1 "${addr}" + done + + ovpn_cmd_fail "reject duplicate IPv4 address on peer update" \ + ovpn_set_peer_vpn_addr 2 "${peer1_addr4}" + + ovpn_cmd_ok "add peer IPv6 address" \ + ovpn_set_peer_vpn_addr 1 "${test_peer_addr6}" + ovpn_cmd_fail "reject duplicate IPv6 address on peer creation" \ + ovpn_new_test_peer "${test_peer_id}" "${test_peer_addr6}" + ovpn_cmd_fail "reject duplicate IPv6 address on peer update" \ + ovpn_set_peer_vpn_addr 2 "${test_peer_addr6}" + + ovpn_cmd_ok "clear peer IPv4 address" \ + ovpn_set_peer_vpn_addr 1 0.0.0.0 + ovpn_cmd_fail "reject clearing the remaining peer IPv6 address" \ + ovpn_set_peer_vpn_addr 1 :: + ovpn_cmd_ok "restore peer IPv4 address" \ + ovpn_set_peer_vpn_addr 1 "${peer1_addr4}" + ovpn_cmd_ok "clear peer IPv6 address" \ + ovpn_set_peer_vpn_addr 1 :: +} + ovpn_run_basic_traffic() { local p local header1 @@ -293,15 +363,16 @@ trap ovpn_stage_err ERR ktap_print_header if [ "${OVPN_FLOAT}" == "1" ]; then - ktap_set_plan 13 + ktap_set_plan 14 else - ktap_set_plan 12 + ktap_set_plan 13 fi ovpn_cleanup modprobe -q ovpn || true ovpn_run_stage "setup network topology" ovpn_prepare_network +ovpn_run_stage "validate peer VPN addresses" ovpn_run_vpn_addr_validation ovpn_run_stage "run baseline data traffic" ovpn_run_basic_traffic ovpn_run_stage "run LAN traffic behind peer1" ovpn_run_lan_traffic [ "${OVPN_FLOAT}" == "1" ] && ovpn_run_stage "run floating peer checks" \ From e6bae5034ef4a61f3f062b18140d4f58c8b9a149 Mon Sep 17 00:00:00 2001 From: Niklas Cassel Date: Fri, 18 Sep 2026 14:40:32 +0200 Subject: [PATCH 270/694] ata: libata-core: Extend Samsung LPM quirk to AMD controllers A Samsung SSD 870 QVO 8TB connected to an AMD 600 Series chipset SATA controller is reported to time out on STANDBY IMMEDIATE during system suspend with med_power_with_dipm enabled. The command completes when using max_performance instead. The existing Samsung LPM quirk only matches ATI controllers, leaving AMD controllers unaffected. Rename it to ATA_QUIRK_NO_LPM_ON_ATI_AND_AMD and extend the vendor check to AMD for the same Samsung SSD model patterns. Keep LPM behavior unchanged for other controller vendors, including Intel. Leave ATA_QUIRK_NO_NCQ_ON_ATI restricted to ATI, since the reported AMD issue concerns LPM rather than NCQ. Link: https://bugzilla.kernel.org/show_bug.cgi?id=221986 Reviewed-by: Damien Le Moal Reviewed-by: Mario Limonciello (AMD) > --- Link: https://lore.kernel.org/r/20260918124030.1962773-5-cassel@kernel.org Signed-off-by: Niklas Cassel --- drivers/ata/libata-core.c | 15 ++++++++------- include/linux/libata.h | 4 ++-- 2 files changed, 10 insertions(+), 9 deletions(-) diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c index f482c0a6d7e9..92233fb25051 100644 --- a/drivers/ata/libata-core.c +++ b/drivers/ata/libata-core.c @@ -2972,9 +2972,10 @@ static void ata_dev_config_lpm(struct ata_device *dev) (dev->id[ATA_ID_SATA_CAPABILITY] & 0xe) == 0x2) dev->quirks |= ATA_QUIRK_NOLPM; - /* ATI specific quirk */ - if ((dev->quirks & ATA_QUIRK_NO_LPM_ON_ATI) && - ata_dev_check_adapter(dev, PCI_VENDOR_ID_ATI)) + /* ATI and AMD specific quirk */ + if ((dev->quirks & ATA_QUIRK_NO_LPM_ON_ATI_AND_AMD) && + (ata_dev_check_adapter(dev, PCI_VENDOR_ID_ATI) || + ata_dev_check_adapter(dev, PCI_VENDOR_ID_AMD))) dev->quirks |= ATA_QUIRK_NOLPM; } @@ -4136,7 +4137,7 @@ static const char * const ata_quirk_names[] = { [__ATA_QUIRK_MAX_SEC] = "maxsec", [__ATA_QUIRK_MAX_TRIM_128M] = "maxtrim128m", [__ATA_QUIRK_NO_NCQ_ON_ATI] = "noncqonati", - [__ATA_QUIRK_NO_LPM_ON_ATI] = "nolpmonati", + [__ATA_QUIRK_NO_LPM_ON_ATI_AND_AMD] = "nolpmonatiandamd", [__ATA_QUIRK_NO_ID_DEV_LOG] = "noiddevlog", [__ATA_QUIRK_NO_LOG_DIR] = "nologdir", [__ATA_QUIRK_NO_FUA] = "nofua", @@ -4420,15 +4421,15 @@ static const struct ata_dev_quirks_entry __ata_dev_quirks[] = { { "Samsung SSD 860*", NULL, ATA_QUIRK_NO_NCQ_TRIM | ATA_QUIRK_ZERO_AFTER_TRIM | ATA_QUIRK_NO_NCQ_ON_ATI | - ATA_QUIRK_NO_LPM_ON_ATI }, + ATA_QUIRK_NO_LPM_ON_ATI_AND_AMD }, { "Samsung SSD 870*", NULL, ATA_QUIRK_NO_NCQ_TRIM | ATA_QUIRK_ZERO_AFTER_TRIM | ATA_QUIRK_NO_NCQ_ON_ATI | - ATA_QUIRK_NO_LPM_ON_ATI }, + ATA_QUIRK_NO_LPM_ON_ATI_AND_AMD }, { "SAMSUNG*MZ7LH*", NULL, ATA_QUIRK_NO_NCQ_TRIM | ATA_QUIRK_ZERO_AFTER_TRIM | ATA_QUIRK_NO_NCQ_ON_ATI | - ATA_QUIRK_NO_LPM_ON_ATI }, + ATA_QUIRK_NO_LPM_ON_ATI_AND_AMD }, { "FCCT*M500*", NULL, ATA_QUIRK_NO_NCQ_TRIM | ATA_QUIRK_ZERO_AFTER_TRIM }, diff --git a/include/linux/libata.h b/include/linux/libata.h index 313e96173b19..48bde275968d 100644 --- a/include/linux/libata.h +++ b/include/linux/libata.h @@ -76,7 +76,7 @@ enum ata_quirks { __ATA_QUIRK_MAX_SEC, /* Limit max sectors */ __ATA_QUIRK_MAX_TRIM_128M, /* Limit max trim size to 128M */ __ATA_QUIRK_NO_NCQ_ON_ATI, /* Disable NCQ on ATI chipset */ - __ATA_QUIRK_NO_LPM_ON_ATI, /* Disable LPM on ATI chipset */ + __ATA_QUIRK_NO_LPM_ON_ATI_AND_AMD, /* Disable LPM on ATI and AMD chipsets */ __ATA_QUIRK_NO_ID_DEV_LOG, /* Identify device log missing */ __ATA_QUIRK_NO_LOG_DIR, /* Do not read log directory */ __ATA_QUIRK_NO_FUA, /* Do not use FUA */ @@ -115,7 +115,7 @@ enum { ATA_QUIRK_MAX_SEC = BIT_ULL(__ATA_QUIRK_MAX_SEC), ATA_QUIRK_MAX_TRIM_128M = BIT_ULL(__ATA_QUIRK_MAX_TRIM_128M), ATA_QUIRK_NO_NCQ_ON_ATI = BIT_ULL(__ATA_QUIRK_NO_NCQ_ON_ATI), - ATA_QUIRK_NO_LPM_ON_ATI = BIT_ULL(__ATA_QUIRK_NO_LPM_ON_ATI), + ATA_QUIRK_NO_LPM_ON_ATI_AND_AMD = BIT_ULL(__ATA_QUIRK_NO_LPM_ON_ATI_AND_AMD), ATA_QUIRK_NO_ID_DEV_LOG = BIT_ULL(__ATA_QUIRK_NO_ID_DEV_LOG), ATA_QUIRK_NO_LOG_DIR = BIT_ULL(__ATA_QUIRK_NO_LOG_DIR), ATA_QUIRK_NO_FUA = BIT_ULL(__ATA_QUIRK_NO_FUA), From 88a0474d92ba102fff860db3cae9da87a0964fbf Mon Sep 17 00:00:00 2001 From: Niklas Cassel Date: Fri, 18 Sep 2026 14:40:33 +0200 Subject: [PATCH 271/694] ata: libata: Correct libata.force parameter documentation Align the documented libata.force options with their implementation. The force table accepts PIO modes 0 through 6, not mode 7, and ncqati controls NCQ generally rather than only queued TRIM. The max_sec_1024 and max_sec_lba48 options only set transfer size limits. Remove the misleading claim that they can also clear them. Reviewed-by: Damien Le Moal Link: https://lore.kernel.org/r/20260918124030.1962773-6-cassel@kernel.org Signed-off-by: Niklas Cassel --- Documentation/admin-guide/kernel-parameters.txt | 10 ++++------ 1 file changed, 4 insertions(+), 6 deletions(-) diff --git a/Documentation/admin-guide/kernel-parameters.txt b/Documentation/admin-guide/kernel-parameters.txt index 33cd30996e47..24459f7ff810 100644 --- a/Documentation/admin-guide/kernel-parameters.txt +++ b/Documentation/admin-guide/kernel-parameters.txt @@ -3491,7 +3491,7 @@ Kernel parameters * SATA link speed limit: 1.5Gbps or 3.0Gbps. - * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7]. + * Transfer mode: pio[0-6], mwdma[0-4] and udma[0-7]. udma[/][16,25,33,44,66,100,133] notation is also allowed. @@ -3509,7 +3509,7 @@ Kernel parameters * [no]ncqtrim: Enable or disable queued DSM TRIM. - * [no]ncqati: Enable or disable NCQ trim on ATI chipset. + * [no]ncqati: Enable or disable NCQ on ATI chipsets. * [no]trim: Enable or disable (unqueued) TRIM. @@ -3540,11 +3540,9 @@ Kernel parameters * max_sec_128: Set transfer size limit to 128 sectors. - * max_sec_1024: Set or clear transfer size limit to - 1024 sectors. + * max_sec_1024: Set transfer size limit to 1024 sectors. - * max_sec_lba48: Set or clear transfer size limit to - 65535 sectors. + * max_sec_lba48: Set transfer size limit to 65535 sectors. * external: Mark port as external (hotplug-capable). From 437f3727b4fd715266c296435c116288f82666fd Mon Sep 17 00:00:00 2001 From: Zhiling Zou Date: Mon, 3 Aug 2026 11:28:22 +0800 Subject: [PATCH 272/694] wifi: mac80211: count only matching reservations in reserved switch ieee80211_vif_use_reserved_switch() validates a replacement by counting assigned links on ctx->replace_ctx and treating reservations too broadly. A link can still be assigned to the old channel context while holding a reservation that belongs to a different replacement operation. That can make the current replacement appear complete too early. If another link finalizes first, the function can free the WILL_BE_REPLACED chanctx while the unmatched link and driver still use it, leading to a use-after-free in later beacon generation. Require an in-place reservation to point at the current replacement context, and require the reserved and assigned contexts to be the matching replacement pair, before counting or moving the link. Fixes: 5bcae31d9cb1 ("mac80211: implement multi-vif in-place reservations") Cc: stable@vger.kernel.org Reported-by: Vega Signed-off-by: Zhiling Zou Link: https://patch.msgid.link/4b6c4b8648ab1920c38a40be32c0b1a2797b4ac6.1785726007.git.zhilinz@nebusec.ai Signed-off-by: Johannes Berg --- net/mac80211/chan.c | 30 +++++++++++++++++++++++------- 1 file changed, 23 insertions(+), 7 deletions(-) diff --git a/net/mac80211/chan.c b/net/mac80211/chan.c index 75bb204ad743..17b935448f13 100644 --- a/net/mac80211/chan.c +++ b/net/mac80211/chan.c @@ -1772,7 +1772,8 @@ ieee80211_link_use_reserved_assign(struct ieee80211_link_data *link) } static bool -ieee80211_link_has_in_place_reservation(struct ieee80211_link_data *link) +ieee80211_link_has_in_place_reservation(struct ieee80211_link_data *link, + struct ieee80211_chanctx *ctx) { struct ieee80211_sub_if_data *sdata = link->sdata; struct ieee80211_chanctx *old_ctx, *new_ctx; @@ -1782,6 +1783,9 @@ ieee80211_link_has_in_place_reservation(struct ieee80211_link_data *link) new_ctx = link->reserved_chanctx; old_ctx = ieee80211_link_get_chanctx(link); + if (new_ctx != ctx) + return false; + if (!old_ctx) return false; @@ -1794,6 +1798,12 @@ ieee80211_link_has_in_place_reservation(struct ieee80211_link_data *link) if (new_ctx->replace_state != IEEE80211_CHANCTX_REPLACES_OTHER) return false; + if (new_ctx->replace_ctx != old_ctx) + return false; + + if (old_ctx->replace_ctx != new_ctx) + return false; + return true; } @@ -1823,7 +1833,8 @@ static int ieee80211_chsw_switch_vifs(struct ieee80211_local *local, } for_each_chanctx_user_reserved(local, ctx, &iter) { - if (!ieee80211_link_has_in_place_reservation(iter.link)) + if (!ieee80211_link_has_in_place_reservation(iter.link, + ctx)) continue; old_ctx = ieee80211_link_get_chanctx(iter.link); @@ -1925,7 +1936,9 @@ static int ieee80211_vif_use_reserved_switch(struct ieee80211_local *local) for_each_chanctx_user_assigned(local, ctx->replace_ctx, &iter) { n_assigned++; - if (iter.link && iter.link->reserved_chanctx) { + if (iter.link && + ieee80211_link_has_in_place_reservation(iter.link, + ctx)) { n_reserved++; if (iter.link->reserved_ready) n_ready++; @@ -1950,7 +1963,8 @@ static int ieee80211_vif_use_reserved_switch(struct ieee80211_local *local) use_reserved: ctx->conf.radar_enabled = false; for_each_chanctx_user_reserved(local, ctx, &iter) { - if (ieee80211_link_has_in_place_reservation(iter.link) && + if (ieee80211_link_has_in_place_reservation(iter.link, + ctx) && !iter.link->reserved_ready) return -EAGAIN; @@ -1991,7 +2005,8 @@ static int ieee80211_vif_use_reserved_switch(struct ieee80211_local *local) } for_each_chanctx_user_reserved(local, ctx, &iter) { - if (!ieee80211_link_has_in_place_reservation(iter.link)) + if (!ieee80211_link_has_in_place_reservation(iter.link, + ctx)) continue; ieee80211_chan_bw_change(local, @@ -2040,7 +2055,7 @@ static int ieee80211_vif_use_reserved_switch(struct ieee80211_local *local) struct ieee80211_bss_conf *link_conf = link->conf; u64 changed = 0; - if (!ieee80211_link_has_in_place_reservation(link)) + if (!ieee80211_link_has_in_place_reservation(link, ctx)) continue; rcu_assign_pointer(link_conf->chanctx_conf, @@ -2091,7 +2106,8 @@ static int ieee80211_vif_use_reserved_switch(struct ieee80211_local *local) for_each_chanctx_user_reserved(local, ctx, &iter) { struct ieee80211_link_data *link = iter.link; - if (WARN_ON(ieee80211_link_has_in_place_reservation(link))) + if (WARN_ON(ieee80211_link_has_in_place_reservation(link, + ctx))) continue; if (!link->reserved_ready) From 3412e9a800786d1dec4a58367db1cba275e7ef11 Mon Sep 17 00:00:00 2001 From: Zhiling Zou Date: Mon, 3 Aug 2026 12:35:10 +0800 Subject: [PATCH 273/694] wifi: mac80211: keep paired old chanctx alive An in-place replacement keeps reciprocal replace_ctx pointers between the old WILL_BE_REPLACED chanctx and the new REPLACES_OTHER chanctx. However, the early free paths only consider assigned and reserved link users when deciding whether to free the old chanctx. That lets the old chanctx be freed too early while the replacement partner still points back to it, either when the last assigned link is released or when that link successfully reassigns to another existing context. Later unreserve and switch-finalization paths can then follow a stale replace_ctx pointer. Keep a paired old chanctx alive until the replacement pair has been torn down. Apply the same check to both __ieee80211_link_release_channel() and ieee80211_link_use_reserved_reassign(). If the replacement is later abandoned, ieee80211_link_unreserve_chanctx() already tears down the pairing before freeing the old chanctx once no users remain. Fixes: 5bcae31d9cb1 ("mac80211: implement multi-vif in-place reservations") Cc: stable@vger.kernel.org Reported-by: Vega Signed-off-by: Zhiling Zou Link: https://patch.msgid.link/b9b582b0db3814f641648b8886bb50f9d68f93ac.1785730815.git.zhilinz@nebusec.ai Signed-off-by: Johannes Berg --- net/mac80211/chan.c | 20 ++++++++++++++++++-- 1 file changed, 18 insertions(+), 2 deletions(-) diff --git a/net/mac80211/chan.c b/net/mac80211/chan.c index 17b935448f13..9cba79507767 100644 --- a/net/mac80211/chan.c +++ b/net/mac80211/chan.c @@ -231,6 +231,13 @@ int ieee80211_chanctx_refcount(struct ieee80211_local *local, return num; } +static bool +ieee80211_chanctx_has_replace_partner(struct ieee80211_chanctx *ctx) +{ + return ctx->replace_state == IEEE80211_CHANCTX_WILL_BE_REPLACED && + ctx->replace_ctx; +} + static int ieee80211_num_chanctx(struct ieee80211_local *local, int radio_idx) { struct ieee80211_chanctx *ctx; @@ -1405,6 +1412,7 @@ void ieee80211_link_unreserve_chanctx(struct ieee80211_link_data *link) { struct ieee80211_sub_if_data *sdata = link->sdata; struct ieee80211_chanctx *ctx = link->reserved_chanctx; + struct ieee80211_chanctx *old_ctx = NULL; lockdep_assert_wiphy(sdata->local->hw.wiphy); @@ -1418,6 +1426,7 @@ void ieee80211_link_unreserve_chanctx(struct ieee80211_link_data *link) if (WARN_ON(!ctx->replace_ctx)) return; + old_ctx = ctx->replace_ctx; WARN_ON(ctx->replace_ctx->replace_state != IEEE80211_CHANCTX_WILL_BE_REPLACED); WARN_ON(ctx->replace_ctx->replace_ctx != ctx); @@ -1428,6 +1437,11 @@ void ieee80211_link_unreserve_chanctx(struct ieee80211_link_data *link) list_del_rcu(&ctx->list); kfree_rcu(ctx, rcu_head); + + if (ieee80211_chanctx_refcount(sdata->local, + old_ctx) == 0) + ieee80211_free_chanctx(sdata->local, old_ctx, + false); } else { ieee80211_free_chanctx(sdata->local, ctx, false); } @@ -1707,7 +1721,8 @@ ieee80211_link_use_reserved_reassign(struct ieee80211_link_data *link) ieee80211_check_fast_xmit_iface(sdata); - if (ieee80211_chanctx_refcount(local, old_ctx) == 0) + if (ieee80211_chanctx_refcount(local, old_ctx) == 0 && + !ieee80211_chanctx_has_replace_partner(old_ctx)) ieee80211_free_chanctx(local, old_ctx, false); ieee80211_recalc_chanctx_min_def(local, new_ctx); @@ -2192,7 +2207,8 @@ void __ieee80211_link_release_channel(struct ieee80211_link_data *link, } ieee80211_assign_link_chanctx(link, NULL, false); - if (ieee80211_chanctx_refcount(local, ctx) == 0) + if (ieee80211_chanctx_refcount(local, ctx) == 0 && + !ieee80211_chanctx_has_replace_partner(ctx)) ieee80211_free_chanctx(local, ctx, skip_idle_recalc); link->radar_required = false; From 7c1611780823b76219cb1478db53dfd890c4a9a2 Mon Sep 17 00:00:00 2001 From: Tianchu Chen Date: Fri, 4 Sep 2026 15:23:34 +0000 Subject: [PATCH 274/694] wifi: cw1200: fix TX padding OOB read leaking heap data to device The TX path rounds the transfer size up to the bus alignment (align_size(), up to 511 bytes for SDIO, 1 byte for SPI) but sends the padded length straight from the frame buffer, which has no room reserved for the padding. cw1200_data_write() therefore reads past the end of the frame and the bus transfer carries those bytes to the device. A bogus device can collect up to 511 bytes of kernel heap data per TX frame. The RX path does not have this problem: it allocates the skb with the aligned length (alloc_len) before reading. Compute the exact padding with the bus align_size() callback, make sure queued TX frames have tailroom for it, and zero the padding bytes before the transfer so nothing past the frame is exposed. This is not expected to change driver behavior on most cases: frame contents and transfer sizes are unchanged, only the content of the padding bytes (previously undefined stale memory, now zero), which the device never reads since the WSM header carries the real frame length. The skb expansion can only fail under memory exhaustion, in which case the frame is dropped like on the existing error paths of cw1200_tx(). Discovered by Atuin - Automated Vulnerability Discovery Engine. Fixes: a910e4a94f69 ("cw1200: add driver for the ST-E CW1100 & CW1200 WLAN chipsets") Cc: stable@vger.kernel.org Assisted-by: LLM Signed-off-by: Tianchu Chen Link: https://patch.msgid.link/7982b4a7e64f74e2068790ec389ca680a081f730@linux.dev Signed-off-by: Johannes Berg --- drivers/net/wireless/st/cw1200/bh.c | 8 +++++++- drivers/net/wireless/st/cw1200/txrx.c | 14 ++++++++++++++ 2 files changed, 21 insertions(+), 1 deletion(-) diff --git a/drivers/net/wireless/st/cw1200/bh.c b/drivers/net/wireless/st/cw1200/bh.c index b034bab4b489..9367a981f1de 100644 --- a/drivers/net/wireless/st/cw1200/bh.c +++ b/drivers/net/wireless/st/cw1200/bh.c @@ -344,6 +344,7 @@ static int cw1200_bh_tx_helper(struct cw1200_common *priv, int *tx_burst) { size_t tx_len; + size_t aligned_len; u8 *data; int ret; struct wsm_hdr *wsm; @@ -376,9 +377,14 @@ static int cw1200_bh_tx_helper(struct cw1200_common *priv, atomic_inc(&priv->bh_tx); - tx_len = priv->hwbus_ops->align_size( + aligned_len = priv->hwbus_ops->align_size( priv->hwbus_priv, tx_len); + /* Zero the bus alignment padding so no stale data is sent. */ + if (aligned_len > tx_len) + memset(data + tx_len, 0, aligned_len - tx_len); + tx_len = aligned_len; + /* Check if not exceeding CW1200 capabilities */ if (WARN_ON_ONCE(tx_len > EFFECTIVE_BUF_SIZE)) pr_debug("Write aligned len: %zu\n", tx_len); diff --git a/drivers/net/wireless/st/cw1200/txrx.c b/drivers/net/wireless/st/cw1200/txrx.c index 084d52b11f5b..e13695667a00 100644 --- a/drivers/net/wireless/st/cw1200/txrx.c +++ b/drivers/net/wireless/st/cw1200/txrx.c @@ -14,6 +14,7 @@ #include "wsm.h" #include "bh.h" #include "sta.h" +#include "hwbus.h" #include "debug.h" #define CW1200_INVALID_RATE_ID (0xFF) @@ -716,6 +717,7 @@ void cw1200_tx(struct ieee80211_hw *dev, struct ieee80211_sta *sta; struct wsm_tx *wsm; bool tid_update = false; + size_t pad_len; u8 flags = 0; int ret; @@ -764,6 +766,18 @@ void cw1200_tx(struct ieee80211_hw *dev, sta = t.sta; + /* + * The bus transfer is padded up to the bus alignment and the + * padding is transferred from the frame buffer. Make sure the + * buffer has room for the padding so the device never receives + * data past the end of the frame. + */ + pad_len = priv->hwbus_ops->align_size(priv->hwbus_priv, skb->len) - + skb->len; + if (skb_tailroom(skb) < pad_len && + pskb_expand_head(skb, 0, pad_len, GFP_ATOMIC)) + goto drop; + spin_lock_bh(&priv->ps_state_lock); { tid_update = cw1200_tx_h_pm_state(priv, &t); From ce67ebe9d8629776ce38e6995ae72fb903c828db Mon Sep 17 00:00:00 2001 From: Aamir Ahmed Date: Sat, 12 Sep 2026 14:47:55 +0100 Subject: [PATCH 275/694] wifi: wfx: validate MIB read length before copying to caller wfx_hif_read_mib() copies reply->length bytes from the firmware response into the caller-provided buffer without checking that reply->length does not exceed val_len. A firmware response with a length field larger than expected overruns the caller buffer and reads past the end of the reply allocation. Replace the dead -ENOMEM check with an active validation that reply->length fits within val_len before the memcpy. Fixes: f95a29d40782 ("staging: wfx: add HIF commands helpers") Assisted-by: LLM Signed-off-by: Aamir Ahmed Link: https://patch.msgid.link/AS8P251MB000155034FC391A2A54B9803C8BD2@AS8P251MB0001.EURP251.PROD.OUTLOOK.COM Signed-off-by: Johannes Berg --- drivers/net/wireless/silabs/wfx/hif_tx.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/net/wireless/silabs/wfx/hif_tx.c b/drivers/net/wireless/silabs/wfx/hif_tx.c index 9f403d275cb1..9efb108bd877 100644 --- a/drivers/net/wireless/silabs/wfx/hif_tx.c +++ b/drivers/net/wireless/silabs/wfx/hif_tx.c @@ -207,9 +207,11 @@ int wfx_hif_read_mib(struct wfx_dev *wdev, int vif_id, u16 mib_id, void *val, si dev_warn(wdev->dev, "%s: confirmation mismatch request\n", __func__); ret = -EIO; } - if (ret == -ENOMEM) + if (!ret && le16_to_cpu(reply->length) > val_len) { dev_err(wdev->dev, "buffer is too small to receive %s (%zu < %d)\n", wfx_get_mib_name(mib_id), val_len, le16_to_cpu(reply->length)); + ret = -EIO; + } if (!ret) memcpy(val, &reply->mib_data, le16_to_cpu(reply->length)); else From f8ebe286394c0bbb4a6dbacb182a34ab6376922d Mon Sep 17 00:00:00 2001 From: Guangshuo Li Date: Wed, 16 Sep 2026 15:37:36 +0800 Subject: [PATCH 276/694] wifi: iwlegacy: 3945: free EEPROM data on remove il3945_pci_probe() calls il_eeprom_init(), which allocates memory for the EEPROM data. The probe failure paths release this allocation with il_eeprom_free(), but the remove path does not free it after a successful probe. As a result, the EEPROM data is leaked when the PCI device is removed. Free the EEPROM data in il3945_pci_remove() before releasing the mac80211 hardware structure. This issue was found by manual code inspection. Fixes: 073d3f5f1b3b ("iwlwifi: changing EEPROM layout handling") Cc: stable@vger.kernel.org Signed-off-by: Guangshuo Li Acked-by: Stanislaw Gruszka Link: https://patch.msgid.link/20260916073736.2952028-1-lgs201920130244@gmail.com Signed-off-by: Johannes Berg --- drivers/net/wireless/intel/iwlegacy/3945-mac.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/net/wireless/intel/iwlegacy/3945-mac.c b/drivers/net/wireless/intel/iwlegacy/3945-mac.c index cbaf250626c5..f7ca182d1494 100644 --- a/drivers/net/wireless/intel/iwlegacy/3945-mac.c +++ b/drivers/net/wireless/intel/iwlegacy/3945-mac.c @@ -3807,6 +3807,7 @@ il3945_pci_remove(struct pci_dev *pdev) il_free_channel_map(il); il_free_geos(il); + il_eeprom_free(il); kfree(il->scan_cmd); dev_kfree_skb(il->beacon_skb); ieee80211_free_hw(il->hw); From ea4debcd8016f73c5dee3a29250a3d7977f015ef Mon Sep 17 00:00:00 2001 From: Sk Anirban Date: Wed, 9 Sep 2026 17:19:32 +0530 Subject: [PATCH 277/694] drm/xe/gt_throttle: Report power brake as a throttle reason on CRI CRI defines bit 5 of the perf limit reasons register as a power brake (PWRBRK) indicator. Add PWRBRK_MASK and a reason_pwrbrk sysfs attribute for CRI in place of reason_ratl. Signed-off-by: Sk Anirban Fixes: 8578e6d0546c ("drm/xe/gt_throttle: Drop individual show functions") Reviewed-by: Raag Jadav Signed-off-by: Matthew Brost Link: https://patch.msgid.link/20260909114931.1039331-2-sk.anirban@intel.com (cherry picked from commit e199c851c0ab608a0ca89e7be1756a1461b41a2c) Signed-off-by: Rodrigo Vivi --- drivers/gpu/drm/xe/regs/xe_gt_regs.h | 1 + drivers/gpu/drm/xe/xe_gt_throttle.c | 5 +++-- 2 files changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/xe/regs/xe_gt_regs.h b/drivers/gpu/drm/xe/regs/xe_gt_regs.h index 08251c7a1a4b..247a736a54aa 100644 --- a/drivers/gpu/drm/xe/regs/xe_gt_regs.h +++ b/drivers/gpu/drm/xe/regs/xe_gt_regs.h @@ -651,6 +651,7 @@ #define MEM_THERMAL_MASK REG_BIT(2) #define VR_THERMAL_MASK REG_BIT(3) #define ICCMAX_MASK REG_BIT(4) +#define PWRBRK_MASK REG_BIT(5) #define SOC_AVG_THERMAL_MASK REG_BIT(6) #define FASTVMODE_MASK REG_BIT(7) #define PSYS_PL1_MASK REG_BIT(12) diff --git a/drivers/gpu/drm/xe/xe_gt_throttle.c b/drivers/gpu/drm/xe/xe_gt_throttle.c index 1e7e3a31aa69..c0af5484611d 100644 --- a/drivers/gpu/drm/xe/xe_gt_throttle.c +++ b/drivers/gpu/drm/xe/xe_gt_throttle.c @@ -39,7 +39,7 @@ * - ``reason_mem_thermal``: Memory thermal * - ``reason_vr_thermal``: VR thermal * - ``reason_iccmax``: ICCMAX - * - ``reason_ratl``: RATL thermal algorithm + * - ``reason_pwrbrk``: Power brake * - ``reason_soc_avg_thermal``: SoC average temp * - ``reason_fastvmode``: VR is hitting FastVMode * - ``reason_psys_pl1``: PSYS PL1 @@ -200,6 +200,7 @@ static THROTTLE_ATTR_RO(reason_psys_pl1, PSYS_PL1_MASK); static THROTTLE_ATTR_RO(reason_psys_pl2, PSYS_PL2_MASK); static THROTTLE_ATTR_RO(reason_p0_freq, P0_FREQ_MASK); static THROTTLE_ATTR_RO(reason_psys_crit, PSYS_CRIT_MASK); +static THROTTLE_ATTR_RO(reason_pwrbrk, PWRBRK_MASK); static struct attribute *cri_throttle_attrs[] = { /* Common */ @@ -209,12 +210,12 @@ static struct attribute *cri_throttle_attrs[] = { &attr_reason_pl2.attr.attr, &attr_reason_pl4.attr.attr, &attr_reason_prochot.attr.attr, - &attr_reason_ratl.attr.attr, /* CRI */ &attr_reason_vr_thermal.attr.attr, &attr_reason_soc_thermal.attr.attr, &attr_reason_mem_thermal.attr.attr, &attr_reason_iccmax.attr.attr, + &attr_reason_pwrbrk.attr.attr, &attr_reason_soc_avg_thermal.attr.attr, &attr_reason_fastvmode.attr.attr, &attr_reason_psys_pl1.attr.attr, From c7a925c84704ec431598f9411dd89c0e16ae34ae Mon Sep 17 00:00:00 2001 From: Shuicheng Lin Date: Mon, 14 Sep 2026 21:53:18 +0000 Subject: [PATCH 278/694] drm/xe/tlb_inval: Treat wedged-device invalidations as complete A TLB invalidation issued on a wedged device fails with -ENOTRECOVERABLE. xe_tlb_inval_issue() squashes only -ECANCELED, so the error reaches callers that treat it as unexpected and WARN, tainting the kernel on a wedge that was deliberately caused: ggtt_invalidate_gt_tlb() drivers/gpu/drm/xe/xe_ggtt.c xe_svm_invalidate() drivers/gpu/drm/xe/xe_svm.c xe_bo_trigger_rebind() drivers/gpu/drm/xe/xe_bo.c xe_vma_userptr_do_inval() drivers/gpu/drm/xe/xe_userptr.c igt@xe_exec_reset@gt-reset-fault-injection hits the GGTT one, turning an otherwise passing run into an abort: *ERROR* SIGID=102 FATAL (-EIO) WEDGED: Device declared wedged! Tile0: GT1: Failed to invalidate GGTT (-ENOTRECOVERABLE) WARNING: drivers/gpu/drm/xe/xe_ggtt.c:588 at ggtt_invalidate_gt_tlb Workqueue: xe-guc-destroy-wq __guc_exec_queue_destroy_async [xe] ggtt_node_remove+0xe3/0x100 [xe] xe_ggtt_remove_bo+0x89/0x2c0 [xe] xe_ttm_bo_destroy+0xcb/0x330 [xe] ... xe_lrc_destroy+0x74/0x90 [xe] xe_exec_queue_fini+0x2d/0x60 [xe] -ECANCELED and -ENOTRECOVERABLE mean the same thing at this layer: the message was dropped rather than delivered, and the fence has already been signalled before the error is returned, so there is nothing left to wait for. Squash both. A wedged device is only recovered by a fresh initialisation, so the error return in xe_bo_trigger_rebind() becomes unreachable by design. v2: fix all invalidation paths. (Sashiko) Fixes: 50fa9acac26f ("drm/xe/guc: distinguish wedged from recoverable cancellation") Assisted-by: Claude:claude-opus-5 Cc: Sk Anirban Cc: Matthew Brost Signed-off-by: Shuicheng Lin Reviewed-by: Matthew Brost Signed-off-by: Matthew Brost Link: https://patch.msgid.link/20260914215318.200603-1-shuicheng.lin@intel.com (cherry picked from commit af14e3705345cb57c53b237169873233052a5c16) Signed-off-by: Rodrigo Vivi --- drivers/gpu/drm/xe/xe_tlb_inval.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/xe/xe_tlb_inval.c b/drivers/gpu/drm/xe/xe_tlb_inval.c index bbd21d393062..67b205bf1291 100644 --- a/drivers/gpu/drm/xe/xe_tlb_inval.c +++ b/drivers/gpu/drm/xe/xe_tlb_inval.c @@ -280,7 +280,8 @@ static void xe_tlb_inval_fence_prep(struct xe_tlb_inval_fence *fence) xe_tlb_inval_fence_signal_unlocked((__fence)); \ mutex_unlock(&(__tlb_inval)->seqno_lock); \ \ - __ret == -ECANCELED ? 0 : __ret; \ + /* Undelivered: fence already signalled, report done */ \ + (__ret == -ECANCELED || __ret == -ENOTRECOVERABLE) ? 0 : __ret; \ }) /** From be1df8badae513e01d9575398438716cfae18655 Mon Sep 17 00:00:00 2001 From: Matthew Brost Date: Thu, 17 Sep 2026 13:31:58 -0700 Subject: [PATCH 279/694] drm/xe: Keep walking on SVM eviction failure The desired behavior for SVM eviction failures, which can occur due to various uncontrollable races, is for TTM to continue walking the LRU list and look for another eviction candidate. This is expressed by returning -ENOSPC from the ->move() callback. Adjust the SVM eviction failure path because of races in ->move() to return -ENOSPC so that TTM continues searching for another buffer to evict. Fixes: 3ca608dc7561 ("drm/xe: Basic SVM BO eviction") Cc: stable@vger.kernel.org Signed-off-by: Matthew Brost Reviewed-by: Himal Prasad Ghimiray Link: https://patch.msgid.link/20260917203158.292823-1-matthew.brost@intel.com Signed-off-by: Rodrigo Vivi (cherry picked from commit 36a86c23588b8f57c9d20feb4cf5a2ab27e3baba) Signed-off-by: Rodrigo Vivi --- drivers/gpu/drm/xe/xe_bo.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/drivers/gpu/drm/xe/xe_bo.c b/drivers/gpu/drm/xe/xe_bo.c index dde309821237..b4921a627ff3 100644 --- a/drivers/gpu/drm/xe/xe_bo.c +++ b/drivers/gpu/drm/xe/xe_bo.c @@ -1037,6 +1037,13 @@ static int xe_bo_move(struct ttm_buffer_object *ttm_bo, bool evict, } else { drm_dbg(&xe->drm, "Evict system allocator BO failed=%pe\n", ERR_PTR(ret)); + /* + * The semantic we want upon SVM eviction failure + * because of racing access is keep walking for + * eviction, which is -ENOSPC. + */ + if (ret == -EBUSY) + ret = -ENOSPC; } goto out; From 24a22fb3c731474b68e986af6804db450fb88617 Mon Sep 17 00:00:00 2001 From: Matthew Auld Date: Fri, 18 Sep 2026 14:10:35 +0100 Subject: [PATCH 280/694] drm/xe/vm: nuke PTs only after unlinking contested VMAs MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit In xe_vm_close_and_put(), external-BO VMAs are queued on the contested list for deferred destruction via xe_vma_destroy_unlocked(). However, xe_vm_pt_destroy() was previously invoked before processing contested VMAs, destroying vm->pt_root while those VMAs were still linked to their respective buffer objects (vm_bo->list.gpuva). If a concurrent thread evicts one of those shared buffer objects, xe_bo_trigger_rebind() holding only bo->resv walks the BO's VMAs and, in fault mode, calls xe_vm_invalidate_vma() -> xe_pt_zap_ptes(). Because vm->pt_root[tile->id] is already NULL, dereferencing pt->level causes a NULL ptr deref. Fix this by deferring xe_vm_free_scratch() and xe_vm_pt_destroy() until after all contested VMAs have been unlinked and destroyed. User is reporting hitting a NULL ptr deref in xe_pt_zap_ptes(), which could be explained by this race. Assisted-by: LLM Fixes: b06d47be7c83 ("drm/xe: Port Xe to GPUVA") Link: https://gitlab.freedesktop.org/drm/xe/kernel/-/work_items/9290 Signed-off-by: Matthew Auld Cc: Thomas Hellström Cc: Matthew Brost Cc: # v6.12+ Reviewed-by: Thomas Hellström Reviewed-by: Matthew Brost Link: https://patch.msgid.link/20260918131034.598078-2-matthew.auld@intel.com (cherry picked from commit c2863648959489767f08892fd6e90577d2ea0b6a) Signed-off-by: Rodrigo Vivi --- drivers/gpu/drm/xe/xe_vm.c | 21 +++++++++------------ 1 file changed, 9 insertions(+), 12 deletions(-) diff --git a/drivers/gpu/drm/xe/xe_vm.c b/drivers/gpu/drm/xe/xe_vm.c index 23952ad8951e..ef20e205a734 100644 --- a/drivers/gpu/drm/xe/xe_vm.c +++ b/drivers/gpu/drm/xe/xe_vm.c @@ -1947,21 +1947,13 @@ void xe_vm_close_and_put(struct xe_vm *vm) vma->gpuva.flags |= XE_VMA_DESTROYED; } - /* - * All vm operations will add shared fences to resv. - * The only exception is eviction for a shared object, - * but even so, the unbind when evicted would still - * install a fence to resv. Hence it's safe to - * destroy the pagetables immediately. - */ - xe_vm_free_scratch(vm); - xe_vm_pt_destroy(vm); xe_vm_unlock(vm); /* - * VM is now dead, cannot re-add nodes to vm->vmas if it's NULL - * Since we hold a refcount to the bo, we can remove and free - * the members safely without locking. + * Unlink and destroy all contested external-BO VMAs before destroying + * the page tables. Otherwise, concurrent eviction holding only bo->resv + * can walk the BO's VMAs and attempt to invalidate/zap page tables that + * have already been freed. */ list_for_each_entry_safe(vma, next_vma, &contested, combined_links.destroy) { @@ -1969,6 +1961,11 @@ void xe_vm_close_and_put(struct xe_vm *vm) xe_vma_destroy_unlocked(vma); } + xe_vm_lock(vm, false); + xe_vm_free_scratch(vm); + xe_vm_pt_destroy(vm); + xe_vm_unlock(vm); + xe_svm_fini(vm); up_write(&vm->lock); From f0ca020cbb9bb7f3f4ea8ba1dfcf30a282aec91e Mon Sep 17 00:00:00 2001 From: Hui Peng Date: Sat, 19 Sep 2026 22:17:38 +0000 Subject: [PATCH 281/694] Bluetooth: bnep: fix out-of-bounds reads on short RX/TX frames and control fallthrough Fix multiple out-of-bounds reads in Bluetooth BNEP frame processing: 1. In bnep_rx_frame() and bnep_ctrl_frame() (net/bluetooth/bnep/core.c), use pskb_may_pull() to verify the BNEP header, control type byte, filter count, and extension headers exist before reading them, and return 0 after handling BNEP_CONTROL instead of falling through to Ethernet frame submission when no extension headers follow. 2. In bnep_net_xmit() (net/bluetooth/bnep/netdev.c), verify skb->len >= ETH_HLEN with pskb_may_pull() before reading the 14-byte Ethernet header to prevent an out-of-bounds heap read and infoleak on short AF_PACKET TX frames. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Assisted-by: LLM Signed-off-by: Hui Peng Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/bnep/core.c | 17 ++++++++++++++++- net/bluetooth/bnep/netdev.c | 8 +++++++- 2 files changed, 23 insertions(+), 2 deletions(-) diff --git a/net/bluetooth/bnep/core.c b/net/bluetooth/bnep/core.c index f7d88c33e23e..ad24d2486665 100644 --- a/net/bluetooth/bnep/core.c +++ b/net/bluetooth/bnep/core.c @@ -270,9 +270,14 @@ static int bnep_rx_extension(struct bnep_session *s, struct sk_buff *skb) BT_DBG("type 0x%x len %u", h->type, h->len); + if (skb->len < h->len) { + err = -EILSEQ; + break; + } + switch (h->type & BNEP_TYPE_MASK) { case BNEP_EXT_CONTROL: - bnep_rx_control(s, skb->data, skb->len); + bnep_rx_control(s, skb->data, h->len); break; default: @@ -373,6 +378,11 @@ static int bnep_rx_frame(struct bnep_session *s, struct sk_buff *skb) goto badframe; } + if ((type & BNEP_TYPE_MASK) == BNEP_CONTROL) { + kfree_skb(skb); + return 0; + } + /* Strip 802.1p header */ if (ntohs(s->eh.h_proto) == ETH_P_8021Q) { if (!skb_pull(skb, 4)) @@ -451,6 +461,11 @@ static int bnep_tx_frame(struct bnep_session *s, struct sk_buff *skb) goto send; } + if (skb->len < ETH_HLEN) { + kfree_skb(skb); + return 0; + } + iv[il++] = (struct kvec) { &type, 1 }; len++; diff --git a/net/bluetooth/bnep/netdev.c b/net/bluetooth/bnep/netdev.c index ee1e39a3daff..b451ef457741 100644 --- a/net/bluetooth/bnep/netdev.c +++ b/net/bluetooth/bnep/netdev.c @@ -166,6 +166,12 @@ static netdev_tx_t bnep_net_xmit(struct sk_buff *skb, BT_DBG("skb %p, dev %p", skb, dev); + if (!pskb_may_pull(skb, ETH_HLEN)) { + dev->stats.tx_dropped++; + kfree_skb(skb); + return NETDEV_TX_OK; + } + #ifdef CONFIG_BT_BNEP_MC_FILTER if (bnep_net_mc_filter(skb, s)) { kfree_skb(skb); @@ -218,7 +224,7 @@ void bnep_net_setup(struct net_device *dev) dev->addr_len = ETH_ALEN; ether_setup(dev); - dev->min_mtu = 0; + dev->min_mtu = ETH_MIN_MTU; dev->max_mtu = ETH_MAX_MTU; dev->priv_flags &= ~IFF_TX_SKB_SHARING; dev->netdev_ops = &bnep_netdev_ops; From 37a11129345337efd6eef8e62b03b6348cd0dd8b Mon Sep 17 00:00:00 2001 From: Ravindra Date: Tue, 15 Sep 2026 10:42:15 +0530 Subject: [PATCH 282/694] Bluetooth: btintel_pcie: validate device-supplied DMA indices In btintel_pcie_msix_rx_handle(), the driver processes RX completion descriptors (urbd1) written by the PCIe device into DMA-coherent memory. urbd1->frbd_tag (a 16-bit field fully controlled by the device firmware via DMA) is used directly as an array index into rxq->bufs[] without any bounds check. rxq->bufs[] has only BTINTEL_PCIE_RX_DESCS_COUNT (64) entries, while frbd_tag can be any value 0-65535. A malicious or malfunctioning device can write an out-of-range frbd_tag, causing the driver to dereference an out-of-bounds data_buf pointer. Additionally, cr_hia is read from a DMA-shared index array also writable by the device; if the device sets cr_hia >= rxq->count, the while-loop never terminates because cr_tia is wrapped via modulo rxq->count and can never equal an out-of-range cr_hia. Add bounds validation for cr_hia and frbd_tag in the RX path, and cr_hia in the TX path. Log invalid values with bt_dev_err before returning. Fixes: c2b636b3f788 ("Bluetooth: btintel_pcie: Add support for PCIe transport") Signed-off-by: Ravindra Signed-off-by: Luiz Augusto von Dentz --- drivers/bluetooth/btintel_pcie.c | 16 ++++++++++++++++ 1 file changed, 16 insertions(+) diff --git a/drivers/bluetooth/btintel_pcie.c b/drivers/bluetooth/btintel_pcie.c index 6e6e2b19815c..2819e001797b 100644 --- a/drivers/bluetooth/btintel_pcie.c +++ b/drivers/bluetooth/btintel_pcie.c @@ -1099,6 +1099,11 @@ static void btintel_pcie_msix_tx_handle(struct btintel_pcie_data *data) txq = &data->txq; + if (cr_hia >= txq->count) { + bt_dev_err(data->hdev, "TXQ: invalid cr_hia %u", cr_hia); + return; + } + while (cr_tia != cr_hia) { data->tx_wait_done = true; wake_up(&data->tx_wait_q); @@ -1650,6 +1655,11 @@ static void btintel_pcie_msix_rx_handle(struct btintel_pcie_data *data) rxq = &data->rxq; + if (cr_hia >= rxq->count) { + bt_dev_err(hdev, "RXQ: invalid cr_hia %u", cr_hia); + return; + } + /* The firmware sends multiple CD in a single MSI-X and it needs to * process all received CDs in this interrupt. */ @@ -1657,6 +1667,12 @@ static void btintel_pcie_msix_rx_handle(struct btintel_pcie_data *data) urbd1 = &rxq->urbd1s[cr_tia]; ipc_print_urbd1(data->hdev, urbd1, cr_tia); + if (urbd1->frbd_tag >= rxq->count) { + bt_dev_err(hdev, "RXQ: invalid frbd_tag %u", + urbd1->frbd_tag); + return; + } + buf = &rxq->bufs[urbd1->frbd_tag]; if (!buf) { bt_dev_err(hdev, "RXQ: failed to get the DMA buffer for %d", From 46f8ffd0a1f1eb6cbc94946a92c11ef601e228a1 Mon Sep 17 00:00:00 2001 From: Hui Peng Date: Sat, 19 Sep 2026 11:25:18 +0000 Subject: [PATCH 283/694] Bluetooth: RFCOMM: fix NULL dereference of dlc->session in RFCOMM_CONNINFO The RFCOMM_CONNINFO getsockopt handler accepts a socket that is not connected as long as deferred setup is enabled: if (sk->sk_state != BT_CONNECTED && !rfcomm_pi(sk)->dlc->defer_setup) { err = -ENOTCONN; break; } l2cap_sk = rfcomm_pi(sk)->dlc->session->sock->sk; dlc->defer_setup is set in rfcomm_sock_init() when rfcomm_connect_ind() creates a child socket for an incoming connection on a listening socket that has BT_DEFER_SETUP enabled. It is never cleared afterwards. The session, however, can go away underneath it. rfcomm_recv_disc() forces the dlc state before tearing it down: d->state = BT_CLOSED; __rfcomm_dlc_close(d, err); The RFCOMM_DEFER_SETUP early return in __rfcomm_dlc_close() only covers BT_CONNECT, BT_CONFIG, BT_OPEN and BT_CONNECT2, so with the state already BT_CLOSED that switch does not match and the function falls through to rfcomm_dlc_unlink(), which sets d->session = NULL, while d->defer_setup stays 1. A getsockopt(SOL_RFCOMM, RFCOMM_CONNINFO) on the accepted socket after that point therefore skips the -ENOTCONN path -- sk->sk_state is BT_CLOSED, but dlc->defer_setup is still set -- and dereferences the NULL session. No race is needed: once the DISC has been processed, the dereference is unconditional. Reproduced on a KASAN kernel under QEMU with a BR/EDR peer emulated over /dev/vhci: the peer brings up an ACL link, opens L2CAP on the RFCOMM PSM, starts a session and sends SABM for a channel bound with BT_DEFER_SETUP, and sends DISC for that dlci after the socket has been accepted. getsockopt(SOL_RFCOMM, RFCOMM_CONNINFO) on the accepted socket then hits: Oops: general protection fault, probably for non-canonical address 0xdffffc0000000002: 0000 [#1] SMP KASAN PTI KASAN: null-ptr-deref in range [0x0000000000000010-0x0000000000000017] CPU: 1 UID: 0 PID: 150 Comm: init Tainted: G B 7.3.0-rc3-g5dd1818b15d9 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996) RIP: 0010:rfcomm_sock_getsockopt+0x529/0x780 Call Trace: do_sock_getsockopt+0x3ad/0x7d0 __sys_getsockopt+0x10e/0x1b0 __x64_sys_getsockopt+0xc2/0x160 do_syscall_64+0xda/0x4b0 entry_SYSCALL_64_after_hwframe+0x77/0x7f 0x10 is the offset of sock in struct rfcomm_session; rfcomm_sock_getsockopt_old() is inlined into rfcomm_sock_getsockopt(). Commit 43a556b2fd43 ("Bluetooth: RFCOMM: take rfcomm_mutex for the deferred setup accept") fixed the same "a remote DISC clears the session while deferred setup is still flagged" problem in rfcomm_dlc_accept(); this is the remaining instance of it, in the getsockopt path. Deferred setup only leaves a socket usable here once it has reached BT_CONNECT2, so restrict the exception to that state and check that a session is actually present before following it. Fixes: bb23c0ab8246 ("Bluetooth: Add support for deferring RFCOMM connection setup") Cc: stable@vger.kernel.org Assisted-by: LLM Signed-off-by: Hui Peng Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/rfcomm/sock.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/net/bluetooth/rfcomm/sock.c b/net/bluetooth/rfcomm/sock.c index e2486bc11cbc..fb924d0e34ec 100644 --- a/net/bluetooth/rfcomm/sock.c +++ b/net/bluetooth/rfcomm/sock.c @@ -786,8 +786,10 @@ static int rfcomm_sock_getsockopt_old(struct socket *sock, int optname, break; case RFCOMM_CONNINFO: - if (sk->sk_state != BT_CONNECTED && - !rfcomm_pi(sk)->dlc->defer_setup) { + if ((sk->sk_state != BT_CONNECTED && + !(sk->sk_state == BT_CONNECT2 && + rfcomm_pi(sk)->dlc->defer_setup)) || + !rfcomm_pi(sk)->dlc->session) { err = -ENOTCONN; break; } From 6d91041bb38b97e2feb625123cc0529d7b83a0e1 Mon Sep 17 00:00:00 2001 From: Hui Peng Date: Sat, 19 Sep 2026 11:25:14 +0000 Subject: [PATCH 284/694] Bluetooth: RFCOMM: Reject short EA=0 frames in rfcomm_recv_frame() While rfcomm_recv_frame() verifies that skb->len is at least sizeof(*hdr) + 1 (4 bytes: 3-byte header + 1-byte FCS), an RFCOMM frame with an extended 2-byte length field (!__test_ea(hdr->len)) has a 4-byte header plus a 1-byte FCS (5 bytes minimum, sizeof(*hdr) + 2). When a 4-byte RFCOMM frame with EA == 0 arrives: 1. The initial skb->len < sizeof(*hdr) + 1 check passes (4 < 4 is false). 2. Trimming the FCS byte decrements skb->len to 3. 3. If __check_fcs() succeeds, skb_pull(skb, 4) fails (4 > 3) and returns NULL without advancing skb->data. 4. Because the return value of skb_pull() is ignored, the un-pulled 3-byte struct rfcomm_hdr remains at skb->data and is either queued as application payload via rfcomm_recv_data() or parsed as a multiplexer control command via rfcomm_recv_mcc() on DLCI 0. Fix this by extending the length check in rfcomm_recv_frame() to also require skb->len >= sizeof(*hdr) + 2 when !__test_ea(hdr->len). Fixes: b230e5bf501c ("Bluetooth: RFCOMM: validate skb length in rfcomm_recv_frame") Assisted-by: LLM Signed-off-by: Hui Peng Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/rfcomm/core.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/net/bluetooth/rfcomm/core.c b/net/bluetooth/rfcomm/core.c index f7463f092283..d91e2a6ee26c 100644 --- a/net/bluetooth/rfcomm/core.c +++ b/net/bluetooth/rfcomm/core.c @@ -1817,7 +1817,8 @@ static struct rfcomm_session *rfcomm_recv_frame(struct rfcomm_session *s, return s; } - if (skb->len < sizeof(*hdr) + 1) { + if (skb->len < sizeof(*hdr) + 1 || + (!__test_ea(hdr->len) && skb->len < sizeof(*hdr) + 2)) { kfree_skb(skb); return s; } From 7325a44474f20c1393a1c114c6811abce6ddfaf4 Mon Sep 17 00:00:00 2001 From: Zhao Li Date: Wed, 16 Sep 2026 19:36:48 +0800 Subject: [PATCH 285/694] wifi: mac80211: release mesh path quota on failed additions Repeated failed or duplicate mesh path additions can exhaust MESH_MAX_MPATHS while the path table holds fewer paths, which stops new paths to reachable destinations from being created. mesh_path_add() reserves a slot before allocating and inserting a path. If allocation fails, the function returns without releasing the slot. If the rhashtable insertion reports an error or an existing destination, the function discards its candidate but retains the reservation. No new path is installed in any of these cases. Release the reservation whenever the candidate is not installed. Fixes: ae76eef027f7 ("mac80211: return new mpath from mesh_path_add()") Fixes: 60854fd94573 ("mac80211: mesh: convert path table to rhashtable") Cc: stable@vger.kernel.org Assisted-by: LLM sparse kasan Signed-off-by: Zhao Li Link: https://patch.msgid.link/20260916113649.28315-2-enderaoelyther@gmail.com Signed-off-by: Johannes Berg --- net/mac80211/mesh_pathtbl.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/net/mac80211/mesh_pathtbl.c b/net/mac80211/mesh_pathtbl.c index 03171cf00855..770fc16b439b 100644 --- a/net/mac80211/mesh_pathtbl.c +++ b/net/mac80211/mesh_pathtbl.c @@ -694,8 +694,10 @@ struct mesh_path *mesh_path_add(struct ieee80211_sub_if_data *sdata, return ERR_PTR(-ENOSPC); new_mpath = mesh_path_new(sdata, dst, GFP_ATOMIC); - if (!new_mpath) + if (!new_mpath) { + atomic_dec(&sdata->u.mesh.mpaths); return ERR_PTR(-ENOMEM); + } tbl = &sdata->u.mesh.mesh_paths; spin_lock_bh(&tbl->walk_lock); @@ -708,6 +710,7 @@ struct mesh_path *mesh_path_add(struct ieee80211_sub_if_data *sdata, if (mpath) { kfree(new_mpath); + atomic_dec(&sdata->u.mesh.mpaths); if (IS_ERR(mpath)) return mpath; From a5f35d4b112af1415ff06bd501c8cc26cc08aa8b Mon Sep 17 00:00:00 2001 From: Zhao Li Date: Wed, 16 Sep 2026 19:36:49 +0800 Subject: [PATCH 286/694] wifi: mac80211: account proxy paths against the mesh path limit Proxy-path churn can drive the interface path counter negative and let later mesh path additions exceed MESH_MAX_MPATHS. mesh_path_free_rcu() decrements the interface mpaths counter for entries from both the mesh and proxy path tables, but mpp_path_add() never reserves a slot in that counter. Removing or expiring a proxy path therefore returns a slot that was never charged. mesh_path_add() uses the same counter to enforce MESH_MAX_MPATHS, so once it falls below the number of installed paths the limit no longer holds. Reserve the shared quota before allocating a proxy path and release it on every unsuccessful addition. Mesh and proxy paths now share one 1024-entry budget, so mpp_path_add() can return -ENOSPC at that limit. Fixes: ece1a2e7e860 ("mac80211: Remove mesh paths when an interface is removed") Cc: stable@vger.kernel.org Assisted-by: LLM sparse kasan Signed-off-by: Zhao Li Link: https://patch.msgid.link/20260916113649.28315-3-enderaoelyther@gmail.com Signed-off-by: Johannes Berg --- net/mac80211/mesh_pathtbl.c | 13 ++++++++++--- 1 file changed, 10 insertions(+), 3 deletions(-) diff --git a/net/mac80211/mesh_pathtbl.c b/net/mac80211/mesh_pathtbl.c index 770fc16b439b..1c55b14c2ac0 100644 --- a/net/mac80211/mesh_pathtbl.c +++ b/net/mac80211/mesh_pathtbl.c @@ -736,10 +736,15 @@ int mpp_path_add(struct ieee80211_sub_if_data *sdata, if (is_multicast_ether_addr(dst)) return -EOPNOTSUPP; + if (!atomic_add_unless(&sdata->u.mesh.mpaths, 1, MESH_MAX_MPATHS)) + return -ENOSPC; + new_mpath = mesh_path_new(sdata, dst, GFP_ATOMIC); - if (!new_mpath) + if (!new_mpath) { + atomic_dec(&sdata->u.mesh.mpaths); return -ENOMEM; + } memcpy(new_mpath->mpp, mpp, ETH_ALEN); tbl = &sdata->u.mesh.mpp_paths; @@ -752,10 +757,12 @@ int mpp_path_add(struct ieee80211_sub_if_data *sdata, hlist_add_head_rcu(&new_mpath->walk_list, &tbl->walk_head); spin_unlock_bh(&tbl->walk_lock); - if (ret) + if (ret) { kfree(new_mpath); - else + atomic_dec(&sdata->u.mesh.mpaths); + } else { mesh_fast_tx_flush_addr(sdata, dst); + } sdata->u.mesh.mpp_paths_generation++; return ret; From 67a9e80c3ca20c9044cd6ee72cd6278dd19b5023 Mon Sep 17 00:00:00 2001 From: Zhao Li Date: Wed, 16 Sep 2026 19:37:35 +0800 Subject: [PATCH 287/694] wifi: mac80211: drain PS delivery work during station teardown Removing an AP or mesh station can notify the driver and wake TXQs after the station has already been removed from the driver, and can update num_sta_ps and the power-save queues after station cleanup has purged them. sta_deliver_ps_frames() tests sta->dead only before it starts delivery. A worker that passes that test can be delayed while teardown sets sta->dead and removes the station from the driver; when it resumes it still runs the full delivery. The existing cancel_work_sync() sits near the end of __cleanup_single_sta(), after the final driver state transition and after PS accounting and queue purging, so it drains the worker only once the state it has to protect has already changed. Before commit d34ba2168a3c ("mac80211: don't delay station destruction") cleanup was deferred through the same ordered workqueue as the delivery work, which kept the two ordered. Cancel the work after setting sta->dead and before moving the driver state, and move the existing cancellation to the start of __cleanup_single_sta(). The first drain closes the driver-lifetime window; driver callbacks during the remaining state transitions can still queue drv_deliver_wk through ieee80211_sta_block_awake(), so retain the second drain at the start of __cleanup_single_sta(). That drain also covers insertion failures, which reach cleanup without setting sta->dead. Fixes: d34ba2168a3c ("mac80211: don't delay station destruction") Cc: stable@vger.kernel.org Assisted-by: LLM sparse smatch coccinelle kasan Signed-off-by: Zhao Li Link: https://patch.msgid.link/20260916113735.31029-1-enderaoelyther@gmail.com Signed-off-by: Johannes Berg --- net/mac80211/sta_info.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/net/mac80211/sta_info.c b/net/mac80211/sta_info.c index fdf00cbf49d8..e81cd155197b 100644 --- a/net/mac80211/sta_info.c +++ b/net/mac80211/sta_info.c @@ -137,6 +137,8 @@ static void __cleanup_single_sta(struct sta_info *sta) struct ieee80211_local *local = sdata->local; struct ps_data *ps; + cancel_work_sync(&sta->drv_deliver_wk); + if (test_sta_flag(sta, WLAN_STA_PS_STA) || test_sta_flag(sta, WLAN_STA_PS_DRIVER) || test_sta_flag(sta, WLAN_STA_PS_DELIVER)) { @@ -166,8 +168,6 @@ static void __cleanup_single_sta(struct sta_info *sta) if (ieee80211_vif_is_mesh(&sdata->vif)) mesh_sta_cleanup(sta); - cancel_work_sync(&sta->drv_deliver_wk); - /* * Destroy aggregation state here. It would be nice to wait for the * driver to finish aggregation stop and then clean up, but for now @@ -1582,6 +1582,8 @@ static void __sta_info_destroy_part2(struct sta_info *sta, bool recalc) sta->dead = true; + cancel_work_sync(&sta->drv_deliver_wk); + local->num_sta--; local->sta_generation++; From a8b8881d0c1aff544fae190469d86db7b07f43a9 Mon Sep 17 00:00:00 2001 From: Wentao Liang Date: Thu, 17 Sep 2026 12:23:25 +0000 Subject: [PATCH 288/694] wifi: wlcore: Fix runtime PM leak in wlcore_remove() pm_runtime_get_sync() keeps the runtime PM usage counter elevated even when it returns an error, so the reference taken at the beginning of wlcore_remove() has to be dropped on every exit path. When the device was not initialized the function returned early without doing so, leaking the reference. Release it with pm_runtime_put_noidle() as done on the other early error paths of this driver. Fixes: fa2648a34e73f ("wlcore: Add support for runtime PM") Cc: stable@vger.kernel.org Signed-off-by: Wentao Liang Link: https://patch.msgid.link/20260917122325.2150910-1-vulab@iscas.ac.cn Signed-off-by: Johannes Berg --- drivers/net/wireless/ti/wlcore/main.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/net/wireless/ti/wlcore/main.c b/drivers/net/wireless/ti/wlcore/main.c index edf6ca23c6c3..26aabf4ea127 100644 --- a/drivers/net/wireless/ti/wlcore/main.c +++ b/drivers/net/wireless/ti/wlcore/main.c @@ -6809,8 +6809,10 @@ void wlcore_remove(struct platform_device *pdev) if (pdev_data->family && pdev_data->family->nvs_name) wait_for_completion(&wl->nvs_loading_complete); - if (!wl->initialized) + if (!wl->initialized) { + pm_runtime_put_noidle(wl->dev); return; + } if (wl->wakeirq >= 0) { dev_pm_clear_wake_irq(wl->dev); From 5e4f3509793b4db2c0835340224743db680e861f Mon Sep 17 00:00:00 2001 From: Yuchao Zhang Date: Fri, 18 Sep 2026 11:40:04 +0800 Subject: [PATCH 289/694] wifi: mac80211: drop oversized fragments to avoid extra_len overflow In ieee80211_rx_h_defragment(), fragment payloads are accumulated into entry->extra_len as subsequent fragments arrive: entry->extra_len += rx->skb->len; When the final fragment arrives, the head fragment is dequeued, and pskb_expand_head() is called with entry->extra_len to ensure sufficient tailroom for all queued fragments before copying: if (skb_tailroom(rx->skb) < entry->extra_len) { if (unlikely(pskb_expand_head(rx->skb, 0, entry->extra_len, GFP_ATOMIC))) { ... } } while ((skb = __skb_dequeue(&entry->skb_list))) { skb_put_data(rx->skb, skb->data, skb->len); dev_kfree_skb(skb); } In commit 69f132236827 ("mac80211: shrink struct ieee80211_fragment_entry"), entry->extra_len was narrowed from unsigned int to u16 in order to reduce structure memory footprint. However, an IEEE 802.11 frame sequence can contain up to 16 fragments (frag numbers 0..15). If a sequence of large fragments arrives (e.g. from a malicious peer or rogue AP), the sum of fragment lengths can exceed 65535 bytes (for example, 15 fragments of 4400 bytes total 66000 bytes). Because extra_len is a u16, this addition overflows and wraps around modulo 65536 (e.g. 66000 wraps to 464). Consequently, pskb_expand_head() allocates only the truncated amount of tailroom (or is skipped entirely if the head fragment already has >= 464 bytes of tailroom). When skb_put_data() subsequently iterates through the queued skb list, it appends the full payload into the undersized buffer, causing skb_put() to trigger skb_over_panic() and crash the kernel in softirq context. A legitimate MSDU in IEEE 802.11 is at most 2304 bytes (or up to 7935/11454 bytes for A-MSDU, which is not fragmented), well below U16_MAX. Fix this without increasing the size of struct ieee80211_fragment_entry by checking whether adding the incoming fragment length would exceed U16_MAX. If so, purge the queued fragments and drop the frame. Fixes: 69f132236827 ("mac80211: shrink struct ieee80211_fragment_entry") Signed-off-by: Yuchao Zhang Link: https://patch.msgid.link/20260918034004.85078-2-ndaugoing@gmail.com [assign a separate drop reason] Signed-off-by: Johannes Berg --- net/mac80211/drop.h | 1 + net/mac80211/rx.c | 6 ++++++ 2 files changed, 7 insertions(+) diff --git a/net/mac80211/drop.h b/net/mac80211/drop.h index f06a8aa905c5..f433ee17cce5 100644 --- a/net/mac80211/drop.h +++ b/net/mac80211/drop.h @@ -108,6 +108,7 @@ typedef unsigned int __bitwise ieee80211_rx_result; R(RX_DROP_U_UNHANDLED_PREQ) \ R(RX_DROP_U_UNHANDLED_MGMT_STYPE) \ R(RX_DROP_U_NO_LINK) \ + R(RX_DROP_U_DEFRAG_OVERFLOW) \ /* this line for the trailing \ - add before this */ /* having two enums allows for checking ieee80211_rx_result use with sparse */ diff --git a/net/mac80211/rx.c b/net/mac80211/rx.c index 5e26be8e27d8..7ead509304eb 100644 --- a/net/mac80211/rx.c +++ b/net/mac80211/rx.c @@ -2499,6 +2499,12 @@ ieee80211_rx_h_defragment(struct ieee80211_rx_data *rx) } skb_pull(rx->skb, ieee80211_hdrlen(fc)); + if (unlikely((u32)entry->extra_len + rx->skb->len > U16_MAX)) { + I802_DEBUG_INC(rx->local->rx_handlers_drop_defrag); + __skb_queue_purge(&entry->skb_list); + return RX_DROP_U_DEFRAG_OVERFLOW; + } + __skb_queue_tail(&entry->skb_list, rx->skb); entry->last_frag = frag; entry->extra_len += rx->skb->len; From ee3aadc43ea69d98f666431e787137ac0badbb85 Mon Sep 17 00:00:00 2001 From: Zihan Xi Date: Fri, 18 Sep 2026 05:26:51 +0000 Subject: [PATCH 290/694] wifi: mac80211: fix mesh fast xmit path deletion UAF mesh_fast_tx_cache() stores raw mesh_path pointers. Path deletion flushes the cache and then frees the path with kfree_rcu(), but a lookup that already holds the path can insert a new cache entry after the flush. The cache then points at freed memory. Set MESH_PATH_DELETED before flushing, and skip inserting a cache entry if the path or MPP path is already deleted. Check this under the cache walk lock so it is ordered with the flush. Use WRITE_ONCE() for the deletion flag updates because the cache check reads flags without taking state_lock. Fixes: d5edb9ae8d56 ("wifi: mac80211: mesh fast xmit support") Cc: stable@vger.kernel.org Reported-by: Vega Assisted-by: LLM Co-developed-by: Luxing Yin Signed-off-by: Luxing Yin Signed-off-by: Zihan Xi Link: https://patch.msgid.link/1daa7a98199fe6965c2da7c42717073a6fd39a0b.1789615568.git.zihanx@nebusec.ai Signed-off-by: Johannes Berg --- net/mac80211/mesh_pathtbl.c | 12 +++++++++++- 1 file changed, 11 insertions(+), 1 deletion(-) diff --git a/net/mac80211/mesh_pathtbl.c b/net/mac80211/mesh_pathtbl.c index 1c55b14c2ac0..07a21dfa8674 100644 --- a/net/mac80211/mesh_pathtbl.c +++ b/net/mac80211/mesh_pathtbl.c @@ -577,6 +577,12 @@ void mesh_fast_tx_cache(struct ieee80211_sub_if_data *sdata, goto unlock_sta; spin_lock(&cache->walk_lock); + if ((READ_ONCE(mpath->flags) & MESH_PATH_DELETED) || + (mppath && (READ_ONCE(mppath->flags) & MESH_PATH_DELETED))) { + kfree(entry); + goto unlock_cache; + } + prev = rhashtable_lookup_get_insert_fast(&cache->rht, &entry->rhash, fast_tx_rht_params); @@ -808,7 +814,8 @@ static void mesh_path_free_rcu(struct mesh_table *tbl, struct ieee80211_sub_if_data *sdata = mpath->sdata; spin_lock_bh(&mpath->state_lock); - mpath->flags |= MESH_PATH_RESOLVING | MESH_PATH_DELETED; + WRITE_ONCE(mpath->flags, + mpath->flags | MESH_PATH_RESOLVING | MESH_PATH_DELETED); mesh_gate_del(tbl, mpath); spin_unlock_bh(&mpath->state_lock); timer_shutdown_sync(&mpath->timer); @@ -822,6 +829,9 @@ static void __mesh_path_del(struct mesh_table *tbl, struct mesh_path *mpath) { hlist_del_rcu(&mpath->walk_list); rhashtable_remove_fast(&tbl->rhead, &mpath->rhash, mesh_rht_params); + spin_lock_bh(&mpath->state_lock); + WRITE_ONCE(mpath->flags, mpath->flags | MESH_PATH_DELETED); + spin_unlock_bh(&mpath->state_lock); if (tbl == &mpath->sdata->u.mesh.mpp_paths) mesh_fast_tx_flush_addr(mpath->sdata, mpath->dst); else From db7b86bd97b380ece8fa39cfdd7502a9fd35e56f Mon Sep 17 00:00:00 2001 From: Heyang Tan Date: Fri, 18 Sep 2026 16:33:49 +0800 Subject: [PATCH 291/694] wifi: p54: validate firmware record lengths The p54 firmware parser accesses component ID and component version records without checking that their lengths cover the fields being read. A malformed record can therefore make the parser read past the record. Validate the component ID and component version record lengths before accessing their data. The descriptor record is also checked against the shortest descriptor length found in the supported firmware images. Reject descriptor values that underflow the RX offset or produce an empty RX range before updating the driver state. Assisted-by: LLM Codex Signed-off-by: Heyang Tan Tested-by: Christian Lamparter Link: https://patch.msgid.link/20260918083349.891-1-thy15333007817@163.com Signed-off-by: Johannes Berg --- drivers/net/wireless/intersil/p54/fwio.c | 42 +++++++++++++++++++++--- 1 file changed, 37 insertions(+), 5 deletions(-) diff --git a/drivers/net/wireless/intersil/p54/fwio.c b/drivers/net/wireless/intersil/p54/fwio.c index a3d9053f043c..f028a7bfa734 100644 --- a/drivers/net/wireless/intersil/p54/fwio.c +++ b/drivers/net/wireless/intersil/p54/fwio.c @@ -52,6 +52,13 @@ int p54_parse_firmware(struct ieee80211_hw *dev, const struct firmware *fw) u32 code = le32_to_cpu(bootrec->code); switch (code) { case BR_CODE_COMPONENT_ID: + if (len < sizeof(struct bootrec_comp_id) / + sizeof(*bootrec->data)) { + wiphy_err(priv->hw->wiphy, + "firmware component ID is too short\n"); + return -EINVAL; + } + priv->fw_interface = be32_to_cpup((__be32 *) bootrec->data); switch (priv->fw_interface) { @@ -72,16 +79,41 @@ int p54_parse_firmware(struct ieee80211_hw *dev, const struct firmware *fw) } break; case BR_CODE_COMPONENT_VERSION: + if (len < DIV_ROUND_UP(sizeof(struct bootrec_comp_ver), + sizeof(*bootrec->data))) { + wiphy_err(priv->hw->wiphy, + "firmware component version is too short\n"); + return -EINVAL; + } + /* 24 bytes should be enough for all firmwares */ - if (strnlen((unsigned char *) bootrec->data, 24) < 24) - fw_version = (unsigned char *) bootrec->data; + if (strnlen((unsigned char *)bootrec->data, + sizeof(struct bootrec_comp_ver)) < + sizeof(struct bootrec_comp_ver)) + fw_version = (unsigned char *)bootrec->data; break; case BR_CODE_DESCR: { struct bootrec_desc *desc = (struct bootrec_desc *)bootrec->data; - priv->rx_start = le32_to_cpu(desc->rx_start); - /* FIXME add sanity checking */ - priv->rx_end = le32_to_cpu(desc->rx_end) - 0x3500; + u32 rx_start, rx_end; + + /* 0xa is the shortest descriptor in supported firmware. */ + if (len < 0xa) { + wiphy_err(priv->hw->wiphy, + "firmware descriptor is too short\n"); + return -EINVAL; + } + + rx_start = le32_to_cpu(desc->rx_start); + rx_end = le32_to_cpu(desc->rx_end); + if (rx_end < 0x3500 || rx_end - 0x3500 <= rx_start) { + wiphy_err(priv->hw->wiphy, + "firmware descriptor has invalid RX range\n"); + return -EINVAL; + } + + priv->rx_start = rx_start; + priv->rx_end = rx_end - 0x3500; priv->headroom = desc->headroom; priv->tailroom = desc->tailroom; priv->privacy_caps = desc->privacy_caps; From 15fb9e3cea55fbc378261ce35a32b0f8828f0000 Mon Sep 17 00:00:00 2001 From: Yuqi Xu Date: Sat, 19 Sep 2026 16:46:20 +0800 Subject: [PATCH 292/694] wifi: mac80211: minstrel_ht: validate fixed rate index The fixed_rate_idx debugfs file is a plain u32 attribute, so any value can be written to it. minstrel_ht_update_stats() then copies the value into mi->max_tp_rate[] and mi->max_prob_rate, and minstrel_ht_set_rate() uses MI_RATE_GROUP()/MI_RATE_IDX() to index minstrel_mcs_groups[] (42 entries) and mi->groups[].rates[] (10 entries) with it. Writing e.g. 65535 selects group 4095 and rate index 15, far outside both arrays. The out-of-bounds entries are dereferenced and updated as struct minstrel_rate_stats (retry counts, etc.), so the invalid index corrupts adjacent kernel memory. With CONFIG_UBSAN_BOUNDS the access is reported as an array-index-out-of-bounds in minstrel_ht_set_rate(). Only a valid group/rate pair, or U32_MAX to disable fixed rate processing, can be used safely, so reject any other value when the debugfs file is written. The file is only reachable through a root-only debugfs mount, so this is not a privilege boundary. The out-of-bounds access is still a bug that must not be reachable through a writable debugfs attribute. Fixes: 24f7580e852b ("minstrel_ht: fixed rate mode through debugfs") Cc: stable@vger.kernel.org Reported-by: Vega Assisted-by: LLM Signed-off-by: Yuqi Xu Reviewed-by: Ren Wei Link: https://patch.msgid.link/288e6f6d30dc82ff40da502755e6666982f6b735.1789798000.git.xuyuqiabc@gmail.com Signed-off-by: Johannes Berg --- net/mac80211/rc80211_minstrel_ht.c | 30 ++++++++++++++++++++++++++++-- 1 file changed, 28 insertions(+), 2 deletions(-) diff --git a/net/mac80211/rc80211_minstrel_ht.c b/net/mac80211/rc80211_minstrel_ht.c index b73ef3adfcc5..c26221287515 100644 --- a/net/mac80211/rc80211_minstrel_ht.c +++ b/net/mac80211/rc80211_minstrel_ht.c @@ -7,6 +7,7 @@ #include #include #include +#include #include #include #include @@ -1947,14 +1948,39 @@ minstrel_ht_alloc(struct ieee80211_hw *hw) } #ifdef CONFIG_MAC80211_DEBUGFS +static int minstrel_ht_fixed_rate_idx_get(void *data, u64 *val) +{ + *val = *(u32 *)data; + return 0; +} + +static int minstrel_ht_fixed_rate_idx_set(void *data, u64 val) +{ + u32 idx = val; + + /* U32_MAX is the default and keeps fixed rate processing disabled */ + if (val != U32_MAX && + (val > U16_MAX || + MI_RATE_GROUP(idx) >= ARRAY_SIZE(minstrel_mcs_groups) || + MI_RATE_IDX(idx) >= MCS_GROUP_RATES)) + return -EINVAL; + + *(u32 *)data = idx; + return 0; +} + +DEFINE_DEBUGFS_ATTRIBUTE(minstrel_ht_fixed_rate_idx_fops, + minstrel_ht_fixed_rate_idx_get, + minstrel_ht_fixed_rate_idx_set, "%llu\n"); + static void minstrel_ht_add_debugfs(struct ieee80211_hw *hw, void *priv, struct dentry *debugfsdir) { struct minstrel_priv *mp = priv; mp->fixed_rate_idx = (u32) -1; - debugfs_create_u32("fixed_rate_idx", S_IRUGO | S_IWUGO, debugfsdir, - &mp->fixed_rate_idx); + debugfs_create_file("fixed_rate_idx", S_IRUGO | S_IWUGO, debugfsdir, + &mp->fixed_rate_idx, &minstrel_ht_fixed_rate_idx_fops); } #endif From 5bfd4b0b40b79781d900cb2f7da0685070f53c67 Mon Sep 17 00:00:00 2001 From: Yuqi Xu Date: Sat, 19 Sep 2026 16:48:12 +0800 Subject: [PATCH 293/694] wifi: mac80211: handle empty FILS association request payload fils_encrypt_assoc_req() accepts a (Re)Association Request frame whose last element is the FILS Session element. crypt_len is then zero and aes_siv_encrypt() is called with an empty plaintext. kmemdup() returns ZERO_SIZE_PTR for a zero-length allocation, so the unconditional sg_init_one() puts an invalid page into the scatterlist. With CONFIG_DEBUG_SG this triggers a kernel BUG in sg_set_buf(); without the debug check the bogus page is handed to the crypto backend. AES-SIV is defined for empty plaintext and the S2V result is then the complete output, so handle this case before the CTR step, which has nothing to encrypt. Fixes: 39404feee691 ("mac80211: FILS AEAD protection for station mode association frames") Cc: stable@vger.kernel.org Reported-by: Vega Assisted-by: LLM Signed-off-by: Yuqi Xu Reviewed-by: Ren Wei Link: https://patch.msgid.link/325c048982d637a3af80d74ac5b36e9369b93f1e.1789751126.git.xuyuqiabc@gmail.com Signed-off-by: Johannes Berg --- net/mac80211/fils_aead.c | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/net/mac80211/fils_aead.c b/net/mac80211/fils_aead.c index 293590976489..9c4f00d61981 100644 --- a/net/mac80211/fils_aead.c +++ b/net/mac80211/fils_aead.c @@ -89,6 +89,14 @@ static int aes_siv_encrypt(const u8 *key, size_t key_len, if (res) return res; + /* With an empty plaintext there is no ciphertext to produce and + * the S2V result is the complete output. + */ + if (!plain_len) { + memcpy(out, v, AES_BLOCK_SIZE); + return 0; + } + /* Use a temporary buffer of the plaintext to handle need for * overwriting this during AES-CTR. */ From 9eda41e32263806a66835e275efdcc0c5f46bbc4 Mon Sep 17 00:00:00 2001 From: Stanislaw Gruszka Date: Sat, 19 Sep 2026 11:24:54 +0200 Subject: [PATCH 294/694] wifi: cfg80211: fix RTS threshold setting for single-radio PHY Due to an incorrect old_radio_rts_threshold check, we do not set rdev->wiphy.rts_threshold for a single-radio PHY, and the threshold value is not propagated to lower layers. Cc: stable@vger.kernel.org Fixes: 264637941cf4 ("wifi: cfg80211: Add Support to Set RTS Threshold for each Radio") Signed-off-by: Stanislaw Gruszka Link: https://patch.msgid.link/20260919092454.50266-1-stf_xl@wp.pl Signed-off-by: Johannes Berg --- net/wireless/nl80211.c | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c index 9fafd7673d85..ddce61ec016a 100644 --- a/net/wireless/nl80211.c +++ b/net/wireless/nl80211.c @@ -4563,12 +4563,13 @@ static int nl80211_set_wiphy(struct sk_buff *skb, struct genl_info *info) rdev->wiphy.retry_long = retry_long; if (changed & WIPHY_PARAM_FRAG_THRESHOLD) rdev->wiphy.frag_threshold = frag_threshold; - if ((changed & WIPHY_PARAM_RTS_THRESHOLD) && - old_radio_rts_threshold) { + if (changed & WIPHY_PARAM_RTS_THRESHOLD) { rdev->wiphy.rts_threshold = rts_threshold; - for (i = 0 ; i < rdev->wiphy.n_radio; i++) - rdev->wiphy.radio_cfg[i].rts_threshold = - rdev->wiphy.rts_threshold; + if (old_radio_rts_threshold) { + for (i = 0; i < rdev->wiphy.n_radio; i++) + rdev->wiphy.radio_cfg[i].rts_threshold = + rdev->wiphy.rts_threshold; + } } if (changed & WIPHY_PARAM_COVERAGE_CLASS) rdev->wiphy.coverage_class = coverage_class; From 7e0ad05ebc72fffba2bdf2400400902d94ed7272 Mon Sep 17 00:00:00 2001 From: Nicolas Escande Date: Wed, 12 Aug 2026 16:58:05 +0200 Subject: [PATCH 295/694] wifi: ath11k: reset ar->num_stations on hardware start When (multiple) hw restart occurs, we end up in a situation where we cannot accept new stations / mesh peers. This seems to be the 'num_stations' in 'struct ath11k' that did not get reset properly in this case. In ath11k_mac_op_start(), it was indeed the only accounting var that did not get reset so let's clear it too. This avoids problems like those: [1248828.088023] WARNING: CPU: 0 PID: 12281 at net/mac80211/util.c:1682 ieee80211_reconfig_stations+0x98/0xc0 [1248828.088115] Call trace: [1248828.088116] ieee80211_reconfig_stations+0x98/0xc0 (P) [1248828.088119] ieee80211_reconfig+0x774/0x1438 [1248828.088123] ieee80211_restart_work+0x10c/0x178 [1248828.088127] process_one_work+0x178/0x3c0 [1248828.088130] worker_thread+0x280/0x440 [1248828.088133] kthread+0xe0/0xf8 [1248828.088137] ret_from_fork+0x10/0x20 [1248828.088141] ---[ end trace 0000000000000000 ]--- [1248836.686604] ath11k_pci 0000:04:00.0: refusing to associate station: too many connected already (128) [1248836.686609] ath11k_pci 0000:04:00.0: Failed to add station: xx:xx:xx:xx:xx:xx for VDEV: 1 Tested-on: QCN9074 hw1.0 PCI WLAN.HK.2.9.0.1-01977-QCAHKSWPL_SILICONZ-1 Fixes: d5c65159f289 ("ath11k: driver for Qualcomm IEEE 802.11ax devices") Signed-off-by: Nicolas Escande Reviewed-by: Rameshkumar Sundaram Reviewed-by: Baochen Qiang Link: https://patch.msgid.link/20260812-num-station-v3-1-871d20e043f6@gmail.com Signed-off-by: Jeff Johnson --- drivers/net/wireless/ath/ath11k/mac.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/net/wireless/ath/ath11k/mac.c b/drivers/net/wireless/ath/ath11k/mac.c index ae91b57c8422..c1fa42edd363 100644 --- a/drivers/net/wireless/ath/ath11k/mac.c +++ b/drivers/net/wireless/ath/ath11k/mac.c @@ -6716,6 +6716,7 @@ static int ath11k_mac_op_start(struct ieee80211_hw *hw) ar->num_started_vdevs = 0; ar->num_created_vdevs = 0; ar->num_peers = 0; + ar->num_stations = 0; ar->allocated_vdev_map = 0; /* Configure monitor status ring with default rx_filter to get rx status From 1af8dc9ca726cf53767f9c0661171396f95c146a Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Toke=20H=C3=B8iland-J=C3=B8rgensen?= Date: Tue, 8 Sep 2026 19:05:18 +0200 Subject: [PATCH 296/694] wifi: ath9k: Clean up device initialisation guards MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The ath9k driver has a couple of guards against the RX path running until device initialisation has completed. However, they are problematic for several reasons: - They use smp_wmb() paired with a data_race() annotation, instead of the idiomatic smp_store_release()/smp_load_acquire() pair. - The guard in ath9k_wmi_event_tasklet() is weirdly late in the function. - There are two of them when there's really no reason the rx init gets to run before the whole initialisation is complete. Consolidate the guards on the single ath9k_htc_priv->initialized variable and use smp_store_release()/smp_load_acquire() to access it. Move the store into ath9k_init_device() to make sure it's set in all paths in the driver. Fixes: 24355fcb0d4c ("wifi: ath9k: delay all of ath9k_wmi_event_tasklet() until init is complete") Fixes: b0ec7e55fce6 ("ath9k_htc: fix NULL pointer dereference at ath9k_htc_rxep()") Signed-off-by: Toke Høiland-Jørgensen Link: https://patch.msgid.link/20260908170520.112946-1-toke@redhat.com Signed-off-by: Jeff Johnson --- drivers/net/wireless/ath/ath9k/htc.h | 1 - drivers/net/wireless/ath/ath9k/htc_drv_init.c | 7 +++---- drivers/net/wireless/ath/ath9k/htc_drv_txrx.c | 8 ++------ drivers/net/wireless/ath/ath9k/wmi.c | 12 ++++++------ 4 files changed, 11 insertions(+), 17 deletions(-) diff --git a/drivers/net/wireless/ath/ath9k/htc.h b/drivers/net/wireless/ath/ath9k/htc.h index 6c33e898b300..4d2b75d0212e 100644 --- a/drivers/net/wireless/ath/ath9k/htc.h +++ b/drivers/net/wireless/ath/ath9k/htc.h @@ -281,7 +281,6 @@ struct ath9k_htc_rxbuf { struct ath9k_htc_rx { struct list_head rxbuf; spinlock_t rxbuflock; - bool initialized; }; #define ATH9K_HTC_TX_CLEANUP_INTERVAL 50 /* ms */ diff --git a/drivers/net/wireless/ath/ath9k/htc_drv_init.c b/drivers/net/wireless/ath/ath9k/htc_drv_init.c index 6de78ae85726..911eaa4f776a 100644 --- a/drivers/net/wireless/ath/ath9k/htc_drv_init.c +++ b/drivers/net/wireless/ath/ath9k/htc_drv_init.c @@ -910,6 +910,9 @@ static int ath9k_init_device(struct ath9k_htc_priv *priv, ath9k_init_leds(priv); ath9k_start_rfkill_poll(priv); + /* signal completion to ath9k_htc_rxep() and ath9k_wmi_event_tasklet() */ + smp_store_release(&priv->initialized, true); + return 0; err_world: @@ -966,10 +969,6 @@ int ath9k_htc_probe_device(struct htc_target *htc_handle, struct device *dev, htc_handle->drv_priv = priv; - /* Allow ath9k_wmi_event_tasklet() to operate. */ - smp_wmb(); - priv->initialized = true; - return 0; err_init: diff --git a/drivers/net/wireless/ath/ath9k/htc_drv_txrx.c b/drivers/net/wireless/ath/ath9k/htc_drv_txrx.c index bed7ea2425a0..95f48c7dca15 100644 --- a/drivers/net/wireless/ath/ath9k/htc_drv_txrx.c +++ b/drivers/net/wireless/ath/ath9k/htc_drv_txrx.c @@ -1141,8 +1141,8 @@ void ath9k_htc_rxep(void *drv_priv, struct sk_buff *skb, struct ath9k_htc_rxbuf *rxbuf = NULL, *tmp_buf = NULL; unsigned long flags; - /* Check if ath9k_rx_init() completed. */ - if (!data_race(priv->rx.initialized)) + /* Check if ath9k_init_device() completed. */ + if (!smp_load_acquire(&priv->initialized)) goto err; spin_lock_irqsave(&priv->rx.rxbuflock, flags); @@ -1200,10 +1200,6 @@ int ath9k_rx_init(struct ath9k_htc_priv *priv) list_add_tail(&rxbuf->list, &priv->rx.rxbuf); } - /* Allow ath9k_htc_rxep() to operate. */ - smp_wmb(); - priv->rx.initialized = true; - return 0; err: diff --git a/drivers/net/wireless/ath/ath9k/wmi.c b/drivers/net/wireless/ath/ath9k/wmi.c index 284e8c13b043..44e91815a159 100644 --- a/drivers/net/wireless/ath/ath9k/wmi.c +++ b/drivers/net/wireless/ath/ath9k/wmi.c @@ -146,6 +146,12 @@ void ath9k_wmi_event_tasklet(struct tasklet_struct *t) unsigned long flags; u16 cmd_id; + /* Check if ath9k_init_device() completed. */ + if (!smp_load_acquire(&priv->initialized)) { + tasklet_schedule(t); + return; + } + do { spin_lock_irqsave(&wmi->wmi_lock, flags); skb = __skb_dequeue(&wmi->wmi_event_queue); @@ -155,12 +161,6 @@ void ath9k_wmi_event_tasklet(struct tasklet_struct *t) } spin_unlock_irqrestore(&wmi->wmi_lock, flags); - /* Check if ath9k_htc_probe_device() completed. */ - if (!data_race(priv->initialized)) { - kfree_skb(skb); - continue; - } - hdr = (struct wmi_cmd_hdr *) skb->data; cmd_id = be16_to_cpu(hdr->command_id); wmi_event = skb_pull(skb, sizeof(struct wmi_cmd_hdr)); From 93ce8ea70451e800601bc9444aa33043de15d67b Mon Sep 17 00:00:00 2001 From: Pengpeng Hou Date: Fri, 14 Aug 2026 15:58:37 +0800 Subject: [PATCH 297/694] wifi: ath9k: reject short WMI command responses ath9k_wmi_rsp_callback() removes the validated WMI header and copies the number of bytes expected by the waiting command into its response buffer. A device response shorter than that expectation can therefore be read beyond the skb payload. Record -EMSGSIZE for a short response, complete the waiter, and return the stored status from the command path. This also prevents callers from consuming stale response bytes as a successful reply. Fixes: fb9987d0f748 ("ath9k_htc: Support for AR9271 chipset.") Assisted-by: LLM Signed-off-by: Pengpeng Hou Link: https://patch.msgid.link/20260814075837.20005-1-pengpeng@iscas.ac.cn Signed-off-by: Jeff Johnson --- drivers/net/wireless/ath/ath9k/wmi.c | 16 ++++++++++++++-- drivers/net/wireless/ath/ath9k/wmi.h | 1 + 2 files changed, 15 insertions(+), 2 deletions(-) diff --git a/drivers/net/wireless/ath/ath9k/wmi.c b/drivers/net/wireless/ath/ath9k/wmi.c index 44e91815a159..4a08babb6d9d 100644 --- a/drivers/net/wireless/ath/ath9k/wmi.c +++ b/drivers/net/wireless/ath/ath9k/wmi.c @@ -206,9 +206,16 @@ static void ath9k_wmi_rsp_callback(struct wmi *wmi, struct sk_buff *skb) { skb_pull(skb, sizeof(struct wmi_cmd_hdr)); - if (wmi->cmd_rsp_buf != NULL && wmi->cmd_rsp_len != 0) + if (wmi->cmd_rsp_buf && wmi->cmd_rsp_len) { + if (skb->len < wmi->cmd_rsp_len) { + wmi->cmd_rsp_status = -EMSGSIZE; + goto complete; + } memcpy(wmi->cmd_rsp_buf, skb->data, wmi->cmd_rsp_len); + } + wmi->cmd_rsp_status = 0; +complete: complete(&wmi->cmd_wait); } @@ -300,6 +307,7 @@ static int ath9k_wmi_cmd_issue(struct wmi *wmi, /* record the rsp buffer and length */ wmi->cmd_rsp_buf = rsp_buf; wmi->cmd_rsp_len = rsp_len; + wmi->cmd_rsp_status = 0; wmi->last_seq_id = wmi->tx_seq_id; spin_unlock_irqrestore(&wmi->wmi_lock, flags); @@ -356,9 +364,13 @@ int ath9k_wmi_cmd(struct wmi *wmi, enum wmi_cmd_id cmd_id, return -ETIMEDOUT; } + spin_lock_irqsave(&wmi->wmi_lock, flags); + ret = wmi->cmd_rsp_status; + spin_unlock_irqrestore(&wmi->wmi_lock, flags); + mutex_unlock(&wmi->op_mutex); - return 0; + return ret; out: ath_dbg(common, WMI, "WMI failure for: %s\n", wmi_cmd_to_name(cmd_id)); diff --git a/drivers/net/wireless/ath/ath9k/wmi.h b/drivers/net/wireless/ath/ath9k/wmi.h index 5c3b710b8f31..0bb9d31f1a1c 100644 --- a/drivers/net/wireless/ath/ath9k/wmi.h +++ b/drivers/net/wireless/ath/ath9k/wmi.h @@ -157,6 +157,7 @@ struct wmi { u16 tx_seq_id; u8 *cmd_rsp_buf; u32 cmd_rsp_len; + int cmd_rsp_status; bool stopped; struct list_head pending_tx_events; From aeaca27c1db3106967ffbb2b517d98c835f6d500 Mon Sep 17 00:00:00 2001 From: Georgios Karantzas Date: Fri, 11 Sep 2026 03:11:43 +0300 Subject: [PATCH 298/694] wifi: ath9k_htc: bound TX aggregation to MAX_TX_BUF_SIZE MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit __hif_usb_tx() dequeues up to MAX_TX_AGGR_NUM (20) frames into a single tx_buf of MAX_TX_BUF_SIZE (32768) bytes, limiting the batch by record count but never by cumulative byte length. With large frames (MTU 2304), 20 aggregated frames of 2292 bytes each exceed the allocation (20 * 2296 = 45920 bytes), so the memcpy() in the loop writes up to 13152 bytes past tx_buf->buf before usb_submit_urb(). Peek the queue head and stop before copying any record that would cross MAX_TX_BUF_SIZE, then dispatch the current batch. Leftover skbs remain queued and are drained on the next URB completion. The byte bound changes the loop's exit semantics: it can now exit before i == tx_skb_cnt - 1. Stock only finalized tx_buf->len on that last index (len += offset), so an early break would submit a URB holding only the last record's length while every dequeued skb is freed on completion, silently dropping frames. Make tx_buf->len a running total and advance tx_buf->offset per record instead; the stride round_up(nskb->len + 4, 4) is identical to the stock stride when the loop runs to completion. Tested on hardware with an MTU 2304 flood: the loop stops at record 15 (len = 32144, offset = 32144, within 32768), no oversized URB is submitted, and MTU 1500 pings pass 20/20. Fixes: fb9987d0f748 ("ath9k_htc: Support for AR9271 chipset.") Signed-off-by: Georgios Karantzas Acked-by: Toke Høiland-Jørgensen Link: https://patch.msgid.link/20260911001143.4507-1-gck.kara@gmail.com Signed-off-by: Jeff Johnson --- drivers/net/wireless/ath/ath9k/hif_usb.c | 21 ++++++++++----------- 1 file changed, 10 insertions(+), 11 deletions(-) diff --git a/drivers/net/wireless/ath/ath9k/hif_usb.c b/drivers/net/wireless/ath/ath9k/hif_usb.c index d3491ff08e6e..7d2f292e20c5 100644 --- a/drivers/net/wireless/ath/ath9k/hif_usb.c +++ b/drivers/net/wireless/ath/ath9k/hif_usb.c @@ -14,6 +14,7 @@ * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. */ +#include #include #include "htc.h" @@ -328,11 +329,14 @@ static int __hif_usb_tx(struct hif_device_usb *hif_dev) tx_skb_cnt = min_t(u16, hif_dev->tx.tx_skb_cnt, MAX_TX_AGGR_NUM); for (i = 0; i < tx_skb_cnt; i++) { + nskb = skb_peek(&hif_dev->tx.tx_skb_queue); + if (!nskb) + break; + + if (tx_buf->offset + nskb->len + 4 > MAX_TX_BUF_SIZE) + break; + nskb = __skb_dequeue(&hif_dev->tx.tx_skb_queue); - - /* Should never be NULL */ - BUG_ON(!nskb); - hif_dev->tx.tx_skb_cnt--; buf = tx_buf->buf; @@ -342,13 +346,8 @@ static int __hif_usb_tx(struct hif_device_usb *hif_dev) *hdr++ = cpu_to_le16(ATH_USB_TX_STREAM_MODE_TAG); buf += 4; memcpy(buf, nskb->data, nskb->len); - tx_buf->len = nskb->len + 4; - - if (i < (tx_skb_cnt - 1)) - tx_buf->offset += (((tx_buf->len - 1) / 4) + 1) * 4; - - if (i == (tx_skb_cnt - 1)) - tx_buf->len += tx_buf->offset; + tx_buf->len = tx_buf->offset + nskb->len + 4; + tx_buf->offset += round_up(nskb->len + 4, 4); __skb_queue_tail(&tx_buf->skb_queue, nskb); TX_STAT_INC(hif_dev, skb_queued); From 0ee5c5d804d592a8328a425b9274487d393d3a05 Mon Sep 17 00:00:00 2001 From: Liu Dalin Date: Thu, 27 Aug 2026 09:54:44 +0800 Subject: [PATCH 299/694] rtc: dev: zero-initialize struct rtc_wkalrm to prevent information leak The struct rtc_wkalrm has padding holes after the 'pending' member. When returned to userspace via the RTC_WKALRM_RD ioctl, uninitialized kernel stack data residing in these holes may be leaked. Zero-initialize the struct at declaration to eliminate this information leak risk. Fixes smatch warnings: - drivers/rtc/dev.c:392 rtc_dev_ioctl() warn: check that 'alarm' doesn't leak information (struct has a hole after 'pending') Fixes: 36e14f5fdfdf ("rtc: rename core files") Assisted-by: smatch:2.0 [static analysis] Signed-off-by: Liu Dalin Link: https://patch.msgid.link/B28B7DE18F3A9447+20260827015444.3314227-1-liudalin@kylinsec.com.cn Signed-off-by: Alexandre Belloni --- drivers/rtc/dev.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/rtc/dev.c b/drivers/rtc/dev.c index 8ba7c25d2565..be173c9511fa 100644 --- a/drivers/rtc/dev.c +++ b/drivers/rtc/dev.c @@ -216,7 +216,7 @@ static long rtc_dev_ioctl(struct file *file, struct rtc_device *rtc = file->private_data; const struct rtc_class_ops *ops = rtc->ops; struct rtc_time tm; - struct rtc_wkalrm alarm; + struct rtc_wkalrm alarm = { 0 }; struct rtc_param param; void __user *uarg = (void __user *)arg; From b430d1f6d80c005b486f028910e1ac3c91812710 Mon Sep 17 00:00:00 2001 From: "Johnny.Fan" Date: Mon, 31 Aug 2026 22:00:14 +0800 Subject: [PATCH 300/694] rtc: efi: restore alarm support with runtime capability probe Restore the EFI RTC alarm functionality (read_alarm/set_alarm and procfs output) that was removed by commit 18a3510bc87d ("rtc: efi: Remove wakeup functionality"). The removal was motivated by the observation that many EFI platforms lack RTC wakeup support while also omitting the RT_PROP table, causing the wakeup capability to be incorrectly advertised. While that is true, removing the feature entirely also breaks platforms where the wakeup runtime service actually works. On CIX SoCs, the RTC is an I2C device owned by firmware, and in the ACPI boot configuration the I2C bus it sits on is not exposed to the OS, so the EFI runtime services are the only way Linux can access the RTC at all. Both Linux and Windows use this same interface by design, so the RTC state is consistent across the operating systems installed on the machine, and Windows relies on the EFI wakeup services for its own RTC alarm support. With the alarm support removed from rtc-efi, Linux loses the equivalent capability entirely on these platforms: rtcwake fails for both suspend-to-RAM and power-off wakeup, as there is no other path to program the RTC alarm. This is a real product requirement that CIX customers have explicitly requested, not a hypothetical use case. To avoid exposing a broken alarm on firmware that does not implement the wakeup runtime services, probe the capability at ->probe() time by actually calling GetWakeupTime() -- if it fails, clear RTC_FEATURE_ALARM and the device behaves as before the removal. Also set RTC_FEATURE_ALARM_WAKEUP_ONLY since EFI wakeup time is a wakeup-only alarm, not a periodic/update interrupt source. Acked-by: Ard Biesheuvel Reviewed-by: Feng Tang Reviewed-by: Fugang Duan Signed-off-by: Johnny.Fan Link: https://patch.msgid.link/20260831140014.839222-1-johnny.fan@cixtech.com Signed-off-by: Alexandre Belloni --- drivers/rtc/rtc-efi.c | 80 +++++++++++++++++++++++++++++++++++++++++-- 1 file changed, 78 insertions(+), 2 deletions(-) diff --git a/drivers/rtc/rtc-efi.c b/drivers/rtc/rtc-efi.c index b4f44999ef0f..a8c18ebc5599 100644 --- a/drivers/rtc/rtc-efi.c +++ b/drivers/rtc/rtc-efi.c @@ -52,6 +52,7 @@ compute_wday(efi_time_t *eft, int yday) static void convert_to_efi_time(struct rtc_time *wtime, efi_time_t *eft) { + memset(eft, 0, sizeof(*eft)); eft->year = wtime->tm_year + 1900; eft->month = wtime->tm_mon + 1; eft->day = wtime->tm_mday; @@ -112,6 +113,35 @@ convert_from_efi_time(efi_time_t *eft, struct rtc_time *wtime) return true; } +static int efi_read_alarm(struct device *dev, struct rtc_wkalrm *wkalrm) +{ + efi_time_t eft; + efi_status_t status; + + memset(&eft, 0, sizeof(eft)); + status = efi.get_wakeup_time((efi_bool_t *)&wkalrm->enabled, + (efi_bool_t *)&wkalrm->pending, &eft); + if (status != EFI_SUCCESS) + return -EINVAL; + + if (!convert_from_efi_time(&eft, &wkalrm->time)) + return -EIO; + + return rtc_valid_tm(&wkalrm->time); +} + +static int efi_set_alarm(struct device *dev, struct rtc_wkalrm *wkalrm) +{ + efi_time_t eft; + efi_status_t status; + + convert_to_efi_time(&wkalrm->time, &eft); + + status = efi.set_wakeup_time((efi_bool_t)!!wkalrm->enabled, &eft); + + return status == EFI_SUCCESS ? 0 : -EINVAL; +} + static int efi_read_time(struct device *dev, struct rtc_time *tm) { efi_status_t status; @@ -146,13 +176,21 @@ static int efi_set_time(struct device *dev, struct rtc_time *tm) static int efi_procfs(struct device *dev, struct seq_file *seq) { - efi_time_t eft; + efi_time_t eft, alm; efi_time_cap_t cap; + efi_bool_t enabled, pending; + struct rtc_device *rtc = dev_get_drvdata(dev); memset(&eft, 0, sizeof(eft)); + memset(&alm, 0, sizeof(alm)); memset(&cap, 0, sizeof(cap)); efi.get_time(&eft, &cap); + if (test_bit(RTC_FEATURE_ALARM, rtc->features) && + efi.get_wakeup_time(&enabled, &pending, &alm) != EFI_SUCCESS) { + enabled = 0; + pending = 0; + } seq_printf(seq, "Time\t\t: %u:%u:%u.%09u\n" @@ -168,6 +206,25 @@ static int efi_procfs(struct device *dev, struct seq_file *seq) /* XXX fixme: convert to string? */ seq_printf(seq, "Timezone\t: %u\n", eft.timezone); + if (test_bit(RTC_FEATURE_ALARM, rtc->features)) { + seq_printf(seq, + "Alarm Time\t: %u:%u:%u.%09u\n" + "Alarm Date\t: %u-%u-%u\n" + "Alarm Daylight\t: %u\n" + "Enabled\t\t: %s\n" + "Pending\t\t: %s\n", + alm.hour, alm.minute, alm.second, alm.nanosecond, + alm.year, alm.month, alm.day, + alm.daylight, + enabled == 1 ? "yes" : "no", + pending == 1 ? "yes" : "no"); + + if (alm.timezone == EFI_UNSPECIFIED_TIMEZONE) + seq_puts(seq, "Alarm Timezone\t: unspecified\n"); + else + seq_printf(seq, "Alarm Timezone\t: %d\n", alm.timezone); + } + /* * now prints the capabilities */ @@ -183,6 +240,8 @@ static int efi_procfs(struct device *dev, struct seq_file *seq) static const struct rtc_class_ops efi_rtc_ops = { .read_time = efi_read_time, .set_time = efi_set_time, + .read_alarm = efi_read_alarm, + .set_alarm = efi_set_alarm, .proc = efi_procfs, }; @@ -191,6 +250,7 @@ static int __init efi_rtc_probe(struct platform_device *dev) struct rtc_device *rtc; efi_time_t eft; efi_time_cap_t cap; + efi_bool_t enabled, pending; /* First check if the RTC is usable */ if (efi.get_time(&eft, &cap) != EFI_SUCCESS) @@ -203,7 +263,23 @@ static int __init efi_rtc_probe(struct platform_device *dev) platform_set_drvdata(dev, rtc); rtc->ops = &efi_rtc_ops; - clear_bit(RTC_FEATURE_ALARM, rtc->features); + clear_bit(RTC_FEATURE_UPDATE_INTERRUPT, rtc->features); + + /* + * The EFI_RT_SUPPORTED_WAKEUP_SERVICES bit defaults to enabled + * and only gets cleared when the RT_PROP table explicitly says + * wakeup is unsupported. Many platforms lack an RT_PROP table + * even though they don't implement the wakeup runtime service, + * so probe by actually calling GetWakeupTime() to avoid exposing + * a broken alarm to userspace. + */ + if (efi_rt_services_supported(EFI_RT_SUPPORTED_WAKEUP_SERVICES) && + efi.get_wakeup_time(&enabled, &pending, &eft) == EFI_SUCCESS) { + set_bit(RTC_FEATURE_ALARM, rtc->features); + set_bit(RTC_FEATURE_ALARM_WAKEUP_ONLY, rtc->features); + } else { + clear_bit(RTC_FEATURE_ALARM, rtc->features); + } device_init_wakeup(&dev->dev, true); From 6a4117eee82dd85f11bf898bf66026764011eeb6 Mon Sep 17 00:00:00 2001 From: Aamir Ahmed Date: Sat, 5 Sep 2026 19:38:08 +0100 Subject: [PATCH 301/694] rtc: ac100: Assign .num before accessing .hws Commit f316cdff8d67 ("clk: Annotate struct clk_hw_onecell_data with __counted_by") annotated the hws member of 'struct clk_hw_onecell_data' with __counted_by, which informs the bounds sanitizer (UBSAN_BOUNDS) about the number of elements in .hws[], so that it can warn when .hws[] is accessed out of bounds. As noted in that change, the __counted_by member must be initialized with the number of elements before the first array access happens, otherwise there will be a warning from each access prior to the initialization because the number of elements is zero. This occurs in ac100_rtc_register_clks() due to .num being assigned only after every clkout clock has been stored in .hws[]. With CONFIG_UBSAN_BOUNDS and a compiler that implements __counted_by (GCC 15.1+ or Clang 20.1+), this triggers an array-index-out-of-bounds report during probe, and with CONFIG_UBSAN_TRAP the first store traps. Initialize .num with AC100_CLKOUT_NUM, the number of elements .hws[] was allocated with, right after the allocation. That is the value the loop counter ends up at on the success path anyway, so the provider's behaviour is unchanged. Cc: stable@vger.kernel.org Fixes: f316cdff8d67 ("clk: Annotate struct clk_hw_onecell_data with __counted_by") Assisted-by: LLM Signed-off-by: Aamir Ahmed Reviewed-by: Chen-Yu Tsai Reviewed-by: Gustavo A. R. Silva Link: https://patch.msgid.link/AS8P251MB00013E724A77A355668B6CCEC8B42@AS8P251MB0001.EURP251.PROD.OUTLOOK.COM Signed-off-by: Alexandre Belloni --- drivers/rtc/rtc-ac100.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/rtc/rtc-ac100.c b/drivers/rtc/rtc-ac100.c index bba7115ff3ad..a2f465438fd0 100644 --- a/drivers/rtc/rtc-ac100.c +++ b/drivers/rtc/rtc-ac100.c @@ -317,6 +317,8 @@ static int ac100_rtc_register_clks(struct ac100_rtc_dev *chip) if (!chip->clk_data) return -ENOMEM; + chip->clk_data->num = AC100_CLKOUT_NUM; + chip->rtc_32k_clk = clk_hw_register_fixed_rate(chip->dev, AC100_RTC_32K_NAME, NULL, 0, @@ -360,7 +362,6 @@ static int ac100_rtc_register_clks(struct ac100_rtc_dev *chip) chip->clk_data->hws[i] = &clk->hw; } - chip->clk_data->num = i; ret = of_clk_add_hw_provider(np, of_clk_hw_onecell_get, chip->clk_data); if (ret) goto err_unregister_rtc_32k; From fcf0b58e976e83adf246b014518e49c662b96f5e Mon Sep 17 00:00:00 2001 From: Aamir Ahmed Date: Sat, 5 Sep 2026 21:41:52 +0100 Subject: [PATCH 302/694] rtc: ac100: Fix clock provider use-after-free on probe failure ac100_rtc_register_clks() registers the RTC-32k clock with clk_hw_register_fixed_rate() and the clock provider with of_clk_add_hw_provider(). Neither is device managed, and both are released only from ac100_rtc_remove(). The driver core does not call remove() when probe() fails: really_probe() reaches probe_failed below the device_remove() call and goes straight to releasing the device's managed resources. ac100_rtc_probe() ends with return devm_rtc_register_device(chip->rtc); so if that fails after the clocks have been registered, the provider is left on the global of_clk_providers list holding chip->clk_data, which was allocated with devm_kzalloc() and is freed while probe() unwinds. A later lookup on this device tree node then reads freed memory in of_clk_hw_onecell_get(). Consumers that do exactly that exist in tree: the wifi power sequence nodes on sun8i-a83t-bananapi-m3 and sun8i-a83t-cubietruck-plus take <&ac100_rtc 1>, and the sun9i-a80 boards route osc32k through <&ac100_rtc 0>. The window is narrow, as devm_rtc_register_device() can only fail with -ENOMEM here, but the provider is left dangling whenever it does. The RTC-32k clock is leaked on the same path. Since clk_core_lookup() searches a global list that is not scoped per device, __clk_register() rejects the duplicate name with -EEXIST, so the leak also makes a later probe of the same device fail. The error path that returns -EINVAL when the ADDA 4M parent clock cannot be found leaks the RTC-32k clock in the same way. No provider has been registered at that point, so that one is a leak rather than a use-after-free. Register both with devm_clk_hw_register_fixed_rate() and devm_of_clk_add_hw_provider() so that they are released whenever the device goes away, on a failed probe as well as on unbind. Devres releases in reverse order of acquisition, so the provider is removed before chip->clk_data is freed, and the clkout children are now unregistered before their parent rather than after it. ac100_rtc_unregister_clks() and the remove callback then have nothing left to do and are removed. Fixes: d00a18a42c14 ("rtc: ac100: Add RTC driver for X-Powers AC100") Reported-by: Sashiko AI Closes: https://lore.kernel.org/linux-rtc/20260905184936.155E11F00A3A@smtp.kernel.org/ Assisted-by: LLM Signed-off-by: Aamir Ahmed Link: https://patch.msgid.link/AS8P251MB0001A489CC45A591ABBB9BDCC8B42@AS8P251MB0001.EURP251.PROD.OUTLOOK.COM Signed-off-by: Alexandre Belloni --- drivers/rtc/rtc-ac100.c | 36 +++++++----------------------------- 1 file changed, 7 insertions(+), 29 deletions(-) diff --git a/drivers/rtc/rtc-ac100.c b/drivers/rtc/rtc-ac100.c index a2f465438fd0..c579c4bf5578 100644 --- a/drivers/rtc/rtc-ac100.c +++ b/drivers/rtc/rtc-ac100.c @@ -319,10 +319,10 @@ static int ac100_rtc_register_clks(struct ac100_rtc_dev *chip) chip->clk_data->num = AC100_CLKOUT_NUM; - chip->rtc_32k_clk = clk_hw_register_fixed_rate(chip->dev, - AC100_RTC_32K_NAME, - NULL, 0, - AC100_RTC_32K_RATE); + chip->rtc_32k_clk = devm_clk_hw_register_fixed_rate(chip->dev, + AC100_RTC_32K_NAME, + NULL, 0, + AC100_RTC_32K_RATE); if (IS_ERR(chip->rtc_32k_clk)) { ret = PTR_ERR(chip->rtc_32k_clk); dev_err(chip->dev, "Failed to register RTC-32k clock: %d\n", @@ -356,28 +356,14 @@ static int ac100_rtc_register_clks(struct ac100_rtc_dev *chip) if (ret) { dev_err(chip->dev, "Failed to register clk '%s': %d\n", init.name, ret); - goto err_unregister_rtc_32k; + return ret; } chip->clk_data->hws[i] = &clk->hw; } - ret = of_clk_add_hw_provider(np, of_clk_hw_onecell_get, chip->clk_data); - if (ret) - goto err_unregister_rtc_32k; - - return 0; - -err_unregister_rtc_32k: - clk_unregister_fixed_rate(chip->rtc_32k_clk->clk); - - return ret; -} - -static void ac100_rtc_unregister_clks(struct ac100_rtc_dev *chip) -{ - of_clk_del_provider(chip->dev->of_node); - clk_unregister_fixed_rate(chip->rtc_32k_clk->clk); + return devm_of_clk_add_hw_provider(chip->dev, of_clk_hw_onecell_get, + chip->clk_data); } /* @@ -615,13 +601,6 @@ static int ac100_rtc_probe(struct platform_device *pdev) return devm_rtc_register_device(chip->rtc); } -static void ac100_rtc_remove(struct platform_device *pdev) -{ - struct ac100_rtc_dev *chip = platform_get_drvdata(pdev); - - ac100_rtc_unregister_clks(chip); -} - static const struct of_device_id ac100_rtc_match[] = { { .compatible = "x-powers,ac100-rtc" }, { }, @@ -630,7 +609,6 @@ MODULE_DEVICE_TABLE(of, ac100_rtc_match); static struct platform_driver ac100_rtc_driver = { .probe = ac100_rtc_probe, - .remove = ac100_rtc_remove, .driver = { .name = "ac100-rtc", .of_match_table = of_match_ptr(ac100_rtc_match), From ac41b05d15db3134e839148290d67415ee0bdb58 Mon Sep 17 00:00:00 2001 From: Manush Prajwal Date: Sun, 6 Sep 2026 16:44:09 +0530 Subject: [PATCH 303/694] rtc: mpfs: fix unchecked devm_clk_get() error pointer in probe() devm_clk_get(&pdev->dev, "rtcref")'s return value was passed straight into clk_get_rate() without checking it for an error first, unlike the "rtc" clock a few lines above which is correctly checked with IS_ERR(). clk_get_rate() only guards against a NULL clk, not an error pointer: if (!clk) return 0; ... rate = clk_core_get_rate_recalc(clk->core); so if devm_clk_get() ever returns an error pointer here (for example ERR_PTR(-EPROBE_DEFER), which is the normal, expected outcome if the clkcfg clock-provider this RTC depends on has not registered its clocks yet by the time this driver probes), clk_get_rate() dereferences that error pointer instead of returning 0, crashing instead of letting probe defer. Capture the clock in the existing 'clk' local and check it with IS_ERR() before calling clk_get_rate(), matching the handling already used for the "rtc" clock in this same function. Signed-off-by: Manush Prajwal Fixes: 0b31d703598d ("rtc: Add driver for Microchip PolarFire SoC") Reviewed-by: Conor Dooley Link: https://patch.msgid.link/6a9d4b02.7d74b517.206c4c.75f8@mx.google.com Signed-off-by: Alexandre Belloni --- drivers/rtc/rtc-mpfs.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/drivers/rtc/rtc-mpfs.c b/drivers/rtc/rtc-mpfs.c index ece6de4a6adb..60596b0ea9d8 100644 --- a/drivers/rtc/rtc-mpfs.c +++ b/drivers/rtc/rtc-mpfs.c @@ -256,8 +256,12 @@ static int mpfs_rtc_probe(struct platform_device *pdev) return ret; } + clk = devm_clk_get(&pdev->dev, "rtcref"); + if (IS_ERR(clk)) + return PTR_ERR(clk); + /* prescaler hardware adds 1 to reg value */ - prescaler = clk_get_rate(devm_clk_get(&pdev->dev, "rtcref")) - 1; + prescaler = clk_get_rate(clk) - 1; if (prescaler > MAX_PRESCALER_COUNT) { dev_dbg(&pdev->dev, "invalid prescaler %lu\n", prescaler); return -EINVAL; From 94b7e3a7e871ae27d4935c76959dfc61829f27f9 Mon Sep 17 00:00:00 2001 From: Helge Deller Date: Sat, 19 Sep 2026 22:09:20 +0200 Subject: [PATCH 304/694] parisc: Increase kernel stack size to 32kb For 64-bit Linux kernels, increase the default kernel stack size (THREAD_SIZE_ORDER) to 32 kB, in order to avoid kernel crashes which have been triggered recently when building the debian vtk9 package with gcc 17: stackcheck: kworker/u128:0 will most likely overflow kernel stack (sp:179a83af0, stk bottom-top:179a80000-179a84000) Kernel panic - not syncing: low stack detected by irq handler - check messages CPU: 2 UID: 0 PID: 30760 Comm: kworker/u128:0 Tainted: G W 6.18.46-dirty #1 NONE Tainted: [W]=WARN Hardware name: 9000/800/rp3440 Workqueue: writeback wb_workfn (flush-259:0) Backtrace: [<000000004022f050>] show_stack+0x70/0x90 [<000000004022378c>] dump_stack_lvl+0x124/0x190 [<000000004022382c>] dump_stack+0x34/0x48 [<000000004020212c>] vpanic+0x204/0x648 [<00000000402025c4>] panic+0x54/0x58 [<0000000040232230>] do_cpu_irq_mask+0x3f8/0x440 [<0000000040227070>] intr_return+0x0/0xc Signed-off-by: Helge Deller Reported-by: John David Anglin Cc: stable@vger.kernel.org # v6.18+ --- arch/parisc/include/asm/thread_info.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/parisc/include/asm/thread_info.h b/arch/parisc/include/asm/thread_info.h index b283738bb6da..249370706242 100644 --- a/arch/parisc/include/asm/thread_info.h +++ b/arch/parisc/include/asm/thread_info.h @@ -24,7 +24,7 @@ struct thread_info { /* thread information allocation */ -#ifdef CONFIG_IRQSTACKS +#if defined(CONFIG_IRQSTACKS) && !defined(CONFIG_64BIT) #define THREAD_SIZE_ORDER 2 /* PA-RISC requires at least 16k stack */ #else #define THREAD_SIZE_ORDER 3 /* PA-RISC requires at least 32k stack */ From 055ef5ce9f67a6a3a1363fa663007f8196cc0fb8 Mon Sep 17 00:00:00 2001 From: Runyu Xiao Date: Wed, 2 Sep 2026 16:03:18 +0800 Subject: [PATCH 305/694] rtc: spear: initialize IRQ state before requesting alarm IRQ devm_request_irq() enables the interrupt before it returns, so the handler may run while probe is still initializing the device. Initialize the MMIO address and spinlock before requesting the alarm IRQ. Fixes: 0942a71e435f ("rtc: add support for spear rtc") Cc: stable@vger.kernel.org Assisted-by: Codex:GPT-5 Signed-off-by: Runyu Xiao Link: https://patch.msgid.link/20260902080318.3498434-1-runyu.xiao@seu.edu.cn Signed-off-by: Alexandre Belloni --- drivers/rtc/rtc-spear.c | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) diff --git a/drivers/rtc/rtc-spear.c b/drivers/rtc/rtc-spear.c index 959acff8faff..d12bccb3cf5e 100644 --- a/drivers/rtc/rtc-spear.c +++ b/drivers/rtc/rtc-spear.c @@ -363,6 +363,12 @@ static int spear_rtc_probe(struct platform_device *pdev) if (irq < 0) return irq; + config->ioaddr = devm_platform_ioremap_resource(pdev, 0); + if (IS_ERR(config->ioaddr)) + return PTR_ERR(config->ioaddr); + + spin_lock_init(&config->lock); + status = devm_request_irq(&pdev->dev, irq, spear_rtc_irq, 0, pdev->name, config); if (status) { @@ -371,10 +377,6 @@ static int spear_rtc_probe(struct platform_device *pdev) return status; } - config->ioaddr = devm_platform_ioremap_resource(pdev, 0); - if (IS_ERR(config->ioaddr)) - return PTR_ERR(config->ioaddr); - config->clk = devm_clk_get(&pdev->dev, NULL); if (IS_ERR(config->clk)) return PTR_ERR(config->clk); @@ -383,7 +385,6 @@ static int spear_rtc_probe(struct platform_device *pdev) if (status < 0) return status; - spin_lock_init(&config->lock); platform_set_drvdata(pdev, config); config->rtc->ops = &spear_rtc_ops; From 79a9172f3ab4ad8392c5e5c8944b9b7710ade620 Mon Sep 17 00:00:00 2001 From: Xu Yunxiang Date: Mon, 21 Sep 2026 05:04:21 +0800 Subject: [PATCH 306/694] bpf: Reject non-negative offsets in stack_slot_obj_get_spi() bpf_get_spi() computes (-off - 1) / BPF_REG_SIZE using C division, which truncates toward zero. For off == 0, this produces spi 0, the same index used by the valid stack slot at fp-8. stack_slot_obj_get_spi() currently checks alignment and the resulting spi bounds, but does not reject the non-negative offset itself. It can therefore validate a PTR_TO_STACK register holding fp+0 against an iterator stored at fp-8 even though the runtime receives the actual fp+0 pointer. An effectful iterator kfunc can then interpret memory outside the BPF stack as iterator state. Reject non-negative offsets before converting the offset to an spi. All valid stack objects begin at a negative offset from the frame pointer. Fixes: 06accc8779c1 ("bpf: add support for open-coded iterator loops") Signed-off-by: Xu Yunxiang Signed-off-by: Andrii Nakryiko Reviewed-by: Sun Jian Link: https://lore.kernel.org/bpf/20260920210423.345636-2-xyx2021@mail.ustc.edu.cn --- kernel/bpf/verifier.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 02be326f235c..dd8bb179d39e 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -567,7 +567,7 @@ static int stack_slot_obj_get_spi(struct bpf_verifier_env *env, struct bpf_reg_s } off = reg->var_off.value; - if (off % BPF_REG_SIZE) { + if (off >= 0 || off % BPF_REG_SIZE) { verbose(env, "cannot pass in %s at an offset=%d\n", obj_kind, off); return -EINVAL; } From 8244668cbbffc8e242b2c5204a2dea20bfca096c Mon Sep 17 00:00:00 2001 From: Xu Yunxiang Date: Mon, 21 Sep 2026 05:04:22 +0800 Subject: [PATCH 307/694] selftests/bpf: Reject iterator destruction through fp+0 Add a verifier regression test that initializes a numeric iterator at fp-8 and attempts to destroy it through fp+0. The verifier must reject the non-negative offset instead of treating it as the initialized stack slot. Check the offset diagnostic to ensure rejection happens at the stack object address check. The numeric iterator destroy operation is a no-op; this test checks verifier rejection and does not run the program. Signed-off-by: Xu Yunxiang Signed-off-by: Andrii Nakryiko Reviewed-by: Sun Jian Link: https://lore.kernel.org/bpf/20260920210423.345636-3-xyx2021@mail.ustc.edu.cn --- .../selftests/bpf/progs/iters_state_safety.c | 22 +++++++++++++++++++ 1 file changed, 22 insertions(+) diff --git a/tools/testing/selftests/bpf/progs/iters_state_safety.c b/tools/testing/selftests/bpf/progs/iters_state_safety.c index 646026430e9b..e5bb9fe6d5e5 100644 --- a/tools/testing/selftests/bpf/progs/iters_state_safety.c +++ b/tools/testing/selftests/bpf/progs/iters_state_safety.c @@ -52,6 +52,28 @@ int create_and_destroy(void *ctx) return 0; } +/* fp+0 is not a stack slot. bpf_get_spi(0) used to alias spi 0 (fp-8). */ +SEC("?raw_tp") +__failure __msg("cannot pass in iter at an offset=0") +int destroy_fp0_fail(void *ctx) +{ + struct bpf_iter_num iter; + + asm volatile ("r1 = %[iter];" + "r2 = 0;" + "r3 = 1000;" + "call %[bpf_iter_num_new];" + /* r10 is fp+0, one byte above the top of the BPF stack */ + "r1 = r10;" + "call %[bpf_iter_num_destroy];" + : + : __imm_ptr(iter), ITER_HELPERS + : __clobber_common + ); + + return 0; +} + SEC("?raw_tp") __failure __msg("Unreleased reference id=1") int create_and_forget_to_destroy_fail(void *ctx) From d06f2ebf67ff2962fe00d687e4f0d4703eb41a12 Mon Sep 17 00:00:00 2001 From: Ratheesh Kannoth Date: Wed, 16 Sep 2026 07:51:11 +0530 Subject: [PATCH 308/694] octeontx2-af: Fix memory scaling limitation in SR-IOV mode The original code used DMA_ATTR_FORCE_CONTIGUOUS, which could exhaust the CMA pool when a large number of VFs were requested. Fix this by switching to the DMA streaming API. This is equivalent on Octeon platforms, which provide full I/O coherency via the SMMU. Cc: Leon Romanovsky Fixes: 73d33dbc0723 ("octeontx2-af: Use DMA_ATTR_FORCE_CONTIGUOUS attribute in DMA alloc") Signed-off-by: Ratheesh Kannoth Reviewed-by: Leon Romanovsky Link: https://patch.msgid.link/20260916022111.1083017-1-rkannoth@marvell.com Signed-off-by: Jakub Kicinski --- .../ethernet/marvell/octeontx2/af/common.h | 45 ++++++++++++++++--- 1 file changed, 39 insertions(+), 6 deletions(-) diff --git a/drivers/net/ethernet/marvell/octeontx2/af/common.h b/drivers/net/ethernet/marvell/octeontx2/af/common.h index 779413a383b7..78e42549d990 100644 --- a/drivers/net/ethernet/marvell/octeontx2/af/common.h +++ b/drivers/net/ethernet/marvell/octeontx2/af/common.h @@ -7,6 +7,10 @@ #ifndef COMMON_H #define COMMON_H +#include +#include +#include + #include "rvu_struct.h" #define OTX2_ALIGN 128 /* Align to cacheline */ @@ -44,6 +48,33 @@ struct qmem { u32 qsize; }; +static inline void *otx2_dma_alloc_coherent(struct device *dev, size_t size, + dma_addr_t *dma_handle) +{ + dma_addr_t dma_addr; + void *vaddr; + + vaddr = kzalloc(size, GFP_KERNEL); + if (!vaddr) + return NULL; + + dma_addr = dma_map_single(dev, vaddr, size, DMA_BIDIRECTIONAL); + if (dma_mapping_error(dev, dma_addr)) { + kfree(vaddr); + return NULL; + } + + *dma_handle = dma_addr; + return vaddr; +} + +static inline void otx2_dma_free_coherent(struct device *dev, size_t size, + void *vaddr, dma_addr_t dma_handle) +{ + dma_unmap_single(dev, dma_handle, size, DMA_BIDIRECTIONAL); + kfree(vaddr); +} + static inline int qmem_alloc(struct device *dev, struct qmem **q, int qsize, int entry_sz) { @@ -60,8 +91,11 @@ static inline int qmem_alloc(struct device *dev, struct qmem **q, qmem->entry_sz = entry_sz; qmem->alloc_sz = (qsize * entry_sz) + OTX2_ALIGN; - qmem->base = dma_alloc_attrs(dev, qmem->alloc_sz, &qmem->iova, - GFP_KERNEL, DMA_ATTR_FORCE_CONTIGUOUS); + + if (get_order(PAGE_ALIGN(qmem->alloc_sz)) > MAX_PAGE_ORDER) + return -ENOMEM; + + qmem->base = otx2_dma_alloc_coherent(dev, qmem->alloc_sz, &qmem->iova); if (!qmem->base) return -ENOMEM; @@ -80,10 +114,9 @@ static inline void qmem_free(struct device *dev, struct qmem *qmem) return; if (qmem->base) - dma_free_attrs(dev, qmem->alloc_sz, - qmem->base - qmem->align, - qmem->iova - qmem->align, - DMA_ATTR_FORCE_CONTIGUOUS); + otx2_dma_free_coherent(dev, qmem->alloc_sz, + qmem->base - qmem->align, + qmem->iova - qmem->align); devm_kfree(dev, qmem); } From 0346ec2f080b40d95ed05b853bb9226289e75212 Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Fri, 18 Sep 2026 08:22:05 +0000 Subject: [PATCH 309/694] ipv6: Prevent rt6_insert_exception() for dying fib6_info. Before the cited commit, fib6_nh_flush_exceptions() always set from->exception_bucket_flushed = 1 under rt6_exception_lock to prevent rt6_insert_exception() from inserting a new exception for a dying fib6_info. The flag was replaced with the FIB6_EXCEPTION_BUCKET_FLUSHED bit stored in nh->rt6i_exception_bucket. The problem is that now the bit is only set when the bucket is not NULL and fib6_nh_flush_exceptions() is called from fib6_nh_release() after fib6_ref has already reached zero. If rt6_insert_exception() is called while the target fib6_info is being removed via fib6_purge_rt(), a new exception could be created successfully because rt6_flush_exceptions() no longer sets the bit. This creates a reference cycle between the fib6_info and the exception route, leaking the fib6_info, its nexthop device, and all per-CPU routes in fib6_nh->rt6i_pcpu, which stalls netdev unregistration. [ 34.680602] unregister_netdevice: waiting for gre6 to become free. Usage count = 68 [ 44.920675] unregister_netdevice: waiting for gre6 to become free. Usage count = 68 [ 55.176582] unregister_netdevice: waiting for gre6 to become free. Usage count = 68 Let's call fib6_drop_pcpu_from() before rt6_flush_exceptions(), to set fib6_destroying before rt6_exception_lock, and check f6i->fib6_destroying in rt6_insert_exception(). Note that FIB6_EXCEPTION_BUCKET_FLUSHED logic is dead and we can clean it up in net-next. Fixes: cc5c073a693f ("ipv6: Move exception bucket to fib6_nh") Signed-off-by: Kuniyuki Iwashima Reviewed-by: Ido Schimmel Link: https://patch.msgid.link/20260918082209.2853582-1-kuniyu@google.com Signed-off-by: Jakub Kicinski --- net/ipv6/ip6_fib.c | 2 +- net/ipv6/route.c | 5 +++++ 2 files changed, 6 insertions(+), 1 deletion(-) diff --git a/net/ipv6/ip6_fib.c b/net/ipv6/ip6_fib.c index 9ea75703b38d..9ff761962b45 100644 --- a/net/ipv6/ip6_fib.c +++ b/net/ipv6/ip6_fib.c @@ -1043,8 +1043,8 @@ static void fib6_purge_rt(struct fib6_info *rt, struct fib6_node *fn, struct fib6_table *table = rt->fib6_table; /* Flush all cached dst in exception table */ - rt6_flush_exceptions(rt); fib6_drop_pcpu_from(rt); + rt6_flush_exceptions(rt); if (rt->nh) { spin_lock(&rt->nh->lock); diff --git a/net/ipv6/route.c b/net/ipv6/route.c index 08bd68f1b5bb..884d9ab0d50d 100644 --- a/net/ipv6/route.c +++ b/net/ipv6/route.c @@ -1729,6 +1729,11 @@ static int rt6_insert_exception(struct rt6_info *nrt, spin_lock_bh(&rt6_exception_lock); + if (f6i->fib6_destroying) { + err = -ENOENT; + goto out; + } + bucket = rcu_dereference_protected(nh->rt6i_exception_bucket, lockdep_is_held(&rt6_exception_lock)); if (!bucket) { From 31995571219c8ac30913d9c0dccad033fbb0b3da Mon Sep 17 00:00:00 2001 From: Nicolai Buchwitz Date: Fri, 18 Sep 2026 11:55:40 +0200 Subject: [PATCH 310/694] net: don't require the hwtstamp NDOs when a PHY provides timestamping Removing the legacy ioctl fallback made both hwtstamp NDOs mandatory. A device that only timestamps in its PHY implements neither, so SIOCSHWTSTAMP fails with EOPNOTSUPP before anything looks at the PHY and PTP stops working there. The check only ever picked the legacy path. That path is gone, so drop it and test where the NDOs are actually called. SIOCGHWTSTAMP is new here, not restored. The old path went through phy_mii_ioctl(), which only handled SIOCSHWTSTAMP. Such a device now returns -ENODEV while absent instead of -EOPNOTSUPP, like the ones that do implement the NDOs. Fixes: 5062245a5a7f ("net: remove legacy way to get/set HW timestamp config") Signed-off-by: Nicolai Buchwitz Reviewed-by: Kory Maincent Link: https://patch.msgid.link/20260918095540.34286-1-nb@tipi-net.de Signed-off-by: Jakub Kicinski --- net/core/dev_ioctl.c | 25 +++++++++---------------- 1 file changed, 9 insertions(+), 16 deletions(-) diff --git a/net/core/dev_ioctl.c b/net/core/dev_ioctl.c index a320e264eaaf..164643140a52 100644 --- a/net/core/dev_ioctl.c +++ b/net/core/dev_ioctl.c @@ -276,19 +276,18 @@ int dev_get_hwtstamp_phylib(struct net_device *dev, if (phy_is_default_hwtstamp(dev->phydev)) return phy_hwtstamp_get(dev->phydev, cfg); + if (!dev->netdev_ops->ndo_hwtstamp_get) + return -EOPNOTSUPP; + return dev->netdev_ops->ndo_hwtstamp_get(dev, cfg); } static int dev_get_hwtstamp(struct net_device *dev, struct ifreq *ifr) { - const struct net_device_ops *ops = dev->netdev_ops; struct kernel_hwtstamp_config kernel_cfg = {}; struct hwtstamp_config cfg; int err; - if (!ops->ndo_hwtstamp_get) - return -EOPNOTSUPP; - if (!netif_device_present(dev)) return -ENODEV; @@ -359,12 +358,18 @@ int dev_set_hwtstamp_phylib(struct net_device *dev, cfg->source = phy_ts ? HWTSTAMP_SOURCE_PHYLIB : HWTSTAMP_SOURCE_NETDEV; if (phy_ts && dev->see_all_hwtstamp_requests) { + if (!ops->ndo_hwtstamp_get) + return -EOPNOTSUPP; + err = ops->ndo_hwtstamp_get(dev, &old_cfg); if (err) return err; } if (!phy_ts || dev->see_all_hwtstamp_requests) { + if (!ops->ndo_hwtstamp_set) + return -EOPNOTSUPP; + err = ops->ndo_hwtstamp_set(dev, cfg, extack); if (err) { if (extack->_msg) @@ -390,7 +395,6 @@ int dev_set_hwtstamp_phylib(struct net_device *dev, static int dev_set_hwtstamp(struct net_device *dev, struct ifreq *ifr) { - const struct net_device_ops *ops = dev->netdev_ops; struct kernel_hwtstamp_config kernel_cfg = {}; struct netlink_ext_ack extack = {}; struct hwtstamp_config cfg; @@ -413,9 +417,6 @@ static int dev_set_hwtstamp(struct net_device *dev, struct ifreq *ifr) return err; } - if (!ops->ndo_hwtstamp_set) - return -EOPNOTSUPP; - if (!netif_device_present(dev)) return -ENODEV; @@ -441,15 +442,11 @@ static int dev_set_hwtstamp(struct net_device *dev, struct ifreq *ifr) int generic_hwtstamp_get_lower(struct net_device *dev, struct kernel_hwtstamp_config *kernel_cfg) { - const struct net_device_ops *ops = dev->netdev_ops; int err; if (!netif_device_present(dev)) return -ENODEV; - if (!ops->ndo_hwtstamp_get) - return -EOPNOTSUPP; - netdev_lock_ops(dev); err = dev_get_hwtstamp_phylib(dev, kernel_cfg); netdev_unlock_ops(dev); @@ -462,15 +459,11 @@ int generic_hwtstamp_set_lower(struct net_device *dev, struct kernel_hwtstamp_config *kernel_cfg, struct netlink_ext_ack *extack) { - const struct net_device_ops *ops = dev->netdev_ops; int err; if (!netif_device_present(dev)) return -ENODEV; - if (!ops->ndo_hwtstamp_set) - return -EOPNOTSUPP; - netdev_lock_ops(dev); err = dev_set_hwtstamp_phylib(dev, kernel_cfg, extack); netdev_unlock_ops(dev); From 7cce782d8327b7291334c4a304cf3fd909a74d9d Mon Sep 17 00:00:00 2001 From: Ivan Vecera Date: Thu, 17 Sep 2026 16:37:36 +0200 Subject: [PATCH 311/694] dpll: use exact lookup for reference sync pin id dpll_pin_ref_sync_state_set() looks up the reference sync pin in the pin->ref_sync_pins xarray, which is keyed by the sync pin's id (see dpll_pin_ref_sync_pair_add() using xa_insert() with ref_sync_pin->id). The pin id to operate on is supplied by userspace via DPLL_A_PIN_ID. The lookup however used xa_find() with a ULONG_MAX limit, which returns the first present entry with an index greater than or equal to the requested id, not the entry stored exactly at that id. If userspace passes an id that is not paired as a reference sync pin, but another pin with a higher id is present in the xarray, xa_find() silently returns that wrong pin and the subsequent ref_sync_set() operates on it. The request only fails when the given id is larger than every present key. Use xa_load() for an exact-key lookup instead, mirroring the deletion path in dpll_pin_ref_sync_pair_del(). Fixes: 58256a26bfb3 ("dpll: add reference sync get/set") Signed-off-by: Ivan Vecera Link: https://patch.msgid.link/20260917143736.526221-1-ivecera@redhat.com Signed-off-by: Jakub Kicinski --- drivers/dpll/dpll_netlink.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/drivers/dpll/dpll_netlink.c b/drivers/dpll/dpll_netlink.c index 45365214fbef..fb24fd53f2e1 100644 --- a/drivers/dpll/dpll_netlink.c +++ b/drivers/dpll/dpll_netlink.c @@ -1210,8 +1210,7 @@ dpll_pin_ref_sync_state_set(struct dpll_pin *pin, struct dpll_device *dpll; int ret; - ref_sync_pin = xa_find(&pin->ref_sync_pins, &ref_sync_pin_idx, - ULONG_MAX, XA_PRESENT); + ref_sync_pin = xa_load(&pin->ref_sync_pins, ref_sync_pin_idx); if (!ref_sync_pin) { NL_SET_ERR_MSG(extack, "reference sync pin not found"); return -EINVAL; From c06bde80ae7a7b595732f7cabcb92cf08db9d56a Mon Sep 17 00:00:00 2001 From: Pengpeng Hou Date: Sun, 20 Sep 2026 11:47:45 +0800 Subject: [PATCH 312/694] net: usb: sr9700: include receive overhead in the length check The receive fixup subtracts the Ethernet CRC from the reported packet length, but compares that payload length against the whole remaining receive buffer. The following copy starts after the three-byte header, and the cursor advance consumes both that header and the four-byte CRC. Require the payload to fit after SR_RX_OVERHEAD before copying it or advancing to the next packet. The loop already ensures that the remaining buffer is larger than the overhead, so the subtraction is safe. The issue was found by our static-analysis tool. Fixes: c9b37458e956 ("USB2NET : SR9700 : One chip USB 1.1 USB2NET SR9700Device Driver Support") Reviewed-by: Ethan Nelson-Moore Tested-by: Ethan Nelson-Moore Signed-off-by: Pengpeng Hou Link: https://patch.msgid.link/20260920034745.18468-1-hppiscas@163.com Signed-off-by: Jakub Kicinski --- drivers/net/usb/sr9700.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/net/usb/sr9700.c b/drivers/net/usb/sr9700.c index 937e6fef3ac6..50981a28376a 100644 --- a/drivers/net/usb/sr9700.c +++ b/drivers/net/usb/sr9700.c @@ -355,7 +355,8 @@ static int sr9700_rx_fixup(struct usbnet *dev, struct sk_buff *skb) /* ignore the CRC length */ len = (skb->data[1] | (skb->data[2] << 8)) - 4; - if (len > ETH_FRAME_LEN || len > skb->len || len < 0) + if (len > ETH_FRAME_LEN || len < 0 || + len > skb->len - SR_RX_OVERHEAD) return 0; /* the last packet of current skb */ From be581d6635579489eadff9cea4caee142ec6d8bc Mon Sep 17 00:00:00 2001 From: Yuya Kusakabe Date: Fri, 18 Sep 2026 22:49:30 +0900 Subject: [PATCH 313/694] selftests: net: fix CONFIG_SYSCTL sort order in configs Commit 8d75c338f0bc ("sysctl: remove CONFIG_PROC_SYSCTL, it just mirrors CONFIG_SYSCTL") renamed CONFIG_PROC_SYSCTL to CONFIG_SYSCTL in place, which left the entry out of alphabetical order in the net and packetdrill configs. The netdev CI check for sorted selftest configs now fails for every patch that touches either file. Fixes: 8d75c338f0bc ("sysctl: remove CONFIG_PROC_SYSCTL, it just mirrors CONFIG_SYSCTL") Signed-off-by: Yuya Kusakabe Reviewed-by: Joel Granados Link: https://patch.msgid.link/20260918-selftests-net-config-sort-v1-1-968ea6e8c1b7@gmail.com Signed-off-by: Jakub Kicinski --- tools/testing/selftests/net/config | 2 +- tools/testing/selftests/net/packetdrill/config | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/tools/testing/selftests/net/config b/tools/testing/selftests/net/config index 30d5fcb09a83..737e7e6327b3 100644 --- a/tools/testing/selftests/net/config +++ b/tools/testing/selftests/net/config @@ -118,10 +118,10 @@ CONFIG_NFT_NAT=m CONFIG_NUMA=y CONFIG_OPENVSWITCH=m CONFIG_PAGE_POOL_STATS=y -CONFIG_SYSCTL=y CONFIG_PSAMPLE=m CONFIG_RPS=y CONFIG_SYN_COOKIES=y +CONFIG_SYSCTL=y CONFIG_SYSFS=y CONFIG_TAP=m CONFIG_TCP_CONG_DCTCP=y diff --git a/tools/testing/selftests/net/packetdrill/config b/tools/testing/selftests/net/packetdrill/config index 83dde525c53c..b7df8bf30920 100644 --- a/tools/testing/selftests/net/packetdrill/config +++ b/tools/testing/selftests/net/packetdrill/config @@ -4,8 +4,8 @@ CONFIG_IPV6=y CONFIG_NET_NS=y CONFIG_NET_SCH_FIFO=y CONFIG_NET_SCH_FQ=y -CONFIG_SYSCTL=y CONFIG_SYN_COOKIES=y +CONFIG_SYSCTL=y CONFIG_TCP_CONG_CUBIC=y CONFIG_TCP_MD5SIG=y CONFIG_TUN=y From 10de7ed8ef4840da9ca21de4c29578657ac367db Mon Sep 17 00:00:00 2001 From: Andrea Parri Date: Thu, 17 Sep 2026 13:55:42 +0200 Subject: [PATCH 314/694] net/mlx5e: fix swapped IPv6 IPsec policy masks IPv6 XFRM policies may use different source and destination prefix lengths. mlx5e_ipsec_policy_mask() builds the corresponding masks independently, but setup_fte_addr6() installs each mask in the opposite address field. When the prefix lengths differ, this makes the source match use the destination prefix and the destination match use the source prefix. The resulting hardware rule can both miss traffic covered by the policy and match traffic outside it. Install each mask in its corresponding match field. Fixes: ca7992f52c2c ("net/mlx5e: Properly match IPsec subnet addresses") Cc: stable@vger.kernel.org Signed-off-by: Andrea Parri Reviewed-by: Tariq Toukan Link: https://patch.msgid.link/20260917115542.177675-1-parri.andrea@gmail.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_fs.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_fs.c b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_fs.c index 329608c59313..8ffa8068e90a 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_fs.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_fs.c @@ -1564,14 +1564,14 @@ static void setup_fte_addr6(struct mlx5_flow_spec *spec, memcpy(MLX5_ADDR_OF(fte_match_param, spec->match_value, outer_headers.src_ipv4_src_ipv6.ipv6_layout.ipv6), saddr, 16); memcpy(MLX5_ADDR_OF(fte_match_param, spec->match_criteria, - outer_headers.src_ipv4_src_ipv6.ipv6_layout.ipv6), dmask, 16); + outer_headers.src_ipv4_src_ipv6.ipv6_layout.ipv6), smask, 16); } if (!addr6_all_zero(daddr)) { memcpy(MLX5_ADDR_OF(fte_match_param, spec->match_value, outer_headers.dst_ipv4_dst_ipv6.ipv6_layout.ipv6), daddr, 16); memcpy(MLX5_ADDR_OF(fte_match_param, spec->match_criteria, - outer_headers.dst_ipv4_dst_ipv6.ipv6_layout.ipv6), smask, 16); + outer_headers.dst_ipv4_dst_ipv6.ipv6_layout.ipv6), dmask, 16); } } From 67f4c1c6a1b51e203d986779299824d1c2c590a6 Mon Sep 17 00:00:00 2001 From: Zihan Xi Date: Wed, 16 Sep 2026 15:29:24 +0000 Subject: [PATCH 315/694] smb: client: fix create context out-of-bounds reads smb2_parse_contexts() validates the complete create-context area but does not limit each record to its Next field before dispatching it. A malformed chain can therefore expose bytes beyond the current context to a handler. The QFid handler also used a full response-structure cast although it only reads DiskFileId. The SMB2/SMB3 lease parsers made the same layout assumption: they read LeaseState and LeaseFlags at canonical offsets rather than at DataOffset. A valid non-canonical DataOffset could therefore yield unrelated in-bounds data, while a short DataLength was still accepted. Limit each context to its Next value, reject offsets before the context header, and reject malformed chains. Bound the name range by the current context and do not dispatch a known handler when DataLength is zero. Read the QFid DiskFileId only when the context data covers that field. Parse the lease context from DataOffset and require DataLength to match the v1 or v2 lease_context size used by ksmbd. A size mismatch skips lease parsing without failing the open. Fixes: b8c32dbb0deb ("CIFS: Request SMB2.1 leases") Fixes: f047390a097e ("CIFS: Add create lease v2 context for SMB3") Fixes: 89a5bfa350fa ("smb3: optimize open to not send query file internal info") Cc: stable@vger.kernel.org Reported-by: Vega Assisted-by: LLM Co-developed-by: Luxing Yin Signed-off-by: Luxing Yin Signed-off-by: Zihan Xi Tested-by: Frank Sorenson Signed-off-by: Paulo Alcantara --- fs/smb/client/smb2ops.c | 28 ++++++++++++++++++-------- fs/smb/client/smb2pdu.c | 44 +++++++++++++++++++++++++++++++---------- 2 files changed, 54 insertions(+), 18 deletions(-) diff --git a/fs/smb/client/smb2ops.c b/fs/smb/client/smb2ops.c index 3464470d3297..aa142420dae2 100644 --- a/fs/smb/client/smb2ops.c +++ b/fs/smb/client/smb2ops.c @@ -4572,25 +4572,37 @@ smb3_create_lease_buf(u8 *lease_key, u8 oplock, u8 *parent_lease_key, __le32 fla static __u8 smb2_parse_lease_buf(void *buf, __u16 *epoch, char *lease_key) { - struct create_lease *lc = (struct create_lease *)buf; + struct create_context *cc = buf; + struct lease_context lc; *epoch = 0; /* not used */ - if (lc->lcontext.LeaseFlags & SMB2_LEASE_FLAG_BREAK_IN_PROGRESS_LE) + if (le32_to_cpu(cc->DataLength) != sizeof(lc)) + return 0; + + memcpy(&lc, (u8 *)cc + le16_to_cpu(cc->DataOffset), sizeof(lc)); + if (lc.LeaseFlags & SMB2_LEASE_FLAG_BREAK_IN_PROGRESS_LE) return SMB2_OPLOCK_LEVEL_NOCHANGE; - return le32_to_cpu(lc->lcontext.LeaseState); + return le32_to_cpu(lc.LeaseState); } static __u8 smb3_parse_lease_buf(void *buf, __u16 *epoch, char *lease_key) { - struct create_lease_v2 *lc = (struct create_lease_v2 *)buf; + struct create_context *cc = buf; + struct lease_context_v2 lc; - *epoch = le16_to_cpu(lc->lcontext.Epoch); - if (lc->lcontext.LeaseFlags & SMB2_LEASE_FLAG_BREAK_IN_PROGRESS_LE) + if (le32_to_cpu(cc->DataLength) != sizeof(lc)) { + *epoch = 0; + return 0; + } + + memcpy(&lc, (u8 *)cc + le16_to_cpu(cc->DataOffset), sizeof(lc)); + *epoch = le16_to_cpu(lc.Epoch); + if (lc.LeaseFlags & SMB2_LEASE_FLAG_BREAK_IN_PROGRESS_LE) return SMB2_OPLOCK_LEVEL_NOCHANGE; if (lease_key) - memcpy(lease_key, &lc->lcontext.LeaseKey, SMB2_LEASE_KEY_SIZE); - return le32_to_cpu(lc->lcontext.LeaseState); + memcpy(lease_key, lc.LeaseKey, SMB2_LEASE_KEY_SIZE); + return le32_to_cpu(lc.LeaseState); } static unsigned int diff --git a/fs/smb/client/smb2pdu.c b/fs/smb/client/smb2pdu.c index 880ce12f50c4..4046500dbe93 100644 --- a/fs/smb/client/smb2pdu.c +++ b/fs/smb/client/smb2pdu.c @@ -2379,11 +2379,17 @@ create_reconnect_durable_buf(struct cifs_fid *fid) static void parse_query_id_ctxt(struct create_context *cc, struct smb2_file_all_info *buf) { - struct create_disk_id_rsp *pdisk_id = (struct create_disk_id_rsp *)cc; + u16 doff = le16_to_cpu(cc->DataOffset); + u32 dlen = le32_to_cpu(cc->DataLength); + u8 *beg; - cifs_dbg(FYI, "parse query id context 0x%llx 0x%llx\n", - pdisk_id->DiskFileId, pdisk_id->VolumeId); - buf->IndexNumber = pdisk_id->DiskFileId; + if (dlen < sizeof(__le64)) + return; + + beg = (u8 *)cc + doff; + memcpy(&buf->IndexNumber, beg, sizeof(__le64)); + cifs_dbg(FYI, "parse query id context 0x%llx\n", + le64_to_cpu(buf->IndexNumber)); } static void @@ -2431,6 +2437,7 @@ int smb2_parse_contexts(struct TCP_Server_Info *server, struct smb2_create_rsp *rsp = rsp_iov->iov_base; struct create_context *cc; size_t rem, off, len; + size_t cc_len; size_t doff, dlen; size_t noff, nlen; char *name; @@ -2453,29 +2460,41 @@ int smb2_parse_contexts(struct TCP_Server_Info *server, buf->IndexNumber = 0; while (rem >= sizeof(*cc)) { + off = le32_to_cpu(cc->Next); + if (off) { + if ((off & 0x7) || off >= rem || off < sizeof(*cc)) + return -EINVAL; + cc_len = off; + } else { + cc_len = rem; + } + doff = le16_to_cpu(cc->DataOffset); dlen = le32_to_cpu(cc->DataLength); - if (check_add_overflow(doff, dlen, &len) || len > rem) + if (doff < sizeof(*cc) || + check_add_overflow(doff, dlen, &len) || len > cc_len) return -EINVAL; noff = le16_to_cpu(cc->NameOffset); nlen = le16_to_cpu(cc->NameLength); - if (noff + nlen > doff) + if (noff < sizeof(*cc) || + check_add_overflow(noff, nlen, &len) || len > cc_len || + (dlen && len > doff)) return -EINVAL; name = (char *)cc + noff; switch (nlen) { case 4: - if (!strncmp(name, SMB2_CREATE_REQUEST_LEASE, 4)) { + if (dlen && !strncmp(name, SMB2_CREATE_REQUEST_LEASE, 4)) { *oplock = server->ops->parse_lease_buf(cc, epoch, lease_key); - } else if (buf && + } else if (dlen && buf && !strncmp(name, SMB2_CREATE_QUERY_ON_DISK_ID, 4)) { parse_query_id_ctxt(cc, buf); } break; case 16: - if (posix && !memcmp(name, smb3_create_tag_posix, 16)) + if (dlen && posix && !memcmp(name, smb3_create_tag_posix, 16)) parse_posix_ctxt(cc, buf, posix); break; default: @@ -2487,13 +2506,18 @@ int smb2_parse_contexts(struct TCP_Server_Info *server, } off = le32_to_cpu(cc->Next); - if (!off) + if (!off) { + rem = 0; break; + } if (check_sub_overflow(rem, off, &rem)) return -EINVAL; cc = (struct create_context *)((u8 *)cc + off); } + if (rem) + return -EINVAL; + if (rsp->OplockLevel != SMB2_OPLOCK_LEVEL_LEASE) *oplock = rsp->OplockLevel; From fa2e9900dd2a3f5a1e7ef5a8c5e8d435feedbfcc Mon Sep 17 00:00:00 2001 From: Zihan Xi Date: Wed, 16 Sep 2026 15:29:25 +0000 Subject: [PATCH 316/694] smb: client: validate POSIX create context length parse_posix_ctxt() reads the fixed nlink, reparse_tag, and mode fields before checking that the POSIX create context contains them. A short context can pass the generic checks and still make these fixed-width reads run past its declared data. The current in-tree smb2_open_file() path passes a NULL posix pointer, so this handler is not reached on the ordinary open path. Still require the POSIX data to cover all three fields before reading them because the helper performs those unguarded reads. Keep the existing soft-failure behavior so malformed optional metadata does not fail the open. Fixes: 69dda3059e7a ("cifs: add SMB2_open() arg to return POSIX data") Cc: stable@vger.kernel.org Reported-by: Vega Assisted-by: LLM Co-developed-by: Luxing Yin Signed-off-by: Luxing Yin Signed-off-by: Zihan Xi Tested-by: Frank Sorenson Signed-off-by: Paulo Alcantara --- fs/smb/client/smb2pdu.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/fs/smb/client/smb2pdu.c b/fs/smb/client/smb2pdu.c index 4046500dbe93..538d708b0404 100644 --- a/fs/smb/client/smb2pdu.c +++ b/fs/smb/client/smb2pdu.c @@ -2396,12 +2396,15 @@ static void parse_posix_ctxt(struct create_context *cc, struct smb2_file_all_info *info, struct create_posix_rsp *posix) { - int sid_len; u8 *beg = (u8 *)cc + le16_to_cpu(cc->DataOffset); - u8 *end = beg + le32_to_cpu(cc->DataLength); + u32 dlen = le32_to_cpu(cc->DataLength); + u8 *end = beg + dlen; + int sid_len; u8 *sid; memset(posix, 0, sizeof(*posix)); + if (dlen < 3 * sizeof(__le32)) + return; posix->nlink = get_unaligned_le32(beg); posix->reparse_tag = get_unaligned_le32(beg + 4); From 566820af017e81497fb5e9d3ad6e7ffe2828bc8b Mon Sep 17 00:00:00 2001 From: Zihan Xi Date: Wed, 16 Sep 2026 15:29:26 +0000 Subject: [PATCH 317/694] smb: client: close handle after create-context parsing failure SMB2_open() accounts a successful CREATE response as a remote open before parsing its create contexts. If smb2_parse_contexts() rejects malformed context data, SMB2_open() returns without closing the handle, leaving the server-side handle open and num_remote_opens elevated. Close the handle after a post-CREATE context parsing failure so the error path releases the remote resource and balances the open count. Fixes: af1689a9b770 ("smb: client: fix potential OOBs in smb2_parse_contexts()") Cc: stable@vger.kernel.org Reported-by: Vega Assisted-by: LLM Co-developed-by: Luxing Yin Signed-off-by: Luxing Yin Signed-off-by: Zihan Xi Tested-by: Frank Sorenson Signed-off-by: Paulo Alcantara --- fs/smb/client/smb2pdu.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/fs/smb/client/smb2pdu.c b/fs/smb/client/smb2pdu.c index 538d708b0404..3d7ead36d1a0 100644 --- a/fs/smb/client/smb2pdu.c +++ b/fs/smb/client/smb2pdu.c @@ -3416,6 +3416,9 @@ SMB2_open(const unsigned int xid, struct cifs_open_parms *oparms, __le16 *path, rc = smb2_parse_contexts(server, &rsp_iov, &oparms->fid->epoch, oparms->fid->lease_key, oplock, file_info, posix); + if (rc) + SMB2_close(xid, tcon, oparms->fid->persistent_fid, + oparms->fid->volatile_fid); trace_smb3_open_done(xid, rsp->PersistentFileId, tcon->tid, ses->Suid, oparms->create_options, oparms->desired_access, From d2ff5fb93ea83034025850266b5eed391f96b825 Mon Sep 17 00:00:00 2001 From: Zihan Xi Date: Wed, 16 Sep 2026 15:29:27 +0000 Subject: [PATCH 318/694] smb: client: clean up failed cached directory opens open_cached_dir() sends CREATE and QUERY_INFO as a compound request. If the CREATE succeeds but a later command returns an error, the function must retain the CREATE FID so common cleanup can issue SMB2_close(). It also must not treat a response error as a valid CREATE. Validate the CREATE response before using its fields, record the FIDs, and mark the handle open before handling errors from later compound commands. Move the -EREMCHG reconnect handling before response validation so a missing response does not hide the reconnect request. Count the handle when it is marked open; confirmed close responses decrement the counter, while existing close retry behavior remains best effort on transport failures. Fixes: b0f6df737a1c ("cifs: cache FILE_ALL_INFO for the shared root handle") Cc: stable@vger.kernel.org Reported-by: Vega Assisted-by: LLM Co-developed-by: Luxing Yin Signed-off-by: Luxing Yin Signed-off-by: Zihan Xi Tested-by: Frank Sorenson Signed-off-by: Paulo Alcantara --- fs/smb/client/cached_dir.c | 32 ++++++++++++++++++++++---------- 1 file changed, 22 insertions(+), 10 deletions(-) diff --git a/fs/smb/client/cached_dir.c b/fs/smb/client/cached_dir.c index 88d5e9a32f28..647fa26da4d2 100644 --- a/fs/smb/client/cached_dir.c +++ b/fs/smb/client/cached_dir.c @@ -8,6 +8,7 @@ #include #include "cifsglob.h" #include "cifsproto.h" +#include "../common/smb2status.h" #include "cifs_debug.h" #include "smb2proto.h" #include "cached_dir.h" @@ -323,25 +324,37 @@ int open_cached_dir(unsigned int xid, struct cifs_tcon *tcon, rc = compound_send_recv(xid, ses, server, flags, 2, rqst, resp_buftype, rsp_iov); - if (rc) { - if (rc == -EREMCHG) { - tcon->need_reconnect = true; - pr_warn_once("server share %s deleted\n", - tcon->tree_name); - } + if (rc == -EREMCHG) { + tcon->need_reconnect = true; + pr_warn_once("server share %s deleted\n", + tcon->tree_name); + } + + if (!rsp_iov[0].iov_base || rsp_iov[0].iov_len < sizeof(*o_rsp)) { + if (!rc) + rc = -EIO; goto oshr_free; } - cfid->is_open = true; - - spin_lock(&cfids->cfid_list_lock); o_rsp = (struct smb2_create_rsp *)rsp_iov[0].iov_base; + if (o_rsp->hdr.Status != STATUS_SUCCESS) { + if (!rc) + rc = -EIO; + goto oshr_free; + } + oparms.fid->persistent_fid = o_rsp->PersistentFileId; oparms.fid->volatile_fid = o_rsp->VolatileFileId; #ifdef CONFIG_CIFS_DEBUG2 oparms.fid->mid = le64_to_cpu(o_rsp->hdr.MessageId); #endif /* CIFS_DEBUG2 */ + cfid->is_open = true; + atomic_inc(&tcon->num_remote_opens); + if (rc) + goto oshr_free; + + spin_lock(&cfids->cfid_list_lock); if (o_rsp->OplockLevel != SMB2_OPLOCK_LEVEL_LEASE) { spin_unlock(&cfids->cfid_list_lock); @@ -408,7 +421,6 @@ int open_cached_dir(unsigned int xid, struct cifs_tcon *tcon, close_cached_dir(cfid); } else { *ret_cfid = cfid; - atomic_inc(&tcon->num_remote_opens); } kfree(utf16_path); From 6c5c547f037bc18f0b8d0b5db5a648f8f630ce85 Mon Sep 17 00:00:00 2001 From: Zihan Xi Date: Wed, 16 Sep 2026 15:29:28 +0000 Subject: [PATCH 319/694] smb: client: close completed creates on compound wait errors compound_send_recv() waits for responses in order. If a later wait is interrupted, or if a later MID fails during response synchronization, an earlier CREATE may already have opened a remote handle. The earlier mid is then released without invoking handle_cancelled_mid(), leaving the remote handle open because no FID was copied to the caller. Mark completed earlier mids as cancelled when a compound wait or MID synchronization aborts. Keep their response buffers attached while the MIDs are synchronized, and transfer them only after synchronization of the processed responses, so the release path can inspect successful CREATE responses and queue SMB2_close() after a later failure. Account for a remote open only after the close work is allocated and before it is queued, since the caller has not yet updated num_remote_opens. Mark the create+close compound used by smb2_unlink() so it is not closed again. Non-CREATE responses and compounds that already include a close keep their existing behavior. Fixes: e0bba0b85481 ("cifs: add compound_send_recv()") Cc: stable@vger.kernel.org Reported-by: Vega Assisted-by: LLM Co-developed-by: Luxing Yin Signed-off-by: Luxing Yin Signed-off-by: Zihan Xi Tested-by: Frank Sorenson Signed-off-by: Paulo Alcantara --- fs/smb/client/smb2inode.c | 2 +- fs/smb/client/smb2misc.c | 9 ++++-- fs/smb/client/transport.c | 67 +++++++++++++++++++++++++++++++-------- 3 files changed, 61 insertions(+), 17 deletions(-) diff --git a/fs/smb/client/smb2inode.c b/fs/smb/client/smb2inode.c index 13fe8e3b48f3..ecd7a65cab08 100644 --- a/fs/smb/client/smb2inode.c +++ b/fs/smb/client/smb2inode.c @@ -1121,7 +1121,7 @@ smb2_unlink(const unsigned int xid, struct cifs_tcon *tcon, const char *name, struct kvec close_iov; int resp_buftype[2]; struct cifs_fid fid; - int flags = 0; + int flags = CIFS_CP_CREATE_CLOSE_OP; __u8 oplock; int rc; diff --git a/fs/smb/client/smb2misc.c b/fs/smb/client/smb2misc.c index 0cfe60ae42c3..5e5cf92d1eb3 100644 --- a/fs/smb/client/smb2misc.c +++ b/fs/smb/client/smb2misc.c @@ -834,7 +834,8 @@ smb2_cancelled_close_fid(struct work_struct *work) */ static int __smb2_handle_cancelled_cmd(struct cifs_tcon *tcon, __u16 cmd, __u64 mid, - __u64 persistent_fid, __u64 volatile_fid) + __u64 persistent_fid, __u64 volatile_fid, + bool account_remote_open) { struct close_cancelled_open *cancelled; @@ -848,6 +849,8 @@ __smb2_handle_cancelled_cmd(struct cifs_tcon *tcon, __u16 cmd, __u64 mid, cancelled->cmd = cmd; cancelled->mid = mid; INIT_WORK(&cancelled->work, smb2_cancelled_close_fid); + if (account_remote_open) + atomic_inc(&tcon->num_remote_opens); WARN_ON(queue_work(cifsiod_wq, &cancelled->work) == false); return 0; @@ -884,7 +887,7 @@ smb2_handle_cancelled_close(struct cifs_tcon *tcon, __u64 persistent_fid, spin_unlock(&tcon->tc_lock); rc = __smb2_handle_cancelled_cmd(tcon, SMB2_CLOSE_HE, 0, - persistent_fid, volatile_fid); + persistent_fid, volatile_fid, false); if (rc) cifs_put_tcon(tcon, netfs_trace_tcon_ref_put_cancelled_close); @@ -912,7 +915,7 @@ smb2_handle_cancelled_mid(struct mid_q_entry *mid, struct TCP_Server_Info *serve le16_to_cpu(hdr->Command), le64_to_cpu(hdr->MessageId), rsp->PersistentFileId, - rsp->VolatileFileId); + rsp->VolatileFileId, true); if (rc) cifs_put_tcon(tcon, netfs_trace_tcon_ref_put_cancelled_mid); diff --git a/fs/smb/client/transport.c b/fs/smb/client/transport.c index 7df5b3447aea..6e21b5f8754a 100644 --- a/fs/smb/client/transport.c +++ b/fs/smb/client/transport.c @@ -805,6 +805,18 @@ cifs_cancelled_callback(struct TCP_Server_Info *server, struct mid_q_entry *mid) release_mid(server, mid); } +static void +cifs_mark_compound_mids_cancelled(struct mid_q_entry **mid, int count) +{ + int i; + + for (i = 0; i < count; i++) { + spin_lock(&mid[i]->mid_lock); + mid[i]->wait_cancelled = true; + spin_unlock(&mid[i]->mid_lock); + } +} + /* * cifs_pick_channel - pick an eligible channel for network operations * @@ -865,6 +877,7 @@ compound_send_recv(const unsigned int xid, struct cifs_ses *ses, int *resp_buf_type, struct kvec *resp_iov) { int i, j, optype, rc = 0; + int num_processed = 0; struct mid_q_entry *mid[MAX_COMPOUND]; bool cancelled_mid[MAX_COMPOUND] = {false}; struct cifs_credits credits[MAX_COMPOUND] = { @@ -1015,6 +1028,14 @@ compound_send_recv(const unsigned int xid, struct cifs_ses *ses, break; } if (rc != 0) { + /* + * A completed CREATE earlier in the compound chain may have + * opened a remote handle even though a later wait was + * interrupted. Mark it cancelled so __release_mid() invokes + * the existing unmatched-open cleanup. + */ + cifs_mark_compound_mids_cancelled(mid, i); + for (; i < num_rqst; i++) { cifs_server_dbg(FYI, "Cancelling wait for mid %llu cmd: %d\n", mid[i]->mid, le16_to_cpu(mid[i]->command)); @@ -1037,6 +1058,14 @@ compound_send_recv(const unsigned int xid, struct cifs_ses *ses, rc = cifs_sync_mid_result(mid[i], server); if (rc != 0) { + /* + * A previous CREATE may have completed before this + * response failed. Mark it cancelled so its remote + * handle is closed when the mid is released. + */ + cifs_mark_compound_mids_cancelled(mid, i); + /* Keep their response buffers for cancelled-mid cleanup. */ + num_processed = 0; /* mark this mid as cancelled to not free it below */ cancelled_mid[i] = true; goto out; @@ -1046,13 +1075,24 @@ compound_send_recv(const unsigned int xid, struct cifs_ses *ses, mid[i]->mid_state != MID_RESPONSE_READY) { rc = smb_EIO1(smb_eio_trace_rx_mid_unready, mid[i]->mid_state); cifs_dbg(FYI, "Bad MID state?\n"); + cifs_mark_compound_mids_cancelled(mid, i); + num_processed = 0; goto out; } rc = server->ops->check_receive(mid[i], server, flags & CIFS_LOG_ERROR); + num_processed = i + 1; + } - if (resp_iov) { +out: + /* + * Delay moving response buffers out of their mids until response + * synchronization completes. This lets cancelled-mid cleanup inspect + * an earlier CREATE response if a later MID fails. + */ + if (resp_iov) { + for (i = 0; i < num_processed; i++) { buf = (char *)mid[i]->resp_buf; resp_iov[i].iov_base = buf; resp_iov[i].iov_len = mid[i]->resp_buf_size; @@ -1071,21 +1111,22 @@ compound_send_recv(const unsigned int xid, struct cifs_ses *ses, /* * Compounding is never used during session establish. */ - spin_lock(&ses->ses_lock); - if ((ses->ses_status == SES_NEW) || (optype & CIFS_NEG_OP) || (optype & CIFS_SESS_OP)) { - struct kvec iov = { - .iov_base = resp_iov[0].iov_base, - .iov_len = resp_iov[0].iov_len - }; - spin_unlock(&ses->ses_lock); - cifs_server_lock(server); - smb311_update_preauth_hash(ses, server, &iov, 1); - cifs_server_unlock(server); + if (num_processed == num_rqst) { spin_lock(&ses->ses_lock); + if ((ses->ses_status == SES_NEW) || (optype & CIFS_NEG_OP) || (optype & CIFS_SESS_OP)) { + struct kvec iov = { + .iov_base = resp_iov[0].iov_base, + .iov_len = resp_iov[0].iov_len + }; + spin_unlock(&ses->ses_lock); + cifs_server_lock(server); + smb311_update_preauth_hash(ses, server, &iov, 1); + cifs_server_unlock(server); + spin_lock(&ses->ses_lock); + } + spin_unlock(&ses->ses_lock); } - spin_unlock(&ses->ses_lock); -out: /* * This will dequeue all mids. After this it is important that the * demultiplex_thread will not process any of these mids any further. From 2e828035d5d736d904c238ae7ec3f77c0d6270bf Mon Sep 17 00:00:00 2001 From: Zihan Xi Date: Wed, 16 Sep 2026 15:29:29 +0000 Subject: [PATCH 320/694] smb: client: preserve create-context parsing errors smb2_compound_op() saves the result from compound_send_recv() in tmp_rc. For SMB2_OP_OPEN_QUERY it then parses the CREATE contexts, but the final assignment of rc from tmp_rc discards a parsing error. A malformed create-context response can therefore be reported as successful to smb2_query_path_info(). Keep a create-context parsing error in tmp_rc so it survives per-command response processing and is returned to the caller. Fixes: b07687edee99 ("cifs: Improve SMB2+ stat() to work also without FILE_READ_ATTRIBUTES") Cc: stable@vger.kernel.org Reported-by: Vega Assisted-by: LLM Co-developed-by: Luxing Yin Signed-off-by: Luxing Yin Signed-off-by: Zihan Xi Tested-by: Frank Sorenson Signed-off-by: Paulo Alcantara --- fs/smb/client/smb2inode.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/fs/smb/client/smb2inode.c b/fs/smb/client/smb2inode.c index ecd7a65cab08..f46a62eae659 100644 --- a/fs/smb/client/smb2inode.c +++ b/fs/smb/client/smb2inode.c @@ -598,8 +598,10 @@ static int smb2_compound_op(const unsigned int xid, struct cifs_tcon *tcon, /* smb2_parse_contexts() fills idata->fi.IndexNumber */ rc = smb2_parse_contexts(server, &rsp_iov[0], &oparms->fid->epoch, oparms->fid->lease_key, &oplock, &idata->fi, NULL); - if (rc) + if (rc) { cifs_dbg(VFS, "rc: %d parsing context of compound op\n", rc); + tmp_rc = rc; + } } for (i = 0; i < num_cmds; i++) { From be31fe6333f534155e6b408f1ef6d77974bb41aa Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Sun, 20 Sep 2026 19:14:32 +0000 Subject: [PATCH 321/694] ipv6: Fix dst leak for uncached routes. ip6_route_output_flags(), ip6_rt_put_flags(), and ip6_dst_check() detect an uncached route by list_empty(&rt->dst.rt_uncached), which replaced the static DST_NOCACHE flag check in commit a4c2fd7f7891 ("net: remove DST_NOCACHE flag"). When a device is unregistered, rt6_uncached_list_flush_dev() unlinks uncached routes tied to the device from rt6_uncached_list. Previously, they were moved to another list with list_move() (__list_del_entry() + list_add()), and since commit 98aa546af5e4 ("inet: remove (struct uncached_list)->quarantine"), the routes are just unlinked with list_del_init(). If list_del_init() runs concurrently, list_empty() evaluates to true; ip6_route_output_flags() calls dst_hold_safe() incorrectly and ip6_rt_put_flags() skips ip6_rt_put(), leaking dst, and thus dev tied via rt->from as well. The same race is partially fixed by commit 9a6f0c4d5796 ("dst: fix races in rt6_uncached_list_del() and rt_del_uncached_list()"). Let's check rt6->dst.rt_uncached_list instead. Note that IPv4 does not have the same issue. Fixes: 98aa546af5e4 ("inet: remove (struct uncached_list)->quarantine") Signed-off-by: Kuniyuki Iwashima Reviewed-by: Hangbin Liu Reviewed-by: Xuanqiang Luo Reviewed-by: Ido Schimmel Reviewed-by: Eric Dumazet Link: https://patch.msgid.link/20260920191558.2990636-1-kuniyu@google.com Signed-off-by: Jakub Kicinski --- include/net/ip6_route.h | 4 ++-- net/ipv6/route.c | 7 ++++--- 2 files changed, 6 insertions(+), 5 deletions(-) diff --git a/include/net/ip6_route.h b/include/net/ip6_route.h index b9e8d2b759e9..0f9b7a260d25 100644 --- a/include/net/ip6_route.h +++ b/include/net/ip6_route.h @@ -101,12 +101,12 @@ static inline struct dst_entry *ip6_route_output(struct net *net, } /* Only conditionally release dst if flags indicates - * !RT6_LOOKUP_F_DST_NOREF or dst is in uncached_list. + * !RT6_LOOKUP_F_DST_NOREF or dst is uncached. */ static inline void ip6_rt_put_flags(struct rt6_info *rt, int flags) { if (!(flags & RT6_LOOKUP_F_DST_NOREF) || - !list_empty(&rt->dst.rt_uncached)) + rt->dst.rt_uncached_list) ip6_rt_put(rt); } diff --git a/net/ipv6/route.c b/net/ipv6/route.c index 884d9ab0d50d..153ce16628c1 100644 --- a/net/ipv6/route.c +++ b/net/ipv6/route.c @@ -139,6 +139,7 @@ void rt6_uncached_list_add(struct rt6_info *rt) { struct uncached_list *ul = raw_cpu_ptr(&rt6_uncached_list); + /* Set once and never cleared: non-NULL marks an uncached route. */ rt->dst.rt_uncached_list = ul; spin_lock_bh(&ul->lock); @@ -2726,8 +2727,8 @@ struct dst_entry *ip6_route_output_flags(struct net *net, rcu_read_lock(); dst = ip6_route_output_flags_noref(net, sk, fl6, flags); rt6 = dst_rt6_info(dst); - /* For dst cached in uncached_list, refcnt is already taken. */ - if (list_empty(&rt6->dst.rt_uncached) && !dst_hold_safe(dst)) { + /* For an uncached dst, refcnt is already taken. */ + if (!rt6->dst.rt_uncached_list && !dst_hold_safe(dst)) { dst = &net->ipv6.ip6_null_entry->dst; dst_hold(dst); } @@ -2836,7 +2837,7 @@ INDIRECT_CALLABLE_SCOPE struct dst_entry *ip6_dst_check(struct dst_entry *dst, from = rcu_dereference(rt->from); if (from && (rt->rt6i_flags & RTF_PCPU || - unlikely(!list_empty(&rt->dst.rt_uncached)))) + unlikely(rt->dst.rt_uncached_list))) dst_ret = rt6_dst_from_check(rt, from, cookie); else dst_ret = rt6_check(rt, from, cookie); From 6c43c72748fffd29dec15cd1f31e9a32949bc437 Mon Sep 17 00:00:00 2001 From: Melody Wang Date: Mon, 14 Sep 2026 00:57:17 +0000 Subject: [PATCH 322/694] x86/sev: Make vTPM SVSM calls preemption-safe Two functions in the SVSM vTPM guest implementation do not disable preemption when fetching the SVSM Calling Area Address (CAA). The SVSM CAA is a per-CPU structure. When a thread is preempted and migrated to a different CPU after fetching the per-CPU CAA, the SVSM call will execute on the new CPU with the original CPU's CAA. Which is wrong. Move the CAA fetching operation inside svsm_perform_call_protocol() which disables interrupts around the SVSM call and thus runs preemption-safe. Fixes: 770de678bc28 ("x86/sev: Add SVSM vTPM probe/send_command functions") Signed-off-by: Melody Wang Signed-off-by: Borislav Petkov (AMD) Reviewed-by: Stefano Garzarella Cc: stable@vger.kernel.org Link: https://patch.msgid.link/a5bc0d4a2c462a0089109e145c21626b244b2ff0.1789345277.git.huibo.wang@amd.com --- arch/x86/coco/sev/svsm.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) diff --git a/arch/x86/coco/sev/svsm.c b/arch/x86/coco/sev/svsm.c index 916d62cd17dc..2d493d55ff2e 100644 --- a/arch/x86/coco/sev/svsm.c +++ b/arch/x86/coco/sev/svsm.c @@ -74,6 +74,14 @@ int svsm_perform_call_protocol(struct svsm_call *call) flags = native_local_irq_save(); + /* + * 'caa' is a per-CPU variable. To avoid using a stale or incorrect + * 'caa' if the task is preempted or migrated to another CPU after it + * is fetched, always fetch 'caa' and then issue the SVSM call with + * interrupts disabled. This ensures the correct 'caa' is used. + */ + call->caa = svsm_get_caa(); + ghcb = __sev_get_ghcb(&state); do { @@ -321,7 +329,6 @@ int snp_svsm_vtpm_send_command(u8 *buffer) { struct svsm_call call = {}; - call.caa = svsm_get_caa(); call.rax = SVSM_VTPM_CALL(SVSM_VTPM_CMD); call.rcx = __pa(buffer); @@ -345,7 +352,6 @@ bool snp_svsm_vtpm_probe(void) if (!snp_vmpl) return false; - call.caa = svsm_get_caa(); call.rax = SVSM_VTPM_CALL(SVSM_VTPM_QUERY); if (svsm_perform_call_protocol(&call)) From 72b782097e534ab48152dd25aaba40dda5f25f9e Mon Sep 17 00:00:00 2001 From: Jakub Pawlak Date: Mon, 14 Sep 2026 10:43:01 +0200 Subject: [PATCH 323/694] accel/ivpu: Use separate flag for job timeout Use separate flag to mark a job timeout as a reason of starting context_abort_work. This allows to distinguish engine reset reason and clearly adjust reset procedure flow. The flag is cleared in ivpu_prepare_for_reset(), which every recovery and suspend path already funnels through, so that the state is clean after recovery. Cc: stable@vger.kernel.org # v7.1+ Fixes: ade00a6c903f ("accel/ivpu: Perform engine reset instead of device recovery on TDR") Signed-off-by: Jakub Pawlak Reviewed-by: Dawid Osuchowski Signed-off-by: Karol Wachowski Link: https://patch.msgid.link/20260914084301.894028-1-karol.wachowski@linux.intel.com --- drivers/accel/ivpu/ivpu_drv.c | 3 ++- drivers/accel/ivpu/ivpu_drv.h | 2 +- drivers/accel/ivpu/ivpu_job.c | 7 +++---- drivers/accel/ivpu/ivpu_mmu.c | 1 - drivers/accel/ivpu/ivpu_pm.c | 1 + 5 files changed, 7 insertions(+), 7 deletions(-) diff --git a/drivers/accel/ivpu/ivpu_drv.c b/drivers/accel/ivpu/ivpu_drv.c index 95120957f42a..8c1c87e69f91 100644 --- a/drivers/accel/ivpu/ivpu_drv.c +++ b/drivers/accel/ivpu/ivpu_drv.c @@ -515,6 +515,7 @@ void ivpu_prepare_for_reset(struct ivpu_device *vdev) { ivpu_hw_irq_disable(vdev); disable_irq(vdev->irq); + atomic_set(&vdev->job_timeout_detected, 0); flush_work(&vdev->irq_dct_work); flush_work(&vdev->context_abort_work); flush_work(&vdev->job_destroy_work); @@ -710,7 +711,7 @@ static int ivpu_dev_init(struct ivpu_device *vdev) vdev->context_xa_limit.max = IVPU_USER_CONTEXT_MAX_SSID; atomic64_set(&vdev->unique_id_counter, 0); atomic_set(&vdev->job_timeout_counter, 0); - atomic_set(&vdev->faults_detected, 0); + atomic_set(&vdev->job_timeout_detected, 0); xa_init_flags(&vdev->context_xa, XA_FLAGS_ALLOC | XA_FLAGS_LOCK_IRQ); xa_init_flags(&vdev->submitted_jobs_xa, XA_FLAGS_ALLOC1); xa_init_flags(&vdev->db_xa, XA_FLAGS_ALLOC1); diff --git a/drivers/accel/ivpu/ivpu_drv.h b/drivers/accel/ivpu/ivpu_drv.h index 86d7c9966cac..6f4012926478 100644 --- a/drivers/accel/ivpu/ivpu_drv.h +++ b/drivers/accel/ivpu/ivpu_drv.h @@ -171,7 +171,7 @@ struct ivpu_device { struct xarray submitted_jobs_xa; struct ivpu_ipc_consumer job_done_consumer; atomic_t job_timeout_counter; - atomic_t faults_detected; + atomic_t job_timeout_detected; atomic64_t unique_id_counter; diff --git a/drivers/accel/ivpu/ivpu_job.c b/drivers/accel/ivpu/ivpu_job.c index ebb2c865b09a..4689b8ab519d 100644 --- a/drivers/accel/ivpu/ivpu_job.c +++ b/drivers/accel/ivpu/ivpu_job.c @@ -621,7 +621,6 @@ bool ivpu_job_handle_engine_error(struct ivpu_device *vdev, u32 job_id, u32 job_ * status and ensure both are handled in the same way */ job->file_priv->has_mmu_faults = true; - atomic_set(&vdev->faults_detected, 1); queue_work(system_percpu_wq, &vdev->context_abort_work); return true; } @@ -1175,10 +1174,10 @@ static int reset_engine_and_mark_faulty_contexts(struct ivpu_device *vdev) return ret; /* - * If faults are detected, ignore guilty contexts from engine reset as NPU may not be stuck - * and could return currently running good context and faulty contexts are already marked + * If job timeout is detected, read guilty context from engine reset, for other reasons + * faulty context is already known */ - if (atomic_cmpxchg(&vdev->faults_detected, 1, 0) == 1) + if (atomic_cmpxchg(&vdev->job_timeout_detected, 1, 0) == 0) return 0; num_impacted_contexts = resp.payload.engine_reset_done.num_impacted_contexts; diff --git a/drivers/accel/ivpu/ivpu_mmu.c b/drivers/accel/ivpu/ivpu_mmu.c index 41efd8985fa6..b2025274f91d 100644 --- a/drivers/accel/ivpu/ivpu_mmu.c +++ b/drivers/accel/ivpu/ivpu_mmu.c @@ -964,7 +964,6 @@ void ivpu_mmu_irq_evtq_handler(struct ivpu_device *vdev) file_priv = xa_load(&vdev->context_xa, ssid); if (file_priv) { if (!READ_ONCE(file_priv->has_mmu_faults)) { - atomic_set(&vdev->faults_detected, 1); ivpu_mmu_dump_event(vdev, event); WRITE_ONCE(file_priv->has_mmu_faults, true); } diff --git a/drivers/accel/ivpu/ivpu_pm.c b/drivers/accel/ivpu/ivpu_pm.c index c1ce8329790e..de0becbfdffb 100644 --- a/drivers/accel/ivpu/ivpu_pm.c +++ b/drivers/accel/ivpu/ivpu_pm.c @@ -229,6 +229,7 @@ static void ivpu_job_timeout_work(struct work_struct *work) ivpu_jsm_state_dump(vdev); ivpu_dev_coredump(vdev); + atomic_set(&vdev->job_timeout_detected, 1); queue_work(system_percpu_wq, &vdev->context_abort_work); } From 6bdcdb31318c10e6369cff8e5b32c37397e98091 Mon Sep 17 00:00:00 2001 From: Yuqi Xu Date: Mon, 21 Sep 2026 10:52:04 +0800 Subject: [PATCH 324/694] wifi: mac80211: validate TX status rate metadata minstrel_ht_tx_status() converts the HT/VHT rate reported in a TX status into a group and rate index and uses those to index the minstrel_ht rate tables. The existing validation only checked that an entry was present and had tries recorded; it did not check that the reported MCS, spatial stream count and bandwidth are representable by the tables, so malformed metadata passed to minstrel_ht_get_stats() or minstrel_ht_ri_get_stats() could be used as an out-of-bounds index. Validate the rate metadata in the common TX status path instead of only in minstrel_ht, since the values are used by more than rate control. The legacy ieee80211_tx_rate array and the rate_info based entries are sanitized in ieee80211_tx_status_ext() and ieee80211_tx_rate_update() before any consumer runs: an entry that does not describe a valid rate for its encoding is dropped. minstrel_ht keeps the checks that depend on its own tables, i.e. the supported number of spatial streams, the MCS group size and the supported bandwidths. This does not take away the driver's responsibility to report sane values, it only prevents a single bad report from corrupting mac80211 state. No WARN_ON() is added: with panic_on_warn a driver bug would otherwise turn into a denial of service. Fixes: a5f69d94d8e2 ("mac80211: Get rid of search loop for rate group index") Fixes: 9208247d74bc ("mac80211: minstrel_ht: add basic support for VHT rates <= 3SS@80MHz") Reported-by: Vega Assisted-by: LLM Signed-off-by: Yuqi Xu Reviewed-by: Ren Wei Link: https://patch.msgid.link/607e985d71fbbe63ebfc182a37f882f04f980f7f.1789958876.git.xuyuqiabc@gmail.com Signed-off-by: Johannes Berg --- net/mac80211/rc80211_minstrel_ht.c | 59 ++++++++++++++++++++++--- net/mac80211/status.c | 70 ++++++++++++++++++++++++++++++ 2 files changed, 123 insertions(+), 6 deletions(-) diff --git a/net/mac80211/rc80211_minstrel_ht.c b/net/mac80211/rc80211_minstrel_ht.c index c26221287515..45e7a7ea7690 100644 --- a/net/mac80211/rc80211_minstrel_ht.c +++ b/net/mac80211/rc80211_minstrel_ht.c @@ -1194,6 +1194,54 @@ minstrel_ht_update_stats(struct minstrel_priv *mp, struct minstrel_ht_sta *mi) mi->sample_time = jiffies; } +/* + * Check whether an HT/VHT rate from a TX status entry maps to an entry + * in the minstrel_ht rate tables. Values that cannot be represented + * there must not be used for indexing mi->groups[] and the MCS groups. + */ +static bool +minstrel_ht_txstat_rate_valid(struct ieee80211_tx_rate *rate) +{ + unsigned int bw; + + if (!(rate->flags & (IEEE80211_TX_RC_MCS | IEEE80211_TX_RC_VHT_MCS))) + return true; + + if (rate->flags & IEEE80211_TX_RC_MCS) { + /* minstrel_ht supports up to MINSTREL_MAX_STREAMS streams */ + return rate->idx < MINSTREL_MAX_STREAMS * 8; + } + + /* minstrel_ht has no VHT groups for 160 MHz and wider */ + bw = !!(rate->flags & IEEE80211_TX_RC_40_MHZ_WIDTH) + + 2 * !!(rate->flags & IEEE80211_TX_RC_80_MHZ_WIDTH); + if ((rate->flags & IEEE80211_TX_RC_160_MHZ_WIDTH) || bw > BW_80) + return false; + + return ieee80211_rate_get_vht_nss(rate) <= MINSTREL_MAX_STREAMS && + ieee80211_rate_get_vht_mcs(rate) < MCS_GROUP_RATES; +} + +static bool +minstrel_ht_ri_txstat_rate_valid(struct rate_info *rate) +{ + if (!(rate->flags & (RATE_INFO_FLAGS_MCS | RATE_INFO_FLAGS_VHT_MCS))) + return true; + + if (rate->flags & RATE_INFO_FLAGS_MCS) { + /* minstrel_ht supports up to MINSTREL_MAX_STREAMS streams */ + return rate->mcs < MINSTREL_MAX_STREAMS * 8; + } + + /* minstrel_ht has VHT groups only for 20/40/80 MHz */ + if (rate->bw != RATE_INFO_BW_20 && rate->bw != RATE_INFO_BW_40 && + rate->bw != RATE_INFO_BW_80) + return false; + + return rate->nss <= MINSTREL_MAX_STREAMS && + rate->mcs < MCS_GROUP_RATES; +} + static bool minstrel_ht_txstat_valid(struct minstrel_priv *mp, struct minstrel_ht_sta *mi, struct ieee80211_tx_rate *rate) @@ -1206,9 +1254,8 @@ minstrel_ht_txstat_valid(struct minstrel_priv *mp, struct minstrel_ht_sta *mi, if (!rate->count) return false; - if (rate->flags & IEEE80211_TX_RC_MCS || - rate->flags & IEEE80211_TX_RC_VHT_MCS) - return true; + if (rate->flags & (IEEE80211_TX_RC_MCS | IEEE80211_TX_RC_VHT_MCS)) + return minstrel_ht_txstat_rate_valid(rate); for (i = 0; i < ARRAY_SIZE(mp->cck_rates); i++) if (rate->idx == mp->cck_rates[i]) @@ -1236,9 +1283,9 @@ minstrel_ht_ri_txstat_valid(struct minstrel_priv *mp, if (!rate_status->try_count) return false; - if (rate_status->rate_idx.flags & RATE_INFO_FLAGS_MCS || - rate_status->rate_idx.flags & RATE_INFO_FLAGS_VHT_MCS) - return true; + if (rate_status->rate_idx.flags & + (RATE_INFO_FLAGS_MCS | RATE_INFO_FLAGS_VHT_MCS)) + return minstrel_ht_ri_txstat_rate_valid(&rate_status->rate_idx); for (i = 0; i < ARRAY_SIZE(mp->cck_rates); i++) { if (rate_status->rate_idx.legacy == diff --git a/net/mac80211/status.c b/net/mac80211/status.c index 3d811f652603..414979e2a1d9 100644 --- a/net/mac80211/status.c +++ b/net/mac80211/status.c @@ -1164,6 +1164,73 @@ void ieee80211_tx_status_skb(struct ieee80211_hw *hw, struct sk_buff *skb) } EXPORT_SYMBOL(ieee80211_tx_status_skb); +/* + * Check whether the HT/VHT rate information in a TX status entry is + * valid for its encoding. This is not specific to any rate control + * algorithm; all status consumers rely on the values being sane. + */ +static bool ieee80211_tx_status_rate_info_valid(const struct rate_info *rate) +{ + if (rate->flags & RATE_INFO_FLAGS_MCS) + return rate->mcs < 32; + + if (rate->flags & RATE_INFO_FLAGS_VHT_MCS) + return rate->nss >= 1 && rate->nss <= 8 && rate->mcs <= 11; + + return true; +} + +static bool +ieee80211_tx_status_tx_rate_valid(const struct ieee80211_tx_rate *rate) +{ + if (rate->flags & IEEE80211_TX_RC_MCS) + return rate->idx < 32; + + if (rate->flags & IEEE80211_TX_RC_VHT_MCS) + return ieee80211_rate_get_vht_mcs(rate) <= 11; + + return true; +} + +/* + * Drop TX status rate entries that don't describe a valid rate. The + * status information is used by rate control and by other mac80211 + * code, and a malformed entry must not be able to corrupt state beyond + * the driver that reported it. + */ +static void +ieee80211_tx_status_drop_invalid_rates(struct ieee80211_tx_status *status) +{ + int i; + + for (i = 0; i < status->n_rates; i++) { + struct ieee80211_rate_status *rs = &status->rates[i]; + + if (ieee80211_tx_status_rate_info_valid(&rs->rate_idx)) + continue; + + rs->try_count = 0; + memset(&rs->rate_idx, 0, sizeof(rs->rate_idx)); + } + + if (!status->info) + return; + + for (i = 0; i < IEEE80211_TX_MAX_RATES; i++) { + struct ieee80211_tx_rate *rate; + + rate = &status->info->status.rates[i]; + if (rate->idx < 0) + break; + + if (ieee80211_tx_status_tx_rate_valid(rate)) + continue; + + rate->idx = -1; + rate->count = 0; + } +} + void ieee80211_tx_status_ext(struct ieee80211_hw *hw, struct ieee80211_tx_status *status) { @@ -1176,6 +1243,8 @@ void ieee80211_tx_status_ext(struct ieee80211_hw *hw, bool acked, noack_success, ack_signal_valid; u16 tx_time_est; + ieee80211_tx_status_drop_invalid_rates(status); + if (pubsta) { sta = container_of(pubsta, struct sta_info, sta); @@ -1300,6 +1369,7 @@ void ieee80211_tx_rate_update(struct ieee80211_hw *hw, .sta = pubsta, }; + ieee80211_tx_status_drop_invalid_rates(&status); rate_control_tx_status(local, &status); if (ieee80211_hw_check(&local->hw, HAS_RATE_CONTROL)) From e588e83e5c7a0f6d7a77dbbb58927c8c8cfd598e Mon Sep 17 00:00:00 2001 From: Zhao Li Date: Thu, 10 Sep 2026 14:54:02 +0800 Subject: [PATCH 325/694] wifi: mac80211: shut down RX BA session timer on teardown Tearing down an RX BA session can leave the session timer queued on a freed tid_rx. Later timer-wheel processing and the timer callback itself then run against freed memory. An RX BAR handler can dereference tid_rx under RCU before teardown unpublishes it and stay in its read-side critical section while teardown deletes the session timer. The stale handler can then rearm the timer after timer_delete_sync() returns and before the RCU callback frees tid_rx. Use timer_shutdown_sync() so rearm attempts become ineffective once teardown starts. Keep reorder_timer unchanged: teardown marks the session removed under reorder_lock and its release path checks that state before rearming. session_timer has no equivalent guard because ieee80211_rx_h_ctrl() rearms it before taking reorder_lock. Fixes: a87f736d942c ("mac80211: use RCU for RX aggregation") Cc: stable@kernel.org # 5.15+ Assisted-by: LLM sparse kasan Signed-off-by: Zhao Li Signed-off-by: Johannes Berg --- net/mac80211/agg-rx.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/mac80211/agg-rx.c b/net/mac80211/agg-rx.c index 9629e00069a1..01d6703767c6 100644 --- a/net/mac80211/agg-rx.c +++ b/net/mac80211/agg-rx.c @@ -102,7 +102,7 @@ void __ieee80211_stop_rx_ba_session(struct sta_info *sta, u16 tid, if (!tid_rx) return; - timer_delete_sync(&tid_rx->session_timer); + timer_shutdown_sync(&tid_rx->session_timer); /* make sure ieee80211_sta_reorder_release() doesn't re-arm the timer */ spin_lock_bh(&tid_rx->reorder_lock); From a10a2f80476d166fd81c361ec0319493671b694e Mon Sep 17 00:00:00 2001 From: Zhao Li Date: Thu, 10 Sep 2026 14:52:51 +0800 Subject: [PATCH 326/694] wifi: mac80211: keep fallback association elements alive Associating to a non-transmitted MBSSID profile reads freed memory while determining the operating channel and applying HT capabilities. When an association response omits WMM, HT, or VHT information, ieee80211_assoc_config_link() parses the cached BSS IEs and copies the missing element pointers out of that fallback parse. For a non-transmitted MBSSID profile those pointers refer to merged-profile storage owned by the parse object, and the fallback block frees that object at its end. The copied pointers are used afterwards to determine the operating channel, configure bandwidth, and apply HT capabilities. Keep the fallback parse result alive until association setup finishes. Fixes: 5023b14cf4df ("mac80211: support profile split between elements") Cc: stable@kernel.org Assisted-by: LLM sparse kasan [reorder variable declaration] Signed-off-by: Zhao Li Signed-off-by: Johannes Berg --- net/mac80211/mlme.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index f51167f0fc46..3c31c16860ab 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -5877,6 +5877,7 @@ static bool ieee80211_assoc_config_link(struct ieee80211_link_data *link, bool is_6ghz = cbss->channel->band == NL80211_BAND_6GHZ; bool is_s1g = cbss->channel->band == NL80211_BAND_S1GHZ; const struct cfg80211_bss_ies *bss_ies = NULL; + struct ieee802_11_elems *bss_elems = NULL; struct ieee80211_supported_band *sband; struct ieee802_11_elems *elems; u16 capab_info; @@ -6007,7 +6008,6 @@ static bool ieee80211_assoc_config_link(struct ieee80211_link_data *link, (is_5ghz && link->u.mgd.conn.mode >= IEEE80211_CONN_MODE_VHT && (!elems->vht_cap_elem || !elems->vht_operation)))) { const struct cfg80211_bss_ies *ies; - struct ieee802_11_elems *bss_elems; rcu_read_lock(); ies = rcu_dereference(cbss->ies); @@ -6069,7 +6069,6 @@ static bool ieee80211_assoc_config_link(struct ieee80211_link_data *link, "AP bug: VHT operation missing from AssocResp\n"); } } - kfree(bss_elems); } /* @@ -6342,6 +6341,7 @@ static bool ieee80211_assoc_config_link(struct ieee80211_link_data *link, ret = true; out: kfree(elems); + kfree(bss_elems); kfree(bss_ies); return ret; } From cc45f313d85533d647ce619326302aad6f797c14 Mon Sep 17 00:00:00 2001 From: Zhao Li Date: Mon, 14 Sep 2026 23:28:25 +0800 Subject: [PATCH 327/694] wifi: cfg80211: preserve hidden-group beacon IE ownership A hidden-group probe-response entry borrows its beacon IEs from the group's beacon entry. cfg80211_update_assoc_bss_entry() can pass such an entry to cfg80211_update_known_bss() when merging scan entries after a channel switch. Copying the borrowed beacon_ies into the associated entry gives the same allocation two owners. Releasing the group's beacon entry can then leave the associated entry pointing to freed IEs, which are freed again when that entry is released. Skip beacon IE replacement when the source entry belongs to a hidden group. Preserve probe-response processing, the subsequent metadata updates, and the existing rejection path for conflicting hidden beacon updates. Fixes: 0afd425b1b64 ("cfg80211: fix duplicated scan entries after channel switch") Reported-by: Yilin Zhang Closes: https://lore.kernel.org/linux-wireless/20260901080427.1983015-1-yilinzhang@moonshot.ai/ Assisted-by: Codex:gpt-6-astra Signed-off-by: Zhao Li Signed-off-by: Johannes Berg --- net/wireless/scan.c | 22 ++++++++++++---------- 1 file changed, 12 insertions(+), 10 deletions(-) diff --git a/net/wireless/scan.c b/net/wireless/scan.c index caa9c6495f20..d35c826c4b1a 100644 --- a/net/wireless/scan.c +++ b/net/wireless/scan.c @@ -1992,20 +1992,22 @@ cfg80211_update_known_bss(struct cfg80211_registered_device *rdev, return false; } - old = rcu_access_pointer(known->pub.beacon_ies); + if (!new->pub.hidden_beacon_bss) { + old = rcu_access_pointer(known->pub.beacon_ies); - rcu_assign_pointer(known->pub.beacon_ies, new->pub.beacon_ies); + rcu_assign_pointer(known->pub.beacon_ies, new->pub.beacon_ies); - /* Override IEs if they were from a beacon before */ - if (old == rcu_access_pointer(known->pub.ies)) - rcu_assign_pointer(known->pub.ies, new->pub.beacon_ies); + /* Override IEs if they were from a beacon before */ + if (old == rcu_access_pointer(known->pub.ies)) + rcu_assign_pointer(known->pub.ies, new->pub.beacon_ies); - cfg80211_update_hidden_bsses(known, - rcu_access_pointer(new->pub.beacon_ies), - old); + cfg80211_update_hidden_bsses(known, + rcu_access_pointer(new->pub.beacon_ies), + old); - if (old) - kfree_rcu((struct cfg80211_bss_ies *)old, rcu_head); + if (old) + kfree_rcu((struct cfg80211_bss_ies *)old, rcu_head); + } } known->pub.beacon_interval = new->pub.beacon_interval; From 7f93ca31f7fd9b42e7d692f59514436d448ac908 Mon Sep 17 00:00:00 2001 From: Felix Fietkau Date: Wed, 16 Sep 2026 11:48:32 +0200 Subject: [PATCH 328/694] wifi: mac80211: prevent AP VLAN tx from other interfaces Since the station lookup was factored out of ieee80211_build_hdr(), ieee80211_lookup_ra_sta() uses sta_info_get_bss() for all AP and AP VLAN frames, not only for control port frames. A unicast frame sent on one interface thus matches a station of a different VLAN of the same BSS and is transmitted to it. Only match stations of the interface the frame is sent on, except for control port frames, since those may be sent for VLAN stations directly on the BSS interface. Fixes: 97ffe75791b3 ("mac80211: factor out station lookup from ieee80211_build_hdr()") Signed-off-by: Felix Fietkau Link: https://patch.msgid.link/20260916094833.164219-1-nbd@nbd.name Signed-off-by: Johannes Berg --- net/mac80211/ieee80211_i.h | 2 +- net/mac80211/tx.c | 24 +++++++++++++++--------- 2 files changed, 16 insertions(+), 10 deletions(-) diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h index d05f59467399..9374c2942489 100644 --- a/net/mac80211/ieee80211_i.h +++ b/net/mac80211/ieee80211_i.h @@ -2475,7 +2475,7 @@ static inline bool ieee80211_require_encrypted_assoc(__le16 fc, /* sta_out needs to be checked for ERR_PTR() before using */ int ieee80211_lookup_ra_sta(struct ieee80211_sub_if_data *sdata, struct sk_buff *skb, - struct sta_info **sta_out); + struct sta_info **sta_out, bool bss); static inline void ieee80211_tx_skb_tid_band(struct ieee80211_sub_if_data *sdata, diff --git a/net/mac80211/tx.c b/net/mac80211/tx.c index 814399989b5e..c8217f6d146e 100644 --- a/net/mac80211/tx.c +++ b/net/mac80211/tx.c @@ -2532,7 +2532,7 @@ static inline bool ieee80211_is_tdls_setup(struct sk_buff *skb) int ieee80211_lookup_ra_sta(struct ieee80211_sub_if_data *sdata, struct sk_buff *skb, - struct sta_info **sta_out) + struct sta_info **sta_out, bool bss) { struct sta_info *sta; @@ -2553,7 +2553,10 @@ int ieee80211_lookup_ra_sta(struct ieee80211_sub_if_data *sdata, *sta_out = ERR_PTR(-ENOENT); return 0; } - sta = sta_info_get_bss(sdata, skb->data); + if (bss) + sta = sta_info_get_bss(sdata, skb->data); + else + sta = sta_info_get(sdata, skb->data); break; #ifdef CONFIG_MAC80211_MESH case NL80211_IFTYPE_MESH_POINT: @@ -4419,7 +4422,8 @@ void __ieee80211_subif_start_xmit(struct sk_buff *skb, ieee80211_mesh_xmit_fast(sdata, skb, ctrl_flags)) goto out; - if (ieee80211_lookup_ra_sta(sdata, skb, &sta)) + if (ieee80211_lookup_ra_sta(sdata, skb, &sta, + skb->protocol == sdata->control_port_protocol)) goto out_free; if (IS_ERR(sta)) @@ -4848,7 +4852,10 @@ static void __ieee80211_subif_start_xmit_8023(struct sk_buff *skb, rcu_read_lock(); - if (ieee80211_lookup_ra_sta(sdata, skb, &sta)) { + if (unlikely(sdata->control_port_protocol == ehdr->h_proto)) + goto skip_offload; + + if (ieee80211_lookup_ra_sta(sdata, skb, &sta, false)) { kfree_skb(skb); goto out; } @@ -4869,8 +4876,7 @@ static void __ieee80211_subif_start_xmit_8023(struct sk_buff *skb, link = &sdata->deflink; key = rcu_dereference(link->default_multicast_key); } else if (unlikely(IS_ERR_OR_NULL(sta) || !sta->uploaded || - !test_sta_flag(sta, WLAN_STA_AUTHORIZED) || - sdata->control_port_protocol == ehdr->h_proto)) { + !test_sta_flag(sta, WLAN_STA_AUTHORIZED))) { goto skip_offload; } else { key = rcu_dereference(sta->ptk[sta->ptk_idx]); @@ -4931,7 +4937,7 @@ ieee80211_build_data_template(struct ieee80211_sub_if_data *sdata, rcu_read_lock(); - if (ieee80211_lookup_ra_sta(sdata, skb, &sta)) { + if (ieee80211_lookup_ra_sta(sdata, skb, &sta, false)) { kfree_skb(skb); skb = ERR_PTR(-EINVAL); goto out; @@ -5002,7 +5008,7 @@ static bool ieee80211_tx_pending_skb(struct ieee80211_local *local, } result = ieee80211_tx(sdata, NULL, skb, true); } else if (info->flags & IEEE80211_TX_CTL_HW_80211_ENCAP) { - if (ieee80211_lookup_ra_sta(sdata, skb, &sta)) { + if (ieee80211_lookup_ra_sta(sdata, skb, &sta, true)) { dev_kfree_skb(skb); return true; } @@ -6639,7 +6645,7 @@ int ieee80211_tx_control_port(struct wiphy *wiphy, struct net_device *dev, * AF_PACKET */ rcu_read_lock(); - err = ieee80211_lookup_ra_sta(sdata, skb, &sta); + err = ieee80211_lookup_ra_sta(sdata, skb, &sta, true); if (err) { dev_kfree_skb(skb); rcu_read_unlock(); From 0d6526f82c3cdefcca47f73f5fc08dc6f335eac6 Mon Sep 17 00:00:00 2001 From: Tim Chen Date: Mon, 21 Sep 2026 17:37:22 -0700 Subject: [PATCH 329/694] sched/cache: Keep nr_pref_llc_running in the runnable domain, to fix LLC mis-scheduling bug alb_break_llc() decides whether to break LLC preference during active load balance. It does so by testing that every runnable fair task on the source rq prefers its LLC: env->src_rq->nr_pref_llc_running == env->src_rq->cfs.h_nr_runnable But the two counters cover different sets. nr_pref_llc_running is updated in account_llc_enqueue()/account_llc_dequeue(), next to cfs_rq->nr_queued, so it follows queued tasks. h_nr_runnable is updated in set_delayed()/ clear_delayed() and drops delay-dequeued tasks. So under DELAY_DEQUEUE, a preferring task that goes to sleep stays counted in nr_pref_llc_running while h_nr_runnable falls. The equality then breaks, alb_break_llc() returns false, and active balance is free to pull a task off its preferred LLC. Active balance only moves runnable tasks, and this is the only LLC check it consults: once the stopper runs, LBF_ACTIVE_LB skips the per-task test in can_migrate_task(). The runnable set is the one we want. Fix it on the counter side. A task should be counted in nr_pref_llc_running exactly while it is both queued on its preferred LLC (pref_llc_queued) and runnable (!sched_delayed). Define that membership once in task_pref_llc_runnable(), and adjust the counter only through pref_llc_running_inc()/pref_llc_running_dec() from the four sites that change either input: account_llc_enqueue(), account_llc_dequeue(), set_delayed() and clear_delayed(). Gating every update on the same predicate keeps the delay, wake and dequeue paths from double-counting or underflowing; see the comments at those sites for the ordering. nr_llc_running and sd->llc_counts are not touched and stay on queued semantics. Fixes: 714059f79ff0 ("sched/cache: Handle moving single tasks to/from their preferred LLC") Closes: https://lore.kernel.org/lkml/20260827135000.735138-1-zhanxusheng@xiaomi.com/ Reported-by: Zhan Xusheng Suggested-by: Chen Yu Signed-off-by: Tim Chen Signed-off-by: Peter Zijlstra (Intel) Signed-off-by: Ingo Molnar Reviewed-by: Kayra Cizmeci Cc: # v7.2.x Link: https://patch.msgid.link/06af61afedac32e6477f57feb4d658f6c411c3af.1790035273.git.tim.c.chen@linux.intel.com --- kernel/sched/fair.c | 63 +++++++++++++++++++++++++++++++++++++++++---- 1 file changed, 58 insertions(+), 5 deletions(-) diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index 7455a83a6a99..de3d589fa8eb 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -1546,6 +1546,28 @@ static bool invalid_llc_nr(struct mm_struct *mm, struct task_struct *p, (scale * per_cpu(sd_llc_size, cpu))); } +/* + * A task counts in nr_pref_llc_running while it is queued on its preferred + * LLC (pref_llc_queued) and runnable (!sched_delayed), keeping the counter in + * the runnable domain so alb_break_llc() can compare it with h_nr_runnable. + */ +static bool task_pref_llc_runnable(struct task_struct *p) +{ + return p->pref_llc_queued && !p->se.sched_delayed; +} + +static void pref_llc_running_inc(struct rq *rq, struct task_struct *p) +{ + if (task_pref_llc_runnable(p)) + rq->nr_pref_llc_running++; +} + +static void pref_llc_running_dec(struct rq *rq, struct task_struct *p) +{ + if (task_pref_llc_runnable(p)) + rq->nr_pref_llc_running--; +} + static void account_llc_enqueue(struct rq *rq, struct task_struct *p) { int pref_llc, pref_llc_queued; @@ -1557,7 +1579,6 @@ static void account_llc_enqueue(struct rq *rq, struct task_struct *p) pref_llc_queued = (pref_llc == task_llc(p)); rq->nr_llc_running++; - rq->nr_pref_llc_running += pref_llc_queued; /* * Record whether p is enqueued on its preferred @@ -1575,6 +1596,9 @@ static void account_llc_enqueue(struct rq *rq, struct task_struct *p) */ p->pref_llc_queued = pref_llc_queued; + /* Skipped while delayed; clear_delayed() adds it back on wake. */ + pref_llc_running_inc(rq, p); + sd = rcu_dereference_all(rq->sd); if (sd && (unsigned int)pref_llc < sd->llc_max) sd->llc_counts[pref_llc]++; @@ -1591,7 +1615,12 @@ static void account_llc_dequeue(struct rq *rq, struct task_struct *p) rq->nr_llc_running--; if (p->pref_llc_queued) { - rq->nr_pref_llc_running--; + /* + * Skipped if still delayed (set_delayed() already removed it); + * clearing pref_llc_queued below also stops clear_delayed() + * from re-adding it. + */ + pref_llc_running_dec(rq, p); /* * Update the status in case * other logic might query @@ -1995,6 +2024,7 @@ void init_sched_mm(struct task_struct *p) * polluting account_llc_enqueue(). */ p->preferred_llc = -1; + p->pref_llc_queued = 0; } #else /* CONFIG_SCHED_CACHE */ @@ -2016,6 +2046,10 @@ static void account_llc_enqueue(struct rq *rq, struct task_struct *p) {} static void account_llc_dequeue(struct rq *rq, struct task_struct *p) {} +static void pref_llc_running_inc(struct rq *rq, struct task_struct *p) {} + +static void pref_llc_running_dec(struct rq *rq, struct task_struct *p) {} + #endif /* CONFIG_SCHED_CACHE */ /* @@ -6390,15 +6424,27 @@ static __always_inline void return_cfs_rq_runtime(struct cfs_rq *cfs_rq); static void set_delayed(struct sched_entity *se) { - se->sched_delayed = 1; - /* * Delayed se of cfs_rq have no tasks queued on them. * Do not adjust h_nr_runnable since __dequeue_task() * will account it for blocked tasks. + * + * This check can be removed because when flat pick + * patches get merged as only task can get delayed, + * same for clear_delayed(). */ - if (!entity_is_task(se)) + if (!entity_is_task(se)) { + se->sched_delayed = 1; return; + } + + /* + * Drop a task leaving the runnable set. + * Needs to be called before sched_delayed is set. + * clear_delayed() mirrors this after clearing the flag. + */ + pref_llc_running_dec(rq_of(cfs_rq_of(se)), task_of(se)); + se->sched_delayed = 1; for_each_sched_entity(se) { struct cfs_rq *cfs_rq = cfs_rq_of(se); @@ -6420,6 +6466,13 @@ static void clear_delayed(struct sched_entity *se) if (!entity_is_task(se)) return; + /* + * Re-add on wake, after sched_delayed is cleared. On a final delayed + * dequeue account_llc_dequeue() already cleared pref_llc_queued, so + * this does nothing. + */ + pref_llc_running_inc(rq_of(cfs_rq_of(se)), task_of(se)); + for_each_sched_entity(se) { struct cfs_rq *cfs_rq = cfs_rq_of(se); From d6013e2465d98d524b030a81c1223882a1bb7e4c Mon Sep 17 00:00:00 2001 From: Lu Wang Date: Mon, 21 Sep 2026 17:37:23 -0700 Subject: [PATCH 330/694] sched/cache: Honor migrate_llc_task semantics in active load balance, to fix LLC mis-scheduling bug Cache aware scheduling introduced the migrate_llc_task migration type to direct tasks toward their preferred LLC, but its semantics can be lost when passive load balance falls back to active load balance (ALB). This may allow ALB to select a candidate whose preferred LLC does not match the destination, moving it away from its preferred LLC. Example scenario: src_rq has two runnable tasks, p1 and p2. p1 prefers dst_rq (dst_llc), while p2 prefers src_rq (src_llc). In this case, migrate_llc_task is set because src_rq has at least one task, p1, that wants to migrate to dst_rq. In ALB, can_migrate_task() finds p2 and returns true for it, thus moving p2 out of its preferred LLC. Solution: The CPU stopper in ALB constructs a fresh lb_env that does not inherit migration_type from the passive load-balance pass. Two approaches are possible: (a) Add a new member to struct rq so ALB can inherit migrate_llc_task from the passive LB that triggered it. (b) Define a new flag LBF_ACTIVE_LB_LLC and select the stopper callback at kick time to preserve the migration semantics across the asynchronous boundary. We choose (b) because it avoids passing migration_type through the stopper, which would affect the meaning of migration_type for delayed-dequeue tasks. Fixes: e4c9a4cb244a ("sched/cache: Add migrate_llc_task migration type for cache-aware balancing") Suggested-by: Chen Yu Signed-off-by: Lu Wang Signed-off-by: Tim Chen Signed-off-by: Peter Zijlstra (Intel) Signed-off-by: Ingo Molnar Reviewed-by: Tim Chen Reviewed-by: Chen Yu Cc: # v7.2.x Link: https://patch.msgid.link/cb39f64a17fc2b76097264aaec74a2d6dfff4315.1790035273.git.tim.c.chen@linux.intel.com --- kernel/sched/fair.c | 57 ++++++++++++++++++++++++++++++++++++++++----- 1 file changed, 51 insertions(+), 6 deletions(-) diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index de3d589fa8eb..514bd54ccd56 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -10448,6 +10448,7 @@ enum migration_type { #define LBF_SOME_PINNED 0x08 #define LBF_ACTIVE_LB 0x10 #define LBF_LLC_PINNED 0x20 +#define LBF_ACTIVE_LB_LLC 0x40 struct lb_env { struct sched_domain *sd; @@ -10866,6 +10867,21 @@ alb_break_llc(struct lb_env *env) return false; } +/* + * Returns true if p's preferred LLC does not match the destination CPU + * under migrate_llc_task semantics. Passive LB passes migrate_llc_task + * in env->migration_type, while active LB carries LBF_ACTIVE_LB_LLC in + * env->flags to avoid overwriting env->migration_type. + */ +static inline bool +migrate_llc_task_wrong_dst(struct task_struct *p, struct lb_env *env) +{ + return sched_cache_enabled() && + (env->migration_type == migrate_llc_task || + env->flags & LBF_ACTIVE_LB_LLC) && + READ_ONCE(p->preferred_llc) != llc_id(env->dst_cpu); +} + /* * Check if migrating task p from env->src_cpu to * env->dst_cpu breaks LLC localiy. @@ -10894,8 +10910,7 @@ static bool migrate_degrades_llc(struct task_struct *p, struct lb_env *env) * run on env->dst_cpu, skip the tasks do not prefer * env->dst_cpu, and find the one that prefers. */ - if (env->migration_type == migrate_llc_task && - READ_ONCE(p->preferred_llc) != llc_id(env->dst_cpu)) + if (migrate_llc_task_wrong_dst(p, env)) return true; if (can_migrate_llc_task(env, p) != mig_forbid) @@ -10917,6 +10932,12 @@ alb_break_llc(struct lb_env *env) return false; } +static inline bool +migrate_llc_task_wrong_dst(struct task_struct *p, struct lb_env *env) +{ + return false; +} + static inline bool migrate_degrades_llc(struct task_struct *p, struct lb_env *env) { @@ -11016,7 +11037,7 @@ int can_migrate_task(struct task_struct *p, struct lb_env *env) * 4) too many balance attempts have failed. */ if (env->flags & LBF_ACTIVE_LB) - return 1; + return !migrate_llc_task_wrong_dst(p, env); degrades = migrate_degrades_locality(p, env); if (!degrades) { @@ -13415,6 +13436,20 @@ static int need_active_balance(struct lb_env *env) } static int active_load_balance_cpu_stop(void *data); +static int active_load_balance_llc_cpu_stop(void *data); + +/* + * migration_type is checked elsewhere to decide migration policy, so + * it shouldn't be repurposed just to flag an LLC-directed active + * balance across the stopper. Pick the callback here instead. + */ +static inline cpu_stop_fn_t alb_stop_fn(struct lb_env *env) +{ + if (env->migration_type == migrate_llc_task) + return active_load_balance_llc_cpu_stop; + + return active_load_balance_cpu_stop; +} static int should_we_balance(struct lb_env *env) { @@ -13760,7 +13795,7 @@ static int sched_balance_rq(int this_cpu, struct rq *this_rq, } if (active_balance) { stop_one_cpu_nowait(cpu_of(busiest), - active_load_balance_cpu_stop, busiest, + alb_stop_fn(&env), busiest, &busiest->active_balance_work); } preempt_enable(); @@ -13865,7 +13900,7 @@ update_next_balance(struct sched_domain *sd, unsigned long *next_balance) * least 1 task to be running on each physical CPU where possible, and * avoids physical / logical imbalances. */ -static int active_load_balance_cpu_stop(void *data) +static int __active_load_balance_cpu_stop(void *data, unsigned int lb_flags) { struct rq *busiest_rq = data; int busiest_cpu = cpu_of(busiest_rq); @@ -13915,7 +13950,7 @@ static int active_load_balance_cpu_stop(void *data) .src_cpu = busiest_rq->cpu, .src_rq = busiest_rq, .idle = CPU_IDLE, - .flags = LBF_ACTIVE_LB, + .flags = LBF_ACTIVE_LB | lb_flags, }; schedstat_inc(sd->alb_count); @@ -13943,6 +13978,16 @@ static int active_load_balance_cpu_stop(void *data) return 0; } +static int active_load_balance_cpu_stop(void *data) +{ + return __active_load_balance_cpu_stop(data, 0); +} + +static int active_load_balance_llc_cpu_stop(void *data) +{ + return __active_load_balance_cpu_stop(data, LBF_ACTIVE_LB_LLC); +} + /* * Scale the max sched_balance_rq interval with the number of CPUs in the system. * This trades load-balance latency on larger machines for less cross talk. From 28f9c0e0a0b94c5d3e1b634db545f6e1f94858c5 Mon Sep 17 00:00:00 2001 From: Tim Chen Date: Mon, 21 Sep 2026 17:37:24 -0700 Subject: [PATCH 331/694] sched/cache: Decouple sched_cache_group from mm to fix UAF Currently the sched cache grouping is by mm and the scheduling statistics sched_cache_stat lives in the mm structure. This ties the life cycle of scheduling stats with mm. In account_mm_sched(), the scheduling stats are accessed by task->mm->sc_stat. However, a task may be switching mm on one CPU when another CPU is running account_mm_sched(), and possibly accessing the old mm that was freed. This problem was found when running tests with KASAN by Hyunwoo: https://lore.kernel.org/lkml/apPb-Dr4nPYuHQOK@v4bel/ Instead of serializing the mm access by introducing extra acquisition of rq lock in the mm free path, extract sched_cache_stat from mm_struct, rename it as sched_cache_group and manage its life cycle apart from mm_struct with its own ref counting. This allows us in the next patch access sched_cache_group directly from task, and add a refcount on sched_cache_group when a task links to it. This prevents the use after free issue when accessing stale and released old mm and its sched cache stat a task switches to a new mm while account_mm_sched() is done elsewhere. The other benefit of this restructure is in the future, the grouping of tasks to a LLC would have the flexibility to be associated with a user defined grouping, or cgroup, cookie group, numa_group or others instead of just with a single mm address space. Rename sched_cache_stat to sched_cache_group and turn it into a refcounted object allocated from mm_struct. The mm_struct now holds a pointer (sched_cache_grp) to this object instead of embedding it. Fixes: df0d98475954 ("sched/cache: Introduce infrastructure for cache-aware load balancing") Closes: https://lore.kernel.org/lkml/apPb-Dr4nPYuHQOK@v4bel/ Closes: https://lore.kernel.org/all/343a7e07-7fad-4979-9c9b-82ec038c293c@linux.dev/ Reported-by: Hyunwoo Kim Reported-by: Zenghui Yu (Huawei) Co-developed-by: Chen Yu Signed-off-by: Chen Yu Signed-off-by: Tim Chen Signed-off-by: Peter Zijlstra (Intel) Signed-off-by: Ingo Molnar Cc: #7.2.x Link: https://patch.msgid.link/91fd1e3266707c865bc9abecfb3e17bc676712df.1790035273.git.tim.c.chen@linux.intel.com --- include/linux/mm_types.h | 15 ++-- include/linux/sched.h | 6 +- kernel/exit.c | 11 ++- kernel/sched/fair.c | 173 ++++++++++++++++++++++++++++----------- 4 files changed, 144 insertions(+), 61 deletions(-) diff --git a/include/linux/mm_types.h b/include/linux/mm_types.h index 6d815f6440c9..f3e5a2fadbe5 100644 --- a/include/linux/mm_types.h +++ b/include/linux/mm_types.h @@ -1226,7 +1226,7 @@ struct mm_struct { struct mm_mm_cid mm_cid; /* sched_cache related statistics */ - struct sched_cache_stat sc_stat; + struct sched_cache_group *sched_cache_grp; #ifdef CONFIG_MMU atomic_long_t pgtables_bytes; /* size of all page tables */ #endif @@ -1624,8 +1624,9 @@ static inline unsigned int mm_cid_size(void) #endif /* CONFIG_SCHED_MM_CID */ #ifdef CONFIG_SCHED_CACHE -void mm_init_sched(struct mm_struct *mm, - struct sched_cache_time __percpu *pcpu_sched); +int mm_init_sched(struct mm_struct *mm, + struct sched_cache_time __percpu *pcpu_sched); +void mm_destroy_sched(struct mm_struct *mm); static inline int mm_alloc_sched_noprof(struct mm_struct *mm) { @@ -1635,17 +1636,11 @@ static inline int mm_alloc_sched_noprof(struct mm_struct *mm) if (!pcpu_sched) return -ENOMEM; - mm_init_sched(mm, pcpu_sched); - return 0; + return mm_init_sched(mm, pcpu_sched); } #define mm_alloc_sched(...) alloc_hooks(mm_alloc_sched_noprof(__VA_ARGS__)) -static inline void mm_destroy_sched(struct mm_struct *mm) -{ - free_percpu(mm->sc_stat.pcpu_sched); - mm->sc_stat.pcpu_sched = NULL; -} #else /* !CONFIG_SCHED_CACHE */ static inline int mm_alloc_sched(struct mm_struct *mm) { return 0; } diff --git a/include/linux/sched.h b/include/linux/sched.h index 705970d07614..e14ad43522c8 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -2405,7 +2405,7 @@ struct sched_cache_time { unsigned long epoch; }; -struct sched_cache_stat { +struct sched_cache_group { struct sched_cache_time __percpu *pcpu_sched; raw_spinlock_t lock; unsigned long epoch; @@ -2413,11 +2413,13 @@ struct sched_cache_stat { unsigned long next_scan; unsigned long footprint; int cpu; + refcount_t refcnt; + struct rcu_head rcu; } ____cacheline_aligned_in_smp; #else -struct sched_cache_stat { }; +struct sched_cache_group { }; #endif diff --git a/kernel/exit.c b/kernel/exit.c index 424c44a42a4d..024350e9b48c 100644 --- a/kernel/exit.c +++ b/kernel/exit.c @@ -558,18 +558,23 @@ void mm_update_next_owner(struct mm_struct *mm) */ static void exit_mm_sched_cache(struct mm_struct *mm) { + struct sched_cache_group *grp; unsigned long fp, sub; if (!current->total_numa_faults) return; /* * No lock protection due to performance considerations. - * Make sure mm->sc_stat.footprint does not become + * Make sure the group footprint does not become * negative. */ - fp = READ_ONCE(mm->sc_stat.footprint); + grp = READ_ONCE(mm->sched_cache_grp); + if (!grp) + return; + + fp = READ_ONCE(grp->footprint); sub = min(fp, current->total_numa_faults); - WRITE_ONCE(mm->sc_stat.footprint, fp - sub); + WRITE_ONCE(grp->footprint, fp - sub); } #else static inline void exit_mm_sched_cache(struct mm_struct *mm) diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index 514bd54ccd56..f0a9586332fd 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -1492,12 +1492,17 @@ static bool exceed_llc_capacity(struct mm_struct *mm, int cpu) return true; if (static_branch_likely(&sched_numa_balancing)) { + struct sched_cache_group *grp = READ_ONCE(mm->sched_cache_grp); + + if (!grp) + return true; + /* * TBD: RDT exclusive LLC ways reserved should be * excluded. */ llc = sd->llc_bytes; - footprint = READ_ONCE(mm->sc_stat.footprint); + footprint = READ_ONCE(grp->footprint); /* * Scale the LLC size by 256*llc_aggr_tolerance @@ -1529,6 +1534,7 @@ static bool exceed_llc_capacity(struct mm_struct *mm, int cpu) static bool invalid_llc_nr(struct mm_struct *mm, struct task_struct *p, int cpu) { + struct sched_cache_group *grp; int scale; if (get_nr_threads(p) <= 1) @@ -1542,7 +1548,11 @@ static bool invalid_llc_nr(struct mm_struct *mm, struct task_struct *p, if (scale == INT_MAX) return false; - return !fits_capacity((mm->sc_stat.nr_running_avg * cpu_smt_num_threads), + grp = READ_ONCE(mm->sched_cache_grp); + if (!grp) + return true; + + return !fits_capacity((READ_ONCE(grp->nr_running_avg) * cpu_smt_num_threads), (scale * per_cpu(sd_llc_size, cpu))); } @@ -1648,12 +1658,20 @@ static void account_llc_dequeue(struct rq *rq, struct task_struct *p) } } -void mm_init_sched(struct mm_struct *mm, - struct sched_cache_time __percpu *_pcpu_sched) +int mm_init_sched(struct mm_struct *mm, + struct sched_cache_time __percpu *_pcpu_sched) { + struct sched_cache_group *grp; unsigned long epoch = 0; int i; + grp = kzalloc_obj(*grp); + if (!grp) { + free_percpu(_pcpu_sched); + mm->sched_cache_grp = NULL; + return -ENOMEM; + } + for_each_possible_cpu(i) { struct sched_cache_time *pcpu_sched = per_cpu_ptr(_pcpu_sched, i); struct rq *rq = cpu_rq(i); @@ -1664,18 +1682,51 @@ void mm_init_sched(struct mm_struct *mm, epoch = rq->cpu_epoch; } - raw_spin_lock_init(&mm->sc_stat.lock); - mm->sc_stat.epoch = epoch; - mm->sc_stat.cpu = -1; - mm->sc_stat.next_scan = jiffies; - mm->sc_stat.nr_running_avg = 0; - mm->sc_stat.footprint = 0; + raw_spin_lock_init(&grp->lock); + grp->epoch = epoch; + grp->cpu = -1; + grp->next_scan = jiffies; + grp->nr_running_avg = 0; + grp->footprint = 0; + refcount_set(&grp->refcnt, 1); /* - * The update to mm->sc_stat should not be reordered - * before initialization to mm's other fields, in case + * The update to grp->pcpu_sched should not be reordered + * before initialization to grp's other fields, in case * the readers may get invalid mm_sched_epoch, etc. */ - smp_store_release(&mm->sc_stat.pcpu_sched, _pcpu_sched); + smp_store_release(&grp->pcpu_sched, _pcpu_sched); + /* + * Publish the group last. Not every reader qualifies it by + * grp->pcpu_sched - can_migrate_llc_task() only checks that the + * pointer is non-NULL before reading grp->footprint and + * grp->nr_running_avg - so a reachable group must already be + * fully initialized. + */ + smp_store_release(&mm->sched_cache_grp, grp); + return 0; +} + +static void sched_cache_group_free_rcu(struct rcu_head *rcu) +{ + struct sched_cache_group *grp = + container_of(rcu, struct sched_cache_group, rcu); + + free_percpu(grp->pcpu_sched); + kfree(grp); +} + +static void sched_cache_group_put(struct sched_cache_group *grp) +{ + if (!grp || !refcount_dec_and_test(&grp->refcnt)) + return; + + call_rcu(&grp->rcu, sched_cache_group_free_rcu); +} + +void mm_destroy_sched(struct mm_struct *mm) +{ + sched_cache_group_put(mm->sched_cache_grp); + mm->sched_cache_grp = NULL; } /* because why would C be fully specified */ @@ -1729,11 +1780,16 @@ static unsigned long fraction_mm_sched(struct rq *rq, static int get_pref_llc(struct task_struct *p, struct mm_struct *mm) { int mm_sched_llc = -1, mm_sched_cpu; + struct sched_cache_group *grp; if (!mm) return -1; - mm_sched_cpu = READ_ONCE(mm->sc_stat.cpu); + grp = READ_ONCE(mm->sched_cache_grp); + if (!grp) + return -1; + + mm_sched_cpu = READ_ONCE(grp->cpu); if (mm_sched_cpu != -1) { mm_sched_llc = llc_id(mm_sched_cpu); @@ -1764,6 +1820,7 @@ static inline void account_mm_sched(struct rq *rq, struct task_struct *p, s64 delta_exec) { struct sched_cache_time *pcpu_sched; + struct sched_cache_group *grp; struct mm_struct *mm = p->mm; int mm_sched_llc = -1; unsigned long epoch; @@ -1777,10 +1834,14 @@ void account_mm_sched(struct rq *rq, struct task_struct *p, s64 delta_exec) * init_task, kthreads and user thread created * by user_mode_thread() don't have mm. */ - if (!mm || !mm->sc_stat.pcpu_sched) + if (!mm) return; - pcpu_sched = per_cpu_ptr(mm->sc_stat.pcpu_sched, cpu_of(rq)); + grp = READ_ONCE(mm->sched_cache_grp); + if (!grp || !grp->pcpu_sched) + return; + + pcpu_sched = per_cpu_ptr(grp->pcpu_sched, cpu_of(rq)); scoped_guard (raw_spinlock, &rq->cpu_epoch_lock) { __update_mm_sched(rq, pcpu_sched); @@ -1793,11 +1854,11 @@ void account_mm_sched(struct rq *rq, struct task_struct *p, s64 delta_exec) * If this process hasn't hit task_cache_work() for a while invalidate * its preferred state. */ - if ((long)(epoch - READ_ONCE(mm->sc_stat.epoch)) > llc_epoch_affinity_timeout || + if ((long)(epoch - READ_ONCE(grp->epoch)) > llc_epoch_affinity_timeout || invalid_llc_nr(mm, p, cpu_of(rq)) || exceed_llc_capacity(mm, cpu_of(rq))) { - if (READ_ONCE(mm->sc_stat.cpu) != -1) - WRITE_ONCE(mm->sc_stat.cpu, -1); + if (READ_ONCE(grp->cpu) != -1) + WRITE_ONCE(grp->cpu, -1); } mm_sched_llc = get_pref_llc(p, mm); @@ -1814,30 +1875,35 @@ void account_mm_sched(struct rq *rq, struct task_struct *p, s64 delta_exec) static void task_tick_cache(struct rq *rq, struct task_struct *p) { struct callback_head *work = &p->cache_work; + struct sched_cache_group *grp; struct mm_struct *mm = p->mm; unsigned long epoch; if (!sched_cache_enabled()) return; - if (!mm || p->flags & PF_KTHREAD || - !mm->sc_stat.pcpu_sched) + if (!mm || p->flags & PF_KTHREAD) + return; + + grp = READ_ONCE(mm->sched_cache_grp); + if (!grp || !grp->pcpu_sched) return; epoch = rq->cpu_epoch; /* avoid moving backwards */ - if (time_after_eq(mm->sc_stat.epoch, epoch)) + if (time_after_eq(grp->epoch, epoch)) return; - guard(raw_spinlock)(&mm->sc_stat.lock); + guard(raw_spinlock)(&grp->lock); if (work->next == work) { task_work_add(p, work, TWA_RESUME); - WRITE_ONCE(mm->sc_stat.epoch, epoch); + WRITE_ONCE(grp->epoch, epoch); } } -static void get_scan_cpumasks(cpumask_var_t cpus, struct task_struct *p) +static void get_scan_cpumasks(cpumask_var_t cpus, struct task_struct *p, + struct sched_cache_group *grp) { #ifdef CONFIG_NUMA_BALANCING int cpu, curr_cpu, nid, pref_nid; @@ -1845,7 +1911,7 @@ static void get_scan_cpumasks(cpumask_var_t cpus, struct task_struct *p) if (!static_branch_likely(&sched_numa_balancing)) goto out; - cpu = READ_ONCE(p->mm->sc_stat.cpu); + cpu = READ_ONCE(grp->cpu); if (cpu != -1) nid = cpu_to_node(cpu); curr_cpu = task_cpu(p); @@ -1906,6 +1972,7 @@ static void task_cache_work(struct callback_head *work) unsigned long next_scan, now = jiffies; struct task_struct *p = current, *cur; unsigned long curr_m_a_occ = 0; + struct sched_cache_group *grp; struct mm_struct *mm = p->mm; unsigned long m_a_occ = 0; cpumask_var_t cpus; @@ -1917,12 +1984,16 @@ static void task_cache_work(struct callback_head *work) if (p->flags & PF_EXITING) return; - next_scan = READ_ONCE(mm->sc_stat.next_scan); + grp = READ_ONCE(mm->sched_cache_grp); + if (!grp) + return; + + next_scan = READ_ONCE(grp->next_scan); if (time_before(now, next_scan)) return; /* only 1 thread is allowed to scan */ - if (!try_cmpxchg(&mm->sc_stat.next_scan, &next_scan, + if (!try_cmpxchg(&grp->next_scan, &next_scan, now + max_t(unsigned long, READ_ONCE(llc_epoch_period), 1))) return; @@ -1930,8 +2001,8 @@ static void task_cache_work(struct callback_head *work) curr_cpu = task_cpu(p); if (invalid_llc_nr(mm, p, curr_cpu) || exceed_llc_capacity(mm, curr_cpu)) { - if (READ_ONCE(mm->sc_stat.cpu) != -1) - WRITE_ONCE(mm->sc_stat.cpu, -1); + if (READ_ONCE(grp->cpu) != -1) + WRITE_ONCE(grp->cpu, -1); return; } @@ -1942,7 +2013,7 @@ static void task_cache_work(struct callback_head *work) scoped_guard (cpus_read_lock) { guard(rcu)(); - get_scan_cpumasks(cpus, p); + get_scan_cpumasks(cpus, p, grp); for_each_cpu(cpu, cpus) { /* XXX sched_cluster_active */ @@ -1955,7 +2026,7 @@ static void task_cache_work(struct callback_head *work) for_each_cpu(i, sched_domain_span(sd)) { occ = fraction_mm_sched(cpu_rq(i), - per_cpu_ptr(mm->sc_stat.pcpu_sched, i)); + per_cpu_ptr(grp->pcpu_sched, i)); a_occ += occ; if (occ > m_occ) { m_occ = occ; @@ -1988,7 +2059,7 @@ static void task_cache_work(struct callback_head *work) m_a_cpu = m_cpu; } - if (llc_id(cpu) == llc_id(READ_ONCE(mm->sc_stat.cpu))) + if (llc_id(cpu) == llc_id(READ_ONCE(grp->cpu))) curr_m_a_occ = a_occ; cpumask_andnot(cpus, cpus, sched_domain_span(sd)); @@ -1997,7 +2068,7 @@ static void task_cache_work(struct callback_head *work) if (m_a_occ > (2 * curr_m_a_occ)) { /* - * Avoid switching sc_stat.cpu too fast. + * Avoid switching sched_cache_grp->cpu too fast. * The reason to choose 2X is because: * 1. It is better to keep the preferred LLC stable, * rather than changing it frequently and cause migrations @@ -2006,10 +2077,10 @@ static void task_cache_work(struct callback_head *work) * 3. 2X is chosen based on test results, as it delivers * the optimal performance gain so far. */ - WRITE_ONCE(mm->sc_stat.cpu, m_a_cpu); + WRITE_ONCE(grp->cpu, m_a_cpu); } - update_avg_scale(&mm->sc_stat.nr_running_avg, nr_running); + update_avg_scale(&grp->nr_running_avg, nr_running); free_cpumask_var(cpus); } @@ -3726,6 +3797,7 @@ static int preferred_group_nid(struct task_struct *p, int nid) static void task_numa_placement(struct task_struct *p) __context_unsafe(/* conditional locking */) { + struct sched_cache_group __maybe_unused *grp; int seq, nid, max_nid = NUMA_NO_NODE; unsigned long max_faults = 0; unsigned long fault_types[2] = { 0, 0 }; @@ -3818,19 +3890,23 @@ static void task_numa_placement(struct task_struct *p) * heuristic and occasional lost updates are tolerable. * * If a task exits, its corresponding footprint must - * be subtracted from the mm->sc_stat.footprint, otherwise - * the mm->sc_stat.footprint will not converge: - * the exiting thread's footprint remains unchanged/undecayed - * in mm->sc_stat.footprint. See exit_mm(). + * be subtracted from the mm->sched_cache_grp->footprint, + * otherwise the mm->sched_cache_grp->footprint will not + * converge: the exiting thread's footprint remains + * unchanged/undecayed in mm->sched_cache_grp->footprint. + * See exit_mm(). * * Lost updates and unsynchronized subtraction * in exit_mm() can cause footprint + diff to * go negative. Clamp to zero to prevent the * unsigned footprint from wrapping. */ - new_fp = (long)READ_ONCE(p->mm->sc_stat.footprint) + diff; - WRITE_ONCE(p->mm->sc_stat.footprint, - max(new_fp, 0L)); + grp = READ_ONCE(p->mm->sched_cache_grp); + if (!grp) + continue; + + new_fp = (long)READ_ONCE(grp->footprint) + diff; + WRITE_ONCE(grp->footprint, max(new_fp, 0L)); #endif } @@ -10778,6 +10854,7 @@ static inline bool task_misfits_asym_cpu(struct lb_env *env, struct task_struct static enum llc_mig can_migrate_llc_task(struct lb_env *env, struct task_struct *p) { + struct sched_cache_group *grp; struct mm_struct *mm; bool to_pref; int cpu, src_cpu, dst_cpu; @@ -10791,15 +10868,19 @@ static enum llc_mig can_migrate_llc_task(struct lb_env *env, if (!mm) return mig_unrestricted; - cpu = READ_ONCE(mm->sc_stat.cpu); + grp = READ_ONCE(mm->sched_cache_grp); + if (!grp) + return mig_unrestricted; + + cpu = READ_ONCE(grp->cpu); if (cpu < 0 || cpus_share_cache(src_cpu, dst_cpu)) return mig_unrestricted; /* skip cache aware load balance for too many threads */ if (invalid_llc_nr(mm, p, dst_cpu) || exceed_llc_capacity(mm, dst_cpu)) { - if (READ_ONCE(mm->sc_stat.cpu) != -1) - WRITE_ONCE(mm->sc_stat.cpu, -1); + if (READ_ONCE(grp->cpu) != -1) + WRITE_ONCE(grp->cpu, -1); return mig_unrestricted; } From b636fef85bda7d1bab9c0a45067ab1508d79d946 Mon Sep 17 00:00:00 2001 From: Tim Chen Date: Mon, 21 Sep 2026 17:37:25 -0700 Subject: [PATCH 332/694] sched/cache: Introduce task_struct->sched_cache_grp to fix UAF Add a sched_cache_grp pointer to task_struct so that scheduler code can access the cache group directly via the task, without going through mm->sched_cache_grp. This decouples the scheduler's hot-path accesses from the mm_struct. Each task holds its own refcount on the sched_cache_group, separate from the reference held by its mm_struct. The reference is acquired in copy_mm() (fork) and exec_mmap() (exec), and released in exit_mm(). This fixes the use-after-free when account_mm_sched() reaches the group through a task whose mm is being switched, as reported by Hyunwoo: https://lore.kernel.org/lkml/apPb-Dr4nPYuHQOK@v4bel/ Convert all scheduler code in fair.c and exit.c to use p->sched_cache_grp instead of p->mm->sched_cache_grp. Keep the fork/exec/exit reference management out of the generic mm paths: add sched_cache_fork(), sched_cache_fork_cleanup(), sched_cache_exec_mmap() and sched_cache_exit_mm() in kernel/sched/cache_sched.c (with empty stubs for !CONFIG_SCHED_CACHE), so fs/exec.c, kernel/fork.c and kernel/exit.c each call one helper instead of open-coding the refcounting under #ifdef. Also add sched_cache_group_get() and task_cache_group_get(). Fixes: df0d98475954 ("sched/cache: Introduce infrastructure for cache-aware load balancing") Closes: https://lore.kernel.org/lkml/apPb-Dr4nPYuHQOK@v4bel/ Closes: https://lore.kernel.org/all/343a7e07-7fad-4979-9c9b-82ec038c293c@linux.dev/ Reported-by: Hyunwoo Kim Reported-by: Zenghui Yu (Huawei) Co-developed-by: Chen Yu Signed-off-by: Chen Yu Signed-off-by: Tim Chen Signed-off-by: Peter Zijlstra (Intel) Signed-off-by: Ingo Molnar Cc: #7.2.x Link: https://patch.msgid.link/ae7081dc54736bf115215f9867abb2711a7403fb.1790035273.git.tim.c.chen@linux.intel.com --- fs/exec.c | 1 + include/linux/sched.h | 14 +++ kernel/exit.c | 33 +------ kernel/fork.c | 2 + kernel/sched/fair.c | 196 +++++++++++++++++++++++++++++------------- 5 files changed, 154 insertions(+), 92 deletions(-) diff --git a/fs/exec.c b/fs/exec.c index 819643408e6d..a5269b5e00df 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -882,6 +882,7 @@ static int exec_mmap(struct linux_binprm *bprm) active_mm = tsk->active_mm; tsk->active_mm = mm; tsk->mm = mm; + sched_cache_exec_mmap(tsk, mm); mm_init_cid(mm, tsk); exec_state = task_exec_state_replace(tsk, exec_state); /* diff --git a/include/linux/sched.h b/include/linux/sched.h index e14ad43522c8..d35ae49a991f 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -1433,6 +1433,7 @@ struct task_struct { #ifdef CONFIG_SCHED_CACHE struct callback_head cache_work; + struct sched_cache_group __rcu *sched_cache_grp; int preferred_llc; /* 1: task was enqueued to its preferred LLC, 0 otherwise */ int pref_llc_queued; @@ -2417,10 +2418,23 @@ struct sched_cache_group { struct rcu_head rcu; } ____cacheline_aligned_in_smp; +struct sched_cache_group *sched_cache_group_get(struct sched_cache_group *grp); +struct sched_cache_group *task_cache_group_get(struct task_struct *p); + +void sched_cache_fork(struct task_struct *p); +void sched_cache_fork_cleanup(struct task_struct *p); +void sched_cache_exec_mmap(struct task_struct *p, struct mm_struct *mm); +void sched_cache_exit_mm(struct task_struct *p); + #else struct sched_cache_group { }; +static inline void sched_cache_fork(struct task_struct *p) { } +static inline void sched_cache_fork_cleanup(struct task_struct *p) { } +static inline void sched_cache_exec_mmap(struct task_struct *p, struct mm_struct *mm) { } +static inline void sched_cache_exit_mm(struct task_struct *p) { } + #endif #ifndef MODULE diff --git a/kernel/exit.c b/kernel/exit.c index 024350e9b48c..282328d2b4cf 100644 --- a/kernel/exit.c +++ b/kernel/exit.c @@ -551,37 +551,6 @@ void mm_update_next_owner(struct mm_struct *mm) } #endif /* CONFIG_MEMCG */ -#if defined(CONFIG_SCHED_CACHE) && defined(CONFIG_NUMA_BALANCING) -/* - * Subtract the memory footprint of the current task from - * mm. - */ -static void exit_mm_sched_cache(struct mm_struct *mm) -{ - struct sched_cache_group *grp; - unsigned long fp, sub; - - if (!current->total_numa_faults) - return; - /* - * No lock protection due to performance considerations. - * Make sure the group footprint does not become - * negative. - */ - grp = READ_ONCE(mm->sched_cache_grp); - if (!grp) - return; - - fp = READ_ONCE(grp->footprint); - sub = min(fp, current->total_numa_faults); - WRITE_ONCE(grp->footprint, fp - sub); -} -#else -static inline void exit_mm_sched_cache(struct mm_struct *mm) -{ -} -#endif /* CONFIG_SCHED_CACHE CONFIG_NUMA_BALANCING */ - /* * Turn us into a lazy TLB process if we * aren't already.. @@ -594,7 +563,7 @@ static void exit_mm(void) if (!mm) return; - exit_mm_sched_cache(mm); + sched_cache_exit_mm(current); mmap_read_lock(mm); mmgrab_lazy_tlb(mm); diff --git a/kernel/fork.c b/kernel/fork.c index 5ef413368912..10f2d05d816a 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -1599,6 +1599,7 @@ static int copy_mm(u64 clone_flags, struct task_struct *tsk) tsk->mm = mm; tsk->active_mm = mm; + sched_cache_fork(tsk); return 0; } @@ -2602,6 +2603,7 @@ __latent_entropy struct task_struct *copy_process( bad_fork_cleanup_namespaces: exit_nsproxy_namespaces(p); bad_fork_cleanup_mm: + sched_cache_fork_cleanup(p); if (p->mm) { mm_clear_owner(p->mm, p); mmput(p->mm); diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index f0a9586332fd..974a7dfe3215 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -1478,7 +1478,7 @@ static inline int get_sched_cache_scale(int mul) return (1 + (tol - 1) * mul); } -static bool exceed_llc_capacity(struct mm_struct *mm, int cpu) +static bool exceed_llc_capacity(struct sched_cache_group *grp, int cpu) { #ifdef CONFIG_NUMA_BALANCING unsigned long llc, footprint; @@ -1492,11 +1492,6 @@ static bool exceed_llc_capacity(struct mm_struct *mm, int cpu) return true; if (static_branch_likely(&sched_numa_balancing)) { - struct sched_cache_group *grp = READ_ONCE(mm->sched_cache_grp); - - if (!grp) - return true; - /* * TBD: RDT exclusive LLC ways reserved should be * excluded. @@ -1531,10 +1526,9 @@ static bool exceed_llc_capacity(struct mm_struct *mm, int cpu) return false; } -static bool invalid_llc_nr(struct mm_struct *mm, struct task_struct *p, +static bool invalid_llc_nr(struct sched_cache_group *grp, struct task_struct *p, int cpu) { - struct sched_cache_group *grp; int scale; if (get_nr_threads(p) <= 1) @@ -1548,10 +1542,6 @@ static bool invalid_llc_nr(struct mm_struct *mm, struct task_struct *p, if (scale == INT_MAX) return false; - grp = READ_ONCE(mm->sched_cache_grp); - if (!grp) - return true; - return !fits_capacity((READ_ONCE(grp->nr_running_avg) * cpu_smt_num_threads), (scale * per_cpu(sd_llc_size, cpu))); } @@ -1723,6 +1713,96 @@ static void sched_cache_group_put(struct sched_cache_group *grp) call_rcu(&grp->rcu, sched_cache_group_free_rcu); } +DEFINE_FREE(sched_cache_group_put, struct sched_cache_group *, + sched_cache_group_put(_T)); + +#define rcu_deref_sched_cache_grp(tsk) \ + rcu_dereference_check((tsk)->sched_cache_grp, (tsk) == current) + +static struct sched_cache_group *sched_cache_replace_grp(struct task_struct *p, + struct sched_cache_group *new) +{ + struct sched_cache_group *old; + + old = rcu_deref_sched_cache_grp(p); + rcu_assign_pointer(p->sched_cache_grp, new); + + return old; +} + +struct sched_cache_group *sched_cache_group_get(struct sched_cache_group *grp) +{ + /* + * refcount_inc_not_zero() is the acquire primitive for lockless + * (RCU) lookups; plain refcount_inc() would scribble the count if + * it already reached zero. Return NULL in that case. + */ + if (grp && !refcount_inc_not_zero(&grp->refcnt)) + grp = NULL; + + return grp; +} + +struct sched_cache_group *task_cache_group_get(struct task_struct *p) +{ + guard(rcu)(); + return sched_cache_group_get(rcu_dereference(p->sched_cache_grp)); +} + +void sched_cache_fork(struct task_struct *p) +{ + /* + * The child takes its own reference on the mm's cache group, separate + * from the reference held by the mm. @p is not yet visible to readers, + * so a plain initializing store is enough. + */ + RCU_INIT_POINTER(p->sched_cache_grp, + sched_cache_group_get(p->mm->sched_cache_grp)); +} + +void sched_cache_fork_cleanup(struct task_struct *p) +{ + /* + * A fork that fails after sched_cache_fork() never reaches exit_mm(), + * so drop the reference here. @p never became visible, so there are no + * concurrent readers and the reference we hold keeps the group alive. + */ + sched_cache_group_put(rcu_access_pointer(p->sched_cache_grp)); + RCU_INIT_POINTER(p->sched_cache_grp, NULL); +} + +void sched_cache_exec_mmap(struct task_struct *p, struct mm_struct *mm) +{ + struct sched_cache_group *old; + + /* + * Acquire the new reference before publishing the pointer, then drop + * the old one. @p is current and the only writer of its own pointer. + */ + old = sched_cache_replace_grp(p, sched_cache_group_get(mm->sched_cache_grp)); + sched_cache_group_put(old); +} + +void sched_cache_exit_mm(struct task_struct *p) +{ + struct sched_cache_group *grp = sched_cache_replace_grp(p, NULL); + +#ifdef CONFIG_NUMA_BALANCING + /* + * Subtract this task's footprint from the group before dropping the + * reference, so the group footprint converges as its threads exit. + * Unlocked for performance; clamp to avoid underflow. + */ + if (grp && p->total_numa_faults) { + unsigned long fp = READ_ONCE(grp->footprint); + unsigned long sub = min(fp, p->total_numa_faults); + + WRITE_ONCE(grp->footprint, fp - sub); + } +#endif + sched_cache_group_put(grp); +} + void mm_destroy_sched(struct mm_struct *mm) { sched_cache_group_put(mm->sched_cache_grp); @@ -1777,15 +1857,10 @@ static unsigned long fraction_mm_sched(struct rq *rq, return div64_u64(NICE_0_LOAD * pcpu_sched->runtime, rq->cpu_runtime + 1); } -static int get_pref_llc(struct task_struct *p, struct mm_struct *mm) +static int get_pref_llc(struct task_struct *p, struct sched_cache_group *grp) { int mm_sched_llc = -1, mm_sched_cpu; - struct sched_cache_group *grp; - if (!mm) - return -1; - - grp = READ_ONCE(mm->sched_cache_grp); if (!grp) return -1; @@ -1819,9 +1894,8 @@ static unsigned int task_running_on_cpu(int cpu, struct task_struct *p); static inline void account_mm_sched(struct rq *rq, struct task_struct *p, s64 delta_exec) { + struct sched_cache_group *grp = rcu_dereference_all(p->sched_cache_grp); struct sched_cache_time *pcpu_sched; - struct sched_cache_group *grp; - struct mm_struct *mm = p->mm; int mm_sched_llc = -1; unsigned long epoch; @@ -1832,12 +1906,8 @@ void account_mm_sched(struct rq *rq, struct task_struct *p, s64 delta_exec) return; /* * init_task, kthreads and user thread created - * by user_mode_thread() don't have mm. + * by user_mode_thread() don't have a cache group. */ - if (!mm) - return; - - grp = READ_ONCE(mm->sched_cache_grp); if (!grp || !grp->pcpu_sched) return; @@ -1855,13 +1925,13 @@ void account_mm_sched(struct rq *rq, struct task_struct *p, s64 delta_exec) * its preferred state. */ if ((long)(epoch - READ_ONCE(grp->epoch)) > llc_epoch_affinity_timeout || - invalid_llc_nr(mm, p, cpu_of(rq)) || - exceed_llc_capacity(mm, cpu_of(rq))) { + invalid_llc_nr(grp, p, cpu_of(rq)) || + exceed_llc_capacity(grp, cpu_of(rq))) { if (READ_ONCE(grp->cpu) != -1) WRITE_ONCE(grp->cpu, -1); } - mm_sched_llc = get_pref_llc(p, mm); + mm_sched_llc = get_pref_llc(p, grp); /* task not on rq accounted later in account_entity_enqueue() */ if (task_running_on_cpu(rq->cpu, p) && @@ -1874,19 +1944,15 @@ void account_mm_sched(struct rq *rq, struct task_struct *p, s64 delta_exec) static void task_tick_cache(struct rq *rq, struct task_struct *p) { + struct sched_cache_group *grp = rcu_dereference_all(p->sched_cache_grp); struct callback_head *work = &p->cache_work; - struct sched_cache_group *grp; - struct mm_struct *mm = p->mm; unsigned long epoch; if (!sched_cache_enabled()) return; - if (!mm || p->flags & PF_KTHREAD) - return; - - grp = READ_ONCE(mm->sched_cache_grp); - if (!grp || !grp->pcpu_sched) + if (!grp || p->flags & PF_KTHREAD || + !grp->pcpu_sched) return; epoch = rq->cpu_epoch; @@ -1968,14 +2034,13 @@ static inline void update_avg_scale(u64 *avg, u64 sample) static void task_cache_work(struct callback_head *work) { + struct sched_cache_group *grp __free(sched_cache_group_put) = NULL; + cpumask_var_t cpus __free(free_cpumask_var) = CPUMASK_VAR_NULL; int cpu, m_a_cpu = -1, nr_running = 0, curr_cpu; unsigned long next_scan, now = jiffies; struct task_struct *p = current, *cur; unsigned long curr_m_a_occ = 0; - struct sched_cache_group *grp; - struct mm_struct *mm = p->mm; unsigned long m_a_occ = 0; - cpumask_var_t cpus; WARN_ON_ONCE(work != &p->cache_work); @@ -1984,7 +2049,12 @@ static void task_cache_work(struct callback_head *work) if (p->flags & PF_EXITING) return; - grp = READ_ONCE(mm->sched_cache_grp); + /* + * A reference makes sure grp is not released by others. The rcu + * lock can not be held till after zalloc_cpumask_var() below, + * because the latter might sleep. + */ + grp = task_cache_group_get(p); if (!grp) return; @@ -1999,8 +2069,8 @@ static void task_cache_work(struct callback_head *work) return; curr_cpu = task_cpu(p); - if (invalid_llc_nr(mm, p, curr_cpu) || - exceed_llc_capacity(mm, curr_cpu)) { + if (invalid_llc_nr(grp, p, curr_cpu) || + exceed_llc_capacity(grp, curr_cpu)) { if (READ_ONCE(grp->cpu) != -1) WRITE_ONCE(grp->cpu, -1); @@ -2033,9 +2103,13 @@ static void task_cache_work(struct callback_head *work) m_cpu = i; } + /* + * rcu_access_pointer() is used because the + * pointer is only compared, never dereferenced. + */ cur = rcu_dereference_all(cpu_rq(i)->curr); if (cur && !(cur->flags & (PF_EXITING | PF_KTHREAD)) && - cur->mm == mm) + rcu_access_pointer(cur->sched_cache_grp) == grp) nr_running++; } @@ -2081,7 +2155,6 @@ static void task_cache_work(struct callback_head *work) } update_avg_scale(&grp->nr_running_avg, nr_running); - free_cpumask_var(cpus); } void init_sched_mm(struct task_struct *p) @@ -2090,6 +2163,13 @@ void init_sched_mm(struct task_struct *p) init_task_work(work, task_cache_work); work->next = work; + /* + * dup_task_struct() copies the parent's task_struct, including its + * sched_cache_grp, for which the child holds no reference. Clear it + * here - before copy_mm() runs - so the child never carries a + * borrowed pointer that the fork error path would put. + */ + RCU_INIT_POINTER(p->sched_cache_grp, NULL); /* * Reset new task's preference to avoid * polluting account_llc_enqueue(). @@ -3890,10 +3970,9 @@ static void task_numa_placement(struct task_struct *p) * heuristic and occasional lost updates are tolerable. * * If a task exits, its corresponding footprint must - * be subtracted from the mm->sched_cache_grp->footprint, - * otherwise the mm->sched_cache_grp->footprint will not - * converge: the exiting thread's footprint remains - * unchanged/undecayed in mm->sched_cache_grp->footprint. + * be subtracted from p->sched_cache_grp->footprint, + * otherwise the footprint will not converge: the + * exiting thread's footprint remains unchanged/undecayed. * See exit_mm(). * * Lost updates and unsynchronized subtraction @@ -3901,12 +3980,14 @@ static void task_numa_placement(struct task_struct *p) * go negative. Clamp to zero to prevent the * unsigned footprint from wrapping. */ - grp = READ_ONCE(p->mm->sched_cache_grp); - if (!grp) - continue; + scoped_guard(rcu) { + grp = rcu_dereference(p->sched_cache_grp); - new_fp = (long)READ_ONCE(grp->footprint) + diff; - WRITE_ONCE(grp->footprint, max(new_fp, 0L)); + if (grp) { + new_fp = (long)READ_ONCE(grp->footprint) + diff; + WRITE_ONCE(grp->footprint, max(new_fp, 0L)); + } + } #endif } @@ -10855,7 +10936,6 @@ static enum llc_mig can_migrate_llc_task(struct lb_env *env, struct task_struct *p) { struct sched_cache_group *grp; - struct mm_struct *mm; bool to_pref; int cpu, src_cpu, dst_cpu; @@ -10864,11 +10944,7 @@ static enum llc_mig can_migrate_llc_task(struct lb_env *env, src_cpu = env->src_cpu; dst_cpu = env->dst_cpu; - mm = p->mm; - if (!mm) - return mig_unrestricted; - - grp = READ_ONCE(mm->sched_cache_grp); + grp = rcu_dereference_all(p->sched_cache_grp); if (!grp) return mig_unrestricted; @@ -10877,8 +10953,8 @@ static enum llc_mig can_migrate_llc_task(struct lb_env *env, return mig_unrestricted; /* skip cache aware load balance for too many threads */ - if (invalid_llc_nr(mm, p, dst_cpu) || - exceed_llc_capacity(mm, dst_cpu)) { + if (invalid_llc_nr(grp, p, dst_cpu) || + exceed_llc_capacity(grp, dst_cpu)) { if (READ_ONCE(grp->cpu) != -1) WRITE_ONCE(grp->cpu, -1); return mig_unrestricted; From 65efcccddc83d6a19e8a2e2a6117811e39e589bf Mon Sep 17 00:00:00 2001 From: Chen Yu Date: Mon, 21 Sep 2026 17:37:26 -0700 Subject: [PATCH 333/694] sched/cache: Skip kernel threads for cache aware scheduling to rubustify the code Kernel thread should not be covered by cache aware scheduling as it borrows the statistics from the user space thread. Filter the kernel thread in account_mm_sched(). In theory a kernel thread does not have any valid cache group, so !grp should gate the kernel thread. Add the PF_KTHREAD check explicitly here for safety reasons, to guard against future modifications and to pair with task_tick_cache(). Fixes: df0d98475954 ("sched/cache: Introduce infrastructure for cache-aware load balancing") Signed-off-by: Chen Yu Signed-off-by: Tim Chen Signed-off-by: Peter Zijlstra (Intel) Signed-off-by: Ingo Molnar Cc: # 7.2.x Link: https://patch.msgid.link/058f0c6ea7b991c177a17de347fa3157f25489a7.1790035273.git.tim.c.chen@linux.intel.com --- kernel/sched/fair.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index 974a7dfe3215..57360f5cdde4 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -1907,8 +1907,14 @@ void account_mm_sched(struct rq *rq, struct task_struct *p, s64 delta_exec) /* * init_task, kthreads and user thread created * by user_mode_thread() don't have a cache group. + * In theory a kernel thread does not have any valid + * cache group, because sched_cache_fork() is not + * invoked for a kernel thread - !grp should gate the + * kernel thread. Use the PF_KTHREAD check explicitly + * here for safety reasons, to guard against future + * modifications and to pair with task_tick_cache(). */ - if (!grp || !grp->pcpu_sched) + if (p->flags & PF_KTHREAD || !grp || !grp->pcpu_sched) return; pcpu_sched = per_cpu_ptr(grp->pcpu_sched, cpu_of(rq)); From 3cb0243767fd033bdce95f4f1b5882172a2f8119 Mon Sep 17 00:00:00 2001 From: Davi Chaves Azevedo Date: Mon, 21 Sep 2026 17:37:27 -0700 Subject: [PATCH 334/694] sched/cache: Refresh LLC capacity across CPU hotplug, to fix capacity underestimation bug The scheduler scales LLC capacity by the fraction of cache-sharing CPUs covered by a domain: llc_bytes = cache_size * span_weight / shared_weight During CPU teardown, sched_cpu_deactivate() rebuilds scheduler domains before cacheinfo_cpu_pre_down() removes the CPU from shared_cpu_map. The new domains therefore use the old sharing weight. The later call to sched_update_llc_bytes() looks up the departing CPU's sd_llc, which has already been detached, and returns without correcting the surviving CPUs. On a Ryzen 5 7535U with twelve logical CPUs sharing a 16 MiB LLC, offlining one SMT sibling left the remaining CPUs with: llc_bytes = floor(16777216 * 11 / 12) = 15379114 bytes The correct capacity is still 16777216 bytes. On systems with active cache-aware scheduling, an underestimated capacity can cause exceed_llc_capacity() to reject aggregation for a process whose footprint would fit. Unchanged cpuset partitions sharing the physical cache can also retain stale capacity when a CPU comes online in another partition. Pass the cache-sharing mask already retained by cacheinfo to the scheduler update. Refresh every surviving CPU using its own LLC domain so that each partition receives the correct share. This also preserves the correction needed as cache-sharing maps grow during boot. Keep the existing CPU-hotplug and scheduler-domain synchronization. The update remains on the hotplug path; no steady-state scheduling operation or persistent allocation is added. Fixes: 7030513a0877 ("sched/cache: Calculate the LLC size and store it in sched_domain") Signed-off-by: Davi Chaves Azevedo Signed-off-by: Tim Chen Signed-off-by: Peter Zijlstra (Intel) Signed-off-by: Ingo Molnar Reviewed-by: Chen Yu Reviewed-by: Tim Chen Reviewed-by: K Prateek Nayak Tested-by: Chen Yu Tested-by: K Prateek Nayak Cc: # v7.2.x Link: https://patch.msgid.link/6751d93e15889e624796c74db0bfe66603d60b1b.1790035273.git.tim.c.chen@linux.intel.com --- drivers/base/cacheinfo.c | 11 ++++++----- include/linux/sched/topology.h | 4 ++-- kernel/sched/topology.c | 22 +++++++++++++--------- 3 files changed, 21 insertions(+), 16 deletions(-) diff --git a/drivers/base/cacheinfo.c b/drivers/base/cacheinfo.c index 9f9c72727a05..7a47a392568a 100644 --- a/drivers/base/cacheinfo.c +++ b/drivers/base/cacheinfo.c @@ -1040,9 +1040,10 @@ static int cacheinfo_cpu_online(unsigned int cpu) rc = cache_add_dev(cpu); if (rc) goto err; - if (cpu_map_shared_cache(true, cpu, &cpu_map)) + if (cpu_map_shared_cache(true, cpu, &cpu_map)) { update_per_cpu_data_slice_size(true, cpu, cpu_map); - sched_update_llc_bytes(cpu); + sched_update_llc_bytes(cpu_map); + } return 0; err: free_cache_attributes(cpu); @@ -1059,10 +1060,10 @@ static int cacheinfo_cpu_pre_down(unsigned int cpu) cpu_cache_sysfs_exit(cpu); free_cache_attributes(cpu); - if (nr_shared > 1) + if (nr_shared > 1) { update_per_cpu_data_slice_size(false, cpu, cpu_map); - - sched_update_llc_bytes(cpu); + sched_update_llc_bytes(cpu_map); + } return 0; } diff --git a/include/linux/sched/topology.h b/include/linux/sched/topology.h index b5d9d7c2b8ad..f96812d71c51 100644 --- a/include/linux/sched/topology.h +++ b/include/linux/sched/topology.h @@ -281,9 +281,9 @@ static inline int task_node(const struct task_struct *p) } #ifdef CONFIG_SCHED_CACHE -extern void sched_update_llc_bytes(unsigned int cpu); +extern void sched_update_llc_bytes(const struct cpumask *cpus); #else -static inline void sched_update_llc_bytes(unsigned int cpu) { } +static inline void sched_update_llc_bytes(const struct cpumask *cpus) { } #endif #endif /* _LINUX_SCHED_TOPOLOGY_H */ diff --git a/kernel/sched/topology.c b/kernel/sched/topology.c index 0248227d983a..3dab0253976f 100644 --- a/kernel/sched/topology.c +++ b/kernel/sched/topology.c @@ -985,8 +985,8 @@ void sched_cache_active_set(void) } /* - * Update the bottom sched_domain's llc_bytes for @cpu and all its - * LLC siblings. Called from cacheinfo_cpu_online() or + * Update the bottom sched_domain's llc_bytes for @cpus sharing a physical + * LLC. Called from cacheinfo_cpu_online() or * cacheinfo_cpu_pre_down() with cpu hotplug lock held. * * Note: get_effective_llc_bytes() returns 0 on PowerPC. @@ -996,17 +996,13 @@ void sched_cache_active_set(void) * and does not populates the per-CPU struct cpu_cacheinfo array * that get_cpu_cacheinfo_llc() reads. */ -void sched_update_llc_bytes(unsigned int cpu) +void sched_update_llc_bytes(const struct cpumask *cpus) { struct sched_domain *sd, *sdp; unsigned int i; sched_domains_mutex_lock(); - sdp = rcu_dereference_sched_domain(per_cpu(sd_llc, cpu)); - if (!sdp) - goto unlock; - /* * ci->shared_cpu_map is built incrementally as CPUs come * online, so the first CPU in an LLC initially sees @@ -1014,14 +1010,22 @@ void sched_update_llc_bytes(unsigned int cpu) * get_effective_llc_bytes(). Re-evaluating every LLC * sibling on each online event corrects this once the full * shared_cpu_map is known. + * + * The departing CPU's domains have already been detached when + * cacheinfo removes it. Use the surviving cache siblings instead. + * They may belong to different cpuset partitions, so use each CPU's + * own LLC domain to scale its share of the physical cache. */ - for_each_cpu(i, sched_domain_span(sdp)) { + for_each_cpu(i, cpus) { + sdp = rcu_dereference_sched_domain(per_cpu(sd_llc, i)); + if (!sdp) + continue; + sd = rcu_dereference_sched_domain(cpu_rq(i)->sd); if (sd) sd->llc_bytes = get_effective_llc_bytes(i, sdp); } -unlock: sched_domains_mutex_unlock(); } From 5fd0783b99d4af98f65cd58b56ec203d1d426104 Mon Sep 17 00:00:00 2001 From: Shardul Bankar Date: Thu, 17 Sep 2026 14:55:32 +0530 Subject: [PATCH 335/694] udp: relocate a connected socket in the 4-tuple hash table on re-connect A connected UDP socket that connects again to a different peer is not re-filed in the 4-tuple hash table: sk binds to 127.0.0.1:21001 sk connects to 127.0.0.2:20001 // filed under hash(sk, peer1) sk connects to 127.0.0.3:20002 // still filed under hash(sk, peer1) packet from 127.0.0.3:20002 // hash(sk, peer2) misses, so the // lookup falls back to scoring the // hash2 chain for this address // and port udp_lib_hash4() returns early when the socket is already hashed, assuming ->rehash() relocates it. ->rehash() runs from __ip{4,6}_datagram_connect() only while the receive address is unset, which a second connect never is: the first connect assigns it, whether the socket was bound to a specific address or to the wildcard. commit 644f9108f3a5 ("udp: Make rehash4 independent in udp_lib_rehash()") added that early return and named connect(AF_UNSPEC) as the way around it. That workaround does not help a socket with both SOCK_BINDADDR_LOCK and SOCK_BINDPORT_LOCK set, because __udp_disconnect() skips ->rehash() for the first and ->unhash() for the second. Delivery is correct either way. Relocate the socket when the hash it is filed under differs from the one requested, which is what commit 78c91ae2c6de ("ipv4/udp: Add 4-tuple hash for connected socket") did before the early return became unconditional. It is done here under hslot->lock, which that version did not take, to match udp_lib_rehash() and udp_lib_unhash(). hslot2 is unchanged, so hash4_cnt needs no adjustment, as in udp_lib_rehash(). A first connect is unaffected, and IPv6 shares the code. With 500 sockets on the port, a re-connected socket measured 522,553 pps without this change and 2,055,078 with it. The UDP side was noted as remaining work in [1]. Link: https://lore.kernel.org/netdev/apnHqmYZQ4yzOP4N@v4bel/ [1] Fixes: 644f9108f3a5 ("udp: Make rehash4 independent in udp_lib_rehash()") Assisted-by: LLM Signed-off-by: Shardul Bankar Reviewed-by: Kuniyuki Iwashima Link: https://patch.msgid.link/20260917-udp_hash4_fix_v1-v1-1-718891af0d7a@mpiricsoftware.com Signed-off-by: Paolo Abeni --- net/ipv4/udp.c | 21 +++++++++++++++------ 1 file changed, 15 insertions(+), 6 deletions(-) diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c index bb8cfc62cb00..0fa3cdbdcc21 100644 --- a/net/ipv4/udp.c +++ b/net/ipv4/udp.c @@ -617,14 +617,23 @@ void udp_lib_hash4(struct sock *sk, u16 hash) struct net *net = sock_net(sk); struct udp_table *udptable; - /* Connected udp socket can re-connect to another remote address, which - * will be handled by rehash. Thus no need to redo hash4 here. - */ - if (udp_hashed4(sk)) - return; - udptable = net->ipv4.udp_table; hslot = udp_hashslot(udptable, net, udp_sk(sk)->udp_port_hash); + + /* A connected socket can re-connect to another address. rehash() + * relocates it, but only runs when the local address changes, so a + * socket bound to a specific address would stay filed under the + * previous peer's hash. Move it here. + */ + if (udp_hashed4(sk)) { + if (udp_sk(sk)->udp_lrpa_hash != hash) { + spin_lock_bh(&hslot->lock); + udp_rehash4(udptable, sk, hash); + spin_unlock_bh(&hslot->lock); + } + return; + } + hslot2 = udp_hashslot2(udptable, udp_sk(sk)->udp_portaddr_hash); hslot4 = udp_hashslot4(udptable, hash); udp_sk(sk)->udp_lrpa_hash = hash; From 9e95b1a94c9c49b4ba722251bbca2759b9c51737 Mon Sep 17 00:00:00 2001 From: Shardul Bankar Date: Thu, 17 Sep 2026 14:55:33 +0530 Subject: [PATCH 336/694] udp: remove a disconnected socket from the 4-tuple hash table A UDP socket bound to a specific address and port keeps its entry in the 4-tuple hash table after it is disconnected: sk binds to 127.0.0.1:21001 sk connects to 127.0.0.2:20001 // filed in the 4-tuple table sk disconnects, connect(AF_UNSPEC) // still filed, peer now 0.0.0.0:0 __udp_disconnect() takes a socket out of that table only as a side effect of ->rehash() or ->unhash(), and it skips ->rehash() when SOCK_BINDADDR_LOCK is set and ->unhash() when SOCK_BINDPORT_LOCK is set. commit 6996a2d2d0a6 ("udp: Unhash auto-bound connected sk from 4-tuple hash table when disconnected.") fixed the same end state for a wildcard-bound socket, by a path this one does not take. The entry is counted whether or not anything hits it. hash4_cnt on the hash2 slot stays raised for as long as the socket lives, so udp_has_hash4() keeps sending every packet for that address and port through the 4-tuple lookup first. On IPv6 it can also be hit. __udp_disconnect() does not clear sk_v6_daddr, so udp_v6_rehash() files the entry under the peer the socket was connected to with a zero dport, and inet6_match() compares that same field: a datagram from the former peer with a zero source port matches, and source port zero is accepted on receive. On IPv4 the peer is cleared, so a match would need a zero source address as well, which the routing layer rejects as martian. The stale sk_v6_daddr is a separate defect, not addressed here; removing the entry closes this path either way. The entry can also be relocated. __udp_disconnect() clears sk_bound_dev_if, so a subsequent SO_BINDTODEVICE calls ->rehash(), and because the receive address is still specific udp_lib_rehash() moves the entry instead of removing it, into the bucket that (rcv_saddr, num, 0, 0) hashes to -- a pure function of the address and port, so every socket reaching this state on one address and port collects in one bucket. The bucket cannot be chosen from outside, as udp_ehashfn() is seeded with a per-boot secret. This last one became reachable only with commit 644f9108f3a5 ("udp: Make rehash4 independent in udp_lib_rehash()"), which moved the hash4 handling out of a branch a disconnected socket does not take; the stale entry itself dates from the commit in Fixes. Take the socket out of the table before __udp_disconnect() runs, while it still matches how it was filed. This also reaches the wildcard case ahead of udp_lib_rehash()'s udp_unhash4() branch, leaving that branch unreachable from udp_disconnect(); removing it belongs in net-next. udp_disconnect() and udp_abort() are the only UDP entries into __udp_disconnect(), which is shared with raw, ping and l2tp sockets that are not struct udp_sock: ping_prot.obj_size is sizeof(struct inet_sock), so udp_hashed4() on one would read past the allocation. Fixes: 78c91ae2c6de ("ipv4/udp: Add 4-tuple hash for connected socket") Assisted-by: LLM Signed-off-by: Shardul Bankar Reviewed-by: Kuniyuki Iwashima Link: https://patch.msgid.link/20260917-udp_hash4_fix_v1-v1-2-718891af0d7a@mpiricsoftware.com Signed-off-by: Paolo Abeni --- net/ipv4/udp.c | 23 +++++++++++++++++++++++ 1 file changed, 23 insertions(+) diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c index 0fa3cdbdcc21..b090bd1f59e8 100644 --- a/net/ipv4/udp.c +++ b/net/ipv4/udp.c @@ -2206,9 +2206,31 @@ int __udp_disconnect(struct sock *sk, int flags) } EXPORT_SYMBOL(__udp_disconnect); +/* __udp_disconnect() takes a socket out of the 4-tuple hash table only via + * ->rehash() or ->unhash(), and neither runs for a socket bound to a + * specific address and port. Remove it here, before its peer is cleared. + */ +static void udp_unhash4_on_disconnect(struct sock *sk) +{ + struct net *net = sock_net(sk); + struct udp_table *udptable; + struct udp_hslot *hslot; + + if (!udp_hashed4(sk)) + return; + + udptable = net->ipv4.udp_table; + hslot = udp_hashslot(udptable, net, udp_sk(sk)->udp_port_hash); + + spin_lock_bh(&hslot->lock); + udp_unhash4(udptable, sk); + spin_unlock_bh(&hslot->lock); +} + int udp_disconnect(struct sock *sk, int flags) { lock_sock(sk); + udp_unhash4_on_disconnect(sk); __udp_disconnect(sk, flags); release_sock(sk); return 0; @@ -3140,6 +3162,7 @@ int udp_abort(struct sock *sk, int err) sk->sk_err = err; sk_error_report(sk); + udp_unhash4_on_disconnect(sk); __udp_disconnect(sk, 0); out: From cec38d5c098a350dcf084d345025136ade7e6d1e Mon Sep 17 00:00:00 2001 From: Sean Christopherson Date: Mon, 21 Sep 2026 12:14:09 -0700 Subject: [PATCH 337/694] perf/x86/intel: Ensure KVM guest PEBS path doesn't set unwanted PERF_GLOBAL_CTRL bits When reinstating PEBS counters into PERF_GLOBAL_CTRL for a KVM guest, mask the value with perf's desired/original PERF_GLOBAL_CTRL value to ensure KVM doesn't unintentionally set reserved bits in PERF_GLOBAL_CTRL. E.g. if the guest's PEBS_ENABLE value had bit 63, "Enable Precise Store", set, then using the raw guest PEBS value would propagate bit 63 to the guest's PERF_GLOBAL_CTRL value (which thankfully would be a failed VM-Entry, not a VMX Abort). The only reason this bug isn't reachable is because KVM doesn't support "Enable Precise Store" (which is probably a KVM bug?), i.e. bit 63 can't be set in kvm_pmu->pebs_enable and thus not in arr[pebs_enable].guest. In other words, this _should_ be a glorified NOP in the current code base. Fixes: c59a1f106f5c ("KVM: x86/pmu: Add IA32_PEBS_ENABLE MSR emulation for extended PEBS") Signed-off-by: Sean Christopherson Signed-off-by: Peter Zijlstra (Intel) Signed-off-by: Ingo Molnar Reviewed-by: Dapeng Mi Link: https://patch.msgid.link/20260921191418.950933-2-seanjc@google.com --- arch/x86/events/intel/core.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/x86/events/intel/core.c b/arch/x86/events/intel/core.c index 1ac2ca35db53..0400b5111db9 100644 --- a/arch/x86/events/intel/core.c +++ b/arch/x86/events/intel/core.c @@ -5359,7 +5359,7 @@ static struct perf_guest_switch_msr *intel_guest_get_msrs(int *nr, void *data) arr[pebs_enable].guest &= ~kvm_pmu->host_cross_mapped_mask; arr[global_ctrl].guest &= ~kvm_pmu->host_cross_mapped_mask; /* Set hw GLOBAL_CTRL bits for PEBS counter when it runs for guest */ - arr[global_ctrl].guest |= arr[pebs_enable].guest; + arr[global_ctrl].guest |= intel_ctrl & arr[pebs_enable].guest; } return arr; From 4b64dbdc5861477f148e13d1ed127e7fe7182e4f Mon Sep 17 00:00:00 2001 From: Sean Christopherson Date: Mon, 21 Sep 2026 12:14:10 -0700 Subject: [PATCH 338/694] perf/x86/intel: Don't write PEBS_ENABLED on host<=>guest xfers if CPU has PEBS isolation, to fix stuck PEBS_ENABLED When filling the list of MSRs to be loaded by KVM on VM-Enter and VM-Exit, *never* insert an entry for PEBS_ENABLED if the CPU properly isolates PEBS events, in which case disabling counters via PERF_GLOBAL_CTRL is sufficient to prevent unwanted PEBS events in the guest (or host). Because perf loads PEBS_ENABLE with the unfiltered cpu_hw_events.pebs_enabled, i.e. with both host and guest masks, there is no need to load different values for the guest versus host, perf+KVM can and should simply control which counters are enabled/disabled via PERF_GLOBAL_CTRL. Avoiding touching PEBS_ENABLED "fixes" a bug where PEBS_ENABLED can end up with "stuck" bits if a PEBS event is throttled between generating the list and actually entering the guest (Intel CPUs can't arbtitrarily block NMIs). Fixes in quotes because leaving PEBS_ENABLED as-is doesn't fix the underlying problem of perf (via PMIs) being able to modify state after the perf<=>KVM handoff. But not writing PEBS_ENABLED is desirable no matter what, as stating the obvious, leaving PEBS_ENABLED as-is avoids three MSR writes on every VMX transition: one each on entry/exit, and one more explicit WRMSR to zero PEBS_ENABLED before VM-Entry (KVM assumes the only reason PEBS_ENABLED is in the load list is if the CPU lacks PEBS isolation and thus needs a quiescent period). Opportunistically add comments to (better) explain the rules for generating the set of PEBS counters that will be active while the guest is running, along with a FIXME for the suspected hack-a-fix where perf disables guest PEBS if _any_ PEBS event is configured to count in the host (commit 854250329c02 ("KVM: x86/pmu: Disable guest PEBS temporarily in two rare situations") doesn't explain the motivation, at all). Fixes: c59a1f106f5c ("KVM: x86/pmu: Add IA32_PEBS_ENABLE MSR emulation for extended PEBS") Signed-off-by: Sean Christopherson Signed-off-by: Peter Zijlstra (Intel) Signed-off-by: Ingo Molnar Reviewed-by: Dapeng Mi Link: https://patch.msgid.link/20260921191418.950933-3-seanjc@google.com --- arch/x86/events/intel/core.c | 55 ++++++++++++++++++++++++------------ 1 file changed, 37 insertions(+), 18 deletions(-) diff --git a/arch/x86/events/intel/core.c b/arch/x86/events/intel/core.c index 0400b5111db9..5ac8fe31871b 100644 --- a/arch/x86/events/intel/core.c +++ b/arch/x86/events/intel/core.c @@ -5292,12 +5292,15 @@ static struct perf_guest_switch_msr *intel_guest_get_msrs(int *nr, void *data) struct kvm_pmu *kvm_pmu = (struct kvm_pmu *)data; u64 intel_ctrl = hybrid(cpuc->pmu, intel_ctrl); u64 pebs_mask = cpuc->pebs_enabled & x86_pmu.pebs_capable; - int global_ctrl, pebs_enable; + u64 guest_pebs_mask; + int global_ctrl; /* * In addition to obeying exclude_guest/exclude_host, remove bits being * used for PEBS when running a guest, because PEBS writes to virtual - * addresses (not physical addresses). + * addresses (not physical addresses). If the guest wants to utilize + * PEBS, and PEBS can be safely enabled in the guest, bits for the guest's + * PEBS-enabled counters will be OR'd back in as appropriate. */ *nr = 0; global_ctrl = (*nr)++; @@ -5344,24 +5347,40 @@ static struct perf_guest_switch_msr *intel_guest_get_msrs(int *nr, void *data) }; } - pebs_enable = (*nr)++; - arr[pebs_enable] = (struct perf_guest_switch_msr){ - .msr = MSR_IA32_PEBS_ENABLE, - .host = cpuc->pebs_enabled & ~cpuc->intel_ctrl_guest_mask, - .guest = pebs_mask & ~cpuc->intel_ctrl_host_mask & kvm_pmu->pebs_enable, - }; + /* + * Restrict guest PEBS events to counters that (a) perf supports, (b) + * the guest wants to use for PEBS, (c) are not excluded from counting + * in the guest, and (d) _are_ excluded from counting in the host. + */ + guest_pebs_mask = pebs_mask & intel_ctrl & kvm_pmu->pebs_enable & + ~cpuc->intel_ctrl_host_mask & + cpuc->intel_ctrl_guest_mask; - if (arr[pebs_enable].host) { - /* Disable guest PEBS if host PEBS is enabled. */ - arr[pebs_enable].guest = 0; - } else { - /* Disable guest PEBS thoroughly for cross-mapped PEBS counters. */ - arr[pebs_enable].guest &= ~kvm_pmu->host_cross_mapped_mask; - arr[global_ctrl].guest &= ~kvm_pmu->host_cross_mapped_mask; - /* Set hw GLOBAL_CTRL bits for PEBS counter when it runs for guest */ - arr[global_ctrl].guest |= intel_ctrl & arr[pebs_enable].guest; - } + /* + * Disable counters where the guest PMC is different than the host PMC + * being used on behalf of the guest, as the PEBS record includes + * PERF_GLOBAL_STATUS, i.e. the guest will see overflow status for the + * wrong counter(s). + */ + guest_pebs_mask &= ~kvm_pmu->host_cross_mapped_mask; + /* + * FIXME: Allow guest and host usage of PEBS events to co-exist instead + * of disabling guest PEBS entirely if the host is using PEBS. + * What exactly goes wrong if guest and host are using PEBS is + * unknown. + */ + if (pebs_mask & ~cpuc->intel_ctrl_guest_mask) + guest_pebs_mask = 0; + + /* + * Do NOT mess with PEBS_ENABLED. As above, disabling counters via + * PERF_GLOBAL_CTRL is sufficient, and loading a stale PEBS_ENABLED, + * e.g. on VM-Exit, can put the system in a bad state. Simply enable + * counters in PERF_GLOBAL_CTRL, as perf load PEBS_ENABLED with the + * full value, i.e. perf *also* relies on PERF_GLOBAL_CTRL. + */ + arr[global_ctrl].guest |= guest_pebs_mask; return arr; } From d06260e99eb93d2942b7af4ccd789eb8a6c829d3 Mon Sep 17 00:00:00 2001 From: Sean Christopherson Date: Mon, 21 Sep 2026 12:14:11 -0700 Subject: [PATCH 339/694] perf/x86/intel: Don't pointlessly context switch DS_AREA (and PEBS config) if PEBS is unused When filling the list of MSRs to be loaded by KVM on VM-Enter and VM-Exit, load the guest values for DS_AREA and (conditionally) MSR_PEBS_DATA_CFG if and only if PEBS will be active in the guest, i.e. only if a PEBS record may be generated while running the guest. As shown by the !pebs_ept path, it's perfectly safe to run with the host's DS_AREA, so long as PEBS-enabled counters are disabled via PERF_GLOBAL_CTRL. Omitting DS_AREA and MSR_PEBS_DATA_CFG when PEBS is unused saves two MSR writes per MSR on each VMX transition, i.e. eliminates two/four pointless MSR writes on each VMX roundtrip when PEBS isn't being used by the guest. Fixes: c59a1f106f5c ("KVM: x86/pmu: Add IA32_PEBS_ENABLE MSR emulation for extended PEBS") Signed-off-by: Sean Christopherson Signed-off-by: Peter Zijlstra (Intel) Signed-off-by: Ingo Molnar Reviewed-by: Jim Mattson Reviewed-by: Dapeng Mi Link: https://patch.msgid.link/20260921191418.950933-4-seanjc@google.com --- arch/x86/events/intel/core.c | 39 +++++++++++++++++++++++------------- 1 file changed, 25 insertions(+), 14 deletions(-) diff --git a/arch/x86/events/intel/core.c b/arch/x86/events/intel/core.c index 5ac8fe31871b..75c74e8fe2f4 100644 --- a/arch/x86/events/intel/core.c +++ b/arch/x86/events/intel/core.c @@ -5330,23 +5330,14 @@ static struct perf_guest_switch_msr *intel_guest_get_msrs(int *nr, void *data) return arr; } + /* + * If the guest won't use PEBS or the CPU doesn't support PEBS in the + * guest, then there's nothing more to do as disabling PMCs via + * PERF_GLOBAL_CTRL is sufficient on CPUs with guest/host isolation. + */ if (!kvm_pmu || !x86_pmu.pebs_ept) return arr; - arr[(*nr)++] = (struct perf_guest_switch_msr){ - .msr = MSR_IA32_DS_AREA, - .host = (unsigned long)cpuc->ds, - .guest = kvm_pmu->ds_area, - }; - - if (x86_pmu.intel_cap.pebs_baseline) { - arr[(*nr)++] = (struct perf_guest_switch_msr){ - .msr = MSR_PEBS_DATA_CFG, - .host = cpuc->active_pebs_data_cfg, - .guest = kvm_pmu->pebs_data_cfg, - }; - } - /* * Restrict guest PEBS events to counters that (a) perf supports, (b) * the guest wants to use for PEBS, (c) are not excluded from counting @@ -5373,6 +5364,26 @@ static struct perf_guest_switch_msr *intel_guest_get_msrs(int *nr, void *data) if (pebs_mask & ~cpuc->intel_ctrl_guest_mask) guest_pebs_mask = 0; + /* + * Context switch DS_AREA and PEBS_DATA_CFG if and only if PEBS will be + * active in the guest; if no records will be generated while the guest + * is running, then simply keep the host values resident in hardware. + */ + arr[(*nr)++] = (struct perf_guest_switch_msr){ + .msr = MSR_IA32_DS_AREA, + .host = (unsigned long)cpuc->ds, + .guest = guest_pebs_mask ? kvm_pmu->ds_area : (unsigned long)cpuc->ds, + }; + + if (x86_pmu.intel_cap.pebs_baseline) { + arr[(*nr)++] = (struct perf_guest_switch_msr){ + .msr = MSR_PEBS_DATA_CFG, + .host = cpuc->active_pebs_data_cfg, + .guest = guest_pebs_mask ? kvm_pmu->pebs_data_cfg : + cpuc->active_pebs_data_cfg, + }; + } + /* * Do NOT mess with PEBS_ENABLED. As above, disabling counters via * PERF_GLOBAL_CTRL is sufficient, and loading a stale PEBS_ENABLED, From a391618e1d563f099e4c2a704f45d08329ccdf7c Mon Sep 17 00:00:00 2001 From: Sean Christopherson Date: Mon, 21 Sep 2026 12:14:12 -0700 Subject: [PATCH 340/694] perf/x86/intel: Make @data a mandatory param for intel_guest_get_msrs() Drop "support" for passing a NULL @data/@kvm_pmu param when getting guest MSRs. KVM, the only in-tree user, unconditionally passes a non-NULL pointer, and carrying code that suggests @data may be NULL is confusing, e.g. incorrectly implies that there are scenarios where KVM doesn't pass a PMU context. Fixes: 8183a538cd95 ("KVM: x86/pmu: Add IA32_DS_AREA MSR emulation to support guest DS") Signed-off-by: Sean Christopherson Signed-off-by: Peter Zijlstra (Intel) Signed-off-by: Ingo Molnar Reviewed-by: Jim Mattson Reviewed-by: Dapeng Mi Link: https://patch.msgid.link/20260921191418.950933-5-seanjc@google.com --- arch/x86/events/intel/core.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/arch/x86/events/intel/core.c b/arch/x86/events/intel/core.c index 75c74e8fe2f4..757d4082de60 100644 --- a/arch/x86/events/intel/core.c +++ b/arch/x86/events/intel/core.c @@ -5331,11 +5331,11 @@ static struct perf_guest_switch_msr *intel_guest_get_msrs(int *nr, void *data) } /* - * If the guest won't use PEBS or the CPU doesn't support PEBS in the - * guest, then there's nothing more to do as disabling PMCs via - * PERF_GLOBAL_CTRL is sufficient on CPUs with guest/host isolation. + * If the CPU doesn't support PEBS in the guest, then there's nothing + * more to do as disabling PMCs via PERF_GLOBAL_CTRL is sufficient on + * CPUs with guest/host isolation. */ - if (!kvm_pmu || !x86_pmu.pebs_ept) + if (!x86_pmu.pebs_ept) return arr; /* From a92e1a412c53dc0d9ad639e7abf8b3fc70a5b6ad Mon Sep 17 00:00:00 2001 From: Myeonghun Pak Date: Thu, 17 Sep 2026 14:33:36 -0400 Subject: [PATCH 341/694] tg3: clean up PHYLIB resources on probe failure tg3_get_invariants() can register an MDIO bus and connect a PHY for USE_PHYLIB devices. If tg3_init_one() later fails, its common error path releases the mappings and netdev without undoing those PHYLIB resources. Disconnect the PHY and unregister the MDIO bus before the remaining teardown. Guard PHY cleanup with USE_PHYLIB to match tg3_phy_init(), and call tg3_mdio_fini() unconditionally to match tg3_mdio_init(). The existing IS_CONNECTED and MDIOBUS_INITED flags make both helpers safe when initialization only completed partially. This issue was identified during our ongoing static-analysis research while reviewing kernel code. Fixes: 158d7abdae85 ("tg3: Add mdio bus registration") Assisted-by: OpenAI:GPT-5.6 Co-developed-by: Ijae Kim Signed-off-by: Ijae Kim Signed-off-by: Myeonghun Pak Link: https://patch.msgid.link/20260917183336.36239-1-mhun512@gmail.com Signed-off-by: Paolo Abeni --- drivers/net/ethernet/broadcom/tg3.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/drivers/net/ethernet/broadcom/tg3.c b/drivers/net/ethernet/broadcom/tg3.c index 73a4b569b03e..caa7a6caa6e2 100644 --- a/drivers/net/ethernet/broadcom/tg3.c +++ b/drivers/net/ethernet/broadcom/tg3.c @@ -18047,6 +18047,10 @@ static int tg3_init_one(struct pci_dev *pdev, return 0; err_out_apeunmap: + if (tg3_flag(tp, USE_PHYLIB)) + tg3_phy_fini(tp); + tg3_mdio_fini(tp); + if (tp->aperegs) { iounmap(tp->aperegs); tp->aperegs = NULL; From 89dc568e8c0be60e05e5fcd0b528c79077d7b84b Mon Sep 17 00:00:00 2001 From: Dapeng Mi Date: Thu, 17 Sep 2026 09:52:23 +0800 Subject: [PATCH 342/694] perf/x86/intel: Fix GRT PEBS load/store direction for latency events, to fix sample classification On Gracemont, intel_grt_pebs_event_constraints[] applies LAT_CONSTRAINT constraints to MEM_UOPS_RETIRED.{LOAD,STORE}_LATENCY, but does not set explicit LOAD/STORE flags for those events. The PEBS latency path (pebs_latency_data(), via __grt_latency_data()) uses the event flags to determine memory operation direction. Without an explicit STORE flag, samples from MEM_UOPS_RETIRED.STORE_LATENCY can be misclassified as LOADs. Set explicit LOAD/STORE flags in intel_grt_pebs_event_constraints[] for: - MEM_UOPS_RETIRED.LOAD_LATENCY - MEM_UOPS_RETIRED.STORE_LATENCY Also update __grt_latency_data() to explicitly interpret these flags when assigning the sampled memory operation direction. This fixes incorrect STORE sample classification. Fixes: 39a41278f041 ("perf/x86/intel: Fix PEBS memory access info encoding for ADL") Signed-off-by: Dapeng Mi Signed-off-by: Peter Zijlstra (Intel) Signed-off-by: Ingo Molnar Cc: # v7.2+ Link: https://patch.msgid.link/20260917015234.981153-2-dapeng1.mi@linux.intel.com --- arch/x86/events/intel/ds.c | 16 +++++++++++++--- 1 file changed, 13 insertions(+), 3 deletions(-) diff --git a/arch/x86/events/intel/ds.c b/arch/x86/events/intel/ds.c index b98029b44052..fff98f457576 100644 --- a/arch/x86/events/intel/ds.c +++ b/arch/x86/events/intel/ds.c @@ -455,6 +455,7 @@ static inline void pebs_set_tlb_lock(u64 *val, bool tlb, bool lock) static u64 __grt_latency_data(struct perf_event *event, u64 status, u8 dse, bool tlb, bool lock, bool blk) { + union perf_mem_data_src src; u64 val; WARN_ON_ONCE(is_hybrid() && @@ -470,7 +471,16 @@ static u64 __grt_latency_data(struct perf_event *event, u64 status, else val |= P(BLK, NA); - return val; + src.val = val; + + if (event->hw.flags & + (PERF_X86_EVENT_PEBS_LDLAT | PERF_X86_EVENT_PEBS_LD_HSW)) + src.mem_op = P(OP, LOAD); + if (event->hw.flags & + (PERF_X86_EVENT_PEBS_STLAT | PERF_X86_EVENT_PEBS_ST_HSW)) + src.mem_op = P(OP, STORE); + + return src.val; } u64 grt_latency_data(struct perf_event *event, u64 status) @@ -1291,8 +1301,8 @@ struct event_constraint intel_glm_pebs_event_constraints[] = { struct event_constraint intel_grt_pebs_event_constraints[] = { /* Allow all events as PEBS with no flags */ - INTEL_HYBRID_LAT_CONSTRAINT(0x5d0, 0x3), - INTEL_HYBRID_LAT_CONSTRAINT(0x6d0, 0x3f), + INTEL_HYBRID_LDLAT_CONSTRAINT(0x5d0, 0x3), + INTEL_HYBRID_STLAT_CONSTRAINT(0x6d0, 0x3f), EVENT_CONSTRAINT_END }; From e961d6db42d1f1b66c81438969a648f08573bd9c Mon Sep 17 00:00:00 2001 From: Dapeng Mi Date: Thu, 17 Sep 2026 09:52:24 +0800 Subject: [PATCH 343/694] perf/x86/intel: Fix CMT PEBS load/store direction for latency events, to fix sample classification The same bug exists on Crestmont as on Gracemont: intel_cmt_pebs_event_constraints[] applies LAT_CONSTRAINT constraints to MEM_UOPS_RETIRED.{LOAD,STORE}_LATENCY, but does not set explicit LOAD/STORE flags for those events. The PEBS latency path (pebs_latency_data(), via cmt_latency_data) uses the event flags to determine memory operation direction. Without an explicit STORE flag, samples from MEM_UOPS_RETIRED.STORE_LATENCY can be misclassified as LOADs. Set explicit LOAD/STORE flags in intel_cmt_pebs_event_constraints[] for: - MEM_UOPS_RETIRED.LOAD_LATENCY - MEM_UOPS_RETIRED.STORE_LATENCY This fixes incorrect STORE sample classification. Fixes: e99fb45436ea ("perf/x86/intel: Update event constraints and cache_extra_regsfor MTL") Signed-off-by: Dapeng Mi Signed-off-by: Peter Zijlstra (Intel) Signed-off-by: Ingo Molnar Cc: # v7.2+ Link: https://patch.msgid.link/20260917015234.981153-3-dapeng1.mi@linux.intel.com --- arch/x86/events/intel/ds.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/x86/events/intel/ds.c b/arch/x86/events/intel/ds.c index fff98f457576..fa7e0268b4f2 100644 --- a/arch/x86/events/intel/ds.c +++ b/arch/x86/events/intel/ds.c @@ -1308,8 +1308,8 @@ struct event_constraint intel_grt_pebs_event_constraints[] = { struct event_constraint intel_cmt_pebs_event_constraints[] = { /* Allow all events as PEBS with no flags */ - INTEL_HYBRID_LAT_CONSTRAINT(0x5d0, 0x3), - INTEL_HYBRID_LAT_CONSTRAINT(0x6d0, 0xff), + INTEL_HYBRID_LDLAT_CONSTRAINT(0x5d0, 0x3), + INTEL_HYBRID_STLAT_CONSTRAINT(0x6d0, 0xff), EVENT_CONSTRAINT_END }; From 2777ec9852277a06ae68fee0c4f1a32783e4a999 Mon Sep 17 00:00:00 2001 From: Nemesa Garg Date: Wed, 9 Sep 2026 16:33:32 +0530 Subject: [PATCH 344/694] drm/i915/psr: Clear stale sel fetch enable bits on sel fetch disable MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Selective fetch is dropped while pipe CRC is active, and the planes keep their SEL_FETCH_PLANE_CTL / SEL_FETCH_CUR_CTL enable bit set in hardware over that. A plane disabled while selective fetch is off never gets the bit cleared, as the disable path is guarded by enable_psr2_sel_fetch. Once selective fetch comes back the hardware resumes fetching for a plane that is no longer enabled and keeps its DDB range reserved. Clear the bits as selective fetch is turned off instead. Atomic check has both the old and the new crtc state, so record the transition there and let the plane and cursor arm paths write the registers to 0 for that commit. v2: Drop the old_crtc_state->hw.active check. [Jouni] Fixes: b1f5279b5981 ("drm/i915/psr: Move plane sel fetch configuration into plane source files") Closes: https://gitlab.freedesktop.org/drm/xe/kernel/-/work_items/8739 Assisted-by: Copilot:Claude-Opus-5 Signed-off-by: Nemesa Garg Reviewed-by: Jouni Högander Signed-off-by: Suraj Kandpal Link: https://patch.msgid.link/20260909110332.3528029-3-nemesa.garg@intel.com (cherry picked from commit a4c0e7f80429eda6990960971aebd4e4b9533cc6) Signed-off-by: Jani Nikula --- drivers/gpu/drm/i915/display/intel_cursor.c | 7 +++++-- .../gpu/drm/i915/display/intel_display_types.h | 2 ++ drivers/gpu/drm/i915/display/intel_psr.c | 15 +++++++++++++++ .../gpu/drm/i915/display/skl_universal_plane.c | 9 ++++----- 4 files changed, 26 insertions(+), 7 deletions(-) diff --git a/drivers/gpu/drm/i915/display/intel_cursor.c b/drivers/gpu/drm/i915/display/intel_cursor.c index 0673f16f6fd0..824ffeef0103 100644 --- a/drivers/gpu/drm/i915/display/intel_cursor.c +++ b/drivers/gpu/drm/i915/display/intel_cursor.c @@ -536,7 +536,8 @@ static void i9xx_cursor_disable_sel_fetch_arm(struct intel_dsb *dsb, struct intel_display *display = to_intel_display(plane); enum pipe pipe = plane->pipe; - if (!crtc_state->enable_psr2_sel_fetch) + if (!crtc_state->enable_psr2_sel_fetch && + !crtc_state->clear_psr2_sel_fetch) return; intel_de_write_dsb(display, dsb, SEL_FETCH_CUR_CTL(pipe), 0); @@ -569,8 +570,10 @@ static void i9xx_cursor_update_sel_fetch_arm(struct intel_dsb *dsb, struct intel_display *display = to_intel_display(plane); enum pipe pipe = plane->pipe; - if (!crtc_state->enable_psr2_sel_fetch) + if (!crtc_state->enable_psr2_sel_fetch) { + i9xx_cursor_disable_sel_fetch_arm(dsb, plane, crtc_state); return; + } if (drm_rect_height(&plane_state->psr2_sel_fetch_area) > 0) { if (crtc_state->enable_psr2_su_region_et) { diff --git a/drivers/gpu/drm/i915/display/intel_display_types.h b/drivers/gpu/drm/i915/display/intel_display_types.h index 43d53a98dae7..f6a9b0de1ade 100644 --- a/drivers/gpu/drm/i915/display/intel_display_types.h +++ b/drivers/gpu/drm/i915/display/intel_display_types.h @@ -1187,6 +1187,8 @@ struct intel_crtc_state { bool has_sel_update; bool enable_psr2_sel_fetch; bool enable_psr2_su_region_et; + /* Drop the stale selective fetch enable bits as selective fetch is turned off */ + bool clear_psr2_sel_fetch; bool req_psr2_sdp_prior_scanline; bool has_panel_replay; bool link_off_after_as_sdp_when_pr_active; diff --git a/drivers/gpu/drm/i915/display/intel_psr.c b/drivers/gpu/drm/i915/display/intel_psr.c index 40e3d7095996..c28bcb1a0184 100644 --- a/drivers/gpu/drm/i915/display/intel_psr.c +++ b/drivers/gpu/drm/i915/display/intel_psr.c @@ -2883,6 +2883,8 @@ int intel_psr2_sel_fetch_update(struct intel_atomic_state *state, struct intel_crtc *crtc) { struct intel_display *display = to_intel_display(state); + const struct intel_crtc_state *old_crtc_state = + intel_atomic_get_old_crtc_state(state, crtc); struct intel_crtc_state *crtc_state = intel_atomic_get_new_crtc_state(state, crtc); struct intel_plane_state *new_plane_state, *old_plane_state; struct intel_plane *plane; @@ -2895,6 +2897,19 @@ int intel_psr2_sel_fetch_update(struct intel_atomic_state *state, bool full_update = false, su_area_changed; int i, ret; + /* + * Selective fetch is not always usable, for instance it is dropped + * while pipe CRC is active. The planes keep their selective fetch + * enable bit set in hardware over that, and a plane disabled while + * selective fetch is off never gets the bit cleared. Once selective + * fetch comes back the hardware would resume fetching for a plane that + * is no longer enabled and keep its DDB range reserved, so have the + * plane update drop the bit for every plane of the pipe as selective + * fetch is turned off. + */ + crtc_state->clear_psr2_sel_fetch = old_crtc_state->enable_psr2_sel_fetch && + !crtc_state->enable_psr2_sel_fetch; + if (!crtc_state->enable_psr2_sel_fetch) return 0; diff --git a/drivers/gpu/drm/i915/display/skl_universal_plane.c b/drivers/gpu/drm/i915/display/skl_universal_plane.c index 07a683293352..eb5ed981b40f 100644 --- a/drivers/gpu/drm/i915/display/skl_universal_plane.c +++ b/drivers/gpu/drm/i915/display/skl_universal_plane.c @@ -885,7 +885,8 @@ static void icl_plane_disable_sel_fetch_arm(struct intel_dsb *dsb, struct intel_display *display = to_intel_display(plane); enum pipe pipe = plane->pipe; - if (!crtc_state->enable_psr2_sel_fetch) + if (!crtc_state->enable_psr2_sel_fetch && + !crtc_state->clear_psr2_sel_fetch) return; intel_de_write_dsb(display, dsb, SEL_FETCH_PLANE_CTL(pipe, plane->id), 0); @@ -1634,10 +1635,8 @@ static void icl_plane_update_sel_fetch_arm(struct intel_dsb *dsb, struct intel_display *display = to_intel_display(plane); enum pipe pipe = plane->pipe; - if (!crtc_state->enable_psr2_sel_fetch) - return; - - if (drm_rect_height(&plane_state->psr2_sel_fetch_area) > 0) + if (crtc_state->enable_psr2_sel_fetch && + drm_rect_height(&plane_state->psr2_sel_fetch_area) > 0) intel_de_write_dsb(display, dsb, SEL_FETCH_PLANE_CTL(pipe, plane->id), SEL_FETCH_PLANE_CTL_ENABLE); else From 5271d81f99dd01d983d439930eb056952485e15e Mon Sep 17 00:00:00 2001 From: Ankit Nautiyal Date: Mon, 7 Sep 2026 09:15:55 +0530 Subject: [PATCH 345/694] drm/i915/quirks: Limit eDP rate to HBR2 on HP Pavilion Plus 14-ew1 The eDP panel on the HP Pavilion Plus Laptop 14-ew1xxx advertises HBR3 while leaving the TPS4 support bit clear. The output however flickers, once link is trained with HBR3. Until commit 8c9006283e4b ("Revert "drm/i915/dp: Reject HBR3 when sink doesn't support TPS4"") such sinks were capped at HBR2 by the TPS4 check which incidentally kept this panel stable. That check was reverted because other panels legitimately need HBR3 without advertising TPS4, and the per-machine QUIRK_EDP_LIMIT_RATE_HBR2 was introduced to handle the affected machines instead. Add the machine to the list of devices that need the QUIRK_EDP_LIMIT_RATE_HBR2. Fixes: 8c9006283e4b ("Revert "drm/i915/dp: Reject HBR3 when sink doesn't support TPS4"") Reported-by: Annoy Cc Closes: https://gitlab.freedesktop.org/drm/i915/kernel/-/work_items/16743 Cc: # v6.18+ Tested-by: Annoy Cc Signed-off-by: Ankit Nautiyal Reviewed-by: Nemesa Garg Link: https://patch.msgid.link/20260907034555.2753846-1-ankit.k.nautiyal@intel.com (cherry picked from commit 550b703fdbb2a2022faa75b4b11ab135241afbd9) Signed-off-by: Jani Nikula --- drivers/gpu/drm/i915/display/intel_quirks.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/gpu/drm/i915/display/intel_quirks.c b/drivers/gpu/drm/i915/display/intel_quirks.c index 33245f44c0d5..7d7db774d8c7 100644 --- a/drivers/gpu/drm/i915/display/intel_quirks.c +++ b/drivers/gpu/drm/i915/display/intel_quirks.c @@ -257,6 +257,9 @@ static struct intel_quirk intel_quirks[] = { /* Dell XPS 13 7390 2-in-1 */ { 0x8a52, 0x1028, 0x08b0, quirk_edp_limit_rate_hbr2 }, + /* HP Pavilion Plus Laptop 14-ew1xxx */ + { 0x7d55, 0x103c, 0x8c31, quirk_edp_limit_rate_hbr2 }, + /* Xiaomi Book Pro 14 2026 */ { 0xb081, 0x1d72, 0x2424, quirk_disable_psr2 }, }; From 8302c5f475fa5a4ed36a7d32c7b965023d5d7066 Mon Sep 17 00:00:00 2001 From: Dapeng Mi Date: Thu, 17 Sep 2026 09:52:25 +0800 Subject: [PATCH 346/694] perf/x86/intel: Fix DKT PEBS load/store direction for latency events, to fix sample classification Same bug exists on Darkmont as on Gracemont: intel_dkt_pebs_event_constraints[] applies LAT_CONSTRAINT constraints to MEM_UOPS_RETIRED.{LOAD,STORE}_LATENCY, but does not set explicit LOAD/STORE flags for those events. The PEBS latency path (pebs_latency_data(), via cmt_latency_data) uses the event flags to determine memory operation direction. Without an explicit STORE flag, samples from MEM_UOPS_RETIRED.STORE_LATENCY can be misclassified as LOADs. Set explicit LOAD/STORE flags in intel_dkt_pebs_event_constraints[] for: - MEM_UOPS_RETIRED.LOAD_LATENCY - MEM_UOPS_RETIRED.STORE_LATENCY This fixes incorrect STORE sample classification. Additionally remove INTEL_HYBRID_LAT_CONSTRAINT() since no one uses it anymore. Fixes: 65fd435095bb ("perf/x86/intel: Update event constraints for PTL") Signed-off-by: Dapeng Mi Signed-off-by: Peter Zijlstra (Intel) Signed-off-by: Ingo Molnar Cc: # v7.2+ Link: https://patch.msgid.link/20260917015234.981153-4-dapeng1.mi@linux.intel.com --- arch/x86/events/intel/ds.c | 4 ++-- arch/x86/events/perf_event.h | 4 ---- 2 files changed, 2 insertions(+), 6 deletions(-) diff --git a/arch/x86/events/intel/ds.c b/arch/x86/events/intel/ds.c index fa7e0268b4f2..ea5b331c58d1 100644 --- a/arch/x86/events/intel/ds.c +++ b/arch/x86/events/intel/ds.c @@ -1315,8 +1315,8 @@ struct event_constraint intel_cmt_pebs_event_constraints[] = { struct event_constraint intel_dkt_pebs_event_constraints[] = { /* Allow all events as PEBS with no flags */ - INTEL_HYBRID_LAT_CONSTRAINT(0x5d0, 0xff), - INTEL_HYBRID_LAT_CONSTRAINT(0x6d0, 0xff), + INTEL_HYBRID_LDLAT_CONSTRAINT(0x5d0, 0xff), + INTEL_HYBRID_STLAT_CONSTRAINT(0x6d0, 0xff), EVENT_CONSTRAINT_END }; diff --git a/arch/x86/events/perf_event.h b/arch/x86/events/perf_event.h index 4680cba91340..fab9da78a5c7 100644 --- a/arch/x86/events/perf_event.h +++ b/arch/x86/events/perf_event.h @@ -517,10 +517,6 @@ struct cpu_hw_events { __EVENT_CONSTRAINT(c, n, INTEL_ARCH_EVENT_MASK|X86_ALL_EVENT_FLAGS, \ HWEIGHT(n), 0, PERF_X86_EVENT_PEBS_ST) -#define INTEL_HYBRID_LAT_CONSTRAINT(c, n) \ - __EVENT_CONSTRAINT(c, n, INTEL_ARCH_EVENT_MASK|X86_ALL_EVENT_FLAGS, \ - HWEIGHT(n), 0, PERF_X86_EVENT_PEBS_LAT_HYBRID) - #define INTEL_HYBRID_LDLAT_CONSTRAINT(c, n) \ __EVENT_CONSTRAINT(c, n, INTEL_ARCH_EVENT_MASK|X86_ALL_EVENT_FLAGS, \ HWEIGHT(n), 0, PERF_X86_EVENT_PEBS_LAT_HYBRID|PERF_X86_EVENT_PEBS_LD_HSW) From acbe9a3b60b9a6ace8ef11fe898f586251c84592 Mon Sep 17 00:00:00 2001 From: Imre Deak Date: Mon, 7 Sep 2026 20:44:12 +0300 Subject: [PATCH 347/694] drm/i915/dp_mst: Fix configuring FEC for a disconnected stream During an atomic commit after all the MST stream CRTC state is computed the driver ensures that the FEC is configured the same way (enabled or disabled) for all the streams on a given MST topology's link. drm_dp_mst_port_downstream_of_parent() used to determine if a stream is downstream of an MST port will return false if the whole topology is disconnected, since in that case it can't verify that the port/ parent_port passed to it is in the given MST topology. This is a problem during the above FEC configuration check, since intel_dp_mst_check_dsc_change()->get_pipes_downstream_of_mst_ports() will not return all the stream CRTCs/pipes for the topology as expected. Since passing parent_port==NULL to get_pipes_downstream_of_mst_port() is meant to return all the streams for the given topology (i.e. mst_mgr) skip checking if an MST port is downstream of a parent port in this case. This fixes a problem where the FEC configuration check explained above failed to ensure that all streams' FEC is configured the same way if the topology was disconnected, leading to a FEC state mismatch error. Cc: stable@vger.kernel.org # v6.10+ Closes: https://gitlab.freedesktop.org/drm/i915/kernel/-/work_items/16073 Closes: https://gitlab.freedesktop.org/drm/i915/kernel/-/work_items/16384 Reviewed-by: Luca Coelho Signed-off-by: Imre Deak Link: https://patch.msgid.link/20260907174413.741851-1-imre.deak@intel.com (cherry picked from commit 270681fbffbba2b6ccf5b7e3c34b8b563b36167f) Signed-off-by: Jani Nikula --- drivers/gpu/drm/i915/display/intel_dp_mst.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/display/intel_dp_mst.c b/drivers/gpu/drm/i915/display/intel_dp_mst.c index 57daed0b0b36..9e65b4a1c71d 100644 --- a/drivers/gpu/drm/i915/display/intel_dp_mst.c +++ b/drivers/gpu/drm/i915/display/intel_dp_mst.c @@ -852,7 +852,8 @@ static u8 get_pipes_downstream_of_mst_port(struct intel_atomic_state *state, if (&connector->mst.dp->mst.mgr != mst_mgr) continue; - if (connector->mst.port != parent_port && + if (parent_port && + connector->mst.port != parent_port && !drm_dp_mst_port_downstream_of_parent(mst_mgr, connector->mst.port, parent_port)) From a443e0b8d647c1401b110d9f919d8c6cb8607260 Mon Sep 17 00:00:00 2001 From: Imre Deak Date: Mon, 7 Sep 2026 20:44:13 +0300 Subject: [PATCH 348/694] drm/i915/dp_mst: Fix configuring TUs for a disconnected stream During an atomic commit after all the MST stream CRTC state is computed the driver ensures that the sum of TUs of all the streams on a given MST topology link is within limits (63 for 8b10 and 64 for 128b132b). For a disconnected stream the DRM MST core's BW verification doesn't ensure this, because the topology state it uses for this is destroyed as soon as the stream (i.e. MST connector/port) is disconnected. The driver should keep the link state valid even for such disconnected streams, as userspace may disable them one-by-one only in a deferred way. Ensure the link's sum of TUs stays within limits in this case by simply reusing the maximum link BPP limit from the stream's (i.e. CRTC's) old state. The disconnection can happen either via the whole topology getting disconnected or via only the given stream's port getting disconnected. Check for both of these conditions separately, as a connector gets unregistered after a link disconnect event only in a deferred way. Cc: stable@vger.kernel.org # v6.10+ Link: https://gitlab.freedesktop.org/drm/i915/kernel/-/work_items/16073 Link: https://gitlab.freedesktop.org/drm/i915/kernel/-/work_items/16384 Reviewed-by: Luca Coelho Signed-off-by: Imre Deak Link: https://patch.msgid.link/20260907174413.741851-2-imre.deak@intel.com (cherry picked from commit ee00f8fbb2b202002ab90834e02e9ba372773a36) Signed-off-by: Jani Nikula --- drivers/gpu/drm/i915/display/intel_dp_mst.c | 21 ++++++++++++++++++++ drivers/gpu/drm/i915/display/intel_dp_mst.h | 2 ++ drivers/gpu/drm/i915/display/intel_link_bw.c | 3 ++- 3 files changed, 25 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/display/intel_dp_mst.c b/drivers/gpu/drm/i915/display/intel_dp_mst.c index 9e65b4a1c71d..fb3942f56b1f 100644 --- a/drivers/gpu/drm/i915/display/intel_dp_mst.c +++ b/drivers/gpu/drm/i915/display/intel_dp_mst.c @@ -2168,6 +2168,27 @@ bool intel_dp_mst_crtc_needs_modeset(struct intel_atomic_state *state, return false; } +bool intel_dp_mst_stream_disconnected(struct intel_atomic_state *state, + const struct intel_crtc *crtc) +{ + struct intel_connector *connector; + + connector = get_connector_in_state_for_crtc(state, crtc); + if (!connector) + return false; + + if (!connector->mst.dp) + return false; + + if (!connector->mst.dp->mst.mgr.mst_state) + return true; + + if (drm_connector_is_unregistered(&connector->base)) + return true; + + return false; +} + /** * intel_dp_mst_prepare_probe - Prepare an MST link for topology probing * @intel_dp: DP port object diff --git a/drivers/gpu/drm/i915/display/intel_dp_mst.h b/drivers/gpu/drm/i915/display/intel_dp_mst.h index ab09b487c6bb..8ce89242c05c 100644 --- a/drivers/gpu/drm/i915/display/intel_dp_mst.h +++ b/drivers/gpu/drm/i915/display/intel_dp_mst.h @@ -28,6 +28,8 @@ int intel_dp_mst_atomic_check_link(struct intel_atomic_state *state, struct intel_link_bw_limits *limits); bool intel_dp_mst_crtc_needs_modeset(struct intel_atomic_state *state, struct intel_crtc *crtc); +bool intel_dp_mst_stream_disconnected(struct intel_atomic_state *state, + const struct intel_crtc *crtc); void intel_dp_mst_prepare_probe(struct intel_dp *intel_dp); bool intel_dp_mst_verify_dpcd_state(struct intel_dp *intel_dp); diff --git a/drivers/gpu/drm/i915/display/intel_link_bw.c b/drivers/gpu/drm/i915/display/intel_link_bw.c index b47474a3e9fe..e71e76d6fd3e 100644 --- a/drivers/gpu/drm/i915/display/intel_link_bw.c +++ b/drivers/gpu/drm/i915/display/intel_link_bw.c @@ -64,7 +64,8 @@ void intel_link_bw_init_limits(struct intel_atomic_state *state, intel_atomic_get_new_crtc_state(state, crtc); int forced_bpp_x16 = get_forced_link_bpp_x16(state, crtc); - if (state->base.duplicated && crtc_state) { + if ((state->base.duplicated && crtc_state) || + intel_dp_mst_stream_disconnected(state, crtc)) { limits->max_bpp_x16[pipe] = crtc_state->max_link_bpp_x16; if (intel_dsc_enabled_on_link(crtc_state)) limits->link_dsc_pipes |= BIT(pipe); From bb2635be7646a6a9e40a27becb936fe3cdcccf8d Mon Sep 17 00:00:00 2001 From: Jani Nikula Date: Tue, 15 Sep 2026 19:06:20 +0300 Subject: [PATCH 349/694] drm/i915/dp: use EXPORT_SYMBOL_IF_KUNIT() for kunit helpers Use EXPORT_SYMBOL_IF_KUNIT() instead of the regular EXPORT_SYMBOL() to export the symbols to the kunit namespace. Otherwise, the symbols get exported for all the kernel to see, and the corresponding MODULE_IMPORT_NS("EXPORTED_FOR_KUNIT_TESTING") in the tests is meaningless. Fixes: 2eb9982ff179 ("drm/i915/kunit: Export link training and caps funcs for testing") Cc: Imre Deak Reviewed-by: Imre Deak Link: https://patch.msgid.link/20260915160620.779372-1-jani.nikula@intel.com Signed-off-by: Jani Nikula (cherry picked from commit 4ffdb772716e4279d63dfaaadf965da73e799aeb) --- drivers/gpu/drm/i915/display/intel_dp_link_caps.c | 6 ++++-- drivers/gpu/drm/i915/display/intel_dp_link_training.c | 4 ++-- 2 files changed, 6 insertions(+), 4 deletions(-) diff --git a/drivers/gpu/drm/i915/display/intel_dp_link_caps.c b/drivers/gpu/drm/i915/display/intel_dp_link_caps.c index 98657aa4d3d5..abec3e2519ca 100644 --- a/drivers/gpu/drm/i915/display/intel_dp_link_caps.c +++ b/drivers/gpu/drm/i915/display/intel_dp_link_caps.c @@ -3,6 +3,8 @@ * Copyright © 2026 Intel Corporation */ +#include + #include #include #include @@ -1302,14 +1304,14 @@ void intel_dp_link_caps_cleanup(struct intel_dp_link_caps *link_caps) const struct intel_dp_link_caps_test_ops i915_display_dp_link_caps_test_ops = { INTEL_DP_LINK_CAPS_TEST_OPS_INIT }; -EXPORT_SYMBOL(i915_display_dp_link_caps_test_ops); +EXPORT_SYMBOL_IF_KUNIT(i915_display_dp_link_caps_test_ops); #else const struct intel_dp_link_caps_test_ops intel_display_dp_link_caps_test_ops = { INTEL_DP_LINK_CAPS_TEST_OPS_INIT }; -EXPORT_SYMBOL(intel_display_dp_link_caps_test_ops); +EXPORT_SYMBOL_IF_KUNIT(intel_display_dp_link_caps_test_ops); #endif /* I915 */ diff --git a/drivers/gpu/drm/i915/display/intel_dp_link_training.c b/drivers/gpu/drm/i915/display/intel_dp_link_training.c index cb92cff90614..9a692f4fdfee 100644 --- a/drivers/gpu/drm/i915/display/intel_dp_link_training.c +++ b/drivers/gpu/drm/i915/display/intel_dp_link_training.c @@ -2825,14 +2825,14 @@ void intel_dp_link_training_cleanup(struct intel_dp_link_training *link_training const struct intel_dp_link_training_test_ops i915_display_dp_link_training_test_ops = { INTEL_DP_LINK_TRAINING_TEST_OPS_INIT }; -EXPORT_SYMBOL(i915_display_dp_link_training_test_ops); +EXPORT_SYMBOL_IF_KUNIT(i915_display_dp_link_training_test_ops); #else const struct intel_dp_link_training_test_ops intel_display_dp_link_training_test_ops = { INTEL_DP_LINK_TRAINING_TEST_OPS_INIT }; -EXPORT_SYMBOL(intel_display_dp_link_training_test_ops); +EXPORT_SYMBOL_IF_KUNIT(intel_display_dp_link_training_test_ops); #endif /* I915 */ From d2da6696e0c4e60414706e607029d0bb0330c67e Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Christian=20K=C3=B6nig?= Date: Thu, 3 Sep 2026 13:36:21 +0200 Subject: [PATCH 350/694] drm/i915: fix incorrect RCU teardown order MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit i915_gem_busy_ioctl uses dma_resv_for_each_fence_unlocked() to iterate over the fences in an GEM object without holding a reference but only the RCU read side lock. What can happen here is that the GEM object is destroyed concurrently while i915_gem_busy_ioctl is still running. This won't free the GEM objects memory, but still drops all the dma_fence references. Now when dma_resv_for_each_fence_unlocked() sees a destroyed dma_fence it assumes that a new fence list was installed and re-starts the loop. But in the case of a destroyed GEM object a new fence list is never installed, only the old one freed and therefore the iteration never finishes resulting in an endless loop. The solution is to drop the fence references only after the RCU grace period. The fixes tag is not necessary the patch introducing the problem, but the one making it so worse that we need to address it. This problem was pointed out by Sashiko-bot. Signed-off-by: Christian König Fixes: 912ff2ebd695 ("drm/i915: use the new iterator in i915_gem_busy_ioctl v2") CC: stable@vger.kernel.org Reviewed-by: Tvrtko Ursulin Signed-off-by: Tvrtko Ursulin Link: https://lore.kernel.org/r/20260903113621.54660-1-christian.koenig@amd.com (cherry picked from commit 5113479556025093bf8133bb2dcaa33be2d50921) Signed-off-by: Jani Nikula --- drivers/gpu/drm/i915/gem/i915_gem_object.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/gem/i915_gem_object.c b/drivers/gpu/drm/i915/gem/i915_gem_object.c index 5172d3982654..9e01f8b2079a 100644 --- a/drivers/gpu/drm/i915/gem/i915_gem_object.c +++ b/drivers/gpu/drm/i915/gem/i915_gem_object.c @@ -89,6 +89,7 @@ struct drm_i915_gem_object *i915_gem_object_alloc(void) void i915_gem_object_free(struct drm_i915_gem_object *obj) { + dma_resv_fini(&obj->base._resv); return kmem_cache_free(slab_objects, obj); } @@ -144,7 +145,6 @@ void __i915_gem_object_fini(struct drm_i915_gem_object *obj) { mutex_destroy(&obj->mm.get_page.lock); mutex_destroy(&obj->mm.get_dma_page.lock); - dma_resv_fini(&obj->base._resv); } /** From 335b0642812ee0f2b7798b634ce507b8b7c9fe0c Mon Sep 17 00:00:00 2001 From: Dapeng Mi Date: Thu, 17 Sep 2026 09:52:26 +0800 Subject: [PATCH 351/694] perf/x86/intel: Update arw_latency_data() mem-op direction handling Align arw_latency_data() with other *_latency_data() helpers by explicitly decoding LOAD/STORE event flags when setting the sampled memory operation direction. This keeps the latency data path behavior consistent across platforms and avoids relying on implicit direction inference. Signed-off-by: Dapeng Mi Signed-off-by: Peter Zijlstra (Intel) Signed-off-by: Ingo Molnar Link: https://patch.msgid.link/20260917015234.981153-5-dapeng1.mi@linux.intel.com --- arch/x86/events/intel/ds.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/arch/x86/events/intel/ds.c b/arch/x86/events/intel/ds.c index ea5b331c58d1..c0004b08b087 100644 --- a/arch/x86/events/intel/ds.c +++ b/arch/x86/events/intel/ds.c @@ -538,7 +538,11 @@ static u64 arw_latency_data(struct perf_event *event, u64 status) val |= P(BLK, NA); src.val = val; - if (event->hw.flags & PERF_X86_EVENT_PEBS_ST_HSW) + if (event->hw.flags & + (PERF_X86_EVENT_PEBS_LDLAT | PERF_X86_EVENT_PEBS_LD_HSW)) + src.mem_op = P(OP, LOAD); + if (event->hw.flags & + (PERF_X86_EVENT_PEBS_STLAT | PERF_X86_EVENT_PEBS_ST_HSW)) src.mem_op = P(OP, STORE); return src.val; From 7c944595cc43a664019edc22505b6d6f6039be5e Mon Sep 17 00:00:00 2001 From: Dapeng Mi Date: Thu, 17 Sep 2026 09:52:27 +0800 Subject: [PATCH 352/694] perf/x86/intel: Remove incorrect LionCove PEBS data-source constraints On Lion Cove, PEBS data source is valid only for these events: - MEM_TRANS_RETIRED.LOAD_LATENCY (0x1cd) - MEM_TRANS_RETIRED.STORE_SAMPLE (0x2cd) The perfmon database (https://github.com/intel/perfmon) previously tagged additional memory events such as MEM_INST_RETIRED.STLB_MISS_LOADS with L1_Hit_Indication, implying PEBS data-source support, which is incorrect. The database has since been fixed, but intel_lnc_pebs_event_constraints[] still follows the old definition and marks those events as data-source capable. As a result, get_data_src() may decode data-source information for events that do not provide valid PEBS data-source data and mislead users. Remove those non-data-source memory events from the Lion Cove PEBS constraint table so matching falls back to the regular non-PEBS constraints, which already provide the same counter constraints. Also update lnc_latency_data() to decode LOAD/STORE flags explicitly when setting memory operation direction, for consistency with other *_latency_data() helpers. Fixes: a932aa0e868f ("perf/x86: Add Lunar Lake and Arrow Lake support") Signed-off-by: Dapeng Mi Signed-off-by: Peter Zijlstra (Intel) Signed-off-by: Ingo Molnar Cc: Link: https://patch.msgid.link/20260917015234.981153-6-dapeng1.mi@linux.intel.com --- arch/x86/events/intel/ds.c | 22 +++++----------------- 1 file changed, 5 insertions(+), 17 deletions(-) diff --git a/arch/x86/events/intel/ds.c b/arch/x86/events/intel/ds.c index c0004b08b087..cbcc0b36536b 100644 --- a/arch/x86/events/intel/ds.c +++ b/arch/x86/events/intel/ds.c @@ -577,7 +577,11 @@ static u64 lnc_latency_data(struct perf_event *event, u64 status) val |= P(BLK, NA); src.val = val; - if (event->hw.flags & PERF_X86_EVENT_PEBS_ST_HSW) + if (event->hw.flags & + (PERF_X86_EVENT_PEBS_LDLAT | PERF_X86_EVENT_PEBS_LD_HSW)) + src.mem_op = P(OP, LOAD); + if (event->hw.flags & + (PERF_X86_EVENT_PEBS_STLAT | PERF_X86_EVENT_PEBS_ST_HSW)) src.mem_op = P(OP, STORE); return src.val; @@ -1510,24 +1514,8 @@ struct event_constraint intel_lnc_pebs_event_constraints[] = { INTEL_FLAGS_UEVENT_CONSTRAINT(0x012a, 0x1), /* OCR.* events */ INTEL_FLAGS_UEVENT_CONSTRAINT(0x012b, 0x1), /* OCR.* events */ - INTEL_FLAGS_UEVENT_CONSTRAINT(0x04a4, 0x1), /* TOPDOWN.BAD_SPEC_SLOTS */ - INTEL_FLAGS_UEVENT_CONSTRAINT(0x08a4, 0x1), /* TOPDOWN.BR_MISPREDICT_SLOTS */ - INTEL_FLAGS_UEVENT_CONSTRAINT(0x10a4, 0x8), /* TOPDOWN.MEMORY_BOUND_SLOTS */ - INTEL_HYBRID_LDLAT_CONSTRAINT(0x1cd, 0x3fc), INTEL_HYBRID_STLAT_CONSTRAINT(0x2cd, 0x3), - INTEL_FLAGS_UEVENT_CONSTRAINT_DATALA_LD(0x11d0, 0xf), /* MEM_INST_RETIRED.STLB_MISS_LOADS */ - INTEL_FLAGS_UEVENT_CONSTRAINT_DATALA_ST(0x12d0, 0xf), /* MEM_INST_RETIRED.STLB_MISS_STORES */ - INTEL_FLAGS_UEVENT_CONSTRAINT_DATALA_LD(0x21d0, 0xf), /* MEM_INST_RETIRED.LOCK_LOADS */ - INTEL_FLAGS_UEVENT_CONSTRAINT_DATALA_LD(0x41d0, 0xf), /* MEM_INST_RETIRED.SPLIT_LOADS */ - INTEL_FLAGS_UEVENT_CONSTRAINT_DATALA_ST(0x42d0, 0xf), /* MEM_INST_RETIRED.SPLIT_STORES */ - INTEL_FLAGS_UEVENT_CONSTRAINT_DATALA_LD(0x81d0, 0xf), /* MEM_INST_RETIRED.ALL_LOADS */ - INTEL_FLAGS_UEVENT_CONSTRAINT_DATALA_ST(0x82d0, 0xf), /* MEM_INST_RETIRED.ALL_STORES */ - INTEL_FLAGS_UEVENT_CONSTRAINT(0x87d0, 0x3ff), /* MEM_INST_RETIRED.ANY */ - - INTEL_FLAGS_EVENT_CONSTRAINT_DATALA_LD_RANGE(0xd1, 0xd4, 0xf), - - INTEL_FLAGS_EVENT_CONSTRAINT(0xd0, 0xf), /* * Everything else is handled by PMU_FL_PEBS_ALL, because we From 9f93d33ad65af9853974c1ec4ba1f937e8ba1376 Mon Sep 17 00:00:00 2001 From: Dapeng Mi Date: Thu, 17 Sep 2026 09:52:28 +0800 Subject: [PATCH 353/694] perf/x86/intel: Remove incorrect Panther Cove PEBS data-source constraints Same issue exists on Panther Cove, PEBS data source is valid only for these events: - MEM_TRANS_RETIRED.LOAD_LATENCY (0x1cd) - MEM_TRANS_RETIRED.STORE_SAMPLE (0x2cd) The perfmon database (https://github.com/intel/perfmon) previously tagged additional memory events such as MEM_INST_RETIRED.STLB_MISS_LOADS with L1_Hit_Indication, implying PEBS data-source support, which is incorrect. The database has since been fixed, but intel_pnc_pebs_event_constraints[] still follows the old definition and marks those events as data-source capable. As a result, get_data_src() may decode data-source information for events that do not provide valid PEBS data-source data and mislead users. Remove those non-data-source memory events from the Pather Cove PEBS constraint table so matching falls back to the regular non-PEBS constraints, which already provide the same counter constraints. Also update pnc_latency_data() to decode LOAD/STORE flags explicitly when setting memory operation direction, for consistency with other *_latency_data() helpers. Fixes: d345b6bb8860 ("perf/x86/intel: Add core PMU support for DMR") Signed-off-by: Dapeng Mi Signed-off-by: Peter Zijlstra (Intel) Signed-off-by: Ingo Molnar Cc: # v7.0+ Link: https://patch.msgid.link/20260917015234.981153-7-dapeng1.mi@linux.intel.com --- arch/x86/events/intel/ds.c | 18 +++++------------- 1 file changed, 5 insertions(+), 13 deletions(-) diff --git a/arch/x86/events/intel/ds.c b/arch/x86/events/intel/ds.c index cbcc0b36536b..72179bcf6d0d 100644 --- a/arch/x86/events/intel/ds.c +++ b/arch/x86/events/intel/ds.c @@ -639,7 +639,11 @@ u64 pnc_latency_data(struct perf_event *event, u64 status) val |= P(BLK, NA); src.val = val; - if (event->hw.flags & PERF_X86_EVENT_PEBS_ST_HSW) + if (event->hw.flags & + (PERF_X86_EVENT_PEBS_LDLAT | PERF_X86_EVENT_PEBS_LD_HSW)) + src.mem_op = P(OP, LOAD); + if (event->hw.flags & + (PERF_X86_EVENT_PEBS_STLAT | PERF_X86_EVENT_PEBS_ST_HSW)) src.mem_op = P(OP, STORE); return src.val; @@ -1528,18 +1532,6 @@ struct event_constraint intel_lnc_pebs_event_constraints[] = { struct event_constraint intel_pnc_pebs_event_constraints[] = { INTEL_HYBRID_LDLAT_CONSTRAINT(0x1cd, 0xfc), INTEL_HYBRID_STLAT_CONSTRAINT(0x2cd, 0x3), - INTEL_FLAGS_UEVENT_CONSTRAINT_DATALA_LD(0x11d0, 0xf), /* MEM_INST_RETIRED.STLB_MISS_LOADS */ - INTEL_FLAGS_UEVENT_CONSTRAINT_DATALA_ST(0x12d0, 0xf), /* MEM_INST_RETIRED.STLB_MISS_STORES */ - INTEL_FLAGS_UEVENT_CONSTRAINT_DATALA_LD(0x21d0, 0xf), /* MEM_INST_RETIRED.LOCK_LOADS */ - INTEL_FLAGS_UEVENT_CONSTRAINT_DATALA_LD(0x41d0, 0xf), /* MEM_INST_RETIRED.SPLIT_LOADS */ - INTEL_FLAGS_UEVENT_CONSTRAINT_DATALA_ST(0x42d0, 0xf), /* MEM_INST_RETIRED.SPLIT_STORES */ - INTEL_FLAGS_UEVENT_CONSTRAINT_DATALA_LD(0x81d0, 0xf), /* MEM_INST_RETIRED.ALL_LOADS */ - INTEL_FLAGS_UEVENT_CONSTRAINT_DATALA_ST(0x82d0, 0xf), /* MEM_INST_RETIRED.ALL_STORES */ - - INTEL_FLAGS_EVENT_CONSTRAINT_DATALA_LD_RANGE(0xd1, 0xd4, 0xf), - - INTEL_FLAGS_EVENT_CONSTRAINT(0xd0, 0xf), - INTEL_FLAGS_EVENT_CONSTRAINT(0xd6, 0xf), /* * Everything else is handled by PMU_FL_PEBS_ALL, because we From 0ac5d6ca2c3b7d047963b67dccef17902dc6c017 Mon Sep 17 00:00:00 2001 From: Dapeng Mi Date: Thu, 17 Sep 2026 09:52:29 +0800 Subject: [PATCH 354/694] perf/x86/intel: Fix Panther Cove PEBS data-source snoop states For Panther Cove, the snoop states for the data source encodings "Prefetch Promotion" and "Cross Core Prefetch Promotion" should be SNOOP_NONE instead of SNOOP_MISS. Fix the incorrect snooping states for Panther Cove. Fixes: d2bdcde9626c ("perf/x86/intel: Add support for PEBS memory auxiliary info field in DMR") Signed-off-by: Dapeng Mi Signed-off-by: Peter Zijlstra (Intel) Signed-off-by: Ingo Molnar Cc: # v7.0+ Link: https://patch.msgid.link/20260917015234.981153-8-dapeng1.mi@linux.intel.com --- arch/x86/events/intel/ds.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/x86/events/intel/ds.c b/arch/x86/events/intel/ds.c index 72179bcf6d0d..d0329a7eb8a5 100644 --- a/arch/x86/events/intel/ds.c +++ b/arch/x86/events/intel/ds.c @@ -277,8 +277,8 @@ static u64 pnc_pebs_l2_hit_data_source[PNC_PEBS_DATA_SOURCE_MAX] = { 0, /* 0x06: Reserved */ OP_LH | P(LVL, L2) | LEVEL(L2) | P(SNOOP, HIT), /* 0x07: L2 Hit Snoop HIT */ OP_LH | P(LVL, L2) | LEVEL(L2) | P(SNOOP, HITM), /* 0x08: L2 Hit Snoop Hit Modified */ - OP_LH | P(LVL, L2) | LEVEL(L2) | P(SNOOP, MISS), /* 0x09: Prefetch Promotion */ - OP_LH | P(LVL, L2) | LEVEL(L2) | P(SNOOP, MISS), /* 0x0a: Cross Core Prefetch Promotion */ + OP_LH | P(LVL, L2) | LEVEL(L2) | P(SNOOP, NONE), /* 0x09: Prefetch Promotion */ + OP_LH | P(LVL, L2) | LEVEL(L2) | P(SNOOP, NONE), /* 0x0a: Cross Core Prefetch Promotion */ 0, /* 0x0b: Reserved */ 0, /* 0x0c: Reserved */ 0, /* 0x0d: Reserved */ From 858b37ca19d3f695f7fa94cd14be5a856fd8e7d7 Mon Sep 17 00:00:00 2001 From: Dapeng Mi Date: Thu, 17 Sep 2026 09:52:30 +0800 Subject: [PATCH 355/694] perf/x86/intel: Delete dead NVL PEBS data-source initcall Nova Lake now uses the OMR data-source table for PEBS data-source decoding and no longer depends on the legacy static pebs_data_source[] mapping. Remove the dead intel_pmu_pebs_data_source_lnl() initialization call for NVL. Fixes: c847a208f43b ("perf/x86/intel: Add core PMU support for Novalake") Signed-off-by: Dapeng Mi Signed-off-by: Peter Zijlstra (Intel) Signed-off-by: Ingo Molnar Link: https://patch.msgid.link/20260917015234.981153-9-dapeng1.mi@linux.intel.com --- arch/x86/events/intel/core.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/arch/x86/events/intel/core.c b/arch/x86/events/intel/core.c index 757d4082de60..a7dbbed8f94e 100644 --- a/arch/x86/events/intel/core.c +++ b/arch/x86/events/intel/core.c @@ -8824,8 +8824,6 @@ __init int intel_pmu_init(void) /* Initialize Atom core specific PerfMon capabilities.*/ pmu = &x86_pmu.hybrid_pmu[X86_HYBRID_PMU_ATOM_IDX]; intel_pmu_init_arw(&pmu->pmu); - - intel_pmu_pebs_data_source_lnl(); break; default: From 04a7ef3b7aa3af34202285043e3423a2e3983595 Mon Sep 17 00:00:00 2001 From: Dapeng Mi Date: Thu, 17 Sep 2026 09:52:31 +0800 Subject: [PATCH 356/694] perf/x86/intel: Constrain Panther Cove UOPS_DISPATCHED events to PMCs 0-3 Per the latest Panther Cove event definitions, the following events are only supported on PMCs 0-3: - UOPS_DISPATCHED.INT_EU_ALL (0x1b2) - UOPS_DISPATCHED.ALU (0x2b2) Add explicit event constraints for these two events so scheduling does not place them on unsupported counters. Fixes: d345b6bb8860 ("perf/x86/intel: Add core PMU support for DMR") Signed-off-by: Dapeng Mi Signed-off-by: Peter Zijlstra (Intel) Signed-off-by: Ingo Molnar Cc: # v7.0+ Link: https://patch.msgid.link/20260917015234.981153-10-dapeng1.mi@linux.intel.com --- arch/x86/events/intel/core.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/arch/x86/events/intel/core.c b/arch/x86/events/intel/core.c index a7dbbed8f94e..ecf6d6325dfe 100644 --- a/arch/x86/events/intel/core.c +++ b/arch/x86/events/intel/core.c @@ -506,6 +506,8 @@ static struct event_constraint intel_pnc_event_constraints[] = { INTEL_EVENT_CONSTRAINT(0xce, 0x1), INTEL_UEVENT_CONSTRAINT(0x01b1, 0x8), + INTEL_UEVENT_CONSTRAINT(0x01b2, 0xf), + INTEL_UEVENT_CONSTRAINT(0x02b2, 0xf), INTEL_UEVENT_CONSTRAINT(0x0847, 0xf), INTEL_UEVENT_CONSTRAINT(0x0446, 0xf), INTEL_UEVENT_CONSTRAINT(0x0846, 0xf), From ff1621adfdd7cfb73f2ccdd3856cd3462ac98ac5 Mon Sep 17 00:00:00 2001 From: Dapeng Mi Date: Thu, 17 Sep 2026 09:52:32 +0800 Subject: [PATCH 357/694] perf/x86/intel: Fix precise OMR event scheduling for DMR/NVL The latest perfmon event database introduces below precise OMR event support for DMR/NVL: - MEM_LOAD_L2_MISS_RETIRED.* (event 0xd6) - MEM_STORE_L2_MISS_RETIRED.* (event 0x4f) These events use the same OMR MSRs as the existing OMR events, but they are not listed in intel_pnc_extra_regs[]. As a result, perf cannot assign the required OMR extra registers when scheduling them. Add the new precise OMR events to intel_pnc_extra_regs[] so they can be scheduled with the correct OMR MSRs. MEM_LOAD_L2_MISS_RETIRED.* remains limited to GP counters 0-3, while MEM_STORE_L2_MISS_RETIRED.* is available on all GP counters. Signed-off-by: Dapeng Mi Signed-off-by: Peter Zijlstra (Intel) Signed-off-by: Ingo Molnar Link: https://patch.msgid.link/20260917015234.981153-11-dapeng1.mi@linux.intel.com --- arch/x86/events/intel/core.c | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/arch/x86/events/intel/core.c b/arch/x86/events/intel/core.c index ecf6d6325dfe..22715602f7f8 100644 --- a/arch/x86/events/intel/core.c +++ b/arch/x86/events/intel/core.c @@ -522,6 +522,14 @@ static struct extra_reg intel_pnc_extra_regs[] __read_mostly = { INTEL_UEVENT_EXTRA_REG(0x022a, MSR_OMR_1, 0x40ffffff0000ffffull, OMR_1), INTEL_UEVENT_EXTRA_REG(0x042a, MSR_OMR_2, 0x40ffffff0000ffffull, OMR_2), INTEL_UEVENT_EXTRA_REG(0x082a, MSR_OMR_3, 0x40ffffff0000ffffull, OMR_3), + INTEL_UEVENT_EXTRA_REG(0x014f, MSR_OMR_0, 0x40ffffff0000ffffull, OMR_0), + INTEL_UEVENT_EXTRA_REG(0x024f, MSR_OMR_1, 0x40ffffff0000ffffull, OMR_1), + INTEL_UEVENT_EXTRA_REG(0x044f, MSR_OMR_2, 0x40ffffff0000ffffull, OMR_2), + INTEL_UEVENT_EXTRA_REG(0x084f, MSR_OMR_3, 0x40ffffff0000ffffull, OMR_3), + INTEL_UEVENT_EXTRA_REG(0x01d6, MSR_OMR_0, 0x40ffffff0000ffffull, OMR_0), + INTEL_UEVENT_EXTRA_REG(0x02d6, MSR_OMR_1, 0x40ffffff0000ffffull, OMR_1), + INTEL_UEVENT_EXTRA_REG(0x04d6, MSR_OMR_2, 0x40ffffff0000ffffull, OMR_2), + INTEL_UEVENT_EXTRA_REG(0x08d6, MSR_OMR_3, 0x40ffffff0000ffffull, OMR_3), INTEL_UEVENT_PEBS_LDLAT_EXTRA_REG(0x01cd), INTEL_UEVENT_EXTRA_REG(0x02c6, MSR_PEBS_FRONTEND, 0x9, FE), INTEL_UEVENT_EXTRA_REG(0x03c6, MSR_PEBS_FRONTEND, 0x7fff1f, FE), From 0102c8c7fdfc4bd700971daff2b94470f4eafae4 Mon Sep 17 00:00:00 2001 From: Dapeng Mi Date: Thu, 17 Sep 2026 09:52:33 +0800 Subject: [PATCH 358/694] perf/x86/intel: Rename DMR offcore_rsp attribute to offmodule_rsp DMR introduces Offmodule Response events in place of the legacy Offcore Response events, but it still exposes the inherited offcore_rsp PMU attribute for programming the corresponding MSR data. Rename the DMR PMU attribute to offmodule_rsp so the sysfs interface matches the underlying event name and avoids user & tooling confusion. Signed-off-by: Dapeng Mi Signed-off-by: Peter Zijlstra (Intel) Signed-off-by: Ingo Molnar Link: https://patch.msgid.link/20260917015234.981153-12-dapeng1.mi@linux.intel.com --- arch/x86/events/intel/core.c | 23 ++++++++++++++++++++--- 1 file changed, 20 insertions(+), 3 deletions(-) diff --git a/arch/x86/events/intel/core.c b/arch/x86/events/intel/core.c index 22715602f7f8..3843dbe530a5 100644 --- a/arch/x86/events/intel/core.c +++ b/arch/x86/events/intel/core.c @@ -6601,6 +6601,8 @@ static void intel_pmu_filter(struct pmu *pmu, int cpu, bool *ret) PMU_FORMAT_ATTR(offcore_rsp, "config1:0-63"); +PMU_FORMAT_ATTR(offmodule_rsp, "config1:0-63"); + PMU_FORMAT_ATTR(ldlat, "config1:0-15"); PMU_FORMAT_ATTR(frontend, "config1:0-23"); @@ -6649,6 +6651,20 @@ static struct attribute *skl_format_attr[] = { NULL, }; +static struct attribute *pnc_format_attr_rtm[] = { + &format_attr_in_tx.attr, + &format_attr_in_tx_cp.attr, + &format_attr_offmodule_rsp.attr, + &format_attr_ldlat.attr, + NULL +}; + +static struct attribute *pnc_format_attr[] = { + &format_attr_offmodule_rsp.attr, + &format_attr_ldlat.attr, + NULL +}; + static __initconst const struct x86_pmu core_pmu = { .name = "core", .handle_irq = x86_pmu_handle_irq, @@ -8589,6 +8605,8 @@ __init int intel_pmu_init(void) case INTEL_DIAMONDRAPIDS_X: intel_pmu_init_pnc(NULL); x86_pmu.pebs_latency_data = pnc_latency_data; + extra_attr = boot_cpu_has(X86_FEATURE_RTM) ? + pnc_format_attr_rtm : pnc_format_attr; pr_cont("Panthercove events, "); name = "panthercove"; @@ -8597,13 +8615,12 @@ __init int intel_pmu_init(void) glc_common: intel_pmu_init_glc(NULL); intel_pmu_pebs_data_source_skl(true); - + extra_attr = boot_cpu_has(X86_FEATURE_RTM) ? + hsw_format_attr : nhm_format_attr; glc_base: x86_pmu.pebs_ept = 1; x86_pmu.hw_config = hsw_hw_config; x86_pmu.get_event_constraints = glc_get_event_constraints; - extra_attr = boot_cpu_has(X86_FEATURE_RTM) ? - hsw_format_attr : nhm_format_attr; extra_skl_attr = skl_format_attr; mem_attr = glc_events_attrs; td_attr = glc_td_events_attrs; From d4d9ccbad527af115b8e83a7a2b74785c0e8dfea Mon Sep 17 00:00:00 2001 From: Dapeng Mi Date: Thu, 17 Sep 2026 09:52:34 +0800 Subject: [PATCH 359/694] perf/x86/intel: Rename NVL offcore_rsp attribute to offmodule_rsp NVL introduces Offmodule Response events in place of the legacy Offcore Response events, but it still exposes the inherited offcore_rsp PMU attribute for programming the corresponding MSR data. Rename the NVL PMU attribute to offmodule_rsp so the sysfs interface matches the underlying event name and avoids user & tooling confusion. Signed-off-by: Dapeng Mi Signed-off-by: Peter Zijlstra (Intel) Signed-off-by: Ingo Molnar Link: https://patch.msgid.link/20260917015234.981153-13-dapeng1.mi@linux.intel.com --- arch/x86/events/intel/core.c | 20 +++++++++++++++++++- 1 file changed, 19 insertions(+), 1 deletion(-) diff --git a/arch/x86/events/intel/core.c b/arch/x86/events/intel/core.c index 3843dbe530a5..3ef80882843e 100644 --- a/arch/x86/events/intel/core.c +++ b/arch/x86/events/intel/core.c @@ -7530,6 +7530,7 @@ static struct attribute *adl_hybrid_tsx_attrs[] = { FORMAT_ATTR_HYBRID(in_tx, hybrid_big); FORMAT_ATTR_HYBRID(in_tx_cp, hybrid_big); FORMAT_ATTR_HYBRID(offcore_rsp, hybrid_big_small_tiny); +FORMAT_ATTR_HYBRID(offmodule_rsp, hybrid_big_small_tiny); FORMAT_ATTR_HYBRID(ldlat, hybrid_big_small_tiny); FORMAT_ATTR_HYBRID(frontend, hybrid_big); @@ -7568,6 +7569,23 @@ static struct attribute *mtl_hybrid_extra_attr[] = { NULL }; +static struct attribute *nvl_hybrid_extra_attr_rtm[] = { + ADL_HYBRID_RTM_FORMAT_ATTR, + FORMAT_HYBRID_PTR(offmodule_rsp), + FORMAT_HYBRID_PTR(ldlat), + FORMAT_HYBRID_PTR(frontend), + FORMAT_HYBRID_PTR(snoop_rsp), + NULL +}; + +static struct attribute *nvl_hybrid_extra_attr[] = { + FORMAT_HYBRID_PTR(offmodule_rsp), + FORMAT_HYBRID_PTR(ldlat), + FORMAT_HYBRID_PTR(frontend), + FORMAT_HYBRID_PTR(snoop_rsp), + NULL +}; + static bool is_attr_for_this_pmu(struct kobject *kobj, struct attribute *attr) { struct device *dev = kobj_to_dev(kobj); @@ -8842,7 +8860,7 @@ __init int intel_pmu_init(void) mem_attr = mtl_hybrid_mem_attrs; tsx_attr = adl_hybrid_tsx_attrs; extra_attr = boot_cpu_has(X86_FEATURE_RTM) ? - mtl_hybrid_extra_attr_rtm : mtl_hybrid_extra_attr; + nvl_hybrid_extra_attr_rtm : nvl_hybrid_extra_attr; /* Initialize big core specific PerfMon capabilities.*/ pmu = &x86_pmu.hybrid_pmu[X86_HYBRID_PMU_CORE_IDX]; From 09b7040a1b79f4f61cdad6d9af972e045bb7c498 Mon Sep 17 00:00:00 2001 From: Niklas Schnelle Date: Wed, 16 Sep 2026 17:14:10 +0200 Subject: [PATCH 360/694] s390/pci: Fix leak of struct pci_dev reference in zpci_report_status() In zpci_report_status(), a reference to the pdev associated with the zdev being reported about is acquired using pci_get_slot(). This reference needs to be dropped with pci_dev_put(), but this call is missing, thus leaking the reference. On subsequent hot unplug, this will cause the struct pci_dev to not be released, leaking memory and preventing reattach. At the same time, the only existing caller already holds a pdev reference. So instead of reacquiring and then dropping another reference, simply pass the existing pdev pointer to zpci_report_status(). This gets rid of the need for pci_get_slot() as well as the zdev->zbus check. Cc: stable@vger.kernel.org Fixes: 4ec6054e7321 ("s390/pci: Report PCI error recovery results via SCLP") Signed-off-by: Niklas Schnelle Reviewed-by: Benjamin Block Reviewed-by: Farhan Ali Signed-off-by: Heiko Carstens --- arch/s390/pci/pci_event.c | 2 +- arch/s390/pci/pci_report.c | 11 +++++------ arch/s390/pci/pci_report.h | 4 +++- 3 files changed, 9 insertions(+), 8 deletions(-) diff --git a/arch/s390/pci/pci_event.c b/arch/s390/pci/pci_event.c index f317a1465dad..3b4941b65840 100644 --- a/arch/s390/pci/pci_event.c +++ b/arch/s390/pci/pci_event.c @@ -298,7 +298,7 @@ static pci_ers_result_t zpci_event_attempt_error_recovery(struct pci_dev *pdev, pci_uevent_ers(pdev, PCI_ERS_RESULT_RECOVERED); out_unlock: device_unlock(&pdev->dev); - zpci_report_status(zdev, "recovery", status_str); + zpci_report_status(zdev, pdev, "recovery", status_str); return ers_res; } diff --git a/arch/s390/pci/pci_report.c b/arch/s390/pci/pci_report.c index 7030f7052926..867419779219 100644 --- a/arch/s390/pci/pci_report.c +++ b/arch/s390/pci/pci_report.c @@ -89,7 +89,8 @@ static struct debug_view debug_log_view = { /** * zpci_report_status - Report the status of operations on a PCI device - * @zdev: The PCI device for which to report status + * @zdev: The zPCI device for which to report status + * @pdev: The PCI device associated with the zdev if any, NULL otherwise * @operation: A string representing the operation reported * @status: A string representing the status of the operation * @@ -103,15 +104,15 @@ static struct debug_view debug_log_view = { * * Return: 0 on success an error code < 0 otherwise. */ -int zpci_report_status(struct zpci_dev *zdev, const char *operation, const char *status) +int zpci_report_status(struct zpci_dev *zdev, struct pci_dev *pdev, + const char *operation, const char *status) { struct zpci_report_error *report; struct pci_driver *driver = NULL; - struct pci_dev *pdev = NULL; char *buf, *end; int ret; - if (!zdev || !zdev->zbus) + if (!zdev) return -ENODEV; /* Protected virtualization hosts get nothing from us */ @@ -121,8 +122,6 @@ int zpci_report_status(struct zpci_dev *zdev, const char *operation, const char report = (void *)get_zeroed_page(GFP_KERNEL); if (!report) return -ENOMEM; - if (zdev->zbus->bus) - pdev = pci_get_slot(zdev->zbus->bus, zdev->devfn); if (pdev) driver = to_pci_driver(pdev->dev.driver); diff --git a/arch/s390/pci/pci_report.h b/arch/s390/pci/pci_report.h index e08003d51a97..dd7b0b05001c 100644 --- a/arch/s390/pci/pci_report.h +++ b/arch/s390/pci/pci_report.h @@ -8,9 +8,11 @@ */ #ifndef __S390_PCI_REPORT_H #define __S390_PCI_REPORT_H +#include struct zpci_dev; -int zpci_report_status(struct zpci_dev *zdev, const char *operation, const char *status); +int zpci_report_status(struct zpci_dev *zdev, struct pci_dev *pdev, + const char *operation, const char *status); #endif /* __S390_PCI_REPORT_H */ From 0261aef4b15efcee2860ab857e5cb05e9bfa47b0 Mon Sep 17 00:00:00 2001 From: Niklas Schnelle Date: Wed, 16 Sep 2026 17:14:11 +0200 Subject: [PATCH 361/694] s390/pci: Fix missing device lock in zpci_report_status() When pdev is non-NULL, zpci_report_status() accesses the device's driver. To get a consistent state matching the recovery, the device lock needs to be held. Do so by expanding the existing device lock critical section. The lock only needs to be held when the pdev is non-NULL, so extract the pdev-specific reporting into a helper function which also adds a lockdep assertion to detect calls without the device lock held. Cc: stable@vger.kernel.org Fixes: 4ec6054e7321 ("s390/pci: Report PCI error recovery results via SCLP") Signed-off-by: Niklas Schnelle Reviewed-by: Benjamin Block Reviewed-by: Farhan Ali Signed-off-by: Heiko Carstens --- arch/s390/pci/pci_event.c | 2 +- arch/s390/pci/pci_report.c | 21 +++++++++++++++------ 2 files changed, 16 insertions(+), 7 deletions(-) diff --git a/arch/s390/pci/pci_event.c b/arch/s390/pci/pci_event.c index 3b4941b65840..ec93f34b6e19 100644 --- a/arch/s390/pci/pci_event.c +++ b/arch/s390/pci/pci_event.c @@ -297,8 +297,8 @@ static pci_ers_result_t zpci_event_attempt_error_recovery(struct pci_dev *pdev, driver->err_handler->resume(pdev); pci_uevent_ers(pdev, PCI_ERS_RESULT_RECOVERED); out_unlock: - device_unlock(&pdev->dev); zpci_report_status(zdev, pdev, "recovery", status_str); + device_unlock(&pdev->dev); return ers_res; } diff --git a/arch/s390/pci/pci_report.c b/arch/s390/pci/pci_report.c index 867419779219..72ecabf7003c 100644 --- a/arch/s390/pci/pci_report.c +++ b/arch/s390/pci/pci_report.c @@ -87,6 +87,19 @@ static struct debug_view debug_log_view = { NULL }; +static ssize_t zpci_report_pdev(struct pci_dev *pdev, char *buf, size_t size) +{ + struct pci_driver *driver; + const char *start = buf; + char *end = buf + size; + + device_lock_assert(&pdev->dev); + buf += scnprintf(buf, end - buf, "state: %s\n", zpci_state_str(pdev->error_state)); + driver = to_pci_driver(pdev->dev.driver); + buf += scnprintf(buf, end - buf, "driver: %s\n", (driver) ? driver->name : "n/a"); + return buf - start; +} + /** * zpci_report_status - Report the status of operations on a PCI device * @zdev: The zPCI device for which to report status @@ -108,7 +121,6 @@ int zpci_report_status(struct zpci_dev *zdev, struct pci_dev *pdev, const char *operation, const char *status) { struct zpci_report_error *report; - struct pci_driver *driver = NULL; char *buf, *end; int ret; @@ -122,16 +134,13 @@ int zpci_report_status(struct zpci_dev *zdev, struct pci_dev *pdev, report = (void *)get_zeroed_page(GFP_KERNEL); if (!report) return -ENOMEM; - if (pdev) - driver = to_pci_driver(pdev->dev.driver); buf = report->data.log_data; end = report->data.log_data + ZPCI_REPORT_DATA_SIZE; buf += scnprintf(buf, end - buf, "report: %s\n", operation); buf += scnprintf(buf, end - buf, "status: %s\n", status); - buf += scnprintf(buf, end - buf, "state: %s\n", - (pdev) ? zpci_state_str(pdev->error_state) : "n/a"); - buf += scnprintf(buf, end - buf, "driver: %s\n", (driver) ? driver->name : "n/a"); + if (pdev) + buf += zpci_report_pdev(pdev, buf, end - buf); ret = debug_dump(pci_debug_msg_id, &debug_log_view, buf, end - buf, true); if (ret < 0) pr_err("Reading PCI debug messages failed with code %d\n", ret); From a1120bea9bc8ea9d9ab2f9904a63b9a228bf2ffc Mon Sep 17 00:00:00 2001 From: Niklas Schnelle Date: Wed, 16 Sep 2026 17:14:12 +0200 Subject: [PATCH 362/694] s390/pci: Report SCLP status on error events when no pdev is associated With commit 4ec6054e7321 ("s390/pci: Report PCI error recovery results via SCLP") SCLP reports are generated when recovery is performed in response to an error event. If such an error event arrives but no pdev is currently associated with the zdev, e.g. because it was removed or not yet probed, no report is generated. Fix this by generating a report specific to an error event for a zdev without an associated pdev. Cc: stable@vger.kernel.org Fixes: 4ec6054e7321 ("s390/pci: Report PCI error recovery results via SCLP") Signed-off-by: Niklas Schnelle Reviewed-by: Benjamin Block Reviewed-by: Farhan Ali Signed-off-by: Heiko Carstens --- arch/s390/pci/pci_event.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/arch/s390/pci/pci_event.c b/arch/s390/pci/pci_event.c index ec93f34b6e19..7b538c6ae11b 100644 --- a/arch/s390/pci/pci_event.c +++ b/arch/s390/pci/pci_event.c @@ -361,8 +361,10 @@ static void __zpci_event_error(struct zpci_ccdf_err *ccdf) __zpci_event_print_error(pdev, ccdf); - if (!pdev) + if (!pdev) { + zpci_report_status(zdev, NULL, "error event", "no pdev bound"); goto no_pdev; + } switch (ccdf->pec) { case 0x002a: /* Error event concerns FMB */ From 3a43be7a1fd06a35cf9e621b88283b3b6e7d281c Mon Sep 17 00:00:00 2001 From: Niklas Schnelle Date: Wed, 16 Sep 2026 17:14:13 +0200 Subject: [PATCH 363/694] s390/pci: Don't report recovery success on skipped recovery When a PCI device is already in the permanent failure state, recovery is skipped, but the SCLP recovery report still shows success. Fix this by changing the status string to explicitly state that recovery was skipped due to permanent failure. Cc: stable@vger.kernel.org Fixes: 4ec6054e7321 ("s390/pci: Report PCI error recovery results via SCLP") Signed-off-by: Niklas Schnelle Reviewed-by: Benjamin Block Reviewed-by: Farhan Ali Signed-off-by: Heiko Carstens --- arch/s390/pci/pci_event.c | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/s390/pci/pci_event.c b/arch/s390/pci/pci_event.c index 7b538c6ae11b..d6af4015223e 100644 --- a/arch/s390/pci/pci_event.c +++ b/arch/s390/pci/pci_event.c @@ -226,6 +226,7 @@ static pci_ers_result_t zpci_event_attempt_error_recovery(struct pci_dev *pdev, device_lock(&pdev->dev); if (pdev->error_state == pci_channel_io_perm_failure) { ers_res = PCI_ERS_RESULT_DISCONNECT; + status_str = "skipped (permanent failure)"; goto out_unlock; } pdev->error_state = pci_channel_io_frozen; From f4d04425e66af2ecee9d1a49ae0484436f3c2fd1 Mon Sep 17 00:00:00 2001 From: Peter Oberparleiter Date: Mon, 21 Sep 2026 15:16:48 +0200 Subject: [PATCH 364/694] s390/cmf: Fix virtual vs physical address confusion The measurement block address is an absolute address. Define the associated schib_config and schib fields as dma64_t to enable automatic detection of incorrect assignments. Also add the missing virt_to_dma64() translation. Without this fix, a wrong address will be used by firmware when storing extended format channel measurement data on kernels built with CONFIG_RANDOMIZE_IDENTITY_BASE=y. Fixes: 14edd0d73bfe ("s390/cmf: fix virtual vs physical address confusion") Cc: stable@vger.kernel.org Signed-off-by: Peter Oberparleiter Reviewed-by: Heiko Carstens Signed-off-by: Heiko Carstens --- drivers/s390/cio/cio.h | 5 +++-- drivers/s390/cio/cmf.c | 2 +- 2 files changed, 4 insertions(+), 3 deletions(-) diff --git a/drivers/s390/cio/cio.h b/drivers/s390/cio/cio.h index bad142c536e1..6d28a62bc67a 100644 --- a/drivers/s390/cio/cio.h +++ b/drivers/s390/cio/cio.h @@ -7,6 +7,7 @@ #include #include #include +#include #include #include #include @@ -49,7 +50,7 @@ struct pmcw { /* Target SCHIB configuration. */ struct schib_config { - u64 mba; + dma64_t mba; u32 intparm; u16 mbi; u32 isc:3; @@ -66,7 +67,7 @@ struct schib_config { struct schib { struct pmcw pmcw; /* path management control word */ union scsw scsw; /* subchannel status word */ - __u64 mba; /* measurement block address */ + dma64_t mba; /* measurement block address */ __u8 mda[4]; /* model dependent area */ } __attribute__ ((packed,aligned(4))); diff --git a/drivers/s390/cio/cmf.c b/drivers/s390/cio/cmf.c index 92ab3d546fe4..66b14fedbd18 100644 --- a/drivers/s390/cio/cmf.c +++ b/drivers/s390/cio/cmf.c @@ -183,7 +183,7 @@ static int set_schib(struct ccw_device *cdev, u32 mme, int mbfc, sch->config.mbfc = mbfc; /* address can be either a block address or a block index */ if (mbfc) - sch->config.mba = address; + sch->config.mba = address ? virt_to_dma64((void *)address) : 0; else sch->config.mbi = address; From 4467df89dbca6a3e9dbc343a315324bb192603d6 Mon Sep 17 00:00:00 2001 From: Mikhail Zaslonko Date: Wed, 16 Sep 2026 18:53:30 +0200 Subject: [PATCH 365/694] s390/debug: Reject NULL debug info in debug_dump() debug_dump() passes id on to debug_info_copy(), which dereferences in->name unchecked. debug_unregister(), debug_set_size() and debug_register_view() guard against a NULL id but debug_dump() does not. Nothing reaches this today, but add the check for consistency. Reported-by: Sashiko Closes: https://lore.kernel.org/all/20260910110147.96E851F000FF@smtp.kernel.org/ Signed-off-by: Mikhail Zaslonko Reviewed-by: Heiko Carstens Signed-off-by: Heiko Carstens --- arch/s390/kernel/debug.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/arch/s390/kernel/debug.c b/arch/s390/kernel/debug.c index b5bf8284dbfc..354ce78fc0cf 100644 --- a/arch/s390/kernel/debug.c +++ b/arch/s390/kernel/debug.c @@ -823,6 +823,9 @@ ssize_t debug_dump(debug_info_t *id, struct debug_view *view, file_private_info_t *p_info; size_t size, offset = 0; + if (!id) + return -EINVAL; + /* Need space for '\0' byte */ if (buf_size < 1) return 0; From 28e29992b034acffc9342df216c06097825ce610 Mon Sep 17 00:00:00 2001 From: Mikhail Zaslonko Date: Fri, 18 Sep 2026 17:13:51 +0200 Subject: [PATCH 366/694] s390/debug: Do not register views for failed static debug areas __REGISTER_STATIC_DEBUG_INFO() calls debug_register_view() unconditionally, even when debug_register_static() has failed. In that case _debug_register() was never reached and id->debugfs_root_entry is still NULL, so debugfs_create_file() places the view file in the debugfs root directory. For sclp_err this leaves a /sys/kernel/debug/hex_ascii file with nothing to indicate which debug log it belongs to. debug_register_static() is not exported and the macro is its only caller, so let it return an error code and skip the view registration when it fails. No debugfs files are created for such an area then. Reproduce by booting with s390dbf=sclp_err::100000000. The sclp_err registration fails, no s390dbf/sclp_err/ directory is created, and a hex_ascii file appears in the debugfs root instead. Fixes: d72541f94512 ("s390/debug: add early tracing support") Signed-off-by: Mikhail Zaslonko Reviewed-by: Heiko Carstens Signed-off-by: Heiko Carstens --- arch/s390/include/asm/debug.h | 8 ++++++-- arch/s390/kernel/debug.c | 12 +++++++++--- 2 files changed, 15 insertions(+), 5 deletions(-) diff --git a/arch/s390/include/asm/debug.h b/arch/s390/include/asm/debug.h index 39d484c59774..ad438d6352c8 100644 --- a/arch/s390/include/asm/debug.h +++ b/arch/s390/include/asm/debug.h @@ -460,7 +460,11 @@ static int VNAME(var, active_entries)[EARLY_AREAS] __initdata #define __REGISTER_STATIC_DEBUG_INFO(var, name, pages, areas, view) \ static int __init VNAME(var, reg)(void) \ { \ - debug_register_static(&var, (pages), (areas)); \ + int rc; \ + \ + rc = debug_register_static(&var, (pages), (areas)); \ + if (rc) \ + return rc; \ debug_register_view(&var, (view)); \ return 0; \ } \ @@ -493,7 +497,7 @@ static debug_info_t __refdata var = \ static debug_info_t __used __section(".s390dbf_info") *VNAME(var, info) = &var; \ __REGISTER_STATIC_DEBUG_INFO(var, name, pages, nr_areas, view) -void debug_register_static(debug_info_t *id, int pages_per_area, int nr_areas); +int debug_register_static(debug_info_t *id, int pages_per_area, int nr_areas); #endif /* MODULE */ diff --git a/arch/s390/kernel/debug.c b/arch/s390/kernel/debug.c index 354ce78fc0cf..142dc1487187 100644 --- a/arch/s390/kernel/debug.c +++ b/arch/s390/kernel/debug.c @@ -953,8 +953,12 @@ EXPORT_SYMBOL(debug_register); * * Note: This function is called automatically via an initcall generated by * DEFINE_STATIC_DEBUG_INFO. + * + * Return: + * - 0 on success + * - negative error code on failure */ -void debug_register_static(debug_info_t *id, int pages_per_area, int nr_areas) +int debug_register_static(debug_info_t *id, int pages_per_area, int nr_areas) { unsigned long flags; debug_info_t *copy; @@ -962,7 +966,7 @@ void debug_register_static(debug_info_t *id, int pages_per_area, int nr_areas) if (!initialized) { pr_err("Tried to register debug feature %s too early\n", id->name); - return; + return -EINVAL; } debug_get_param(id->name, &id->level, &pages_per_area, false); @@ -978,7 +982,7 @@ void debug_register_static(debug_info_t *id, int pages_per_area, int nr_areas) id->active_entries = NULL; raw_spin_unlock_irqrestore(&id->lock, flags); - return; + return -ENOMEM; } /* Replace static trace area with dynamic copy. */ @@ -996,6 +1000,8 @@ void debug_register_static(debug_info_t *id, int pages_per_area, int nr_areas) mutex_lock(&debug_mutex); _debug_register(id); mutex_unlock(&debug_mutex); + + return 0; } /* Remove debugfs entries. */ From a644f09b2090ad22a13fbcf9d141084f573108ef Mon Sep 17 00:00:00 2001 From: Wentao Liang Date: Thu, 17 Sep 2026 11:01:35 +0000 Subject: [PATCH 367/694] fsl/fman: Fix clk reference leak in read_dts_node() of_clk_get() returns a clock with its reference count incremented, but read_dts_node() only uses it to read the rate and never calls clk_put(). The clock is not stored anywhere, so the reference cannot be released later either. Release the clock once its rate has been read, which also covers the error path taken when the rate is zero. Fixes: 414fd46e7762 ("fsl/fman: Add FMan support") Cc: stable@vger.kernel.org Signed-off-by: Wentao Liang Reviewed-by: Simon Horman Link: https://patch.msgid.link/20260917110135.2148068-1-vulab@iscas.ac.cn Signed-off-by: Paolo Abeni --- drivers/net/ethernet/freescale/fman/fman.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/net/ethernet/freescale/fman/fman.c b/drivers/net/ethernet/freescale/fman/fman.c index 299bab043175..46cc28895e56 100644 --- a/drivers/net/ethernet/freescale/fman/fman.c +++ b/drivers/net/ethernet/freescale/fman/fman.c @@ -2734,6 +2734,7 @@ static struct fman *read_dts_node(struct platform_device *of_dev) } clk_rate = clk_get_rate(clk); + clk_put(clk); if (!clk_rate) { err = -EINVAL; dev_err(&of_dev->dev, "%s: Failed to determine FM%d clock rate\n", From 012bfcd5a51082d5a65f096dfb9ca652b5267965 Mon Sep 17 00:00:00 2001 From: Mikhail Zaslonko Date: Wed, 16 Sep 2026 18:06:26 +0200 Subject: [PATCH 368/694] s390/debug: Fix NULL pointer dereference in debug_info_copy() When debug_register_static() fails, it clears areas, active_pages and active_entries but leaves the area bounds unchanged. Copying such an area, either by opening its view file or via debug_dump(), makes debug_info_copy() dereference the NULL pointers. Skip the copy loop when the source has no areas. Closes: https://lore.kernel.org/r/20260903132123.12F271F00A3F@smtp.kernel.org Fixes: d72541f94512 ("s390/debug: add early tracing support") Signed-off-by: Mikhail Zaslonko Reviewed-by: Peter Oberparleiter Acked-by: Heiko Carstens Signed-off-by: Heiko Carstens --- arch/s390/kernel/debug.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/arch/s390/kernel/debug.c b/arch/s390/kernel/debug.c index 142dc1487187..3b68f5c58333 100644 --- a/arch/s390/kernel/debug.c +++ b/arch/s390/kernel/debug.c @@ -434,7 +434,8 @@ static debug_info_t *debug_info_copy(debug_info_t *in, int mode) debug_info_free(rc); } while (1); - if (mode == NO_AREAS) + /* debug_register_static() failure leaves areas NULL, bounds intact */ + if (mode == NO_AREAS || !in->areas) goto out; for (i = 0; i < in->nr_areas; i++) { From 2d14720beb58870b15a52b236c3ab0be0e06e915 Mon Sep 17 00:00:00 2001 From: Muhammad Bilal Date: Sun, 20 Sep 2026 00:19:37 +0500 Subject: [PATCH 369/694] net: spacemit: clear TX descriptor on fragment mapping failure emac_tx_mem_map() writes TX_DESC_0_OWN into the ring descriptor for every slot beyond old_head as soon as that slot's memset()'d local copy is committed with "*tx_desc_addr = tx_desc", i.e. before the buffers for that slot have necessarily all been mapped successfully. If emac_tx_map_frag() then fails on a later fragment, the err_free_skb path calls emac_free_tx_buf() to unmap and drop the skb, but leaves the already-written descriptor memory untouched, and tx_ring->head is never advanced past old_head (the "tx_ring->head = head" store is skipped by the goto). So a slot between old_head and the rolled-back head can be left with TX_DESC_0_OWN set and buffer_addr_{1,2} pointing at DMA mappings that emac_free_tx_buf() just tore down, while software considers that slot free again. The next successful emac_tx_mem_map() call only rebuilds old_head itself; if the DMA engine auto-advances into the following descriptor once it finishes old_head's packet, it will fetch that stale, already-unmapped address. emac_tx_clean_desc() already treats emac_free_tx_buf() and clearing the descriptor as a pair when reclaiming completed descriptors; do the same in the mapping failure path. Fixes: bfec6d7f2001 ("net: spacemit: Add K1 Ethernet MAC") Signed-off-by: Muhammad Bilal Reviewed-by: Vivian Wang Reviewed-by: Troy Mitchell Link: https://patch.msgid.link/20260919191937.271202-1-meatuni001@gmail.com Signed-off-by: Paolo Abeni --- drivers/net/ethernet/spacemit/k1_emac.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/net/ethernet/spacemit/k1_emac.c b/drivers/net/ethernet/spacemit/k1_emac.c index f7f16397a2c2..d641ac26a1e8 100644 --- a/drivers/net/ethernet/spacemit/k1_emac.c +++ b/drivers/net/ethernet/spacemit/k1_emac.c @@ -803,6 +803,9 @@ static void emac_tx_mem_map(struct emac_priv *priv, struct sk_buff *skb) while (i != head) { emac_free_tx_buf(priv, i); + tx_desc_addr = &((struct emac_desc *)tx_ring->desc_addr)[i]; + memset(tx_desc_addr, 0, sizeof(*tx_desc_addr)); + if (++i == tx_ring->total_cnt) i = 0; } From ac4334522e4ba4a3b6710dd5d4cc98092824b8ca Mon Sep 17 00:00:00 2001 From: bui duc phuc Date: Fri, 18 Sep 2026 11:28:04 +0700 Subject: [PATCH 370/694] net: ethernet: ti: netcp: fix pm_runtime usage counter leak on error pm_runtime_get_sync() leaves the runtime PM usage counter incremented even when it fails, but the error path in netcp_probe() does not call pm_runtime_put_noidle() to balance it, leaking a reference each time resume fails. Use pm_runtime_resume_and_get() instead, which automatically drops the usage counter on failure, fixing the leak. Fixes: 84640e27f230 ("net: netcp: Add Keystone NetCP core ethernet driver") Signed-off-by: bui duc phuc Reviewed-by: Simon Horman Link: https://patch.msgid.link/20260918042804.13101-1-phucduc.bui@gmail.com Signed-off-by: Paolo Abeni --- drivers/net/ethernet/ti/netcp_core.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/ethernet/ti/netcp_core.c b/drivers/net/ethernet/ti/netcp_core.c index eb8fc2ed05f4..4f9e20468bbb 100644 --- a/drivers/net/ethernet/ti/netcp_core.c +++ b/drivers/net/ethernet/ti/netcp_core.c @@ -2225,7 +2225,7 @@ static int netcp_probe(struct platform_device *pdev) return -ENOMEM; pm_runtime_enable(&pdev->dev); - ret = pm_runtime_get_sync(&pdev->dev); + ret = pm_runtime_resume_and_get(&pdev->dev); if (ret < 0) { dev_err(dev, "Failed to enable NETCP power-domain\n"); pm_runtime_disable(&pdev->dev); From 999e8295bc41d6ce45b8e54f88150efa96f3f01e Mon Sep 17 00:00:00 2001 From: Wentao Liang Date: Thu, 17 Sep 2026 11:08:28 +0000 Subject: [PATCH 371/694] net: hisilicon: hns_dsaf_mac: fix mdio device leak in hns_mac_register_phy() hns_dsaf_find_platform_device() returns the mdio platform device with its reference count incremented. hns_mac_register_phy() never drops that reference, so the mdio device can not be released. Release the reference on both the deferred probe and the normal path. Fixes: 1d1afa2ebf82 ("net: hns: register phy device in each mac initial sequence") Cc: stable@vger.kernel.org Signed-off-by: Wentao Liang Reviewed-by: Simon Horman Link: https://patch.msgid.link/20260917110828.2148390-1-vulab@iscas.ac.cn Signed-off-by: Paolo Abeni --- drivers/net/ethernet/hisilicon/hns/hns_dsaf_mac.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/net/ethernet/hisilicon/hns/hns_dsaf_mac.c b/drivers/net/ethernet/hisilicon/hns/hns_dsaf_mac.c index bc6b269be299..f1cb6d56e4b9 100644 --- a/drivers/net/ethernet/hisilicon/hns/hns_dsaf_mac.c +++ b/drivers/net/ethernet/hisilicon/hns/hns_dsaf_mac.c @@ -793,6 +793,7 @@ static int hns_mac_register_phy(struct hns_mac_cb *mac_cb) dev_err(mac_cb->dev, "mac%d mdio is NULL, dsaf will probe again later\n", mac_cb->mac_id); + put_device(&pdev->dev); return -EPROBE_DEFER; } @@ -801,6 +802,8 @@ static int hns_mac_register_phy(struct hns_mac_cb *mac_cb) dev_dbg(mac_cb->dev, "mac%d register phy addr:%d\n", mac_cb->mac_id, addr); + put_device(&pdev->dev); + return rc; } From 23d42b9a3bcd55b17d3b371544fedc708c2397e9 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Th=C3=A9o=20Lebrun?= Date: Fri, 18 Sep 2026 21:53:52 +0200 Subject: [PATCH 372/694] net: macb: fix dma_alloc_coherent() leak on macb_alloc() error paths MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Fix 3 leaks in macb_alloc() error paths: - Tx buffer allocated but crossing a 4G boundary: Tx leaked. - Rx buffer allocation fails: Tx leaked. - Rx buffer allocated but crossing a 4G boundary: Tx & Rx leaked. This is because our error handling calls macb_free(bp) which in turn frees the buffers stored in bp->queues[0], but nothing has been stored in there. Fix by storing allocated buffers into bp->queues[0] ASAP. Fixes: 78d901897b3c ("net: macb: single dma_alloc_coherent() for DMA descriptors") Cc: stable@vger.kernel.org Signed-off-by: Théo Lebrun Reviewed-by: Nicolai Buchwitz Link: https://patch.msgid.link/20260918-macb-alloc-leak-v1-1-aba9a3d4f6e3@bootlin.com Signed-off-by: Paolo Abeni --- drivers/net/ethernet/cadence/macb_main.c | 14 ++++++++++++-- 1 file changed, 12 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/cadence/macb_main.c b/drivers/net/ethernet/cadence/macb_main.c index b8234ac4b602..8e5c034dc3a4 100644 --- a/drivers/net/ethernet/cadence/macb_main.c +++ b/drivers/net/ethernet/cadence/macb_main.c @@ -2749,14 +2749,24 @@ static int macb_alloc(struct macb *bp) size = bp->num_queues * macb_tx_ring_size_per_queue(bp); tx = dma_alloc_coherent(dev, size, &tx_dma, GFP_KERNEL); - if (!tx || upper_32_bits(tx_dma) != upper_32_bits(tx_dma + size - 1)) + if (!tx) + goto out_err; + /* Record the buffer so that the error path frees it. */ + bp->queues[0].tx_ring = tx; + bp->queues[0].tx_ring_dma = tx_dma; + if (upper_32_bits(tx_dma) != upper_32_bits(tx_dma + size - 1)) goto out_err; netdev_dbg(bp->netdev, "Allocated %zu bytes for %u TX rings at %08lx (mapped %p)\n", size, bp->num_queues, (unsigned long)tx_dma, tx); size = bp->num_queues * macb_rx_ring_size_per_queue(bp); rx = dma_alloc_coherent(dev, size, &rx_dma, GFP_KERNEL); - if (!rx || upper_32_bits(rx_dma) != upper_32_bits(rx_dma + size - 1)) + if (!rx) + goto out_err; + /* Record the buffer so that the error path frees it. */ + bp->queues[0].rx_ring = rx; + bp->queues[0].rx_ring_dma = rx_dma; + if (upper_32_bits(rx_dma) != upper_32_bits(rx_dma + size - 1)) goto out_err; netdev_dbg(bp->netdev, "Allocated %zu bytes for %u RX rings at %08lx (mapped %p)\n", size, bp->num_queues, (unsigned long)rx_dma, rx); From 261e8a37ecbaf462cdf9c336d2b2f5056088401a Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Fri, 18 Sep 2026 15:29:48 -0700 Subject: [PATCH 373/694] genetlink: report the real command id for dump-only ops in policy dumps The op-to-policy map a CTRL_CMD_GETPOLICY dump returns is the only way for userspace to find out which policy index belongs to which command. ctrl_dumppolicy_put_op() tags the nest with doit->cmd, but an op which only has a dumpit has no doit and every path which fills the split ops in zeroes it out, so those entries all claim to be command 0. nlctrl's own CTRL_CMD_GETPOLICY and NETDEV_CMD_QSTATS_GET are both in that group: [{'family-id': 16, 'op-policy': {'do': 0, 'dump': 0, 'op-id': 3}}, {'family-id': 16, 'op-policy': {'dump': 1, 'op-id': 0}}, ctrl_fill_info() gets this right - it uses the iterator's cmd for CTRL_ATTR_OP_ID - so the two introspection interfaces of the same family contradict each other today. Pass the command in rather than reconstructing it from doit->cmd | dumpit->cmd inside the helper, both callers already have it. Fixes: 26588edbef60 ("genetlink: support split policies in ctrl_dumppolicy_put_op()") Signed-off-by: Jakub Kicinski Link: https://patch.msgid.link/20260918222949.4190284-1-kuba@kernel.org Signed-off-by: Paolo Abeni --- net/netlink/genetlink.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/net/netlink/genetlink.c b/net/netlink/genetlink.c index 41d37442f186..5cc1037d4917 100644 --- a/net/netlink/genetlink.c +++ b/net/netlink/genetlink.c @@ -1656,7 +1656,7 @@ static void *ctrl_dumppolicy_prep(struct sk_buff *skb, } static int ctrl_dumppolicy_put_op(struct sk_buff *skb, - struct netlink_callback *cb, + struct netlink_callback *cb, u32 cmd, struct genl_split_ops *doit, struct genl_split_ops *dumpit) { @@ -1677,7 +1677,7 @@ static int ctrl_dumppolicy_put_op(struct sk_buff *skb, if (!nest_pol) goto err; - nest_op = nla_nest_start(skb, doit->cmd); + nest_op = nla_nest_start(skb, cmd); if (!nest_op) goto err; @@ -1721,7 +1721,8 @@ static int ctrl_dumppolicy(struct sk_buff *skb, struct netlink_callback *cb) &doit, &dumpit))) return -ENOENT; - if (ctrl_dumppolicy_put_op(skb, cb, &doit, &dumpit)) + if (ctrl_dumppolicy_put_op(skb, cb, ctx->op, + &doit, &dumpit)) return skb->len; /* done with the per-op policy index list */ @@ -1730,6 +1731,7 @@ static int ctrl_dumppolicy(struct sk_buff *skb, struct netlink_callback *cb) while (ctx->dump_map) { if (ctrl_dumppolicy_put_op(skb, cb, + ctx->op_iter->cmd, &ctx->op_iter->doit, &ctx->op_iter->dumpit)) return skb->len; From a87529034b9c3c3f3bb71c33e2d6d94658e06383 Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Fri, 18 Sep 2026 15:29:49 -0700 Subject: [PATCH 374/694] selftests: net: nl_nlctrl: check the op ids in the policy map Validate that the op map in the policy dump is correct. Signed-off-by: Jakub Kicinski Link: https://patch.msgid.link/20260918222949.4190284-2-kuba@kernel.org Signed-off-by: Paolo Abeni --- tools/testing/selftests/net/nl_nlctrl.py | 116 +++++++++++++++++++---- 1 file changed, 99 insertions(+), 17 deletions(-) diff --git a/tools/testing/selftests/net/nl_nlctrl.py b/tools/testing/selftests/net/nl_nlctrl.py index fe1f66dc9435..237b3d273260 100755 --- a/tools/testing/selftests/net/nl_nlctrl.py +++ b/tools/testing/selftests/net/nl_nlctrl.py @@ -9,40 +9,86 @@ from lib.py import ksft_run, ksft_exit from lib.py import ksft_eq, ksft_ge, ksft_true, ksft_in, ksft_not_in from lib.py import NetdevFamily, EthtoolFamily, NlctrlFamily +# Families we can expect to always be around, and which between them +# cover ops with a do, with a dump, and with both. +FAMILIES = ('nlctrl', 'netdev') -def getfamily_do(ctrl) -> None: - """Query a single family by name and validate its ops.""" - fam = ctrl.getfamily({'family-name': 'netdev'}) - ksft_eq(fam['family-name'], 'netdev') + +def _get_ops(ctrl, name): + """Get the ops of a family, keyed by command id.""" + fam = ctrl.getfamily({'family-name': name}) + ksft_eq(fam['family-name'], name) ksft_true(fam['family-id'] > 0) # The format of ops is quite odd, [{$idx: {"id"...}}, {$idx: {"id"...}}] # Discard the indices and re-key by command id. ops_by_id = {v['id']: v for op in fam['ops'] for v in op.values()} - ksft_eq(len(ops_by_id), len(fam['ops'])) + ksft_eq(len(ops_by_id), len(fam['ops']), + comment=f"{name} lists a command twice") + return ops_by_id - # All ops should have a policy (either do or dump has one) - for op in ops_by_id.values(): - ksft_in('cmd-cap-haspol', op['flags'], - comment=f"op {op['id']} missing haspol") + +def _get_policy_map(ctrl, req): + """ + The policy map in the Netlink replies looks like this: + + [{'family-id': 16, 'op-policy': {'do': 0, 'dump': 0, 'op-id': 3}}, + {'family-id': 16, 'op-policy': {'dump': 1, 'op-id': 4}}, ...] + + Return the mapping: + + {3:{'do','dump'}, 4:{'dump'}} + + The policy itself is discarded here, only return which command has policy. + """ + pol_map = {} + for msg in ctrl.getpolicy(req, dump=True): + if 'op-policy' not in msg: + continue + modes = dict(msg['op-policy']) + cmd = modes.pop('op-id') + ksft_not_in(cmd, pol_map, comment=f"command {cmd} reported twice") + pol_map[cmd] = set(modes.keys()) + return pol_map + + +def getfamily_do(ctrl) -> None: + """Query single families by name and validate their ops.""" + ops = {name: _get_ops(ctrl, name) for name in FAMILIES} + + for name, ops_by_id in ops.items(): + for op in ops_by_id.values(): + # All ops in nlctrl and netdev have a policy + ksft_in('cmd-cap-haspol', op['flags'], + comment=f"{name} op {op['id']} missing haspol") + ksft_true(op['flags'] & {'cmd-cap-do', 'cmd-cap-dump'}, + comment=f"{name} op {op['id']} has no handler") + + # nlctrl getfamily (id 3) does both, getpolicy (id 10) is dump-only + ksft_in('cmd-cap-do', ops['nlctrl'][3]['flags']) + ksft_in('cmd-cap-dump', ops['nlctrl'][3]['flags']) + ksft_not_in('cmd-cap-do', ops['nlctrl'][10]['flags']) + ksft_in('cmd-cap-dump', ops['nlctrl'][10]['flags']) + + netdev = ops['netdev'] # dev-get (id 1) should support both do and dump - ksft_in('cmd-cap-do', ops_by_id[1]['flags']) - ksft_in('cmd-cap-dump', ops_by_id[1]['flags']) + ksft_in('cmd-cap-do', netdev[1]['flags']) + ksft_in('cmd-cap-dump', netdev[1]['flags']) # qstats-get (id 12) is dump-only - ksft_not_in('cmd-cap-do', ops_by_id[12]['flags']) - ksft_in('cmd-cap-dump', ops_by_id[12]['flags']) + ksft_not_in('cmd-cap-do', netdev[12]['flags']) + ksft_in('cmd-cap-dump', netdev[12]['flags']) # napi-set (id 14) is do-only and requires admin - ksft_in('cmd-cap-do', ops_by_id[14]['flags']) - ksft_not_in('cmd-cap-dump', ops_by_id[14]['flags']) - ksft_in('admin-perm', ops_by_id[14]['flags']) + ksft_in('cmd-cap-do', netdev[14]['flags']) + ksft_not_in('cmd-cap-dump', netdev[14]['flags']) + ksft_in('admin-perm', netdev[14]['flags']) # Notification-only commands (dev-add/del/change-ntf etc.) must # not appear in the ops list since they have no do/dump handlers. for ntf_id in [2, 3, 4, 6, 7, 8]: - ksft_not_in(ntf_id, ops_by_id, + ksft_not_in(ntf_id, netdev, comment=f"ntf-only cmd {ntf_id} should not be in ops") @@ -103,6 +149,41 @@ def getpolicy_dump(_ctrl) -> None: comment="linkinfo-set should not have a dump policy") +def getpolicy_op_map(ctrl) -> None: + """Check the op-to-policy map consistency. Each op with 'haspol' flag + has to have a policy. The policy back-references must name only + real ops that exist, have given modes (do vs dump) and have 'haspol'. + """ + for name in FAMILIES: + ops_by_id = _get_ops(ctrl, name) + haspol = {cmd for cmd, op in ops_by_id.items() + if 'cmd-cap-haspol' in op['flags']} + + pol_map = _get_policy_map(ctrl, {'family-name': name}) + ksft_eq(set(pol_map), haspol, + comment=f"{name} policy map does not match the op list") + + # Walk the op list rather than the map, the map may be missing + # the very op we are after. Asking for a command the family does + # not have is an error, so it must not come from the map either. + for cmd in sorted(haspol): + modes = pol_map.get(cmd, set()) + + # The kernel only reports a mode the op actually has. + if 'do' in modes: + ksft_in('cmd-cap-do', ops_by_id[cmd]['flags'], + comment=f"{name} cmd {cmd} has no do") + if 'dump' in modes: + ksft_in('cmd-cap-dump', ops_by_id[cmd]['flags'], + comment=f"{name} cmd {cmd} has no dump") + + # Asking for one op builds the map in a different place in + # the kernel, it has to report what the full dump did. + single = _get_policy_map(ctrl, {'family-name': name, 'op': cmd}) + ksft_eq(single, {cmd: modes}, + comment=f"{name} cmd {cmd} policy differs from the dump") + + def getpolicy_by_op(_ctrl) -> None: """Query policy for specific ops, check attr names are resolved.""" ndev = NetdevFamily() @@ -122,6 +203,7 @@ def main() -> None: ksft_run([getfamily_do, getfamily_dump, getpolicy_dump, + getpolicy_op_map, getpolicy_by_op], args=(ctrl, )) ksft_exit() From 9892d71cf0ce3ff3d4fed2d9a3968fd4feb1c918 Mon Sep 17 00:00:00 2001 From: Coia Prant Date: Sun, 20 Sep 2026 01:20:21 +0800 Subject: [PATCH 375/694] net: pcs: xpcs: fix clock reference leak on xpcs_init_clks failure xpcs_init_clks() takes references with clk_bulk_get_optional() and then enables them with clk_bulk_prepare_enable(). If the enable step fails, the function returns without dropping the references. xpcs_create() handles the failure through out_free_data, which calls xpcs_free_data() but never xpcs_clear_clks(), so the clk references are leaked. Add the missing clk_bulk_put() on the enable failure path. The prepare/enable side is already rolled back by clk_bulk_prepare_enable() itself. Fixes: f6bb3e9d98c2 ("net: pcs: xpcs: Add Synopsys DW xPCS platform device driver") Signed-off-by: Coia Prant Reviewed-by: Simon Horman Link: https://patch.msgid.link/20260919172021.2336748-1-coiaprant@gmail.com Signed-off-by: Paolo Abeni --- drivers/net/pcs/pcs-xpcs.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/net/pcs/pcs-xpcs.c b/drivers/net/pcs/pcs-xpcs.c index 0337e2bcc012..b415b93d77c1 100644 --- a/drivers/net/pcs/pcs-xpcs.c +++ b/drivers/net/pcs/pcs-xpcs.c @@ -1545,8 +1545,10 @@ static int xpcs_init_clks(struct dw_xpcs *xpcs) return dev_err_probe(dev, ret, "Failed to get clocks\n"); ret = clk_bulk_prepare_enable(DW_XPCS_NUM_CLKS, xpcs->clks); - if (ret) + if (ret) { + clk_bulk_put(DW_XPCS_NUM_CLKS, xpcs->clks); return dev_err_probe(dev, ret, "Failed to enable clocks\n"); + } return 0; } From 17741334d00bf5ebd37f8c1c36bc9c146a351deb Mon Sep 17 00:00:00 2001 From: Wentao Liang Date: Thu, 17 Sep 2026 11:58:11 +0000 Subject: [PATCH 376/694] net: usb: lan78xx: Fix URB reference leak in lan78xx_submit_deferred_urbs() usb_get_from_anchor() hands over a reference to the URB, which the caller must release. lan78xx_submit_deferred_urbs() never does, so every deferred Tx URB keeps an extra reference: the counter grows on each suspend/resume cycle and the URBs are never freed when the buffers are released. Drop the reference after submitting, and on the path that drops the packet instead of submitting it. Fixes: 5f4cc6e25148 ("lan78xx: Fix race conditions in suspend/resume handling") Cc: stable@vger.kernel.org Signed-off-by: Wentao Liang Link: https://patch.msgid.link/20260917115811.2150119-1-vulab@iscas.ac.cn Signed-off-by: Paolo Abeni --- drivers/net/usb/lan78xx.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/net/usb/lan78xx.c b/drivers/net/usb/lan78xx.c index cb782d81d84f..5655941f1478 100644 --- a/drivers/net/usb/lan78xx.c +++ b/drivers/net/usb/lan78xx.c @@ -5239,10 +5239,12 @@ static bool lan78xx_submit_deferred_urbs(struct lan78xx_net *dev) !netif_carrier_ok(dev->net) || pipe_halted) { lan78xx_release_tx_buf(dev, skb); + usb_put_urb(urb); continue; } ret = usb_submit_urb(urb, GFP_ATOMIC); + usb_put_urb(urb); if (ret == 0) { netif_trans_update(dev->net); From 10180a277549339020b08000206092c07e0bff5a Mon Sep 17 00:00:00 2001 From: Sean Christopherson Date: Mon, 21 Sep 2026 14:16:07 -0700 Subject: [PATCH 377/694] KVM: x86: Re-pend GET_NESTED_STATE_PAGES if getting said pages fails Re-pend GET_NESTED_STATE_PAGES before exiting to userspace if getting the nested pages fails in the KVM_RUN path. If userspace re-runs the vCPU, and vmcs02 holds valid PFNs from the *previous* run of L2, then KVM could re-enter L2 with stale, unpinned PFNs mapped into e.g. the vAPIC page. Note, both SVM and VMX (as of commit 11722439fb20 ("KVM: nVMX: Ensure KVM_REQ_GET_NESTED_STATE_PAGES is cleared on VM-Exit") ensure the request is cleared on VM-Exit (including the "forced" case), i.e. there is no risk of double-mapping due to emulated VMLAUNCH/VMRESUME/VMRUN *and* the request trying to map the nested pages. Fixes: 671ddc700fd0 ("KVM: nVMX: Don't leak L1 MMIO regions to L2") Cc: stable@vger.kernel.org Reported-by: Jinwoo Lee Closes: https://lore.kernel.org/all/20260813043932.3214460-1-rkskek9254@gmail.com Reported-by: Stefan Teodorescu Link: https://patch.msgid.link/20260921211608.1030158-2-seanjc@google.com Signed-off-by: Sean Christopherson --- arch/x86/kvm/x86.c | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index a137dc6dd8c6..4ec17aaff413 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -8072,6 +8072,7 @@ static int vcpu_enter_guest(struct kvm_vcpu *vcpu) if (kvm_check_request(KVM_REQ_GET_NESTED_STATE_PAGES, vcpu)) { if (unlikely(!kvm_nested_call(get_nested_state_pages)(vcpu))) { + kvm_make_request(KVM_REQ_GET_NESTED_STATE_PAGES, vcpu); r = 0; goto out; } From c1214f293d77c6e425a379f90d09bdb4815993a8 Mon Sep 17 00:00:00 2001 From: Sean Christopherson Date: Mon, 21 Sep 2026 14:16:08 -0700 Subject: [PATCH 378/694] KVM: x86: Fill kvm_run exit fields in common get_nested_state_pages() error paths Fill kvm_run with "internal error, emulation" in the common error handling paths for getting nested state pages, as requiring each check to manually fill kvm_run is error prone and requires a non-trivial amount of copy+paste. Specifically, both SVM and VMX fail to fill kvm_run if load_pdptrs() fails, and SVM fails to fill kvm_run if kvm_hv_verify_vp_assist() fails. If those flows fail, the *best* case scenario is that KVM will exit to userspace with KVM_EXIT_UNKNOWN. The worst case scenario is that KVM exits with a stale exit_reason and confuses userspace. Note, SVM never exits to userspace if something goes sideways when dealing with vmcb12 assets while emulating VMRUN, i.e. lack of SVM-specific code is not a bug. Fixes: 0f85722341b0 ("KVM: nVMX: delay loading of PDPTRs to KVM_REQ_GET_NESTED_STATE_PAGES") Fixes: 232f75d3b4b5 ("KVM: nSVM: call nested_svm_load_cr3 on nested state load") Fixes: 3f4a812edf5c ("KVM: nSVM: hyper-v: Enable L2 TLB flush") Cc: stable@vger.kernel.org Reported-by: Jinwoo Lee Closes: https://lore.kernel.org/all/20260813043932.3214460-1-rkskek9254@gmail.com Reported-by: Stefan Teodorescu Link: https://patch.msgid.link/20260921211608.1030158-3-seanjc@google.com Signed-off-by: Sean Christopherson --- arch/x86/kvm/svm/nested.c | 7 +------ arch/x86/kvm/vmx/nested.c | 15 +++++---------- arch/x86/kvm/x86.c | 3 +++ 3 files changed, 9 insertions(+), 16 deletions(-) diff --git a/arch/x86/kvm/svm/nested.c b/arch/x86/kvm/svm/nested.c index 73f37b050d0a..f9090b601efa 100644 --- a/arch/x86/kvm/svm/nested.c +++ b/arch/x86/kvm/svm/nested.c @@ -2125,13 +2125,8 @@ static bool svm_get_nested_state_pages(struct kvm_vcpu *vcpu) return false; } - if (!nested_svm_merge_msrpm(vcpu)) { - vcpu->run->exit_reason = KVM_EXIT_INTERNAL_ERROR; - vcpu->run->internal.suberror = - KVM_INTERNAL_ERROR_EMULATION; - vcpu->run->internal.ndata = 0; + if (!nested_svm_merge_msrpm(vcpu)) return false; - } if (kvm_hv_verify_vp_assist(vcpu)) return false; diff --git a/arch/x86/kvm/vmx/nested.c b/arch/x86/kvm/vmx/nested.c index 151873407abd..40c1a5f6fa8a 100644 --- a/arch/x86/kvm/vmx/nested.c +++ b/arch/x86/kvm/vmx/nested.c @@ -3465,10 +3465,6 @@ static bool nested_get_vmcs12_pages(struct kvm_vcpu *vcpu) } else { pr_debug_ratelimited("%s: no backing for APIC-access address in vmcs12\n", __func__); - vcpu->run->exit_reason = KVM_EXIT_INTERNAL_ERROR; - vcpu->run->internal.suberror = - KVM_INTERNAL_ERROR_EMULATION; - vcpu->run->internal.ndata = 0; return false; } } @@ -3539,11 +3535,6 @@ static bool vmx_get_nested_state_pages(struct kvm_vcpu *vcpu) if (!nested_get_evmcs_page(vcpu)) { pr_debug_ratelimited("%s: enlightened vmptrld failed\n", __func__); - vcpu->run->exit_reason = KVM_EXIT_INTERNAL_ERROR; - vcpu->run->internal.suberror = - KVM_INTERNAL_ERROR_EMULATION; - vcpu->run->internal.ndata = 0; - return false; } #endif @@ -3915,8 +3906,12 @@ static int nested_vmx_run(struct kvm_vcpu *vcpu, bool launch) vmentry_failed: vcpu->arch.nested_run_pending = 0; - if (status == NVMX_VMENTRY_KVM_INTERNAL_ERROR) + if (status == NVMX_VMENTRY_KVM_INTERNAL_ERROR) { + vcpu->run->exit_reason = KVM_EXIT_INTERNAL_ERROR; + vcpu->run->internal.suberror = KVM_INTERNAL_ERROR_EMULATION; + vcpu->run->internal.ndata = 0; return 0; + } if (status == NVMX_VMENTRY_VMEXIT) return 1; WARN_ON_ONCE(status != NVMX_VMENTRY_VMFAIL); diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index 4ec17aaff413..aad065d035fb 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -8072,6 +8072,9 @@ static int vcpu_enter_guest(struct kvm_vcpu *vcpu) if (kvm_check_request(KVM_REQ_GET_NESTED_STATE_PAGES, vcpu)) { if (unlikely(!kvm_nested_call(get_nested_state_pages)(vcpu))) { + vcpu->run->exit_reason = KVM_EXIT_INTERNAL_ERROR; + vcpu->run->internal.suberror = KVM_INTERNAL_ERROR_EMULATION; + vcpu->run->internal.ndata = 0; kvm_make_request(KVM_REQ_GET_NESTED_STATE_PAGES, vcpu); r = 0; goto out; From 277d3623d99a4fc2623bfb7d191b649ca380605d Mon Sep 17 00:00:00 2001 From: Zeng Chi Date: Mon, 21 Sep 2026 18:24:42 +0800 Subject: [PATCH 379/694] KVM: Don't treat reserved xarray entries as having memory attributes kvm_vm_set_mem_attributes() reserves an xarray entry for every gfn in the range before storing the new attributes, so that the store loop can't fail partway through. If one of the reservations fails, e.g. with -ENOMEM, the entries that were already reserved are left in the array. That is harmless as far as xa_reserve() is concerned, as the reserved entries read back as NULL via xa_load(), but it confuses the "does this range have no attributes at all" check: if (!attrs) return !xas_find(&xas, end - 1); A reserved entry is XA_ZERO_ENTRY, not NULL, and xas_find() returns it as present. So a leftover reservation makes KVM report that a fully shared range has attributes even though kvm_get_memory_attributes() returns none for every gfn in the range. On x86, the next time mixed-attribute tracking is recomputed for the range (memslot creation, or a later attribute change that straddles the 2MiB page), hugepage_has_attrs() treats a fully shared 2MiB range as mixed and refuses to map it with a hugepage, until userspace happens to set attributes on the range again. Drop the shortcut and handle the !attrs case in the per-index loop, using xas_next_entry() to find the next non-NULL entry. xas_next_entry() is essentially an optimized xas_find(), so the effective change is that the !attrs lookup now goes through xas_retry() like the attrs != 0 case, i.e. reserved entries are skipped and retry entries restart the walk. Don't check the index when no entry is found, as the xarray leaves the xas index in a bogus state in that case; no entry simply means the rest of the range has no attributes. KVM never stores a non-NULL entry with a value of zero (clearing stores NULL), but such an entry would be returned by xas_next_entry() and trip the index check, so WARN if one is ever seen. Fixes: 5a475554db1e ("KVM: Introduce per-page memory attributes") Cc: stable@vger.kernel.org Suggested-by: Sean Christopherson Cc: David Ballesteros Signed-off-by: Zeng Chi Link: https://patch.msgid.link/20260921102442.1232375-1-zeng_chi911@163.com [sean: expand comment to elaborate on xarray APIs, split optimization out] Signed-off-by: Sean Christopherson --- virt/kvm/kvm_main.c | 28 +++++++++++++++++++++++++--- 1 file changed, 25 insertions(+), 3 deletions(-) diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index 65eb26a0520d..2a046e95f95e 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -2447,14 +2447,36 @@ bool kvm_range_has_memory_attributes(struct kvm *kvm, gfn_t start, gfn_t end, return (kvm_get_memory_attributes(kvm, start) & mask) == attrs; guard(rcu)(); - if (!attrs) - return !xas_find(&xas, end - 1); + /* + * Lookup the entry for each index instead of iterating over the xarray + * as KVM deletes/nullifies entries to represent "no attributes", and + * the xas index is effectively invalid when no entry is found. I.e. + * matching non-zero attributes for *every* entry effectively requires + * a manually lookup for each index. + * + * Skip pre-allocated, reserved entries, or restart the lookup if the + * xarray was concurrently modified, via xas_retry() ("retry" means the + * entry holds an internal xarray value, i.e. is either invalid or NULL + * from the caller's perspective). + * + * Use xas_next() when looking for non-zero attributes to optimize for + * the case where the start of the range (or the entire range) doesn't + * have any attributes, as xas_next() returns literally the next entry, + * whereas xas_next_entry() returns the next non-NULL entry (bounded by + * a maximum index). + */ for (index = start; index < end; index++) { do { - entry = xas_next(&xas); + entry = attrs ? xas_next(&xas) : + xas_next_entry(&xas, end - 1); } while (xas_retry(&xas, entry)); + if (!entry) + return !attrs; + + WARN_ON_ONCE(!xa_to_value(entry)); + if (xas.xa_index != index || (xa_to_value(entry) & mask) != attrs) return false; From 382e5d514b6f35bdda2ab9044b4eed23d2ec4254 Mon Sep 17 00:00:00 2001 From: David Ballesteros Date: Tue, 15 Sep 2026 17:53:57 +0000 Subject: [PATCH 380/694] KVM: Ensure memory attributes xarray nodes are accounted to the caller's memcg Explicitly instantiate the memory attributes xarray with XA_FLAGS_ACCOUNT to ensure that all allocations are accounted to the memcg. Frustratingly, memory allocations done in the "fastpath" do not honor the passed in gfp, even for an explicit xa_reserve(). Only the rare, slow path __xas_nomem() honors the original gfp. E.g. xa_reserve(..., GFP_KERNEL_ACCOUNT) | -> ... | -> __xa_cmpxchg_raw() | -> xas_store() <== does not take @gfp | -> xas_create() | -> xas_alloc() The bug was confirmed by observing that a process in a cgroup limited to 256 MiB grew radix_tree_node slab by ~512 MiB while its memory.current stayed near 0. Fixes: 5a475554db1e ("KVM: Introduce per-page memory attributes") Cc: stable@vger.kernel.org Assisted-by: Claude-Code:claude-opus-5 Signed-off-by: David Ballesteros Link: https://patch.msgid.link/20260915175335.138547-4-davimaba.v@proton.me [sean: rewrite changelog, tag for stable] Signed-off-by: Sean Christopherson --- virt/kvm/kvm_main.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index 2a046e95f95e..df643ec3ea75 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -1117,7 +1117,7 @@ static struct kvm *kvm_create_vm(unsigned long type, const char *fdname) rcuwait_init(&kvm->mn_memslots_update_rcuwait); xa_init(&kvm->vcpu_array); #ifdef CONFIG_KVM_GENERIC_MEMORY_ATTRIBUTES - xa_init(&kvm->mem_attr_array); + xa_init_flags(&kvm->mem_attr_array, XA_FLAGS_ACCOUNT); #endif INIT_LIST_HEAD(&kvm->gpc_list); From 119e9db233ad0f4cbd4cfb9f9385a7bca378b37d Mon Sep 17 00:00:00 2001 From: Zeng Chi Date: Mon, 21 Sep 2026 18:24:42 +0800 Subject: [PATCH 381/694] KVM: Don't pre-reserve xarray entries when storing empty/NULL attributes Skip the xarray reservation loop when clearing all memory attributes, as storing NULL only erases the entry and never needs to allocate, so no reservation (and no cleanup of a failed one) is required in that case. Suggested-by: Sean Christopherson Cc: David Ballesteros Signed-off-by: Zeng Chi Link: https://patch.msgid.link/20260921102442.1232375-1-zeng_chi911@163.com [sean: split to separate patch] Signed-off-by: Sean Christopherson --- virt/kvm/kvm_main.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index df643ec3ea75..85f42289748d 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -2593,9 +2593,10 @@ static int kvm_vm_set_mem_attributes(struct kvm *kvm, gfn_t start, gfn_t end, /* * Reserve memory ahead of time to avoid having to deal with failures - * partway through setting the new attributes. + * partway through setting the new attributes. Storing NULL never + * allocates, so no reservations are needed when clearing. */ - for (i = start; i < end; i++) { + for (i = start; entry && i < end; i++) { r = xa_reserve(&kvm->mem_attr_array, i, GFP_KERNEL_ACCOUNT); if (r) goto out_unlock; From 141008dec73521ccf64878517460cec8b3297251 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Szymon=20Aceda=C5=84ski?= Date: Wed, 16 Sep 2026 19:30:30 +0200 Subject: [PATCH 382/694] drm/xe: Limit sg segment size to PAGE_SIZE on Xen PV MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Fix display corruption on Xen PV dom0, where DMA buffers are not guaranteed machine-contiguous, in which case bounce buffering kicks in, breaking xe's memory coherency assumptions. Apply the same workaround i915 carries in i915_sg_segment_size() since commit 78a07fe777c4 ("drm/i915: stop abusing swiotlb_max_segment"). Fixes: dd08ebf6c352 ("drm/xe: Introduce a new DRM driver for Intel GPUs") Reported-by: Marek Marczykowski-Górecki Closes: https://gitlab.freedesktop.org/drm/xe/kernel/-/work_items/8382 Link: https://lore.kernel.org/xen-devel/aYtznP_tT6xNPwf-@mail-itl/ Link: https://lore.kernel.org/all/20221020110308.1582518-1-hch@lst.de/ # i915 counterpart Cc: Christoph Hellwig Cc: Robert Beckett Cc: stable@vger.kernel.org # v6.8+ Signed-off-by: Szymon Acedański Reviewed-by: Thomas Hellström Signed-off-by: Thomas Hellström Link: https://patch.msgid.link/20260916173030.3223833-1-accek@invisiblethingslab.com (cherry picked from commit 77f704158f099b952681f207478a22d5b8218edb) Signed-off-by: Rodrigo Vivi --- drivers/gpu/drm/xe/xe_bo.h | 19 +++++++++++++++++++ 1 file changed, 19 insertions(+) diff --git a/drivers/gpu/drm/xe/xe_bo.h b/drivers/gpu/drm/xe/xe_bo.h index e8081af5bfc1..152bfcffe399 100644 --- a/drivers/gpu/drm/xe/xe_bo.h +++ b/drivers/gpu/drm/xe/xe_bo.h @@ -9,6 +9,8 @@ #include #include +#include + #include "xe_bo_types.h" #include "xe_ggtt.h" #include "xe_macros.h" @@ -575,6 +577,23 @@ static inline unsigned int xe_sg_segment_size(struct device *dev) struct scatterlist __maybe_unused sg; size_t max = BIT_ULL(sizeof(sg.length) * 8) - 1; + /* + * For Xen PV guests pages aren't contiguous in DMA (machine) address + * space. The DMA API takes care of that both in dma_alloc_* (by + * calling into the hypervisor to make the pages contiguous) and in + * dma_map_* (by bounce buffering). But xe (like i915, see commit + * 78a07fe777c4) ignores the coherency aspects of the DMA API and thus + * can't cope with bounce buffering actually happening, so add a hack + * here to force small allocations and mappings when running in PV + * mode on Xen. + * + * Note this will still break if bounce buffering is required for other + * reasons, like confidential computing hypervisors or PCIe root ports + * with addressing limitations. + */ + if (xen_pv_domain()) + return PAGE_SIZE; + max = min_t(size_t, max, dma_max_mapping_size(dev)); /* From 90f467577ebc28c5bc4ba2f0f1b83a4419fb0740 Mon Sep 17 00:00:00 2001 From: Tangudu Tilak Tirumalesh Date: Wed, 16 Sep 2026 15:35:44 +0530 Subject: [PATCH 383/694] drm/xe: harden adjust_idledly() against divide-by-zero and overflow adjust_idledly() has several corner-case issues flagged during review: 1. If xe_gt_clock_init() failed to recognise the crystal clock, gt->info.timestamp_base is 0, which makes idledly_units_ps also 0. The subsequent DIV_ROUND_CLOSEST(..., idledly_units_ps) is then a divide-by-zero and panics the kernel. 2. The tick-to-ns conversions are done in u32: idledly * idledly_units_ps, (maxcnt - 1) * 1000 Both overflow u32 before DIV_ROUND_CLOSEST() sees them. 3. If IDLE_WAIT_TIME reads back as 0, maxcnt evaluates to 0 and the maxcnt - 1 clamp wraps to 0xFFFFFFFF in u32. 4. The register only stores whole ticks, so the clamped ns value has to be converted to ticks and back. DIV_ROUND_CLOSEST() can round that conversion up past maxcnt: maxcnt = 640 ns, one tick = 666664 ps clamp: maxcnt - 1 = 639 ns ns -> ticks: 639000 / 666664 = 0.958 -> rounds to 1 tick tick -> ns: 1 * 666664 / 1000 = 667 ns 667 ns is programmed into RING_IDLEDLY, but 667 >= maxcnt (640), so xe_gt_WARN_ON() fires again on every subsequent init. Return early if timestamp_base is 0 (the unknown-crystal path). Do the conversions in u64 via the *_ULL() helpers so they cannot wrap. Clamp with a floor (DIV_ROUND_DOWN_ULL) so the programmed delay stays strictly below maxcnt, and guard the maxcnt == 0 case with a zero delay while still writing RING_IDLEDLY so INHIBIT_SWITCH_UNTIL_PREEMPTED is cleared. v2: Drop the redundant warn on the timestamp_base == 0 path; xe_gt_clock_init() already warns on an unrecognised crystal clock. Keep the early return to avoid the divide-by-zero. - Vinay v3: Field-mask the RING_IDLEDLY write with REG_FIELD_PREP(IDLE_DELAY, ...) instead of writing the raw tick count, which could clobber INHIBIT_SWITCH_UNTIL_PREEMPTED and reserved bits. Split the inhibit-switch clear from the maxcnt clamp so a set inhibit bit no longer forces a needless delay overwrite when the delay itself is already valid. Use gt_to_xe(gt) instead of gt_to_xe(hwe->gt). Fixes: d2de4410a88f ("drm/xe: Apply Wa_16023105232") Cc: stable@vger.kernel.org Assisted-by: GitHub_Copilot:claude-opus-4.8 Signed-off-by: Tangudu Tilak Tirumalesh Reviewed-by: Vinay Belgaumkar Link: https://patch.msgid.link/20260916100545.779894-2-tilak.tirumalesh.tangudu@intel.com Signed-off-by: Matt Roper (cherry picked from commit d864065ea25e9d12897c175de9176ce46677e176) Signed-off-by: Rodrigo Vivi --- drivers/gpu/drm/xe/xe_hw_engine.c | 27 +++++++++++++++++++++------ 1 file changed, 21 insertions(+), 6 deletions(-) diff --git a/drivers/gpu/drm/xe/xe_hw_engine.c b/drivers/gpu/drm/xe/xe_hw_engine.c index 010499766fce..d868af8e3301 100644 --- a/drivers/gpu/drm/xe/xe_hw_engine.c +++ b/drivers/gpu/drm/xe/xe_hw_engine.c @@ -592,22 +592,37 @@ static void adjust_idledly(struct xe_hw_engine *hwe) u32 idledly_units_ps = 8 * gt->info.timestamp_base; u32 maxcnt_units_ns = 640; bool inhibit_switch = 0; + bool wa_applied = false; + + if (!IS_SRIOV_VF(gt_to_xe(gt)) && XE_GT_WA(gt, 16023105232)) { + /* xe_gt_clock_init() warns and zeroes timestamp_base on unknown crystal clock. */ + if (!idledly_units_ps) + return; - if (!IS_SRIOV_VF(gt_to_xe(hwe->gt)) && XE_GT_WA(gt, 16023105232)) { idledly = xe_mmio_read32(>->mmio, RING_IDLEDLY(hwe->mmio_base)); maxcnt = xe_mmio_read32(>->mmio, RING_PWRCTX_MAXCNT(hwe->mmio_base)); inhibit_switch = idledly & INHIBIT_SWITCH_UNTIL_PREEMPTED; idledly = REG_FIELD_GET(IDLE_DELAY, idledly); - idledly = DIV_ROUND_CLOSEST(idledly * idledly_units_ps, 1000); + idledly = DIV_ROUND_CLOSEST_ULL((u64)idledly * idledly_units_ps, 1000); maxcnt = REG_FIELD_GET(IDLE_WAIT_TIME, maxcnt); maxcnt *= maxcnt_units_ns; - if (xe_gt_WARN_ON(gt, idledly >= maxcnt || inhibit_switch)) { - idledly = DIV_ROUND_CLOSEST(((maxcnt - 1) * 1000), - idledly_units_ps); - xe_mmio_write32(>->mmio, RING_IDLEDLY(hwe->mmio_base), idledly); + /* Clear the inhibit switch without disturbing a valid delay. */ + if (inhibit_switch) + wa_applied = true; + + if (xe_gt_WARN_ON(gt, idledly >= maxcnt)) { + /* Floor below maxcnt; write 0 to still clear the inhibit bit. */ + idledly = maxcnt ? + DIV_ROUND_DOWN_ULL((u64)(maxcnt - 1) * 1000, + idledly_units_ps) : 0; + wa_applied = true; } + + if (wa_applied) + xe_mmio_write32(>->mmio, RING_IDLEDLY(hwe->mmio_base), + REG_FIELD_PREP(IDLE_DELAY, idledly)); } } From bd3a19800dd1ba1ba9a4c307d9dfe50eac443c01 Mon Sep 17 00:00:00 2001 From: Tingmao Wang Date: Sun, 8 Feb 2026 23:54:48 +0000 Subject: [PATCH 384/694] landlock: Add counted_by in landlock_domain MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit For a domain, this array stores the access masks for each layer (of which there are num_layers of them). This annotation serves as useful documentation. Signed-off-by: Tingmao Wang Reviewed-by: Günther Noack Link: https://patch.msgid.link/20260208235449.1124354-1-m@maowtm.org [mic: Rebase on the ruleset/domain split, and update commit message] Signed-off-by: Mickaël Salaün --- security/landlock/domain.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/security/landlock/domain.h b/security/landlock/domain.h index 5ce2f91488d5..caa3d19d2c43 100644 --- a/security/landlock/domain.h +++ b/security/landlock/domain.h @@ -243,7 +243,8 @@ struct landlock_domain { * overlapping access rights. These layers are set once * and never changed for the lifetime of the domain. */ - struct access_masks handled_masks[]; + struct access_masks + handled_masks[] __counted_by(num_layers); }; }; }; From e7e0a54300a896e731dffd3e2e8dae5631d6243d Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Micka=C3=ABl=20Sala=C3=BCn?= Date: Tue, 22 Sep 2026 15:26:14 +0200 Subject: [PATCH 385/694] landlock: Widen ruleset versions to 64 bits MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Tracepoint consumers use a ruleset ID and version to identify the successful landlock_add_rule(2) call prefix used to create a domain. LANDLOCK_MAX_NUM_RULES bounds distinct stored rules, not successful calls: re-adding already-present rights for an object or port succeeds without increasing num_rules. Because every successful call increments the version, these calls can wrap the 32-bit counter and give different prefixes the same trace identity. Widen the counter and its trace fields to 64 bits so the counter cannot wrap in practice, while preserving the successful-call semantics. Saturating would alias all subsequent histories, while rejecting a call at the limit would change otherwise valid syscall behavior solely for trace metadata. Cc: Günther Noack Cc: Steven Rostedt Fixes: 63747c94774d ("landlock: Add landlock_add_rule_fs and landlock_add_rule_net tracepoints") Reviewed-by: Günther Noack Link: https://patch.msgid.link/20260922132615.1025945-1-mic@digikod.net Signed-off-by: Mickaël Salaün --- include/trace/events/landlock.h | 20 ++++++++++---------- security/landlock/ruleset.h | 5 +++-- 2 files changed, 13 insertions(+), 12 deletions(-) diff --git a/include/trace/events/landlock.h b/include/trace/events/landlock.h index 523ba5ea9870..3a43638c9bc2 100644 --- a/include/trace/events/landlock.h +++ b/include/trace/events/landlock.h @@ -344,7 +344,7 @@ TRACE_EVENT(landlock_create_ruleset, TP_STRUCT__entry( __field( u64, ruleset_id ) - __field( u32, ruleset_version ) + __field( u64, ruleset_version ) __field( access_mask_t, handled_fs ) __field( access_mask_t, handled_net ) __field( access_mask_t, scoped ) @@ -358,7 +358,7 @@ TRACE_EVENT(landlock_create_ruleset, __entry->scoped = ruleset->handled_masks.scope; ), - TP_printk("ruleset=%llx.%u handled_fs=%s handled_net=%s scoped=%s", + TP_printk("ruleset=%llx.%llu handled_fs=%s handled_net=%s scoped=%s", __entry->ruleset_id, __entry->ruleset_version, __print_flags(__entry->handled_fs, "|", _LANDLOCK_ACCESS_FS_NAMES), __print_flags(__entry->handled_net, "|", _LANDLOCK_ACCESS_NET_NAMES), @@ -384,7 +384,7 @@ TRACE_EVENT(landlock_free_ruleset, TP_STRUCT__entry( __field( u64, ruleset_id ) - __field( u32, ruleset_version ) + __field( u64, ruleset_version ) ), TP_fast_assign( @@ -392,7 +392,7 @@ TRACE_EVENT(landlock_free_ruleset, __entry->ruleset_version = ruleset->version; ), - TP_printk("ruleset=%llx.%u", + TP_printk("ruleset=%llx.%llu", __entry->ruleset_id, __entry->ruleset_version) ); @@ -423,7 +423,7 @@ TRACE_EVENT(landlock_add_rule_path_beneath, TP_STRUCT__entry( __field( u64, ruleset_id ) - __field( u32, ruleset_version ) + __field( u64, ruleset_version ) __field( access_mask_t, access_rights ) __field( dev_t, dev ) __field( ino_t, ino ) @@ -444,7 +444,7 @@ TRACE_EVENT(landlock_add_rule_path_beneath, __assign_str(pathname); ), - TP_printk("ruleset=%llx.%u access_rights=%s dev=%u:%u ino=%lu path=%s", + TP_printk("ruleset=%llx.%llu access_rights=%s dev=%u:%u ino=%lu path=%s", __entry->ruleset_id, __entry->ruleset_version, __print_flags(__entry->access_rights, "|", _LANDLOCK_ACCESS_FS_NAMES), MAJOR(__entry->dev), MINOR(__entry->dev), __entry->ino, @@ -477,7 +477,7 @@ TRACE_EVENT(landlock_add_rule_net_port, TP_STRUCT__entry( __field( u64, ruleset_id ) - __field( u32, ruleset_version ) + __field( u64, ruleset_version ) __field( access_mask_t, access_rights ) __field( u64, port ) ), @@ -490,7 +490,7 @@ TRACE_EVENT(landlock_add_rule_net_port, __entry->port = port; ), - TP_printk("ruleset=%llx.%u access_rights=%s port=%llu", + TP_printk("ruleset=%llx.%llu access_rights=%s port=%llu", __entry->ruleset_id, __entry->ruleset_version, __print_flags(__entry->access_rights, "|", _LANDLOCK_ACCESS_NET_NAMES), __entry->port) @@ -526,7 +526,7 @@ TRACE_EVENT(landlock_create_domain, __field( u64, domain_id ) __field( u64, parent_id ) __field( u64, ruleset_id ) - __field( u32, ruleset_version ) + __field( u64, ruleset_version ) ), TP_fast_assign( @@ -538,7 +538,7 @@ TRACE_EVENT(landlock_create_domain, __entry->ruleset_version = ruleset->version; ), - TP_printk("domain=%llx parent=%llx ruleset=%llx.%u", + TP_printk("domain=%llx parent=%llx ruleset=%llx.%llu", __entry->domain_id, __entry->parent_id, __entry->ruleset_id, __entry->ruleset_version) ); diff --git a/security/landlock/ruleset.h b/security/landlock/ruleset.h index b536fa0425b7..cf77f1806a95 100644 --- a/security/landlock/ruleset.h +++ b/security/landlock/ruleset.h @@ -171,9 +171,10 @@ struct landlock_ruleset { * @version: Counter incremented on each successful * landlock_add_rule(2), including when it only extends an existing * rule's access rights. Used by tracepoints to correlate a domain with - * the exact ruleset state it was created from. Protected by @lock. + * the exact successful rule history it was created from. Protected by + * @lock. */ - u32 version; + u64 version; /** * @id: Unique identifier for this ruleset, used for tracing. */ From 41112a787f9182c7f2d27122817861e3f22ef928 Mon Sep 17 00:00:00 2001 From: Florian Schmaus Date: Sun, 20 Sep 2026 16:35:25 +0200 Subject: [PATCH 386/694] PM: hibernate: Freeze kernel threads after image preallocation Commit 783c81098445 ("PM: hibernate: call preallocate_image() after freeze prepare") moved hibernate_preallocate_memory() after dpm_prepare() so that device drivers have the opportunity to release pinned/unswappable memory during their ->prepare() callback before memory is preallocated for the snapshot image. However, that commit also placed hibernate_preallocate_memory() after freeze_kernel_threads(). While it was assumed during review that swap I/O submitted via submit_bio() is synchronous and would not depend on frozen kernel threads, this does not hold in practice. Calling hibernate_preallocate_memory() with kernel threads frozen leads to intermittent deadlocks during hibernation. Inside hibernate_preallocate_memory(), shrink_all_memory() is invoked with .may_writepage = 1 and .may_swap = 1 to aggressively reclaim and swap out pages. Any writeback or swap I/O that relies on freezable kernel threads, block device helpers, or WQ_FREEZABLE workqueues (such as those in storage drivers, device mapper, or filesystems) deadlocks waiting on tasks that are stuck in the refrigerator. Fix this by reordering hibernation_snapshot(): 1. Call dpm_prepare(PMSG_FREEZE) first, allowing device drivers to release pinned resources while kernel threads are still active. 2. Call hibernate_preallocate_memory() second, performing page reclaim and swapout while storage layers, workqueues, and kernel threads are alive. 3. Call freeze_kernel_threads() third, only after all memory preallocation and swap I/O have completed. Additionally, restore the call to swsusp_free() in the cleanup path so that preallocated image memory is properly freed if freeze_kernel_threads() fails or if TEST_FREEZER is enabled. Fixes: 783c81098445 ("PM: hibernate: call preallocate_image() after freeze prepare") Signed-off-by: Florian Schmaus Reviewed-by: Mario Limonciello (AMD) Tested-by: Matthew Leach Reviewed-by: Matthew Leach Link: https://patch.msgid.link/20260920-fix-hibernation-v1-1-f9940c2d7d7f@geekplace.eu Signed-off-by: Rafael J. Wysocki --- kernel/power/hibernate.c | 26 ++++++++++++++------------ 1 file changed, 14 insertions(+), 12 deletions(-) diff --git a/kernel/power/hibernate.c b/kernel/power/hibernate.c index d2479c69d71a..c13f68ab7f6e 100644 --- a/kernel/power/hibernate.c +++ b/kernel/power/hibernate.c @@ -408,9 +408,18 @@ int hibernation_snapshot(int platform_mode) if (error) goto Close; + error = dpm_prepare(PMSG_FREEZE); + if (error) + goto Complete; + + /* Preallocate image memory before freezing kernel threads and shutting down devices. */ + error = hibernate_preallocate_memory(); + if (error) + goto Complete; + error = freeze_kernel_threads(); if (error) - goto Close; + goto Cleanup; if (hibernation_test(TEST_FREEZER)) { @@ -422,15 +431,6 @@ int hibernation_snapshot(int platform_mode) goto Thaw; } - error = dpm_prepare(PMSG_FREEZE); - if (error) - goto Complete; - - /* Preallocate image memory before shutting down devices. */ - error = hibernate_preallocate_memory(); - if (error) - goto Complete; - console_suspend_all(); pm_restrict_gfp_mask(); @@ -464,10 +464,12 @@ int hibernation_snapshot(int platform_mode) platform_end(platform_mode); return error; - Complete: - dpm_complete(PMSG_RECOVER); Thaw: thaw_kernel_threads(); + Cleanup: + swsusp_free(); + Complete: + dpm_complete(PMSG_RECOVER); goto Close; } From ec0d89150a9381d591344a9f6f5428655c227a7f Mon Sep 17 00:00:00 2001 From: Manaf Meethalavalappu Pallikunhi Date: Tue, 22 Sep 2026 17:49:02 +0530 Subject: [PATCH 387/694] thermal: gov_step_wise: Fix stale mitigation vote with non-zero lower bounds MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit When two or more thermal zones bind to a common cooling device and one zone uses a non-zero instance->lower value, there is a bug where the instance holds a stale mitigation vote even after its trip is cleared. Problem scenario: - thermal-zone1: Trip at 50°C, cooling-map with lower=0 - thermal-zone2: Trip at 55°C, cooling-map with lower=2 - Both zones share the same cooling device (e.g., CPU) Issue flow: 1. Both trips trigger, zone1 requests state 5, zone2 also mitigates 2. Zone2 trip clears (temp < 53°C due to hysteresis) 3. When throttle=false and trend=THERMAL_TREND_DROPPING: - Current code checks: if (cur_state <= instance->lower) return THERMAL_NO_TARGET - Since cur_state (5) > instance->lower (2), it returns instance->lower (2) - This is the BUG where it returns instance->lower even though trip is cleared 4. Zone2's passive polling stops (tz->passive reaches 0) - no more updates for zone2 5. Zone2's stale vote of 2 persists indefinitely 6. Even when zone1 wants to reduce cooling to state, the cooling device cannot go below state 2 due to zone2's stale vote When a trip is cleared (throttle == false), always return THERMAL_NO_TARGET instead of instance->lower. Remove the unnecessary check comparing cur_state with instance->lower. Since passive polling is already deactivated when the trip is cleared, the instance should always be deactivated regardless of its current cooling state. This ensures that instances with non-zero lower bounds do not retain stale mitigation votes after their trips are cleared. Fixes: 042a3d80f118 ("thermal: core: Move passive polling management to the core") Signed-off-by: Manaf Meethalavalappu Pallikunhi Link: https://patch.msgid.link/20260922-step_wise_multi_zone_stale_vote_fix-v1-1-789f68dab229@oss.qualcomm.com Signed-off-by: Rafael J. Wysocki --- drivers/thermal/gov_step_wise.c | 10 ++++------ 1 file changed, 4 insertions(+), 6 deletions(-) diff --git a/drivers/thermal/gov_step_wise.c b/drivers/thermal/gov_step_wise.c index ea277c466d8d..4fa4377f0d42 100644 --- a/drivers/thermal/gov_step_wise.c +++ b/drivers/thermal/gov_step_wise.c @@ -65,14 +65,12 @@ static unsigned long get_target_state(struct thermal_instance *instance, min(instance->lower + 1, instance->upper), instance->upper); } else if (trend == THERMAL_TREND_DROPPING) { - if (cur_state <= instance->lower) - return THERMAL_NO_TARGET; - /* - * If 'throttle' is false, no mitigation is necessary, so - * request the lower state for this instance. + * If 'throttle' is false, no mitigation is necessary and + * passive polling is already deactivated, so clear this + * instance state by returning THERMAL_NO_TARGET. */ - return instance->lower; + return THERMAL_NO_TARGET; } return instance->target; From ed6eec97b534979dcf28b40c389cee57bd6561d4 Mon Sep 17 00:00:00 2001 From: Emil Tsalapatis Date: Tue, 22 Sep 2026 17:20:18 +0000 Subject: [PATCH 388/694] bpf: Fix bounds check for skb-backed dynptrs The skb_pointer_if_linear() function checks whether a memory region of length len starting at offset off into the skb is in the linear area, and returns a pointer to the region if so. The check currently subtracts between skb_headlen and offset of the check, and since skb_headlen is unsigned the subtraction can underflow. This causes the bounds check to spuriously pass and generate an arbitrary pointer of the form *(skb->data + off). The only user of this helper is currently skb-backed BPF dynptr code. Returning the wrong pointer leads to the dynptr erroneously being backed with invalid memory. Ensure the subtraction cannot underflow, and fail the check if it would. Use u64 arithmetic to also prevent overflow when calculating (skb_headlen(skb) - off) since off is unsigned. Fixes: 6f5a630d7c57 ("bpf, net: Introduce skb_pointer_if_linear().") Reported-by: Nicholas Carlini Signed-off-by: Emil Tsalapatis Signed-off-by: Alexei Starovoitov Reviewed-by: Jiayuan Chen Link: https://patch.msgid.link/20260922172028.6269-2-emil@etsalapatis.com --- include/linux/skbuff.h | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index 421f6fc45451..c8e21903074c 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -4372,7 +4372,10 @@ skb_header_pointer_careful(const struct sk_buff *skb, int offset, static inline void * __must_check skb_pointer_if_linear(const struct sk_buff *skb, int offset, int len) { - if (likely(skb_headlen(skb) - offset >= len)) + unsigned int uoffset = (unsigned int)offset; + + if (likely(uoffset <= skb_headlen(skb) && + (unsigned int)len <= skb_headlen(skb) - uoffset)) return skb->data + offset; return NULL; } From 4fd72eb9f1c939ce33bb714c6f745a125ac5f40c Mon Sep 17 00:00:00 2001 From: Emil Tsalapatis Date: Tue, 22 Sep 2026 17:20:19 +0000 Subject: [PATCH 389/694] selftests/bpf: Test dynptr slices past end of skb Add a selftest to ensure dynptr slices cannot include past the end of the linear area of an skb. Signed-off-by: Emil Tsalapatis Signed-off-by: Alexei Starovoitov Link: https://patch.msgid.link/20260922172028.6269-3-emil@etsalapatis.com --- .../testing/selftests/bpf/prog_tests/dynptr.c | 10 ++++++++++ .../selftests/bpf/progs/dynptr_success.c | 20 +++++++++++++++++++ 2 files changed, 30 insertions(+) diff --git a/tools/testing/selftests/bpf/prog_tests/dynptr.c b/tools/testing/selftests/bpf/prog_tests/dynptr.c index 5fda11590708..4396560365e8 100644 --- a/tools/testing/selftests/bpf/prog_tests/dynptr.c +++ b/tools/testing/selftests/bpf/prog_tests/dynptr.c @@ -9,6 +9,7 @@ enum test_setup_type { SETUP_SYSCALL_SLEEP, SETUP_SKB_PROG, + SETUP_SKB_PROG_NONLINEAR, SETUP_SKB_PROG_TP, SETUP_XDP_PROG, }; @@ -32,6 +33,7 @@ static struct { {"test_ringbuf", SETUP_SYSCALL_SLEEP}, {"test_skb_readonly", SETUP_SKB_PROG}, {"test_dynptr_skb_data", SETUP_SKB_PROG}, + {"test_dynptr_skb_slice_non_linear", SETUP_SKB_PROG_NONLINEAR}, {"test_dynptr_skb_meta_data", SETUP_SKB_PROG}, {"test_dynptr_skb_meta_flags", SETUP_SKB_PROG}, {"test_adjust", SETUP_SYSCALL_SLEEP}, @@ -94,7 +96,9 @@ static void verify_success(const char *prog_name, enum test_setup_type setup_typ bpf_link__destroy(link); break; case SETUP_SKB_PROG: + case SETUP_SKB_PROG_NONLINEAR: { + struct __sk_buff ctx = {}; int prog_fd; char buf[64]; @@ -106,6 +110,12 @@ static void verify_success(const char *prog_name, enum test_setup_type setup_typ .repeat = 1, ); + if (setup_type == SETUP_SKB_PROG_NONLINEAR) { + ctx.data_end = ETH_HLEN + sizeof(struct iphdr); + topts.ctx_in = &ctx; + topts.ctx_size_in = sizeof(ctx); + } + prog_fd = bpf_program__fd(prog); if (!ASSERT_GE(prog_fd, 0, "prog_fd")) goto cleanup; diff --git a/tools/testing/selftests/bpf/progs/dynptr_success.c b/tools/testing/selftests/bpf/progs/dynptr_success.c index e0745b6e467e..b668ebd61fc7 100644 --- a/tools/testing/selftests/bpf/progs/dynptr_success.c +++ b/tools/testing/selftests/bpf/progs/dynptr_success.c @@ -10,6 +10,7 @@ #include "errno.h" #define PAGE_SIZE_64K 65536 +#define TEST_SKB_LINEAR_SIZE (sizeof(struct ethhdr) + sizeof(struct iphdr)) char _license[] SEC("license") = "GPL"; @@ -211,6 +212,25 @@ int test_dynptr_skb_data(struct __sk_buff *skb) return 1; } +SEC("?tc") +int test_dynptr_skb_slice_non_linear(struct __sk_buff *skb) +{ + struct bpf_dynptr ptr; + void *data; + + if (bpf_dynptr_from_skb(skb, 0, &ptr)) { + err = 1; + return 1; + } + + /* Ensure we cannot read past the end of the buffer. */ + data = bpf_dynptr_slice(&ptr, TEST_SKB_LINEAR_SIZE + 1, NULL, 1); + if (data) + err = 2; + + return 1; +} + SEC("?tc") int test_dynptr_skb_meta_data(struct __sk_buff *skb) { From 4a4852376e3a2727ea40e61143d6d7c22bb6dfad Mon Sep 17 00:00:00 2001 From: Emil Tsalapatis Date: Tue, 22 Sep 2026 17:20:20 +0000 Subject: [PATCH 390/694] bpf: Fix bpf_sock context code generation Currently, the ctx access code reads the rx_queue_mapping field with either a 4-byte or 2-byte load. The rest of the bits in the register are marked known zero by the verifier. However, the emitted ctx access code places in the register on certain the special value (-1) using BPF_MOV_IMM64, which gets sign-extended to turn on all the bits in the register. By shifting this value right, the program ends up with a value at runtime above what the verifier assumes is possible. Fix this by ensuring the read value is as wide as the assumed size. Use MOV32 instructions instead of MOV64 instructions to keep the upper bits zero as assumed by the verifier. Also properly report the size of the destination variable (the bpf_sock field, 4 bytes) instead of the source (the socket field, 2 bytes). Fixes: c3c16f2ea6d2 ("bpf: Add rx_queue_mapping to bpf_sock") Reported-by: Nicholas Carlini Suggested-by: Nicholas Carlini Signed-off-by: Emil Tsalapatis Signed-off-by: Alexei Starovoitov Reviewed-by: Jiayuan Chen Link: https://patch.msgid.link/20260922172028.6269-4-emil@etsalapatis.com --- net/core/filter.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/net/core/filter.c b/net/core/filter.c index 532405988fd9..70dc621672f2 100644 --- a/net/core/filter.c +++ b/net/core/filter.c @@ -10566,11 +10566,12 @@ u32 bpf_sock_convert_ctx_access(enum bpf_access_type type, target_size)); *insn++ = BPF_JMP_IMM(BPF_JNE, si->dst_reg, NO_QUEUE_MAPPING, 1); - *insn++ = BPF_MOV64_IMM(si->dst_reg, -1); + *insn++ = BPF_MOV32_IMM(si->dst_reg, -1); #else - *insn++ = BPF_MOV64_IMM(si->dst_reg, -1); - *target_size = 2; + *insn++ = BPF_MOV32_IMM(si->dst_reg, -1); #endif + *target_size = sizeof_field(struct bpf_sock, rx_queue_mapping); + break; } From dec0c209a6808fe6de2e4787538e02786a16a4ef Mon Sep 17 00:00:00 2001 From: Emil Tsalapatis Date: Tue, 22 Sep 2026 17:20:21 +0000 Subject: [PATCH 391/694] selftests/bpf: Add selftests for rx_queue_mapping context access Add tests to ensure the verifier properly tracks the 0 bit state and width of the rx_queue_mapping field read from struct sock. Signed-off-by: Emil Tsalapatis Signed-off-by: Alexei Starovoitov Link: https://patch.msgid.link/20260922172028.6269-5-emil@etsalapatis.com --- .../selftests/bpf/progs/verifier_sock.c | 38 +++++++++++++++++++ 1 file changed, 38 insertions(+) diff --git a/tools/testing/selftests/bpf/progs/verifier_sock.c b/tools/testing/selftests/bpf/progs/verifier_sock.c index 4f2f3209eec8..bf9f6fb6582c 100644 --- a/tools/testing/selftests/bpf/progs/verifier_sock.c +++ b/tools/testing/selftests/bpf/progs/verifier_sock.c @@ -88,6 +88,44 @@ l0_%=: r0 = *(u32*)(r1 + %[bpf_sock_family]); \ : __clobber_all); } +SEC("socket") +__description("skb->sk: sk->rx_queue_mapping [no sign extension]") +__success __success_unpriv __retval(0) +__naked void sk_rx_queue_mapping_no_sign_ext(void) +{ + asm volatile (" \ + r1 = *(u64*)(r1 + %[__sk_buff_sk]); \ + if r1 != 0 goto l0_%=; \ + r0 = 0xdead; \ + exit; \ +l0_%=: r0 = *(u32*)(r1 + %[bpf_sock_rx_queue_mapping]); \ + r0 >>= 32; \ + exit; \ +" : + : __imm_const(__sk_buff_sk, offsetof(struct __sk_buff, sk)), + __imm_const(bpf_sock_rx_queue_mapping, offsetof(struct bpf_sock, rx_queue_mapping)) + : __clobber_all); +} + +SEC("socket") +__description("skb->sk: sk->rx_queue_mapping [narrow load mask]") +__success __success_unpriv __retval(0) +__naked void sk_rx_queue_mapping_narrow_load_mask(void) +{ + asm volatile (" \ + r1 = *(u64*)(r1 + %[__sk_buff_sk]); \ + if r1 != 0 goto l0_%=; \ + r0 = 0xdead; \ + exit; \ +l0_%=: r0 = *(u16*)(r1 + %[bpf_sock_rx_queue_mapping]); \ + r0 >>= 16; \ + exit; \ +" : + : __imm_const(__sk_buff_sk, offsetof(struct __sk_buff, sk)), + __imm_const(bpf_sock_rx_queue_mapping, offsetof(struct bpf_sock, rx_queue_mapping)) + : __clobber_all); +} + SEC("cgroup/skb") __description("skb->sk: sk->type [fullsock field]") __failure __msg("invalid sock_common access") From a6c1edfbe240e4377038a0e1d233ad81fde9a21b Mon Sep 17 00:00:00 2001 From: Emil Tsalapatis Date: Tue, 22 Sep 2026 17:20:22 +0000 Subject: [PATCH 392/694] bpf: Reject pkt arguments in mutating subprogs The verifier tracks changes in how PTR_TO_PACKET registers' bounds are modified across subprog boundaries. PTR_TO_PACKET registers are actually passed as PTR_TO_MEM, which is assumed valid for the entire call. This is not the case with packet memory, where a pskb_* call may invalidate its memory region. Reject BPF code that passes PTR_TO_PACKET pointers to subprogs that may mutate a packet. We cannot pass the pointer as a true PTR_TO_PACKET because we would also need to somehow pass the PTR_TO_PACKET_META or PTR_TO_PACKET_END to the subprog. Since we cannot avoid representing the pointer in the subprog as PTR_TO_MEM, only permit it if the subprog is guaranteed not to mutate the packet. Fixes: 80f281664f5a ("bpf: Support pointers in global func args") Reported-by: Nicholas Carlini Suggested-by: Nicholas Carlini Signed-off-by: Emil Tsalapatis Signed-off-by: Alexei Starovoitov Link: https://patch.msgid.link/20260922172028.6269-6-emil@etsalapatis.com --- kernel/bpf/verifier.c | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index dd8bb179d39e..7ffbb804184f 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -9750,6 +9750,16 @@ static int btf_check_func_arg_match(struct bpf_verifier_env *env, int subprog, if (check_mem_reg(env, reg, argno, arg->mem_size, BPF_READ | BPF_WRITE, NULL, NULL)) return -EINVAL; + /* + * PTR_TO_PACKET get passed as PTR_TO_MEM, preventing + * us from adjusting bounds tracking info. + */ + if ((reg_is_pkt_pointer_any(reg) || reg_is_dynptr_slice_pkt(reg)) && + sub->changes_pkt_data) { + bpf_log(log, "%s is a packet pointer, but func#%d may change packet data\n", + reg_arg_name(env, argno), subprog); + return -EINVAL; + } if (!(arg->arg_type & PTR_MAYBE_NULL) && (type_may_be_null(reg->type) || bpf_register_is_null(reg))) { bpf_log(log, "%s is expected to be non-NULL\n", From 1ed69a54d31848618131aaf3311a78adbe78ced0 Mon Sep 17 00:00:00 2001 From: Emil Tsalapatis Date: Tue, 22 Sep 2026 17:20:23 +0000 Subject: [PATCH 393/694] selftests/bpf: Test rejection of pkt args to mutating subprogs Add a selftests that ensures that PTR_TO_PACKET arguments can only be passed to subprogs that will never adjust the underlying packet memory, and are rejected otherwise. Signed-off-by: Emil Tsalapatis Signed-off-by: Alexei Starovoitov Link: https://patch.msgid.link/20260922172028.6269-7-emil@etsalapatis.com --- .../bpf/progs/verifier_global_ptr_args.c | 53 +++++++++++++++++++ 1 file changed, 53 insertions(+) diff --git a/tools/testing/selftests/bpf/progs/verifier_global_ptr_args.c b/tools/testing/selftests/bpf/progs/verifier_global_ptr_args.c index a3d2af8dc839..dcc2dd46751a 100644 --- a/tools/testing/selftests/bpf/progs/verifier_global_ptr_args.c +++ b/tools/testing/selftests/bpf/progs/verifier_global_ptr_args.c @@ -350,4 +350,57 @@ int anything_to_untrusted_mem(void *ctx) return 0; } +struct pkt_arg { + __u64 x; + __u8 pad[56]; +}; + +__weak int subprog_pkt_ptr_no_change(struct pkt_arg *p) +{ + if (!p) + return 0; + + return p->x; +} + +SEC("?tc") +__success +int pkt_ptr_to_global_mem_arg_no_change(struct __sk_buff *skb) +{ + void *data = (void *)(long)skb->data; + void *data_end = (void *)(long)skb->data_end; + struct pkt_arg *p = data; + + if ((void *)(p + 1) > data_end) + return 0; + + return subprog_pkt_ptr_no_change(p); +} + +__weak int subprog_pkt_ptr_changes_data(struct __sk_buff *skb __arg_ctx, + struct pkt_arg *p) +{ + if (!p) + return 0; + + bpf_skb_pull_data(skb, 0); + return p->x; +} + +SEC("?tc") +__failure __log_level(2) +__msg("R2 is a packet pointer, but func#{{[0-9]+}} may change packet data") +__msg("Caller passes invalid args into func#{{[0-9]+}} ('subprog_pkt_ptr_changes_data')") +int pkt_ptr_to_global_mem_arg_changes_data(struct __sk_buff *skb) +{ + void *data = (void *)(long)skb->data; + void *data_end = (void *)(long)skb->data_end; + struct pkt_arg *p = data; + + if ((void *)(p + 1) > data_end) + return 0; + + return subprog_pkt_ptr_changes_data(skb, p); +} + char _license[] SEC("license") = "GPL"; From f85f5917aa2fbd861c72ea71ecb14a2fa915c3f6 Mon Sep 17 00:00:00 2001 From: Emil Tsalapatis Date: Tue, 22 Sep 2026 17:20:24 +0000 Subject: [PATCH 394/694] bpf: Prevent variable arena/non-arena register contents The verifier marks ALU instructions that include at least one arena operand with needs_zext: These instructions are fixed up after verification to be ALU32 instructions to ensure that the result is a valid offset into an arena. However, different code paths may provide two non-arena 64-bit arguments to the same instruction. The result of the operation in that code path is wrong, since it is now unexpectedly truncated to 32 bits and zero-extended. Add logic to the verifier to ensure every instruction either always has at least one PTR_TO_ARENA argument, or never does. Since needs_zext already tracks the first scenario, add a prevent_zext field in bpf_insn_aux to track the latter. Reject instructions that use arena arguments and have prevent_zext set, or do not have arena arguments and have needs_zext set. Fixes: 6082b6c328b5 ("bpf: Recognize addr_space_cast instruction in the verifier.") Reported-by: Nicholas Carlini Suggested-by: Nicholas Carlini Signed-off-by: Emil Tsalapatis Signed-off-by: Alexei Starovoitov Link: https://patch.msgid.link/20260922172028.6269-8-emil@etsalapatis.com --- include/linux/bpf_verifier.h | 1 + kernel/bpf/verifier.c | 24 +++++++++++++++++++++++- 2 files changed, 24 insertions(+), 1 deletion(-) diff --git a/include/linux/bpf_verifier.h b/include/linux/bpf_verifier.h index 6fe8e5dc57aa..b83ec99f1a13 100644 --- a/include/linux/bpf_verifier.h +++ b/include/linux/bpf_verifier.h @@ -680,6 +680,7 @@ struct bpf_insn_aux_data { bool nospec_result; /* result is unsafe under speculation, nospec must follow */ bool zext_dst; /* this insn zero extends dst reg */ bool needs_zext; /* alu op needs to clear upper bits */ + bool prevent_zext; /* alu op cannot be zext (already used with 64-bit scalars) */ bool non_sleepable; /* helper/kfunc may be called from non-sleepable context */ bool is_iter_next; /* bpf_iter__next() kfunc call */ bool call_with_percpu_alloc_ptr; /* {this,per}_cpu_ptr() with prog percpu alloc */ diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 7ffbb804184f..41b49c56e123 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -15741,6 +15741,7 @@ static int adjust_reg_min_max_vals(struct bpf_verifier_env *env, struct bpf_reg_state *regs = state->regs, *dst_reg, *src_reg; struct bpf_reg_state *ptr_reg = NULL, off_reg = {0}; bool alu32 = (BPF_CLASS(insn->code) != BPF_ALU64); + struct bpf_insn_aux_data *aux = cur_aux(env); u8 opcode = BPF_OP(insn->code); int err; @@ -15752,12 +15753,23 @@ static int adjust_reg_min_max_vals(struct bpf_verifier_env *env, /* Case where at least one operand is an arena. */ if (dst_reg->type == PTR_TO_ARENA || (src_reg && src_reg->type == PTR_TO_ARENA)) { - struct bpf_insn_aux_data *aux = cur_aux(env); if (dst_reg->type != PTR_TO_ARENA) *dst_reg = *src_reg; if (BPF_CLASS(insn->code) == BPF_ALU64) { + /* + * Only arena pointers set needs_zext, but doing so + * modifies the instruction at fixup time to an ALU32 + * and makes it unsuitable for 64-bit scalar args. We + * prevent zext from being set if the instruction has + * been previously called with non-arena registers. + */ + if (aux->prevent_zext) { + verbose(env, "same insn cannot be used with and without arena pointer\n"); + return -EINVAL; + } + /* * 32-bit operations zero upper bits automatically. * 64-bit operations need to be converted to 32. @@ -15770,6 +15782,16 @@ static int adjust_reg_min_max_vals(struct bpf_verifier_env *env, return 0; } + /* Prevent the instruction from being used with arena pointers (see above). */ + if (env->prog->aux->arena && BPF_CLASS(insn->code) == BPF_ALU64) { + if (aux->needs_zext) { + verbose(env, "same insn cannot be used with and without arena pointer\n"); + return -EINVAL; + } + + aux->prevent_zext = true; + } + if (dst_reg->type != SCALAR_VALUE) ptr_reg = dst_reg; From a9e86dd9de4f933b69f5cca6293fd4c8919224ec Mon Sep 17 00:00:00 2001 From: Emil Tsalapatis Date: Tue, 22 Sep 2026 17:20:25 +0000 Subject: [PATCH 395/694] selftests/bpf: Test for mixed arena/nonarena code paths Add a selftest to confirm the verifier rejects ALU operations that return arena or non-arena results depending on code path. Signed-off-by: Emil Tsalapatis Signed-off-by: Alexei Starovoitov Link: https://patch.msgid.link/20260922172028.6269-9-emil@etsalapatis.com --- .../selftests/bpf/progs/verifier_arena.c | 49 +++++++++++++++++++ 1 file changed, 49 insertions(+) diff --git a/tools/testing/selftests/bpf/progs/verifier_arena.c b/tools/testing/selftests/bpf/progs/verifier_arena.c index 815f342eb4b0..3e33766547c0 100644 --- a/tools/testing/selftests/bpf/progs/verifier_arena.c +++ b/tools/testing/selftests/bpf/progs/verifier_arena.c @@ -561,6 +561,55 @@ int arena_ptr_add_arena_ptr(void *ctx) return 0; } +SEC("syscall") +__failure __msg("same insn cannot be used with and without arena pointer") +int mixed_arena_scalar_alu64_scalar_first(void *ctx) +{ + volatile register __u64 reg asm("r3"); + __u32 pick_arena = bpf_get_prandom_u32(); + + reg = 1ULL << 32; + + if (pick_arena) { + asm volatile ( + "r9 = %[arena] ll;" + "%[reg] = 0;" + "%[reg] = addr_space_cast(%[reg], 0x0, 0x1);" + : [reg] "=r"(reg) + : __imm_addr(arena) + : "r9" + ); + } + + reg += 1; + + return 0; +} + +SEC("syscall") +__failure __msg("same insn cannot be used with and without arena pointer") +int mixed_arena_scalar_alu64_arena_first(void *ctx) +{ + volatile register __u64 reg asm("r3"); + __u32 pick_scalar = bpf_get_prandom_u32(); + + asm volatile ( + "r9 = %[arena] ll;" + "%[reg] = 0;" + "%[reg] = addr_space_cast(%[reg], 0x0, 0x1);" + : [reg] "=r"(reg) + : __imm_addr(arena) + : "r9" + ); + + if (pick_scalar) + reg = 1ULL << 32; + + reg += 1; + + return 0; +} + SEC("syscall") __success __retval(0) int scalar_xor_arena_ptr(void *ctx) From cc319238e3f6668f867beb381ce93727c69b7317 Mon Sep 17 00:00:00 2001 From: Tangudu Tilak Tirumalesh Date: Wed, 16 Sep 2026 15:35:45 +0530 Subject: [PATCH 396/694] drm/xe: Add wa_14025941587 to xe2, xe3 and xe3p platforms Avoid programming the IDLEDLY timer to less than 5 microseconds. Apply wa_14025941587 to Graphics Versions 20.01 to 35.11 and Media Versions 13.01 to 35.03 v2: Use xe_rtp_match_not_sriov_vf, move to local variable Remove warn and other knits - Matt R v3: Add verbose comment - Tejas v4: Restore IDLE_DLY register on engine reset. Add it to GUC save-restore list. -Vivek v5: Extend WA to Media Versions 13.01 to 35.03 - Vinay v6: Avoid clearing inhibit switch - Bala Refactor code accordingly by adding idle_reg_val. v7: Rebased with the divide-by-zero/overflow guards living in a separate hardening patch. v8: Preserve the Wa_16023105232 floor (DIV_ROUND_DOWN_ULL) and the maxcnt == 0 guard from the hardening patch. Round up (DIV_ROUND_UP_ULL) the Wa_14025941587 minimum conversion instead, so the tick-quantized delay cannot round back below 5 us. v9: Evaluate the Wa_16023105232 xe_gt_WARN_ON() against the value read from hardware instead of the Wa_14025941587-bumped value, so it no longer fires on the driver's own floor. Re-check the rounded-up tick value against maxcnt and floor it if tick quantization pushed it back to/above maxcnt, logging via xe_gt_dbg since this is the driver's own value, not a hardware anomaly. Assisted-by: GitHub_Copilot:claude-opus-4.8 Signed-off-by: Tangudu Tilak Tirumalesh Reviewed-by: Vinay Belgaumkar Link: https://patch.msgid.link/20260916100545.779894-3-tilak.tirumalesh.tangudu@intel.com Signed-off-by: Matt Roper (cherry picked from commit 9453c528fc909076468ff10df1c2e334ca5a9b00) Signed-off-by: Rodrigo Vivi --- drivers/gpu/drm/xe/xe_guc_ads.c | 2 +- drivers/gpu/drm/xe/xe_hw_engine.c | 95 ++++++++++++++++++++++++------ drivers/gpu/drm/xe/xe_wa_oob.rules | 2 + 3 files changed, 80 insertions(+), 19 deletions(-) diff --git a/drivers/gpu/drm/xe/xe_guc_ads.c b/drivers/gpu/drm/xe/xe_guc_ads.c index ff8eee3831aa..58a3f6293ce4 100644 --- a/drivers/gpu/drm/xe/xe_guc_ads.c +++ b/drivers/gpu/drm/xe/xe_guc_ads.c @@ -864,7 +864,7 @@ static unsigned int guc_mmio_regset_write(struct xe_guc_ads *ads, } } - if (XE_GT_WA(hwe->gt, 16023105232)) + if (XE_GT_WA(hwe->gt, 16023105232) || XE_GT_WA(hwe->gt, 14025941587)) guc_mmio_regset_write_one(ads, regset_map, RING_IDLEDLY(hwe->mmio_base), count++); diff --git a/drivers/gpu/drm/xe/xe_hw_engine.c b/drivers/gpu/drm/xe/xe_hw_engine.c index d868af8e3301..9680d2a5adec 100644 --- a/drivers/gpu/drm/xe/xe_hw_engine.c +++ b/drivers/gpu/drm/xe/xe_hw_engine.c @@ -585,44 +585,103 @@ static void hw_engine_init_early(struct xe_gt *gt, struct xe_hw_engine *hwe, xe_reg_whitelist_process_engine(hwe); } +static u32 idledly_floor_ticks(u32 idledly_ns, u32 idledly_units_ps) +{ + return DIV_ROUND_DOWN_ULL((u64)idledly_ns * 1000, idledly_units_ps); +} + static void adjust_idledly(struct xe_hw_engine *hwe) { struct xe_gt *gt = hwe->gt; - u32 idledly, maxcnt; + u32 idledly, idledly_hw, idledly_reg_val, maxcnt; u32 idledly_units_ps = 8 * gt->info.timestamp_base; u32 maxcnt_units_ns = 640; - bool inhibit_switch = 0; + bool inhibit_switch = false; bool wa_applied = false; + bool clamped_below_maxcnt = false; + + if ((!IS_SRIOV_VF(gt_to_xe(gt)) && XE_GT_WA(gt, 16023105232)) || + XE_GT_WA(gt, 14025941587)) { + u32 mincnt_idledly_ns = 5000; - if (!IS_SRIOV_VF(gt_to_xe(gt)) && XE_GT_WA(gt, 16023105232)) { /* xe_gt_clock_init() warns and zeroes timestamp_base on unknown crystal clock. */ if (!idledly_units_ps) return; - idledly = xe_mmio_read32(>->mmio, RING_IDLEDLY(hwe->mmio_base)); + idledly_reg_val = xe_mmio_read32(>->mmio, RING_IDLEDLY(hwe->mmio_base)); maxcnt = xe_mmio_read32(>->mmio, RING_PWRCTX_MAXCNT(hwe->mmio_base)); - inhibit_switch = idledly & INHIBIT_SWITCH_UNTIL_PREEMPTED; - idledly = REG_FIELD_GET(IDLE_DELAY, idledly); + inhibit_switch = idledly_reg_val & INHIBIT_SWITCH_UNTIL_PREEMPTED; + idledly = REG_FIELD_GET(IDLE_DELAY, idledly_reg_val); idledly = DIV_ROUND_CLOSEST_ULL((u64)idledly * idledly_units_ps, 1000); + idledly_hw = idledly; maxcnt = REG_FIELD_GET(IDLE_WAIT_TIME, maxcnt); maxcnt *= maxcnt_units_ns; - /* Clear the inhibit switch without disturbing a valid delay. */ - if (inhibit_switch) - wa_applied = true; - - if (xe_gt_WARN_ON(gt, idledly >= maxcnt)) { - /* Floor below maxcnt; write 0 to still clear the inhibit bit. */ - idledly = maxcnt ? - DIV_ROUND_DOWN_ULL((u64)(maxcnt - 1) * 1000, - idledly_units_ps) : 0; + /* + * Wa_14025941587 is applied before Wa_16023105232, which takes + * priority if the two ever conflict (not expected in practice). + */ + if (XE_GT_WA(gt, 14025941587) && + idledly < mincnt_idledly_ns) { + idledly = mincnt_idledly_ns; wa_applied = true; } - if (wa_applied) - xe_mmio_write32(>->mmio, RING_IDLEDLY(hwe->mmio_base), - REG_FIELD_PREP(IDLE_DELAY, idledly)); + if (XE_GT_WA(gt, 16023105232)) { + /* Clear the inhibit switch without disturbing a valid delay. */ + if (inhibit_switch) { + idledly_reg_val &= ~INHIBIT_SWITCH_UNTIL_PREEMPTED; + wa_applied = true; + } + + /* Warn only on the value read from hardware. */ + xe_gt_WARN_ON(gt, idledly_hw >= maxcnt); + + if (idledly >= maxcnt) { + /* maxcnt may be 0 if IDLE_WAIT_TIME is unprogrammed. */ + idledly = maxcnt ? maxcnt - 1 : 0; + clamped_below_maxcnt = true; + wa_applied = true; + } + } + + if (wa_applied) { + u32 idledly_ticks; + + /* + * Wa_16023105232 requires idledly < maxcnt, so floor + * that clamp; otherwise round up to guarantee the + * Wa_14025941587 minimum survives tick quantization. + */ + if (clamped_below_maxcnt) + idledly_ticks = idledly_floor_ticks(idledly, idledly_units_ps); + else + idledly_ticks = DIV_ROUND_UP_ULL((u64)idledly * 1000, + idledly_units_ps); + + /* + * Tick quantization can still push the rounded-up value + * to/above maxcnt; re-floor here so Wa_16023105232 keeps + * priority even in that case. + */ + if (!clamped_below_maxcnt && XE_GT_WA(gt, 16023105232) && + (u64)idledly_ticks * idledly_units_ps >= (u64)maxcnt * 1000) { + xe_gt_dbg(gt, "idledly %s: %u ticks would exceed maxcnt=%u, so flooring\n", + hwe->name, idledly_ticks, maxcnt); + idledly = maxcnt ? maxcnt - 1 : 0; + idledly_ticks = idledly_floor_ticks(idledly, idledly_units_ps); + } + + idledly_reg_val &= ~IDLE_DELAY; + idledly_reg_val |= REG_FIELD_PREP(IDLE_DELAY, idledly_ticks); + xe_gt_dbg(gt, "idledly %s: set %u max=%u inh=%u ts=%u\n", + hwe->name, idledly, maxcnt, + !!inhibit_switch, gt->info.timestamp_base); + xe_mmio_write32(>->mmio, + RING_IDLEDLY(hwe->mmio_base), + idledly_reg_val); + } } } diff --git a/drivers/gpu/drm/xe/xe_wa_oob.rules b/drivers/gpu/drm/xe/xe_wa_oob.rules index dd69ad07f7a9..3001155f8d09 100644 --- a/drivers/gpu/drm/xe/xe_wa_oob.rules +++ b/drivers/gpu/drm/xe/xe_wa_oob.rules @@ -72,3 +72,5 @@ 16029897822 MEDIA_VERSION(3500) GRAPHICS_VERSION(3510) 14027054324 GRAPHICS_VERSION(3511) +14025941587 GRAPHICS_VERSION_RANGE(2001, 3511), FUNC(xe_rtp_match_not_sriov_vf) + MEDIA_VERSION_RANGE(1301, 3503), FUNC(xe_rtp_match_not_sriov_vf) From 814a81c842bd88f6bd8a4ce550d560df071a5d03 Mon Sep 17 00:00:00 2001 From: Zhao Gongyi Date: Thu, 17 Sep 2026 20:10:16 +0800 Subject: [PATCH 397/694] bpf, sockmap: Reject max_entries > INT_MAX in sock_map_alloc sock_map_alloc() only rejects max_entries == 0 and otherwise allows any u32 value. sock_map_free() then walks the sks[] array with a signed int iterator: int i; for (i = 0; i < stab->map.max_entries; i++) struct sock **psk = &stab->sks[i]; When a SOCKMAP is created with max_entries = 0xffffffff (UINT_MAX), the allocation of 32 GiB can succeed on large-memory hosts. During free the counter reaches 0x80000000, wraps to INT_MIN, is sign-extended by movslq and turned into a ~16 GiB negative offset from stab->sks, pointing far below the allocation. The faulting access is an xchg() write in sock_map_free(). Without KASAN, the same out-of-bounds write can fault on an unmapped vmalloc page or corrupt an unrelated allocation if that vmalloc address is populated. On a KASAN kernel with CONFIG_KASAN_VMALLOC=y, the shadow check for that address hits an unmapped shadow page and oopses first: BUG: unable to handle page fault for address: fffff521b59c5a00 RIP: 0010:kasan_check_range+0x107/0x190 Call Trace: sock_map_free+0x93/0x190 map_create+0x68d/0xb30 __sys_bpf+0x21e/0x2e70 Vmcore confirmed stab->map.max_entries == 0xffffffff, stab->sks == 0xffffc911ace2d000, and the faulting address sks + (s64)INT_MIN * 8 exactly at 0xffffc90dace2d000. The same buggy path is reached on the normal close()/bpf_map_free_deferred() path whenever such a map is destroyed. sock_map_alloc() used to bound its allocation size through bpf_map_charge_init(), but the bound was dropped when rlimit-based memory accounting was removed. Reject max_entries > INT_MAX at creation time so the signed iterator in sock_map_free() never sees a value that would overflow. Triggered by syzkaller and reproduced on both a 6.6-based KASAN kernel and the upstream v7.3-rc2 kernel. Fixes: 0d2c4f964050 ("bpf: Eliminate rlimit-based memory accounting for sockmap and sockhash maps") Signed-off-by: Zhao Gongyi Signed-off-by: Alexei Starovoitov Link: https://patch.msgid.link/20260917121016.48171-1-zhaogongyi@bytedance.com --- net/core/sock_map.c | 1 + 1 file changed, 1 insertion(+) diff --git a/net/core/sock_map.c b/net/core/sock_map.c index ca49bc7f8687..38df84284328 100644 --- a/net/core/sock_map.c +++ b/net/core/sock_map.c @@ -41,6 +41,7 @@ static struct bpf_map *sock_map_alloc(union bpf_attr *attr) struct bpf_stab *stab; if (attr->max_entries == 0 || + attr->max_entries > INT_MAX || attr->key_size != 4 || (attr->value_size != sizeof(u32) && attr->value_size != sizeof(u64)) || From c6b51091cafff9ce6c03c1416aa13864d17ab97c Mon Sep 17 00:00:00 2001 From: Takashi Sakamoto Date: Tue, 22 Sep 2026 22:26:39 +0900 Subject: [PATCH 398/694] firewire: cdev: fix back-transition for iso_resource_auto client resource The todo member of iso_resource_auto structure represents the state of the client resource and normally transitions in the following order: ISO_RES_AUTO_ALLOC -> ISO_RES_AUTO_REALLOC -> ISO_RES_AUTO_DEALLOC However, concurrent access from the work item and the file descriptor release function can cause the state to transition backwards from ISO_RES_AUTO_DEALLOC to ISO_RES_AUTO_REALLOC. Prevent the back-transition by checking the current state before updating it in the work item. Fixes: fcabbf40fae5 ("firewire: core: move allocation/reallocation paths into specific branch after isoc resource management in cdev") Link: https://lore.kernel.org/r/20260922132639.191593-1-o-takashi@sakamocchi.jp Signed-off-by: Takashi Sakamoto --- drivers/firewire/core-cdev.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/firewire/core-cdev.c b/drivers/firewire/core-cdev.c index e49d8a58be09..664952a67a11 100644 --- a/drivers/firewire/core-cdev.c +++ b/drivers/firewire/core-cdev.c @@ -1397,8 +1397,10 @@ static void iso_resource_auto_work(struct work_struct *work) } else { // Transit from allocation to reallocation, except if the client requested // deallocation in the meantime. - scoped_guard(spinlock_irq, &client->lock) - r->todo = ISO_RES_AUTO_REALLOC; + scoped_guard(spinlock_irq, &client->lock) { + if (r->todo == ISO_RES_AUTO_ALLOC) + r->todo = ISO_RES_AUTO_REALLOC; + } if (channel >= 0) r->params.channels_mask = BIT_ULL(channel); From 35e6f970f553954d92ba20afa885139a8e7dd0d7 Mon Sep 17 00:00:00 2001 From: Bernardo Soares Date: Fri, 18 Sep 2026 10:59:30 +0100 Subject: [PATCH 399/694] net/mlx5: Bridge, don't fail switchdev events of sibling eswitch ports mlx5 registers the bridge offload switchdev notifiers once per eswitch instance, but the notifier chains are global, so every instance sees every event and must filter out the ones that aren't its own. The existing filter, mlx5_esw_bridge_dev_same_hw(), only checks that the event netdevice sits on the same HCA - intentional for merged eswitch, where one bridge can span representors of several eswitches on one HCA - but same-HCA doesn't mean the instance actually has that port: peer ports are only created reactively from NETDEV_CHANGEUPPER, so an instance brought up after a sibling PF's port was already enslaved has none. The port object and attribute handlers claim the event anyway once same-HW passes, then fail the port lookup and return -EINVAL, which gets reported to user space even though the owning instance already handled it (e.g. "bridge vlan add ... RTNETLINK answers: Invalid argument"). Fix by filtering on the tracked port instead. The same gap exists in the generic recursive lower-device walk used by attribute changes on a bridge with more than one representor enslaved directly: mlx5_esw_bridge_lower_rep_vport_num_vhca_id_get() is entered with the bridge master netdevice, falls through to its generic netdev_for_each_lower_dev() loop, and returns as soon as the recursion into any one lower device yields a non-NULL rep - the underlying base case, mlx5_esw_bridge_rep_vport_num_vhca_id_get(), only checks mlx5_esw_bridge_dev_same_hw(), not ownership by the calling instance's br_offloads. mlx5_esw_bridge_lag_rep_get(), used for the LAG-master case, already filters on mlx5_esw_bridge_dev_same_esw() per candidate and so cannot select a sibling's rep; it is not the source of this bug. On a merged-eswitch HCA with a bridge spanning representors of more than one eswitch instance directly, the walk can return a sibling's rep instead of continuing to the one the calling instance actually owns, so the attribute change fails the same way as above. Fix by checking mlx5_esw_bridge_port_exists() at the point each rep is picked, same as the previous fix did for the notifier filter. Fixes: c358ea1741bc ("net/mlx5: Bridge, allow merged eswitch connectivity") Signed-off-by: Bernardo Soares Cc: Vlad Buslov Cc: Saeed Mahameed Reviewed-by: Mark Bloch Link: https://patch.msgid.link/20260918095931.29792-2-bsoares.it@gmail.com Signed-off-by: Jakub Kicinski --- .../mellanox/mlx5/core/en/rep/bridge.c | 45 +++++++++++++++---- .../ethernet/mellanox/mlx5/core/esw/bridge.c | 6 +++ .../ethernet/mellanox/mlx5/core/esw/bridge.h | 2 + 3 files changed, 44 insertions(+), 9 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/rep/bridge.c b/drivers/net/ethernet/mellanox/mlx5/core/en/rep/bridge.c index baac38bece14..4b7b0a0fc2b2 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/rep/bridge.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/rep/bridge.c @@ -85,9 +85,16 @@ mlx5_esw_bridge_lower_rep_vport_num_vhca_id_get(struct net_device *dev, struct m struct net_device *lower_dev; struct list_head *iter; - if (netif_is_lag_master(dev) || mlx5e_eswitch_rep(dev)) - return mlx5_esw_bridge_rep_vport_num_vhca_id_get(dev, esw, vport_num, - esw_owner_vhca_id); + if (netif_is_lag_master(dev) || mlx5e_eswitch_rep(dev)) { + struct net_device *rep; + + rep = mlx5_esw_bridge_rep_vport_num_vhca_id_get(dev, esw, vport_num, + esw_owner_vhca_id); + if (rep && !mlx5_esw_bridge_port_exists(*vport_num, *esw_owner_vhca_id, + esw->br_offloads)) + return NULL; + return rep; + } netdev_for_each_lower_dev(dev, lower_dev, iter) { struct net_device *rep; @@ -104,6 +111,28 @@ mlx5_esw_bridge_lower_rep_vport_num_vhca_id_get(struct net_device *dev, struct m return NULL; } +static bool mlx5_esw_bridge_rep_port_lookup(struct net_device *dev, + struct mlx5_esw_bridge_offloads *br_offloads, + u16 *vport_num, u16 *esw_owner_vhca_id) +{ + if (!mlx5_esw_bridge_rep_vport_num_vhca_id_get(dev, br_offloads->esw, vport_num, + esw_owner_vhca_id)) + return false; + + return mlx5_esw_bridge_port_exists(*vport_num, *esw_owner_vhca_id, br_offloads); +} + +static bool mlx5_esw_bridge_lower_rep_port_lookup(struct net_device *dev, + struct mlx5_esw_bridge_offloads *br_offloads, + u16 *vport_num, u16 *esw_owner_vhca_id) +{ + if (!mlx5_esw_bridge_lower_rep_vport_num_vhca_id_get(dev, br_offloads->esw, vport_num, + esw_owner_vhca_id)) + return false; + + return mlx5_esw_bridge_port_exists(*vport_num, *esw_owner_vhca_id, br_offloads); +} + static bool mlx5_esw_bridge_is_local(struct net_device *dev, struct net_device *rep, struct mlx5_eswitch *esw) { @@ -218,8 +247,7 @@ mlx5_esw_bridge_port_obj_add(struct net_device *dev, u16 vport_num, esw_owner_vhca_id; int err; - if (!mlx5_esw_bridge_rep_vport_num_vhca_id_get(dev, br_offloads->esw, &vport_num, - &esw_owner_vhca_id)) + if (!mlx5_esw_bridge_rep_port_lookup(dev, br_offloads, &vport_num, &esw_owner_vhca_id)) return 0; port_obj_info->handled = true; @@ -251,8 +279,7 @@ mlx5_esw_bridge_port_obj_del(struct net_device *dev, const struct switchdev_obj_port_mdb *mdb; u16 vport_num, esw_owner_vhca_id; - if (!mlx5_esw_bridge_rep_vport_num_vhca_id_get(dev, br_offloads->esw, &vport_num, - &esw_owner_vhca_id)) + if (!mlx5_esw_bridge_rep_port_lookup(dev, br_offloads, &vport_num, &esw_owner_vhca_id)) return 0; port_obj_info->handled = true; @@ -283,8 +310,8 @@ mlx5_esw_bridge_port_obj_attr_set(struct net_device *dev, u16 vport_num, esw_owner_vhca_id; int err = 0; - if (!mlx5_esw_bridge_lower_rep_vport_num_vhca_id_get(dev, br_offloads->esw, &vport_num, - &esw_owner_vhca_id)) + if (!mlx5_esw_bridge_lower_rep_port_lookup(dev, br_offloads, &vport_num, + &esw_owner_vhca_id)) return 0; port_attr_info->handled = true; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/esw/bridge.c b/drivers/net/ethernet/mellanox/mlx5/core/esw/bridge.c index 87b5fd349594..ac90ccda1272 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/esw/bridge.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/esw/bridge.c @@ -1686,6 +1686,12 @@ int mlx5_esw_bridge_vport_peer_unlink(struct net_device *br_netdev, u16 vport_nu extack); } +bool mlx5_esw_bridge_port_exists(u16 vport_num, u16 esw_owner_vhca_id, + struct mlx5_esw_bridge_offloads *br_offloads) +{ + return mlx5_esw_bridge_port_lookup(vport_num, esw_owner_vhca_id, br_offloads); +} + int mlx5_esw_bridge_port_vlan_add(u16 vport_num, u16 esw_owner_vhca_id, u16 vid, u16 flags, struct mlx5_esw_bridge_offloads *br_offloads, struct netlink_ext_ack *extack) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/esw/bridge.h b/drivers/net/ethernet/mellanox/mlx5/core/esw/bridge.h index d6f539161993..a4e59cc21089 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/esw/bridge.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/esw/bridge.h @@ -80,6 +80,8 @@ int mlx5_esw_bridge_vlan_proto_set(u16 vport_num, u16 esw_owner_vhca_id, u16 pro struct mlx5_esw_bridge_offloads *br_offloads); int mlx5_esw_bridge_mcast_set(u16 vport_num, u16 esw_owner_vhca_id, bool enable, struct mlx5_esw_bridge_offloads *br_offloads); +bool mlx5_esw_bridge_port_exists(u16 vport_num, u16 esw_owner_vhca_id, + struct mlx5_esw_bridge_offloads *br_offloads); int mlx5_esw_bridge_port_vlan_add(u16 vport_num, u16 esw_owner_vhca_id, u16 vid, u16 flags, struct mlx5_esw_bridge_offloads *br_offloads, struct netlink_ext_ack *extack); From 2e51097c982b7b22382fda3202ba29f0ea33e8c0 Mon Sep 17 00:00:00 2001 From: Bernardo Soares Date: Fri, 18 Sep 2026 10:59:31 +0100 Subject: [PATCH 400/694] net/mlx5: Bridge, don't fail unlink of untracked/unsupported peer ports mlx5_esw_bridge_vport_unlink() returns -EINVAL when the port isn't tracked by this instance's br_offloads. This is reachable on a sibling instance that registered its notifier after the port was already enslaved: it never saw the NETDEV_CHANGEUPPER link event, so peer_link() never created a peer port for it, but it does see the later unlink event and fails. Return 0 instead, and give mlx5_esw_bridge_vport_peer_unlink() the same merged_eswitch capability guard peer_link() already has, since without it peer_link() likewise never creates a port to unlink. This also matters beyond the -EINVAL itself: mlx5_esw_bridge_switchdev_port_event() runs on the per-netns netdev_chain, and notifier_from_errno(-EINVAL) sets NOTIFY_STOP_MASK, which call_netdevice_notifiers_info() checks to stop calling further listeners on that chain - so the old -EINVAL silently dropped the event for any listener registered later on the same chain, even though none of it was visible to user space since __netdev_upper_dev_unlink() discards the return value. Fixes: c358ea1741bc ("net/mlx5: Bridge, allow merged eswitch connectivity") Signed-off-by: Bernardo Soares Reviewed-by: Mark Bloch Link: https://patch.msgid.link/20260918095931.29792-3-bsoares.it@gmail.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/mellanox/mlx5/core/esw/bridge.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/esw/bridge.c b/drivers/net/ethernet/mellanox/mlx5/core/esw/bridge.c index ac90ccda1272..b4cf3c5ac0dd 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/esw/bridge.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/esw/bridge.c @@ -1649,10 +1649,8 @@ int mlx5_esw_bridge_vport_unlink(struct net_device *br_netdev, u16 vport_num, int err; port = mlx5_esw_bridge_port_lookup(vport_num, esw_owner_vhca_id, br_offloads); - if (!port) { - NL_SET_ERR_MSG_MOD(extack, "Port is not attached to any bridge"); - return -EINVAL; - } + if (!port) + return 0; if (port->bridge->ifindex != br_netdev->ifindex) { NL_SET_ERR_MSG_MOD(extack, "Port is attached to another bridge"); return -EINVAL; @@ -1682,6 +1680,9 @@ int mlx5_esw_bridge_vport_peer_unlink(struct net_device *br_netdev, u16 vport_nu struct mlx5_esw_bridge_offloads *br_offloads, struct netlink_ext_ack *extack) { + if (!MLX5_CAP_ESW(br_offloads->esw->dev, merged_eswitch)) + return 0; + return mlx5_esw_bridge_vport_unlink(br_netdev, vport_num, esw_owner_vhca_id, br_offloads, extack); } From 0bf6bb567f0edaa771e7dd208ef98da50e6a4485 Mon Sep 17 00:00:00 2001 From: Wentao Liang Date: Thu, 17 Sep 2026 11:31:31 +0000 Subject: [PATCH 401/694] net/mlx5: Fix rev_entry reference leak in mlx5_tc_ct_shared_counter_get() When the reverse entry is found but its counter is already being released, refcount_inc_not_zero() fails and the reference taken by mlx5_tc_ct_entry_get() is never dropped before falling through to create_counter. Drop it so the reverse entry is not kept alive forever by a shared counter lookup that did not use it. Fixes: 1edae2335adf ("net/mlx5e: CT: Use the same counter for both directions") Cc: stable@vger.kernel.org Signed-off-by: Wentao Liang Reviewed-by: Tariq Toukan Link: https://patch.msgid.link/20260917113131.2149024-1-vulab@iscas.ac.cn Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c index 6c87a1c7db09..c1841ce74d9a 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c @@ -1082,6 +1082,9 @@ mlx5_tc_ct_shared_counter_get(struct mlx5_tc_ct_priv *ct_priv, spin_unlock_bh(&ct_priv->ht_lock); + if (rev_entry) + mlx5_tc_ct_entry_put(rev_entry); + create_counter: shared_counter = mlx5_tc_ct_counter_create(ct_priv); From a1259e92e1e892f009bbebf23e1207b02e2d5708 Mon Sep 17 00:00:00 2001 From: ZhangGuoDong Date: Mon, 21 Sep 2026 09:41:28 +0800 Subject: [PATCH 402/694] smb: client: update POSIX extension specification references The POSIX extension specifications have now been published: https://smb3posix.org/ Suggested-by: Paulo Alcantara Reviewed-by: ChenXiaoSong Signed-off-by: ZhangGuoDong Reviewed-by: Namjae Jeon Signed-off-by: Paulo Alcantara --- fs/smb/client/smb2pdu.h | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) diff --git a/fs/smb/client/smb2pdu.h b/fs/smb/client/smb2pdu.h index ab6c667bebc0..b37a1e3941a1 100644 --- a/fs/smb/client/smb2pdu.h +++ b/fs/smb/client/smb2pdu.h @@ -224,8 +224,7 @@ struct smb2_file_id_extd_directory_info { extern char smb2_padding[7]; /* - * See POSIX-SMB2 2.2.14.2.16 - * Link: https://gitlab.com/samba-team/smb3-posix-spec/-/blob/master/smb3_posix_extensions.md + * See POSIX-SMB2 2.1.3.2.1 */ struct create_posix_rsp { u32 nlink; @@ -238,6 +237,7 @@ struct create_posix_rsp { #define SMB2_QUERY_DIRECTORY_IOV_SIZE 2 /* + * See POSIX-FSCC 2.2.1 * SMB2-only POSIX info level for query dir * * See posix_info_sid_size(), posix_info_extra_size() and @@ -256,13 +256,17 @@ struct smb2_posix_info { __le64 Inode; __le32 DeviceId; __le32 Zero; - /* beginning of POSIX Create Context Response */ + /* + * Beginning of POSIX Create Context Response + * See POSIX-SMB2 2.1.3.2.1 + */ __le32 HardLinks; __le32 ReparseTag; __le32 Mode; /* * var sized owner SID * var sized group SID + * End of POSIX Create Context Response * le32 filenamelength * u8 filename[] */ From 031fe051abb3c1f36c2ba6346931fcdb3cddaeae Mon Sep 17 00:00:00 2001 From: Fredric Cover Date: Tue, 22 Sep 2026 10:37:59 -0700 Subject: [PATCH 403/694] smb: client: use GFP_KERNEL in get_targets() Currently, get_targets() uses GFP_ATOMIC to allocate the cache target iterator and to duplicate t->name. The callers of get_targets() are in a sleepable context; therefore, switch to GFP_KERNEL to reduce risk of failure and reduce pressure on emergency pools in low-memory scenarios. Signed-off-by: Fredric Cover Signed-off-by: Paulo Alcantara --- fs/smb/client/dfs_cache.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/fs/smb/client/dfs_cache.c b/fs/smb/client/dfs_cache.c index 29dfd7595941..26956e33037e 100644 --- a/fs/smb/client/dfs_cache.c +++ b/fs/smb/client/dfs_cache.c @@ -798,13 +798,13 @@ static int get_targets(struct cache_entry *ce, struct dfs_cache_tgt_list *tl) INIT_LIST_HEAD(head); list_for_each_entry(t, &ce->tlist, list) { - it = kzalloc_obj(*it, GFP_ATOMIC); + it = kzalloc_obj(*it, GFP_KERNEL); if (!it) { rc = -ENOMEM; goto err_free_it; } - it->it_name = kstrdup(t->name, GFP_ATOMIC); + it->it_name = kstrdup(t->name, GFP_KERNEL); if (!it->it_name) { kfree(it); rc = -ENOMEM; From 4498467a8af06cfa3d71cb04bd7c4170dec8f449 Mon Sep 17 00:00:00 2001 From: Aohan Mei Date: Mon, 21 Sep 2026 17:37:04 +0800 Subject: [PATCH 404/694] sctp: discard the rest of the packet on a stale-cookie error When an association is in COOKIE-ECHOED state and the peer sends a bundled [ERROR(Stale Cookie)][DATA] packet from one of its non-primary addresses, processing the ERROR chunk takes the non-fatal stale-cookie retry path sctp_sf_do_5_2_6_stale(), which queues SCTP_CMD_DEL_NON_PRIMARY while keeping the association alive. sctp_cmd_del_non_primary() removes every non-primary transport - including the very transport this packet arrived on, which is still referenced by the receive lookup and shared by all chunks of the packet via chunk->transport. sctp_assoc_rm_peer() does redirect asoc->peer.last_data_from away from the removed transport, but right afterwards the bundled DATA chunk makes sctp_assoc_bh_rcv() re-register asoc->peer.last_data_from = chunk->transport unconditionally, undoing the redirection with the just-removed transport. Once the packet is done, the receive reference is dropped and the transport is RCU-freed, while the surviving association keeps the dangling last_data_from. A later FWD-TSN (or the delayed SACK timer) makes sctp_gen_sack() dereference it (->param_flags and friends), and sctp_make_sack()/sctp_outq_select_transport() may write to the freed object and link it into the live transport list. This is a use-after-free triggerable by any malicious SCTP peer (or a local unprivileged user acting as one) with no capabilities required: BUG: KASAN: slab-use-after-free in sctp_do_sm+0x498a/0x5660 Read of size 4 at addr ffff88800e1e356c by task poc/115 Call Trace: sctp_do_sm <- sctp_assoc_bh_rcv <- sctp_inq_push <- sctp_rcv <- ip_protocol_deliver_rcu <- ip_rcv Allocated: sctp_transport_new <- sctp_assoc_add_peer <- sctp_process_init (INIT-ACK processing) Freed: kfree <- sctp_transport_destroy_rcu <- rcu_core (call_rcu queued by sctp_transport_put at end of sctp_rcv) The buggy address is located 364 bytes inside of freed 1024-byte region [ffff88800e1e3400, ffff88800e1e3800), cache kmalloc-1k Note that commit 03a9d10ecf71 ("sctp: drop a chunk if its transport was removed") only covers the window between the receive lookup and the chunk processing (e.g. an ASCONF DEL-IP racing the socket backlog); here the transport is removed *while* the packet is being processed, by an earlier chunk of the same packet, so the drop in sctp_inq_push() does not reach this path. Verified with the bundled [ERROR(Stale Cookie)][DATA] + FWD-TSN reproducer: the KASAN report above still fires with that commit applied, and is gone with this patch on top. Fix it by discarding the rest of the packet on this path, as suggested by Xin. After the stale-cookie ERROR has sent the association back to COOKIE-WAIT and removed the non-primary transports, the remaining chunks of the packet can only run against the restarted handshake while referencing the removed arrival transport through chunk->transport: besides the last_data_from registration above, sctp_cmd_setup_t2() and the sctp_make_*() reply builders would also copy that pointer into association-lifetime state that sctp_assoc_rm_peer() has already sanitized. Let the peer retransmit them, in line with what sctp_inq_push() does for chunks whose transport was removed before processing. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Suggested-by: Xin Long Reported-by: TencentOS Corvus AI Cc: stable@vger.kernel.org Signed-off-by: Aohan Mei Acked-by: Xin Long Link: https://patch.msgid.link/20260921093707.1432184-1-ljp1205831794@gmail.com Signed-off-by: Jakub Kicinski --- net/sctp/sm_statefuns.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/net/sctp/sm_statefuns.c b/net/sctp/sm_statefuns.c index c701cff6ea67..8dc65498763c 100644 --- a/net/sctp/sm_statefuns.c +++ b/net/sctp/sm_statefuns.c @@ -2654,6 +2654,8 @@ static enum sctp_disposition sctp_sf_do_5_2_6_stale( sctp_add_cmd_sf(commands, SCTP_CMD_REPLY, SCTP_CHUNK(reply)); + sctp_add_cmd_sf(commands, SCTP_CMD_DISCARD_PACKET, SCTP_NULL()); + return SCTP_DISPOSITION_CONSUME; nomem: From 07e1a9408b6c2f9d0cfb757b67dabb52da7a32b2 Mon Sep 17 00:00:00 2001 From: Willem de Bruijn Date: Fri, 18 Sep 2026 20:47:28 -0400 Subject: [PATCH 405/694] virtio_net: copy zerocopy frags in start_xmit without NAPI Virtio-net without NAPI frees completed skbs lazily on the next start_xmit. Senders waiting for in-flight zerocopy buffers can deadlock if they cannot transmit more packets, as then no completed packets will be freed. When !use_napi, virtio-net already calls skb_orphan to avoid waiting up for transmitted skbs to be freed. For zerocopy packets that require deep copying on orphan (i.e. those that do not set SKBFL_DONT_ORPHAN, such as PACKET_TX_RING), call skb_orphan_frags before orphaning to release the buffers. This fixes the tpacket_snd slot reuse bug on skb_orphan for virtio-net, and prevents PACKET_TX_RING from running out of slots. This fix also touches vhost_net zerocopy packets, which also do not set SKBFL_DONT_ORPHAN. This is fine: vhost_net packets only encounter virtio-net in nested virtualization, and only if napi_tx is explicitly disabled (it has been default-enabled since Linux 4.12). In that rare case, copying the frags is desirable anyway to prevent holding guest descriptors pinned across unbounded intervals. This is a prerequisite for the next patch, which converts PACKET_TX_RING to standard zerocopy completion. Without this patch first, a bounded ring sender can stall indefinitely behind a virtio-net virtqueue that cannot reclaim. Fixes: 5cd8d46ea156 ("packet: copy user buffers before orphan or clone") Cc: stable@vger.kernel.org Cc: mst@redhat.com Cc: jasowangio@gmail.com Signed-off-by: Willem de Bruijn Link: https://patch.msgid.link/20260919004748.1463985-2-willemdebruijn.kernel@gmail.com Signed-off-by: Jakub Kicinski --- drivers/net/virtio_net.c | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c index e34c52d059d3..bf82ef9874ab 100644 --- a/drivers/net/virtio_net.c +++ b/drivers/net/virtio_net.c @@ -3349,6 +3349,14 @@ static netdev_tx_t start_xmit(struct sk_buff *skb, struct net_device *dev) else virtqueue_disable_cb(sq->vq); + if (!use_napi && + unlikely(skb_orphan_frags(skb, GFP_ATOMIC))) { + DEV_STATS_INC(dev, tx_dropped); + dev_kfree_skb_any(skb); + kick = !xmit_more || netif_xmit_stopped(txq); + goto kick_vq; + } + /* timestamp packet in software */ skb_tx_timestamp(skb); @@ -3381,6 +3389,7 @@ static netdev_tx_t start_xmit(struct sk_buff *skb, struct net_device *dev) kick = use_napi ? __netdev_tx_sent_queue(txq, skb->len, xmit_more) : !xmit_more || netif_xmit_stopped(txq); +kick_vq: if (kick) { if (virtqueue_kick_prepare(sq->vq) && virtqueue_notify(sq->vq)) { u64_stats_update_begin(&sq->stats.syncp); From 9518405613863d0bf0700927a21367f5942cf058 Mon Sep 17 00:00:00 2001 From: Willem de Bruijn Date: Fri, 18 Sep 2026 20:47:29 -0400 Subject: [PATCH 406/694] packet: use ubuf_info completion for TX_RING packets tpacket_snd sends skbs with frags pointing into its ring slots. Slots are released when skb->destructor is called. A call to skb_orphan calls skb->destructor before the skb is freed. This can cause the slot to be reused while still linked into the skb. Switch to standard zerocopy completion (ubuf_info) so the slot is only released once all references to the payload are freed or copied. Restore skb->destructor to standard sock_wfree. The ubuf_info completion callback can be called with a NULL skb, but only from net_zcopy_put and related API, used by zerocopy implementations that hold their own reference on the uarg, such as MSG_ZEROCOPY. This uarg is only ever completed from skb_zcopy_clear, so skb is always set. To prevent userspace from aliasing in-flight state on shared ring slots, allocate tpacket_uarg per packet, rather than per slot. This adds a small allocation to the transmit path. Use standard kmalloc to allow backporting to stable kernels. The uarg holds an sk_wmem_alloc reference, rather than an sk_refcnt reference. packet_free_tx_ring waits on sk_wmem_alloc before freeing the ring pages. Always allocate vec->deferred for tx_ring so page-backed rings also wait on sk_wmem_alloc when skb_copy_ubufs drops page refs before calling tpacket_ubuf_complete. Drop the tx_ring.pg_vec test that tpacket_destruct_skb performed before accessing the slot. The sk_wmem_alloc reference now guarantees that the slot is valid. The test is also not sufficient by itself, as it reads pg_vec without pg_vec_lock, so it can race with packet_set_ring. As a result a slot is released when its payload is copied, which can be before transmission (e.g., in skb_orphan_frags_rx). If copied before skb_tx_timestamp() is called, no slot timestamp is recorded, similar to when skb_orphan() was called early in the datapath before this patch. Revert the now unused previous skb_zcopy_.._nouarg infra. Depends on commit 992cc9f94ca9 ("net/packet: defer vmalloc TX_RING free until skbs finish"). Reported-by: Katherine Leaver Reported-by: Bjoern Doebel Closes: https://lore.kernel.org/netdev/20260909085542.3370986-1-doebel@amazon.de/ Fixes: 5cd8d46ea156 ("packet: copy user buffers before orphan or clone") Cc: stable@vger.kernel.org Signed-off-by: Willem de Bruijn Link: https://patch.msgid.link/20260919004748.1463985-3-willemdebruijn.kernel@gmail.com Signed-off-by: Jakub Kicinski --- include/linux/skbuff.h | 19 +------- net/packet/af_packet.c | 102 ++++++++++++++++++++++++++--------------- 2 files changed, 65 insertions(+), 56 deletions(-) diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index 421f6fc45451..b14d6be7370b 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -1834,22 +1834,6 @@ static inline void skb_zcopy_set(struct sk_buff *skb, struct ubuf_info *uarg, } } -static inline void skb_zcopy_set_nouarg(struct sk_buff *skb, void *val) -{ - skb_shinfo(skb)->destructor_arg = (void *)((uintptr_t) val | 0x1UL); - skb_shinfo(skb)->flags |= SKBFL_ZEROCOPY_FRAG; -} - -static inline bool skb_zcopy_is_nouarg(struct sk_buff *skb) -{ - return (uintptr_t) skb_shinfo(skb)->destructor_arg & 0x1UL; -} - -static inline void *skb_zcopy_get_nouarg(struct sk_buff *skb) -{ - return (void *)((uintptr_t) skb_shinfo(skb)->destructor_arg & ~0x1UL); -} - static inline void net_zcopy_put(struct ubuf_info *uarg) { if (uarg) @@ -1872,8 +1856,7 @@ static inline void skb_zcopy_clear(struct sk_buff *skb, bool zerocopy_success) struct ubuf_info *uarg = skb_zcopy(skb); if (uarg) { - if (!skb_zcopy_is_nouarg(skb)) - uarg->ops->complete(skb, uarg, zerocopy_success); + uarg->ops->complete(skb, uarg, zerocopy_success); skb_shinfo(skb)->flags &= ~SKBFL_ALL_ZEROCOPY; } diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c index 50cae32ae269..64b501db660a 100644 --- a/net/packet/af_packet.c +++ b/net/packet/af_packet.c @@ -2530,26 +2530,6 @@ static int tpacket_rcv(struct sk_buff *skb, struct net_device *dev, goto drop_n_restore; } -static void tpacket_destruct_skb(struct sk_buff *skb) -{ - struct packet_sock *po = pkt_sk(skb->sk); - - if (likely(po->tx_ring.pg_vec)) { - void *ph; - __u32 ts; - - ph = skb_zcopy_get_nouarg(skb); - - ts = __packet_set_timestamp(po, ph, skb); - __packet_set_status(po, ph, TP_STATUS_AVAILABLE | ts); - - packet_dec_pending(&po->tx_ring); - complete(&po->skb_completion); - } - - sock_wfree(skb); -} - static int __packet_snd_vnet_parse(struct virtio_net_hdr *vnet_hdr, size_t len) { if ((vnet_hdr->flags & VIRTIO_NET_HDR_F_NEEDS_CSUM) && @@ -2589,27 +2569,56 @@ static int packet_snd_vnet_parse(struct msghdr *msg, size_t *len, return 0; } +struct tpacket_uarg { + struct ubuf_info ubuf; + struct packet_sock *po; + void *ph; +}; + +static void tpacket_ubuf_complete(struct sk_buff *skb, struct ubuf_info *uarg, + bool success) +{ + struct tpacket_uarg *tu = container_of(uarg, struct tpacket_uarg, ubuf); + struct packet_sock *po = tu->po; + void *ph = tu->ph; + __u32 ts; + + DEBUG_NET_WARN_ON_ONCE(!skb); + + if (!refcount_dec_and_test(&uarg->refcnt)) + return; + + ts = __packet_set_timestamp(po, ph, skb); + __packet_set_status(po, ph, TP_STATUS_AVAILABLE | ts); + + packet_dec_pending(&po->tx_ring); + complete(&po->skb_completion); + + kfree(tu); + sk_free(&po->sk); +} + +static const struct ubuf_info_ops tpacket_ubuf_ops = { + .complete = tpacket_ubuf_complete, +}; + static int tpacket_fill_skb(struct packet_sock *po, struct sk_buff *skb, - void *frame, struct net_device *dev, void *data, int tp_len, + struct net_device *dev, void *data, int tp_len, __be16 proto, unsigned char *addr, int hlen, int copylen, int hard_header_len, const struct sockcm_cookie *sockc) { - union tpacket_uhdr ph; int to_write, offset, len, nr_frags, len_max; struct socket *sock = po->sk.sk_socket; struct page *page; int err; - ph.raw = frame; - skb->protocol = proto; skb->dev = dev; skb->priority = sockc->priority; skb->mark = sockc->mark; skb_set_delivery_type_by_clockid(skb, sockc->transmit_time, po->sk.sk_clockid); skb_setup_tx_timestamp(skb, sockc); - skb_zcopy_set_nouarg(skb, ph.raw); skb_reserve(skb, hlen); skb_reset_network_header(skb); @@ -2749,6 +2758,7 @@ static int tpacket_snd(struct packet_sock *po, struct msghdr *msg) struct virtio_net_hdr vnet_hdr; bool has_vnet_hdr = false; struct sockcm_cookie sockc; + struct tpacket_uarg *uarg; __be16 proto; int err, reserve = 0; void *ph; @@ -2876,7 +2886,7 @@ static int tpacket_snd(struct packet_sock *po, struct msghdr *msg) err = len_sum; goto out_status; } - tp_len = tpacket_fill_skb(po, skb, ph, dev, data, tp_len, proto, + tp_len = tpacket_fill_skb(po, skb, dev, data, tp_len, proto, addr, hlen, copylen, hard_header_len, &sockc); if (likely(tp_len >= 0) && @@ -2908,7 +2918,24 @@ static int tpacket_snd(struct packet_sock *po, struct msghdr *msg) virtio_net_hdr_set_proto(skb, &vnet_hdr); } - skb->destructor = tpacket_destruct_skb; + uarg = kmalloc(sizeof(*uarg), GFP_KERNEL); + if (unlikely(!uarg)) { + if (likely(len_sum > 0)) + err = len_sum; + else + err = -ENOMEM; + goto out_status; + } + uarg->po = po; + uarg->ph = ph; + uarg->ubuf.ops = &tpacket_ubuf_ops; + uarg->ubuf.flags = SKBFL_ZEROCOPY_FRAG; + refcount_set(&uarg->ubuf.refcnt, 1); + + /* Hold a sk_wmem_alloc reference until completion */ + refcount_inc(&po->sk.sk_wmem_alloc); + skb_zcopy_init(skb, &uarg->ubuf); + __packet_set_status(po, ph, TP_STATUS_SENDING); packet_inc_pending(&po->tx_ring); @@ -4486,21 +4513,20 @@ static struct pgv *alloc_pg_vec(struct tpacket_req *req, int order, bool tx_ring vec->len = block_nr; pg_vec = vec->pg_vec; + if (tx_ring) { + vec->deferred = kzalloc_obj(*vec->deferred, + GFP_KERNEL | __GFP_NOWARN); + if (!vec->deferred) + goto out_free_pgvec; + vec->deferred->vec = vec; + INIT_DELAYED_WORK(&vec->deferred->work, + packet_free_pg_vec_work); + } + for (i = 0; i < block_nr; i++) { pg_vec[i].buffer = alloc_one_pg_vec_page(order); if (unlikely(!pg_vec[i].buffer)) goto out_free_pgvec; - - if (tx_ring && !vec->deferred && - is_vmalloc_addr(pg_vec[i].buffer)) { - vec->deferred = kzalloc_obj(*vec->deferred, - GFP_KERNEL | __GFP_NOWARN); - if (!vec->deferred) - goto out_free_pgvec; - vec->deferred->vec = vec; - INIT_DELAYED_WORK(&vec->deferred->work, - packet_free_pg_vec_work); - } } out: From cae23ae3f7887a1cf8a75da38edcebeef695040f Mon Sep 17 00:00:00 2001 From: Xin Long Date: Mon, 21 Sep 2026 14:03:45 -0400 Subject: [PATCH 407/694] sctp: hold asoc or transport before mod_timer() in timer handlers Take the association or transport reference before rearming a timer in the timer handlers. The existing code calls mod_timer() before taking the reference needed by the rearmed timer without holding the sock lock. This creates a race with timer cleanup: if the timer is deleted after mod_timer() returns but before the reference is taken, the cleanup path can drop the timer's reference and destroy the transport or association. The timer handler then takes a reference on the already freed object and eventually drops it, causing a refcount underflow. Hold the object before mod_timer() and drop the reference if mod_timer() reports that the timer was already pending in timer handlers. Apply the same ordering to the proto-unreachable path, which can rearm a transport timer outside the timer handlers without holding the sock lock. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Reported-by: Tangxin Xie Signed-off-by: Xin Long Link: https://patch.msgid.link/c31b5e3ee2b7274e804f5eba2f21e2412e7eef7a.1790013825.git.lucien.xin@gmail.com Signed-off-by: Jakub Kicinski --- net/sctp/input.c | 7 ++++--- net/sctp/sm_sideeffect.c | 37 ++++++++++++++++++++++--------------- 2 files changed, 26 insertions(+), 18 deletions(-) diff --git a/net/sctp/input.c b/net/sctp/input.c index 864741fae418..9494cfa51106 100644 --- a/net/sctp/input.c +++ b/net/sctp/input.c @@ -436,9 +436,10 @@ void sctp_icmp_proto_unreachable(struct sock *sk, if (timer_pending(&t->proto_unreach_timer)) return; else { - if (!mod_timer(&t->proto_unreach_timer, - jiffies + (HZ/20))) - sctp_transport_hold(t); + sctp_transport_hold(t); + if (mod_timer(&t->proto_unreach_timer, + jiffies + (HZ / 20))) + sctp_transport_put(t); } } else { struct net *net = sock_net(sk); diff --git a/net/sctp/sm_sideeffect.c b/net/sctp/sm_sideeffect.c index 0d99b7e8c082..35f540fb15fc 100644 --- a/net/sctp/sm_sideeffect.c +++ b/net/sctp/sm_sideeffect.c @@ -244,8 +244,9 @@ void sctp_generate_t3_rtx_event(struct timer_list *t) pr_debug("%s: sock is busy\n", __func__); /* Try again later. */ - if (!mod_timer(&transport->T3_rtx_timer, jiffies + (HZ/20))) - sctp_transport_hold(transport); + sctp_transport_hold(transport); + if (mod_timer(&transport->T3_rtx_timer, jiffies + (HZ / 20))) + sctp_transport_put(transport); goto out_unlock; } @@ -280,8 +281,9 @@ static void sctp_generate_timeout_event(struct sctp_association *asoc, timeout_type); /* Try again later. */ - if (!mod_timer(&asoc->timers[timeout_type], jiffies + (HZ/20))) - sctp_association_hold(asoc); + sctp_association_hold(asoc); + if (mod_timer(&asoc->timers[timeout_type], jiffies + (HZ / 20))) + sctp_association_put(asoc); goto out_unlock; } @@ -378,8 +380,9 @@ void sctp_generate_heartbeat_event(struct timer_list *t) pr_debug("%s: sock is busy\n", __func__); /* Try again later. */ - if (!mod_timer(&transport->hb_timer, jiffies + (HZ/20))) - sctp_transport_hold(transport); + sctp_transport_hold(transport); + if (mod_timer(&transport->hb_timer, jiffies + (HZ / 20))) + sctp_transport_put(transport); goto out_unlock; } @@ -388,8 +391,9 @@ void sctp_generate_heartbeat_event(struct timer_list *t) timeout = sctp_transport_timeout(transport); if (elapsed < timeout) { elapsed = timeout - elapsed; - if (!mod_timer(&transport->hb_timer, jiffies + elapsed)) - sctp_transport_hold(transport); + sctp_transport_hold(transport); + if (mod_timer(&transport->hb_timer, jiffies + elapsed)) + sctp_transport_put(transport); goto out_unlock; } @@ -422,9 +426,10 @@ void sctp_generate_proto_unreach_event(struct timer_list *t) pr_debug("%s: sock is busy\n", __func__); /* Try again later. */ - if (!mod_timer(&transport->proto_unreach_timer, - jiffies + (HZ/20))) - sctp_transport_hold(transport); + sctp_transport_hold(transport); + if (mod_timer(&transport->proto_unreach_timer, + jiffies + (HZ / 20))) + sctp_transport_put(transport); goto out_unlock; } @@ -458,8 +463,9 @@ void sctp_generate_reconf_event(struct timer_list *t) pr_debug("%s: sock is busy\n", __func__); /* Try again later. */ - if (!mod_timer(&transport->reconf_timer, jiffies + (HZ / 20))) - sctp_transport_hold(transport); + sctp_transport_hold(transport); + if (mod_timer(&transport->reconf_timer, jiffies + (HZ / 20))) + sctp_transport_put(transport); goto out_unlock; } @@ -495,8 +501,9 @@ void sctp_generate_probe_event(struct timer_list *t) pr_debug("%s: sock is busy\n", __func__); /* Try again later. */ - if (!mod_timer(&transport->probe_timer, jiffies + (HZ / 20))) - sctp_transport_hold(transport); + sctp_transport_hold(transport); + if (mod_timer(&transport->probe_timer, jiffies + (HZ / 20))) + sctp_transport_put(transport); goto out_unlock; } From 73b3fc67a493e9b9a8e94a38839f6638d12fe7a6 Mon Sep 17 00:00:00 2001 From: Mina Almasry Date: Mon, 21 Sep 2026 19:54:12 +0000 Subject: [PATCH 408/694] net: devmem: document that bind-tx is unprivileged by design Unlike bind-rx, which configures shared NIC RX queues to steer incoming traffic into the caller's dmabuf and requires CAP_NET_ADMIN (uns-admin-perm), bind-tx only DMA-maps the caller's dmabuf so the caller can transmit from it on their own sockets without affecting other traffic or device configuration. Add a comment in netdev.yaml and above netdev_nl_bind_tx_doit() to make it explicit that NETDEV_CMD_BIND_TX is unprivileged by design. Signed-off-by: Mina Almasry Acked-by: Stanislav Fomichev Acked-by: Daniel Borkmann Link: https://patch.msgid.link/20260921195545.493253-1-almasrymina@google.com Signed-off-by: Jakub Kicinski --- Documentation/netlink/specs/netdev.yaml | 2 ++ net/core/netdev-genl.c | 6 ++++++ 2 files changed, 8 insertions(+) diff --git a/Documentation/netlink/specs/netdev.yaml b/Documentation/netlink/specs/netdev.yaml index 3e3f03bd5c29..0562ba1f567a 100644 --- a/Documentation/netlink/specs/netdev.yaml +++ b/Documentation/netlink/specs/netdev.yaml @@ -851,6 +851,8 @@ operations: name: bind-tx doc: Bind dmabuf to netdev for TX attribute-set: dmabuf + # Intentionally unprivileged (no admin-perm / uns-admin-perm); see + # comment above netdev_nl_bind_tx_doit(). do: request: attributes: diff --git a/net/core/netdev-genl.c b/net/core/netdev-genl.c index cb18db681640..33b9f4eb9565 100644 --- a/net/core/netdev-genl.c +++ b/net/core/netdev-genl.c @@ -1168,6 +1168,12 @@ netdev_find_netmem_tx_dev(struct net_device *dev) return NULL; } +/* Note: NETDEV_CMD_BIND_TX is intentionally unprivileged (no + * GENL_ADMIN_PERM / GENL_UNS_ADMIN_PERM). Unlike bind-rx, which configures + * shared NIC RX queues, bind-tx only DMA-maps the caller's dmabuf so they can + * transmit from it on their own sockets without affecting other traffic or + * device state. + */ int netdev_nl_bind_tx_doit(struct sk_buff *skb, struct genl_info *info) { struct net_devmem_dmabuf_binding *binding; From d68acbf93531abdb5b02b21994cd4c15a3c95b42 Mon Sep 17 00:00:00 2001 From: Maxime Chevallier Date: Thu, 17 Sep 2026 23:53:32 +0200 Subject: [PATCH 409/694] net: stmmac: selftests: Support running selftests on DSA conduits Most stmmac selftests rely on dev_add_pack() to add custom handlers, that validate the packets sent to ourselves through MAC loopback. However, when the stmmac-driven interface is a DSA CPU conduit, all frames that are received have ETH_P_XDSA as a protocol, even though they don't actually contain any tag as they come from the loopback and not the switch. This will prevent any incoming packet to match our packet handlers. Let's register a ETH_P_ALL packet handler when we detect that we're a DSA conduit, and use a proxy packet handler to filter the h_proto. As this allows external frames to be received through our .func(), the packet handler is added after the dev->addr field is populated in our selftest attributes. Note that we may still receive incoming packets from the switch, but these frames shouldn't interfere with the very specific frames used for selftests, and stmmac selftests in general aren't safe against external traffic interferences. This was validated on a WPQ864 devkit for IPQ8064, that has the SoC connected to a QCA8k switch. The ARP offload's packet handler is left alone, this feature is just not implemented in stmmac and due for removal. Fixes: 091810dbded9 ("net: stmmac: Introduce selftests support") Reviewed-by: Nicolai Buchwitz Signed-off-by: Maxime Chevallier Link: https://patch.msgid.link/20260917215339.2022523-2-maxime.chevallier@bootlin.com Signed-off-by: Jakub Kicinski --- .../stmicro/stmmac/stmmac_selftests.c | 89 +++++++++++++++---- 1 file changed, 71 insertions(+), 18 deletions(-) diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c index 6372ec7c3f31..614b5995dec5 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c @@ -12,6 +12,7 @@ #include #include #include +#include #include #include #include @@ -237,6 +238,9 @@ struct stmmac_test_priv { struct stmmac_packet_attrs *packet; struct packet_type pt; struct completion comp; + __be16 packet_type; + int (*func)(struct sk_buff *skb, struct net_device *ndev, + struct packet_type *pt, struct net_device *orig_ndev); int double_vlan; int vlan_id; int ok; @@ -316,6 +320,50 @@ static int stmmac_test_loopback_validate(struct sk_buff *skb, return 0; } +static int stmmac_sft_filter(struct sk_buff *skb, struct net_device *ndev, + struct packet_type *pt, + struct net_device *orig_ndev) +{ + struct stmmac_test_priv *tpriv = pt->af_packet_priv; + struct ethhdr *hdr = eth_hdr(skb); + int ret = 0; + + if (hdr->h_proto == tpriv->packet_type) { + struct sk_buff *nskb = skb_clone(skb, GFP_ATOMIC); + + if (nskb) + ret = tpriv->func(nskb, ndev, pt, orig_ndev); + } + + kfree_skb(skb); + return ret; +} + +static void stmmac_sft_add_pack(struct packet_type *pt) +{ + struct stmmac_test_priv *tpriv = pt->af_packet_priv; + + if (netdev_uses_dsa(tpriv->pt.dev)) { + tpriv->packet_type = tpriv->pt.type; + tpriv->func = tpriv->pt.func; + + /* DSA conduit will report ETH_P_XDSA, so our packet handler + * won't match. Let's register a ETH_P_ALL match and filter + * manually in stmmac_sft_filter. + */ + tpriv->pt.type = htons(ETH_P_ALL); + tpriv->pt.func = stmmac_sft_filter; + tpriv->pt.ignore_outgoing = true; + } + + dev_add_pack(pt); +} + +static void stmmac_sft_remove_pack(struct packet_type *pt) +{ + dev_remove_pack(pt); +} + static int __stmmac_test_loopback(struct stmmac_priv *priv, struct stmmac_packet_attrs *attr) { @@ -337,7 +385,7 @@ static int __stmmac_test_loopback(struct stmmac_priv *priv, tpriv->packet = attr; if (!attr->dont_wait) - dev_add_pack(&tpriv->pt); + stmmac_sft_add_pack(&tpriv->pt); skb = stmmac_test_get_udp_skb(priv, attr); if (!skb) { @@ -360,7 +408,7 @@ static int __stmmac_test_loopback(struct stmmac_priv *priv, cleanup: if (!attr->dont_wait) - dev_remove_pack(&tpriv->pt); + stmmac_sft_remove_pack(&tpriv->pt); kfree(tpriv); return ret; } @@ -767,7 +815,7 @@ static int stmmac_test_flowctrl(struct stmmac_priv *priv) tpriv->pt.func = stmmac_test_flowctrl_validate; tpriv->pt.dev = priv->dev; tpriv->pt.af_packet_priv = tpriv; - dev_add_pack(&tpriv->pt); + stmmac_sft_add_pack(&tpriv->pt); /* Compute minimum number of packets to make FIFO full */ pkt_count = rx_fifo_size; @@ -823,7 +871,7 @@ static int stmmac_test_flowctrl(struct stmmac_priv *priv) cleanup: dev_mc_del(priv->dev, paddr); dev_set_promiscuity(priv->dev, -1); - dev_remove_pack(&tpriv->pt); + stmmac_sft_remove_pack(&tpriv->pt); kfree(tpriv); return ret; } @@ -928,18 +976,20 @@ static int __stmmac_test_vlanfilt(struct stmmac_priv *priv) * HASH values. */ tpriv->vlan_id = 0x123; - dev_add_pack(&tpriv->pt); ret = vlan_vid_add(priv->dev, htons(ETH_P_8021Q), tpriv->vlan_id); if (ret) goto cleanup; + attr.vlan = 1; + attr.dst = priv->dev->dev_addr; + attr.sport = 9; + attr.dport = 9; + + stmmac_sft_add_pack(&tpriv->pt); + for (i = 0; i < 4; i++) { - attr.vlan = 1; attr.vlan_id_out = tpriv->vlan_id + i; - attr.dst = priv->dev->dev_addr; - attr.sport = 9; - attr.dport = 9; skb = stmmac_test_get_udp_skb(priv, &attr); if (!skb) { @@ -966,9 +1016,9 @@ static int __stmmac_test_vlanfilt(struct stmmac_priv *priv) } vlan_del: + stmmac_sft_remove_pack(&tpriv->pt); vlan_vid_del(priv->dev, htons(ETH_P_8021Q), tpriv->vlan_id); cleanup: - dev_remove_pack(&tpriv->pt); kfree(tpriv); return ret; } @@ -1022,18 +1072,20 @@ static int __stmmac_test_dvlanfilt(struct stmmac_priv *priv) * HASH values. */ tpriv->vlan_id = 0x123; - dev_add_pack(&tpriv->pt); ret = vlan_vid_add(priv->dev, htons(ETH_P_8021AD), tpriv->vlan_id); if (ret) goto cleanup; + attr.vlan = 2; + attr.dst = priv->dev->dev_addr; + attr.sport = 9; + attr.dport = 9; + + stmmac_sft_add_pack(&tpriv->pt); + for (i = 0; i < 4; i++) { - attr.vlan = 2; attr.vlan_id_out = tpriv->vlan_id + i; - attr.dst = priv->dev->dev_addr; - attr.sport = 9; - attr.dport = 9; skb = stmmac_test_get_udp_skb(priv, &attr); if (!skb) { @@ -1060,9 +1112,9 @@ static int __stmmac_test_dvlanfilt(struct stmmac_priv *priv) } vlan_del: + stmmac_sft_remove_pack(&tpriv->pt); vlan_vid_del(priv->dev, htons(ETH_P_8021AD), tpriv->vlan_id); cleanup: - dev_remove_pack(&tpriv->pt); kfree(tpriv); return ret; } @@ -1293,7 +1345,6 @@ static int stmmac_test_vlanoff_common(struct stmmac_priv *priv, bool svlan) tpriv->pt.af_packet_priv = tpriv; tpriv->packet = &attr; tpriv->vlan_id = 0x123; - dev_add_pack(&tpriv->pt); ret = vlan_vid_add(priv->dev, htons(proto), tpriv->vlan_id); if (ret) @@ -1301,6 +1352,8 @@ static int stmmac_test_vlanoff_common(struct stmmac_priv *priv, bool svlan) attr.dst = priv->dev->dev_addr; + stmmac_sft_add_pack(&tpriv->pt); + skb = stmmac_test_get_udp_skb(priv, &attr); if (!skb) { ret = -ENOMEM; @@ -1318,9 +1371,9 @@ static int stmmac_test_vlanoff_common(struct stmmac_priv *priv, bool svlan) ret = tpriv->ok ? 0 : -ETIMEDOUT; vlan_del: + stmmac_sft_remove_pack(&tpriv->pt); vlan_vid_del(priv->dev, htons(proto), tpriv->vlan_id); cleanup: - dev_remove_pack(&tpriv->pt); kfree(tpriv); return ret; } From c8c1795aa8106293020a836d01e127e98f442925 Mon Sep 17 00:00:00 2001 From: Maxime Chevallier Date: Thu, 17 Sep 2026 23:53:33 +0200 Subject: [PATCH 410/694] net: stmmac: selftests: Validate EEE based on the actual LPI timer value The EEE selftest is a 2-step test : - It validates that we enter in LPI mode with the irq_tx_path_in_lpi_mode_n counter - It then validates that we exit LPI when sending a frame, with the irq_tx_path_exit_lpi_mode_n counter. The current state of the test lacks 2 main things : - We don't know exactly when was the previous frame sent (it's from the previous selftest) - The timeout is hardcoded, while the LPI is entered after a user-configurable delay. On top of that, the timeout loop uses a pre-decrement iterator (--retries) that actually only iterate nine times, so 900ms while the default LPI value is 1 second. Let's therefore make it more deterministic : - Send a frame at the beginning of the test - Wait for more than the lpi timer value, we timeout after about twice the value, - Then send another frame, and verify that we do go out of LPI, also with a timeout. As LPI timer can get pretty high, bail out if LPI timer is over 5 seconds. Note that the test's goal isn't to validate the LPI timer value itself, only that we enter/leave LPI mode. Fixes: 091810dbded9 ("net: stmmac: Introduce selftests support") Reviewed-by: Nicolai Buchwitz Signed-off-by: Maxime Chevallier Link: https://patch.msgid.link/20260917215339.2022523-3-maxime.chevallier@bootlin.com Signed-off-by: Jakub Kicinski --- .../stmicro/stmmac/stmmac_selftests.c | 44 ++++++++++++++++--- 1 file changed, 37 insertions(+), 7 deletions(-) diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c index 614b5995dec5..2f9f7746c40a 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c @@ -30,6 +30,7 @@ struct stmmachdr { sizeof(struct stmmachdr)) #define STMMAC_TEST_PKT_MAGIC 0xdeadcafecafedeadULL #define STMMAC_LB_TIMEOUT msecs_to_jiffies(200) +#define STMMAC_SFT_MAX_LPI (5 * USEC_PER_SEC) struct stmmac_packet_attrs { int vlan; @@ -462,12 +463,16 @@ static int stmmac_test_mmc(struct stmmac_priv *priv) static int stmmac_test_eee(struct stmmac_priv *priv) { struct stmmac_extra_stats *initial, *final; - int retries = 10; + unsigned long timeout, max_duration; int ret; if (!priv->dma_cap.eee || !priv->eee_active) return -EOPNOTSUPP; + /* Bail out if the configured LPI timer is too long */ + if (priv->tx_lpi_timer > STMMAC_SFT_MAX_LPI) + return -EOPNOTSUPP; + initial = kzalloc_obj(*initial); if (!initial) return -ENOMEM; @@ -478,14 +483,21 @@ static int stmmac_test_eee(struct stmmac_priv *priv) goto out_free_initial; } + /* Snapshot stats, we want to count the in_lpi events. We may enter + * LPI just after the packet was sent. + */ memcpy(initial, &priv->xstats, sizeof(*initial)); + /* Send a frame, then wait to enter LPI */ ret = stmmac_test_mac_loopback(priv); if (ret) goto out_free_final; + max_duration = usecs_to_jiffies(2 * priv->tx_lpi_timer); + /* We have no traffic in the line so, sooner or later it will go LPI */ - while (--retries) { + timeout = jiffies + max_duration; + while (!time_after(jiffies, timeout)) { memcpy(final, &priv->xstats, sizeof(*final)); if (final->irq_tx_path_in_lpi_mode_n > @@ -494,20 +506,38 @@ static int stmmac_test_eee(struct stmmac_priv *priv) msleep(100); } - if (!retries) { + memcpy(final, &priv->xstats, sizeof(*final)); + if (final->irq_tx_path_in_lpi_mode_n <= + initial->irq_tx_path_in_lpi_mode_n) { ret = -ETIMEDOUT; goto out_free_final; } - if (final->irq_tx_path_in_lpi_mode_n <= - initial->irq_tx_path_in_lpi_mode_n) { - ret = -EINVAL; + /* Re-snapshot, as we want to measure exit_lpi events. We should be + * in LPI right now. + */ + memcpy(initial, &priv->xstats, sizeof(*initial)); + + /* TX something so we go out of LPI */ + ret = stmmac_test_mac_loopback(priv); + if (ret) goto out_free_final; + + /* Wait for the exit LPI interrupt */ + timeout = jiffies + max_duration; + while (!time_after(jiffies, timeout)) { + memcpy(final, &priv->xstats, sizeof(*final)); + + if (final->irq_tx_path_exit_lpi_mode_n > + initial->irq_tx_path_exit_lpi_mode_n) + break; + msleep(100); } + memcpy(final, &priv->xstats, sizeof(*final)); if (final->irq_tx_path_exit_lpi_mode_n <= initial->irq_tx_path_exit_lpi_mode_n) { - ret = -EINVAL; + ret = -ETIMEDOUT; goto out_free_final; } From ba804b23d76d278ee475b8427fa7c5623ce5e270 Mon Sep 17 00:00:00 2001 From: Maxime Chevallier Date: Thu, 17 Sep 2026 23:53:34 +0200 Subject: [PATCH 411/694] net: stmmac: selftests: Check the dev->features for S-TAG offload testing The S-TAG offload insertion incorrectly checks the dvlan (double vlan) DMA cap, which is different than S-TAG support. Use NETIF_F_HW_VLAN_STAG_TX to check if the feature is supported instead. Note that this flag isn't set in stmmac yet, but contrary to ARP offload, this is a feature that has a chance to get there eventually so let's leave the selftest here for now. It'll report -EOPNOTSUPP in the meantime. Fixes: 091810dbded9 ("net: stmmac: Introduce selftests support") Reviewed-by: Nicolai Buchwitz Signed-off-by: Maxime Chevallier Link: https://patch.msgid.link/20260917215339.2022523-4-maxime.chevallier@bootlin.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c index 2f9f7746c40a..de02c0da56dc 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c @@ -1415,7 +1415,7 @@ static int stmmac_test_vlanoff(struct stmmac_priv *priv) static int stmmac_test_svlanoff(struct stmmac_priv *priv) { - if (!priv->dma_cap.dvlan) + if (!(priv->dev->features & NETIF_F_HW_VLAN_STAG_TX)) return -EOPNOTSUPP; return stmmac_test_vlanoff_common(priv, true); } From 960db6f65788c21249ea04a947d5c01e38d19294 Mon Sep 17 00:00:00 2001 From: Maxime Chevallier Date: Thu, 17 Sep 2026 23:53:35 +0200 Subject: [PATCH 412/694] net: stmmac: selftests: Capture all packets for vlan checks While we use vlan_vid_add to trigger the tag filtering machinery in the driver, there's no netdev associated to the VLAN. This causes the skb to arrive with empty skb->vlan_tci fields, as the packet is marked OTHERHOST in __netif_receive_skb_core(), and we fail our validation. Let's use the proxy mechanism introduced for DSA, that registers a ETH_P_ALL packet handler that runs earlier, before the vlan netdev lookup, then filters for the correct ethertype before passing an skb clone to our validation function. As we may receive external frames with the right tag from the outside, let's move the address check in the vlan validation function earlier. Fixes: 091810dbded9 ("net: stmmac: Introduce selftests support") Reviewed-by: Nicolai Buchwitz Signed-off-by: Maxime Chevallier Link: https://patch.msgid.link/20260917215339.2022523-5-maxime.chevallier@bootlin.com Signed-off-by: Jakub Kicinski --- .../stmicro/stmmac/stmmac_selftests.c | 19 +++++++++++++------ 1 file changed, 13 insertions(+), 6 deletions(-) diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c index de02c0da56dc..43b8411c5112 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c @@ -242,6 +242,7 @@ struct stmmac_test_priv { __be16 packet_type; int (*func)(struct sk_buff *skb, struct net_device *ndev, struct packet_type *pt, struct net_device *orig_ndev); + bool capture_all; int double_vlan; int vlan_id; int ok; @@ -344,13 +345,15 @@ static void stmmac_sft_add_pack(struct packet_type *pt) { struct stmmac_test_priv *tpriv = pt->af_packet_priv; - if (netdev_uses_dsa(tpriv->pt.dev)) { + if (netdev_uses_dsa(tpriv->pt.dev) || tpriv->capture_all) { tpriv->packet_type = tpriv->pt.type; tpriv->func = tpriv->pt.func; /* DSA conduit will report ETH_P_XDSA, so our packet handler * won't match. Let's register a ETH_P_ALL match and filter - * manually in stmmac_sft_filter. + * manually in stmmac_sft_filter. This is also useful for + * VLAN tests, to capture packets otherwise marked as + * OTHERHOST. */ tpriv->pt.type = htons(ETH_P_ALL); tpriv->pt.func = stmmac_sft_filter; @@ -943,6 +946,11 @@ static int stmmac_test_vlan_validate(struct sk_buff *skb, goto out; if (skb_headlen(skb) < (STMMAC_TEST_PKT_SIZE - ETH_HLEN)) goto out; + + ehdr = (struct ethhdr *)skb_mac_header(skb); + if (!ether_addr_equal_unaligned(ehdr->h_dest, tpriv->packet->dst)) + goto out; + if (tpriv->vlan_id) { if (skb->vlan_proto != htons(proto)) goto out; @@ -954,10 +962,6 @@ static int stmmac_test_vlan_validate(struct sk_buff *skb, } } - ehdr = (struct ethhdr *)skb_mac_header(skb); - if (!ether_addr_equal_unaligned(ehdr->h_dest, tpriv->packet->dst)) - goto out; - ihdr = ip_hdr(skb); if (tpriv->double_vlan) ihdr = (struct iphdr *)(skb_network_header(skb) + 4); @@ -999,6 +1003,7 @@ static int __stmmac_test_vlanfilt(struct stmmac_priv *priv) tpriv->pt.dev = priv->dev; tpriv->pt.af_packet_priv = tpriv; tpriv->packet = &attr; + tpriv->capture_all = true; /* * As we use HASH filtering, false positives may appear. This is a @@ -1095,6 +1100,7 @@ static int __stmmac_test_dvlanfilt(struct stmmac_priv *priv) tpriv->pt.dev = priv->dev; tpriv->pt.af_packet_priv = tpriv; tpriv->packet = &attr; + tpriv->capture_all = true; /* * As we use HASH filtering, false positives may appear. This is a @@ -1375,6 +1381,7 @@ static int stmmac_test_vlanoff_common(struct stmmac_priv *priv, bool svlan) tpriv->pt.af_packet_priv = tpriv; tpriv->packet = &attr; tpriv->vlan_id = 0x123; + tpriv->capture_all = true; ret = vlan_vid_add(priv->dev, htons(proto), tpriv->vlan_id); if (ret) From b42e7012773a0e81e97a2dda6ef907f5147a6658 Mon Sep 17 00:00:00 2001 From: Maxime Chevallier Date: Thu, 17 Sep 2026 23:53:36 +0200 Subject: [PATCH 413/694] net: stmmac: dwmac4: Use the correct bufzise when the len is exactly 8K DMA bufsize selection isn't made on the MTU but the actual frame length, so including the L2 header. On DWMAC4, if the len is exactly BUF_SIZE_8KiB, the next larger size is incorrectly selected. Lets fix the comparison and while at it, rename the parameter from len to mtu. Fixes: c3efed5ad1b0 ("net: stmmac: Enable dwmac4 jumbo frame more than 8KiB"). Signed-off-by: Maxime Chevallier Reviewed-by: Nicolai Buchwitz Link: https://patch.msgid.link/20260917215339.2022523-6-maxime.chevallier@bootlin.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/stmicro/stmmac/dwmac4_descs.c | 4 ++-- drivers/net/ethernet/stmicro/stmmac/hwif.h | 2 +- drivers/net/ethernet/stmicro/stmmac/ring_mode.c | 4 ++-- 3 files changed, 5 insertions(+), 5 deletions(-) diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac4_descs.c b/drivers/net/ethernet/stmicro/stmmac/dwmac4_descs.c index 2994df41ec2c..c6a8f8d73501 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwmac4_descs.c +++ b/drivers/net/ethernet/stmicro/stmmac/dwmac4_descs.c @@ -474,11 +474,11 @@ static void dwmac4_set_sarc(struct dma_desc *p, u32 sarc_type) sarc_type)); } -static int set_16kib_bfsize(int mtu) +static int set_16kib_bfsize(int len) { int ret = 0; - if (unlikely(mtu >= BUF_SIZE_8KiB)) + if (unlikely(len > BUF_SIZE_8KiB)) ret = BUF_SIZE_16KiB; return ret; } diff --git a/drivers/net/ethernet/stmicro/stmmac/hwif.h b/drivers/net/ethernet/stmicro/stmmac/hwif.h index 9314bcb85c22..857f7562c6c6 100644 --- a/drivers/net/ethernet/stmicro/stmmac/hwif.h +++ b/drivers/net/ethernet/stmicro/stmmac/hwif.h @@ -540,7 +540,7 @@ struct stmmac_mode_ops { bool (*is_jumbo_frm)(unsigned int len, bool enh_desc); int (*jumbo_frm)(struct stmmac_tx_queue *tx_q, struct sk_buff *skb, int csum); - int (*set_16kib_bfsize)(int mtu); + int (*set_16kib_bfsize)(int len); void (*init_desc3)(struct dma_desc *p); void (*refill_desc3)(struct stmmac_rx_queue *rx_q, struct dma_desc *p); void (*clean_desc3)(struct stmmac_tx_queue *tx_q, struct dma_desc *p); diff --git a/drivers/net/ethernet/stmicro/stmmac/ring_mode.c b/drivers/net/ethernet/stmicro/stmmac/ring_mode.c index f7949419eb9f..d2f0c321661d 100644 --- a/drivers/net/ethernet/stmicro/stmmac/ring_mode.c +++ b/drivers/net/ethernet/stmicro/stmmac/ring_mode.c @@ -124,10 +124,10 @@ static void clean_desc3(struct stmmac_tx_queue *tx_q, struct dma_desc *p) p->des3 = 0; } -static int set_16kib_bfsize(int mtu) +static int set_16kib_bfsize(int len) { int ret = 0; - if (unlikely(mtu > BUF_SIZE_8KiB)) + if (unlikely(len > BUF_SIZE_8KiB)) ret = BUF_SIZE_16KiB; return ret; } From b8a26d46c0a4254f8bfe143681adb9d18d020299 Mon Sep 17 00:00:00 2001 From: Maxime Chevallier Date: Thu, 17 Sep 2026 23:53:37 +0200 Subject: [PATCH 414/694] net: stmmac: size the RX buffers from the frame length, not the MTU When picking the buffsize to use based on the MTU, we shouldn't check only the MTU value, but also : - ETH_HLEN for the L2 header, - up to 2 VLAN tags, - the FCS, The default bufsize is 1536 bytes, which is enough to contain all the above so this hasn't surfaced before, but the addition of NET_IP_ALIGN to the start of buffer address tripped the Jumbo selftest, leading to this discovery. With that, we don't need the '>=' checks on the buffer len, we can use more consistent comparison operators in stmmac_set_bfsize. Fixes: 286a83721720 ("stmmac: add CHAINED descriptor mode support (V4)") Reviewed-by: Nicolai Buchwitz Signed-off-by: Maxime Chevallier Link: https://patch.msgid.link/20260917215339.2022523-7-maxime.chevallier@bootlin.com Signed-off-by: Jakub Kicinski --- .../net/ethernet/stmicro/stmmac/stmmac_main.c | 20 ++++++++++--------- 1 file changed, 11 insertions(+), 9 deletions(-) diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c index 1fb5f804ea23..d5a984ad864f 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c @@ -1536,17 +1536,17 @@ static unsigned int stmmac_rx_offset(struct stmmac_priv *priv) return NET_SKB_PAD + NET_IP_ALIGN; } -static int stmmac_set_bfsize(int mtu) +static int stmmac_set_bfsize(int len) { int ret; - if (mtu >= BUF_SIZE_8KiB) + if (len > BUF_SIZE_8KiB) ret = BUF_SIZE_16KiB; - else if (mtu >= BUF_SIZE_4KiB) + else if (len > BUF_SIZE_4KiB) ret = BUF_SIZE_8KiB; - else if (mtu >= BUF_SIZE_2KiB) + else if (len > BUF_SIZE_2KiB) ret = BUF_SIZE_4KiB; - else if (mtu > DEFAULT_BUFSIZE) + else if (len > DEFAULT_BUFSIZE) ret = BUF_SIZE_2KiB; else ret = DEFAULT_BUFSIZE; @@ -4063,7 +4063,7 @@ static struct stmmac_dma_conf * stmmac_setup_dma_desc(struct stmmac_priv *priv, unsigned int mtu) { struct stmmac_dma_conf *dma_conf; - int bfsize, ret; + int bfsize, len, ret; u8 chan; dma_conf = kzalloc_obj(*dma_conf); @@ -4073,13 +4073,15 @@ stmmac_setup_dma_desc(struct stmmac_priv *priv, unsigned int mtu) return ERR_PTR(-ENOMEM); } - /* Returns 0 or BUF_SIZE_16KiB if mtu > 8KiB and dwmac4 or ring mode */ - bfsize = stmmac_set_16kib_bfsize(priv, mtu); + len = mtu + ETH_HLEN + 2 * VLAN_HLEN + ETH_FCS_LEN; + + /* Returns 0 or BUF_SIZE_16KiB if len > 8KiB and dwmac4 or ring mode */ + bfsize = stmmac_set_16kib_bfsize(priv, len); if (bfsize < 0) bfsize = 0; if (bfsize < BUF_SIZE_16KiB) - bfsize = stmmac_set_bfsize(mtu); + bfsize = stmmac_set_bfsize(len); dma_conf->dma_buf_sz = bfsize; /* Chose the tx/rx size from the already defined one in the From c4ac6e94eb9423126bda907a7f2933284f7450ff Mon Sep 17 00:00:00 2001 From: Maxime Chevallier Date: Thu, 17 Sep 2026 23:53:38 +0200 Subject: [PATCH 415/694] net: stmmac: selftests: Account for alignment shift on dwmac1000 for Jumbo test On dwmac1000, we currently only support single-descriptor frames. The Jumbo test started failing when NET_IP_ALIGN was added to align the IP header, as this tests tries to send the biggest possible frame. On dwmac1000 the DMA transfer is aligned on 4-bytes, so adding a 2-byte shift at the start-of-buffer address means it takes a whole extra 4-byte DMA burst to receive the Jumbo packet, causing it to spill over the next descriptor. This doesn't seem to happen on dwmac4 and xgmac that appear to correctly handle unaligned xfers (only tested on dwmac4) Let's account for that in the Jumbo test, reduce the size of our big packet by the align size. Fixes: 23680bf5f8c6 ("net: stmmac: restore NET_IP_ALIGN in the RX DMA offset") Reviewed-by: Nicolai Buchwitz Signed-off-by: Maxime Chevallier Link: https://patch.msgid.link/20260917215339.2022523-8-maxime.chevallier@bootlin.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c index 43b8411c5112..c25dc9f89270 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c @@ -1789,6 +1789,9 @@ static int __stmmac_test_jumbo(struct stmmac_priv *priv, u16 queue) struct stmmac_packet_attrs attr = { }; int size = priv->dma_conf.dma_buf_sz; + if (!dwmac_is_xmac(priv->plat->core_type)) + size -= NET_IP_ALIGN; + attr.dst = priv->dev->dev_addr; attr.max_size = size - ETH_FCS_LEN; attr.queue_mapping = queue; From 0160953d8eec75c3c55562158c46442ff1fd410b Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Bj=C3=B6rn=20T=C3=B6pel?= Date: Fri, 18 Sep 2026 13:46:40 +0200 Subject: [PATCH 416/694] eth: fbnic: Avoid rounding zero ring sizes MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit roundup_pow_of_two() is undefined for zero. ethtool permits a zero ring size to reach the driver, where the minimum-size check should reject it. Leave zero unchanged while rounding nonzero ring sizes. The minimum-size check then rejects zero deterministically without changing the established behavior for other values. Fixes: 6cbf18a05c06 ("eth: fbnic: support ring size configuration") Reported-by: Sashiko Link: https://lore.kernel.org/netdev/178971206933.22033.236948278674126701@kernel.org/ Suggested-by: Alexander Duyck Signed-off-by: Björn Töpel Reviewed-by: Joe Damato Link: https://patch.msgid.link/20260918114641.1281172-1-bjorn@kernel.org Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/meta/fbnic/fbnic_ethtool.c | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-) diff --git a/drivers/net/ethernet/meta/fbnic/fbnic_ethtool.c b/drivers/net/ethernet/meta/fbnic/fbnic_ethtool.c index 423f179c9d47..76e9a545bb16 100644 --- a/drivers/net/ethernet/meta/fbnic/fbnic_ethtool.c +++ b/drivers/net/ethernet/meta/fbnic/fbnic_ethtool.c @@ -313,6 +313,11 @@ fbnic_get_ringparam(struct net_device *netdev, struct ethtool_ringparam *ring, kernel_ring->hds_thresh = fbn->hds_thresh; } +static u32 fbnic_ring_size_pow2(u32 size) +{ + return size ? roundup_pow_of_two(size) : 0; +} + static void fbnic_set_rings(struct fbnic_net *fbn, struct ethtool_ringparam *ring, struct kernel_ethtool_ringparam *kernel_ring) @@ -334,10 +339,10 @@ fbnic_set_ringparam(struct net_device *netdev, struct ethtool_ringparam *ring, struct fbnic_net *clone; int err; - ring->rx_pending = roundup_pow_of_two(ring->rx_pending); - ring->rx_mini_pending = roundup_pow_of_two(ring->rx_mini_pending); - ring->rx_jumbo_pending = roundup_pow_of_two(ring->rx_jumbo_pending); - ring->tx_pending = roundup_pow_of_two(ring->tx_pending); + ring->rx_pending = fbnic_ring_size_pow2(ring->rx_pending); + ring->rx_mini_pending = fbnic_ring_size_pow2(ring->rx_mini_pending); + ring->rx_jumbo_pending = fbnic_ring_size_pow2(ring->rx_jumbo_pending); + ring->tx_pending = fbnic_ring_size_pow2(ring->tx_pending); /* These are absolute minimums allowing the device and driver to operate * but not necessarily guarantee reasonable performance. Settings below From 9c572a83037a7dcd653ba3a9cc468c16b857d0c9 Mon Sep 17 00:00:00 2001 From: Bernard Ladenthin Date: Fri, 18 Sep 2026 15:39:37 +0200 Subject: [PATCH 417/694] net/sched: fix potential stack infoleak in em_text_dump() em_text_dump() allocates struct tcf_em_text on the stack without zeroing it. strscpy() writes the algorithm name and a NUL terminator into conf.algo[], leaving the remaining bytes uninitialised. nla_put_nohdr() then copies the full struct to the netlink response. KMSAN on Linux 7.2-rc6 reports two kernel-infoleak splats from this path, one triggered via "tc filter show" and one via a raw RTM_GETTFILTER dump: BUG: KMSAN: kernel-infoleak in _copy_to_iter+0x1c9/0x2620 nla_put_nohdr+0x83/0x130 em_text_dump+0x291/0x550 Local variable conf created at: em_text_dump+0x5d/0x550 Bytes 168-179 of 199 are uninitialized I am not certain whether this constitutes a real security problem in practice: the test was conducted in a controlled KMSAN environment and the leaked stack bytes may or may not carry sensitive data on actual production kernels. I am reporting it because KMSAN flagged it as a kernel-infoleak and the fix is straightforward. I can provide a userspace reproducer on request. The original code used strncpy() which zero-pads to the destination size. Commit b04202d6065c ("net/sched: replace strncpy with strscpy") replaced it with strscpy(), which does not pad, creating this condition. Zero-initialising the struct closes it. Fixes: b04202d6065c ("net/sched: replace strncpy with strscpy") Link: https://lore.kernel.org/netdev/20250327143733.187438-1-richard120310@gmail.com/ Assisted-by: Claude:claude-sonnet-4-6 [KMSAN] Signed-off-by: Bernard Ladenthin Acked-by: Jamal Hadi Salim Link: https://patch.msgid.link/20260918133953.12494-1-bernard.ladenthin@gmail.com Signed-off-by: Jakub Kicinski --- net/sched/em_text.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/sched/em_text.c b/net/sched/em_text.c index 343f1aebeec2..4132f8c3c5fc 100644 --- a/net/sched/em_text.c +++ b/net/sched/em_text.c @@ -113,7 +113,7 @@ static void em_text_destroy(struct tcf_ematch *m) static int em_text_dump(struct sk_buff *skb, struct tcf_ematch *m) { struct text_match *tm = EM_TEXT_PRIV(m); - struct tcf_em_text conf; + struct tcf_em_text conf = {}; strscpy(conf.algo, tm->config->ops->name); conf.from_offset = tm->from_offset; From 6db1ce73e9853f533eb7f413f14ba00f8ec6f80d Mon Sep 17 00:00:00 2001 From: Weiming Shi Date: Sun, 20 Sep 2026 21:23:04 +0800 Subject: [PATCH 418/694] bpf: Reject dev-bound-only programs on other devices __bpf_offload_dev_match() falls back to comparing offdev pointers after an exact netdev mismatch. Bound-only programs normally have NULL offdevs, so unrelated netdevs compare equal. A bound-only program on an offload-registered netdev can instead inherit a real offdev and match a sibling port. With CAP_BPF and CAP_NET_ADMIN, a caller can use bpf(BPF_LINK_CREATE) with a different target ifindex to run metadata kfuncs specialized for the bound driver on the target driver's xdp_buff. Running a veth-bound program on tun reads beyond tun's bare stack xdp_buff as a veth_xdp_buff. Oops: general protection fault, probably for non-canonical address KASAN: null-ptr-deref in range [0x0000000000000010-0x0000000000000017] RIP: 0010:veth_xdp_rx_timestamp (drivers/net/veth.c:1673) Call Trace: ... tun_build_skb (drivers/net/tun.c:1739) tun_get_user (drivers/net/tun.c:1856) tun_chr_write_iter (drivers/net/tun.c:2091) vfs_write (fs/read_write.c:595 fs/read_write.c:687) ksys_write (fs/read_write.c:739) do_syscall_64 (arch/x86/entry/syscall_64.c:84) entry_SYSCALL_64_after_hwframe (arch/x86/entry/entry_64.S:121) Kernel panic - not syncing: Fatal exception in interrupt Restrict non-offloaded programs to exact netdev matches and retain the shared-offdev fallback only for genuinely offloaded multi-port programs. Fixes: 2b3486bc2d23 ("bpf: Introduce device-bound XDP programs") Reported-by: Signed-off-by: Weiming Shi Signed-off-by: Alexei Starovoitov Link: https://lore.kernel.org/bpf/20260917161335.1020405-2-bestswngs@gmail.com/ Link: https://patch.msgid.link/20260920132303.4109240-3-bestswngs@gmail.com --- kernel/bpf/offload.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/kernel/bpf/offload.c b/kernel/bpf/offload.c index 0d6f5569588c..d855399812ee 100644 --- a/kernel/bpf/offload.c +++ b/kernel/bpf/offload.c @@ -698,6 +698,8 @@ static bool __bpf_offload_dev_match(struct bpf_prog *prog, return false; if (offload->netdev == netdev) return true; + if (!bpf_prog_is_offloaded(prog->aux)) + return false; ondev1 = bpf_offload_find_netdev(offload->netdev); ondev2 = bpf_offload_find_netdev(netdev); From 76ebb69da677d2a4c5e59bf758b6424d511f5684 Mon Sep 17 00:00:00 2001 From: Christian Brauner Date: Thu, 17 Sep 2026 13:49:03 +0200 Subject: [PATCH 419/694] Revert "selftests/filesystems: add mntns cleanup test" This reverts commit 3452eecbcc954ef859b7d19309c121a78d6d0e31. The test checks that the mounts of a destroyed mount namespace stay connected. The commit that made them stay connected is reverted next because it leaks superblocks and loop devices, so drop the test with it. Link: https://patch.msgid.link/20260917-work-put_mnt_ns-revert-v1-1-34d9b8679e68@kernel.org Signed-off-by: Christian Brauner (Amutable) --- tools/testing/selftests/Makefile | 1 - .../filesystems/mntns_cleanup/.gitignore | 2 - .../filesystems/mntns_cleanup/Makefile | 6 -- .../mntns_cleanup/mntns_cleanup_test.c | 58 ------------------- 4 files changed, 67 deletions(-) delete mode 100644 tools/testing/selftests/filesystems/mntns_cleanup/.gitignore delete mode 100644 tools/testing/selftests/filesystems/mntns_cleanup/Makefile delete mode 100644 tools/testing/selftests/filesystems/mntns_cleanup/mntns_cleanup_test.c diff --git a/tools/testing/selftests/Makefile b/tools/testing/selftests/Makefile index 330d061f6366..273853937c25 100644 --- a/tools/testing/selftests/Makefile +++ b/tools/testing/selftests/Makefile @@ -49,7 +49,6 @@ TARGETS += filesystems/move_mount TARGETS += filesystems/empty_mntns TARGETS += filesystems/fsmount_ns TARGETS += filesystems/fscontext_ns -TARGETS += filesystems/mntns_cleanup TARGETS += filesystems/xattr TARGETS += firmware TARGETS += fpu diff --git a/tools/testing/selftests/filesystems/mntns_cleanup/.gitignore b/tools/testing/selftests/filesystems/mntns_cleanup/.gitignore deleted file mode 100644 index 493fbcf8d9ec..000000000000 --- a/tools/testing/selftests/filesystems/mntns_cleanup/.gitignore +++ /dev/null @@ -1,2 +0,0 @@ -# SPDX-License-Identifier: GPL-2.0-only -mntns_cleanup_test diff --git a/tools/testing/selftests/filesystems/mntns_cleanup/Makefile b/tools/testing/selftests/filesystems/mntns_cleanup/Makefile deleted file mode 100644 index 0e09e7030a5c..000000000000 --- a/tools/testing/selftests/filesystems/mntns_cleanup/Makefile +++ /dev/null @@ -1,6 +0,0 @@ -# SPDX-License-Identifier: GPL-2.0 -TEST_GEN_PROGS := mntns_cleanup_test - -CFLAGS += -Wall -O2 -g $(KHDR_INCLUDES) - -include ../../lib.mk diff --git a/tools/testing/selftests/filesystems/mntns_cleanup/mntns_cleanup_test.c b/tools/testing/selftests/filesystems/mntns_cleanup/mntns_cleanup_test.c deleted file mode 100644 index 5209712568b1..000000000000 --- a/tools/testing/selftests/filesystems/mntns_cleanup/mntns_cleanup_test.c +++ /dev/null @@ -1,58 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0 - -#define _GNU_SOURCE -#include -#include -#include -#include -#include -#include - -#include "../../kselftest_harness.h" - -FIXTURE(mntns_cleanup) { -}; - -FIXTURE_SETUP(mntns_cleanup) -{ - if (geteuid() != 0) - SKIP(return, "test requires CAP_SYS_ADMIN"); - - ASSERT_EQ(unshare(CLONE_NEWNS), 0); - ASSERT_EQ(mount("", "/", NULL, MS_REC | MS_PRIVATE, NULL), 0); - - rmdir("/mnt_dir"); - ASSERT_EQ(mkdir("/mnt_dir", 0755), 0); - ASSERT_EQ(mount("tmpfs", "/mnt_dir", "tmpfs", 0, NULL), 0); - ASSERT_EQ(mkdir("/mnt_dir/hidden", 0755), 0); - ASSERT_EQ(mkdir("/mnt_dir/hidden/secret", 0755), 0); - ASSERT_EQ(mount("tmpfs", "/mnt_dir/hidden", "tmpfs", 0, NULL), 0); -} - -FIXTURE_TEARDOWN(mntns_cleanup) -{ -} - -/* Mounts must stay connected when a mount namespace is cleaned up. */ -TEST_F(mntns_cleanup, keeps_mounts_connected) -{ - int fd, sfd, err; - - fd = open("/mnt_dir", O_PATH | O_DIRECTORY | O_CLOEXEC); - ASSERT_GE(fd, 0); - - /* Destroy the namespace; the fd keeps /mnt_dir alive. */ - ASSERT_EQ(unshare(CLONE_NEWNS), 0); - - sfd = openat(fd, "hidden/secret", O_RDONLY); - err = errno; - if (sfd >= 0) - close(sfd); - close(fd); - - ASSERT_LT(sfd, 0) - TH_LOG("mount namespace teardown revealed what the overmount covered"); - ASSERT_EQ(err, ENOENT); -} - -TEST_HARNESS_MAIN From 2e2142a35d809c3cc726d3b39e7aeee98d9ab71c Mon Sep 17 00:00:00 2001 From: Christian Brauner Date: Thu, 17 Sep 2026 13:49:04 +0200 Subject: [PATCH 420/694] Revert "put_mnt_ns(): leave mounts connected" This reverts commit 0342482a4d15358fe6931606caf58968de5d1d38. Keeping every mount of a dying namespace attached to its parent changes who owns it. An attached but unmounted mount is owned by its parent: the parent's final mntput() is what drops the child's reference and takes its superblock down. Before that commit only the namespace root was in that position and every other mount dropped its own reference in namespace_unlock(). That ownership rule turns any reference from a child's superblock back to one of its ancestors into a cycle. The obvious case is a loop device: unshare -m sh -c 'mount -o loop /var/tmp/img /var/tmp/mp' mount(8) opens the image from inside the new namespace, so the loop device's backing file pins that namespace's copy of the root mount. When the namespace dies the loop mount stays attached to that copy and is owned by it. The copy can't reach zero because the backing file holds it. The backing file is only put once the ext4 superblock releases the block device and autoclear runs, and that needs the loop mount to go first. Nothing breaks the cycle. The superblock, the loop device and, since the root copy keeps all of its children. systemd-sysext sets up its loop devices in a private mount namespace and relies on autoclear when that namespace goes away, which is how this was found. Anything that holds a file on an ancestor from a mounted filesystem has the same problem. That includes ecryptfs lower paths, erofs file-backed mounts, fuse passthrough backing files. That's a bigger fix and not an -rc change. Revert. Fixes: 0342482a4d15 ("put_mnt_ns(): leave mounts connected") Reported-by: Michael Vogt Link: https://gist.github.com/mvo5/63ef46482349f3b1c3957d463a0c9c6f Link: https://patch.msgid.link/20260917-work-put_mnt_ns-revert-v1-2-34d9b8679e68@kernel.org Signed-off-by: Christian Brauner (Amutable) --- fs/namespace.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/namespace.c b/fs/namespace.c index 5e41021eaa63..580877e46b1a 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -6301,7 +6301,7 @@ void put_mnt_ns(struct mnt_namespace *ns) guard(namespace_excl)(); emptied_ns = ns; guard(mount_writer)(); - umount_tree(ns->root, UMOUNT_CONNECTED); + umount_tree(ns->root, 0); } struct vfsmount *kern_mount(struct file_system_type *type) From 1feb5d39b05afd902ed9fc902ec5b15be03a4bdb Mon Sep 17 00:00:00 2001 From: Bartosz Golaszewski Date: Tue, 22 Sep 2026 10:52:28 +0200 Subject: [PATCH 421/694] gpio: cdev: fix kernel stack leak to user-space in error path If we fail to acquire the GPIO chip guard in gpio_desc_to_lineinfo(), we return immediately before zeroing the info struct we'll end up passing to the user-space later in lineinfo_get_v1(). This may leak the kernel stack contents. Make gpio_desc_to_lineinfo() return int so that the -ENODEV returned on failure to acquire the guard can be propagated to the callers. While not strictly necessary: move the memset() before trying to acquire the SRCU read lock too for good measure. Fixes: d83cee3d2bb1 ("gpio: protect the pointer to gpio_chip in gpio_device with SRCU") Cc: stable@vger.kernel.org Reported-by: Sashiko Closes: https://sashiko.dev/#/patchset/20260912123529.7951-1-tzungbi%40kernel.org?part=3 Reviewed-by: Kent Gibson Link: https://patch.msgid.link/20260922-gpio-cdev-stack-leak-fixes-v3-1-7a0c7a4299d5@oss.qualcomm.com Signed-off-by: Bartosz Golaszewski --- drivers/gpio/gpiolib-cdev.c | 32 +++++++++++++++++++++++++------- 1 file changed, 25 insertions(+), 7 deletions(-) diff --git a/drivers/gpio/gpiolib-cdev.c b/drivers/gpio/gpiolib-cdev.c index d1105b7ae437..5d53bfcdf726 100644 --- a/drivers/gpio/gpiolib-cdev.c +++ b/drivers/gpio/gpiolib-cdev.c @@ -2172,18 +2172,19 @@ static void gpio_v2_line_info_changed_to_v1( #endif /* CONFIG_GPIO_CDEV_V1 */ -static void gpio_desc_to_lineinfo(struct gpio_desc *desc, - struct gpio_v2_line_info *info, bool atomic) +static int gpio_desc_to_lineinfo(struct gpio_desc *desc, + struct gpio_v2_line_info *info, bool atomic) { u32 debounce_period_us; unsigned long dflags; const char *label; + memset(info, 0, sizeof(*info)); + CLASS(gpio_chip_guard, guard)(desc); if (!guard.gc) - return; + return -ENODEV; - memset(info, 0, sizeof(*info)); info->offset = gpiod_hwgpio(desc); if (desc->name) @@ -2258,6 +2259,8 @@ static void gpio_desc_to_lineinfo(struct gpio_desc *desc, debounce_period_us; info->num_attrs++; } + + return 0; } struct gpio_chardev_data { @@ -2309,6 +2312,7 @@ static int lineinfo_get_v1(struct gpio_chardev_data *cdev, void __user *ip, struct gpio_desc *desc; struct gpioline_info lineinfo; struct gpio_v2_line_info lineinfo_v2; + int ret; if (copy_from_user(&lineinfo, ip, sizeof(lineinfo))) return -EFAULT; @@ -2326,7 +2330,10 @@ static int lineinfo_get_v1(struct gpio_chardev_data *cdev, void __user *ip, return -EBUSY; } - gpio_desc_to_lineinfo(desc, &lineinfo_v2, false); + ret = gpio_desc_to_lineinfo(desc, &lineinfo_v2, false); + if (ret) + return ret; + gpio_v2_line_info_to_v1(&lineinfo_v2, &lineinfo); if (copy_to_user(ip, &lineinfo, sizeof(lineinfo))) { @@ -2344,6 +2351,7 @@ static int lineinfo_get(struct gpio_chardev_data *cdev, void __user *ip, { struct gpio_desc *desc; struct gpio_v2_line_info lineinfo; + int ret; if (copy_from_user(&lineinfo, ip, sizeof(lineinfo))) return -EFAULT; @@ -2363,7 +2371,10 @@ static int lineinfo_get(struct gpio_chardev_data *cdev, void __user *ip, if (test_and_set_bit(lineinfo.offset, cdev->watched_lines)) return -EBUSY; } - gpio_desc_to_lineinfo(desc, &lineinfo, false); + + ret = gpio_desc_to_lineinfo(desc, &lineinfo, false); + if (ret) + return ret; if (copy_to_user(ip, &lineinfo, sizeof(lineinfo))) { if (watch) @@ -2489,6 +2500,7 @@ static int lineinfo_changed_notify(struct notifier_block *nb, struct lineinfo_changed_ctx *ctx; struct gpio_desc *desc = data; struct file *fp; + int ret; if (!test_bit(gpiod_hwgpio(desc), cdev->watched_lines)) return NOTIFY_DONE; @@ -2519,7 +2531,13 @@ static int lineinfo_changed_notify(struct notifier_block *nb, ctx->chg.event_type = action; ctx->chg.timestamp_ns = ktime_get_ns(); - gpio_desc_to_lineinfo(desc, &ctx->chg.info, true); + + ret = gpio_desc_to_lineinfo(desc, &ctx->chg.info, true); + if (ret) { + fput(fp); + return NOTIFY_DONE; + } + /* Keep the GPIO device alive until we emit the event. */ ctx->gdev = gpio_device_get(desc->gdev); ctx->cdev = cdev; From 2d2a2d7aa98741b58f54cacc99b52024e4d865f9 Mon Sep 17 00:00:00 2001 From: Christian Brauner Date: Wed, 9 Sep 2026 21:30:34 +0200 Subject: [PATCH 422/694] super: make iterate_supers_type() deletion-safe iterate_supers_type() drops sb_lock while invoking the callback and keeps only a passive reference to the current superblock. That reference keeps the object allocated, but does not keep its s_instances node linked. After the iterator releases s_umount, final teardown can unlink the current s_instances node. The iterator then advances through a reinitialized node. With the current hlist it stops without visiting the remaining superblocks. The unlink moved from generic_shutdown_super() to kill_super_notify(), but the cursor lifetime has been unsafe since the helper was introduced. The CIFS DFS lookup can consequently miss a matching superblock and return -EINVAL. Move removal from fs_supers to put_super(), alongside removal from super_blocks, so a passive reference keeps both list nodes linked. Keep the filesystem module reference until then, since unlinking s_instances may touch type->fs_supers. Make sget_fc() skip SB_DEAD superblocks before invoking test(), and set SB_DEAD under sb_lock to serialize with those callbacks. This allows kernfs to free its private information after kill_anon_super() returns. Keep matching SB_DYING superblocks until SB_DEAD is set so concurrent mounts still wait for teardown before retrying. Fixes: 43e15cdbefea ("new helper: iterate_supers_type()") Reported-by: Karl Mehltretter Closes: https://lore.kernel.org/r/20260903013336.92081-1-kmehltretter@gmail.com Suggested-by: Jan Kara Cc: stable@vger.kernel.org Tested-by: Karl Mehltretter [kmehltretter: supplied the commit message] Signed-off-by: Karl Mehltretter Link: https://patch.msgid.link/20260909193034.7467-1-kmehltretter@gmail.com Reviewed-by: Jan Kara Signed-off-by: Christian Brauner (Amutable) --- fs/kernfs/mount.c | 4 ++-- fs/super.c | 39 +++++++++++++++++++-------------------- 2 files changed, 21 insertions(+), 22 deletions(-) diff --git a/fs/kernfs/mount.c b/fs/kernfs/mount.c index f183a96778b9..a57399021c8b 100644 --- a/fs/kernfs/mount.c +++ b/fs/kernfs/mount.c @@ -434,8 +434,8 @@ void kernfs_kill_sb(struct super_block *sb) up_write(&root->kernfs_supers_rwsem); /* - * Remove the superblock from fs_supers/s_instances - * so we can't find it, before freeing kernfs_super_info. + * Mark the superblock dead so sget_fc() can't find it, + * before freeing kernfs_super_info. */ kill_anon_super(sb); kfree(info); diff --git a/fs/super.c b/fs/super.c index 01db6124e409..e41cae9e5c62 100644 --- a/fs/super.c +++ b/fs/super.c @@ -433,15 +433,19 @@ static struct super_block *alloc_super(struct file_system_type *type, int flags, void put_super(struct super_block *s) { if (refcount_dec_and_test(&s->s_passive)) { + struct file_system_type *type = s->s_type; spin_lock(&sb_lock); list_del_init(&s->s_list); + hlist_del_init(&s->s_instances); spin_unlock(&sb_lock); WARN_ON(s->s_dentry_lru.node); WARN_ON(s->s_inode_lru.node); WARN_ON(s->s_mounts); call_rcu(&s->rcu, destroy_super_rcu); + /* The unlink above may touch type->fs_supers, so drop it last. */ + put_filesystem(type); } } @@ -558,17 +562,6 @@ static void kill_super_notify(struct super_block *sb) if (sb->s_flags & SB_DEAD) return; - /* - * Remove it from @fs_supers so it isn't found by new - * sget_fc() walkers anymore. Any concurrent mounter still - * managing to grab a temporary reference is guaranteed to - * already see SB_DYING and will wait until we notify them about - * SB_DEAD. - */ - spin_lock(&sb_lock); - hlist_del_init(&sb->s_instances); - spin_unlock(&sb_lock); - /* Drop sget_fc()'s claim; a never-registered entry stays with the sb. */ if (sb->s_super_dev->sd_dev) { super_dev_put(sb->s_super_dev); @@ -577,11 +570,15 @@ static void kill_super_notify(struct super_block *sb) /* * Let concurrent mounts know that this thing is really dead. - * We don't need @sb->s_umount here as every concurrent caller - * will see SB_DYING and either discard the superblock or wait - * for SB_DEAD. + * sget_fc() skips SB_DEAD superblocks and calls test() under + * sb_lock, so set it under sb_lock: once we return no test() + * runs on this superblock anymore and none will start. Everyone + * else already saw SB_DYING and either discarded the superblock + * or waits for SB_DEAD. */ + spin_lock(&sb_lock); super_wake(sb, SB_DEAD); + spin_unlock(&sb_lock); } /** @@ -608,7 +605,6 @@ void deactivate_locked_super(struct super_block *s) list_lru_destroy(&s->s_dentry_lru); list_lru_destroy(&s->s_inode_lru); - put_filesystem(fs); put_super(s); } else { super_unlock_excl(s); @@ -795,12 +791,12 @@ void generic_shutdown_super(struct super_block *sb) } /* * Broadcast to everyone that grabbed a temporary reference to this - * superblock before we removed it from @fs_supers that the superblock - * is dying. Every walker of @fs_supers outside of sget_fc() will now - * discard this superblock and treat it as dead. + * superblock that it is dying. Every walker of @fs_supers outside + * of sget_fc() will now discard this superblock and treat it as + * dead. * - * We leave the superblock on @fs_supers so it can be found by - * sget_fc() until we passed sb->kill_sb(). + * sget_fc() keeps finding the superblock until SB_DEAD is set, so + * a concurrent mounter waits until we passed sb->kill_sb(). */ super_wake(sb, SB_DYING); super_unlock_excl(sb); @@ -879,6 +875,9 @@ struct super_block *sget_fc(struct fs_context *fc, spin_lock(&sb_lock); if (test) { hlist_for_each_entry(old, &fc->fs_type->fs_supers, s_instances) { + /* Only unlinked at the last passive reference. */ + if (super_flags(old, SB_DEAD)) + continue; if (test(old, fc)) goto share_extant_sb; } From e9438ab5328a177c9c0e5df87eb92a7162841e98 Mon Sep 17 00:00:00 2001 From: Ridham Khurana Date: Tue, 22 Sep 2026 09:20:58 +0000 Subject: [PATCH 423/694] gpio: zynq: fix runtime PM leak on request error path pm_runtime_get_sync() leaves the usage counter incremented even when it fails, and zynq_gpio_request() returns the error without dropping it. gpiolib does not call ->free() when ->request() fails, so zynq_gpio_free(), which holds the only matching pm_runtime_put(), never runs. The reference is leaked and the controller can no longer runtime-suspend, so its clock stays enabled. Switch to pm_runtime_resume_and_get(), which only increments the usage counter on success. Fixes: 3242ba117e9b ("gpio: Add driver for Zynq GPIO controller") Cc: stable@vger.kernel.org Signed-off-by: Ridham Khurana Link: https://patch.msgid.link/20260922092102.1053513-1-khurana.ridham222@gmail.com Signed-off-by: Bartosz Golaszewski --- drivers/gpio/gpio-zynq.c | 10 +--------- 1 file changed, 1 insertion(+), 9 deletions(-) diff --git a/drivers/gpio/gpio-zynq.c b/drivers/gpio/gpio-zynq.c index 15a79d9a2e9e..13e4c5f0e297 100644 --- a/drivers/gpio/gpio-zynq.c +++ b/drivers/gpio/gpio-zynq.c @@ -798,15 +798,7 @@ static int zynq_gpio_runtime_resume(struct device *dev) static int zynq_gpio_request(struct gpio_chip *chip, unsigned int offset) { - int ret; - - ret = pm_runtime_get_sync(chip->parent); - - /* - * If the device is already active pm_runtime_get() will return 1 on - * success, but gpio_request still needs to return 0. - */ - return ret < 0 ? ret : 0; + return pm_runtime_resume_and_get(chip->parent); } static void zynq_gpio_free(struct gpio_chip *chip, unsigned int offset) From cca4980630b3c7a85f53cb43c6018184ce5d4e37 Mon Sep 17 00:00:00 2001 From: Namhyung Kim Date: Sun, 20 Sep 2026 16:16:39 -0700 Subject: [PATCH 424/694] perf/core: Fix a refcount leak in attach_perf_ctx_data() The attach_perf_ctx_data() can race on global and !global cases. The global case is protected by global_ctx_data_rwsem and shares a single reference count using perf_ctx_data.global field. But when it races with !global case, it may miss to set the global field and result in a reference count leak. CPU1 CPU2 ---------------------------------------------------------------- attach_task_ctx_data(.global=1) attach_task_ctx_data(.global=0) cd1 = alloc_perf_ctx_data(); cd2 = alloc_perf_ctx_data(); // { .global = 0, .refcount = 1 }; try_cmpxchg(); // success, // task->perf_ctx_data = cd2 try_cmpxhg(); // fail; old = cd2 refcount_inc_not_zero(&old->refcount); // success // old.refcount = 2 free_perf_ctx_data(cd1); Then later detach_global_ctx_data() will see the data but it's not marked as global, so it won't call detach_task_ctx_data(). Fixes: 506e64e710ff ("perf: attach/detach PMU specific data") Assisted-by: Sashiko.dev:Gemini-3.1-pro Signed-off-by: Namhyung Kim Signed-off-by: Peter Zijlstra (Intel) Link: https://patch.msgid.link/20260920231639.11910-1-namhyung@kernel.org --- kernel/events/core.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/kernel/events/core.c b/kernel/events/core.c index db7b76d6b68a..e180134bad5e 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -5454,6 +5454,8 @@ attach_task_ctx_data(struct task_struct *task, struct kmem_cache *ctx_cache, } if (refcount_inc_not_zero(&old->refcount)) { + if (global) + old->global = true; free_perf_ctx_data(cd); /* unused */ return 0; } From 36bb85cf36cab15fb611cb44b78a5df06e4e69a2 Mon Sep 17 00:00:00 2001 From: Puranjay Mohan Date: Mon, 10 Aug 2026 06:35:34 -0700 Subject: [PATCH 425/694] perf/core: Fix NULL pmu_ctx passed to pmu->sched_task() perf_pmu_sched_task() returns early when cpuctx->task_ctx is set, and cpc->task_epc is only non-NULL while a task context is scheduled in on this CPU. __perf_pmu_sched_task() therefore always passes NULL: Unable to handle kernel NULL pointer dereference at virtual address 00 pc : armv8pmu_sched_task+0x14/0x50 Call trace: armv8pmu_sched_task+0x14/0x50 (P) perf_pmu_sched_task+0xac/0x108 __perf_event_task_sched_out+0x6c/0xe0 Pass &cpc->epc instead, the CPU-wide context for this PMU, which the function already dereferences a few lines up to find pmu. armv8pmu_sched_task() is the only in-tree implementation that dereferences the argument, and it only reads ->pmu, so the oops needs BRBE, added in v6.17. Fixes: bd2756811766 ("perf: Rewrite core context handling") Signed-off-by: Puranjay Mohan Signed-off-by: Peter Zijlstra (Intel) Tested-by: Yifan Wu Cc: stable@vger.kernel.org Link: https://patch.msgid.link/20260810133540.1947118-2-puranjay@kernel.org --- kernel/events/core.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/kernel/events/core.c b/kernel/events/core.c index e180134bad5e..7ce72f366f6c 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -3914,7 +3914,7 @@ static void __perf_pmu_sched_task(struct perf_cpu_pmu_context *cpc, perf_ctx_lock(cpuctx, cpuctx->task_ctx); perf_pmu_disable(pmu); - pmu->sched_task(cpc->task_epc, task, sched_in); + pmu->sched_task(&cpc->epc, task, sched_in); perf_pmu_enable(pmu); perf_ctx_unlock(cpuctx, cpuctx->task_ctx); From 3d8d74100954a3b17e5c5e37adfe14e16b1db103 Mon Sep 17 00:00:00 2001 From: Puranjay Mohan Date: Mon, 10 Aug 2026 06:35:35 -0700 Subject: [PATCH 426/694] perf/core: Run sched_task() for PMUs with only CPU-wide events perf_pmu_sched_task() returns early when cpuctx->task_ctx is set and leaves the work to perf_ctx_sched_task_cb(), which only walks ctx->pmu_ctx_list. A PMU whose events are all CPU-wide is not on that list, so nothing calls its sched_task(). With perf record -b -e cycles -a -- ls armv8pmu_sched_task() is skipped on every switch to a task that has a perf context but no event on that PMU, and BRBE records leak across the task boundary. intel_pmu_lbr_add() calls perf_sched_cb_inc() unconditionally too, so LBR records leak the same way on x86. Drop the early return and skip only the CPCs that perf_ctx_sched_task_cb() handles. That one needs a gate of its own to make the split exact: it tests cpc->sched_cb_usage, which perf_sched_cb_inc() sets per CPU for every branch stack user, so a task with an event for that PMU pinned to another CPU would be handled twice. On x86 the second __intel_pmu_lbr_restore() finds lbr_stack_state == LBR_NONE and calls intel_pmu_lbr_reset(), throwing away the callstack the first one restored. cpc->task_epc is set only while a task context is scheduled in, and there is one epc per PMU on ctx->pmu_ctx_list, so the two gates are inverses. For the CPCs perf_pmu_sched_task() picks up, the callback now runs outside the perf_ctx_disable() and perf_ctx_enable() pair in perf_event_context_sched_in(). __perf_pmu_sched_task() disables the PMU around the call itself. Fixes: bd2756811766 ("perf: Rewrite core context handling") Signed-off-by: Puranjay Mohan Signed-off-by: Peter Zijlstra (Intel) Tested-by: Yifan Wu Link: https://patch.msgid.link/20260810133540.1947118-3-puranjay@kernel.org Cc: stable@vger.kernel.org --- kernel/events/core.c | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-) diff --git a/kernel/events/core.c b/kernel/events/core.c index 7ce72f366f6c..634d2ccbab82 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -3764,6 +3764,9 @@ static void perf_ctx_sched_task_cb(struct perf_event_context *ctx, list_for_each_entry(pmu_ctx, &ctx->pmu_ctx_list, pmu_ctx_entry) { cpc = this_cpc(pmu_ctx->pmu); + if (cpc->task_epc != pmu_ctx) + continue; + if (cpc->sched_cb_usage && pmu_ctx->pmu->sched_task) pmu_ctx->pmu->sched_task(pmu_ctx, task, sched_in); } @@ -3924,15 +3927,17 @@ static void perf_pmu_sched_task(struct task_struct *prev, struct task_struct *next, bool sched_in) { - struct perf_cpu_context *cpuctx = this_cpu_ptr(&perf_cpu_context); struct perf_cpu_pmu_context *cpc, *cpc2; - /* cpuctx->task_ctx will be handled in perf_event_context_sched_in/out */ - if (prev == next || cpuctx->task_ctx) + if (prev == next) return; - list_for_each_entry_safe(cpc, cpc2, this_cpu_ptr(&sched_cb_list), sched_cb_entry) + list_for_each_entry_safe(cpc, cpc2, this_cpu_ptr(&sched_cb_list), sched_cb_entry) { + if (cpc->task_epc) + continue; + __perf_pmu_sched_task(cpc, sched_in ? next : prev, sched_in); + } } static void perf_event_switch(struct task_struct *task, From 24b620729e53d978b3e425f55bc66efd3bab1f59 Mon Sep 17 00:00:00 2001 From: Puranjay Mohan Date: Mon, 10 Aug 2026 06:35:36 -0700 Subject: [PATCH 427/694] perf/core: Fill branch entries with a single assignment perf_clear_branch_entry_bitfields() clears the bitfields of struct perf_branch_entry one by one and leaves from/to alone, since callers overwrite those straight away. The list has to be kept in sync with the struct by hand and has already fallen behind: new_type and priv were added to perf_branch_entry and never added here. Only BRBE writes those two, and neither for every record. brbe_set_perf_entry_type() leaves new_type alone for a branch type it does not recognise, and priv is not set for source-only records. arm_pmuv3.c allocates the per-CPU branch stack with kmalloc(), so such a record reaches userspace with whatever the slot held: uninitialised kmalloc() data on the first pass over the buffer, the previous record's values after that. Nothing under arch/x86/events/ writes either field, so only arm64 is affected. Assign the whole entry at each site instead. Everything not named is then zero, and there is no list to keep in sync. The bitfields add up to exactly 64 bits, so the struct has no padding to leave undefined. perf_clear_branch_entry_bitfields() has no callers left, so remove it. perf_entry_from_brbe_regset() assigns an empty literal instead, since it fills from/to conditionally. PERF_BR_SPEC_NA is 0, so dropping the explicit spec assignment changes nothing. Fixes: b190bc4ac9e6 ("perf: Extend branch type classification") Fixes: 5402d25aa571 ("perf: Capture branch privilege information") Suggested-by: Peter Zijlstra Signed-off-by: Puranjay Mohan Signed-off-by: Peter Zijlstra (Intel) Tested-by: Yifan Wu Link: https://patch.msgid.link/20260810133540.1947118-4-puranjay@kernel.org --- arch/x86/events/amd/brs.c | 9 +++--- arch/x86/events/amd/lbr.c | 16 +++++----- arch/x86/events/intel/lbr.c | 63 ++++++++++++++++++++----------------- drivers/perf/arm_brbe.c | 2 +- include/linux/perf_event.h | 17 ---------- 5 files changed, 47 insertions(+), 60 deletions(-) diff --git a/arch/x86/events/amd/brs.c b/arch/x86/events/amd/brs.c index dc564688f3d7..54b13faba116 100644 --- a/arch/x86/events/amd/brs.c +++ b/arch/x86/events/amd/brs.c @@ -343,11 +343,10 @@ void amd_brs_drain(void) if (!amd_brs_match_plm(event, from, to)) continue; - perf_clear_branch_entry_bitfields(br+nr); - - br[nr].from = from; - br[nr].to = to; - + br[nr] = (struct perf_branch_entry){ + .from = from, + .to = to, + }; nr++; } empty: diff --git a/arch/x86/events/amd/lbr.c b/arch/x86/events/amd/lbr.c index 9d9c961989d5..a55646fcb846 100644 --- a/arch/x86/events/amd/lbr.c +++ b/arch/x86/events/amd/lbr.c @@ -184,13 +184,6 @@ void amd_pmu_lbr_read(void) entry.to.split.reserved) continue; - perf_clear_branch_entry_bitfields(br + out); - - br[out].from = sign_ext_branch_ip(entry.from.split.ip); - br[out].to = sign_ext_branch_ip(entry.to.split.ip); - br[out].mispred = entry.from.split.mispredict; - br[out].predicted = !br[out].mispred; - /* * Set branch speculation information using the status of * the valid and spec bits. @@ -208,7 +201,14 @@ void amd_pmu_lbr_read(void) * speculative and took the correct path */ idx = (entry.to.split.valid << 1) | entry.to.split.spec; - br[out].spec = lbr_spec_map[idx]; + + br[out] = (struct perf_branch_entry){ + .from = sign_ext_branch_ip(entry.from.split.ip), + .to = sign_ext_branch_ip(entry.to.split.ip), + .mispred = entry.from.split.mispredict, + .predicted = !entry.from.split.mispredict, + .spec = lbr_spec_map[idx], + }; out++; } diff --git a/arch/x86/events/intel/lbr.c b/arch/x86/events/intel/lbr.c index cbe5c762008d..22e2a06d5786 100644 --- a/arch/x86/events/intel/lbr.c +++ b/arch/x86/events/intel/lbr.c @@ -756,10 +756,10 @@ void intel_pmu_lbr_read_32(struct cpu_hw_events *cpuc) rdmsrq(x86_pmu.lbr_from + lbr_idx, msr_lastbranch.lbr); - perf_clear_branch_entry_bitfields(br); - - br->from = msr_lastbranch.from; - br->to = msr_lastbranch.to; + *br = (struct perf_branch_entry){ + .from = msr_lastbranch.from, + .to = msr_lastbranch.to, + }; br++; } cpuc->lbr_stack.nr = i; @@ -847,14 +847,15 @@ void intel_pmu_lbr_read_64(struct cpu_hw_events *cpuc) if (abort && x86_pmu.lbr_double_abort && out > 0) out--; - perf_clear_branch_entry_bitfields(br+out); - br[out].from = from; - br[out].to = to; - br[out].mispred = mis; - br[out].predicted = pred; - br[out].in_tx = in_tx; - br[out].abort = abort; - br[out].cycles = cycles; + br[out] = (struct perf_branch_entry){ + .from = from, + .to = to, + .mispred = mis, + .predicted = pred, + .in_tx = in_tx, + .abort = abort, + .cycles = cycles, + }; out++; } cpuc->lbr_stack.nr = out; @@ -905,6 +906,7 @@ static void intel_pmu_store_lbr(struct cpu_hw_events *cpuc, struct perf_branch_entry *e; struct lbr_entry *lbr; u64 from, to, info; + bool mispred; int i; for (i = 0; i < x86_pmu.lbr_nr; i++) { @@ -921,24 +923,27 @@ static void intel_pmu_store_lbr(struct cpu_hw_events *cpuc, to = rdlbr_to(i, lbr); info = rdlbr_info(i, lbr); - perf_clear_branch_entry_bitfields(e); + mispred = get_lbr_mispred(info); - e->from = from; - e->to = to; - e->mispred = get_lbr_mispred(info); - e->predicted = !e->mispred; - e->in_tx = !!(info & LBR_INFO_IN_TX); - e->abort = !!(info & LBR_INFO_ABORT); - e->cycles = get_lbr_cycles(info); - e->type = get_lbr_br_type(info); - - /* - * Leverage the reserved field of cpuc->lbr_entries[i] to - * temporarily store the branch counters information. - * The later code will decide what content can be disclosed - * to the perf tool. Pleae see intel_pmu_lbr_counters_reorder(). - */ - e->reserved = (info >> LBR_INFO_BR_CNTR_OFFSET) & LBR_INFO_BR_CNTR_FULL_MASK; + *e = (struct perf_branch_entry){ + .from = from, + .to = to, + .mispred = mispred, + .predicted = !mispred, + .in_tx = !!(info & LBR_INFO_IN_TX), + .abort = !!(info & LBR_INFO_ABORT), + .cycles = get_lbr_cycles(info), + .type = get_lbr_br_type(info), + /* + * Leverage the reserved field of + * cpuc->lbr_entries[i] to temporarily store the + * branch counters information. The later code will + * decide what content can be disclosed to the perf + * tool. Pleae see intel_pmu_lbr_counters_reorder(). + */ + .reserved = (info >> LBR_INFO_BR_CNTR_OFFSET) & + LBR_INFO_BR_CNTR_FULL_MASK, + }; } cpuc->lbr_stack.nr = i; diff --git a/drivers/perf/arm_brbe.c b/drivers/perf/arm_brbe.c index ba554e0c846c..254be4da8ae2 100644 --- a/drivers/perf/arm_brbe.c +++ b/drivers/perf/arm_brbe.c @@ -604,7 +604,7 @@ static bool perf_entry_from_brbe_regset(int index, struct perf_branch_entry *ent return false; brbinf = bregs.brbinf; - perf_clear_branch_entry_bitfields(entry); + *entry = (struct perf_branch_entry){ }; if (brbe_record_is_complete(brbinf)) { entry->from = bregs.brbsrc; entry->to = bregs.brbtgt; diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h index 5842552294c1..915c6fd3f084 100644 --- a/include/linux/perf_event.h +++ b/include/linux/perf_event.h @@ -1467,23 +1467,6 @@ static inline u32 perf_sample_data_size(struct perf_sample_data *data, return size; } -/* - * Clear all bitfields in the perf_branch_entry. - * The to and from fields are not cleared because they are - * systematically modified by caller. - */ -static inline void perf_clear_branch_entry_bitfields(struct perf_branch_entry *br) -{ - br->mispred = 0; - br->predicted = 0; - br->in_tx = 0; - br->abort = 0; - br->cycles = 0; - br->type = 0; - br->spec = PERF_BR_SPEC_NA; - br->reserved = 0; -} - extern void perf_output_sample(struct perf_output_handle *handle, struct perf_event_header *header, struct perf_sample_data *data, From 2bc6b218717b9d08f466f88209251d54bc09b207 Mon Sep 17 00:00:00 2001 From: Fuad Tabba Date: Tue, 22 Sep 2026 19:14:30 +0100 Subject: [PATCH 428/694] arm64/boot: Disable trapping of PMZR_EL0 writes to EL2 __init_el2_fgt2() writes one mask to both HDFGRTR2_EL2 and HDFGWTR2_EL2. PMZR_EL0 is write-only, so its trap bit, nPMZR_EL0, exists only in HDFGWTR2_EL2 and is therefore never set: a PMZR_EL0 write from the host traps to EL2, where the nVHE hypervisor has no handler and BUG()s. The kernel never writes PMZR_EL0, but kernel.perf_user_access=1 has the PMU driver set PMUSERENR_EL0.UEN for a task with a user-read event, so a write from EL0 reaches the trap and takes the host down without a panic message. Accumulate the HDFGWTR2_EL2 bits separately, as __init_el2_fgt() already does for HDFGWTR_EL2, and set nPMZR_EL0 with the other FEAT_PMUv3p9 bits. Fixes: 858c7bfcb35e1 ("arm64/boot: Enable EL2 requirements for FEAT_PMUv3p9") Cc: stable@vger.kernel.org Signed-off-by: Fuad Tabba Reviewed-by: Anshuman Khandual Reviewed-by: Oliver Upton Signed-off-by: Will Deacon --- Documentation/arch/arm64/booting.rst | 1 + arch/arm64/include/asm/el2_setup.h | 9 ++++++++- 2 files changed, 9 insertions(+), 1 deletion(-) diff --git a/Documentation/arch/arm64/booting.rst b/Documentation/arch/arm64/booting.rst index 13ef311dace8..3fea4b14ef7c 100644 --- a/Documentation/arch/arm64/booting.rst +++ b/Documentation/arch/arm64/booting.rst @@ -465,6 +465,7 @@ Before jumping into the kernel, the following conditions must be met: - HDFGWTR2_EL2.nPMICNTR_EL0 (bit 2) must be initialised to 0b1. - HDFGWTR2_EL2.nPMICFILTR_EL0 (bit 3) must be initialised to 0b1. - HDFGWTR2_EL2.nPMUACR_EL1 (bit 4) must be initialised to 0b1. + - HDFGWTR2_EL2.nPMZR_EL0 (bit 21) must be initialised to 0b1. For CPUs with SPE data source filtering (FEAT_SPE_FDS): diff --git a/arch/arm64/include/asm/el2_setup.h b/arch/arm64/include/asm/el2_setup.h index aa8ec9df8024..87560d8b254e 100644 --- a/arch/arm64/include/asm/el2_setup.h +++ b/arch/arm64/include/asm/el2_setup.h @@ -418,6 +418,7 @@ b.lt .Lskip_fgt2_\@ mov x0, xzr + mov x2, xzr mrs x1, id_aa64dfr0_el1 ubfx x1, x1, #ID_AA64DFR0_EL1_PMUVer_SHIFT, #4 cmp x1, #ID_AA64DFR0_EL1_PMUVer_V3P9 @@ -426,6 +427,11 @@ orr x0, x0, #HDFGRTR2_EL2_nPMICNTR_EL0 orr x0, x0, #HDFGRTR2_EL2_nPMICFILTR_EL0 orr x0, x0, #HDFGRTR2_EL2_nPMUACR_EL1 + orr x2, x2, #HDFGWTR2_EL2_nPMICNTR_EL0 + orr x2, x2, #HDFGWTR2_EL2_nPMICFILTR_EL0 + orr x2, x2, #HDFGWTR2_EL2_nPMUACR_EL1 + /* PMZR_EL0 is write-only, so it has no read trap to disable */ + orr x2, x2, #HDFGWTR2_EL2_nPMZR_EL0 .Lskip_pmuv3p9_\@: /* If SPE is implemented, */ __spe_vers_imp .Lskip_spefds_\@, ID_AA64DFR0_EL1_PMSVer_IMP, x1 @@ -436,10 +442,11 @@ cbz x1, .Lskip_spefds_\@ /* disable traps of PMSDSFR to EL2. */ orr x0, x0, #HDFGRTR2_EL2_nPMSDSFR_EL1 + orr x2, x2, #HDFGWTR2_EL2_nPMSDSFR_EL1 .Lskip_spefds_\@: msr_s SYS_HDFGRTR2_EL2, x0 - msr_s SYS_HDFGWTR2_EL2, x0 + msr_s SYS_HDFGWTR2_EL2, x2 msr_s SYS_HFGRTR2_EL2, xzr msr_s SYS_HFGWTR2_EL2, xzr msr_s SYS_HFGITR2_EL2, xzr From 40eb4b18ad167b63644c27fa67a7400ba050379a Mon Sep 17 00:00:00 2001 From: Chandrashekar Devegowda Date: Tue, 22 Sep 2026 12:10:20 +0530 Subject: [PATCH 429/694] Bluetooth: btintel: fix buffer over-read in btintel_hw_error() The "Exception info %s" print uses skb->data + 1, but the 12-byte payload is not guaranteed to be NUL-terminated and can be read past its end. Bound the print with "%.*s" and skb->len - 1. Fixes: 973bb97e5aee ("Bluetooth: btintel: Add generic function for handling hardware errors") Cc: stable@vger.kernel.org Signed-off-by: Chandrashekar Devegowda Signed-off-by: Luiz Augusto von Dentz --- drivers/bluetooth/btintel.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/bluetooth/btintel.c b/drivers/bluetooth/btintel.c index 909a265fd906..e35fff90b31d 100644 --- a/drivers/bluetooth/btintel.c +++ b/drivers/bluetooth/btintel.c @@ -289,7 +289,8 @@ void btintel_hw_error(struct hci_dev *hdev, u8 code) goto unlock; } - bt_dev_err(hdev, "Exception info %s", (char *)(skb->data + 1)); + bt_dev_err(hdev, "Exception info %.*s", (int)(skb->len - 1), + (char *)(skb->data + 1)); kfree_skb(skb); From 7dbeb6935ac331b0cba6c41cabd8e9d02d5f64ea Mon Sep 17 00:00:00 2001 From: Chandrashekar Devegowda Date: Tue, 22 Sep 2026 12:10:21 +0530 Subject: [PATCH 430/694] Bluetooth: btintel_pcie: fix plen overflow in btintel_pcie_recv_frame() plen is u16 and computed as HDR_SIZE + dlen. For ACL/ISO frames dlen is a 16-bit field, so a value >= 0xFFFC wraps the sum, bypassing the "skb->len < plen" check and letting skb_trim() truncate the packet. Widen plen to u32 so the addition cannot wrap. Fixes: c2b636b3f788 ("Bluetooth: btintel_pcie: Add support for PCIe transport") Cc: stable@vger.kernel.org Signed-off-by: Chandrashekar Devegowda Signed-off-by: Luiz Augusto von Dentz --- drivers/bluetooth/btintel_pcie.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/bluetooth/btintel_pcie.c b/drivers/bluetooth/btintel_pcie.c index 2819e001797b..d05af27790f9 100644 --- a/drivers/bluetooth/btintel_pcie.c +++ b/drivers/bluetooth/btintel_pcie.c @@ -1197,7 +1197,7 @@ static int btintel_pcie_recv_frame(struct btintel_pcie_data *data, { int ret; u8 pkt_type; - u16 plen; + u32 plen; u32 pcie_pkt_type; void *pdata; struct hci_dev *hdev = data->hdev; From f771f96557a4d953a7e4d045870547271e36e819 Mon Sep 17 00:00:00 2001 From: Chandrashekar Devegowda Date: Tue, 22 Sep 2026 12:10:22 +0530 Subject: [PATCH 431/694] Bluetooth: btintel_pcie: reject oversized TX packets in send_frame() btintel_pcie_send_frame() eventually copies skb->data into a fixed BTINTEL_PCIE_BUFFER_SIZE (4096) DMA buffer via memcpy() in btintel_pcie_prepare_tx(), with a 4-byte PCIe type header prepended. A user-space process with HCI_CHANNEL_USER access can inject an oversized packet and overflow the buffer. Reject packets larger than BTINTEL_PCIE_BUFFER_SIZE - BTINTEL_PCIE_HCI_TYPE_LEN on entry. Fixes: 6e65a09f9275 ("Bluetooth: btintel_pcie: Add *setup* function to download firmware") Cc: stable@vger.kernel.org Signed-off-by: Chandrashekar Devegowda Signed-off-by: Luiz Augusto von Dentz --- drivers/bluetooth/btintel_pcie.c | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/drivers/bluetooth/btintel_pcie.c b/drivers/bluetooth/btintel_pcie.c index d05af27790f9..f333b01f7465 100644 --- a/drivers/bluetooth/btintel_pcie.c +++ b/drivers/bluetooth/btintel_pcie.c @@ -2219,6 +2219,16 @@ static int btintel_pcie_send_frame(struct hci_dev *hdev, if (test_bit(BTINTEL_PCIE_RECOVERY_IN_PROGRESS, &data->flags)) return -ENODEV; + /* Account for the 4-byte PCIe type header prepended before the + * DMA copy. Written as a subtraction to avoid wrap-around on + * attacker-controlled skb->len. + */ + if (skb->len > BTINTEL_PCIE_BUFFER_SIZE - BTINTEL_PCIE_HCI_TYPE_LEN) { + bt_dev_err(hdev, "Packet too large: %u > %u", skb->len, + BTINTEL_PCIE_BUFFER_SIZE - BTINTEL_PCIE_HCI_TYPE_LEN); + return -EMSGSIZE; + } + /* Due to the fw limitation, the type header of the packet should be * 4 bytes unlike 1 byte for UART. In UART, the firmware can read * the first byte to get the packet type and redirect the rest of data From 74847177eca5638827ab74cdf77f0b10fc40ca6c Mon Sep 17 00:00:00 2001 From: Linmao Li Date: Tue, 22 Sep 2026 10:01:41 +0800 Subject: [PATCH 432/694] Bluetooth: hci_core: Fix inquiry cache timestamps on 64-bit systems On 64-bit systems, outgoing BR/EDR connections always fall back to page scan repetition mode R2 with no clock offset once the system has been up for more than five minutes, even when inquiry found the peer only seconds earlier. This lengthens paging and can increase the risk of a Page Timeout. The inquiry cache stores jiffies in __u32 timestamps, but its age helpers subtract them from unsigned long jiffies. INITIAL_JIFFIES casts -300 * HZ through unsigned int, so jiffies crosses 2^32 five minutes after boot on 64-bit systems. Assigning it to __u32 then drops the upper 32 bits. In one trace, a 7.6-second-old entry (HZ=1000) was reported as 2^32 + 7620 ticks old and rejected by hci_acl_create_conn_sync(). hci_inquiry() is affected by the same truncation when checking the whole cache. 32-bit systems are unaffected because unsigned long is 32 bits wide there. Use unsigned long for both timestamps so they have the same width as jiffies on 32-bit and 64-bit systems, and update the debugfs format specifier accordingly. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Cc: stable@vger.kernel.org Signed-off-by: Linmao Li Signed-off-by: Luiz Augusto von Dentz --- include/net/bluetooth/hci_core.h | 4 ++-- net/bluetooth/hci_debugfs.c | 2 +- 2 files changed, 3 insertions(+), 3 deletions(-) diff --git a/include/net/bluetooth/hci_core.h b/include/net/bluetooth/hci_core.h index 4105c446ca98..ca77ad0d9393 100644 --- a/include/net/bluetooth/hci_core.h +++ b/include/net/bluetooth/hci_core.h @@ -62,7 +62,7 @@ struct inquiry_entry { NAME_PENDING, NAME_KNOWN, } name_state; - __u32 timestamp; + unsigned long timestamp; struct inquiry_data data; }; @@ -78,7 +78,7 @@ struct discovery_state { struct list_head all; /* All devices found during inquiry */ struct list_head unknown; /* Name state not known */ struct list_head resolve; /* Name needs to be resolved */ - __u32 timestamp; + unsigned long timestamp; bdaddr_t last_adv_addr; u8 last_adv_addr_type; s8 last_adv_rssi; diff --git a/net/bluetooth/hci_debugfs.c b/net/bluetooth/hci_debugfs.c index aadffaaff20e..2559fb6324d6 100644 --- a/net/bluetooth/hci_debugfs.c +++ b/net/bluetooth/hci_debugfs.c @@ -364,7 +364,7 @@ static int inquiry_cache_show(struct seq_file *f, void *p) list_for_each_entry(e, &cache->all, all) { struct inquiry_data *data = &e->data; - seq_printf(f, "%pMR %d %d %d 0x%.2x%.2x%.2x 0x%.4x %d %d %u\n", + seq_printf(f, "%pMR %d %d %d 0x%.2x%.2x%.2x 0x%.4x %d %d %lu\n", &data->bdaddr, data->pscan_rep_mode, data->pscan_period_mode, data->pscan_mode, data->dev_class[2], From d1ac915fb4184faad1b62148be50f317de7c24b5 Mon Sep 17 00:00:00 2001 From: Sang-Hoon Choi Date: Tue, 22 Sep 2026 03:49:28 +0900 Subject: [PATCH 433/694] Bluetooth: hci_core: free the HCI ID if naming fails hci_register_dev() allocates an ID before calling dev_set_name(). If naming fails, it returns without releasing that ID. The device has not been registered and hdev->id has not been assigned, so the normal unregister path cannot release it. Free the allocated ID before returning the naming error. Fixes: dcda165706b9 ("Bluetooth: hci_core: Fix build warnings") Reported-by: Changyul Lee Assisted-by: LLM Signed-off-by: Sang-Hoon Choi Reported-by: Changyul Lee Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/hci_core.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/net/bluetooth/hci_core.c b/net/bluetooth/hci_core.c index d183efaf9063..bbe391525530 100644 --- a/net/bluetooth/hci_core.c +++ b/net/bluetooth/hci_core.c @@ -2557,8 +2557,10 @@ int hci_register_dev(struct hci_dev *hdev) return id; error = dev_set_name(&hdev->dev, "hci%u", id); - if (error) + if (error) { + ida_free(&hci_index_ida, id); return error; + } hdev->name = dev_name(&hdev->dev); hdev->id = id; From c9c15d4d8956df8c564f7c210e4dc5b9b820d97a Mon Sep 17 00:00:00 2001 From: Sai Teja Aluvala Date: Tue, 22 Sep 2026 18:33:11 +0530 Subject: [PATCH 434/694] Bluetooth: btintel: validate DDC record lengths Parse DDC record lengths in unsigned storage so 0xFF does not wrap. Reject records smaller than the mandatory three-byte header, records exceeding U8_MAX, and records exceeding remaining firmware bytes before issuing Intel_Write_DDC. This issue was reported by Claude Mythos. Fixes: 145f2368c5fd ("Bluetooth: btintel: Add Device Configuration support") Signed-off-by: Sai Teja Aluvala Signed-off-by: Luiz Augusto von Dentz --- drivers/bluetooth/btintel.c | 12 ++++++++++-- 1 file changed, 10 insertions(+), 2 deletions(-) diff --git a/drivers/bluetooth/btintel.c b/drivers/bluetooth/btintel.c index e35fff90b31d..c69a5d226e88 100644 --- a/drivers/bluetooth/btintel.c +++ b/drivers/bluetooth/btintel.c @@ -409,8 +409,16 @@ int btintel_load_ddc_config(struct hci_dev *hdev, const char *ddc_name) /* DDC file contains one or more DDC structure which has * Length (1 byte), DDC ID (2 bytes), and DDC value (Length - 2). */ - while (fw->size > fw_ptr - fw->data) { - u8 cmd_plen = fw_ptr[0] + sizeof(u8); + while (fw->size > (size_t)(fw_ptr - fw->data)) { + size_t remaining = fw->size - (fw_ptr - fw->data); + unsigned int cmd_plen = fw_ptr[0] + 1U; + + if (cmd_plen < 3 || cmd_plen > U8_MAX || cmd_plen > remaining) { + bt_dev_err(hdev, "Malformed DDC record (plen=%u, remaining=%zu)", + cmd_plen, remaining); + release_firmware(fw); + return -EINVAL; + } skb = __hci_cmd_sync(hdev, 0xfc8b, cmd_plen, fw_ptr, HCI_INIT_TIMEOUT); From 4cbe530c0233c7413aaaeb029a4f32dd6aadacbb Mon Sep 17 00:00:00 2001 From: Karl Mehltretter Date: Sat, 19 Sep 2026 19:10:58 +0200 Subject: [PATCH 435/694] gpio: tps65219: Fix GPIO input value reads TPS65219_MFP_GPIO_STATUS_MASK is already BIT(4). Passing it to BIT() again tests bit 16, which cannot be set in the 8-bit MFP_CTRL register, so GPIO0 is always reported low when configured as an input. Test the register value with the mask directly. Fixes: 57e30e00bd5b ("gpio: tps65219: add GPIO support for TPS65219 PMIC") Cc: stable@vger.kernel.org Assisted-by: LLM Signed-off-by: Karl Mehltretter Reviewed-by: Jonathan Cormier Link: https://patch.msgid.link/20260919171100.90430-2-kmehltretter@gmail.com Signed-off-by: Bartosz Golaszewski --- drivers/gpio/gpio-tps65219.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpio/gpio-tps65219.c b/drivers/gpio/gpio-tps65219.c index 457fd8a589e8..b25c6f727768 100644 --- a/drivers/gpio/gpio-tps65219.c +++ b/drivers/gpio/gpio-tps65219.c @@ -79,7 +79,7 @@ static int tps65219_gpio_get(struct gpio_chip *gc, unsigned int offset) if (ret) return ret; - ret = !!(val & BIT(TPS65219_MFP_GPIO_STATUS_MASK)); + ret = !!(val & TPS65219_MFP_GPIO_STATUS_MASK); dev_warn(dev, "GPIO%d = %d, MULTI_DEVICE_ENABLE, not a standard GPIO\n", offset, ret); /* From 93cf8cedeaaa05714f709b539cfb976e0b80c830 Mon Sep 17 00:00:00 2001 From: Karl Mehltretter Date: Sat, 19 Sep 2026 19:10:59 +0200 Subject: [PATCH 436/694] gpio: tps65219: Use the variant-specific direction callback The TPS65214 template installs its own get_direction callback because its direction bit is in GENERAL_CONFIG. The shared get and direction callbacks nevertheless call tps65219_gpio_get_direction() directly and interpret the unrelated TPS65219 MFP bit. On TPS65214 this can reject reads from an input and skip the change from input to output. Call the callback selected by the gpio_chip template instead. Fixes: 1b6ab07c0c80 ("gpio: tps65219: Add support for TI TPS65214 PMIC") Cc: stable@vger.kernel.org Assisted-by: LLM Signed-off-by: Karl Mehltretter Link: https://patch.msgid.link/20260919171100.90430-3-kmehltretter@gmail.com Signed-off-by: Bartosz Golaszewski --- drivers/gpio/gpio-tps65219.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/gpio/gpio-tps65219.c b/drivers/gpio/gpio-tps65219.c index b25c6f727768..479a80ef7654 100644 --- a/drivers/gpio/gpio-tps65219.c +++ b/drivers/gpio/gpio-tps65219.c @@ -87,7 +87,7 @@ static int tps65219_gpio_get(struct gpio_chip *gc, unsigned int offset) * status bit. */ - if (tps65219_gpio_get_direction(gc, offset) == GPIO_LINE_DIRECTION_OUT) + if (gc->get_direction(gc, offset) == GPIO_LINE_DIRECTION_OUT) return -ENOTSUPP; return ret; @@ -176,7 +176,7 @@ static int tps65219_gpio_direction_input(struct gpio_chip *gc, unsigned int offs return -ENOTSUPP; } - if (tps65219_gpio_get_direction(gc, offset) == GPIO_LINE_DIRECTION_IN) + if (gc->get_direction(gc, offset) == GPIO_LINE_DIRECTION_IN) return 0; return gpio->change_dir(gc, offset, GPIO_LINE_DIRECTION_IN); @@ -190,7 +190,7 @@ static int tps65219_gpio_direction_output(struct gpio_chip *gc, unsigned int off if (offset != TPS6521X_GPIO0_IDX) return 0; - if (tps65219_gpio_get_direction(gc, offset) == GPIO_LINE_DIRECTION_OUT) + if (gc->get_direction(gc, offset) == GPIO_LINE_DIRECTION_OUT) return 0; return gpio->change_dir(gc, offset, GPIO_LINE_DIRECTION_OUT); From 270437f3fe62516f16482742a7762a075e7a9457 Mon Sep 17 00:00:00 2001 From: Karl Mehltretter Date: Sat, 19 Sep 2026 19:11:00 +0200 Subject: [PATCH 437/694] gpio: tps65219: Fix TPS65214 GPIO direction programming GPIO_LINE_DIRECTION_OUT and GPIO_LINE_DIRECTION_IN have the values 0 and 1, respectively, while the TPS65214 GPIO_CONFIG field is BIT(1). regmap_update_bits() masks the supplied value, so passing either direction value clears the field and selects input mode. Translate the GPIO direction to the register encoding used by tps65214_gpio_get_direction(), setting GPIO_CONFIG for output and clearing it for input. Fixes: 1b6ab07c0c80 ("gpio: tps65219: Add support for TI TPS65214 PMIC") Cc: stable@vger.kernel.org Assisted-by: LLM Signed-off-by: Karl Mehltretter Link: https://patch.msgid.link/20260919171100.90430-4-kmehltretter@gmail.com Signed-off-by: Bartosz Golaszewski --- drivers/gpio/gpio-tps65219.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/gpio/gpio-tps65219.c b/drivers/gpio/gpio-tps65219.c index 479a80ef7654..6958466455d0 100644 --- a/drivers/gpio/gpio-tps65219.c +++ b/drivers/gpio/gpio-tps65219.c @@ -158,8 +158,10 @@ static int tps65214_gpio_change_direction(struct gpio_chip *gc, unsigned int off if (ret) dev_err(dev, "GPIO%d configured as VSEL, not GPIO\n", offset); + val = direction == GPIO_LINE_DIRECTION_OUT ? + TPS65214_GPIO0_DIR_MASK : 0; ret = regmap_update_bits(gpio->tps->regmap, TPS65219_REG_GENERAL_CONFIG, - TPS65214_GPIO0_DIR_MASK, direction); + TPS65214_GPIO0_DIR_MASK, val); if (ret) dev_err(dev, "Fail to change direction to %u for GPIO%d.\n", direction, offset); From 5b76268dac968612f7283d59b539036de955b7d9 Mon Sep 17 00:00:00 2001 From: Vineeth Vijayan Date: Tue, 22 Sep 2026 22:48:39 +0200 Subject: [PATCH 438/694] s390/cio: Fix NULL pointer dereference in ccw_device_get_util_str() The channel path registry entry associated with a CHPID may be removed while the subchannel's PMCW still references that CHPID. In this case, chpid_to_chp() can return NULL, leading to a NULL pointer dereference. Add the missing NULL check before dereferencing the returned pointer. Fixes: 199652309a4d ("s390/cio: add helper to query utility strings per given ccw device") Cc: stable@vger.kernel.org Signed-off-by: Vineeth Vijayan Reviewed-by: Peter Oberparleiter Signed-off-by: Heiko Carstens --- drivers/s390/cio/device_ops.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/s390/cio/device_ops.c b/drivers/s390/cio/device_ops.c index f2f7f8cba410..1f7e83fd5087 100644 --- a/drivers/s390/cio/device_ops.c +++ b/drivers/s390/cio/device_ops.c @@ -517,6 +517,8 @@ u8 *ccw_device_get_util_str(struct ccw_device *cdev, int chp_idx) chp_id_init(&chpid); chpid.id = sch->schib.pmcw.chpid[chp_idx]; chp = chpid_to_chp(chpid); + if (!chp) + return NULL; util_str = kmalloc(sizeof(chp->desc_fmt3.util_str), GFP_KERNEL); if (!util_str) From 883b776abe48fed8ef615f4ff7e91113a6c4dffb Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Tue, 22 Sep 2026 12:41:02 +0200 Subject: [PATCH 439/694] isofs: Fix handling of directories with tight blocks Thomas has reported that after commit b2eb2e288604 ("isofs: Drop support of directory entries straddling blocks") some isofs images of Fedora miss some entries in some directories. The problem is triggered when directory entries are packed in a block in such a way that they exactly fit the block (which BTW mkisofs doesn't do so I didn't catch this bug when testing with my images). Fix readdir and lookup code to properly transition to the next block when directory block is tightly packed. Link: https://bugzilla.redhat.com/show_bug.cgi?id=2535353 Reported-by: Thomas Schmitt Reported-by: Matthias Goergens Fixes: b2eb2e288604 ("isofs: Drop support of directory entries straddling blocks") Reviewed-by: Thomas Schmitt Signed-off-by: Jan Kara --- fs/isofs/dir.c | 20 ++++++++------------ fs/isofs/namei.c | 16 +++++++++------- 2 files changed, 17 insertions(+), 19 deletions(-) diff --git a/fs/isofs/dir.c b/fs/isofs/dir.c index c7ca7603e97a..28741251d56e 100644 --- a/fs/isofs/dir.c +++ b/fs/isofs/dir.c @@ -110,25 +110,21 @@ static int do_isofs_readdir(struct inode *inode, struct file *file, return 0; } - de = (struct iso_directory_record *) (bh->b_data + offset); - - de_len = *(unsigned char *)de; - + de = (struct iso_directory_record *)(bh->b_data + offset); /* - * If the length byte is zero, we should move on to the next - * CDROM sector. If we are at the end of the directory, we - * kick out of the while loop. + * If we are at the end of a block (or at its zero-padded + * tail), move on to the next CDROM sector. If we are at the + * end of the directory, we'll abort the while loop. */ - - if (de_len == 0) { + if (offset >= bufsize || de->length[0] == 0) { brelse(bh); bh = NULL; - ctx->pos = (ctx->pos + ISOFS_BLOCK_SIZE) & ~(ISOFS_BLOCK_SIZE - 1); - block = ctx->pos >> bufbits; + block++; + ctx->pos = (loff_t)block << bufbits; offset = 0; continue; } - + de_len = de->length[0]; block_saved = block; offset_saved = offset; offset += de_len; diff --git a/fs/isofs/namei.c b/fs/isofs/namei.c index 010682f5901a..4fba1bf7f016 100644 --- a/fs/isofs/namei.c +++ b/fs/isofs/namei.c @@ -74,18 +74,20 @@ isofs_find_entry(struct inode *dir, struct dentry *dentry, return 0; } - de = (struct iso_directory_record *) (bh->b_data + offset); - - de_len = *(unsigned char *) de; - if (!de_len) { + de = (struct iso_directory_record *)(bh->b_data + offset); + /* + * If we are at the end of the block or at its zero-padded + * tail, move to the next block. + */ + if (offset >= bufsize || de->length[0] == 0) { brelse(bh); bh = NULL; - f_pos = (f_pos + ISOFS_BLOCK_SIZE) & ~(ISOFS_BLOCK_SIZE - 1); - block = f_pos >> bufbits; + block++; + f_pos = block << bufbits; offset = 0; continue; } - + de_len = de->length[0]; block_saved = bh->b_blocknr; offset_saved = offset; offset += de_len; From 4409a85735cdca5c4c400c0dad1feee091872eee Mon Sep 17 00:00:00 2001 From: Liang Luo Date: Wed, 23 Sep 2026 19:07:37 +0800 Subject: [PATCH 440/694] sched_ext: Count SCX_EV_SUB_BYPASS_DISPATCH in the dispatch fallback When a descendant scheduler enters bypass mode, its tasks are parked in the bypass DSQs of the nearest non-bypassing ancestor, which is then responsible for running them. On behalf of such a non-bypassing host, scx_dispatch_sched() consumes those bypass DSQs from two places: the attempt made every SCX_BYPASS_HOST_NTH dispatches, and the end-of-dispatch fallback that keeps the CPU from going idle while bypassed descendants still have tasks queued. The former increments SCX_EV_SUB_BYPASS_DISPATCH but the latter does not, even though both perform the same scx_consume_dispatch_q() on the same bypass DSQ. The descendant bypass dispatches done by the fallback are therefore missing from the counter exposed via sysfs, scx_dump_state() and the scx_bpf_events() kfunc, which under-reports the actual number of such dispatches. Add the missing __scx_add_event() so the fallback counts them too. When @sch itself is bypassing, scx_dispatch_sched() takes the earlier self-bypass branch and returns before reaching these host paths; that mode is accounted for by SCX_EV_BYPASS_DISPATCH at enqueue time and is intentionally left unchanged. Fixes: 025b1bd41965 ("sched_ext: Implement hierarchical bypass mode") Signed-off-by: Liang Luo Signed-off-by: Tejun Heo --- kernel/sched/ext/inlines.h | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/kernel/sched/ext/inlines.h b/kernel/sched/ext/inlines.h index ed423bcc26b8..2ff5479334cb 100644 --- a/kernel/sched/ext/inlines.h +++ b/kernel/sched/ext/inlines.h @@ -129,8 +129,10 @@ scx_dispatch_sched(struct scx_sched *sch, struct rq *rq, * scheduler's ops.dispatch() doesn't yield any tasks. */ if (scx_bypass_dsp_enabled(sch) && - scx_consume_dispatch_q(sch, rq, scx_bypass_dsq(sch, cpu), 0)) + scx_consume_dispatch_q(sch, rq, scx_bypass_dsq(sch, cpu), 0)) { + __scx_add_event(sch, SCX_EV_SUB_BYPASS_DISPATCH, 1); return SCX_DSP_LOCAL; + } return SCX_DSP_NONE; } From b8d1d5b63a8ef532038eebd9d97d406860385668 Mon Sep 17 00:00:00 2001 From: "Masami Hiramatsu (Google)" Date: Tue, 22 Sep 2026 13:24:55 +0900 Subject: [PATCH 441/694] x86/mce: Fix hardware debug register corruption on task migration In exc_machine_check_user(), local_db_save() and local_db_restore() are invoked in the outer entry stubs (DEFINE_IDTENTRY_MCE_USER, DEFINE_FREDENTRY_MCE, and DEFINE_IDTENTRY_RAW), surrounding exc_machine_check_user(). However, exc_machine_check_user() calls irqentry_exit_to_user_mode(), which handles pending thread work and may schedule() if TIF_NEED_RESCHED is set. If the task migrates to another CPU during schedule(), local_db_restore() runs on the new CPU with the dr7 state saved from the old CPU. This corrupts the new CPU's DR7 hardware debug register and leaves the old CPU's DR7 disabled. In short, local_db_save() and local_db_restore() pair must be run on the same CPU. To fix this, move local_db_save() and local_db_restore() inside exc_machine_check_user() and exc_machine_check_kernel(). In exc_machine_check_user(), DR7 is saved and restored strictly around do_machine_check() to avoid schedule() during migration. In exc_machine_check_kernel(), local_db_save() is called at the entry point to prevent early memory accesses from triggering nested #DB exceptions, and restored on all exits. Fixes: cd840e424f27 ("x86/entry, mce: Disallow #DB during #MC") Assisted-by: LLM Signed-off-by: Masami Hiramatsu (Google) Signed-off-by: Borislav Petkov (AMD) Acked-by: Peter Zijlstra (Intel) Cc: Link: https://patch.msgid.link/179005109564.388919.3937970081044095776.stgit@devnote2 --- arch/x86/kernel/cpu/mce/core.c | 27 ++++++++++----------------- 1 file changed, 10 insertions(+), 17 deletions(-) diff --git a/arch/x86/kernel/cpu/mce/core.c b/arch/x86/kernel/cpu/mce/core.c index ab469605fc89..39f238952e14 100644 --- a/arch/x86/kernel/cpu/mce/core.c +++ b/arch/x86/kernel/cpu/mce/core.c @@ -2108,6 +2108,9 @@ bool filter_mce(struct mce *m) static __always_inline void exc_machine_check_kernel(struct pt_regs *regs) { irqentry_state_t irq_state; + unsigned long dr7; + + dr7 = local_db_save(); WARN_ON_ONCE(user_mode(regs)); @@ -2116,20 +2119,26 @@ static __always_inline void exc_machine_check_kernel(struct pt_regs *regs) * mce_check_crashing_cpu() for details. */ if (mca_cfg.initialized && mce_check_crashing_cpu()) - return; + goto out; irq_state = irqentry_nmi_enter(regs); do_machine_check(regs); irqentry_nmi_exit(regs, irq_state); +out: + local_db_restore(dr7); } static __always_inline void exc_machine_check_user(struct pt_regs *regs) { + unsigned long dr7; + irqentry_enter_from_user_mode(regs); + dr7 = local_db_save(); do_machine_check(regs); + local_db_restore(dr7); irqentry_exit_to_user_mode(regs); } @@ -2138,21 +2147,13 @@ static __always_inline void exc_machine_check_user(struct pt_regs *regs) /* MCE hit kernel mode */ DEFINE_IDTENTRY_MCE(exc_machine_check) { - unsigned long dr7; - - dr7 = local_db_save(); exc_machine_check_kernel(regs); - local_db_restore(dr7); } /* The user mode variant. */ DEFINE_IDTENTRY_MCE_USER(exc_machine_check) { - unsigned long dr7; - - dr7 = local_db_save(); exc_machine_check_user(regs); - local_db_restore(dr7); } #ifdef CONFIG_X86_FRED @@ -2169,28 +2170,20 @@ DEFINE_IDTENTRY_MCE_USER(exc_machine_check) */ DEFINE_FREDENTRY_MCE(exc_machine_check) { - unsigned long dr7; - - dr7 = local_db_save(); if (user_mode(regs)) exc_machine_check_user(regs); else exc_machine_check_kernel(regs); - local_db_restore(dr7); } #endif #else /* 32bit unified entry point */ DEFINE_IDTENTRY_RAW(exc_machine_check) { - unsigned long dr7; - - dr7 = local_db_save(); if (user_mode(regs)) exc_machine_check_user(regs); else exc_machine_check_kernel(regs); - local_db_restore(dr7); } #endif From 4fde448225123442c5796f54b7a4400e2d3cbaf6 Mon Sep 17 00:00:00 2001 From: Mario Limonciello Date: Tue, 8 Sep 2026 14:05:59 -0500 Subject: [PATCH 442/694] x86/PCI: Disable enhanced atomics on AMD NBIO 7.7 and 7.11 Multiple users report data corruption during 64-bit DMA transfers on systems with AMD NBIO 7.7 and 7.11 controllers. This occurs when BIOS enables AMD "enhanced atomic operations" on PCIe Root Ports. When enhanced atomics are enabled, any 64-bit DMA access may be corrupted. Disable enhanced atomics using SMN for NBIO 7.7 and 7.11 based models. Reported-by: Mikael Etienne Closes: https://lore.kernel.org/178789300872.392066.15963676631650361573@gmail.com/ Reported-by: Arthur Husband Closes: https://lore.kernel.org/20260406222335.379935-1-artmoty@gmail.com/ Reported-by: Alvin Lim Closes: https://lore.kernel.org/20260621100844.1224301-1-alvinwylim@gmail.com/ Signed-off-by: Mario Limonciello [bhelgaas: commit log, s/IOVA/DMA/ in comment] Signed-off-by: Bjorn Helgaas Cc: stable@vger.kernel.org Cc: David Laight Cc: John Smith Cc: Lennert Buytenhek Cc: Niklas Cassel Cc: Roland Waltersson Link: https://patch.msgid.link/20260908190600.226485-2-mario.limonciello@amd.com --- arch/x86/pci/fixup.c | 99 ++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 99 insertions(+) diff --git a/arch/x86/pci/fixup.c b/arch/x86/pci/fixup.c index b301c6c8df75..795d81c7a4de 100644 --- a/arch/x86/pci/fixup.c +++ b/arch/x86/pci/fixup.c @@ -886,6 +886,105 @@ static void quirk_clear_strap_no_soft_reset_dev2_f0(struct pci_dev *dev) } } DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_AMD, 0x15b8, quirk_clear_strap_no_soft_reset_dev2_f0); + +/* + * Enhanced atomic operations can cause corruption with 64-bit DMA + * on these devices. + */ +#define RX_ENH_ATOMIC_EN BIT(8) + +static const u32 nbio_7_7_pcie_smn_addrs[] = { + 0x111401d0, + 0x111411d0, + 0x111421d0, + 0x111431d0, + 0x111441d0, + 0x112401d0, + 0x112411d0, + 0x112421d0, + 0x112431d0, + 0x112441d0, + 0x112451d0, + 0x113401d0, + 0x114401d0, +}; + +static const u32 nbio_7_11_pcie_smn_addrs[] = { + 0x112401d0, + 0x112411d0, + 0x112421d0, + 0x112431d0, + 0x112441d0, + 0x112451d0, + 0x113401d0, + 0x113411d0, + 0x113421d0, + 0x113431d0, + 0x113441d0, + 0x113451d0, +}; + +static void quirk_amd_nbio_enhanced_atomic(struct pci_dev *host_bridge, + const u32 *smn_addrs, + size_t nr_smn_addrs) +{ + bool changed = false; + size_t i; + u32 data; + int ret; + + for (i = 0; i < nr_smn_addrs; i++) { + ret = amd_smn_read(0, smn_addrs[i], &data); + if (ret) + continue; + if (!(data & RX_ENH_ATOMIC_EN)) + continue; + data = data & ~RX_ENH_ATOMIC_EN; + ret = amd_smn_write(0, smn_addrs[i], data); + if (ret) + continue; + if (changed) + continue; + ret = amd_smn_read(0, smn_addrs[i], &data); + if (ret) + continue; + if (data & RX_ENH_ATOMIC_EN) + continue; + changed = true; + } + + if (changed) + pci_info(host_bridge, "enhanced atomics disabled\n"); +} + +static void quirk_amd_nbio_7_7_disable_enhanced_atomic(struct pci_dev *dev) +{ + quirk_amd_nbio_enhanced_atomic(dev, nbio_7_7_pcie_smn_addrs, + ARRAY_SIZE(nbio_7_7_pcie_smn_addrs)); +} + +static void quirk_amd_nbio_7_11_disable_enhanced_atomic(struct pci_dev *dev) +{ + quirk_amd_nbio_enhanced_atomic(dev, nbio_7_11_pcie_smn_addrs, + ARRAY_SIZE(nbio_7_11_pcie_smn_addrs)); +} + +/* Phoenix, Hawk Point (NBIO 7.7) */ +DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_AMD, 0x14E8, + quirk_amd_nbio_7_7_disable_enhanced_atomic); +DECLARE_PCI_FIXUP_RESUME(PCI_VENDOR_ID_AMD, 0x14E8, + quirk_amd_nbio_7_7_disable_enhanced_atomic); + +/* Strix, Krackan, Strix Halo (NBIO 7.11) */ +DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_AMD, 0x1507, + quirk_amd_nbio_7_11_disable_enhanced_atomic); +DECLARE_PCI_FIXUP_RESUME(PCI_VENDOR_ID_AMD, 0x1507, + quirk_amd_nbio_7_11_disable_enhanced_atomic); +DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_AMD, 0x1122, + quirk_amd_nbio_7_11_disable_enhanced_atomic); +DECLARE_PCI_FIXUP_RESUME(PCI_VENDOR_ID_AMD, 0x1122, + quirk_amd_nbio_7_11_disable_enhanced_atomic); + #endif /* From 80320b278fea07ffcda3f57b67b61658e0a4e1ca Mon Sep 17 00:00:00 2001 From: Matthias Goergens Date: Thu, 24 Sep 2026 01:52:03 +0800 Subject: [PATCH 443/694] ata: libata-scsi: bound the ATA passthru sense descriptor writes When an ATA PASS-THROUGH command to an ATAPI device fails, the sense buffer holds the device's REQUEST SENSE reply, and ata_scsi_set_passthru_sense_fields() trusts its additional length byte, sb[7], when adding the ATA Status Return descriptor. A faulty or malicious device can use that to make the kernel read and write past the 96-byte buffer in three ways: - scsi_sense_desc_find() is passed sb[7] + 8 as the buffer length, so its clamp against sb[7] does nothing and the walk runs off the end. - A type-9 descriptor found near the end is filled in unchecked. - A new descriptor at sb[8 + len] needs len + 22 bytes, not len + 14, so len 75..82 writes up to 8 bytes past the end. Reproduced with KASAN under qemu, with the emulated ATAPI REQUEST SENSE reply patched: BUG: KASAN: slab-out-of-bounds in scsi_sense_desc_find+0x1a5/0x210 BUG: KASAN: slab-out-of-bounds in ata_scsi_qc_complete+0x1a15/0x1a50 Both are gone with this patch, and a valid descriptor is still filled in. Fixes: 97981926224a ("ata: libata-scsi: Do not overwrite valid sense data when CK_COND=1") Cc: stable@vger.kernel.org Reviewed-by: Damien Le Moal Signed-off-by: Matthias Goergens Link: https://lore.kernel.org/r/20260923175203.1576825-1-matthias.goergens@gmail.com Signed-off-by: Niklas Cassel --- drivers/ata/libata-scsi.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) diff --git a/drivers/ata/libata-scsi.c b/drivers/ata/libata-scsi.c index 7e22bbc38238..8f9aa97a519d 100644 --- a/drivers/ata/libata-scsi.c +++ b/drivers/ata/libata-scsi.c @@ -261,12 +261,18 @@ static void ata_scsi_set_passthru_sense_fields(struct ata_queued_cmd *qc) /* descriptor format */ len = sb[7]; - desc = (char *)scsi_sense_desc_find(sb, len + 8, 9); + desc = (char *)scsi_sense_desc_find(sb, SCSI_SENSE_BUFFERSIZE, 9); if (!desc) { - if (SCSI_SENSE_BUFFERSIZE < len + 14) + /* + * The descriptor is written at sb[8 + len] and is 14 + * bytes long, so it needs len + 22 bytes of buffer. + */ + if (len + 22 > SCSI_SENSE_BUFFERSIZE) return; sb[7] = len + 14; desc = sb + 8 + len; + } else if (desc - sb > SCSI_SENSE_BUFFERSIZE - 14) { + return; } desc[0] = 9; desc[1] = 12; From c5fd4eaad50d620c7e09ac2082b2fb55ee54170e Mon Sep 17 00:00:00 2001 From: Wentao Liang Date: Wed, 16 Sep 2026 09:49:56 +0000 Subject: [PATCH 444/694] drm/amd/display: Fix dc stream excess put in dm_update_crtc_state() In dm_update_crtc_state(), when a modeset is required the newly created stream is stored in dm_new_crtc_state->stream and an extra reference is taken with dc_stream_retain(). The reference returned by create_validate_stream_for_sink() is released as an extra reference at the skip_modeset label, leaving the stream owned by the new CRTC state. If amdgpu_dm_check_crtc_color_mgmt() fails afterwards, the code jumps to the fail label which releases new_stream again. Since the extra reference was already released at skip_modeset, this drops the reference owned by dm_new_crtc_state->stream and the stream is released while the atomic state still points to it, leading to a premature free of the dc stream. Set new_stream to NULL after releasing the extra reference at the skip_modeset label so that a later goto fail cannot release the reference owned by the new CRTC state. Fixes: 7cd4b70091a5 ("drm/amd/display: Rework CRTC color management") Signed-off-by: Wentao Liang Signed-off-by: Alex Deucher (cherry picked from commit 102a47065a62dc8f6bbbb47cf082a2934282eb08) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c index 36d2f86f000a..91fdf3de7202 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -5583,8 +5583,10 @@ static int dm_update_crtc_state(struct amdgpu_display_manager *dm, skip_modeset: /* Release extra reference */ - if (new_stream) + if (new_stream) { dc_stream_release(new_stream); + new_stream = NULL; + } new_stream = NULL; /* From 3022bdfe3e6d776e9273d6892f7c193138ca0666 Mon Sep 17 00:00:00 2001 From: Prike Liang Date: Fri, 31 Jul 2026 11:44:37 +0800 Subject: [PATCH 445/694] drm/amdgpu: move userq fence wait out of signalling section The eviction fence suspend worker waits for every pending userq fence from inside a dma_fence_begin_signalling() critical section. Waiting on another DMA fence while responsible for signalling one violates the cross-driver fence contract and is reported by lockdep as a dma_fence_map dependency. Move the wait before dma_fence_begin_signalling(). Keep userq_mutex held so queue lifetime remains stable while inspecting last_fence. Fixes: fc61df151617 ("drm/amdgpu: annotate eviction fence signaling path") Signed-off-by: Prike Liang Reviewed-by: Vitaly Prosyak Signed-off-by: Alex Deucher (cherry picked from commit 3bd4fbc5ed89621340b5cd249869092691a9c81f) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/amdgpu/amdgpu_eviction_fence.c | 3 +++ drivers/gpu/drm/amd/amdgpu/amdgpu_userq.c | 4 +--- drivers/gpu/drm/amd/amdgpu/amdgpu_userq.h | 1 + 3 files changed, 5 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_eviction_fence.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_eviction_fence.c index f6b7522c3c82..f8652fd0525d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_eviction_fence.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_eviction_fence.c @@ -68,6 +68,9 @@ amdgpu_eviction_fence_suspend_worker(struct work_struct *work) mutex_lock(&uq_mgr->userq_mutex); + /* Fence waits are not allowed in a fence signalling critical section. */ + amdgpu_userq_wait_for_signal(uq_mgr); + /* * This is intentionally after taking the userq_mutex since we do * allocate memory while holding this lock, but only after ensuring that diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_userq.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_userq.c index e43bda0cab3f..280bdeb38970 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_userq.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_userq.c @@ -1272,7 +1272,7 @@ amdgpu_userq_evict_all(struct amdgpu_userq_mgr *uq_mgr) return ret; } -static void +void amdgpu_userq_wait_for_signal(struct amdgpu_userq_mgr *uq_mgr) { struct amdgpu_usermode_queue *queue; @@ -1291,8 +1291,6 @@ amdgpu_userq_wait_for_signal(struct amdgpu_userq_mgr *uq_mgr) void amdgpu_userq_evict(struct amdgpu_userq_mgr *uq_mgr) { - /* Wait for any pending userqueue fence work to finish */ - amdgpu_userq_wait_for_signal(uq_mgr); amdgpu_userq_evict_all(uq_mgr); } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_userq.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_userq.h index 6412a7f7b6ef..488dc21d7c81 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_userq.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_userq.h @@ -162,6 +162,7 @@ void amdgpu_userq_mgr_cancel_reset_work(struct amdgpu_device *adev); void amdgpu_userq_mgr_cancel_resume(struct amdgpu_userq_mgr *userq_mgr); void amdgpu_userq_mgr_fini(struct amdgpu_userq_mgr *userq_mgr); +void amdgpu_userq_wait_for_signal(struct amdgpu_userq_mgr *uq_mgr); void amdgpu_userq_evict(struct amdgpu_userq_mgr *uq_mgr); void amdgpu_userq_ensure_ev_fence(struct amdgpu_userq_mgr *userq_mgr, From cd195f1616b2bb5fb7765465326c4d5d64a620a0 Mon Sep 17 00:00:00 2001 From: Sunil Khatri Date: Thu, 17 Sep 2026 18:56:45 +0530 Subject: [PATCH 446/694] drm/amdgpu/userq: fix double jiffies conversion in hang detect timeout MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Function amdgpu_userq_start_hang_detect_work() calls msecs_to_jiffies() on adev->gfx_timeout/compute_timeout/sdma_timeout before arming hang_detect_work. These timeout values already hold jiffies values from amdgpu_device_get_job_timeout_settings() at device init. This silently shrinks the real hang-detect deadline to (2 * HZ) ms instead of the intended timeout. e.g. 500ms instead of the 2000ms default on a CONFIG_HZ=250 kernel, only coincidentally correct at HZ=1000. The shortened window is easily exceeded by ordinary fence-completion latency, causing hang_detect_work to fire and trigger a per-queue or full GPU reset for queues that are not actually hung. Pass the jiffies value directly to queue_delayed_work() instead of converting it a second time. Fixes: fc3336be9c62 ("drm/amd/amdgpu: Add independent hang detect work for user queue fence") Signed-off-by: Sunil Khatri Reviewed-by: Christian König Signed-off-by: Alex Deucher (cherry picked from commit 13d44ca033cb74756c2aef0ade54a75cdf2f6271) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/amdgpu/amdgpu_userq.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_userq.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_userq.c index 280bdeb38970..cc8e7af18834 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_userq.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_userq.c @@ -184,27 +184,27 @@ static void amdgpu_userq_hang_detect_work(struct work_struct *work) void amdgpu_userq_start_hang_detect_work(struct amdgpu_usermode_queue *queue) { struct amdgpu_device *adev; - unsigned long timeout_ms; + unsigned long timeout_jiffies; adev = queue->userq_mgr->adev; /* Determine timeout based on queue type */ switch (queue->queue_type) { case AMDGPU_RING_TYPE_GFX: - timeout_ms = adev->gfx_timeout; + timeout_jiffies = adev->gfx_timeout; break; case AMDGPU_RING_TYPE_COMPUTE: - timeout_ms = adev->compute_timeout; + timeout_jiffies = adev->compute_timeout; break; case AMDGPU_RING_TYPE_SDMA: - timeout_ms = adev->sdma_timeout; + timeout_jiffies = adev->sdma_timeout; break; default: - timeout_ms = adev->gfx_timeout; + timeout_jiffies = adev->gfx_timeout; break; } queue_delayed_work(adev->reset_domain->wq, &queue->hang_detect_work, - msecs_to_jiffies(timeout_ms)); + timeout_jiffies); } void amdgpu_userq_process_fence_irq(struct amdgpu_device *adev, u32 doorbell) From f952ed353a27b46c86c9525a39b7a642850b8139 Mon Sep 17 00:00:00 2001 From: Sunil Khatri Date: Thu, 17 Sep 2026 18:56:45 +0530 Subject: [PATCH 447/694] drm/amdgpu/vcn5.0.1: fix video_timeout unit mismatch in jpeg reset wait MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit vcn_v5_0_1_reset_jpeg_pre_helper() passes adev->video_timeout directly to amdgpu_fence_wait_polling(), whose timeout parameter is documented and implemented in usecs (busy-wait loop decrementing by udelay(2)). adev->video_timeout is set in jiffies by amdgpu_device_get_job_timeout_settings(), via msecs_to_jiffies(). Passing it unconverted means the intended ~2s wait for outstanding JPEG fences to complete before the JPEG queue is torn down actually lasts only a couple of microseconds (HZ jiffies interpreted as usecs), so pending jobs are almost never given a real chance to finish before the reset path forces completion in the following helper. Convert the jiffies value to usecs with jiffies_to_usecs() before passing it to amdgpu_fence_wait_polling(). Fixes: fab47d2db5ca ("drm/amdgpu/vcn5.0.1: rework reset handling") Cc: Jesse.Zhang Assisted-by: Claude:claude-sonnet-5 Signed-off-by: Sunil Khatri Reviewed-by: Christian König Signed-off-by: Alex Deucher (cherry picked from commit b8334fec8b90ebffcaa01001a23edca9f29a05e9) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/amdgpu/vcn_v5_0_1.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v5_0_1.c b/drivers/gpu/drm/amd/amdgpu/vcn_v5_0_1.c index 1a07c3bf4425..011afc0fdc87 100644 --- a/drivers/gpu/drm/amd/amdgpu/vcn_v5_0_1.c +++ b/drivers/gpu/drm/amd/amdgpu/vcn_v5_0_1.c @@ -1335,7 +1335,8 @@ static int vcn_v5_0_1_reset_jpeg_pre_helper(struct amdgpu_device *adev, int inst /* if Jobs are still pending after timeout, * We'll handle them in the bottom helper */ - amdgpu_fence_wait_polling(ring, wait_seq, adev->video_timeout); + amdgpu_fence_wait_polling(ring, wait_seq, + jiffies_to_usecs(adev->video_timeout)); } return 0; From 6b13ddbf5bb8deec337f9b4887f579097a953f6a Mon Sep 17 00:00:00 2001 From: Sunil Khatri Date: Thu, 17 Sep 2026 18:56:45 +0530 Subject: [PATCH 448/694] drm/amdgpu/vcn4.0.3: fix video_timeout unit mismatch in jpeg reset wait MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit vcn_v4_0_3_reset_jpeg_pre_helper() passes adev->video_timeout directly to amdgpu_fence_wait_polling(), whose timeout parameter is documented and implemented in usecs (busy-wait loop decrementing by udelay(2)). adev->video_timeout is set in jiffies by amdgpu_device_get_job_timeout_settings(), via msecs_to_jiffies(). Passing it unconverted means the intended ~2s wait for outstanding JPEG fences to complete before the JPEG queue is torn down actually lasts only a couple of microseconds (HZ jiffies interpreted as usecs), so pending jobs are almost never given a real chance to finish before the reset path forces completion in the following helper. Convert the jiffies value to usecs with jiffies_to_usecs() before passing it to amdgpu_fence_wait_polling(). Fixes: d25c67fd9d6f ("drm/amdgpu/vcn4.0.3: rework reset handling") Cc: Jesse.Zhang Assisted-by: Claude:claude-sonnet-5 Signed-off-by: Sunil Khatri Reviewed-by: Christian König Signed-off-by: Alex Deucher (cherry picked from commit 5feabbd673c10ebee22b880e4d812f08974d2ef7) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/amdgpu/vcn_v4_0_3.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_3.c b/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_3.c index 179b892fb410..62e2e04314dc 100644 --- a/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_3.c +++ b/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_3.c @@ -1689,7 +1689,8 @@ static int vcn_v4_0_3_reset_jpeg_pre_helper(struct amdgpu_device *adev, int inst /* if Jobs are still pending after timeout, * We'll handle them in the bottom helper */ - amdgpu_fence_wait_polling(ring, wait_seq, adev->video_timeout); + amdgpu_fence_wait_polling(ring, wait_seq, + jiffies_to_usecs(adev->video_timeout)); } return 0; From c3a31087b1c8df679b653c1a09d9abe5ff7ec8ef Mon Sep 17 00:00:00 2001 From: Asad Kamal Date: Fri, 28 Aug 2026 15:59:21 +0800 Subject: [PATCH 449/694] drm/amdkfd: fix use-after-free and multi-container gap in kfd_dev_mapping kfd_dev_mapping caches the address_space of the first /dev/kfd opener so that the GPU reset path can call unmap_mapping_range() to zap all userspace mappings of doorbell and MMIO ranges. This design has two bugs that both manifest under SRIOV with multiple containers: 1. Use-after-free / rwsem deadlock. The cached pointer refers to an inode owned by the first opener's container. When that container exits and its inode is released, kfd_dev_mapping becomes a dangling pointer. A subsequent GPU reset dereferences it inside unmap_mapping_range(), which takes i_mmap_rwsem on the freed inode, causing a hard hang observable as an uninterruptible rwsem wait. 2. Multi-container gap. Only the first opener's address_space is cached; VMAs created by later openers live in a different address_space and are never reached by unmap_mapping_range(). After a GPU reset those stale mappings keep doorbell and MMIO pages accessible to guest userspace with no GPU behind them, risking PCIe transaction timeouts and NMI panics. Fix both bugs with the same approach used by DRM core (drm_drv.c): create a private pseudo-filesystem at module init time and allocate one anonymous inode from it. In kfd_open() redirect every opener's file->f_mapping to that inode's address_space. The inode is module-owned, lives exactly as long as the amdgpu module, and collects VMAs from all openers in one address_space. A single unmap_mapping_range() call in the reset path then correctly reaches every container's mappings with no dangling pointer risk. The hang manifests as an NMI backtrace on the GPU reset workqueue stuck spinning in rwsem_down_read_slowpath() with a corrupted i_mmap_rwsem: Workqueue: amdgpu-reset-dev xgpu_ai_mailbox_flr_work [amdgpu] Call Trace: kvm_wait+0x1f/0x40 __pv_queued_spin_lock_slowpath+0x31d/0x3a0 _raw_spin_lock_irq+0x51/0x80 rwsem_down_read_slowpath+0xb3/0x550 down_read+0x48/0xd0 unmap_mapping_range+0x71/0x140 kfd_dev_unmap_mapping_range+0x5b/0x140 [amdgpu] amdgpu_amdkfd_clear_kfd_mapping+0xd8/0x190 [amdgpu] amdgpu_device_gpu_recover+0x232/0x450 [amdgpu] xgpu_ai_mailbox_flr_work+0xb5/0xc0 [amdgpu] process_one_work+0x18e/0x3e0 worker_thread+0x2e3/0x420 kthread+0x10a/0x230 Fixes: 70cadefcc616 ("drm/amdgpu: unmap all user mappings of framebuffer and doorbell before mode1 reset") Signed-off-by: Asad Kamal Reviewed-by: Lijo Lazar Signed-off-by: Alex Deucher (cherry picked from commit 1128b4a52de1572e87431de837fd9850cb99542c) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/amdkfd/kfd_chardev.c | 71 +++++++++++++++++++----- 1 file changed, 57 insertions(+), 14 deletions(-) diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c index 504a286368eb..344da6c0e96a 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c @@ -35,6 +35,7 @@ #include #include #include +#include #include #include #include @@ -70,18 +71,54 @@ static const struct class kfd_class = { }; /* - * Cache the address space of the chardev on first open so that the reset - * path can drop all userspace mappings of doorbell and MMIO ranges via - * unmap_mapping_range(). + * Private pseudo-filesystem for KFD, Provides a stable, module-owned + * inode whose address_space is the unmap target for all /dev/kfd + * openers during GPU reset. */ -static struct address_space *kfd_dev_mapping; +static struct vfsmount *kfd_fs_mnt; +static int kfd_fs_cnt; + +static int kfd_fs_init_fs_context(struct fs_context *fc) +{ + return init_pseudo(fc, 0x4b464400 /* "KFD" */) ? 0 : -ENOMEM; +} + +static struct file_system_type kfd_fs_type = { + .name = "kfd", + .init_fs_context = kfd_fs_init_fs_context, + .kill_sb = kill_anon_super, +}; + +static struct inode *kfd_fs_inode_new(void) +{ + struct inode *inode; + int r; + + r = simple_pin_fs(&kfd_fs_type, &kfd_fs_mnt, &kfd_fs_cnt); + if (r < 0) + return ERR_PTR(r); + + inode = alloc_anon_inode(kfd_fs_mnt->mnt_sb); + if (IS_ERR(inode)) + simple_release_fs(&kfd_fs_mnt, &kfd_fs_cnt); + + return inode; +} + +static void kfd_fs_inode_free(struct inode *inode) +{ + if (inode) { + iput(inode); + simple_release_fs(&kfd_fs_mnt, &kfd_fs_cnt); + } +} + +static struct inode *kfd_anon_inode; void kfd_dev_unmap_mapping_range(loff_t const holebegin, loff_t const holelen) { - struct address_space *mapping = READ_ONCE(kfd_dev_mapping); - - if (mapping) - unmap_mapping_range(mapping, holebegin, holelen, 1); + if (kfd_anon_inode) + unmap_mapping_range(kfd_anon_inode->i_mapping, holebegin, holelen, 1); } static inline struct kfd_process_device *kfd_lock_pdd_by_id(struct kfd_process *p, __u32 gpu_id) @@ -107,6 +144,13 @@ int kfd_chardev_init(void) { int err = 0; + kfd_anon_inode = kfd_fs_inode_new(); + if (IS_ERR(kfd_anon_inode)) { + err = PTR_ERR(kfd_anon_inode); + kfd_anon_inode = NULL; + return err; + } + kfd_char_dev_major = register_chrdev(0, kfd_dev_name, &kfd_fops); err = kfd_char_dev_major; if (err < 0) @@ -130,6 +174,8 @@ int kfd_chardev_init(void) err_class_create: unregister_chrdev(kfd_char_dev_major, kfd_dev_name); err_register_chrdev: + kfd_fs_inode_free(kfd_anon_inode); + kfd_anon_inode = NULL; return err; } @@ -138,6 +184,8 @@ void kfd_chardev_exit(void) device_destroy(&kfd_class, MKDEV(kfd_char_dev_major, 0)); class_unregister(&kfd_class); unregister_chrdev(kfd_char_dev_major, kfd_dev_name); + kfd_fs_inode_free(kfd_anon_inode); + kfd_anon_inode = NULL; kfd_device = NULL; } @@ -150,12 +198,7 @@ static int kfd_open(struct inode *inode, struct file *filep) if (iminor(inode) != 0) return -ENODEV; - /* - * /dev/kfd is a single chardev so all opens share one inode. Cache - * its address_space on the first open for use by the reset path. - */ - if (!READ_ONCE(kfd_dev_mapping)) - cmpxchg(&kfd_dev_mapping, NULL, inode->i_mapping); + filep->f_mapping = kfd_anon_inode->i_mapping; is_32bit_user_mode = in_compat_syscall(); From aea841bc62a76242396610d22d8ff40c13065f64 Mon Sep 17 00:00:00 2001 From: Wentao Liang Date: Wed, 16 Sep 2026 10:01:39 +0000 Subject: [PATCH 450/694] drm/amdgpu: Fix vmid_wait fence leak in amdgpu_ring_init() amdgpu_ring_init() initializes ring->vmid_wait with a reference to the stub fence taken via dma_fence_get_stub(). When a later step of the initialization fails, e.g. amdgpu_fence_driver_init_ring(), a writeback slot allocation or the ring buffer allocation, the function returns an error without releasing the stub fence reference and the reference is leaked if the ring is torn down without amdgpu_ring_fini(). Move the stub fence assignment to the end of the initialization, right before the ring is registered with the GPU scheduler, where no further failure is possible. The stub fence is only consumed by command submission handling in amdgpu_ids.c once the ring is up and running, so nothing reads it during the error-prone part of the initialization. Fixes: 48e9fbd1a284 ("drm/amdgpu: initialize the vmid_wait with the stub fence") Signed-off-by: Wentao Liang Signed-off-by: Alex Deucher (cherry picked from commit f2b96986851203e9c50ca0d13aaa3581ca3e8ebd) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c index 686c92e96025..5922406d0a03 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c @@ -254,7 +254,6 @@ int amdgpu_ring_init(struct amdgpu_device *adev, struct amdgpu_ring *ring, ring->adev = adev; ring->num_hw_submission = sched_hw_submission; ring->sched_score = sched_score; - ring->vmid_wait = dma_fence_get_stub(); ring->idx = adev->num_rings++; adev->rings[ring->idx] = ring; @@ -374,6 +373,7 @@ int amdgpu_ring_init(struct amdgpu_device *adev, struct amdgpu_ring *ring, ring->max_dw = max_dw; ring->hw_prio = hw_prio; + ring->vmid_wait = dma_fence_get_stub(); if (!ring->no_scheduler && ring->funcs->type < AMDGPU_HW_IP_NUM) { hw_ip = ring->funcs->type; From a997baa61179b450bd55c4810c7ccfed3b753a54 Mon Sep 17 00:00:00 2001 From: Wentao Liang Date: Wed, 16 Sep 2026 09:54:32 +0000 Subject: [PATCH 451/694] drm/amdgpu: Fix acpi device leak in amdgpu_acpi_enumerate_xcc() amdgpu_acpi_enumerate_xcc() looks up each XCC ACPI device with acpi_dev_get_first_match_dev(), which takes a reference to the device. The reference is dropped with acpi_dev_put() after the XCC info is initialized, but if the kzalloc_obj() allocation of the XCC info fails the function returns -ENOMEM without releasing the reference, leaking the last reference to the ACPI device. Drop the ACPI device reference on the allocation failure path before returning. Fixes: 4d5275ab0b18 ("drm/amdgpu: Add parsing of acpi xcc objects") Reviewed-by: Lijo Lazar Signed-off-by: Wentao Liang Signed-off-by: Alex Deucher (cherry picked from commit 9211ef48b31ec66999cf55e04d0cbc60cd855fd5) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c index 7f5abb03be1b..8b8acf98fdfe 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c @@ -1167,8 +1167,10 @@ int amdgpu_acpi_enumerate_xcc(void) } xcc_info = kzalloc_obj(struct amdgpu_acpi_xcc_info); - if (!xcc_info) + if (!xcc_info) { + acpi_dev_put(acpi_dev); return -ENOMEM; + } INIT_LIST_HEAD(&xcc_info->list); xcc_info->handle = acpi_device_handle(acpi_dev); From b4f7b4459b1b155e4c4977a6482b5df2cf08758c Mon Sep 17 00:00:00 2001 From: Wentao Liang Date: Wed, 16 Sep 2026 09:55:36 +0000 Subject: [PATCH 452/694] drm/amdgpu: Fix last_update fence leak in amdgpu_vm_init() amdgpu_vm_init() initializes vm->last_update, vm->last_unlocked and vm->last_tlb_flush with references to the stub fence taken via dma_fence_get_stub(). The error label at the end of the function releases the last_unlocked and last_tlb_flush references with dma_fence_put(), but the reference stored in vm->last_update is never dropped, so whenever the page table root creation, the reservation of the root BO or the PASID registration fails, the stub fence reference leaks. Drop the vm->last_update reference together with the other stub fence references on the error path. Fixes: 187916e6ed9d ("drm/amdgpu: install stub fence into potential unused fence pointers") Signed-off-by: Wentao Liang Signed-off-by: Alex Deucher (cherry picked from commit e7979c84fc05a176bdf855ee664871b1648404c9) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index bb04101b0fb5..4a63b472f68e 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -2678,6 +2678,7 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm, amdgpu_bo_unref(&root_bo); error_free_delayed: + dma_fence_put(vm->last_update); dma_fence_put(vm->last_tlb_flush); dma_fence_put(vm->last_unlocked); ttm_lru_bulk_move_fini(&adev->mman.bdev, &vm->lru_bulk_move); From 2b86ab1bd6673c525adda88819d7658ba9e784ec Mon Sep 17 00:00:00 2001 From: Wentao Liang Date: Wed, 16 Sep 2026 09:58:05 +0000 Subject: [PATCH 453/694] drm/amdgpu: Fix runtime PM leak in amdgpu_debugfs_test_ib_show() amdgpu_debugfs_test_ib_show() resumes the device with pm_runtime_get_sync() before taking the reset domain semaphore with down_write_killable(). If the write lock acquisition is interrupted, the function returns without calling pm_runtime_put_autosuspend(), leaking the runtime PM reference acquired for the device and keeping the GPU awake. Drop the runtime PM reference on the interrupted down_write_killable() error path before returning. Fixes: 6049db43d6dd ("drm/amdgpu: change reset lock from mutex to rw_semaphore") Signed-off-by: Wentao Liang Signed-off-by: Alex Deucher (cherry picked from commit ec30a576c2d4c0364549e6c04218f50704ef56c8) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c index 132d054900b5..aca1a8045afa 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c @@ -1780,8 +1780,10 @@ static int amdgpu_debugfs_test_ib_show(struct seq_file *m, void *unused) /* Avoid accidently unparking the sched thread during GPU reset */ r = down_write_killable(&adev->reset_domain->sem); - if (r) + if (r) { + pm_runtime_put_autosuspend(dev->dev); return r; + } /* hold on the scheduler */ for (i = 0; i < AMDGPU_MAX_RINGS; i++) { From 0fd5e9ddf362b1253b3c94b858371f90400e5c4a Mon Sep 17 00:00:00 2001 From: Alex Hung Date: Wed, 23 Sep 2026 11:31:18 -0600 Subject: [PATCH 454/694] drm/amd/display: Relax DML frame limit with UBSAN [WHY] UBSAN instrumentation adds checks and handler calls and increases stack usage in the large DML calculation functions, similar to KASAN and KCSAN. With UBSAN enabled these files exceed the default -Wframe-larger-than limit and fail to build when -Werror is in effect. Reproduced with LLVM (make LLVM=1, clang 19.1.1), CONFIG_UBSAN=y, CONFIG_GCOV_PROFILE_ALL=y and CONFIG_DRM_AMDGPU_WERROR=y on x86_64. [HOW] Include CONFIG_UBSAN in the sanitizer check that selects the higher per-file frame warning limit in the dml and dml2_0 Makefiles. Suggested-by: Leo Li Assisted-by: Copilot:Claude-Opus-5.5 Signed-off-by: Alex Hung Signed-off-by: Alex Deucher (cherry picked from commit ebf8b0fd8508b744f85a8eee82b745b1d3502dd0) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/display/dc/dml/Makefile | 2 +- drivers/gpu/drm/amd/display/dc/dml2_0/Makefile | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/amd/display/dc/dml/Makefile b/drivers/gpu/drm/amd/display/dc/dml/Makefile index 10d4ace04d4f..096520bd27c6 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/Makefile +++ b/drivers/gpu/drm/amd/display/dc/dml/Makefile @@ -29,7 +29,7 @@ dml_ccflags := $(CC_FLAGS_FPU) dml_rcflags := $(CC_FLAGS_NO_FPU) ifneq ($(CONFIG_FRAME_WARN),0) - ifeq ($(filter y,$(CONFIG_KASAN)$(CONFIG_KCSAN)),y) + ifeq ($(filter y,$(CONFIG_KASAN)$(CONFIG_KCSAN)$(CONFIG_UBSAN)),y) ifeq ($(CONFIG_CC_IS_CLANG)$(CONFIG_COMPILE_TEST),yy) frame_warn_limit := 4096 else diff --git a/drivers/gpu/drm/amd/display/dc/dml2_0/Makefile b/drivers/gpu/drm/amd/display/dc/dml2_0/Makefile index 5388bf094fbc..4d682e92df1d 100644 --- a/drivers/gpu/drm/amd/display/dc/dml2_0/Makefile +++ b/drivers/gpu/drm/amd/display/dc/dml2_0/Makefile @@ -28,7 +28,7 @@ dml2_ccflags := $(CC_FLAGS_FPU) dml2_rcflags := $(CC_FLAGS_NO_FPU) ifneq ($(CONFIG_FRAME_WARN),0) - ifeq ($(filter y,$(CONFIG_KASAN)$(CONFIG_KCSAN)),y) + ifeq ($(filter y,$(CONFIG_KASAN)$(CONFIG_KCSAN)$(CONFIG_UBSAN)),y) ifeq ($(CONFIG_CC_IS_CLANG)$(CONFIG_COMPILE_TEST),yy) frame_warn_limit := 4096 else From 18779dd84515db093fedb4ebaf0998c9b165a5fb Mon Sep 17 00:00:00 2001 From: Ivan Lipski Date: Fri, 21 Aug 2026 00:04:53 -0400 Subject: [PATCH 455/694] drm/amd/display: Bump frame warning limit for clang builds of dml [Why&How] When building the DML files with clang without any sanitizer or LTO, the following -Wframe-larger-than errors break the build under CONFIG_WERROR: display_mode_vba_30.c: error: stack frame size (2512) exceeds limit (2048) in 'dml30_ModeSupportAndSystemConfigurationFull' display_mode_vba_31.c: error: stack frame size (2416) exceeds limit (2048) in 'dml31_ModeSupportAndSystemConfigurationFull' display_mode_vba_314.c: error: stack frame size (2392) exceeds limit (2048) in 'dml314_ModeSupportAndSystemConfigurationFull' Clang consistently spills more than gcc, pushing the frame past the 2048 byte limit. Apply an existing approach of increasing the warn stack size to the non-sanitizer path so plain clang builds use a 3072 byte limit. Closes: https://gitlab.freedesktop.org/drm/amd/-/work_items/5642 Signed-off-by: Ivan Lipski Signed-off-by: Alex Deucher (cherry picked from commit 21711b6e66bb7b41b1aec67b2d99aafe768c8fcb) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/display/dc/dml/Makefile | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/display/dc/dml/Makefile b/drivers/gpu/drm/amd/display/dc/dml/Makefile index 096520bd27c6..79eeb3721b99 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/Makefile +++ b/drivers/gpu/drm/amd/display/dc/dml/Makefile @@ -36,7 +36,11 @@ ifneq ($(CONFIG_FRAME_WARN),0) frame_warn_limit := 3072 endif else - frame_warn_limit := 2048 + ifeq ($(CONFIG_CC_IS_CLANG),y) + frame_warn_limit := 3072 + else + frame_warn_limit := 2048 + endif endif ifeq ($(call test-lt, $(CONFIG_FRAME_WARN), $(frame_warn_limit)),y) From 686f942332b1667f13f3b8d6a2f50bcfbf42e277 Mon Sep 17 00:00:00 2001 From: Pengpeng Hou Date: Wed, 15 Jul 2026 16:43:25 +0800 Subject: [PATCH 456/694] nfc: nfcmrvl: validate helper command length before pull The firmware download receive path removes the NCI data header and reads the helper command before validating the remaining packet length. A short frame can therefore reach the data access before the malformed packet is rejected. Validate the complete helper command length before stripping the NCI data header. Fixes: 3194c6870158 ("NFC: nfcmrvl: add firmware download support") Signed-off-by: Pengpeng Hou Link: https://patch.msgid.link/20260715084325.40276-1-pengpeng@iscas.ac.cn Signed-off-by: David Heidelberg --- drivers/nfc/nfcmrvl/fw_dnld.c | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-) diff --git a/drivers/nfc/nfcmrvl/fw_dnld.c b/drivers/nfc/nfcmrvl/fw_dnld.c index 2b8f401d8fd7..8b9d5257320d 100644 --- a/drivers/nfc/nfcmrvl/fw_dnld.c +++ b/drivers/nfc/nfcmrvl/fw_dnld.c @@ -263,9 +263,14 @@ static int process_state_fw_dnld(struct nfcmrvl_private *priv, * B8..N: payload */ - /* Remove NCI HDR */ - skb_pull(skb, 3); - if (skb->data[0] != HELPER_CMD_PACKET_FORMAT || skb->len != 5) { + if (skb->len != NCI_DATA_HDR_SIZE + 5) { + nfc_err(priv->dev, "bad command"); + return -EINVAL; + } + + /* Remove NCI header */ + skb_pull(skb, NCI_DATA_HDR_SIZE); + if (skb->data[0] != HELPER_CMD_PACKET_FORMAT) { nfc_err(priv->dev, "bad command"); return -EINVAL; } From a653c01ce447f10c36b901646888c0330363af4f Mon Sep 17 00:00:00 2001 From: Pengpeng Hou Date: Wed, 15 Jul 2026 16:44:05 +0800 Subject: [PATCH 457/694] nfc: st21nfca: validate received frame size st21nfca_hci_i2c_repack() trims a received frame at its EOF marker before removing byte stuffing. It then assumes the truncated frame contains the LLC header and two CRC bytes, and it unconditionally reads the byte after an escape marker. A malformed frame can place EOF immediately after the start marker or can end its data portion with an escape marker. The former leaves too few bytes for check_crc(), while the latter makes the unstuffing loop read past the current skb length. Require the minimum framing bytes both before and after unstuffing. Use separate input and output cursors while removing byte stuffing, and reject an escape marker without its encoded byte. This keeps malformed frames within the received frame boundary before CRC processing. Fixes: 3096e25a3e40 ("NFC: st21nfca: Fix incorrect byte stuffing revocation") Signed-off-by: Pengpeng Hou Link: https://patch.msgid.link/20260715084405.41546-1-pengpeng@iscas.ac.cn Signed-off-by: David Heidelberg --- drivers/nfc/st21nfca/i2c.c | 29 +++++++++++++++++++---------- 1 file changed, 19 insertions(+), 10 deletions(-) diff --git a/drivers/nfc/st21nfca/i2c.c b/drivers/nfc/st21nfca/i2c.c index a4c93ff7c5b0..0f44c783bd04 100644 --- a/drivers/nfc/st21nfca/i2c.c +++ b/drivers/nfc/st21nfca/i2c.c @@ -289,27 +289,36 @@ static int check_crc(u8 *buf, int buflen) */ static int st21nfca_hci_i2c_repack(struct sk_buff *skb) { - int i, j, r, size; + int read, write, r, size; - if (skb->len < 1 || (skb->len > 1 && skb->data[1] != 0)) + if (skb->len < ST21NFCA_FRAME_HEADROOM || + !IS_START_OF_FRAME(skb->data)) return -EBADMSG; size = get_frame_size(skb->data, skb->len); if (size > 0) { + if (size < ST21NFCA_FRAME_HEADROOM + 2) + return -EBADMSG; + skb_trim(skb, size); /* remove ST21NFCA byte stuffing for upper layer */ - for (i = 1, j = 0; i < skb->len; i++) { - if (skb->data[i + j] == + for (read = 1, write = 1; read < skb->len;) { + if (skb->data[read] == (u8) ST21NFCA_ESCAPE_BYTE_STUFFING) { - skb->data[i] = skb->data[i + j + 1] - | ST21NFCA_BYTE_STUFFING_MASK; - i++; - j++; + if (read + 1 == skb->len) + return -EBADMSG; + + skb->data[write++] = skb->data[read + 1] + | ST21NFCA_BYTE_STUFFING_MASK; + read += 2; + } else { + skb->data[write++] = skb->data[read++]; } - skb->data[i] = skb->data[i + j]; } /* remove byte stuffing useless byte */ - skb_trim(skb, i - j); + skb_trim(skb, write); + if (skb->len < ST21NFCA_FRAME_HEADROOM + 2) + return -EBADMSG; /* remove ST21NFCA_SOF_EOF from head */ skb_pull(skb, 1); From bf1460acdf8cf5a07c819f59785d40f20d113099 Mon Sep 17 00:00:00 2001 From: Aldo Ariel Panzardo Date: Thu, 16 Jul 2026 20:26:57 -0300 Subject: [PATCH 458/694] nfc: llcp: Fix list corruption / refcount desync in nfc_llcp_recv_dm() nfc_llcp_recv_dm() handles DM(NOBOUND)/DM(REJ) for a socket that is still linked on local->connecting_sockets: it looks the socket up with nfc_llcp_connecting_sock_get(), sets sk->sk_state = LLCP_CLOSED and returns, without taking the socket lock and without unlinking the socket from the connecting_sockets list. llcp_sock_release() selects the list to unlink from by sk_state: a socket in LLCP_CONNECTING is unlinked from connecting_sockets, otherwise from the sockets list. Because recv_dm left the socket physically on connecting_sockets but in the LLCP_CLOSED state, release() takes the else branch and calls nfc_llcp_sock_unlink(&local->sockets, sk). That runs sk_del_node_init() while holding sockets.lock, i.e. it removes the socket from the connecting_sockets hlist under the wrong lock. A concurrent connect() linking another socket onto connecting_sockets under connecting_sockets.lock then mutates the same hlist unserialized, which corrupts the list and desyncs the sk_add_node()/sk_del_node_init() sock_hold()/__sock_put() pairing. An unprivileged local process holding LLCP sockets, with the DM supplied by the remote peer over an established LLCP link, can drive this to leak kernel sockets without bound (the mis-decrement goes through the non-freeing __sock_put() path, so the object is never released), leading to memory exhaustion / DoS. This is the same class of bug that was fixed in the sibling handler nfc_llcp_recv_cc() by commit b493ea2765cc ("nfc: llcp: Fix use-after-free race in nfc_llcp_recv_cc()"); recv_dm did not receive the equivalent fix. Fix it the same way: take lock_sock(), re-check that the socket is still hashed (release() may have won the race), and for the NOBOUND/REJ case unlink it from connecting_sockets before moving it to LLCP_CLOSED. The unlink drops the connecting_sockets membership reference via sk_del_node_init(), leaving the socket unhashed, so the later nfc_llcp_sock_unlink() in llcp_sock_release() becomes a no-op and no double put occurs. Fixes: a69f32af86e3 ("NFC: Socket linked list") Signed-off-by: Aldo Ariel Panzardo Link: https://patch.msgid.link/20260716232657.203145-1-qwe.aldo@gmail.com Signed-off-by: David Heidelberg --- net/nfc/llcp_core.c | 25 +++++++++++++++++++++++++ 1 file changed, 25 insertions(+) diff --git a/net/nfc/llcp_core.c b/net/nfc/llcp_core.c index cac1b5487064..bd6361e2efa4 100644 --- a/net/nfc/llcp_core.c +++ b/net/nfc/llcp_core.c @@ -1251,6 +1251,7 @@ static void nfc_llcp_recv_dm(struct nfc_llcp_local *local, struct nfc_llcp_sock *llcp_sock; struct sock *sk; u8 dsap, ssap, reason; + bool connecting = false; dsap = nfc_llcp_dsap(skb); ssap = nfc_llcp_ssap(skb); @@ -1262,6 +1263,7 @@ static void nfc_llcp_recv_dm(struct nfc_llcp_local *local, case LLCP_DM_NOBOUND: case LLCP_DM_REJ: llcp_sock = nfc_llcp_connecting_sock_get(local, dsap); + connecting = true; break; default: @@ -1276,10 +1278,33 @@ static void nfc_llcp_recv_dm(struct nfc_llcp_local *local, sk = &llcp_sock->sk; + lock_sock(sk); + + /* Check if socket was destroyed whilst waiting for the lock */ + if (!sk_hashed(sk)) { + release_sock(sk); + nfc_llcp_sock_put(llcp_sock); + return; + } + + /* + * For DM(NOBOUND)/DM(REJ) the socket is still linked on the + * connecting_sockets list. Unlink it here, under the socket lock, + * before moving it to LLCP_CLOSED: llcp_sock_release() selects the + * list to unlink from by sk_state, so leaving a connecting socket + * in the CLOSED state would make it unlink from the wrong list and + * corrupt the connecting_sockets list / desync the socket refcount. + * This mirrors nfc_llcp_recv_cc(). + */ + if (connecting) + nfc_llcp_sock_unlink(&local->connecting_sockets, sk); + sk->sk_err = ENXIO; sk->sk_state = LLCP_CLOSED; sk->sk_state_change(sk); + release_sock(sk); + nfc_llcp_sock_put(llcp_sock); } From 092c6a605cbd6414ef499834c2e0da69c2c3388e Mon Sep 17 00:00:00 2001 From: Doruk Tan Ozturk Date: Sat, 11 Jul 2026 14:36:51 +0200 Subject: [PATCH 459/694] nfc: port100: reject frames whose declared length exceeds the received data port100_recv_response() passes the URB transfer buffer to port100_rx_frame_is_valid(), which checksums le16_to_cpu(frame->datalen) bytes of frame->data. datalen is a 16-bit field supplied by the device and is never checked against the number of bytes actually received (urb->actual_length), so a device reporting a datalen larger than the received frame makes port100_data_checksum() read out of bounds past the transfer buffer. Reject a response whose declared frame size does not fit the received length before validating it. Found by 0sec (https://0sec.ai) using automated source analysis; the missing bound is evident from source. Compile-tested. Fixes: 562d4d59b8a1 ("NFC: Sony Port-100 Series driver") Cc: stable@vger.kernel.org Assisted-by: 0sec:claude-opus-4-8 Signed-off-by: Doruk Tan Ozturk Reviewed-by: Simon Horman Link: https://patch.msgid.link/20260711123651.32595-1-doruk@0sec.ai Signed-off-by: David Heidelberg --- drivers/nfc/port100.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/drivers/nfc/port100.c b/drivers/nfc/port100.c index b613f5e2fd57..e769a30b8b5c 100644 --- a/drivers/nfc/port100.c +++ b/drivers/nfc/port100.c @@ -636,6 +636,13 @@ static void port100_recv_response(struct urb *urb) in_frame = dev->in_urb->transfer_buffer; + if (urb->actual_length < PORT100_FRAME_HEADER_LEN || + urb->actual_length < port100_rx_frame_size(in_frame)) { + nfc_err(&dev->interface->dev, "Received a truncated frame\n"); + cmd->status = -EIO; + goto sched_wq; + } + if (!port100_rx_frame_is_valid(in_frame)) { nfc_err(&dev->interface->dev, "Received an invalid frame\n"); cmd->status = -EIO; From 3d8afc5243ea2ee803d98e69eb4a01748167ac1b Mon Sep 17 00:00:00 2001 From: Lei Zhu Date: Wed, 29 Jul 2026 15:24:26 +0800 Subject: [PATCH 460/694] selftests: nci: Correct pthread_create return value check The pthread_create() functions returns 0 on success and a positive value on failure. Modify the return value check to correctly detect failure cases. Fixes: 72696bd8a09d ("selftests: nci: Extract the start/stop discovery function") Signed-off-by: Lei Zhu Link: https://patch.msgid.link/20260729072426.303484-1-zhulei_szu@163.com Signed-off-by: David Heidelberg --- tools/testing/selftests/nci/nci_dev.c | 12 +++++++----- 1 file changed, 7 insertions(+), 5 deletions(-) diff --git a/tools/testing/selftests/nci/nci_dev.c b/tools/testing/selftests/nci/nci_dev.c index 312f84ee0444..c053f5cf2745 100644 --- a/tools/testing/selftests/nci/nci_dev.c +++ b/tools/testing/selftests/nci/nci_dev.c @@ -438,7 +438,7 @@ FIXTURE_SETUP(NCI) else rc = pthread_create(&thread_t, NULL, virtual_dev_open, (void *)&self->virtual_nci_fd); - ASSERT_GT(rc, -1); + ASSERT_EQ(rc, 0); rc = send_cmd_with_idx(self->sd, self->fid, self->pid, NFC_CMD_DEV_UP, self->dev_idex); @@ -509,7 +509,7 @@ FIXTURE_TEARDOWN(NCI) rc = pthread_create(&thread_t, NULL, virtual_deinit, (void *)&self->virtual_nci_fd); - ASSERT_GT(rc, -1); + ASSERT_EQ(rc, 0); rc = send_cmd_with_idx(self->sd, self->fid, self->pid, NFC_CMD_DEV_DOWN, self->dev_idex); EXPECT_EQ(rc, 0); @@ -590,7 +590,7 @@ int start_polling(int dev_idx, int proto, int virtual_fd, int sd, int fid, int p rc = pthread_create(&thread_t, NULL, virtual_poll_start, (void *)&virtual_fd); - if (rc < 0) + if (rc) return rc; rc = send_cmd_mt_nla(sd, fid, pid, NFC_CMD_START_POLL, 2, nla_start_poll_type, @@ -610,7 +610,7 @@ int stop_polling(int dev_idx, int virtual_fd, int sd, int fid, int pid) rc = pthread_create(&thread_t, NULL, virtual_poll_stop, (void *)&virtual_fd); - if (rc < 0) + if (rc) return rc; rc = send_cmd_with_idx(sd, fid, pid, @@ -830,6 +830,8 @@ int disconnect_tag(int nfc_sock, int virtual_fd) status = pthread_create(&thread_t, NULL, virtual_deactivate_proc, (void *)&virtual_fd); + if (status) + return status; close(nfc_sock); pthread_join(thread_t, (void **)&status); @@ -874,7 +876,7 @@ TEST_F(NCI, deinit) else rc = pthread_create(&thread_t, NULL, virtual_deinit, (void *)&self->virtual_nci_fd); - ASSERT_GT(rc, -1); + ASSERT_EQ(rc, 0); rc = send_cmd_with_idx(self->sd, self->fid, self->pid, NFC_CMD_DEV_DOWN, self->dev_idex); From c3eef2f988a3db9690369d7cef9a3344dd9788d3 Mon Sep 17 00:00:00 2001 From: Lee Jones Date: Wed, 2 Sep 2026 12:30:31 +0000 Subject: [PATCH 461/694] nfc: llcp: Fix race condition in accept_queue lifecycle In nfc_llcp_socket_release(), sockets and listener accept queues are walked under the local sockets rwlock and bh_lock_sock(). However, bh_lock_sock() does not synchronise against process-context lock_sock() held by nfc_llcp_accept_dequeue() during accept(). Because socket_release() does not check sock_owned_by_user(), both paths can concurrently unlink and release the same child socket, resulting in use-after-free or a NULL pointer dereference of child->parent in nfc_llcp_accept_unlink(). Fix this synchronisation race by having nfc_llcp_socket_release() use process-context lock_sock() instead of bh_lock_sock(): 1. Pop sockets from the local sockets list under the write lock using nfc_llcp_sock_list_pop() so lock_sock() can be acquired without holding the rwlock. 2. Because lock_sock() can sleep, defer the final release of the nfc_llcp_local structure to a workqueue (release_work). This avoids a sleeping-in-atomic bug when the last local reference is dropped from softirq context. Additionally, hold a single device reference on local from registration until final destruction. 3. In nfc_llcp_local_get(), use kref_get_unless_zero() to prevent resurrecting a local object whose teardown has been scheduled. 4. In llcp_sock_accept(), verify that the listener socket state is still LLCP_LISTEN after waking from schedule_timeout() to prevent hangs if the listener is closed concurrently. 5. When unlinking unaccepted child sockets during listener release, unlink them from local->sockets, call sock_orphan(), and drop their initial sk_alloc creation reference via sock_put(). 6. Make nfc_llcp_accept_unlink() idempotent by guarding parent access with a NULL check. Fixes: 50b78b2a6500 ("NFC: Fix sleeping in atomic when releasing socket") Signed-off-by: Lee Jones Link: https://patch.msgid.link/20260902123033.1169067-1-lee@kernel.org Signed-off-by: David Heidelberg --- net/nfc/llcp.h | 1 + net/nfc/llcp_core.c | 123 +++++++++++++++++++++++++++----------------- net/nfc/llcp_sock.c | 51 +++++++++++++----- 3 files changed, 116 insertions(+), 59 deletions(-) diff --git a/net/nfc/llcp.h b/net/nfc/llcp.h index d8345ed57c95..23ae7a0112d3 100644 --- a/net/nfc/llcp.h +++ b/net/nfc/llcp.h @@ -91,6 +91,7 @@ struct nfc_llcp_local { struct hlist_head pending_sdreqs; struct timer_list sdreq_timer; struct work_struct sdreq_timeout_work; + struct work_struct release_work; u8 sdreq_next_tid; /* sockets array */ diff --git a/net/nfc/llcp_core.c b/net/nfc/llcp_core.c index bd6361e2efa4..23553e7426ec 100644 --- a/net/nfc/llcp_core.c +++ b/net/nfc/llcp_core.c @@ -20,6 +20,8 @@ static LIST_HEAD(llcp_devices); /* Protects llcp_devices list */ static DEFINE_SPINLOCK(llcp_devices_lock); +static struct workqueue_struct *llcp_wq; + static void nfc_llcp_rx_skb(struct nfc_llcp_local *local, struct sk_buff *skb); void nfc_llcp_sock_link(struct llcp_sock_list *l, struct sock *sk) @@ -63,21 +65,33 @@ static void nfc_llcp_socket_purge(struct nfc_llcp_sock *sock) } } +static struct sock *nfc_llcp_sock_list_pop(struct llcp_sock_list *l) +{ + struct sock *sk; + + write_lock(&l->lock); + sk = sk_head(&l->head); + if (sk) { + sock_hold(sk); + sk_del_node_init(sk); + } + write_unlock(&l->lock); + + return sk; +} + static void nfc_llcp_socket_release(struct nfc_llcp_local *local, bool device, int err) { struct sock *sk; - struct hlist_node *tmp; struct nfc_llcp_sock *llcp_sock; skb_queue_purge(&local->tx_queue); - write_lock(&local->sockets.lock); - - sk_for_each_safe(sk, tmp, &local->sockets.head) { + while ((sk = nfc_llcp_sock_list_pop(&local->sockets))) { llcp_sock = nfc_llcp_sock(sk); - bh_lock_sock(sk); + lock_sock(sk); nfc_llcp_socket_purge(llcp_sock); @@ -91,17 +105,27 @@ static void nfc_llcp_socket_release(struct nfc_llcp_local *local, bool device, list_for_each_entry_safe(lsk, n, &llcp_sock->accept_queue, accept_queue) { + bool put_creation = false; + accept_sk = &lsk->sk; - bh_lock_sock(accept_sk); + lock_sock_nested(accept_sk, + SINGLE_DEPTH_NESTING); - nfc_llcp_accept_unlink(accept_sk); + if (nfc_llcp_sock(accept_sk)->parent == sk) { + nfc_llcp_accept_unlink(accept_sk); + nfc_llcp_sock_unlink(&local->sockets, accept_sk); - if (err) - accept_sk->sk_err = err; - accept_sk->sk_state = LLCP_CLOSED; - accept_sk->sk_state_change(sk); + if (err) + accept_sk->sk_err = err; + accept_sk->sk_state = LLCP_CLOSED; + accept_sk->sk_state_change(accept_sk); + sock_orphan(accept_sk); + put_creation = true; + } - bh_unlock_sock(accept_sk); + release_sock(accept_sk); + if (put_creation) + sock_put(accept_sk); /* creation ref */ } } @@ -110,23 +134,18 @@ static void nfc_llcp_socket_release(struct nfc_llcp_local *local, bool device, sk->sk_state = LLCP_CLOSED; sk->sk_state_change(sk); - bh_unlock_sock(sk); - - sk_del_node_init(sk); + release_sock(sk); + sock_put(sk); } - write_unlock(&local->sockets.lock); - /* If we still have a device, we keep the RAW sockets alive */ if (device == true) return; - write_lock(&local->raw_sockets.lock); - - sk_for_each_safe(sk, tmp, &local->raw_sockets.head) { + while ((sk = nfc_llcp_sock_list_pop(&local->raw_sockets))) { llcp_sock = nfc_llcp_sock(sk); - bh_lock_sock(sk); + lock_sock(sk); nfc_llcp_socket_purge(llcp_sock); @@ -135,26 +154,20 @@ static void nfc_llcp_socket_release(struct nfc_llcp_local *local, bool device, sk->sk_state = LLCP_CLOSED; sk->sk_state_change(sk); - bh_unlock_sock(sk); - - sk_del_node_init(sk); + release_sock(sk); + sock_put(sk); } - - write_unlock(&local->raw_sockets.lock); } static struct nfc_llcp_local *nfc_llcp_local_get(struct nfc_llcp_local *local) { - /* Since using nfc_llcp_local may result in usage of nfc_dev, whenever - * we hold a reference to local, we also need to hold a reference to - * the device to avoid UAF. - */ - if (!nfc_get_device(local->dev->idx)) + if (!local) return NULL; - kref_get(&local->ref); + if (kref_get_unless_zero(&local->ref)) + return local; - return local; + return NULL; } static void local_cleanup(struct nfc_llcp_local *local) @@ -172,30 +185,34 @@ static void local_cleanup(struct nfc_llcp_local *local) nfc_llcp_free_sdp_tlv_list(&local->pending_sdreqs); } +static void local_release_work(struct work_struct *work) +{ + struct nfc_llcp_local *local; + struct nfc_dev *dev; + + local = container_of(work, struct nfc_llcp_local, release_work); + dev = local->dev; + + local_cleanup(local); + kfree(local); + nfc_put_device(dev); +} + static void local_release(struct kref *ref) { struct nfc_llcp_local *local; local = container_of(ref, struct nfc_llcp_local, ref); - local_cleanup(local); - kfree(local); + queue_work(llcp_wq, &local->release_work); } int nfc_llcp_local_put(struct nfc_llcp_local *local) { - struct nfc_dev *dev; - int ret; - - if (local == NULL) + if (!local) return 0; - dev = local->dev; - - ret = kref_put(&local->ref, local_release); - nfc_put_device(dev); - - return ret; + return kref_put(&local->ref, local_release); } static struct nfc_llcp_sock *nfc_llcp_sock_get(struct nfc_llcp_local *local, @@ -1705,6 +1722,7 @@ int nfc_llcp_register_device(struct nfc_dev *ndev) INIT_WORK(&local->rx_work, nfc_llcp_rx_work); INIT_WORK(&local->timeout_work, nfc_llcp_timeout_work); + INIT_WORK(&local->release_work, local_release_work); rwlock_init(&local->sockets.lock); rwlock_init(&local->connecting_sockets.lock); @@ -1748,10 +1766,23 @@ void nfc_llcp_unregister_device(struct nfc_dev *dev) int __init nfc_llcp_init(void) { - return nfc_llcp_sock_init(); + int ret; + + llcp_wq = alloc_workqueue("nfc_llcp_wq", WQ_UNBOUND, 0); + if (!llcp_wq) + return -ENOMEM; + + ret = nfc_llcp_sock_init(); + if (ret) { + destroy_workqueue(llcp_wq); + return ret; + } + + return 0; } void nfc_llcp_exit(void) { nfc_llcp_sock_exit(); + destroy_workqueue(llcp_wq); } diff --git a/net/nfc/llcp_sock.c b/net/nfc/llcp_sock.c index 5558d8a4d48b..ce6875eb58fb 100644 --- a/net/nfc/llcp_sock.c +++ b/net/nfc/llcp_sock.c @@ -392,11 +392,12 @@ void nfc_llcp_accept_unlink(struct sock *sk) pr_debug("state %d\n", sk->sk_state); - list_del_init(&llcp_sock->accept_queue); - sk_acceptq_removed(llcp_sock->parent); - llcp_sock->parent = NULL; - - sock_put(sk); + if (llcp_sock->parent) { + list_del_init(&llcp_sock->accept_queue); + sk_acceptq_removed(llcp_sock->parent); + llcp_sock->parent = NULL; + sock_put(sk); + } } void nfc_llcp_accept_enqueue(struct sock *parent, struct sock *sk) @@ -423,12 +424,20 @@ struct sock *nfc_llcp_accept_dequeue(struct sock *parent, list_for_each_entry_safe(lsk, n, &llcp_parent->accept_queue, accept_queue) { + struct nfc_llcp_local *local; + sk = &lsk->sk; - lock_sock(sk); + lock_sock_nested(sk, SINGLE_DEPTH_NESTING); if (sk->sk_state == LLCP_CLOSED) { - release_sock(sk); + local = nfc_llcp_sock(sk)->local; + nfc_llcp_accept_unlink(sk); + if (local) + nfc_llcp_sock_unlink(&local->sockets, sk); + sock_orphan(sk); + release_sock(sk); + sock_put(sk); continue; } @@ -464,7 +473,7 @@ static int llcp_sock_accept(struct socket *sock, struct socket *newsock, pr_debug("parent %p\n", sk); - lock_sock_nested(sk, SINGLE_DEPTH_NESTING); + lock_sock(sk); if (sk->sk_state != LLCP_LISTEN) { ret = -EBADFD; @@ -490,7 +499,12 @@ static int llcp_sock_accept(struct socket *sock, struct socket *newsock, release_sock(sk); timeo = schedule_timeout(timeo); - lock_sock_nested(sk, SINGLE_DEPTH_NESTING); + lock_sock(sk); + + if (sk->sk_state != LLCP_LISTEN) { + ret = -EBADFD; + break; + } } __set_current_state(TASK_RUNNING); remove_wait_queue(sk_sleep(sk), &wait); @@ -629,13 +643,24 @@ static int llcp_sock_release(struct socket *sock) list_for_each_entry_safe(lsk, n, &llcp_sock->accept_queue, accept_queue) { - accept_sk = &lsk->sk; - lock_sock(accept_sk); + bool put_creation = false; - nfc_llcp_send_disconnect(lsk); - nfc_llcp_accept_unlink(accept_sk); + accept_sk = &lsk->sk; + lock_sock_nested(accept_sk, SINGLE_DEPTH_NESTING); + + if (nfc_llcp_sock(accept_sk)->parent == sk) { + nfc_llcp_send_disconnect(lsk); + nfc_llcp_accept_unlink(accept_sk); + nfc_llcp_sock_unlink(&local->sockets, accept_sk); + + accept_sk->sk_state = LLCP_CLOSED; + sock_orphan(accept_sk); + put_creation = true; + } release_sock(accept_sk); + if (put_creation) + sock_put(accept_sk); /* creation ref */ } } From eda518d2cdb6074a0bcdfa06af291616bcb5c421 Mon Sep 17 00:00:00 2001 From: Chaithanya Lagisetty Date: Tue, 1 Sep 2026 07:06:18 +0000 Subject: [PATCH 462/694] selftests: nci: Fix uninitialized family ID on missing attribute get_family_id() walks the generic netlink CTRL_CMD_GETFAMILY reply looking for the CTRL_ATTR_FAMILY_ID attribute and returns the parsed value in the local variable "id". If the reply does not carry that attribute, the parsing loop never assigns "id" and the function returns an indeterminate stack value, which the caller stores in self->fid and uses for subsequent netlink requests. Initialize "id" to 0 so a missing attribute yields a deterministic (invalid) family ID instead of a garbage value. Fixes: f595cf1242f3 ("selftests: Add nci suite") Signed-off-by: Chaithanya Lagisetty Reviewed-by: Hangbin Liu Link: https://patch.msgid.link/20260901070618.3299012-1-nagachaithanya9911@gmail.com Signed-off-by: David Heidelberg --- tools/testing/selftests/nci/nci_dev.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/testing/selftests/nci/nci_dev.c b/tools/testing/selftests/nci/nci_dev.c index c053f5cf2745..23fd38acfcf4 100644 --- a/tools/testing/selftests/nci/nci_dev.c +++ b/tools/testing/selftests/nci/nci_dev.c @@ -182,7 +182,7 @@ static int get_family_id(int sd, __u32 pid, __u32 *event_group) } ans; struct nlattr *na; int resp_len; - __u16 id; + __u16 id = 0; int len; int rc; From 6be581aeffc215bfc77939cd59902b0dbc4af23e Mon Sep 17 00:00:00 2001 From: Chris Gellermann Date: Fri, 4 Sep 2026 11:59:15 +0200 Subject: [PATCH 463/694] selftests/nci: Fix out-of-bounds store on thread join The NCI test collects the exit status of its helper threads by passing the address of an int to pthread_join(): int status; ... pthread_join(thread_t, (void **) &status); pthread_join() stores a void pointer to the memory location. On 64-bit systems, a void pointer is wider than an int, so the store overruns the 4 bytes of space allocated on the stack for the integer and corrupts the adjacent stack. On our CHERI system, this caused a fault due to a capability bounds violation. Fix this by introducing a helper that joins a thread through a void pointer and converts the result back to an integer, which is what the helper threads return. While here, also fix the logic in disconnect_tag() if the helper thread creation failed. Previously, it would have joined a thread that was never created when pthread_create() failed. Fixes: f595cf1242f3 ("selftests: Add nci suite") Signed-off-by: Chris Gellermann Reviewed-by: Simon Horman Link: https://patch.msgid.link/20260904095915.3372241-1-christian.gellermann@codasip.com Signed-off-by: David Heidelberg --- tools/testing/selftests/nci/nci_dev.c | 31 +++++++++++++++++---------- 1 file changed, 20 insertions(+), 11 deletions(-) diff --git a/tools/testing/selftests/nci/nci_dev.c b/tools/testing/selftests/nci/nci_dev.c index 23fd38acfcf4..07427fa42888 100644 --- a/tools/testing/selftests/nci/nci_dev.c +++ b/tools/testing/selftests/nci/nci_dev.c @@ -8,6 +8,7 @@ #include #include +#include #include #include #include @@ -87,6 +88,16 @@ struct msgtemplate { char buf[MAX_MSG_SIZE]; }; +static int join_thread_status(pthread_t thread) +{ + void *thread_ret = NULL; + + if (pthread_join(thread, &thread_ret)) + return -1; + + return (int)(intptr_t)thread_ret; +} + static int create_nl_socket(void) { int fd; @@ -444,7 +455,7 @@ FIXTURE_SETUP(NCI) NFC_CMD_DEV_UP, self->dev_idex); EXPECT_EQ(rc, 0); - pthread_join(thread_t, (void **)&status); + status = join_thread_status(thread_t); ASSERT_EQ(status, 0); self->open_state = true; } @@ -514,7 +525,7 @@ FIXTURE_TEARDOWN(NCI) NFC_CMD_DEV_DOWN, self->dev_idex); EXPECT_EQ(rc, 0); - pthread_join(thread_t, (void **)&status); + status = join_thread_status(thread_t); ASSERT_EQ(status, 0); } @@ -585,7 +596,6 @@ int start_polling(int dev_idx, int proto, int virtual_fd, int sd, int fid, int p void *nla_start_poll_data[2] = {&dev_idx, &proto}; int nla_start_poll_len[2] = {4, 4}; pthread_t thread_t; - int status; int rc; rc = pthread_create(&thread_t, NULL, virtual_poll_start, @@ -598,14 +608,12 @@ int start_polling(int dev_idx, int proto, int virtual_fd, int sd, int fid, int p if (rc != 0) return rc; - pthread_join(thread_t, (void **)&status); - return status; + return join_thread_status(thread_t); } int stop_polling(int dev_idx, int virtual_fd, int sd, int fid, int pid) { pthread_t thread_t; - int status; int rc; rc = pthread_create(&thread_t, NULL, virtual_poll_stop, @@ -618,8 +626,7 @@ int stop_polling(int dev_idx, int virtual_fd, int sd, int fid, int pid) if (rc != 0) return rc; - pthread_join(thread_t, (void **)&status); - return status; + return join_thread_status(thread_t); } TEST_F(NCI, start_poll) @@ -834,8 +841,10 @@ int disconnect_tag(int nfc_sock, int virtual_fd) return status; close(nfc_sock); - pthread_join(thread_t, (void **)&status); - return status; + if (status) + return -1; + + return join_thread_status(thread_t); } TEST_F(NCI, t4t_tag_read) @@ -882,7 +891,7 @@ TEST_F(NCI, deinit) NFC_CMD_DEV_DOWN, self->dev_idex); EXPECT_EQ(rc, 0); - pthread_join(thread_t, (void **)&status); + status = join_thread_status(thread_t); self->open_state = 0; ASSERT_EQ(status, 0); From 51814683e28fc64eceb415962376956c3cfc75a7 Mon Sep 17 00:00:00 2001 From: Chris Gellermann Date: Fri, 4 Sep 2026 18:42:52 +0200 Subject: [PATCH 464/694] nfc: virtual_ncidev: Add missing ioctl compat handler The compat handler for ioctls to the virtual nci device is missing. So, nci-specific ioctls of a compat task return with -1 and errno set to ENOTTY. Add a handler. The handling of an ioctl() call of a compat task to get the index of virtual nci device (IOCTL_GET_NCIDEV_IDX) lands in the default case of the ioctl compat handler (see fs/ioctl.c): COMPAT_SYSCALL_DEFINE3(ioctl, ...) { ... default: error = do_vfs_ioctl(fd_file(f), fd, cmd, ...); if (error != -ENOIOCTLCMD) break; if (fd_file(f)->f_op->compat_ioctl) error = fd_file(f)->f_op->compat_ioctl(fd_file(f), cmd, arg); if (error == -ENOIOCTLCMD) error = -ENOTTY; ... } There, do_vfs_ioctl() returns -ENOIOCTLCMD and compat_ioctl is not set for virtual_ncidev_fops, i.e. f_op->compat_ioctl == NULL. So, the ioctl() syscall returns with -1 and errno set to ENOTTY to the compat task. To fix this, use the compat_ptr_ioctl helper for compat handling here. It shall be used for ioctls that "either ignore the argument or pass a pointer to a compatible data type". The driver's sole ioctl takes a user void pointer and copies nfc_dev->idx to it, a 4-byte integer across all ABIs. This issue has been found by running the nci_dev kernel selftest as rv64 binary on top of a CHERI kernel, where the ioctl() ends up in the ioctl compat handler, similar to a 32-bit application on top of a 64-bit kernel. Fixes: e624e6c3e777 ("nfc: Add a virtual nci device driver") Signed-off-by: Chris Gellermann Reviewed-by: Simon Horman Link: https://patch.msgid.link/20260904164252.18351-1-christian.gellermann@codasip.com Signed-off-by: David Heidelberg --- drivers/nfc/virtual_ncidev.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/nfc/virtual_ncidev.c b/drivers/nfc/virtual_ncidev.c index 8eeb447ac96e..e51c647b27eb 100644 --- a/drivers/nfc/virtual_ncidev.c +++ b/drivers/nfc/virtual_ncidev.c @@ -195,7 +195,8 @@ static const struct file_operations virtual_ncidev_fops = { .write = virtual_ncidev_write, .open = virtual_ncidev_open, .release = virtual_ncidev_close, - .unlocked_ioctl = virtual_ncidev_ioctl + .unlocked_ioctl = virtual_ncidev_ioctl, + .compat_ioctl = compat_ptr_ioctl, }; static struct miscdevice miscdev = { From 273f9d667cde649f8de9d72b1303cc2f4b658c50 Mon Sep 17 00:00:00 2001 From: Aamir Ahmed Date: Tue, 15 Sep 2026 19:54:27 +0100 Subject: [PATCH 465/694] nfc: llcp: drop truncated I/RR/RNR PDUs in nfc_llcp_recv_hdlc() nfc_llcp_recv_hdlc() reads the sequence byte skb->data[2], via nfc_llcp_ns()/nfc_llcp_nr(), before any length check. The receive path only guarantees the two-byte LLCP header -- __nfc_llcp_recv() checks it with pskb_may_pull() and nfc_llcp_recv_agf() admits two-byte inner PDUs -- so a two-byte I, RR or RNR PDU reads one byte of uninitialised skb tailroom. The byte becomes N(R)/N(S); a peer can already set those with a well-formed PDU, so this is acting on uninitialised memory, not new peer control. Guard the read with pskb_may_pull(), as commit 95674f506c63 ("nfc: llcp: reject PDUs shorter than the LLCP header") did for the two-byte header, so the sequence byte is present and linear before it is read. RR and RNR PDUs are LLCP_HEADER_SIZE + LLCP_SEQUENCE_SIZE bytes and an I PDU is longer, so no valid frame is rejected; a truncated PDU is malformed, so return without a DM reply. Fixes: d646960f7986 ("NFC: Initial LLCP support") Cc: stable@vger.kernel.org Assisted-by: LLM Signed-off-by: Aamir Ahmed Reviewed-by: Simon Horman Link: https://patch.msgid.link/AS8P251MB0001789BBF04B72745C7D96BC8BA2@AS8P251MB0001.EURP251.PROD.OUTLOOK.COM Signed-off-by: David Heidelberg --- net/nfc/llcp_core.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/net/nfc/llcp_core.c b/net/nfc/llcp_core.c index 23553e7426ec..5017f6aa57a0 100644 --- a/net/nfc/llcp_core.c +++ b/net/nfc/llcp_core.c @@ -1091,6 +1091,9 @@ static void nfc_llcp_recv_hdlc(struct nfc_llcp_local *local, struct sock *sk; u8 dsap, ssap, ptype, ns, nr; + if (!pskb_may_pull(skb, LLCP_HEADER_SIZE + LLCP_SEQUENCE_SIZE)) + return; + ptype = nfc_llcp_ptype(skb); dsap = nfc_llcp_dsap(skb); ssap = nfc_llcp_ssap(skb); From 66f4300206b82b0b143ef0d9be90cd8d29f23a47 Mon Sep 17 00:00:00 2001 From: Cong Nguyen Date: Mon, 14 Sep 2026 19:11:29 +0700 Subject: [PATCH 466/694] nfc: llcp: fix sdreq TLV list leak on parse/alloc/send failure nfc_genl_llc_sdreq() builds a list of TLV nodes while walking nested netlink attrs, but 3 error paths (nested-attr parse failure, TLV alloc ENOMEM, nfc_llcp_send_snl_sdreq() failure) all skip freeing what was already queued. Route them through a new free_list label, mirroring the SDRES path in the same file which already does this. Harmless on the success path too -- send_snl_sdreq() drains the list as it moves nodes, so it's already empty by the time free_list runs. Fixes: d9b8d8e19b07 ("NFC: llcp: Service Name Lookup netlink interface") Assisted-by: Claude:claude-opus-4 Signed-off-by: Cong Nguyen Reviewed-by: Simon Horman Link: https://patch.msgid.link/20260914121129.2098606-1-congnt264@gmail.com Signed-off-by: David Heidelberg --- net/nfc/netlink.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/net/nfc/netlink.c b/net/nfc/netlink.c index 0c58824cb150..224bdfa2dd0d 100644 --- a/net/nfc/netlink.c +++ b/net/nfc/netlink.c @@ -1181,7 +1181,7 @@ static int nfc_genl_llc_sdreq(struct sk_buff *skb, struct genl_info *info) if (rc != 0) { rc = -EINVAL; - goto put_local; + goto free_list; } if (!sdp_attrs[NFC_SDP_ATTR_URI]) @@ -1200,7 +1200,7 @@ static int nfc_genl_llc_sdreq(struct sk_buff *skb, struct genl_info *info) sdreq = nfc_llcp_build_sdreq_tlv(tid, uri, uri_len); if (sdreq == NULL) { rc = -ENOMEM; - goto put_local; + goto free_list; } tlvs_len += sdreq->tlv_len; @@ -1215,6 +1215,9 @@ static int nfc_genl_llc_sdreq(struct sk_buff *skb, struct genl_info *info) rc = nfc_llcp_send_snl_sdreq(local, &sdreq_list, tlvs_len); +free_list: + nfc_llcp_free_sdp_tlv_list(&sdreq_list); + put_local: nfc_llcp_local_put(local); From d2acbde7e67df44efa8f0963462d1192e7694ffc Mon Sep 17 00:00:00 2001 From: Myeonghun Pak Date: Sun, 13 Sep 2026 00:26:25 -0400 Subject: [PATCH 467/694] nfc: trf7970a: power down on startup RX gain failure trf7970a_startup() powers up the device before applying the optional RX gain reduction. If the register read or write fails, it returns without undoing that power-up. Probe's unwind only drops the separate regulator references acquired by probe, leaving the additional VIN enable from startup unbalanced. The system resume caller also has no power-down on this error. Call trf7970a_power_down() before returning the RX gain error to deassert the enable GPIOs, release the startup VIN reference and restore the powered-off state. Runtime PM has not been enabled yet, so the full shutdown helper is not appropriate here. Preserve the original SPI error. This issue was identified during our ongoing static-analysis research while reviewing kernel code. Fixes: 5d69351820ea ("NFC: trf7970a: Create device-tree parameter for RX gain reduction") Cc: stable@vger.kernel.org Assisted-by: OpenAI:GPT-5.6 Co-developed-by: Ijae Kim Signed-off-by: Ijae Kim Signed-off-by: Myeonghun Pak Reviewed-by: Paul Geurts Link: https://patch.msgid.link/20260913042625.31296-1-mhun512@gmail.com Signed-off-by: David Heidelberg --- drivers/nfc/trf7970a.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/nfc/trf7970a.c b/drivers/nfc/trf7970a.c index 60883001fa5d..ddfc58c29964 100644 --- a/drivers/nfc/trf7970a.c +++ b/drivers/nfc/trf7970a.c @@ -1997,8 +1997,10 @@ static int trf7970a_startup(struct trf7970a *trf) return ret; ret = trf7970a_update_rx_gain_reduction(trf); - if (ret) + if (ret) { + trf7970a_power_down(trf); return ret; + } pm_runtime_set_active(trf->dev); pm_runtime_enable(trf->dev); From dcab71a7011918f6fdba7adcec02d217dcb84b8d Mon Sep 17 00:00:00 2001 From: Luxiao Xu Date: Wed, 9 Sep 2026 13:19:24 +0800 Subject: [PATCH 468/694] nfc: fix use-after-free in nfc_get_local_general_bytes Commit 6709d4b7bc2e ("net: nfc: Fix use-after-free caused by nfc_llcp_find_local") attempted to fix a use-after-free (UAF) issue by invoking nfc_llcp_local_put(local) after accessing local->gb. However, if the reference count drops to zero, local is freed immediately, leading to a use-after-free when callers access the returned pointer. Alternative approaches using dynamic allocation (e.g. kmemdup) introduced memory leaks because callers consistently treat the returned pointer as borrowed memory. Fix this properly by refactoring nfc_llcp_general_bytes() and nfc_get_local_general_bytes() to accept a caller-provided output buffer (out_gb) and its maximum length (gb_max_len). The general bytes are safely copied into out_gb before calling nfc_llcp_local_put(local), ensuring safe lifetime management without ownership transfer complications. Update all callers across drivers (microread, pn533, pn544, st21nfca, digital_dep, and nci) to provide their own destination buffers and pass them to nfc_get_local_general_bytes(). Fixes: 6709d4b7bc2e ("net: nfc: Fix use-after-free caused by nfc_llcp_find_local") Cc: stable@vger.kernel.org Reported-by: Vega Assisted-by: LLM Signed-off-by: Luxiao Xu Signed-off-by: Ren Wei Reviewed-by: Simon Horman Link: https://patch.msgid.link/3cbaac3bee23f8ff3a3284ed32d347696eb1d208.1788841683.git.rakukuip@gmail.com Signed-off-by: David Heidelberg --- drivers/nfc/microread/microread.c | 6 +++--- drivers/nfc/pn533/pn533.c | 14 ++++++++------ drivers/nfc/pn533/pn533.h | 4 +++- drivers/nfc/pn544/pn544.c | 7 +++---- drivers/nfc/st21nfca/core.c | 8 ++++---- include/net/nfc/hci.h | 2 +- include/net/nfc/nfc.h | 3 ++- net/nfc/core.c | 15 +++++++-------- net/nfc/digital_dep.c | 8 ++++---- net/nfc/llcp_core.c | 17 +++++++++++++---- net/nfc/nci/core.c | 10 +++++----- net/nfc/nfc.h | 3 ++- 12 files changed, 55 insertions(+), 42 deletions(-) diff --git a/drivers/nfc/microread/microread.c b/drivers/nfc/microread/microread.c index dfa2490db545..2bfafa94e83d 100644 --- a/drivers/nfc/microread/microread.c +++ b/drivers/nfc/microread/microread.c @@ -251,9 +251,9 @@ static int microread_start_poll(struct nfc_hci_dev *hdev, param[1] |= (1 << 1); if ((im_protocols | tm_protocols) & NFC_PROTO_NFC_DEP_MASK) { - hdev->gb = nfc_get_local_general_bytes(hdev->ndev, - &hdev->gb_len); - if (hdev->gb == NULL || hdev->gb_len == 0) { + nfc_get_local_general_bytes(hdev->ndev, hdev->gb, + sizeof(hdev->gb), &hdev->gb_len); + if (hdev->gb_len == 0) { im_protocols &= ~NFC_PROTO_NFC_DEP_MASK; tm_protocols &= ~NFC_PROTO_NFC_DEP_MASK; } diff --git a/drivers/nfc/pn533/pn533.c b/drivers/nfc/pn533/pn533.c index f5a6a7c20d5a..b0133e51dce9 100644 --- a/drivers/nfc/pn533/pn533.c +++ b/drivers/nfc/pn533/pn533.c @@ -1357,10 +1357,11 @@ static int pn533_poll_dep(struct nfc_dev *nfc_dev) u8 *next, nfcid3[NFC_NFCID3_MAXSIZE]; u8 passive_data[PASSIVE_DATA_LEN] = {0x00, 0xff, 0xff, 0x00, 0x3}; - if (!dev->gb) { - dev->gb = nfc_get_local_general_bytes(nfc_dev, &dev->gb_len); - - if (!dev->gb || !dev->gb_len) { + if (!dev->gb_len) { + nfc_get_local_general_bytes(nfc_dev, dev->gb, + sizeof(dev->gb), + &dev->gb_len); + if (!dev->gb_len) { dev->poll_dep = 0; queue_work(dev->wq, &dev->rf_work); } @@ -1658,8 +1659,9 @@ static int pn533_start_poll(struct nfc_dev *nfc_dev, } if (tm_protocols) { - dev->gb = nfc_get_local_general_bytes(nfc_dev, &dev->gb_len); - if (dev->gb == NULL) + nfc_get_local_general_bytes(nfc_dev, dev->gb, + sizeof(dev->gb), &dev->gb_len); + if (dev->gb_len == 0) tm_protocols = 0; } diff --git a/drivers/nfc/pn533/pn533.h b/drivers/nfc/pn533/pn533.h index 09e35b8693f5..5ab668e05121 100644 --- a/drivers/nfc/pn533/pn533.h +++ b/drivers/nfc/pn533/pn533.h @@ -6,6 +6,8 @@ * Copyright (C) 2012-2013 Tieto Poland */ +#include + #define PN533_DEVICE_STD 0x1 #define PN533_DEVICE_PASORI 0x2 #define PN533_DEVICE_ACR122U 0x3 @@ -166,7 +168,7 @@ struct pn533 { struct timer_list listen_timer; int cancel_listen; - u8 *gb; + u8 gb[NFC_MAX_GT_LEN]; size_t gb_len; u8 tgt_available_prots; diff --git a/drivers/nfc/pn544/pn544.c b/drivers/nfc/pn544/pn544.c index 9d0a16ac465e..c4fa70e45c14 100644 --- a/drivers/nfc/pn544/pn544.c +++ b/drivers/nfc/pn544/pn544.c @@ -377,10 +377,9 @@ static int pn544_hci_start_poll(struct nfc_hci_dev *hdev, return r; if ((im_protocols | tm_protocols) & NFC_PROTO_NFC_DEP_MASK) { - hdev->gb = nfc_get_local_general_bytes(hdev->ndev, - &hdev->gb_len); - pr_debug("generate local bytes %p\n", hdev->gb); - if (hdev->gb == NULL || hdev->gb_len == 0) { + nfc_get_local_general_bytes(hdev->ndev, hdev->gb, + sizeof(hdev->gb), &hdev->gb_len); + if (hdev->gb_len == 0) { im_protocols &= ~NFC_PROTO_NFC_DEP_MASK; tm_protocols &= ~NFC_PROTO_NFC_DEP_MASK; } diff --git a/drivers/nfc/st21nfca/core.c b/drivers/nfc/st21nfca/core.c index fd39a05c9622..6bfeb8e7ed89 100644 --- a/drivers/nfc/st21nfca/core.c +++ b/drivers/nfc/st21nfca/core.c @@ -351,10 +351,10 @@ static int st21nfca_hci_start_poll(struct nfc_hci_dev *hdev, if (r < 0) return r; } else { - hdev->gb = nfc_get_local_general_bytes(hdev->ndev, - &hdev->gb_len); - - if (hdev->gb == NULL || hdev->gb_len == 0) { + nfc_get_local_general_bytes(hdev->ndev, hdev->gb, + sizeof(hdev->gb), + &hdev->gb_len); + if (hdev->gb_len == 0) { im_protocols &= ~NFC_PROTO_NFC_DEP_MASK; tm_protocols &= ~NFC_PROTO_NFC_DEP_MASK; } diff --git a/include/net/nfc/hci.h b/include/net/nfc/hci.h index 756c11084f65..86ed63e5d533 100644 --- a/include/net/nfc/hci.h +++ b/include/net/nfc/hci.h @@ -144,7 +144,7 @@ struct nfc_hci_dev { data_exchange_cb_t async_cb; void *async_cb_context; - u8 *gb; + u8 gb[NFC_MAX_GT_LEN]; size_t gb_len; unsigned long quirks; diff --git a/include/net/nfc/nfc.h b/include/net/nfc/nfc.h index c54df042db6b..bcafab5c53e5 100644 --- a/include/net/nfc/nfc.h +++ b/include/net/nfc/nfc.h @@ -273,7 +273,8 @@ struct sk_buff *nfc_alloc_recv_skb(unsigned int size, gfp_t gfp); int nfc_set_remote_general_bytes(struct nfc_dev *dev, const u8 *gt, u8 gt_len); -u8 *nfc_get_local_general_bytes(struct nfc_dev *dev, size_t *gb_len); +u8 *nfc_get_local_general_bytes(struct nfc_dev *dev, u8 *out_gb, + size_t gb_max_len, size_t *gb_len); int nfc_fw_download_done(struct nfc_dev *dev, const char *firmware_name, u32 result); diff --git a/net/nfc/core.c b/net/nfc/core.c index a92a6566e6a0..f521669293f0 100644 --- a/net/nfc/core.c +++ b/net/nfc/core.c @@ -279,10 +279,10 @@ static struct nfc_target *nfc_find_target(struct nfc_dev *dev, u32 target_idx) int nfc_dep_link_up(struct nfc_dev *dev, int target_index, u8 comm_mode) { - int rc = 0; - u8 *gb; - size_t gb_len; struct nfc_target *target; + u8 gb[NFC_MAX_GT_LEN]; + size_t gb_len = 0; + int rc = 0; pr_debug("dev_name=%s comm %d\n", dev_name(&dev->dev), comm_mode); @@ -301,7 +301,7 @@ int nfc_dep_link_up(struct nfc_dev *dev, int target_index, u8 comm_mode) goto error; } - gb = nfc_llcp_general_bytes(dev, &gb_len); + nfc_get_local_general_bytes(dev, gb, sizeof(gb), &gb_len); if (gb_len > NFC_MAX_GT_LEN) { rc = -EINVAL; goto error; @@ -644,11 +644,10 @@ int nfc_set_remote_general_bytes(struct nfc_dev *dev, const u8 *gb, u8 gb_len) } EXPORT_SYMBOL(nfc_set_remote_general_bytes); -u8 *nfc_get_local_general_bytes(struct nfc_dev *dev, size_t *gb_len) +u8 *nfc_get_local_general_bytes(struct nfc_dev *dev, u8 *out_gb, + size_t gb_max_len, size_t *gb_len) { - pr_debug("dev_name=%s\n", dev_name(&dev->dev)); - - return nfc_llcp_general_bytes(dev, gb_len); + return nfc_llcp_general_bytes(dev, out_gb, gb_max_len, gb_len); } EXPORT_SYMBOL(nfc_get_local_general_bytes); diff --git a/net/nfc/digital_dep.c b/net/nfc/digital_dep.c index 3982fa084737..968547c306a5 100644 --- a/net/nfc/digital_dep.c +++ b/net/nfc/digital_dep.c @@ -1490,14 +1490,14 @@ static int digital_tg_send_atr_res(struct nfc_digital_dev *ddev, struct digital_atr_req *atr_req) { struct digital_atr_res *atr_res; + u8 gb[NFC_MAX_GT_LEN]; struct sk_buff *skb; - u8 *gb, payload_bits; + u8 payload_bits; size_t gb_len; int rc; - gb = nfc_get_local_general_bytes(ddev->nfc_dev, &gb_len); - if (!gb) - gb_len = 0; + nfc_get_local_general_bytes(ddev->nfc_dev, gb, sizeof(gb), + &gb_len); skb = digital_skb_alloc(ddev, sizeof(struct digital_atr_res) + gb_len); if (!skb) diff --git a/net/nfc/llcp_core.c b/net/nfc/llcp_core.c index 5017f6aa57a0..5ce3ce64baf2 100644 --- a/net/nfc/llcp_core.c +++ b/net/nfc/llcp_core.c @@ -652,23 +652,32 @@ static int nfc_llcp_build_gb(struct nfc_llcp_local *local) return ret; } -u8 *nfc_llcp_general_bytes(struct nfc_dev *dev, size_t *general_bytes_len) +u8 *nfc_llcp_general_bytes(struct nfc_dev *dev, u8 *out_gb, size_t gb_max_len, + size_t *general_bytes_len) { struct nfc_llcp_local *local; + if (!out_gb || !general_bytes_len) + return NULL; + local = nfc_llcp_find_local(dev); - if (local == NULL) { + if (!local) { *general_bytes_len = 0; return NULL; } nfc_llcp_build_gb(local); - *general_bytes_len = local->gb_len; + if (local->gb_len) { + *general_bytes_len = min_t(size_t, local->gb_len, gb_max_len); + memcpy(out_gb, local->gb, *general_bytes_len); + } else { + *general_bytes_len = 0; + } nfc_llcp_local_put(local); - return local->gb; + return out_gb; } int nfc_llcp_set_remote_gb(struct nfc_dev *dev, const u8 *gb, u8 gb_len) diff --git a/net/nfc/nci/core.c b/net/nfc/nci/core.c index 5f46c4b5720f..73e3a96470ac 100644 --- a/net/nfc/nci/core.c +++ b/net/nfc/nci/core.c @@ -780,15 +780,15 @@ static int nci_set_local_general_bytes(struct nfc_dev *nfc_dev) { struct nci_dev *ndev = nfc_get_drvdata(nfc_dev); struct nci_set_config_param param; + u8 gb[NFC_MAX_GT_LEN]; int rc; - param.val = nfc_get_local_general_bytes(nfc_dev, ¶m.len); - if ((param.val == NULL) || (param.len == 0)) + nfc_get_local_general_bytes(nfc_dev, gb, sizeof(gb), + ¶m.len); + if (param.len == 0) return 0; - if (param.len > NFC_MAX_GT_LEN) - return -EINVAL; - + param.val = gb; param.id = NCI_PN_ATR_REQ_GEN_BYTES; rc = nci_request(ndev, nci_set_config_req, ¶m, diff --git a/net/nfc/nfc.h b/net/nfc/nfc.h index 0b1e6466f4fb..82c5dfdad10e 100644 --- a/net/nfc/nfc.h +++ b/net/nfc/nfc.h @@ -49,7 +49,8 @@ void nfc_llcp_mac_is_up(struct nfc_dev *dev, u32 target_idx, int nfc_llcp_register_device(struct nfc_dev *dev); void nfc_llcp_unregister_device(struct nfc_dev *dev); int nfc_llcp_set_remote_gb(struct nfc_dev *dev, const u8 *gb, u8 gb_len); -u8 *nfc_llcp_general_bytes(struct nfc_dev *dev, size_t *general_bytes_len); +u8 *nfc_llcp_general_bytes(struct nfc_dev *dev, u8 *out_gb, size_t gb_max_len, + size_t *general_bytes_len); int nfc_llcp_data_received(struct nfc_dev *dev, struct sk_buff *skb); struct nfc_llcp_local *nfc_llcp_find_local(struct nfc_dev *dev); int nfc_llcp_local_put(struct nfc_llcp_local *local); From 7f2ea5ed588c03d481f0301e6c3d4240132383fb Mon Sep 17 00:00:00 2001 From: Pengpeng Hou Date: Sun, 30 Aug 2026 21:29:58 +0800 Subject: [PATCH 469/694] nfc: st21nfca: validate ISO15693 inventory length The ISO15693 inventory helper removes a two-byte prefix without checking that it exists, then accepts a one-byte remainder before reading data[1] as the DSFID. Require the prefix and at least two remaining bytes before copying the UID data and reading the DSFID. Fixes: 7974728094d3 ("NFC: st21nfca: Add ISO15693 Reader/Writer support") Signed-off-by: Pengpeng Hou Link: https://patch.msgid.link/20260830132958.6397-1-pengpeng@iscas.ac.cn Signed-off-by: David Heidelberg --- drivers/nfc/st21nfca/core.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/drivers/nfc/st21nfca/core.c b/drivers/nfc/st21nfca/core.c index 6bfeb8e7ed89..b5c1ca3acfbe 100644 --- a/drivers/nfc/st21nfca/core.c +++ b/drivers/nfc/st21nfca/core.c @@ -577,9 +577,7 @@ static int st21nfca_get_iso15693_inventory(struct nfc_hci_dev *hdev, if (r < 0) goto exit; - skb_pull(inventory_skb, 2); - - if (inventory_skb->len == 0 || + if (!skb_pull(inventory_skb, 2) || inventory_skb->len < 2 || inventory_skb->len > NFC_ISO15693_UID_MAXSIZE) { r = -EPROTO; goto exit; From c04981e42d94f39c1dba965cc462a046e946a6c5 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?=C3=96mer=20Mete=20Kaya?= Date: Wed, 9 Sep 2026 15:16:22 +0300 Subject: [PATCH 470/694] nfc: llcp: fix -ENOMEM on connect with zero-length service name MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit When service_name_len is 0, kmemdup() returns ZERO_SIZE_PTR which passes the NULL check, causing nfc_llcp_send_connect() to attempt building a zero-length service name TLV and fail with -ENOMEM. Fix by setting service_name to NULL directly when service_name_len is 0. Fixes: d646960f7986 ("NFC: Initial LLCP support") Signed-off-by: Ömer Mete Kaya Link: https://patch.msgid.link/20260909122029.34081-1-omermetekaya0@gmail.com Signed-off-by: David Heidelberg --- net/nfc/llcp_sock.c | 16 ++++++++++------ 1 file changed, 10 insertions(+), 6 deletions(-) diff --git a/net/nfc/llcp_sock.c b/net/nfc/llcp_sock.c index ce6875eb58fb..1e5ee4bcde68 100644 --- a/net/nfc/llcp_sock.c +++ b/net/nfc/llcp_sock.c @@ -759,12 +759,16 @@ static int llcp_sock_connect(struct socket *sock, struct sockaddr_unsized *_addr llcp_sock->service_name_len = min_t(unsigned int, addr->service_name_len, NFC_LLCP_MAX_SERVICE_NAME); - llcp_sock->service_name = kmemdup(addr->service_name, - llcp_sock->service_name_len, - GFP_KERNEL); - if (!llcp_sock->service_name) { - ret = -ENOMEM; - goto sock_llcp_release; + if (llcp_sock->service_name_len == 0) { + llcp_sock->service_name = NULL; + } else { + llcp_sock->service_name = kmemdup(addr->service_name, + llcp_sock->service_name_len, + GFP_KERNEL); + if (!llcp_sock->service_name) { + ret = -ENOMEM; + goto sock_llcp_release; + } } nfc_llcp_sock_link(&local->connecting_sockets, sk); From 408cff6bd60636df201274d320edfdfde9ed41db Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?=C3=96mer=20Mete=20Kaya?= Date: Wed, 9 Sep 2026 15:14:31 +0300 Subject: [PATCH 471/694] nfc: llcp: fix WKS SAP hijacking via prefix match in nfc_llcp_wks_sap() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit nfc_llcp_wks_sap() compares only service_name_len bytes, so a short service_name like "u" matches longer WKS strings like "urn:nfc:sn:snep". Fix by requiring exact length match before strncmp(). Fixes: d646960f7986 ("NFC: Initial LLCP support") Signed-off-by: Ömer Mete Kaya Link: https://patch.msgid.link/20260909121437.33744-1-omermetekaya0@gmail.com Signed-off-by: David Heidelberg --- net/nfc/llcp_core.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/net/nfc/llcp_core.c b/net/nfc/llcp_core.c index 5ce3ce64baf2..def712e1b2b2 100644 --- a/net/nfc/llcp_core.c +++ b/net/nfc/llcp_core.c @@ -369,7 +369,8 @@ static int nfc_llcp_wks_sap(const char *service_name, size_t service_name_len) if (wks[sap] == NULL) continue; - if (strncmp(wks[sap], service_name, service_name_len) == 0) + if (strlen(wks[sap]) == service_name_len && + !strncmp(wks[sap], service_name, service_name_len)) return sap; } From 7dcf371a35632f035baf77bcf2c129165f772ce4 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?=C3=96mer=20Mete=20Kaya?= Date: Tue, 8 Sep 2026 19:18:01 +0300 Subject: [PATCH 472/694] nfc: llcp: fix slab-out-of-bounds reads when logging service names MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit nfc_llcp_wks_sap() and nfc_llcp_build_sdreq_tlv() pass non-null- terminated strings to pr_debug() using the %s format specifier. The buffers are allocated via kmemdup() or come from netlink attributes and are not guaranteed to be null-terminated, causing __dynamic_pr_debug() to read beyond the allocated region: KASAN: slab-out-of-bounds Read in __dynamic_pr_debug Fix both call sites by using %.*s with the explicit length to limit the output to the actual length of the string. Fixes: d9b8d8e19b07 ("NFC: llcp: Service Name Lookup netlink interface") Reported-by: syzbot+1e3df0852e82c21ca418@syzkaller.appspotmail.com Closes: https://syzkaller.appspot.com/bug?extid=1e3df0852e82c21ca418 Signed-off-by: Ömer Mete Kaya Link: https://patch.msgid.link/20260908161952.731468-1-omermetekaya0@gmail.com Signed-off-by: David Heidelberg --- net/nfc/llcp_commands.c | 2 +- net/nfc/llcp_core.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/net/nfc/llcp_commands.c b/net/nfc/llcp_commands.c index ca89fe967d6a..80a00938c869 100644 --- a/net/nfc/llcp_commands.c +++ b/net/nfc/llcp_commands.c @@ -135,7 +135,7 @@ struct nfc_llcp_sdp_tlv *nfc_llcp_build_sdreq_tlv(u8 tid, const char *uri, { struct nfc_llcp_sdp_tlv *sdreq; - pr_debug("uri: %s, len: %zu\n", uri, uri_len); + pr_debug("uri: %.*s, len: %zu\n", (int)uri_len, uri, uri_len); /* sdreq->tlv_len is u8, takes uri_len, + 3 for header, + 1 for NULL */ if (WARN_ON_ONCE(uri_len > U8_MAX - 4)) diff --git a/net/nfc/llcp_core.c b/net/nfc/llcp_core.c index def712e1b2b2..74bf817007cf 100644 --- a/net/nfc/llcp_core.c +++ b/net/nfc/llcp_core.c @@ -358,7 +358,7 @@ static int nfc_llcp_wks_sap(const char *service_name, size_t service_name_len) { int sap, num_wks; - pr_debug("%s\n", service_name); + pr_debug("%.*s\n", (int)service_name_len, service_name); if (service_name == NULL) return -EINVAL; From b61732f47316d45f27706db7812950145d3327b5 Mon Sep 17 00:00:00 2001 From: Deepanshu Kartikey Date: Wed, 23 Sep 2026 09:26:27 +0530 Subject: [PATCH 473/694] nfc: pn533: fix OOB read in pn533_acr122_is_rx_frame_valid() frame->ccid.datalen is read directly from the USB response frame and used, unchecked, as an index into frame->data[]. A malicious or malfunctioning device can set this field to an arbitrary value, causing the driver to read far outside the received buffer. Bound ccid.datalen against the maximum possible ACR122 frame size before using it. This replaces the existing datalen == 0 check, since datalen < 2 already covers that case and additionally rejects datalen == 1, which would still underflow the "datalen - 2" offset used below. Fixes: 9815c7cf22da ("NFC: pn533: Separate physical layer from the core implementation") Reported-by: syzbot+1853daab1a47603d4678@syzkaller.appspotmail.com Closes: https://syzkaller.appspot.com/bug?extid=1853daab1a47603d4678 Tested-by: syzbot+1853daab1a47603d4678@syzkaller.appspotmail.com Assisted-by: LLM Signed-off-by: Deepanshu Kartikey Link: https://patch.msgid.link/20260923035627.6210-1-kartikey406@gmail.com Signed-off-by: David Heidelberg --- drivers/nfc/pn533/usb.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/nfc/pn533/usb.c b/drivers/nfc/pn533/usb.c index efb07f944fce..972eaac09e59 100644 --- a/drivers/nfc/pn533/usb.c +++ b/drivers/nfc/pn533/usb.c @@ -319,7 +319,9 @@ static bool pn533_acr122_is_rx_frame_valid(void *_frame, struct pn533 *dev) if (frame->ccid.type != 0x83) return false; - if (!frame->ccid.datalen) + if (frame->ccid.datalen < 2 || + frame->ccid.datalen > PN533_ACR122_FRAME_MAX_PAYLOAD_LEN + + PN533_ACR122_RX_FRAME_TAIL_LEN) return false; if (frame->data[frame->ccid.datalen - 2] == 0x63) From cfa165cbfbed9d0f4bbc22fef4309f595a3ab187 Mon Sep 17 00:00:00 2001 From: Victor Nogueira Date: Sun, 20 Sep 2026 14:07:01 -0300 Subject: [PATCH 474/694] net/sched: act_gate: budget the per-entry list in get_fill_size tcf_gate_get_fill_size returns only the TCA_GATE_PARMS size, but tcf_gate_dump also emits three 64-bit timestamps, the clock id, flags, priority and the variable-length TCA_GATE_ENTRY_LIST nest. The per-entry nest is unbounded: parse_gate_list places no cap on the number of sched-entries, so a gate with many entries can push the real dump well past the skb that tca_get_fill allocates from this size. RTM_NEWACTION then fails the add-notify with -EINVAL while the action is already committed to the IDR, and a subsequent RTM_GETACTION on the installed gate also returns -EINVAL because its dump no longer fits. Fix this by accounting for the missing fields in tcf_gate_get_fill_size along with all elements in the entries list. Note that sizing the reply from the action lets an oversized gate install cleanly for the first time: with the input unbounded by parse_gate_list, the sized skb can now grow well above NLMSG_GOODSIZE per netlink request (a transient GFP_KERNEL allocation reachable only with namespace-local CAP_NET_ADMIN). Overload from a malicious netns admin is hardening material, not net, per the discussion at https://lore.kernel.org/netdev/20260914191108.55a1a4f1@kernel.org/; a follow-up patch for net-next will cap the sched-entry count. Fixes: 4e76e75d6aba ("net sched actions: calculate add/delete event message size") Reported-by: Sashiko Link: https://netdev-ai.bots.linux.dev/sashiko/#/patchset/20260824153903.4143642-1-victor@mojatatu.com Tested-by: hybris Co-developed-by: Jamal Hadi Salim Signed-off-by: Jamal Hadi Salim Signed-off-by: Victor Nogueira Link: https://patch.msgid.link/QDISC-3BLH.v1.20260914203033@mojatatu.com Signed-off-by: Jakub Kicinski --- net/sched/act_gate.c | 30 +++++++++++++++++++++++++++++- 1 file changed, 29 insertions(+), 1 deletion(-) diff --git a/net/sched/act_gate.c b/net/sched/act_gate.c index 5d228a402204..6d6d45e03c07 100644 --- a/net/sched/act_gate.c +++ b/net/sched/act_gate.c @@ -681,7 +681,35 @@ static void tcf_gate_stats_update(struct tc_action *a, u64 bytes, u64 packets, static size_t tcf_gate_get_fill_size(const struct tc_action *act) { - return nla_total_size(sizeof(struct tc_gate)); + struct tcf_gate *gact = to_gate(act); + const struct tcf_gate_params *p; + struct tcfg_gate_entry *entry; + size_t size = nla_total_size(sizeof(struct tc_gate)) /* TCA_GATE_PARMS */ + + 3 * nla_total_size_64bit(sizeof(u64)) /* TCA_GATE_BASE_TIME + * TCA_GATE_CYCLE_TIME + * TCA_GATE_CYCLE_TIME_EXT + */ + + nla_total_size(sizeof(s32)) /* TCA_GATE_CLOCKID */ + + nla_total_size(sizeof(u32)) /* TCA_GATE_FLAGS */ + + nla_total_size(sizeof(s32)) /* TCA_GATE_PRIORITY */ + + nla_total_size(0); /* TCA_GATE_ENTRY_LIST */ + /* TCA_GATE_TM is budgeted by tcf_action_shared_attrs_size() */ + + rcu_read_lock(); + p = rcu_dereference(gact->param); + if (p) { + list_for_each_entry_rcu(entry, &p->entries, list) + /* TCA_GATE_ONE_ENTRY nest and its attributes */ + size += nla_total_size(0) + + nla_total_size(sizeof(u32)) /* TCA_GATE_ENTRY_INDEX */ + + nla_total_size(0) /* TCA_GATE_ENTRY_GATE */ + + nla_total_size(sizeof(u32)) /* TCA_GATE_ENTRY_INTERVAL */ + + nla_total_size(sizeof(s32)) /* TCA_GATE_ENTRY_MAX_OCTETS */ + + nla_total_size(sizeof(s32)); /* TCA_GATE_ENTRY_IPV */ + } + rcu_read_unlock(); + + return size; } static void tcf_gate_entry_destructor(void *priv) From ab1404ac81154a89fb61ac50ae9a04cd8d4834dc Mon Sep 17 00:00:00 2001 From: Fourie Zhang Date: Sun, 20 Sep 2026 19:08:43 +0800 Subject: [PATCH 475/694] net: bridge: mdb: restart port group walk after deletion br_mdb_flush_pgs() keeps a pointer-to-pointer cursor while walking mp->ports. br_multicast_del_pg() can re-enter the same MDB entry through br_multicast_sg_del_exclude_ports() and unlink other port groups. If the cursor points into one of those groups, the next iteration dereferences a stale cursor and can leave mp->ports pointing at freed memory. A following RTM_GETMDB exposes the dangling pointer: BUG: KASAN: slab-use-after-free in br_mdb_dump Read of size 8 br_mdb_dump rtnl_mdb_dump rtnl_dumpit netlink_dump Reset the cursor to mp->ports after every deletion. The deletion removes at least the selected group, so the restarted walk always makes progress. Fixes: a6acb535afb2 ("bridge: mdb: Add MDB bulk deletion support") Cc: stable@vger.kernel.org Signed-off-by: Fourie Zhang Acked-by: Nikolay Aleksandrov Link: https://patch.msgid.link/20260920110852.60293-1-fouriezhang@tencent.com Signed-off-by: Jakub Kicinski --- net/bridge/br_mdb.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/net/bridge/br_mdb.c b/net/bridge/br_mdb.c index e0c7020b12f5..a01bd280c722 100644 --- a/net/bridge/br_mdb.c +++ b/net/bridge/br_mdb.c @@ -1523,6 +1523,8 @@ static void br_mdb_flush_pgs(struct net_bridge *br, } br_multicast_del_pg(mp, p, pp); + /* br_multicast_del_pg() can remove other groups from this list. */ + pp = &mp->ports; } } From fdfec06ac1eb5cdbc18d556c70a7b26837208218 Mon Sep 17 00:00:00 2001 From: Nicolai Buchwitz Date: Tue, 22 Sep 2026 09:31:40 +0200 Subject: [PATCH 476/694] MAINTAINERS: add Nicolai Buchwitz as GENET maintainer I have been contributing to and reviewing the GENET driver for a while now. Florian asked if I would like to formalize this commitment, so add myself as a maintainer. Signed-off-by: Nicolai Buchwitz Acked-by: Florian Fainelli Acked-by: Justin Chen Link: https://patch.msgid.link/20260922073140.1471858-1-nb@tipi-net.de Signed-off-by: Jakub Kicinski --- MAINTAINERS | 1 + 1 file changed, 1 insertion(+) diff --git a/MAINTAINERS b/MAINTAINERS index 3b2eb2a7a89a..7200905501b8 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -5454,6 +5454,7 @@ F: include/linux/brcmphy.h BROADCOM GENET ETHERNET DRIVER M: Doug Berger M: Florian Fainelli +M: Nicolai Buchwitz R: Broadcom internal kernel review list L: netdev@vger.kernel.org S: Maintained From 7e87508b5c4d81210d0a736ed01962e52f5c4c56 Mon Sep 17 00:00:00 2001 From: Nicolai Buchwitz Date: Tue, 22 Sep 2026 15:06:38 +0200 Subject: [PATCH 477/694] net: bcmgenet: stop Tx NAPI before disabling the queues bcmgenet_netif_stop() and the Wake-on-LAN branch of bcmgenet_suspend() both disable the Tx queues first and stop Tx NAPI several steps later. A completion in flight calls netif_tx_wake_queue() in between, and nothing stops the queue again, so a transmit can reach the rings after they have been freed. Close is safe because dev_deactivate_many() stops the qdisc first. bcmgenet_suspend() does not, so stop Tx NAPI before the queues on both paths. KASAN on a Raspberry Pi CM4, driven from an MTU change because suspend freezes user space before the callback runs: BUG: KASAN: use-after-free in bcmgenet_xmit+0x17f8/0x2258 Write of size 8 at addr ffffff8055844a68 by task ksoftirqd/0/14 bcmgenet_xmit+0x17f8/0x2258 dev_hard_start_xmit+0x13c/0x588 sch_direct_xmit+0x108/0x340 __dev_queue_xmit+0x1190/0x3848 Fixes: 254f3239dd07 ("net: bcmgenet: revise suspend/resume") Signed-off-by: Nicolai Buchwitz Link: https://patch.msgid.link/20260922130639.1660797-1-nb@tipi-net.de Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/broadcom/genet/bcmgenet.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/broadcom/genet/bcmgenet.c b/drivers/net/ethernet/broadcom/genet/bcmgenet.c index b916080f4ff1..ca62041efecd 100644 --- a/drivers/net/ethernet/broadcom/genet/bcmgenet.c +++ b/drivers/net/ethernet/broadcom/genet/bcmgenet.c @@ -3441,6 +3441,8 @@ static void bcmgenet_netif_stop(struct net_device *dev, bool stop_phy) { struct bcmgenet_priv *priv = netdev_priv(dev); + /* Stop completion polling before it can wake a stopped queue */ + bcmgenet_disable_tx_napi(priv); netif_tx_disable(dev); /* Disable MAC receive */ @@ -3455,7 +3457,6 @@ static void bcmgenet_netif_stop(struct net_device *dev, bool stop_phy) /* Disable MAC transmit. TX DMA disabled must be done before this */ umac_enable_set(priv, CMD_TX_EN, false); - bcmgenet_disable_tx_napi(priv); bcmgenet_disable_rx_napi(priv); bcmgenet_intr_disable(priv); @@ -4320,6 +4321,8 @@ static int bcmgenet_suspend(struct device *d) netif_device_detach(dev); if (device_may_wakeup(d) && priv->wolopts) { + /* Stop completion polling before it can wake a stopped queue */ + bcmgenet_disable_tx_napi(priv); netif_tx_disable(dev); /* Suspend non-wake Rx data flows */ @@ -4348,7 +4351,6 @@ static int bcmgenet_suspend(struct device *d) netdev_warn(priv->dev, "Timed out while disabling TX DMA\n"); - bcmgenet_disable_tx_napi(priv); bcmgenet_disable_rx_napi(priv); disable_irq(priv->irq1); bcmgenet_tx_reclaim_all(dev); From 7104a370714346b667712913dc16abf14bbc97ed Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Mon, 21 Sep 2026 16:18:56 -0700 Subject: [PATCH 478/694] veth: manage XDP program pointers during channel resize veth_set_channels() tears down XDP resources for removed RX queues without clearing rq->xdp_prog. If the program is then detached or replaced, those queues keep the old pointer after bpf_prog_put(). A later channel increase can re-enable NAPI and run the freed program. BUG: unable to handle page fault for address: ffffc90000256048 Oops: Oops: 0000 [#1] SMP KASAN NOPTI RIP: veth_xdp_rcv_skb (include/linux/filter.h:779 include/net/xdp.h:696 drivers/net/veth.c:820) Call Trace: veth_xdp_rcv (drivers/net/veth.c:941) veth_poll (drivers/net/veth.c:986) __napi_poll (net/core/dev.c:7787) net_rx_action (net/core/dev.c:7850 net/core/dev.c:8007) handle_softirqs (kernel/softirq.c:645) Kernel panic - not syncing: Fatal exception in interrupt Fixes: 4752eeb3d891 ("veth: implement support for set_channel ethtool op") Signed-off-by: Weiming Shi Acked-by: Stanislav Fomichev Reviewed-by: Jiayuan Chen Reviewed-by: Jason Xing Link: https://patch.msgid.link/20260921231856.1798630-1-kuba@kernel.org Signed-off-by: Jakub Kicinski --- drivers/net/veth.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/net/veth.c b/drivers/net/veth.c index 6ed3ee81153f..71227d0389aa 100644 --- a/drivers/net/veth.c +++ b/drivers/net/veth.c @@ -1054,6 +1054,7 @@ static int __veth_napi_enable_range(struct net_device *dev, int start, int end) for (i = start; i < end; i++) { struct veth_rq *rq = &priv->rq[i]; + rcu_assign_pointer(rq->xdp_prog, priv->_xdp_prog); napi_enable(&rq->xdp_napi); rcu_assign_pointer(priv->rq[i].napi, &priv->rq[i].xdp_napi); } @@ -1088,6 +1089,7 @@ static void veth_napi_del_range(struct net_device *dev, int start, int end) rcu_assign_pointer(priv->rq[i].napi, NULL); napi_disable(&rq->xdp_napi); + rcu_assign_pointer(rq->xdp_prog, NULL); __netif_napi_del(&rq->xdp_napi); } synchronize_net(); From 3b4e0b0c008a8c1b474730248cd5b873026c74bd Mon Sep 17 00:00:00 2001 From: Shihuang Liu Date: Sat, 19 Sep 2026 21:36:04 +0800 Subject: [PATCH 479/694] net: skbuff: fix pull-bound underflow in skb_checksum_setup_ipv6() skb_maybe_pull_tail() subtracts skb_headlen(skb) from the unsigned max argument and passes the result to __pskb_pull_tail() as a signed int. The function does not ensure that max is at least skb_headlen(skb). This can happen while parsing IPv6 extension headers when an skb already has a linear area larger than MAX_IPV6_HDR_LEN. Once the parser needs data beyond the linear area, max - skb_headlen(skb) wraps and is converted to a negative delta. __pskb_pull_tail() then passes that negative length to skb_copy_bits(), where it can become a very large copy length. Pass the requested length itself as the pull bound at the three extension-header call sites, so the delta can no longer go negative. Fixes: 1431fb31ecba ("xen-netback: fix fragment detection in checksum setup") Suggested-by: Eric Dumazet Signed-off-by: Shihuang Liu Link: https://patch.msgid.link/20260919133604.50948-1-shlomojune6@gmail.com Signed-off-by: Jakub Kicinski --- net/core/skbuff.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) diff --git a/net/core/skbuff.c b/net/core/skbuff.c index 609f2c7f4a47..b4edbd06655e 100644 --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@ -5977,7 +5977,8 @@ static int skb_checksum_setup_ipv6(struct sk_buff *skb, bool recalculate) err = skb_maybe_pull_tail(skb, off + sizeof(struct ipv6_opt_hdr), - MAX_IPV6_HDR_LEN); + off + + sizeof(struct ipv6_opt_hdr)); if (err < 0) goto out; @@ -5992,7 +5993,8 @@ static int skb_checksum_setup_ipv6(struct sk_buff *skb, bool recalculate) err = skb_maybe_pull_tail(skb, off + sizeof(struct ip_auth_hdr), - MAX_IPV6_HDR_LEN); + off + + sizeof(struct ip_auth_hdr)); if (err < 0) goto out; @@ -6007,7 +6009,8 @@ static int skb_checksum_setup_ipv6(struct sk_buff *skb, bool recalculate) err = skb_maybe_pull_tail(skb, off + sizeof(struct frag_hdr), - MAX_IPV6_HDR_LEN); + off + + sizeof(struct frag_hdr)); if (err < 0) goto out; From ab7aa05c06ae340e5c7530bb78fa8d23794e460b Mon Sep 17 00:00:00 2001 From: Ido Schimmel Date: Tue, 22 Sep 2026 16:12:39 +0300 Subject: [PATCH 480/694] vrf: Stop corrupting skb->csum when capturing CHECKSUM_COMPLETE packets The VRF device is an Ethernet device but it can have non-Ethernet ports such as IP tunnels. Before the cited commit, capturing packets from such ports on the VRF device resulted in these packets being detected as malformed since they lack an Ethernet header. The cited commit fixed it by pushing a dummy Ethernet header to such packets before the capture and pulling it afterwards. In the case of CHECKSUM_COMPLETE packets it also updated skb->csum with the checksum of the dummy Ethernet header. This is wrong as skb->csum should not include the checksum of the Ethernet header ("checksum of the _whole_ packet as seen by netif_rx()"). This also means that L4 protocols receive a corrupted skb->csum and potentially drop the packet, as is the case with UDP packets whose checksum was completed by software. Fix by removing the unnecessary call to skb_postpush_rcsum(). Fixes: 048939088220 ("vrf: add mac header for tunneled packets when sniffer is attached") Reported-by: Stefano Sasso Closes: https://lore.kernel.org/netdev/CALtE316UtL3x7LL6uxfXzx8rW6AbzYPeDOb478hqJCr_-dj=Wg@mail.gmail.com/ Signed-off-by: Ido Schimmel Reviewed-by: David Ahern Reviewed-by: Eric Dumazet Reviewed-by: Andrea Mayer Link: https://patch.msgid.link/20260922131239.2509494-1-idosch@nvidia.com Signed-off-by: Jakub Kicinski --- drivers/net/vrf.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/drivers/net/vrf.c b/drivers/net/vrf.c index a0557a3a7026..d4dc6d690a75 100644 --- a/drivers/net/vrf.c +++ b/drivers/net/vrf.c @@ -1175,8 +1175,6 @@ static int vrf_prepare_mac_header(struct sk_buff *skb, skb->protocol = eth->h_proto; skb->pkt_type = PACKET_HOST; - skb_postpush_rcsum(skb, skb->data, ETH_HLEN); - skb_pull_inline(skb, ETH_HLEN); return 0; From 2d959c75c27f90e9ec489d18ce5ee6b852ad4741 Mon Sep 17 00:00:00 2001 From: Hui Peng Date: Mon, 21 Sep 2026 04:40:25 +0000 Subject: [PATCH 481/694] ipv6: sr: enforce exact attribute length for SEG6_ATTR_DST In seg6_genl_policy, SEG6_ATTR_DST is defined with .type = NLA_BINARY and .len = sizeof(struct in6_addr). For NLA_BINARY, .len only enforces the maximum payload length and permits shorter payloads (e.g., 0 bytes). When seg6_genl_set_tunsrc() copies sizeof(struct in6_addr) bytes via kmemdup(val, sizeof(*val), GFP_KERNEL), a short SEG6_ATTR_DST attribute triggers a 16-byte out-of-bounds read past skb->tail into uninitialized skb->head memory, which is stored in sdata->tun_src and leaked back to userspace via SEG6_CMD_GET_TUNSRC. Switch SEG6_ATTR_DST in seg6_genl_policy to NLA_POLICY_EXACT_LEN(sizeof(struct in6_addr)) so that generic netlink validation rejects any attribute whose length is not exactly sizeof(struct in6_addr) with -ERANGE. Tested in QEMU against Linux 7.3.0-rc3 by sending a SEG6_CMD_SET_TUNSRC Generic Netlink message with a 0-byte SEG6_ATTR_DST attribute followed by SEG6_CMD_GET_TUNSRC. On the unfixed kernel, SEG6_CMD_SET_TUNSRC succeeds (err = 0) and SEG6_CMD_GET_TUNSRC leaks 16 bytes of uninitialized kernel heap memory (tun_src = 836a61ecc4d25a1042a8d60411cfb378); with this patch applied, SEG6_CMD_SET_TUNSRC is rejected by netlink policy validation with -ERANGE (-34) and tun_src remains zeroed. Fixes: 915d7e5e5930 ("ipv6: sr: add code base for control plane support of SR-IPv6") Cc: stable@vger.kernel.org Signed-off-by: Hui Peng Reviewed-by: Hangbin Liu Reviewed-by: Justin Iurman Reviewed-by: Andrea Mayer Link: https://patch.msgid.link/20260921044025.1535982-1-benquike@gmail.com Signed-off-by: Jakub Kicinski --- net/ipv6/seg6.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/net/ipv6/seg6.c b/net/ipv6/seg6.c index 62a7eb779202..8c2b156c227a 100644 --- a/net/ipv6/seg6.c +++ b/net/ipv6/seg6.c @@ -138,8 +138,8 @@ void seg6_icmp_srh(struct sk_buff *skb, struct inet6_skb_parm *opt) static struct genl_family seg6_genl_family; static const struct nla_policy seg6_genl_policy[SEG6_ATTR_MAX + 1] = { - [SEG6_ATTR_DST] = { .type = NLA_BINARY, - .len = sizeof(struct in6_addr) }, + [SEG6_ATTR_DST] = + NLA_POLICY_EXACT_LEN(sizeof(struct in6_addr)), [SEG6_ATTR_DSTLEN] = { .type = NLA_S32, }, [SEG6_ATTR_HMACKEYID] = { .type = NLA_U32, }, [SEG6_ATTR_SECRET] = { .type = NLA_BINARY, }, From 6b491af01aa5c0633a580e3b11bc7277adc903b2 Mon Sep 17 00:00:00 2001 From: Nicolo Giuliani Date: Mon, 21 Sep 2026 05:56:29 +0200 Subject: [PATCH 482/694] net: dsa: mv88e6xxx: 88E6191X and 88E6193X have no PTP The 88E6191X and 88E6193X are 6393 family devices that share mv88e6393x_ops with the 88E6393X and are marked as ptp_support. Marvell's UMSD driver describes both as parts without AVB (88E6193X: "BGA package - No AVB, No Routing, No Cut-through"), and the register access confirms it on an 88E6193X: the whole indirect AVB register space behind Global 2 registers 0x16 and 0x17 reads zero, for every port, block and address, with the 6390 and with the 6352 command encoding. Writes to the TAI registers, including the clock period register and the TAI global configuration register, read back as zero. Since commit 7e3c18097a70 ("net: dsa: mv88e6xxx: read cycle counter period from hardware") the PTP setup reads the TAI clock period, so the switch fails to probe: mv88e6xxx ...: unexpected cycle counter period of 0 ps Add mv88e6191x_ops, a copy of mv88e6393x_ops without avb_ops and ptp_ops, use it for the 88E6191X and the 88E6193X and stop setting ptp_support for them. The 88E6393X is unchanged. Tested on an 88E6193X (Sophos XGS 107w): the switch probes and the ports work. I do not have an 88E6191X, it is changed because UMSD describes it the same way. Fixes: de776d0d316f ("net: dsa: mv88e6xxx: add support for mv88e6393x family") Suggested-by: Andrew Lunn Signed-off-by: Nicolo Giuliani Reviewed-by: Andrew Lunn Link: https://patch.msgid.link/20260921-send-net-v2-1-031ad720f140@studio.unibo.it Signed-off-by: Jakub Kicinski --- drivers/net/dsa/mv88e6xxx/chip.c | 68 ++++++++++++++++++++++++++++++-- 1 file changed, 64 insertions(+), 4 deletions(-) diff --git a/drivers/net/dsa/mv88e6xxx/chip.c b/drivers/net/dsa/mv88e6xxx/chip.c index 7f68a0c55802..a4a8c7e11bf4 100644 --- a/drivers/net/dsa/mv88e6xxx/chip.c +++ b/drivers/net/dsa/mv88e6xxx/chip.c @@ -5639,6 +5639,68 @@ static const struct mv88e6xxx_ops mv88e6390x_ops = { .pcs_ops = &mv88e6390_pcs_ops, }; +static const struct mv88e6xxx_ops mv88e6191x_ops = { + /* MV88E6XXX_FAMILY_6393 without AVB and PTP: 6191X and 6193X */ + .irl_init_all = mv88e6390_g2_irl_init_all, + .get_eeprom = mv88e6xxx_g2_get_eeprom8, + .set_eeprom = mv88e6xxx_g2_set_eeprom8, + .set_switch_mac = mv88e6xxx_g2_set_switch_mac, + .phy_read = mv88e6xxx_g2_smi_phy_read_c22, + .phy_write = mv88e6xxx_g2_smi_phy_write_c22, + .phy_read_c45 = mv88e6xxx_g2_smi_phy_read_c45, + .phy_write_c45 = mv88e6xxx_g2_smi_phy_write_c45, + .port_set_link = mv88e6xxx_port_set_link, + .port_sync_link = mv88e6xxx_port_sync_link, + .port_set_rgmii_delay = mv88e6390_port_set_rgmii_delay, + .port_set_speed_duplex = mv88e6393x_port_set_speed_duplex, + .port_tag_remap = mv88e6390_port_tag_remap, + .port_set_policy = mv88e6393x_port_set_policy, + .port_set_frame_mode = mv88e6351_port_set_frame_mode, + .port_set_ucast_flood = mv88e6352_port_set_ucast_flood, + .port_set_mcast_flood = mv88e6352_port_set_mcast_flood, + .port_set_ether_type = mv88e6393x_port_set_ether_type, + .port_set_jumbo_size = mv88e6165_port_set_jumbo_size, + .port_egress_rate_limiting = mv88e6097_port_egress_rate_limiting, + .port_pause_limit = mv88e6390_port_pause_limit, + .port_disable_learn_limit = mv88e6xxx_port_disable_learn_limit, + .port_disable_pri_override = mv88e6xxx_port_disable_pri_override, + .port_get_cmode = mv88e6352_port_get_cmode, + .port_set_cmode = mv88e6393x_port_set_cmode, + .port_setup_message_port = mv88e6xxx_setup_message_port, + .port_set_upstream_port = mv88e6393x_port_set_upstream_port, + .port_enable_tcam = mv88e6xxx_port_enable_tcam, + .stats_snapshot = mv88e6390_g1_stats_snapshot, + .stats_set_histogram = mv88e6390_g1_stats_set_histogram, + .stats_get_sset_count = mv88e6320_stats_get_sset_count, + .stats_get_strings = mv88e6320_stats_get_strings, + .stats_get_stat = mv88e6390_stats_get_stat, + /* .set_cpu_port is missing because this family does not support a global + * CPU port, only per port CPU port which is set via + * .port_set_upstream_port method. + */ + .set_egress_port = mv88e6393x_set_egress_port, + .watchdog_ops = &mv88e6393x_watchdog_ops, + .mgmt_rsvd2cpu = mv88e6393x_port_mgmt_rsvd2cpu, + .pot_clear = mv88e6xxx_g2_pot_clear, + .hardware_reset_pre = mv88e6xxx_g2_eeprom_wait, + .hardware_reset_post = mv88e6xxx_g2_eeprom_wait, + .reset = mv88e6352_g1_reset, + .rmu_disable = mv88e6390_g1_rmu_disable, + .atu_get_hash = mv88e6165_g1_atu_get_hash, + .atu_set_hash = mv88e6165_g1_atu_set_hash, + .vtu_getnext = mv88e6390_g1_vtu_getnext, + .vtu_loadpurge = mv88e6390_g1_vtu_loadpurge, + .stu_getnext = mv88e6390_g1_stu_getnext, + .stu_loadpurge = mv88e6390_g1_stu_loadpurge, + .serdes_get_lane = mv88e6393x_serdes_get_lane, + .serdes_irq_mapping = mv88e6390_serdes_irq_mapping, + /* TODO: serdes stats */ + .gpio_ops = &mv88e6352_gpio_ops, + .phylink_get_caps = mv88e6393x_phylink_get_caps, + .pcs_ops = &mv88e6393x_pcs_ops, + .tcam_ops = &mv88e6393_tcam_ops, +}; + static const struct mv88e6xxx_ops mv88e6393x_ops = { /* MV88E6XXX_FAMILY_6393 */ .irl_init_all = mv88e6390_g2_irl_init_all, @@ -6163,8 +6225,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = { .atu_move_port_mask = 0x1f, .pvt = true, .multi_chip = true, - .ptp_support = true, - .ops = &mv88e6393x_ops, + .ops = &mv88e6191x_ops, }, [MV88E6193X] = { @@ -6190,8 +6251,7 @@ static const struct mv88e6xxx_info mv88e6xxx_table[] = { .atu_move_port_mask = 0x1f, .pvt = true, .multi_chip = true, - .ptp_support = true, - .ops = &mv88e6393x_ops, + .ops = &mv88e6191x_ops, }, [MV88E6220] = { From d22609f3d13fc5baacd92c222731b03c593401db Mon Sep 17 00:00:00 2001 From: Hui Peng Date: Mon, 21 Sep 2026 04:59:20 +0000 Subject: [PATCH 483/694] fou: reject omitted FOU_ATTR_IPPROTO on FOU_ENCAP_DIRECT Commit 7a9bc9e3f423 ("fou: Don't allow 0 for FOU_ATTR_IPPROTO.") added NLA_POLICY_MIN(NLA_U8, 1) to fou_nl_policy[FOU_ATTR_IPPROTO], which rejects an explicitly supplied FOU_ATTR_IPPROTO == 0 attribute with -ERANGE. However, FOU_ATTR_IPPROTO is an optional netlink attribute. When a user sends FOU_CMD_ADD with FOU_ATTR_TYPE set to FOU_ENCAP_DIRECT and omits FOU_ATTR_IPPROTO entirely, nla_policy validation succeeds and parse_nl_config() leaves cfg->protocol as 0 (from memset(cfg, 0, sizeof(*cfg))). fou_create() then creates a FOU_ENCAP_DIRECT socket with fou->protocol == 0. In fou_udp_recv(), returning -fou->protocol to udp_queue_rcv_one_skb() triggers IP protocol resubmission when fou->protocol > 0, whereas returning 0 tells the UDP tunnel layer that the skb was consumed without freeing it. When fou->protocol == 0, every packet received on the socket returns 0 from fou_udp_recv() and leaks the sk_buff. Reject FOU_ENCAP_DIRECT when !cfg->protocol in fou_create() so that creating a direct encapsulation port without FOU_ATTR_IPPROTO fails with -EINVAL while leaving FOU_CMD_DEL and FOU_CMD_GET (which share parse_nl_config()) unaffected. Tested in QEMU against Linux 7.3.0-rc3 by sending a FOU_CMD_ADD Generic Netlink request with FOU_ATTR_PORT = 5555 and FOU_ATTR_TYPE = FOU_ENCAP_DIRECT while omitting FOU_ATTR_IPPROTO. On the unfixed kernel, FOU_CMD_ADD succeeds (err = 0), FOU_CMD_GET reports fou->type = 1 and fou->protocol = 0, and sending 4000 UDP packets to 127.0.0.1:5555 leaks all 4000 sk_buffs (SUnreclaim in /proc/meminfo grows from 41456 kB to 59008 kB, +17552 kB); with this patch applied, FOU_CMD_ADD is rejected with -EINVAL (-22). Fixes: 23461551c006 ("fou: Support for foo-over-udp RX path") Fixes: 7a9bc9e3f423 ("fou: Don't allow 0 for FOU_ATTR_IPPROTO.") Cc: stable@vger.kernel.org Signed-off-by: Hui Peng Reviewed-by: Hangbin Liu Link: https://patch.msgid.link/20260921045920.1613098-1-benquike@gmail.com Signed-off-by: Jakub Kicinski --- net/ipv4/fou_core.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/net/ipv4/fou_core.c b/net/ipv4/fou_core.c index 5e867f1b5c1d..076fdca44f51 100644 --- a/net/ipv4/fou_core.c +++ b/net/ipv4/fou_core.c @@ -600,6 +600,10 @@ static int fou_create(struct net *net, struct fou_cfg *cfg, /* Initial for fou type */ switch (cfg->type) { case FOU_ENCAP_DIRECT: + if (!cfg->protocol) { + err = -EINVAL; + goto error; + } tunnel_cfg.encap_rcv = fou_udp_recv; tunnel_cfg.gro_receive = fou_gro_receive; tunnel_cfg.gro_complete = fou_gro_complete; From 89a8a1eef2d441b7825a6c0116ce817235a7ffe6 Mon Sep 17 00:00:00 2001 From: Jonas Jelonek Date: Fri, 18 Sep 2026 21:19:55 +0000 Subject: [PATCH 484/694] net: mdio: realtek-rtl9300: fix RTL931x C22 extended page selection The RTL931x indirect access engine has a separate nine-bit extended page field. The driver leaves it at zero, and otto_emdio_run_cmd() therefore programs extended page zero for every Clause 22 transaction. This overrides page selection made through PHY register 30, causing accesses to private PHY pages to hit extended page zero instead. Set the field to its 0x1ff "do not change" value for RTL931x Clause 22 reads and writes. This preserves extended page selection made through PHY register 30 and restores access to its private register pages. Fixes: 5ebdcac59aff ("net: mdio: realtek-rtl9300: Add support for RTL931x") Signed-off-by: Jonas Jelonek Acked-by: Markus Stockhausen Link: https://patch.msgid.link/20260918211955.3955777-1-jonas@jonasjelonek.de Signed-off-by: Jakub Kicinski --- drivers/net/mdio/mdio-realtek-rtl9300.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/drivers/net/mdio/mdio-realtek-rtl9300.c b/drivers/net/mdio/mdio-realtek-rtl9300.c index afd52a1cd7f8..9ce2b7807532 100644 --- a/drivers/net/mdio/mdio-realtek-rtl9300.c +++ b/drivers/net/mdio/mdio-realtek-rtl9300.c @@ -88,6 +88,8 @@ #define RTL9310_SMI_INDRT_ACCESS_BC_PHYID_CTRL 0x0c14 #define RTL9310_BC_PORT_ID GENMASK(10, 5) #define RTL9310_SMI_INDRT_ACCESS_CTRL_1 0x0c04 +#define RTL9310_SMI_INDRT_EXT_PAGE GENMASK(8, 0) +#define RTL9310_SMI_INDRT_EXT_PAGE_NO_CHANGE 0x1ff #define RTL9310_SMI_INDRT_ACCESS_CTRL_2_LOW 0x0c08 #define RTL9310_SMI_INDRT_ACCESS_CTRL_2_HIGH 0x0c0c #define RTL9310_SMI_INDRT_ACCESS_CTRL_3 0x0c10 /* I/O fields flipped */ @@ -325,6 +327,8 @@ static int otto_emdio_9310_read_c22(struct mii_bus *bus, int port, int regnum, u .broadcast = FIELD_PREP(RTL9310_BC_PORT_ID, port), .c22_data = FIELD_PREP(RTL9310_PHY_CTRL_REG_ADDR, regnum) | FIELD_PREP(RTL9310_PHY_CTRL_MAIN_PAGE, RAW_PAGE(priv)), + .ext_page = FIELD_PREP(RTL9310_SMI_INDRT_EXT_PAGE, + RTL9310_SMI_INDRT_EXT_PAGE_NO_CHANGE), }; return otto_emdio_read_cmd(bus, RTL9310_PHY_CTRL_TYPE_C22, &cmd_data, @@ -337,6 +341,8 @@ static int otto_emdio_9310_write_c22(struct mii_bus *bus, int port, int regnum, struct otto_emdio_cmd_regs cmd_data = { .c22_data = FIELD_PREP(RTL9310_PHY_CTRL_REG_ADDR, regnum) | FIELD_PREP(RTL9310_PHY_CTRL_MAIN_PAGE, RAW_PAGE(priv)), + .ext_page = FIELD_PREP(RTL9310_SMI_INDRT_EXT_PAGE, + RTL9310_SMI_INDRT_EXT_PAGE_NO_CHANGE), .io_data = FIELD_PREP(RTL9310_PHY_CTRL_INDATA, value), .port_mask_high = (u32)(BIT_ULL(port) >> 32), .port_mask_low = (u32)(BIT_ULL(port)), From db762fd96be225bd06161c9631160c755d891693 Mon Sep 17 00:00:00 2001 From: Johan Almbladh Date: Wed, 23 Sep 2026 12:51:57 +0200 Subject: [PATCH 485/694] bpf: Fix immediate JMP JEQ/JNE on MIPS32 An addu instruction was emitted instead of addiu, causing the immediate value 1 to be interpreted as register $at. This made the comparison result invalid when the immediate operand was negative. Note that $at is mapped to BPF_REG_AX, which is used for constant blinding. Fix the instruction to use the immediate form. Found with test_bpf on MIPS32r1 emulated by QEMU. Fixes: eb63cfcd2ee8 ("mips, bpf: Add eBPF JIT for 32-bit MIPS") Signed-off-by: Johan Almbladh Signed-off-by: Alexei Starovoitov Link: https://patch.msgid.link/20260923105158.3514342-1-johan.almbladh@anyfinetworks.com --- arch/mips/net/bpf_jit_comp32.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/mips/net/bpf_jit_comp32.c b/arch/mips/net/bpf_jit_comp32.c index 40a878b672f5..15a2a153dc87 100644 --- a/arch/mips/net/bpf_jit_comp32.c +++ b/arch/mips/net/bpf_jit_comp32.c @@ -1111,7 +1111,7 @@ static void emit_jmp_i64(struct jit_context *ctx, emit(ctx, xor, tmp, lo(dst), tmp); } if (imm < 0) { /* Compare sign extension */ - emit(ctx, addu, MIPS_R_T9, hi(dst), 1); + emit(ctx, addiu, MIPS_R_T9, hi(dst), 1); emit(ctx, or, tmp, tmp, MIPS_R_T9); } else { /* Compare zero extension */ emit(ctx, or, tmp, tmp, hi(dst)); From 8110ba09777873443db286b3cbb89b0e6311c554 Mon Sep 17 00:00:00 2001 From: Johan Almbladh Date: Wed, 23 Sep 2026 12:51:58 +0200 Subject: [PATCH 486/694] bpf: Fix BSWAP 32 and 16 on MIPS64 The 16/32-bit byteswap implementations for MIPS64r1 and earlier do not have an explicit zero extension afterwards. The input is first sign-extended to 64 bits, and the byteswap sequence can then leave the result sign-extended depending on the value of the low bits. Add the missing zero-extension. Found with test_bpf on MIPS64r1 emulated by QEMU. Fixes: fbc802de6b10 ("mips, bpf: Add new eBPF JIT for 64-bit MIPS") Signed-off-by: Johan Almbladh Signed-off-by: Alexei Starovoitov Link: https://patch.msgid.link/20260923105158.3514342-2-johan.almbladh@anyfinetworks.com --- arch/mips/net/bpf_jit_comp64.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/arch/mips/net/bpf_jit_comp64.c b/arch/mips/net/bpf_jit_comp64.c index fa7e9aa37f49..6681ccac9dd9 100644 --- a/arch/mips/net/bpf_jit_comp64.c +++ b/arch/mips/net/bpf_jit_comp64.c @@ -305,8 +305,7 @@ static void emit_bswap_r64(struct jit_context *ctx, u8 dst, u32 width) case 16: emit_sext(ctx, dst, dst); emit_bswap_r(ctx, dst, width); - if (cpu_has_mips64r2 || cpu_has_mips64r6) - emit_zext(ctx, dst); + emit_zext(ctx, dst); break; } clobber_reg(ctx, dst); From d8b6529e80bcb4fb8177121404cbb3377acaebd2 Mon Sep 17 00:00:00 2001 From: Zijie Huang Date: Mon, 21 Sep 2026 01:36:19 +0800 Subject: [PATCH 487/694] net: arp: terminate device name before lookup The ARP ioctl copies a user-provided struct arpreq into a stack object. Its arp_dev field may contain IFNAMSIZ bytes without a NUL terminator. Such input is passed to dev_get_by_name_rcu() or __dev_get_by_name(), where strcmp() can read past the end of the stack object when a matching alternative interface name exists. Terminate the field before the lookup to prevent the out-of-bounds read. Fixes: 36fbf1e52bd3 ("net: rtnetlink: add linkprop commands to add and delete alternative ifnames") Cc: stable@vger.kernel.org Reported-by: Vega Signed-off-by: Zijie Huang Signed-off-by: Ren Wei Reviewed-by: Ido Schimmel Link: https://patch.msgid.link/fabf02a70787d17299e4b3153eadffaf20d154b3.1789910973.git.milkory@outlook.com Signed-off-by: Jakub Kicinski --- net/ipv4/arp.c | 1 + 1 file changed, 1 insertion(+) diff --git a/net/ipv4/arp.c b/net/ipv4/arp.c index d409f606aec0..60009d92e071 100644 --- a/net/ipv4/arp.c +++ b/net/ipv4/arp.c @@ -1278,6 +1278,7 @@ int arp_ioctl(struct net *net, unsigned int cmd, void __user *arg) err = copy_from_user(&r, arg, sizeof(struct arpreq)); if (err) return -EFAULT; + r.arp_dev[IFNAMSIZ - 1] = '\0'; break; default: return -EINVAL; From 4eb3f195ef08c5acaed87958297e41cc49588dde Mon Sep 17 00:00:00 2001 From: Ivan Delalande Date: Fri, 18 Sep 2026 15:47:15 -0700 Subject: [PATCH 488/694] tg3: use random MAC address when tg3_get_device_address fails Some of the tg3 NICs we use (BCM57762) reset the SRAM MAC address to the placeholder address on link flaps, tg3_chip_reset, etc. We've typically fixed it from userspace, but since e4c00ba7274b ("tg3: replace placeholder MAC address with device property") was merged, tg3 just fails probe as we don't have a way to get it through the generic device_get_mac_address infrastructure as fallback on our systems. Make the driver assign a random address in this condition instead of being fatal for probe. Set deferred_probe_reason through dev_warn_probe if the address isn't yet available from the provider. Fixes: e4c00ba7274b ("tg3: replace placeholder MAC address with device property") Suggested-by: Jakub Kicinski Link: https://lore.kernel.org/netdev/20260909191751.651aa5c4@kernel.org/ Signed-off-by: Ivan Delalande Link: https://patch.msgid.link/20260918224715.GA654128@visor Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/broadcom/tg3.c | 11 +++++++---- 1 file changed, 7 insertions(+), 4 deletions(-) diff --git a/drivers/net/ethernet/broadcom/tg3.c b/drivers/net/ethernet/broadcom/tg3.c index caa7a6caa6e2..8b6806a79edf 100644 --- a/drivers/net/ethernet/broadcom/tg3.c +++ b/drivers/net/ethernet/broadcom/tg3.c @@ -17915,11 +17915,14 @@ static int tg3_init_one(struct pci_dev *pdev, err = tg3_get_device_address(tp, addr); if (err) { - dev_err(&pdev->dev, - "Could not obtain valid ethernet address, aborting\n"); - goto err_out_apeunmap; + dev_warn_probe(&pdev->dev, err, + "Could not obtain a valid ethernet address\n"); + if (err == -EPROBE_DEFER) + goto err_out_apeunmap; + eth_hw_addr_random(dev); + } else { + eth_hw_addr_set(dev, addr); } - eth_hw_addr_set(dev, addr); intmbx = MAILBOX_INTERRUPT_0 + TG3_64BIT_REG_LOW; rcvmbx = MAILBOX_RCVRET_CON_IDX_0 + TG3_64BIT_REG_LOW; From 0a7822e34a0bfde31b194ac3da3253e5032b44cc Mon Sep 17 00:00:00 2001 From: Lorenzo Bianconi Date: Mon, 21 Sep 2026 16:46:18 +0200 Subject: [PATCH 489/694] net: stmmac: clear stale buf->page after recycling on skb build failure In stmmac_rx(), when napi_build_skb() fails the descriptor page is recycled back to the page pool with page_pool_recycle_direct(), but buf->page is left pointing at the recycled page, unlike every other consumption site in the function which clears the pointer after handing the page away. With the stale pointer stmmac_rx_refill() skips the replacement allocation and programs the already-recycled page back into the RX descriptor. Clear buf->page on the napi_build_skb() failure path to keep the buffer lifecycle consistent with the other consumption sites. Fixes: df542f669307 ("net: stmmac: Switch to zero-copy in non-XDP RX path") Signed-off-by: Lorenzo Bianconi Reviewed-by: Maxime Chevallier Link: https://patch.msgid.link/20260921-stmmac-fix-napi-build-skb-error-v1-1-3d54bf6d9bb6@oss.qualcomm.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c index d5a984ad864f..3f34d491c959 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c @@ -5889,6 +5889,7 @@ static int stmmac_rx(struct stmmac_priv *priv, int limit, u32 queue) if (!skb) { page_pool_recycle_direct(rx_q->page_pool, buf->page); + buf->page = NULL; rx_dropped++; count++; goto drain_data; From 87cd6b717e4069dca34e0866e57eaeb44d3b173e Mon Sep 17 00:00:00 2001 From: Yuya Kusakabe Date: Tue, 22 Sep 2026 05:49:56 +0900 Subject: [PATCH 490/694] net: ipv6: keep room for the mac header in dst_dev_overhead() The seg6, ioam6 and rpl lwtunnels size their skb_cow_head() request as the length they are about to push plus dst_dev_overhead(), then push the new headers and rebuild the mac header below them with skb_mac_header_rebuild(). That rebuild needs skb->mac_len of headroom, but dst_dev_overhead() leaves LL_RESERVED_SPACE() of the egress device, 16 bytes for plain Ethernet. Where the mac header is longer than that, as it is on ingress through a VLAN device with reorder_hdr off, the rebuild runs out of room: skb_set_mac_header(skb, -skb->mac_len) computes a negative offset, stores it unchecked in the u16 skb->mac_header, and the memmove that follows writes skb->mac_len bytes about 64 KB past skb->head. Forwarding plain ping6 traffic through such a device reproduces it on all five seg6 encapsulation modes and on the rpl and ioam6 inline paths; skb->mac_header comes back as 65534 on a 704-byte head. Return the larger of the two. The helper already returns skb->mac_len when it has no dst, so this only makes the other branch agree, and it covers every caller rather than each call site in turn. Fixes: 40475b63761a ("net: ipv6: seg6_iptunnel: mitigate 2-realloc issue") Fixes: dce525185bc9 ("net: ipv6: ioam6_iptunnel: mitigate 2-realloc issue") Fixes: 985ec6f5e623 ("net: ipv6: rpl_iptunnel: mitigate 2-realloc issue") Suggested-by: Andrea Mayer Signed-off-by: Yuya Kusakabe Reviewed-by: Justin Iurman Reviewed-by: Gabriel Goller Reviewed-by: Eric Dumazet Reviewed-by: Andrea Mayer Link: https://patch.msgid.link/20260922-seg6-maclen-headroom-v3-1-7b2f982ef79d@gmail.com Signed-off-by: Jakub Kicinski --- include/net/dst.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/include/net/dst.h b/include/net/dst.h index 307073eae7f8..dbedfe72e1fd 100644 --- a/include/net/dst.h +++ b/include/net/dst.h @@ -455,7 +455,8 @@ static inline unsigned int dst_dev_overhead(struct dst_entry *dst, struct sk_buff *skb) { if (likely(dst)) - return LL_RESERVED_SPACE(dst->dev); + return max_t(unsigned int, skb->mac_len, + LL_RESERVED_SPACE(dst->dev)); return skb->mac_len; } From 0e2bec77ea62895416600c90588f593516572bca Mon Sep 17 00:00:00 2001 From: Florian Fainelli Date: Mon, 21 Sep 2026 15:00:17 -0700 Subject: [PATCH 491/694] net: bcmgenet: fix 64-bit RTNL stats reading in ethtool on 32-bit systems When bcmgenet was converted to 64-bit statistics, STAT_RTNL members were switched to point into struct rtnl_link_stats64, whose fields are 64-bit (__u64) regardless of architecture. However, bcmgenet_get_ethtool_stats() retained a legacy check: if (sizeof(unsigned long) != sizeof(u32) && s->stat_sizeof == sizeof(unsigned long)) On 32-bit systems, sizeof(unsigned long) == sizeof(u32), causing this condition to evaluate to false. As a result, 64-bit RTNL stats fields were read via *(u32 *)p. On 32-bit Big-Endian systems (such as MIPS BE), this reads the high 32 bits and returns 0 until the counter exceeds 4GB; on 32-bit Little-Endian systems (such as 32-bit ARM), the value is truncated to 32 bits. Fix this by checking if s->stat_sizeof == sizeof(u64) so 64-bit fields are always read as 64-bit values. Fixes: 59aa6e3072aa ("net: bcmgenet: switch to use 64bit statistics") Reviewed-by: Nicolai Buchwitz Signed-off-by: Florian Fainelli Link: https://patch.msgid.link/20260921220021.281418-2-florian.fainelli@broadcom.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/broadcom/genet/bcmgenet.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/drivers/net/ethernet/broadcom/genet/bcmgenet.c b/drivers/net/ethernet/broadcom/genet/bcmgenet.c index ca62041efecd..47a6c073c8d9 100644 --- a/drivers/net/ethernet/broadcom/genet/bcmgenet.c +++ b/drivers/net/ethernet/broadcom/genet/bcmgenet.c @@ -1346,9 +1346,8 @@ static void bcmgenet_get_ethtool_stats(struct net_device *dev, p = (char *)&stats64; p += s->stat_offset; - if (sizeof(unsigned long) != sizeof(u32) && - s->stat_sizeof == sizeof(unsigned long)) - data[i] = *(unsigned long *)p; + if (s->stat_sizeof == sizeof(u64)) + data[i] = *(u64 *)p; else data[i] = *(u32 *)p; } From 3aeaa609fda19c09d5298c9fedaaa3b6229601b5 Mon Sep 17 00:00:00 2001 From: Florian Fainelli Date: Mon, 21 Sep 2026 15:00:18 -0700 Subject: [PATCH 492/694] net: bcmgenet: initialize u64 stats seq counter for all queues bcmgenet_gstrings_stats statically defines ethtool statistics for queues 0 through GENET_MAX_MQ_CNT (4). However, bcmgenet_probe() only initialized the u64_stats_sync seq counter up to priv->hw_params->rx_queues and priv->hw_params->tx_queues. Since priv->hw_params->rx_queues is 0 across all hardware versions (and priv->hw_params->tx_queues is 0 on GENET V1), rings 1..4 have uninitialized u64_stats_sync structures. When ethtool -S is run on 32-bit kernels, bcmgenet_get_ethtool_stats() reads stats from rx_rings[1..4], causing lockdep warnings due to the uninitialized sequence counters. Initialize the sequence counters for all GENET_MAX_MQ_CNT + 1 queues. Fixes: ffc2c8c4a714 ("net: bcmgenet: Initialize u64 stats seq counter") Reviewed-by: Nicolai Buchwitz Signed-off-by: Florian Fainelli Link: https://patch.msgid.link/20260921220021.281418-3-florian.fainelli@broadcom.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/broadcom/genet/bcmgenet.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/broadcom/genet/bcmgenet.c b/drivers/net/ethernet/broadcom/genet/bcmgenet.c index 47a6c073c8d9..2ab37bb031ed 100644 --- a/drivers/net/ethernet/broadcom/genet/bcmgenet.c +++ b/drivers/net/ethernet/broadcom/genet/bcmgenet.c @@ -4135,10 +4135,10 @@ static int bcmgenet_probe(struct platform_device *pdev) priv->rx_rings[i].rx_max_coalesced_frames = 1; /* Initialize u64 stats seq counter for 32bit machines */ - for (i = 0; i <= priv->hw_params->rx_queues; i++) + for (i = 0; i <= GENET_MAX_MQ_CNT; i++) { u64_stats_init(&priv->rx_rings[i].stats64.syncp); - for (i = 0; i <= priv->hw_params->tx_queues; i++) u64_stats_init(&priv->tx_rings[i].stats64.syncp); + } /* libphy will determine the link state */ netif_carrier_off(dev); From cbbc1aee7776c7fa1d89e6cb963a23e58c495dca Mon Sep 17 00:00:00 2001 From: Florian Fainelli Date: Mon, 21 Sep 2026 15:00:19 -0700 Subject: [PATCH 493/694] net: bcmgenet: do not skip WoL power up on GENET V1 bcmgenet_power_up() had an early check for bcmgenet_has_ext(priv) before dispatching by power mode. GENET V1 does not have the EXT block (unlike GENET V2+), which causes bcmgenet_power_up() to immediately return 0. As a consequence, when waking up from GENET_POWER_WOL_MAGIC on GENET V1, bcmgenet_wol_power_up_cfg() is never invoked to disable the WoL clock, clear wake event masks, and restore normal PHY and MAC operations. Move the bcmgenet_has_ext() checks to the GENET_POWER_PASSIVE and GENET_POWER_CABLE_SENSE cases where the EXT registers are actually accessed, allowing GENET_POWER_WOL_MAGIC cleanup to execute on all hardware versions. Fixes: c3ae64ae0c08 ("net: bcmgenet: handle GENET_POWER_WOL_MAGIC") Reviewed-by: Nicolai Buchwitz Signed-off-by: Florian Fainelli Link: https://patch.msgid.link/20260921220021.281418-4-florian.fainelli@broadcom.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/broadcom/genet/bcmgenet.c | 13 ++++++++----- 1 file changed, 8 insertions(+), 5 deletions(-) diff --git a/drivers/net/ethernet/broadcom/genet/bcmgenet.c b/drivers/net/ethernet/broadcom/genet/bcmgenet.c index 2ab37bb031ed..07032e4193ea 100644 --- a/drivers/net/ethernet/broadcom/genet/bcmgenet.c +++ b/drivers/net/ethernet/broadcom/genet/bcmgenet.c @@ -1762,13 +1762,12 @@ static int bcmgenet_power_up(struct bcmgenet_priv *priv, int ret = 0; u32 reg; - if (!bcmgenet_has_ext(priv)) - return ret; - - reg = bcmgenet_ext_readl(priv, EXT_EXT_PWR_MGMT); - switch (mode) { case GENET_POWER_PASSIVE: + if (!bcmgenet_has_ext(priv)) + break; + + reg = bcmgenet_ext_readl(priv, EXT_EXT_PWR_MGMT); reg &= ~(EXT_PWR_DOWN_DLL | EXT_PWR_DOWN_BIAS | EXT_ENERGY_DET_MASK); if (GENET_IS_V5(priv) && !bcmgenet_has_ephy_16nm(priv)) { @@ -1792,8 +1791,12 @@ static int bcmgenet_power_up(struct bcmgenet_priv *priv, break; case GENET_POWER_CABLE_SENSE: + if (!bcmgenet_has_ext(priv)) + break; + /* enable APD */ if (!GENET_IS_V5(priv)) { + reg = bcmgenet_ext_readl(priv, EXT_EXT_PWR_MGMT); reg |= EXT_PWR_DN_EN_LD; bcmgenet_ext_writel(priv, reg, EXT_EXT_PWR_MGMT); } From 273941c85fc2632cd3e56ddff737b9245de7697d Mon Sep 17 00:00:00 2001 From: Florian Fainelli Date: Mon, 21 Sep 2026 15:00:20 -0700 Subject: [PATCH 494/694] net: bcmgenet: validate Ethernet address in bcmgenet_set_mac_addr bcmgenet_set_mac_addr() did not check whether the provided MAC address is a valid Ethernet address before applying it. Userspace could configure an invalid address (such as all zeroes or a multicast address) while the interface is down. Add a call to is_valid_ether_addr() and return -EADDRNOTAVAIL if the MAC address is not valid. Fixes: 1c1008c793fa ("net: bcmgenet: add main driver file") Reviewed-by: Nicolai Buchwitz Signed-off-by: Florian Fainelli Link: https://patch.msgid.link/20260921220021.281418-5-florian.fainelli@broadcom.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/broadcom/genet/bcmgenet.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/net/ethernet/broadcom/genet/bcmgenet.c b/drivers/net/ethernet/broadcom/genet/bcmgenet.c index 07032e4193ea..011376a1678a 100644 --- a/drivers/net/ethernet/broadcom/genet/bcmgenet.c +++ b/drivers/net/ethernet/broadcom/genet/bcmgenet.c @@ -3635,6 +3635,9 @@ static int bcmgenet_set_mac_addr(struct net_device *dev, void *p) if (netif_running(dev)) return -EBUSY; + if (!is_valid_ether_addr(addr->sa_data)) + return -EADDRNOTAVAIL; + eth_hw_addr_set(dev, addr->sa_data); return 0; From d64e277b955be4506931802837499b62c8f3968a Mon Sep 17 00:00:00 2001 From: Florian Fainelli Date: Mon, 21 Sep 2026 15:00:21 -0700 Subject: [PATCH 495/694] net: bcmgenet: mask DMA_TIMEOUT_MASK when reading DMA_RING0_TIMEOUT bcmgenet_get_coalesce() reads DMA_RING0_TIMEOUT to calculate rx_coalesce_usecs without masking out bits outside DMA_TIMEOUT_MASK (16 bits). If upper bits are non-zero or contain status/flags, the computed value of rx_coalesce_usecs returned to userspace via ethtool becomes corrupted. Mask the register read with DMA_TIMEOUT_MASK before computing the timeout in microseconds. Fixes: 4a29645bfe6c ("net: bcmgenet: Implement RX coalescing control knobs") Reviewed-by: Nicolai Buchwitz Signed-off-by: Florian Fainelli Link: https://patch.msgid.link/20260921220021.281418-6-florian.fainelli@broadcom.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/broadcom/genet/bcmgenet.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/broadcom/genet/bcmgenet.c b/drivers/net/ethernet/broadcom/genet/bcmgenet.c index 011376a1678a..21668e41b696 100644 --- a/drivers/net/ethernet/broadcom/genet/bcmgenet.c +++ b/drivers/net/ethernet/broadcom/genet/bcmgenet.c @@ -852,7 +852,8 @@ static int bcmgenet_get_coalesce(struct net_device *dev, ec->rx_max_coalesced_frames = bcmgenet_rdma_ring_readl(priv, 0, DMA_MBUF_DONE_THRESH); ec->rx_coalesce_usecs = - bcmgenet_rdma_readl(priv, DMA_RING0_TIMEOUT) * 8192 / 1000; + (bcmgenet_rdma_readl(priv, DMA_RING0_TIMEOUT) & + DMA_TIMEOUT_MASK) * 8192 / 1000; for (i = 0; i <= priv->hw_params->rx_queues; i++) { ring = &priv->rx_rings[i]; From 4bdee8060d1e4581624e68fbd369b1afb14df4bc Mon Sep 17 00:00:00 2001 From: Myeonghun Pak Date: Mon, 21 Sep 2026 20:09:14 -0400 Subject: [PATCH 496/694] net: airoha: npu: cancel wdt_work after releasing the WDT IRQ airoha_npu_remove() calls cancel_work_sync() on each core's wdt_work, but the watchdog IRQ that queues it is requested with devm_request_irq() and is freed only after .remove() returns. airoha_npu_wdt_handler() can therefore schedule_work() again once the cancel has returned. struct airoha_npu, which contains the work, is devm_kzalloc()'d and is freed in that same unwind, so the late work dereferences freed memory. Register the work with devm_work_autocancel() before devm_request_irq() and drop .remove(). Devres runs in reverse order, so the IRQ is freed before cancel_work_sync(), including when probe fails. A cancel left in .remove() cannot get that order. Initializing the work first also stops a pending watchdog interrupt from queuing an uninitialized work item. Probe currently calls INIT_WORK() only after devm_request_irq(). This issue was identified during our ongoing static-analysis research while reviewing kernel code. Fixes: 23290c7bc190 ("net: airoha: Introduce Airoha NPU support") Cc: stable@vger.kernel.org # 6.15+ Co-developed-by: Ijae Kim Signed-off-by: Ijae Kim Signed-off-by: Myeonghun Pak Acked-by: Lorenzo Bianconi Link: https://patch.msgid.link/20260922000914.542068-1-mhun512@gmail.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/airoha/airoha_npu.c | 18 ++++++------------ 1 file changed, 6 insertions(+), 12 deletions(-) diff --git a/drivers/net/ethernet/airoha/airoha_npu.c b/drivers/net/ethernet/airoha/airoha_npu.c index 5bb4817a898d..4d3195eb00f7 100644 --- a/drivers/net/ethernet/airoha/airoha_npu.c +++ b/drivers/net/ethernet/airoha/airoha_npu.c @@ -5,6 +5,7 @@ */ #include +#include #include #include #include @@ -751,12 +752,15 @@ static int airoha_npu_probe(struct platform_device *pdev) if (irq < 0) return irq; + err = devm_work_autocancel(dev, &core->wdt_work, + airoha_npu_wdt_work); + if (err) + return err; + err = devm_request_irq(dev, irq, airoha_npu_wdt_handler, IRQF_SHARED, "airoha-npu-wdt", core); if (err) return err; - - INIT_WORK(&core->wdt_work, airoha_npu_wdt_work); } /* wlan IRQ lines */ @@ -803,18 +807,8 @@ static int airoha_npu_probe(struct platform_device *pdev) return 0; } -static void airoha_npu_remove(struct platform_device *pdev) -{ - struct airoha_npu *npu = platform_get_drvdata(pdev); - int i; - - for (i = 0; i < ARRAY_SIZE(npu->cores); i++) - cancel_work_sync(&npu->cores[i].wdt_work); -} - static struct platform_driver airoha_npu_driver = { .probe = airoha_npu_probe, - .remove = airoha_npu_remove, .driver = { .name = "airoha-npu", .of_match_table = of_airoha_npu_match, From a3f315be9d30eeb6938d11fa17fd4b32d52f7c42 Mon Sep 17 00:00:00 2001 From: Xuanqiang Luo Date: Mon, 21 Sep 2026 11:18:59 +0800 Subject: [PATCH 497/694] ip_gre: Reject enabling collect metadata through changelink ipgre_netlink_parms() can enable collect_md on an existing GRE, GRETAP or ERSPAN device. Unlike newlink, changelink does not enforce metadata tunnel uniqueness. Converting a non-metadata device can therefore replace the metadata receive entry for another device of the same type in the same netns. Deleting either device then clears the shared entry, breaking metadata receive lookup for the surviving device. If parameter validation fails after collect_md is set, deleting the modified device can also clear an entry it never owned. Reject enabling metadata mode in both changelink callbacks before any encapsulation or tunnel parameters are modified. Allow requests that repeat the metadata attribute on an existing metadata device. Fixes: 2e15ea390e6f ("ip_gre: Add support to collect tunnel metadata.") Signed-off-by: Xuanqiang Luo Reviewed-by: Ido Schimmel Reviewed-by: Hangbin Liu Link: https://patch.msgid.link/20260921031859.9283-1-xuanqiang.luo@linux.dev Signed-off-by: Jakub Kicinski --- net/ipv4/ip_gre.c | 12 ++++++++++++ 1 file changed, 12 insertions(+) diff --git a/net/ipv4/ip_gre.c b/net/ipv4/ip_gre.c index 82309efd417e..e4878e9aa636 100644 --- a/net/ipv4/ip_gre.c +++ b/net/ipv4/ip_gre.c @@ -1464,6 +1464,12 @@ static int ipgre_changelink(struct net_device *dev, struct nlattr *tb[], if (!rtnl_dev_link_net_capable(dev, t->net)) return -EPERM; + if (data && data[IFLA_GRE_COLLECT_METADATA] && !t->collect_md) { + NL_SET_ERR_MSG(extack, + "Enabling collect_md on an existing device is not supported"); + return -EOPNOTSUPP; + } + err = ipgre_newlink_encap_setup(dev, data); if (err) return err; @@ -1496,6 +1502,12 @@ static int erspan_changelink(struct net_device *dev, struct nlattr *tb[], if (!rtnl_dev_link_net_capable(dev, t->net)) return -EPERM; + if (data && data[IFLA_GRE_COLLECT_METADATA] && !t->collect_md) { + NL_SET_ERR_MSG(extack, + "Enabling collect_md on an existing device is not supported"); + return -EOPNOTSUPP; + } + err = ipgre_newlink_encap_setup(dev, data); if (err) return err; From 0f2fd31f63c65c409bd336af3a42d478a9207c1f Mon Sep 17 00:00:00 2001 From: Gilberto Conde Date: Mon, 21 Sep 2026 10:34:26 +0100 Subject: [PATCH 498/694] net: usb: qmi_wwan: add Quectel EG120K-EA Add support for the Quectel EG120K-EA LTE Cat.12 module (USB ID 2c7c:030b). Its QMI interface (interface 4) uses class/subclass/protocol ff/ff/ff like the other recent Quectel modules, so match it the same way. The product ID is shared with the EM060K, which the option driver already knows and uses to claim the serial interfaces. Without a qmi_wwan entry the data interface is left unbound. Tested on a GL.iNet GL-X2000, where the module is soldered down and enumerates at SuperSpeed: cdc-wdm0 and wwan0 appear and ModemManager brings up a data session. T: Bus=02 Lev=01 Prnt=01 Port=00 Cnt=01 Dev#= 2 Spd=5000 MxCh= 0 D: Ver= 3.10 Cls=00(>ifc ) Sub=00 Prot=00 MxPS= 9 #Cfgs= 1 P: Vendor=2c7c ProdID=030b Rev= 5.04 S: Manufacturer=Quectel S: Product=EG120K-EA S: SerialNumber=45546267 C:* #Ifs= 6 Cfg#= 1 Atr=a0 MxPwr=896mA I:* If#= 0 Alt= 0 #EPs= 2 Cls=ff(vend.) Sub=ff Prot=30 Driver=option E: Ad=01(O) Atr=02(Bulk) MxPS=1024 Ivl=0ms E: Ad=81(I) Atr=02(Bulk) MxPS=1024 Ivl=0ms I:* If#= 1 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=00 Prot=40 Driver=option E: Ad=83(I) Atr=03(Int.) MxPS= 10 Ivl=32ms E: Ad=82(I) Atr=02(Bulk) MxPS=1024 Ivl=0ms E: Ad=02(O) Atr=02(Bulk) MxPS=1024 Ivl=0ms I:* If#= 2 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=ff Prot=40 Driver=option E: Ad=85(I) Atr=03(Int.) MxPS= 10 Ivl=32ms E: Ad=84(I) Atr=02(Bulk) MxPS=1024 Ivl=0ms E: Ad=03(O) Atr=02(Bulk) MxPS=1024 Ivl=0ms I:* If#= 3 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=ff Prot=40 Driver=option E: Ad=87(I) Atr=03(Int.) MxPS= 10 Ivl=32ms E: Ad=86(I) Atr=02(Bulk) MxPS=1024 Ivl=0ms E: Ad=04(O) Atr=02(Bulk) MxPS=1024 Ivl=0ms I:* If#= 4 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=ff Prot=ff Driver=qmi_wwan E: Ad=88(I) Atr=03(Int.) MxPS= 8 Ivl=32ms E: Ad=8e(I) Atr=02(Bulk) MxPS=1024 Ivl=0ms E: Ad=0f(O) Atr=02(Bulk) MxPS=1024 Ivl=0ms I:* If#=12 Alt= 0 #EPs= 1 Cls=ff(vend.) Sub=ff Prot=70 Driver=(none) E: Ad=89(I) Atr=02(Bulk) MxPS=1024 Ivl=0ms Signed-off-by: Gilberto Conde Link: https://patch.msgid.link/20260921093426.2870266-1-gilbertorconde@gmail.com Signed-off-by: Jakub Kicinski --- drivers/net/usb/qmi_wwan.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/net/usb/qmi_wwan.c b/drivers/net/usb/qmi_wwan.c index f51cf9cb9421..0e2ab567dd40 100644 --- a/drivers/net/usb/qmi_wwan.c +++ b/drivers/net/usb/qmi_wwan.c @@ -1086,6 +1086,7 @@ static const struct usb_device_id products[] = { {QMI_MATCH_FF_FF_FF(0x2c7c, 0x0125)}, /* Quectel EC25, EC20 R2.0 Mini PCIe */ {QMI_MATCH_FF_FF_FF(0x2c7c, 0x013d)}, /* Quectel RG660QB */ {QMI_MATCH_FF_FF_FF(0x2c7c, 0x0306)}, /* Quectel EP06/EG06/EM06 */ + {QMI_MATCH_FF_FF_FF(0x2c7c, 0x030b)}, /* Quectel EM060K/EG120K-EA */ {QMI_MATCH_FF_FF_FF(0x2c7c, 0x0512)}, /* Quectel EG12/EM12 */ {QMI_MATCH_FF_FF_FF(0x2c7c, 0x0620)}, /* Quectel EM160R-GL */ {QMI_MATCH_FF_FF_FF(0x2c7c, 0x0800)}, /* Quectel RM500Q-GL */ From 481506a756dcd828ef42391cb08f38d8d96d38fc Mon Sep 17 00:00:00 2001 From: Sanghyun Park Date: Fri, 18 Sep 2026 12:26:58 +0900 Subject: [PATCH 499/694] vxlan: use one headroom snapshot for neighbour replies vxlan_na_create() samples LL_RESERVED_SPACE() to size the reply skb and then samples it again to reserve headroom. A concurrent vxlan_changelink() can update needed_headroom between the two reads, creating a TOCTOU race. The second value can exceed the allocation and make the Ethernet header write out of bounds. The race is reproducible on the unpatched kernel. It occurred when vxlan_na_create() generated a neighbour reply while vxlan_changelink() changed the link headroom. KASAN caught a four-byte write two bytes beyond a 704-byte skbuff_small_head allocation. Snapshot the headroom once and use that value for both allocation and reservation. Fixes: 4b29dba9c085 ("vxlan: fix nonfunctional neigh_reduce()") Signed-off-by: Sanghyun Park Link: https://patch.msgid.link/20260918032842.502409-2-sanghyun.park.cnu@gmail.com Signed-off-by: Jakub Kicinski --- drivers/net/vxlan/vxlan_core.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/net/vxlan/vxlan_core.c b/drivers/net/vxlan/vxlan_core.c index c1d54339fa2b..3390e341d1e5 100644 --- a/drivers/net/vxlan/vxlan_core.c +++ b/drivers/net/vxlan/vxlan_core.c @@ -1958,13 +1958,15 @@ static struct sk_buff *vxlan_na_create(struct sk_buff *request, struct ipv6hdr *pip6; u8 *daddr; int na_olen = 8; /* opt hdr + ETH_ALEN for target */ + int headroom; int ns_olen; int i, len; if (dev == NULL || !pskb_may_pull(request, request->len)) return NULL; - len = LL_RESERVED_SPACE(dev) + sizeof(struct ipv6hdr) + + headroom = LL_RESERVED_SPACE(dev); + len = headroom + sizeof(struct ipv6hdr) + sizeof(*na) + na_olen + dev->needed_tailroom; reply = alloc_skb(len, GFP_ATOMIC); if (reply == NULL) @@ -1972,7 +1974,7 @@ static struct sk_buff *vxlan_na_create(struct sk_buff *request, reply->protocol = htons(ETH_P_IPV6); reply->dev = dev; - skb_reserve(reply, LL_RESERVED_SPACE(request->dev)); + skb_reserve(reply, headroom); skb_push(reply, sizeof(struct ethhdr)); skb_reset_mac_header(reply); From 4da3b7b8b50f3e2fde54a4c18a82a8e3f6223910 Mon Sep 17 00:00:00 2001 From: Norbert Szetei Date: Mon, 21 Sep 2026 17:03:57 +0200 Subject: [PATCH 500/694] net: xps: reject an out of range traffic class Only the entries below dev->num_tc are valid in dev->tc_to_txq[], and dev->prio_tc_map[] may only name classes below it. netdev_set_num_tc() lowers dev->num_tc without touching either array. netdev_txq_to_tc() walks all TC_MAX_QUEUE slots and netdev_get_prio_tc_map() returns the entry as it stands, so a leftover entry is handed out as a traffic class >= dev->num_tc. Taking that class from netdev_txq_to_tc(), __netif_set_xps_queue() rejects only a negative one and indexes an XPS map sized for dev->num_tc classes: tci = j * num_tc + tc; RCU_INIT_POINTER(new_dev_maps->attr_map[tci], map); attr_map[] holds nr_ids * num_tc entries and j runs over the ids named in the mask, so a class that is not below num_tc pushes tci past the end of the map for the last ids and the store overruns it. Any caller that lowers num_tc leaves such entries behind, and mqprio_destroy() tears down with netdev_set_num_tc(dev, 0) rather than netdev_reset_tc(). After mqprio with 8 classes then 1, tc_to_txq[1..7] still describe txq 1..7. The splat is from an XPS write to txq 2 on a veth with 8 rx queues: attr_map[] has 8 * 1 entries, tci = j + 2, and j == 6 stores one past the end of the 88-byte map: BUG: KASAN: slab-out-of-bounds in __netif_set_xps_queue (net/core/dev.c:2954) Write of size 8 at addr ffff88813016bc58 by task xps_oob/634 __netif_set_xps_queue (net/core/dev.c:2954) xps_rxqs_store (net/core/net-sysfs.c:1880) netdev_queue_attr_store (net/core/net-sysfs.c:1390) Allocated by task 634: __kmalloc_noprof (mm/slub.c:5439) __netif_set_xps_queue (net/core/dev.c:2937) The buggy address is located 0 bytes to the right of allocated 88-byte region [ffff88813016bc00, ffff88813016bc58) Reject a class the map has no room for. Fixes: 184c449f91fe ("net: Add support for XPS with QoS via traffic classes") Signed-off-by: Norbert Szetei Link: https://patch.msgid.link/162DD16F-54C6-444A-9E09-0B8CB3D591F2@doyensec.com Signed-off-by: Jakub Kicinski --- net/core/dev.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/core/dev.c b/net/core/dev.c index c67900354fa6..0292a16e16c2 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -2901,7 +2901,7 @@ int __netif_set_xps_queue(struct net_device *dev, const unsigned long *mask, dev = netdev_get_tx_queue(dev, index)->sb_dev ? : dev; tc = netdev_txq_to_tc(dev, index); - if (tc < 0) + if (tc < 0 || tc >= num_tc) return -EINVAL; } From a0bb6fac53fa7cf1cadb487b43d4c9276a6b82e3 Mon Sep 17 00:00:00 2001 From: Zhan Xusheng Date: Fri, 18 Sep 2026 21:29:15 +0800 Subject: [PATCH 501/694] sched/core: Account PSI IRQ time to the execution context, not the scheduling context psi_account_irqtime() has two callers which share rq->psi_irq_time, and they disagree about the context: __schedule() passes the outgoing rq->curr, sched_tick() passes rq->donor. Under proxy execution the donor is blocked on a mutex while rq->curr burns the CPU. The tick charges PSI_IRQ_FULL to the donor's cgroup and advances the timestamp, so the call from __schedule() then finds delta <= 0 and charges nothing. The delta is not counted twice, it lands on the wrong cgroup. Pass rq->curr, which is what the call read before commit af0c8b2bf67b ("sched: Split scheduler and execution contexts") renamed 'curr' to 'donor' across sched_tick(). Without CONFIG_SCHED_PROXY_EXEC the two rq members are a union, so this only changes anything where that option is set, and it depends on EXPERT. Fixes: af0c8b2bf67b ("sched: Split scheduler and execution contexts") Signed-off-by: Zhan Xusheng Signed-off-by: Peter Zijlstra (Intel) Signed-off-by: Ingo Molnar Link: https://patch.msgid.link/20260918132915.1236312-1-zhanxusheng@xiaomi.com --- kernel/sched/core.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/kernel/sched/core.c b/kernel/sched/core.c index 0b846a13c628..1fe40de6ebe3 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -5798,7 +5798,7 @@ void sched_tick(void) curr = rq->curr; donor = rq->donor; - psi_account_irqtime(rq, donor, NULL); + psi_account_irqtime(rq, curr, NULL); update_rq_clock(rq); hw_pressure = arch_scale_hw_pressure(cpu_of(rq)); From e47a1958e12abc3a17b5231a4f21c8f1bf662e08 Mon Sep 17 00:00:00 2001 From: Yuqi Xu Date: Sat, 19 Sep 2026 16:45:27 +0800 Subject: [PATCH 502/694] net: ipconfig: bound DHCP option construction ic_dhcp_init_options() appends the hostname (option 12), vendor-class (option 60) and client-ID (option 61) options into the fixed 312-byte bootp_pkt.exten[] buffer. Only the client-ID branch checked the remaining space; the hostname and vendor-class writes were unbounded. A 64-byte hostname together with the maximum 252-byte dhcpclass= identifier needs 18 + (2 + 64) + (2 + 252) = 338 of the 312 available bytes even before the terminating END marker, so the vendor-class memcpy runs past the end of exten[]. With CONFIG_FORTIFY_SOURCE this is reported as a field-spanning write and, when the kernel is booted with panic_on_warn=1, aborts boot with a panic. Route the optional options through a common helper that makes sure the option, its 2-byte header and the END marker all fit and drops an option that would not. Configurations with short options keep sending exactly the same bytes as before. Fixes: 130c0f47fdf9 ("ipconfig: send host-name in DHCP requests") Cc: stable@vger.kernel.org Reported-by: Vega Assisted-by: LLM Signed-off-by: Yuqi Xu Reviewed-by: Ren Wei Reviewed-by: Simon Horman Link: https://patch.msgid.link/7808dfbfa2162dfd0b19f59aff5742d6e0db2abb.1789798023.git.xuyuqiabc@gmail.com Signed-off-by: Paolo Abeni --- net/ipv4/ipconfig.c | 45 ++++++++++++++++++++++++++------------------- 1 file changed, 26 insertions(+), 19 deletions(-) diff --git a/net/ipv4/ipconfig.c b/net/ipv4/ipconfig.c index 155db067eaec..1b8585404a41 100644 --- a/net/ipv4/ipconfig.c +++ b/net/ipv4/ipconfig.c @@ -676,6 +676,24 @@ static const u8 ic_bootp_cookie[4] = { 99, 130, 83, 99 }; #ifdef IPCONFIG_DHCP +static bool __init +ic_dhcp_add_option(u8 **options, const u8 *end, u8 type, const void *value, + int len) +{ + u8 *e = *options; + + /* leave room for the option header and the END marker */ + if (len > U8_MAX || end - e < len + 3) + return false; + + *e++ = type; + *e++ = len; + memcpy(e, value, len); + *options = e + len; + + return true; +} + static void __init ic_dhcp_init_options(u8 *options, struct ic_device *d) { @@ -691,6 +709,7 @@ ic_dhcp_init_options(u8 *options, struct ic_device *d) 42, /* NTP servers */ }; u8 mt = (ic_servaddr == NONE) ? DHCPDISCOVER : DHCPREQUEST; + u8 *end = options + sizeof(((struct bootp_pkt *)0)->exten); u8 *e = options; int len; @@ -721,31 +740,19 @@ ic_dhcp_init_options(u8 *options, struct ic_device *d) e += sizeof(ic_req_params); if (ic_host_name_set) { - *e++ = 12; /* host-name */ len = strlen(utsname()->nodename); - *e++ = len; - memcpy(e, utsname()->nodename, len); - e += len; + ic_dhcp_add_option(&e, end, 12, utsname()->nodename, len); } if (*vendor_class_identifier) { - pr_info("DHCP: sending class identifier \"%s\"\n", - vendor_class_identifier); - *e++ = 60; /* Class-identifier */ len = strlen(vendor_class_identifier); - *e++ = len; - memcpy(e, vendor_class_identifier, len); - e += len; + if (ic_dhcp_add_option(&e, end, 60, vendor_class_identifier, len)) + pr_info("DHCP: sending class identifier \"%s\"\n", + vendor_class_identifier); } len = strlen(dhcp_client_identifier + 1); - /* the minimum length of identifier is 2, include 1 byte type, - * and can not be larger than the length of options - */ - if (len >= 1 && len < 312 - (e - options) - 1) { - *e++ = 61; - *e++ = len + 1; - memcpy(e, dhcp_client_identifier, len + 1); - e += len + 1; - } + /* the minimum length of identifier is 2, include 1 byte type */ + if (len >= 1) + ic_dhcp_add_option(&e, end, 61, dhcp_client_identifier, len + 1); *e++ = 255; /* End of the list */ } From 7b824c293a6b56de8285a97984c507cba56bc4c4 Mon Sep 17 00:00:00 2001 From: Brajesh Gupta Date: Tue, 22 Sep 2026 09:56:55 +0530 Subject: [PATCH 503/694] drm/imagination: Propagate map failures correctly from pvr_mmu_map_sgl() Map failure from pvr_mmu_map_sgl() interface was not returned correctly to pvr_mmu_map() interface. This resulted in pvr_mmu_map() interface to continue instead of returning an error to caller. Fix it by returning a proper error code from pvr_mmu_map_sgl() interface. Call stack for crash: [ 1179.286237] Unable to handle kernel NULL pointer dereference at virtual address 0000000000000008 [ 1179.295067] Mem abort info: [ 1179.297877] ESR = 0x0000000096000004 [ 1179.301656] EC = 0x25: DABT (current EL), IL = 32 bits [ 1179.306987] SET = 0, FnV = 0 [ 1179.310048] EA = 0, S1PTW = 0 [ 1179.313198] FSC = 0x04: level 0 translation fault [ 1179.318096] Data abort info: [ 1179.320993] ISV = 0, ISS = 0x00000004, ISS2 = 0x00000000 [ 1179.326483] CM = 0, WnR = 0, TnD = 0, TagAccess = 0 [ 1179.331546] GCS = 0, Overlay = 0, DirtyBit = 0, Xs = 0 [ 1179.336895] user pgtable: 4k pages, 48-bit VAs, pgdp=000000009822a000 [ 1179.343402] [0000000000000008] pgd=0000000000000000, p4d=0000000000000000 [ 1179.350243] Internal error: Oops: 0000000096000004 [#2] SMP [ 1179.355908] Modules linked in: powervr gpu_sched drm_shmem_helper drm_gpuvm drm_exec xhci_plat_hcd xhci_hcd dwc3 usbcore usb_common snd_soc_simple_card snd_soc_simple_card_utils dwc3_am62 at24 sa2ul sha512 libsha512 sha256 authenc sch_fq_codel fuse dm_mod ipv6 [ 1179.378992] CPU: 1 UID: 1000 PID: 680 Comm: deqp-vk Tainted: G D 6.17.0 #1 PREEMPT [ 1179.388120] Tainted: [D]=DIE [ 1179.390994] Hardware name: Texas Instruments AM625 SK (DT) [ 1179.396467] pstate: 00000005 (nzcv daif -PAN -UAO -TCO -DIT -SSBS BTYPE=--) [ 1179.403415] pc : pvr_mmu_op_context_unmap_curr_page+0x6c/0x134 [powervr] [ 1179.410140] lr : pvr_mmu_op_context_unmap_curr_page+0x58/0x134 [powervr] [ 1179.416848] sp : ffff8000839ab8c0 [ 1179.420153] x29: ffff8000839ab8c0 x28: 0000000000000001 x27: 000000008f386000 [ 1179.427283] x26: ffff000016d1df98 x25: 0000000000247000 x24: 00000000000001e6 [ 1179.434413] x23: 0000000000000002 x22: 000000000000ffff x21: 0000000000000247 [ 1179.441540] x20: 0000000000000245 x19: ffff000016d1df60 x18: 0000000000000002 [ 1179.448668] x17: 0000000000000000 x16: 0000000000000000 x15: 0000000000000001 [ 1179.455793] x14: 0000000000060810 x13: ffff80007fffffff x12: ffff000004190480 [ 1179.462921] x11: ffff8000853f7000 x10: ffff8000811ae000 x9 : ffff0000041900b8 [ 1179.470051] x8 : 0000000000000000 x7 : 00000000990c4001 x6 : 0000000000000007 [ 1179.477177] x5 : ffff000016d1df60 x4 : 0000000000000000 x3 : ffff00000a7d8000 [ 1179.484306] x2 : 00000000000001ff x1 : 0000000000000000 x0 : 0000000000000000 [ 1179.491433] Call trace: [ 1179.493872] pvr_mmu_op_context_unmap_curr_page+0x6c/0x134 [powervr] (P) [ 1179.500582] pvr_mmu_map+0x31c/0x388 [powervr] [ 1179.505027] pvr_vm_gpuva_map+0x40/0x88 [powervr] [ 1179.509732] __drm_gpuvm_sm_map+0x250/0x44c [drm_gpuvm] [ 1179.514952] drm_gpuvm_sm_map+0x48/0x5c [drm_gpuvm] [ 1179.519822] pvr_vm_bind_op_exec+0x64/0x70 [powervr] [ 1179.524785] pvr_vm_map+0x1f8/0x2a8 [powervr] [ 1179.529142] pvr_ioctl_vm_map+0x12c/0x188 [powervr] [ 1179.534018] drm_ioctl_kernel+0xb8/0x128 [ 1179.537941] drm_ioctl+0x21c/0x4ec [ 1179.541337] __arm64_sys_ioctl+0xac/0x108 [ 1179.545344] invoke_syscall+0x44/0x100 [ 1179.549091] el0_svc_common.constprop.0+0x40/0xe0 [ 1179.553790] do_el0_svc+0x1c/0x28 [ 1179.557106] el0_svc+0x34/0xf0 [ 1179.560159] el0t_64_sync_handler+0xd0/0xe4 [ 1179.564334] el0t_64_sync+0x198/0x19c [ 1179.567996] Code: 54000300 35000360 f9402261 79409a62 (f9400421) [ 1179.574081] ---[ end trace 0000000000000000 ]--- Fixes: ff5f643de0bf ("drm/imagination: Add GEM and VM related code") Reviewed-by: Alexandru Dadu Reviewed-by: Alessio Belle Cc: stable@vger.kernel.org Link: https://patch.msgid.link/20260922-mmu_fix-v4-1-12f1a871456a@imgtec.com Signed-off-by: Brajesh Gupta --- drivers/gpu/drm/imagination/pvr_mmu.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/imagination/pvr_mmu.c b/drivers/gpu/drm/imagination/pvr_mmu.c index 3cac482e1034..23261d9ad3fd 100644 --- a/drivers/gpu/drm/imagination/pvr_mmu.c +++ b/drivers/gpu/drm/imagination/pvr_mmu.c @@ -12,6 +12,7 @@ #include "pvr_rogue_mmu_defs.h" #include +#include #include #include #include @@ -2553,7 +2554,9 @@ pvr_mmu_map_sgl(struct pvr_mmu_op_context *op_ctx, struct scatterlist *sgl, err_destroy_pages: memcpy(&op_ctx->curr_page, &ptr_copy, sizeof(op_ctx->curr_page)); - err = pvr_mmu_op_context_unmap_curr_page(op_ctx, page); + if (pvr_mmu_op_context_unmap_curr_page(op_ctx, page)) + drm_err(from_pvr_device(op_ctx->mmu_ctx->pvr_dev), + "%s : Failure in unmapping pages\n", __func__); return err; } From 0a8224058a5835297dcf4a46bbcd16f77a9fe424 Mon Sep 17 00:00:00 2001 From: Brajesh Gupta Date: Tue, 22 Sep 2026 09:56:56 +0530 Subject: [PATCH 504/694] drm/imagination: Fix page count for page table for map() interface The GPU virtual start address wasn't included in the calculation for the amount of page tables required for mapping a BO object in map() interface. It resulted in map failure later due to not enough pages at L0/L1 level. Update pvr_mmu_op_context_create() interface to pass device address as well to allow correct calculation for page table memory. If L0 tables cover 2MB (0x200000), the range defined by device address 0x80001ff000 (general heap at 2MB - 4KB) and size 0x2000 (two 4KB pages) requires two L0 pages to be mapped, but without the base address a range of 0x2000 computes to a single L0 page which is not enough. Fixes: ff5f643de0bf ("drm/imagination: Add GEM and VM related code") Reviewed-by: Alexandru Dadu Reviewed-by: Alessio Belle Cc: stable@vger.kernel.org Link: https://patch.msgid.link/20260922-mmu_fix-v4-2-12f1a871456a@imgtec.com Signed-off-by: Brajesh Gupta --- drivers/gpu/drm/imagination/pvr_mmu.c | 14 ++++++++------ drivers/gpu/drm/imagination/pvr_mmu.h | 2 +- drivers/gpu/drm/imagination/pvr_vm.c | 4 ++-- 3 files changed, 11 insertions(+), 9 deletions(-) diff --git a/drivers/gpu/drm/imagination/pvr_mmu.c b/drivers/gpu/drm/imagination/pvr_mmu.c index 23261d9ad3fd..62eae7fcd5a2 100644 --- a/drivers/gpu/drm/imagination/pvr_mmu.c +++ b/drivers/gpu/drm/imagination/pvr_mmu.c @@ -2336,6 +2336,7 @@ void pvr_mmu_op_context_destroy(struct pvr_mmu_op_context *op_ctx) * pvr_mmu_op_context_create() - Create an MMU op context. * @ctx: MMU context associated with owning VM context. * @sgt: Scatter gather table containing pages pinned for use by this context. + * @device_addr: Virtual device address at the start of the requested mapping. * @sgt_offset: Start offset of the requested device-virtual memory mapping. * @size: Size in bytes of the requested device-virtual memory mapping. For an * unmapping, this should be zero so that no page tables are allocated. @@ -2347,8 +2348,9 @@ void pvr_mmu_op_context_destroy(struct pvr_mmu_op_context *op_ctx) */ struct pvr_mmu_op_context * pvr_mmu_op_context_create(struct pvr_mmu_context *ctx, struct sg_table *sgt, - u64 sgt_offset, u64 size) + u64 device_addr, u64 sgt_offset, u64 size) { + u64 start_addr = device_addr + sgt_offset; int err; struct pvr_mmu_op_context *op_ctx = kzalloc_obj(*op_ctx); @@ -2364,16 +2366,16 @@ pvr_mmu_op_context_create(struct pvr_mmu_context *ctx, struct sg_table *sgt, if (size) { /* * The number of page table objects we need to prealloc is - * indicated by the mapping size, start offset and the sizes + * indicated by the mapping size, start address and the sizes * of the areas mapped per PT or PD. The range calculation is * identical to that for the index into a table for a device * address, so we reuse those functions here. */ - const u32 l1_start_idx = pvr_page_table_l2_idx(sgt_offset); - const u32 l1_end_idx = pvr_page_table_l2_idx(sgt_offset + size); + const u32 l1_start_idx = pvr_page_table_l2_idx(start_addr); + const u32 l1_end_idx = pvr_page_table_l2_idx(start_addr + size); const u32 l1_count = l1_end_idx - l1_start_idx + 1; - const u32 l0_start_idx = pvr_page_table_l1_idx(sgt_offset); - const u32 l0_end_idx = pvr_page_table_l1_idx(sgt_offset + size); + const u32 l0_start_idx = pvr_page_table_l1_idx(start_addr); + const u32 l0_end_idx = pvr_page_table_l1_idx(start_addr + size); const u32 l0_count = l0_end_idx - l0_start_idx + 1; /* diff --git a/drivers/gpu/drm/imagination/pvr_mmu.h b/drivers/gpu/drm/imagination/pvr_mmu.h index a8ecd460168d..2c02d61ba0a2 100644 --- a/drivers/gpu/drm/imagination/pvr_mmu.h +++ b/drivers/gpu/drm/imagination/pvr_mmu.h @@ -99,7 +99,7 @@ dma_addr_t pvr_mmu_get_root_table_dma_addr(struct pvr_mmu_context *ctx); void pvr_mmu_op_context_destroy(struct pvr_mmu_op_context *op_ctx); struct pvr_mmu_op_context * pvr_mmu_op_context_create(struct pvr_mmu_context *ctx, - struct sg_table *sgt, u64 sgt_offset, u64 size); + struct sg_table *sgt, u64 device_addr, u64 sgt_offset, u64 size); int pvr_mmu_map(struct pvr_mmu_op_context *op_ctx, u64 size, u64 flags, u64 device_addr); diff --git a/drivers/gpu/drm/imagination/pvr_vm.c b/drivers/gpu/drm/imagination/pvr_vm.c index ceb78694cd98..55cc999f3708 100644 --- a/drivers/gpu/drm/imagination/pvr_vm.c +++ b/drivers/gpu/drm/imagination/pvr_vm.c @@ -276,7 +276,7 @@ pvr_vm_bind_op_map_init(struct pvr_vm_bind_op *bind_op, goto err_bind_op_fini; bind_op->mmu_op_ctx = - pvr_mmu_op_context_create(vm_ctx->mmu_ctx, sgt, offset, size); + pvr_mmu_op_context_create(vm_ctx->mmu_ctx, sgt, device_addr, offset, size); err = PTR_ERR_OR_ZERO(bind_op->mmu_op_ctx); if (err) { bind_op->mmu_op_ctx = NULL; @@ -318,7 +318,7 @@ pvr_vm_bind_op_unmap_init(struct pvr_vm_bind_op *bind_op, } bind_op->mmu_op_ctx = - pvr_mmu_op_context_create(vm_ctx->mmu_ctx, NULL, 0, 0); + pvr_mmu_op_context_create(vm_ctx->mmu_ctx, NULL, device_addr, 0, 0); err = PTR_ERR_OR_ZERO(bind_op->mmu_op_ctx); if (err) { bind_op->mmu_op_ctx = NULL; From 45585c3aa285854face65293acc95eff73063d6d Mon Sep 17 00:00:00 2001 From: Pengpeng Hou Date: Sun, 20 Sep 2026 11:43:29 +0800 Subject: [PATCH 505/694] drm/imagination: clamp freelist reconstruction requests The firmware reconstruction count controls accesses to the request's fixed freelist ID array and the copy into the fixed response array. Neither access currently bounds the count to those protocol arrays. Clamp the count to the request capacity, which is shared by the response layout, and use that count consistently for reconstruction and response publication. Keep the firmware recovery exchange instead of dropping an oversized request without a response, as discussed with the firmware maintainer. The issue was found by our static-analysis tool. Fixes: 6eedddab733b ("drm/imagination: Implement free list and HWRT create and destroy ioctls") Assisted-by: gpt 5 Signed-off-by: Pengpeng Hou Reviewed-by: Alessio Belle Link: https://patch.msgid.link/20260920034329.16614-1-hppiscas@163.com Signed-off-by: Brajesh Gupta --- drivers/gpu/drm/imagination/pvr_free_list.c | 15 ++++++++++++--- 1 file changed, 12 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/imagination/pvr_free_list.c b/drivers/gpu/drm/imagination/pvr_free_list.c index e85cac83834c..faf5e586d8dc 100644 --- a/drivers/gpu/drm/imagination/pvr_free_list.c +++ b/drivers/gpu/drm/imagination/pvr_free_list.c @@ -8,6 +8,7 @@ #include "pvr_vm.h" #include +#include #include #include #include @@ -612,13 +613,21 @@ pvr_free_list_process_reconstruct_req(struct pvr_device *pvr_dev, }; struct rogue_fwif_freelists_reconstruction_data *resp = &resp_cmd.cmd_data.free_lists_reconstruction_data; + u32 count = min_t(u32, req->freelist_count, + ARRAY_SIZE(req->freelist_ids)); - for (u32 i = 0; i < req->freelist_count; i++) + if (count != req->freelist_count) { + drm_warn_once(from_pvr_device(pvr_dev), + "Requested reconstruction of %u freelists, limiting to %u\n", + req->freelist_count, count); + } + + for (u32 i = 0; i < count; i++) pvr_free_list_reconstruct(pvr_dev, req->freelist_ids[i]); - resp->freelist_count = req->freelist_count; + resp->freelist_count = count; memcpy(resp->freelist_ids, req->freelist_ids, - req->freelist_count * sizeof(resp->freelist_ids[0])); + count * sizeof(resp->freelist_ids[0])); WARN_ON(pvr_kccb_send_cmd(pvr_dev, &resp_cmd, NULL)); } From 00efbbd40bd5fd92c67b7cf1aab8904fa59a96f6 Mon Sep 17 00:00:00 2001 From: David Dai Date: Fri, 18 Sep 2026 16:11:55 -0500 Subject: [PATCH 506/694] bonding: crypto offload enabled, non-offload slave failover, rekey failed Create a bonding device (i.e. bond0) in active-backup mode, 2 slaves. Active slave: offload capable interface (i.e. eth1), primary interface. Backup slave: non-offload capable interface(i.e. eth2). Configure strongswan service swantl.conf child SA "hw_offload = crypto" Start strongswan service IPSec Crytpo Offload is enabled on top of bond0. i.e. ip xfrm state |grep offload crypto offload parameters: dev bond0 dir out mode crypto crypto offload parameters: dev bond0 dir in mode crypto Active slave eth1 takes adavantage of IPSec Crypto Offload capability. If active slave eth1 is down for any reason (i.e. eth1 link down): ip link set down dev eth1 non-offload capable interface eth2 failover to becomes active slave. The existing SAs can continue use software IPsec after failover. Traffic still keeps going properly. However if eth1 link had not recovered yet, strongswan service does new child SA rekey, or uses swanctl command to do new child SA rekey, it will fail because active slave eth2 doesn't support crypto offload. In bond_ipsec_add_sa routine, it returns -EINVAL now, which is treated as fatal error by xfrm_dev_state_add routine in kernel xfrm. To make the non-offload active slave survive the child SA rekey, need to make bond_ipsec_add_sa routine returns -EOPNOTSUPP instead when active slave doesn't support IPsec Crypto offload, the xfrm will gracefully fallback to create new SA using Software IPsec. Network traffic can keep going. After offload capable interface eth1 link is up, becomes active slave, next time strongswan child SA rekey will create a new SA which enables crypto offload again. Fixes: 18cb261afd7b ("bonding: support hardware encryption offload to slaves") Signed-off-by: David Dai Reviewed-by: Hangbin Liu Link: https://patch.msgid.link/20260918211155.1664493-1-zdai@linux.ibm.com Signed-off-by: Paolo Abeni --- drivers/net/bonding/bond_main.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c index a9bff7663eec..de2489c3d9bf 100644 --- a/drivers/net/bonding/bond_main.c +++ b/drivers/net/bonding/bond_main.c @@ -490,7 +490,7 @@ static int bond_ipsec_add_sa(struct net_device *bond_dev, !real_dev->xfrmdev_ops->xdo_dev_state_add || netif_is_bond_master(real_dev)) { NL_SET_ERR_MSG_MOD(extack, "Slave does not support ipsec offload"); - err = -EINVAL; + err = -EOPNOTSUPP; goto out; } From 36c2009d90f2210ef92e6f4f2850e8b57b09e754 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Gajdos=20Tam=C3=A1s?= Date: Mon, 21 Sep 2026 11:13:32 +0200 Subject: [PATCH 507/694] net: atl1c: fix soft lockup on out-of-range tpd_cons read MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The hardware can report an out-of-range tpd_cons (seen as 0xffff) while the PCIe link/MAC is resetting. An out-of-range value can never be reached and the loop below would spin forever. To avoid a soft lockup treat it as "nothing new to clean" instead. Reproduced on two machines, same NIC (Qualcomm Atheros AR8151 v2.0, 4-port), triggered by rebooting a Mikrotik CCR2004 PCIe card that the ports are directly linked to: - Ubuntu 26.04.1 LTS, kernel 7.0.0-31-generic. The link-flap precursor, before the lockup was captured with a full trace elsewhere: atl1c 0000:05:00.0 enp5s0f0: NETDEV WATCHDOG: CPU: 4: transmit queue 2 timed out 489984 ms atl1c 0000:05:00.0: MAC state machine can't be idle since disabled for 10ms second atl1c 0000:05:00.0: atl1c: enp5s0f0 NIC Link is Up<65535 Mbps Full Duplex> 65535 (0xffff) here is the same value tpd_cons reads back once the loop below gets stuck. - Proxmox VE, kernel 7.0.14-11-pve. Same NIC/trigger, this time caught by the soft lockup watchdog with a full stack trace: watchdog: BUG: soft lockup - CPU#12 stuck for 354s! [napi/eth%d-0:329] CPU: 12 UID: 0 PID: 329 Comm: napi/eth%d-0 Tainted: P O L 7.0.14-11-pve #1 PREEMPT(lazy) RIP: 0010:atl1c_clean_tx+0x142/0x2d0 [atl1c] Call Trace: __napi_poll+0x32/0x1e0 napi_threaded_poll_loop+0x286/0x2e0 napi_threaded_poll+0xfd/0x140 kthread+0xf7/0x130 ret_from_fork+0x2da/0x3a0 ret_from_fork_asm+0x1a/0x30 Fixes: 43250ddd75a35d ("atl1c: Atheros L1C Gigabit Ethernet driver") Cc: stable@vger.kernel.org Signed-off-by: Gajdos Tamás Link: https://patch.msgid.link/20260921091334.3571525-2-tamas@rimpianto.com Signed-off-by: Paolo Abeni --- drivers/net/ethernet/atheros/atl1c/atl1c_main.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/net/ethernet/atheros/atl1c/atl1c_main.c b/drivers/net/ethernet/atheros/atl1c/atl1c_main.c index 7efa3fc257b3..e58f1d2c26bd 100644 --- a/drivers/net/ethernet/atheros/atl1c/atl1c_main.c +++ b/drivers/net/ethernet/atheros/atl1c/atl1c_main.c @@ -1602,6 +1602,9 @@ static int atl1c_clean_tx(struct napi_struct *napi, int budget) AT_READ_REGW(&adapter->hw, atl1c_qregs[tpd_ring->num].tpd_cons, &hw_next_to_clean); + if (unlikely(hw_next_to_clean >= tpd_ring->count)) + hw_next_to_clean = next_to_clean; + while (next_to_clean != hw_next_to_clean) { buffer_info = &tpd_ring->buffer_info[next_to_clean]; if (buffer_info->skb) { From 374bf9e4b90f979e052332c4faca2d745c491a12 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Gajdos=20Tam=C3=A1s?= Date: Mon, 21 Sep 2026 11:13:33 +0200 Subject: [PATCH 508/694] net: atl1e: fix soft lockup on out-of-range hw_next_to_clean read MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Same issue as atl1c (see the first commit in this series, "net: atl1c: fix soft lockup on out-of-range tpd_cons read"): the hardware can report an out-of-range hw_next_to_clean (seen as 0xffff) while the PCIe link/MAC is resetting. An out-of-range value can never be reached and the loop below would spin forever. Treat it as "nothing new to clean" instead. Fixes: a6a5325239c202 ("atl1e: Atheros L1E Gigabit Ethernet driver") Cc: stable@vger.kernel.org Signed-off-by: Gajdos Tamás Link: https://patch.msgid.link/20260921091334.3571525-3-tamas@rimpianto.com Signed-off-by: Paolo Abeni --- drivers/net/ethernet/atheros/atl1e/atl1e_main.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/net/ethernet/atheros/atl1e/atl1e_main.c b/drivers/net/ethernet/atheros/atl1e/atl1e_main.c index 40290028580b..437989edb780 100644 --- a/drivers/net/ethernet/atheros/atl1e/atl1e_main.c +++ b/drivers/net/ethernet/atheros/atl1e/atl1e_main.c @@ -1234,6 +1234,9 @@ static bool atl1e_clean_tx_irq(struct atl1e_adapter *adapter) u16 hw_next_to_clean = AT_READ_REGW(&adapter->hw, REG_TPD_CONS_IDX); u16 next_to_clean = atomic_read(&tx_ring->next_to_clean); + if (unlikely(hw_next_to_clean >= tx_ring->count)) + hw_next_to_clean = next_to_clean; + while (next_to_clean != hw_next_to_clean) { tx_buffer = &tx_ring->tx_buffer[next_to_clean]; if (tx_buffer->dma) { From 43e746821f5f5afbbf68e388bf9fbe221e03bfca Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Gajdos=20Tam=C3=A1s?= Date: Mon, 21 Sep 2026 11:13:34 +0200 Subject: [PATCH 509/694] net: atl1: fix soft lockup on out-of-range cmb_tpd_next_to_clean read MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Same issue as atl1c (see the first commit in this series, "net: atl1c: fix soft lockup on out-of-range tpd_cons read"): the hardware can report an out-of-range cmb_tpd_next_to_clean (seen as 0xffff) while the PCIe link/MAC is resetting. An out-of-range value can never be reached and the loop below would spin forever. Treat it as "nothing new to clean" instead. Fixes: f3cc28c797604f ("Add Attansic L1 ethernet driver.") Cc: stable@vger.kernel.org Signed-off-by: Gajdos Tamás Link: https://patch.msgid.link/20260921091334.3571525-4-tamas@rimpianto.com Signed-off-by: Paolo Abeni --- drivers/net/ethernet/atheros/atlx/atl1.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/net/ethernet/atheros/atlx/atl1.c b/drivers/net/ethernet/atheros/atlx/atl1.c index 98a4d089270e..957d5598dda5 100644 --- a/drivers/net/ethernet/atheros/atlx/atl1.c +++ b/drivers/net/ethernet/atheros/atlx/atl1.c @@ -2066,6 +2066,9 @@ static int atl1_intr_tx(struct atl1_adapter *adapter) sw_tpd_next_to_clean = atomic_read(&tpd_ring->next_to_clean); cmb_tpd_next_to_clean = le16_to_cpu(adapter->cmb.cmb->tpd_cons_idx); + if (unlikely(cmb_tpd_next_to_clean >= tpd_ring->count)) + cmb_tpd_next_to_clean = sw_tpd_next_to_clean; + while (cmb_tpd_next_to_clean != sw_tpd_next_to_clean) { buffer_info = &tpd_ring->buffer_info[sw_tpd_next_to_clean]; if (buffer_info->dma) { From 77b1718e39e5c9f6956fb60807326af20baf889d Mon Sep 17 00:00:00 2001 From: Myeonghun Pak Date: Mon, 21 Sep 2026 21:46:05 -0400 Subject: [PATCH 510/694] bna: prevent IOC timer rearm during teardown bna: prevent IOC timer rearm during teardown bnad_pci_remove() and the probe disable_ioceth path call timer_delete_sync() for ioc_timer, sem_timer and hb_timer, but not for iocpf_timer. bnad_iocpf_timeout() then takes bnad->bna_lock after free_netdev() has freed the struct bnad. Deleting iocpf_timer last does not fix this. sem_timer and iocpf_timer rearm each other: bnad_iocpf_sem_timeout() can arm iocpf_timer, and bnad_iocpf_timeout() arms sem_timer from bfa_ioc_hw_sem_get() when the semaphore is busy. timer_delete_sync() only waits out its own callback. bnad_ioceth_disable() can time out and leave that callback live. Shut all four IOC timers down with timer_shutdown_sync() on both paths, so a later mod_timer() is ignored. This issue was identified during our ongoing static-analysis research while reviewing kernel code. Fixes: 1d32f7696286 ("bna: IOC failure auto recovery fix") Cc: stable@vger.kernel.org Assisted-by: LLM Co-developed-by: Ijae Kim Signed-off-by: Ijae Kim Signed-off-by: Myeonghun Pak Link: https://patch.msgid.link/20260922014605.588040-1-mhun512@gmail.com Signed-off-by: Paolo Abeni --- drivers/net/ethernet/brocade/bna/bnad.c | 24 ++++++++++++++++++------ 1 file changed, 18 insertions(+), 6 deletions(-) diff --git a/drivers/net/ethernet/brocade/bna/bnad.c b/drivers/net/ethernet/brocade/bna/bnad.c index 8b75004ba7c9..55dfd4896785 100644 --- a/drivers/net/ethernet/brocade/bna/bnad.c +++ b/drivers/net/ethernet/brocade/bna/bnad.c @@ -2571,6 +2571,22 @@ bnad_ioceth_disable(struct bnad *bnad) return err; } +/* + * The IOC timers rearm one another, so deleting one cannot stop a + * sibling callback from arming it again. Shut them down so a later + * mod_timer() is ignored. + */ +static void +bnad_ioc_timers_shutdown(struct bnad *bnad) +{ + struct bfa_ioc *ioc = &bnad->bna.ioceth.ioc; + + timer_shutdown_sync(&ioc->ioc_timer); + timer_shutdown_sync(&ioc->sem_timer); + timer_shutdown_sync(&ioc->hb_timer); + timer_shutdown_sync(&ioc->iocpf_timer); +} + static int bnad_ioceth_enable(struct bnad *bnad) { @@ -3727,9 +3743,7 @@ bnad_pci_probe(struct pci_dev *pdev, bnad_res_free(bnad, &bnad->mod_res_info[0], BNA_MOD_RES_T_MAX); disable_ioceth: bnad_ioceth_disable(bnad); - timer_delete_sync(&bnad->bna.ioceth.ioc.ioc_timer); - timer_delete_sync(&bnad->bna.ioceth.ioc.sem_timer); - timer_delete_sync(&bnad->bna.ioceth.ioc.hb_timer); + bnad_ioc_timers_shutdown(bnad); spin_lock_irqsave(&bnad->bna_lock, flags); bna_uninit(bna); spin_unlock_irqrestore(&bnad->bna_lock, flags); @@ -3770,9 +3784,7 @@ bnad_pci_remove(struct pci_dev *pdev) mutex_lock(&bnad->conf_mutex); bnad_ioceth_disable(bnad); - timer_delete_sync(&bnad->bna.ioceth.ioc.ioc_timer); - timer_delete_sync(&bnad->bna.ioceth.ioc.sem_timer); - timer_delete_sync(&bnad->bna.ioceth.ioc.hb_timer); + bnad_ioc_timers_shutdown(bnad); spin_lock_irqsave(&bnad->bna_lock, flags); bna_uninit(bna); spin_unlock_irqrestore(&bnad->bna_lock, flags); From 58eb1b3325edac42dc6df72c80962bd53a3c8ca7 Mon Sep 17 00:00:00 2001 From: Dongliang Qin Date: Tue, 22 Sep 2026 11:15:45 +0800 Subject: [PATCH 511/694] rds: ib: Clear the sg list when mapping an MR fails rds_ib_map_frmr() stores the caller's scatterlist in the MR before DMA mapping and registration can fail. On failure, __rds_rdma_map() unpins the pages and frees the scatterlist, but rds_ib_free_frmr() can still return the MR to the pool with the stale pointer set. This leaves the pool with a dangling scatterlist and can lead to local privilege escalation. KASAN detects the resulting use-after-free when the MR is later torn down: BUG: KASAN: slab-use-after-free in __rds_ib_teardown_mr Read of size 8 Call Trace: __rds_ib_teardown_mr rds_ib_unreg_frmr rds_ib_flush_mr_pool rds_ib_flush_mrs rds_free_mr rds_setsockopt Store the scatterlist in the MR only after DMA mapping succeeds. If DMA mapping fails, return directly while the MR fields remain clear; the caller keeps ownership of the scatterlist and its pinned pages. If a later registration step fails, unmap the scatterlist and clear the MR fields before returning. Fixes: 1659185fb4d0 ("RDS: IB: Support Fastreg MR (FRMR) memory registration mode") Cc: stable@vger.kernel.org Signed-off-by: Dongliang Qin Reviewed-by: Allison Henderson Link: https://patch.msgid.link/20260922031546.3874605-1-cccccccccccc777777@gmail.com Signed-off-by: Paolo Abeni --- net/rds/ib_frmr.c | 11 +++++------ 1 file changed, 5 insertions(+), 6 deletions(-) diff --git a/net/rds/ib_frmr.c b/net/rds/ib_frmr.c index bd861191157b..8397aa4a17ac 100644 --- a/net/rds/ib_frmr.c +++ b/net/rds/ib_frmr.c @@ -204,19 +204,16 @@ static int rds_ib_map_frmr(struct rds_ib_device *rds_ibdev, */ rds_ib_teardown_mr(ibmr); - ibmr->sg = sg; - ibmr->sg_len = sg_len; - ibmr->sg_dma_len = 0; frmr->sg_byte_len = 0; - WARN_ON(ibmr->sg_dma_len); - ibmr->sg_dma_len = ib_dma_map_sg(dev, ibmr->sg, ibmr->sg_len, + ibmr->sg_dma_len = ib_dma_map_sg(dev, sg, sg_len, DMA_BIDIRECTIONAL); if (unlikely(!ibmr->sg_dma_len)) { pr_warn("RDS/IB: %s failed!\n", __func__); return -EBUSY; } - frmr->sg_byte_len = 0; + ibmr->sg = sg; + ibmr->sg_len = sg_len; frmr->dma_npages = 0; len = 0; @@ -264,6 +261,8 @@ static int rds_ib_map_frmr(struct rds_ib_device *rds_ibdev, ib_dma_unmap_sg(rds_ibdev->dev, ibmr->sg, ibmr->sg_len, DMA_BIDIRECTIONAL); ibmr->sg_dma_len = 0; + ibmr->sg = NULL; + ibmr->sg_len = 0; return ret; } From c2de369c5c5b8599ca10fd5ca8d11fcd845c1331 Mon Sep 17 00:00:00 2001 From: Haseeb Malik Date: Mon, 21 Sep 2026 16:40:30 -0400 Subject: [PATCH 512/694] macsec: initialize SecY before registering the netdevice Creating a MACsec device with MAC offload over an LRO-capable lower device triggers a warning in rtmsg_ifinfo_build_skb() when IPv4 forwarding is enabled by default. register_netdevice() invokes inetdev_init(), which disables LRO and emits a NETDEV_FEAT_CHANGE notification. This reaches macsec_fill_info() before macsec_add_dev() initializes the SecY. key_len is still zero, so macsec_fill_info() returns -EMSGSIZE and trips the WARN_ON in rtmsg_ifinfo_build_skb(), even though the skb has enough space. Even without the warning, notifications during registration can report uninitialized SecY attributes, including the SCI. This ordering has existed since the driver was introduced. Initialize the SecY and apply the new-link attributes before registration. Move MAC address inheritance into macsec_newlink() so the SCI can also be initialized before registration-time notifications report it. Move the per-CPU statistics and metadata destination allocation into ndo_init(), and release partial allocations on failure. Fixes: c09440f7dcb3 ("macsec: introduce IEEE 802.1AE driver") Reported-by: syzbot+f2f6312ad1b5a0bfe316@syzkaller.appspotmail.com Closes: https://syzkaller.appspot.com/bug?extid=f2f6312ad1b5a0bfe316 Suggested-by: Sabrina Dubroca Link: https://lists.openwall.net/linux-kernel/2026/08/19/552 Signed-off-by: Haseeb Malik Reviewed-by: Sabrina Dubroca Link: https://patch.msgid.link/20260921-fix-macsec-net-v3-1-accf94f93f5e@gmail.com Signed-off-by: Paolo Abeni --- drivers/net/macsec.c | 84 +++++++++++++++++++++++--------------------- 1 file changed, 43 insertions(+), 41 deletions(-) diff --git a/drivers/net/macsec.c b/drivers/net/macsec.c index 6f9f3aceffaa..78a19b134632 100644 --- a/drivers/net/macsec.c +++ b/drivers/net/macsec.c @@ -3539,6 +3539,22 @@ static int macsec_dev_init(struct net_device *dev) if (err) return err; + err = -ENOMEM; + macsec->stats = netdev_alloc_pcpu_stats(struct pcpu_secy_stats); + if (!macsec->stats) + goto destroy_gro_cells; + + macsec->secy.tx_sc.stats = + netdev_alloc_pcpu_stats(struct pcpu_tx_sc_stats); + if (!macsec->secy.tx_sc.stats) + goto free_secy_stats; + + macsec->secy.tx_sc.md_dst = metadata_dst_alloc(0, METADATA_MACSEC, + GFP_KERNEL); + if (!macsec->secy.tx_sc.md_dst) + goto free_tx_sc_stats; + macsec->secy.tx_sc.md_dst->u.macsec_info.sci = macsec->secy.sci; + macsec_inherit_tso_max(dev); dev->hw_features = real_dev->hw_features & MACSEC_OFFLOAD_FEATURES; @@ -3551,8 +3567,6 @@ static int macsec_dev_init(struct net_device *dev) macsec_set_head_tail_room(dev); - if (is_zero_ether_addr(dev->dev_addr)) - eth_hw_addr_inherit(dev, real_dev); if (is_zero_ether_addr(dev->broadcast)) memcpy(dev->broadcast, real_dev->broadcast, dev->addr_len); @@ -3560,6 +3574,14 @@ static int macsec_dev_init(struct net_device *dev) netdev_hold(real_dev, &macsec->dev_tracker, GFP_KERNEL); return 0; + +free_tx_sc_stats: + free_percpu(macsec->secy.tx_sc.stats); +free_secy_stats: + free_percpu(macsec->stats); +destroy_gro_cells: + gro_cells_destroy(&macsec->gro_cells); + return err; } static void macsec_dev_uninit(struct net_device *dev) @@ -4116,26 +4138,11 @@ static sci_t dev_to_sci(struct net_device *dev, __be16 port) return make_sci(dev->dev_addr, port); } -static int macsec_add_dev(struct net_device *dev, sci_t sci, u8 icv_len) +static void macsec_init_secy(struct net_device *dev, sci_t sci, u8 icv_len) { struct macsec_dev *macsec = macsec_priv(dev); struct macsec_secy *secy = &macsec->secy; - macsec->stats = netdev_alloc_pcpu_stats(struct pcpu_secy_stats); - if (!macsec->stats) - return -ENOMEM; - - secy->tx_sc.stats = netdev_alloc_pcpu_stats(struct pcpu_tx_sc_stats); - if (!secy->tx_sc.stats) - return -ENOMEM; - - secy->tx_sc.md_dst = metadata_dst_alloc(0, METADATA_MACSEC, GFP_KERNEL); - if (!secy->tx_sc.md_dst) - /* macsec and secy percpu stats will be freed when unregistering - * net_device in macsec_free_netdev() - */ - return -ENOMEM; - if (sci == MACSEC_UNDEF_SCI) sci = dev_to_sci(dev, MACSEC_PORT_ES); @@ -4149,15 +4156,12 @@ static int macsec_add_dev(struct net_device *dev, sci_t sci, u8 icv_len) secy->xpn = DEFAULT_XPN; secy->sci = sci; - secy->tx_sc.md_dst->u.macsec_info.sci = sci; secy->tx_sc.active = true; secy->tx_sc.encoding_sa = DEFAULT_ENCODING_SA; secy->tx_sc.encrypt = DEFAULT_ENCRYPT; secy->tx_sc.send_sci = DEFAULT_SEND_SCI; secy->tx_sc.end_station = false; secy->tx_sc.scb = false; - - return 0; } static struct lock_class_key macsec_netdev_addr_lock_key; @@ -4220,6 +4224,24 @@ static int macsec_newlink(struct net_device *dev, if (rx_handler && rx_handler != macsec_handle_frame) return -EBUSY; + if (is_zero_ether_addr(dev->dev_addr)) + eth_hw_addr_inherit(dev, real_dev); + + if (data && data[IFLA_MACSEC_SCI]) + sci = nla_get_sci(data[IFLA_MACSEC_SCI]); + else if (data && data[IFLA_MACSEC_PORT]) + sci = dev_to_sci(dev, nla_get_be16(data[IFLA_MACSEC_PORT])); + else + sci = dev_to_sci(dev, MACSEC_PORT_ES); + + /* Registration can notify listeners before returning. */ + macsec_init_secy(dev, sci, icv_len); + if (data) { + err = macsec_changelink_common(dev, data); + if (err) + return err; + } + err = register_netdevice(dev); if (err < 0) return err; @@ -4232,31 +4254,11 @@ static int macsec_newlink(struct net_device *dev, if (err < 0) goto unregister; - /* need to be already registered so that ->init has run and - * the MAC addr is set - */ - if (data && data[IFLA_MACSEC_SCI]) - sci = nla_get_sci(data[IFLA_MACSEC_SCI]); - else if (data && data[IFLA_MACSEC_PORT]) - sci = dev_to_sci(dev, nla_get_be16(data[IFLA_MACSEC_PORT])); - else - sci = dev_to_sci(dev, MACSEC_PORT_ES); - if (rx_handler && sci_exists(real_dev, sci)) { err = -EBUSY; goto unlink; } - err = macsec_add_dev(dev, sci, icv_len); - if (err) - goto unlink; - - if (data) { - err = macsec_changelink_common(dev, data); - if (err) - goto del_dev; - } - /* If h/w offloading is available, propagate to the device */ if (macsec_is_offloaded(macsec)) { const struct macsec_ops *ops; From 15814c01ac57643edf1662a076d2961332f277a5 Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Wed, 23 Sep 2026 21:44:30 -0400 Subject: [PATCH 513/694] drm/amd/display: Bump frame warning limit for all builds of dml Some configs with gcc are also now affected. Signed-off-by: Alex Deucher (cherry picked from commit 76b9706e7fa1a6e374703170128b1be2f590cda7) --- drivers/gpu/drm/amd/display/dc/dml/Makefile | 6 +----- drivers/gpu/drm/amd/display/dc/dml2_0/Makefile | 2 +- 2 files changed, 2 insertions(+), 6 deletions(-) diff --git a/drivers/gpu/drm/amd/display/dc/dml/Makefile b/drivers/gpu/drm/amd/display/dc/dml/Makefile index 79eeb3721b99..91465ac05c97 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/Makefile +++ b/drivers/gpu/drm/amd/display/dc/dml/Makefile @@ -36,11 +36,7 @@ ifneq ($(CONFIG_FRAME_WARN),0) frame_warn_limit := 3072 endif else - ifeq ($(CONFIG_CC_IS_CLANG),y) - frame_warn_limit := 3072 - else - frame_warn_limit := 2048 - endif + frame_warn_limit := 3072 endif ifeq ($(call test-lt, $(CONFIG_FRAME_WARN), $(frame_warn_limit)),y) diff --git a/drivers/gpu/drm/amd/display/dc/dml2_0/Makefile b/drivers/gpu/drm/amd/display/dc/dml2_0/Makefile index 4d682e92df1d..39ee2d1999dd 100644 --- a/drivers/gpu/drm/amd/display/dc/dml2_0/Makefile +++ b/drivers/gpu/drm/amd/display/dc/dml2_0/Makefile @@ -35,7 +35,7 @@ ifneq ($(CONFIG_FRAME_WARN),0) frame_warn_limit := 3072 endif else - frame_warn_limit := 2056 + frame_warn_limit := 3072 endif ifeq ($(call test-lt, $(CONFIG_FRAME_WARN), $(frame_warn_limit)),y) From d6ec384c87cc851cfd13bb18c99ce351ccee6192 Mon Sep 17 00:00:00 2001 From: Fang Xieyan Date: Mon, 21 Sep 2026 20:54:41 +0800 Subject: [PATCH 514/694] net/sched: act_ife: validate metadata length before decoding skbmark_decode(), skbprio_decode() and skbtcindex_decode() read fixed-size values from the TLV payload without validating its length. A malformed IFE frame can declare a shorter payload, causing the decoders to consume bytes beyond the declared metadata value: [TLV type=IFE_META_SKBMARK len=4] -> dlen == 0, but decode reads 4 bytes The decoder may therefore set skb metadata from unintended input. Validate the payload length before decoding and return -EINVAL for invalid lengths. Read the values with get_unaligned_be32() and get_unaligned_be16(), as TLV payloads are not guaranteed to be aligned. Teach tcf_ife_decode() to log a decoder error separately from an unknown metaid; both are counted as overlimits and decoding continues with the remaining metadata. The metadata length issue was found by an automated audit of the IFE decode path at v6.18-rc7 and reproduced with a userspace sanitizer model of the decode path. Compile-tested on x86_64 with defconfig and NET_ACT_IFE=y: act_ife.o and the three act_meta_*.o build warning-free. Fixes: 084e2f6566d2 ("Support to encoding decoding skb mark on IFE action") Fixes: 200e10f46936 ("Support to encoding decoding skb prio on IFE action") Fixes: 408fbc22ef1e ("net sched ife action: Introduce skb tcindex metadata encap decap") Assisted-by: Hawkeye:GLM-5.3-flash Assisted-by: Qoder:Qwen3.8-Max Signed-off-by: Fang Xieyan Link: https://patch.msgid.link/20260921125441.81459-1-fangxy@xiaopeng.com Signed-off-by: Paolo Abeni --- net/sched/act_ife.c | 17 ++++++++++++----- net/sched/act_meta_mark.c | 6 ++++-- net/sched/act_meta_skbprio.c | 6 ++++-- net/sched/act_meta_skbtcindex.c | 6 ++++-- 4 files changed, 24 insertions(+), 11 deletions(-) diff --git a/net/sched/act_ife.c b/net/sched/act_ife.c index 9cea71fc1db3..2afd68983ece 100644 --- a/net/sched/act_ife.c +++ b/net/sched/act_ife.c @@ -737,6 +737,7 @@ static int tcf_ife_decode(struct sk_buff *skb, const struct tc_action *a, u8 *curr_data; u16 mtype; u16 dlen; + int ret; curr_data = ife_tlv_meta_decode(tlv_data, ifehdr_end, &mtype, &dlen, NULL); @@ -745,13 +746,19 @@ static int tcf_ife_decode(struct sk_buff *skb, const struct tc_action *a, return TC_ACT_SHOT; } - if (find_decode_metaid(skb, p, mtype, dlen, curr_data)) { - /* abuse overlimits to count when we receive metadata - * but dont have an ops for it + ret = find_decode_metaid(skb, p, mtype, dlen, curr_data); + if (ret < 0) { + /* abuse overlimits to count metadata we cannot + * decode: no ops for it, or the decoder rejected it */ - pr_info_ratelimited("Unknown metaid %d dlen %d\n", - mtype, dlen); qstats_cpu_overlimit_inc(ife->common.cpu_qstats); + + if (ret == -ENOENT) + pr_info_ratelimited("Unknown metaid %d dlen %d\n", + mtype, dlen); + else + pr_info_ratelimited("Failed to decode metaid %d dlen %d err %d\n", + mtype, dlen, ret); } } diff --git a/net/sched/act_meta_mark.c b/net/sched/act_meta_mark.c index ea0573cb8b2d..e2f61b22bf0f 100644 --- a/net/sched/act_meta_mark.c +++ b/net/sched/act_meta_mark.c @@ -10,6 +10,7 @@ #include #include #include +#include #include #include #include @@ -28,9 +29,10 @@ static int skbmark_encode(struct sk_buff *skb, void *skbdata, static int skbmark_decode(struct sk_buff *skb, void *data, u16 len) { - u32 ifemark = *(u32 *)data; + if (len != sizeof(u32)) + return -EINVAL; - skb->mark = ntohl(ifemark); + skb->mark = get_unaligned_be32(data); return 0; } diff --git a/net/sched/act_meta_skbprio.c b/net/sched/act_meta_skbprio.c index 2df3133ce5ad..5cdb57931eab 100644 --- a/net/sched/act_meta_skbprio.c +++ b/net/sched/act_meta_skbprio.c @@ -10,6 +10,7 @@ #include #include #include +#include #include #include #include @@ -33,9 +34,10 @@ static int skbprio_encode(struct sk_buff *skb, void *skbdata, static int skbprio_decode(struct sk_buff *skb, void *data, u16 len) { - u32 ifeprio = *(u32 *)data; + if (len != sizeof(u32)) + return -EINVAL; - skb->priority = ntohl(ifeprio); + skb->priority = get_unaligned_be32(data); return 0; } diff --git a/net/sched/act_meta_skbtcindex.c b/net/sched/act_meta_skbtcindex.c index 44547caead46..8803710c0905 100644 --- a/net/sched/act_meta_skbtcindex.c +++ b/net/sched/act_meta_skbtcindex.c @@ -10,6 +10,7 @@ #include #include #include +#include #include #include #include @@ -28,9 +29,10 @@ static int skbtcindex_encode(struct sk_buff *skb, void *skbdata, static int skbtcindex_decode(struct sk_buff *skb, void *data, u16 len) { - u16 ifetc_index = *(u16 *)data; + if (len != sizeof(u16)) + return -EINVAL; - skb->tc_index = ntohs(ifetc_index); + skb->tc_index = get_unaligned_be16(data); return 0; } From 7c9f391ec89cb621d7af375ace2eb9a6248e5b9d Mon Sep 17 00:00:00 2001 From: Christian Lamparter Date: Mon, 21 Sep 2026 18:28:17 +0200 Subject: [PATCH 515/694] net: emac: move setting of netops to fix crash fixes the following crash on driver initialization: |BUG: Kernel NULL pointer dereference on read at 0x00000158 |Faulting instruction address: 0xc0566b40 |Oops: Kernel access of bad area, sig: 11 [#1] |BE PAGE_SIZE=4K PowerPC 44x Platform |Modules linked in: |CPU: 0 UID: 0 PID: 1 Comm: swapper/0 Tainted: GW 7.3.0-rc3+ #1 |Tainted: [W]=WARN |Hardware name: MyBook Live APM821XX 0x12c41c83 PowerPC 44x Platform |NIP: c0566b40 LR: c05648f8 CTR: c04c1e9c |REGS: c1053a20 TRAP: 0300 Tainted: GW (7.3.0-rc3+) |MSR: 0002b000 CR: 24008808 XER: 00000000 |DEAR: 00000158 ESR: 00000000 |GPR00: c05648f8 c1053b10 c1063600 c1030000 c5ab3000 00000000 [...] |GPR08: 00000002 00000000 00000000 c1053b40 84002808 00000000 [...] |GPR16: cfffd210 00000002 c0beafcc cfffc960 00000000 c1030644 [...] |GPR24: c0beafbc c1037000 00000000 0000000a 00000000 c1030000 [...] |NIP [c0566b40] phy_link_topo_add_phy+0x2c/0x1d0 |LR [c05648f8] phy_attach_direct+0x1a4/0x368 |Call Trace: |[c1053b10] [c0811e04] klist_put+0x54/0xb4 (unreliable) |[c1053b40] [c05648f8] phy_attach_direct+0x1a4/0x368 |[c1053b70] [c0564ae8] phy_connect_direct+0x2c/0x60 |[c1053b90] [c056c7a8] of_phy_connect+0x50/0x74 |[c1053bc0] [c0572a88] emac_probe+0xd50/0x119c |[c1053c90] [c04cb770] platform_probe+0x74/0xa4 |[c1053cb0] [c04c8f08] really_probe+0x120/0x2b0 |[c1053cd0] [c04c9254] __driver_probe_device+0x1bc/0x1fc |[c1053d00] [c04c9334] driver_probe_device+0x38/0xa8 |[c1053d30] [c04c9560] __driver_attach+0xf4/0x10c |[c1053d50] [c04c6af8] bus_for_each_dev+0x68/0xd0 |[c1053d90] [c04c7c34] bus_add_driver+0xcc/0x1ec |[c1053dc0] [c04c9f6c] driver_register+0xcc/0x110 |[c1053de0] [c0aa3f40] emac_init+0x1c4/0x200 This bug showed up starting with v7.3-rc1. At the NIP in phy_link_topo_add_phy() is a netdev_need_ops_lock() check. This was added by the following commit ded86da4bbb7 ("net: ethtool: relax ethnl_req_get_phydev() locking assertion") The bug shows up because at the time of_phy_connect() was called, the netdev_ops were *not yet* determined. My fix is to move the code that sets netdev_ops+commac.ops+ethtool_ops further up as emac_init_config() derives that by looking at the device-tree and sets the required dev->phy_mode accordingly. During review, the Sashiko bot's AI stated that the commac assignment became a dead store. Great catch! To keep the original behavior as-is, one mentioned option "set dev->commac.ops = &emac_commac_ops only in the non-gige case?" sounded like a great plan. So the dev->commac.ops assignment for the non-gige-case moves into the else block. This patch was tested on a WD MyBook Live (RGMII). the device now works again. Fixes: ded86da4bbb7 ("net: ethtool: relax ethnl_req_get_phydev() locking assertion") Signed-off-by: Christian Lamparter Reviewed-by: Nicolai Buchwitz Reviewed-by: Maxime Chevallier Link: https://patch.msgid.link/49cd7343bc0e507c022071f3e2b5662b053dca73.1790007431.git.chunkeey@gmail.com Signed-off-by: Paolo Abeni --- drivers/net/ethernet/ibm/emac/core.c | 16 +++++++++------- 1 file changed, 9 insertions(+), 7 deletions(-) diff --git a/drivers/net/ethernet/ibm/emac/core.c b/drivers/net/ethernet/ibm/emac/core.c index 1d46cf6c2c12..e7043523457c 100644 --- a/drivers/net/ethernet/ibm/emac/core.c +++ b/drivers/net/ethernet/ibm/emac/core.c @@ -3044,6 +3044,15 @@ static int emac_probe(struct platform_device *ofdev) if (err) goto err_gone; + if (emac_phy_supports_gige(dev->phy_mode)) { + ndev->netdev_ops = &emac_gige_netdev_ops; + dev->commac.ops = &emac_commac_sg_ops; + } else { + ndev->netdev_ops = &emac_netdev_ops; + dev->commac.ops = &emac_commac_ops; + } + ndev->ethtool_ops = &emac_ethtool_ops; + dev->emacp = devm_platform_ioremap_resource(ofdev, 0); if (IS_ERR(dev->emacp)) { err = PTR_ERR(dev->emacp); @@ -3076,7 +3085,6 @@ static int emac_probe(struct platform_device *ofdev) dev->mdio_instance = platform_get_drvdata(dev->mdio_dev); /* Register with MAL */ - dev->commac.ops = &emac_commac_ops; dev->commac.dev = dev; dev->commac.tx_chan_mask = MAL_CHAN_MASK(dev->mal_tx_chan); dev->commac.rx_chan_mask = MAL_CHAN_MASK(dev->mal_rx_chan); @@ -3144,12 +3152,6 @@ static int emac_probe(struct platform_device *ofdev) ndev->features |= ndev->hw_features | NETIF_F_RXCSUM; } ndev->watchdog_timeo = 5 * HZ; - if (emac_phy_supports_gige(dev->phy_mode)) { - ndev->netdev_ops = &emac_gige_netdev_ops; - dev->commac.ops = &emac_commac_sg_ops; - } else - ndev->netdev_ops = &emac_netdev_ops; - ndev->ethtool_ops = &emac_ethtool_ops; /* MTU range: 46 - 1500 or whatever is in OF */ ndev->min_mtu = EMAC_MIN_MTU; From e66cf1625ec4a3fe68346119f371def713fd0a4d Mon Sep 17 00:00:00 2001 From: Namjae Jeon Date: Wed, 23 Sep 2026 20:25:49 +0900 Subject: [PATCH 516/694] smb: client: use finish_no_open() for non-regular inodes An O_CREAT open can find an existing symlink or another non-regular inode. cifs_atomic_open() calls finish_open() on it and attaches a cifsFileInfo. Symlink inodes have no CIFS release operation, so the dentry reference held by cifsFileInfo is leaked. FMODE_OPENED also prevents the VFS from following the symlink. Track whether cifs_do_create() returned an open server handle. For non-regular inodes, close the handle if present, remove the pending open, and call finish_no_open() so the VFS can continue the lookup. Do not set FMODE_CREATED unless a regular file was opened. For O_NOFOLLOW with __O_REGULAR, return -ELOOP before the VFS's -EFTYPE check. Defer closing a legacy POSIX handle on a non-regular inode until after inode lookup. This avoids closing it again if lookup fails. Fixes: d2c127197dfc ("cifs: implement i_op->atomic_open()") Signed-off-by: Namjae Jeon Signed-off-by: Paulo Alcantara Cc: stable@vger.kernel.org --- fs/smb/client/dir.c | 52 +++++++++++++++++++++++++++++++++------------ 1 file changed, 39 insertions(+), 13 deletions(-) diff --git a/fs/smb/client/dir.c b/fs/smb/client/dir.c index 6fa6d48fdfd3..1a56fa4d0e89 100644 --- a/fs/smb/client/dir.c +++ b/fs/smb/client/dir.c @@ -199,7 +199,7 @@ static int __cifs_do_create(struct inode *dir, struct dentry *direntry, struct tcon_link *tlink, unsigned int oflags, umode_t mode, __u32 *oplock, struct cifs_fid *fid, struct cifs_open_info_data *buf, - struct inode **inode) + struct inode **inode, bool *opened) { int rc = -ENOENT; int create_options = CREATE_NOT_DIR; @@ -216,6 +216,7 @@ static int __cifs_do_create(struct inode *dir, struct dentry *direntry, __le32 lease_flags = 0; *inode = NULL; + *opened = false; *oplock = 0; if (tcon->ses->server->oplocks) *oplock = REQ_OPLOCK; @@ -232,6 +233,7 @@ static int __cifs_do_create(struct inode *dir, struct dentry *direntry, oflags, oplock, &fid->netfid, xid); switch (rc) { case 0: + *opened = true; if (newinode == NULL) { /* query inode info */ goto cifs_create_get_file_info; @@ -253,11 +255,9 @@ static int __cifs_do_create(struct inode *dir, struct dentry *direntry, /* * The server may allow us to open things like * FIFOs, but the client isn't set up to deal - * with that. If it's not a regular file, just - * close it and proceed as if it were a normal - * lookup. + * with that. Keep the handle until the caller + * can finish the lookup. */ - CIFSSMBClose(xid, tcon, fid->netfid); goto cifs_create_get_file_info; } /* success, no need to query */ @@ -384,6 +384,7 @@ static int __cifs_do_create(struct inode *dir, struct dentry *direntry, } return rc; } + *opened = true; if (rdwr_for_fscache == 2) cifs_invalidate_cache(dir, FSCACHE_INVAL_DIO_WRITE); @@ -475,7 +476,7 @@ static int __cifs_do_create(struct inode *dir, struct dentry *direntry, return rc; out_err: - if (server->ops->close) + if (*opened && server->ops->close) server->ops->close(xid, tcon, fid); if (newinode) iput(newinode); @@ -487,7 +488,7 @@ static int cifs_do_create(struct inode *dir, struct dentry *direntry, unsigned int oflags, umode_t mode, __u32 *oplock, struct cifs_fid *fid, struct cifs_open_info_data *buf, - struct inode **inode) + struct inode **inode, bool *opened) { void *page = alloc_dentry_path(); const char *full_path; @@ -496,10 +497,11 @@ static int cifs_do_create(struct inode *dir, struct dentry *direntry, full_path = build_path_from_dentry(direntry, page); if (IS_ERR(full_path)) { rc = PTR_ERR(full_path); + *opened = false; } else { rc = __cifs_do_create(dir, direntry, full_path, xid, tlink, oflags, mode, oplock, - fid, buf, inode); + fid, buf, inode, opened); } free_dentry_path(page); return rc; @@ -529,6 +531,8 @@ int cifs_atomic_open(struct inode *dir, struct dentry *direntry, struct inode *inode; unsigned int xid; __u32 oplock; + bool is_regular; + bool opened; int rc; if (unlikely(cifs_forced_shutdown(cifs_sb))) @@ -581,12 +585,26 @@ int cifs_atomic_open(struct inode *dir, struct dentry *direntry, cifs_add_pending_open(&fid, tlink, &open); rc = cifs_do_create(dir, direntry, xid, tlink, oflags, mode, - &oplock, &fid, &buf, &inode); + &oplock, &fid, &buf, &inode, &opened); if (rc) { cifs_del_pending_open(&open); goto out; } + is_regular = S_ISREG(inode->i_mode); + if (!is_regular || !opened) { + if (opened && server->ops->close) + server->ops->close(xid, tcon, &fid); + cifs_del_pending_open(&open); + if (S_ISLNK(inode->i_mode) && + (oflags & (O_NOFOLLOW | __O_REGULAR)) == + (O_NOFOLLOW | __O_REGULAR) && !(oflags & O_EXCL)) { + iput(inode); + rc = -ELOOP; + goto out; + } + } + if (d_in_lookup(direntry)) { alias = d_splice_alias(inode, direntry); if (!IS_ERR_OR_NULL(alias)) @@ -595,9 +613,15 @@ int cifs_atomic_open(struct inode *dir, struct dentry *direntry, d_instantiate(direntry, inode); } - if ((oflags & (O_CREAT | O_EXCL)) == (O_CREAT | O_EXCL)) + if (is_regular && opened && + (oflags & (O_CREAT | O_EXCL)) == (O_CREAT | O_EXCL)) file->f_mode |= FMODE_CREATED; + if (!is_regular || !opened) { + rc = finish_no_open(file, NULL); + goto out; + } + rc = finish_open(file, direntry, generic_file_open); if (rc) { if (server->ops->close) @@ -660,6 +684,7 @@ int cifs_create(struct mnt_idmap *idmap, struct inode *dir, struct inode *inode; struct cifs_fid fid; __u32 oplock; + bool opened; struct cifs_open_info_data buf = {}; cifs_dbg(FYI, "cifs_create parent inode = 0x%p name is: %pd and dentry = 0x%p\n", @@ -682,10 +707,10 @@ int cifs_create(struct mnt_idmap *idmap, struct inode *dir, server->ops->new_lease_key(&fid); rc = cifs_do_create(dir, direntry, xid, tlink, oflags, - mode, &oplock, &fid, &buf, &inode); + mode, &oplock, &fid, &buf, &inode, &opened); if (!rc) { d_instantiate(direntry, inode); - if (server->ops->close) + if (opened && server->ops->close) server->ops->close(xid, tcon, &fid); } @@ -1078,6 +1103,7 @@ int cifs_tmpfile(struct mnt_idmap *idmap, struct inode *dir, struct inode *inode; unsigned int xid; __u32 oplock; + bool opened; int namelen; int rc; @@ -1116,7 +1142,7 @@ int cifs_tmpfile(struct mnt_idmap *idmap, struct inode *dir, namelen = scnprintf(name, namesize, CIFS_TMPNAME_PREFIX "%x", atomic_inc_return(&cifs_tmpcounter)); rc = __cifs_do_create(dir, dentry, path, xid, tlink, oflags, - mode, &oplock, &fid, NULL, &inode); + mode, &oplock, &fid, NULL, &inode, &opened); if (!rc) { rc = d_mark_tmpfile_name(file, &QSTR_LEN(name, namelen)); if (rc) { From c3a66e5f5bab3912e9f84223c5a982bf4333d5a1 Mon Sep 17 00:00:00 2001 From: Donggeun Yoo Date: Thu, 24 Sep 2026 19:23:20 +0900 Subject: [PATCH 517/694] bpf: Zero-fill other CPUs when BPF_F_CPU creates a per-cpu hash element pcpu_init_value() initializes the per-cpu area of a newly created [lru_]percpu_hash element. The area is recycled, so when the value comes from a BPF program (onallcpus == false) it writes the running CPU's slot and zeroes the rest. bpf_percpu_hash_update() passes onallcpus == true, which delegates to pcpu_copy_value(). pcpu_copy_value() writes only the CPU named in map_flags when BPF_F_CPU is set, so on the create path the other slots keep the recycled element's values: update(k1, 0xdeadc0de, BPF_F_ALL_CPUS) every CPU holds 0xdeadc0de delete(k1) element back on the freelist update(k2, 0xc0ffee, BPF_F_CPU | 0) creates, writes CPU 0 only lookup(k2) CPU 0 0xc0ffee, rest 0xdeadc0de Zero-fill the other CPUs on that arm too. Fixes: c6936161fd55 ("bpf: Add BPF_F_CPU and BPF_F_ALL_CPUS flags support for percpu_hash and lru_percpu_hash maps") Signed-off-by: Donggeun Yoo Signed-off-by: Alexei Starovoitov Link: https://patch.msgid.link/20260924102321.2120434-2-donggeunyoo.kernel@gmail.com --- kernel/bpf/hashtab.c | 11 +++++++---- 1 file changed, 7 insertions(+), 4 deletions(-) diff --git a/kernel/bpf/hashtab.c b/kernel/bpf/hashtab.c index 4f495dcbf670..f9464e566f10 100644 --- a/kernel/bpf/hashtab.c +++ b/kernel/bpf/hashtab.c @@ -1054,14 +1054,17 @@ static void pcpu_init_value(struct bpf_htab *htab, void __percpu *pptr, /* When not setting the initial value on all cpus, zero-fill element * values for other cpus. Otherwise, bpf program has no way to ensure * known initial values for cpus other than current one - * (onallcpus=false always when coming from bpf prog). + * (onallcpus=false always when coming from bpf prog, + * map_flags & BPF_F_CPU when coming from syscall but setting + * only one cpu). */ - if (!onallcpus) { - int current_cpu = raw_smp_processor_id(); + if (!onallcpus || (map_flags & BPF_F_CPU)) { + int init_cpu = (map_flags & BPF_F_CPU) ? map_flags >> 32 : + raw_smp_processor_id(); int cpu; for_each_possible_cpu(cpu) { - if (cpu == current_cpu) + if (cpu == init_cpu) copy_map_value(&htab->map, per_cpu_ptr(pptr, cpu), value); else /* Since elem is preallocated, we cannot touch special fields */ zero_map_value(&htab->map, per_cpu_ptr(pptr, cpu)); From 3422808f4e959266ea7790101ac90b8341f48226 Mon Sep 17 00:00:00 2001 From: Donggeun Yoo Date: Thu, 24 Sep 2026 19:23:21 +0900 Subject: [PATCH 518/694] selftests/bpf: Test per-cpu initialization of a BPF_F_CPU created element The existing cpu_flag subtests always prime a key with BPF_F_ALL_CPUS before any BPF_F_CPU write, so the create path is never covered. Add a subtest that creates the element with BPF_F_CPU on a map with max_entries 1, so the key can only reuse the element the previous key released, and check that the CPUs the update did not name read back zero. Run it for PERCPU_HASH preallocated and BPF_F_NO_PREALLOC, whose per-cpu areas come from different allocators, and for LRU_PERCPU_HASH. Under BPF_F_NO_PREALLOC the reuse is only guaranteed on the cpu that ran the delete, so pin the thread across the pair, and name a cpu other than that one in map_flags. Signed-off-by: Donggeun Yoo Signed-off-by: Alexei Starovoitov Link: https://patch.msgid.link/20260924102321.2120434-3-donggeunyoo.kernel@gmail.com --- .../selftests/bpf/prog_tests/percpu_alloc.c | 105 ++++++++++++++++++ 1 file changed, 105 insertions(+) diff --git a/tools/testing/selftests/bpf/prog_tests/percpu_alloc.c b/tools/testing/selftests/bpf/prog_tests/percpu_alloc.c index a72ae0b29f6e..7b4a1e24363b 100644 --- a/tools/testing/selftests/bpf/prog_tests/percpu_alloc.c +++ b/tools/testing/selftests/bpf/prog_tests/percpu_alloc.c @@ -1,4 +1,6 @@ // SPDX-License-Identifier: GPL-2.0 +#define _GNU_SOURCE +#include #include #include "cgroup_helpers.h" #include "percpu_alloc_array.skel.h" @@ -350,6 +352,103 @@ static void test_lru_percpu_hash_cpu_flag(void) test_percpu_map_cpu_flag(BPF_MAP_TYPE_LRU_PERCPU_HASH); } +/* + * A BPF_F_CPU update that creates an element must zero the value on the other + * cpus, rather than leave them holding whatever the recycled element last + * contained. max_entries is 1 so the second key can only reuse the element + * the first one released. + */ +static void test_percpu_map_cpu_flag_create(enum bpf_map_type map_type, __u32 map_flags) +{ + LIBBPF_OPTS(bpf_map_create_opts, opts, .map_flags = map_flags); + const u32 stale = 0xDEADC0DE, fresh = 0xC0FFEE; + int nr_cpus, cpu, map_fd, err, key; + int pinned_cpu, value_cpu; + cpu_set_t old_mask, new_mask; + bool restore_mask = false; + u32 value; + u64 flags; + + nr_cpus = libbpf_num_possible_cpus(); + if (!ASSERT_GT(nr_cpus, 0, "libbpf_num_possible_cpus")) + return; + + if (nr_cpus < 2) { + test__skip(); + return; + } + + map_fd = bpf_map_create(map_type, "cpu_flag_create", sizeof(key), sizeof(value), 1, &opts); + if (!ASSERT_GE(map_fd, 0, "bpf_map_create")) + return; + + /* NO_PREALLOC recycles per cpu, so keep the delete and the create on one cpu. */ + err = sched_getaffinity(0, sizeof(old_mask), &old_mask); + if (!ASSERT_OK(err, "sched_getaffinity")) + goto out; + + pinned_cpu = sched_getcpu(); + if (!ASSERT_GE(pinned_cpu, 0, "sched_getcpu")) + goto out; + + CPU_ZERO(&new_mask); + CPU_SET(pinned_cpu, &new_mask); + err = sched_setaffinity(0, sizeof(new_mask), &new_mask); + if (!ASSERT_OK(err, "sched_setaffinity")) + goto out; + restore_mask = true; + + value_cpu = pinned_cpu ? 0 : 1; + + key = 1; + value = stale; + err = bpf_map_update_elem(map_fd, &key, &value, BPF_F_ALL_CPUS); + if (!ASSERT_OK(err, "bpf_map_update_elem all_cpus")) + goto out; + + err = bpf_map_delete_elem(map_fd, &key); + if (!ASSERT_OK(err, "bpf_map_delete_elem")) + goto out; + + key = 2; + value = fresh; + flags = (u64)value_cpu << 32 | BPF_F_CPU; + err = bpf_map_update_elem(map_fd, &key, &value, flags); + if (!ASSERT_OK(err, "bpf_map_update_elem specified cpu")) + goto out; + + for (cpu = 0; cpu < nr_cpus; cpu++) { + value = 0; + flags = (u64)cpu << 32 | BPF_F_CPU; + err = bpf_map_lookup_elem_flags(map_fd, &key, &value, flags); + if (!ASSERT_OK(err, "bpf_map_lookup_elem_flags specified cpu")) + goto out; + if (!ASSERT_EQ(value, cpu == value_cpu ? fresh : 0, "value on specified cpu")) + goto out; + } + +out: + if (restore_mask) + sched_setaffinity(0, sizeof(old_mask), &old_mask); + close(map_fd); +} + +static void test_percpu_hash_cpu_flag_create(void) +{ + test_percpu_map_cpu_flag_create(BPF_MAP_TYPE_PERCPU_HASH, 0); +} + +static void test_percpu_hash_cpu_flag_create_malloc(void) +{ + test_percpu_map_cpu_flag_create(BPF_MAP_TYPE_PERCPU_HASH, BPF_F_NO_PREALLOC); +} + +static void test_lru_percpu_hash_cpu_flag_create(void) +{ + /* lru without prealloc is -ENOTSUPP, so there is no malloc variant */ + test_percpu_map_cpu_flag_create(BPF_MAP_TYPE_LRU_PERCPU_HASH, 0); +} + static void test_percpu_cgroup_storage_cpu_flag(void) { struct percpu_alloc_array *skel = NULL; @@ -454,6 +553,12 @@ void test_percpu_alloc(void) test_percpu_hash_cpu_flag(); if (test__start_subtest("cpu_flag_lru_percpu_hash")) test_lru_percpu_hash_cpu_flag(); + if (test__start_subtest("cpu_flag_create_percpu_hash")) + test_percpu_hash_cpu_flag_create(); + if (test__start_subtest("cpu_flag_create_percpu_hash_malloc")) + test_percpu_hash_cpu_flag_create_malloc(); + if (test__start_subtest("cpu_flag_create_lru_percpu_hash")) + test_lru_percpu_hash_cpu_flag_create(); if (test__start_subtest("cpu_flag_percpu_cgroup_storage")) test_percpu_cgroup_storage_cpu_flag(); if (test__start_subtest("cpu_flag_array")) From c2cdef41e0b4d8ed23a5b41e6ad4e64594e055e4 Mon Sep 17 00:00:00 2001 From: Aleksei Sviridkin Date: Fri, 18 Sep 2026 04:50:19 +0300 Subject: [PATCH 519/694] net: dsa: mt7530: fix NULL dereference on unbind of MT7531 and MT7621 The core and io supplies are only requested for ID_MT7530: both the devm_regulator_get() in probe and the regulator_enable() in mt7530_setup() are guarded by the switch id, but mt7530_remove() disables them unconditionally. On an MT7621 or an MT7531 both pointers are still NULL from devm_kzalloc(), so rmmod or a sysfs unbind calls regulator_disable() on NULL. Fixes: ddda1ac116c8 ("net: dsa: mt7530: support the 7530 switch on the Mediatek MT7621 SoC") Signed-off-by: Aleksei Sviridkin Link: https://patch.msgid.link/20260918015020.2518315-2-f@lex.la Signed-off-by: Jakub Kicinski --- drivers/net/dsa/mt7530-mdio.c | 18 ++++++++++-------- 1 file changed, 10 insertions(+), 8 deletions(-) diff --git a/drivers/net/dsa/mt7530-mdio.c b/drivers/net/dsa/mt7530-mdio.c index 784dd58a7158..de42f70afcfa 100644 --- a/drivers/net/dsa/mt7530-mdio.c +++ b/drivers/net/dsa/mt7530-mdio.c @@ -227,15 +227,17 @@ mt7530_remove(struct mdio_device *mdiodev) if (!priv) return; - ret = regulator_disable(priv->core_pwr); - if (ret < 0) - dev_err(priv->dev, - "Failed to disable core power: %d\n", ret); + if (priv->id == ID_MT7530) { + ret = regulator_disable(priv->core_pwr); + if (ret < 0) + dev_err(priv->dev, + "Failed to disable core power: %d\n", ret); - ret = regulator_disable(priv->io_pwr); - if (ret < 0) - dev_err(priv->dev, "Failed to disable io pwr: %d\n", - ret); + ret = regulator_disable(priv->io_pwr); + if (ret < 0) + dev_err(priv->dev, "Failed to disable io pwr: %d\n", + ret); + } mt7530_remove_common(priv); From 0d80ba0a204c6a16bd7778b50de578dff107c0fe Mon Sep 17 00:00:00 2001 From: Aleksei Sviridkin Date: Fri, 18 Sep 2026 04:50:20 +0300 Subject: [PATCH 520/694] net: dsa: mt7530: leave the MDIO IRQ mappings to regmap-irq mt7530_remove_common() disposes the per-PHY interrupt mappings from .remove, but the regmap-irq chip that owns the domain is devm-registered, so its parent interrupt is only freed once .remove has returned. The switch's own regmap-irq thread can therefore still dispatch on a mapping that is already gone: irq_find_mapping() returns 0, irq_to_desc() returns NULL and handle_nested_irq() locks desc->lock without checking it. The attached PHYs have not given those interrupts back yet either, which the kernel warns about a moment before the fault. regmap_del_irq_chip() disposes the same mappings itself, after freeing the parent interrupt and before removing the domain, so there is nothing left for the driver to do here. Until it runs the descriptors stay alive, and a late dispatch on one of them is harmless: dsa_unregister_switch() has freed the PHY handlers by then, so handle_nested_irq() finds no action and returns. Fixes: 254f6b272e3b ("dsa: mt7530: Utilize REGMAP_IRQ for interrupt handling") Signed-off-by: Aleksei Sviridkin Link: https://patch.msgid.link/20260918015020.2518315-3-f@lex.la Signed-off-by: Jakub Kicinski --- drivers/net/dsa/mt7530.c | 3 --- 1 file changed, 3 deletions(-) diff --git a/drivers/net/dsa/mt7530.c b/drivers/net/dsa/mt7530.c index 3e61eb3c2b1e..96832852c65a 100644 --- a/drivers/net/dsa/mt7530.c +++ b/drivers/net/dsa/mt7530.c @@ -3593,9 +3593,6 @@ EXPORT_SYMBOL_GPL(mt7530_probe_common); void mt7530_remove_common(struct mt7530_priv *priv) { - if (priv->irq_domain) - mt7530_free_mdio_irq(priv); - dsa_unregister_switch(priv->ds); mutex_destroy(&priv->reg_mutex); From 26cc0e69cce062cd3aa6fae33074684669c35a71 Mon Sep 17 00:00:00 2001 From: Hui Peng Date: Mon, 21 Sep 2026 05:10:01 +0000 Subject: [PATCH 521/694] mctp: route: iterate socket tag list in mctp_lookup_prealloc_tag() When a socket transmits a packet with MCTP_TAG_PREALLOC set, mctp_lookup_prealloc_tag() iterates over the per-netns &mns->keys list and matches netid, req_tag, peer_addr, and manual_alloc, without checking whether tmp->sk == &msk->sk. This allows any MCTP socket in the same network namespace to use and consume another socket's preallocated tag. Iterate the socket's own tag list (&msk->keys via sklist) instead of the namespace-wide &mns->keys list in mctp_lookup_prealloc_tag(), ensuring that only tags allocated by msk are matched. Tested in QEMU against Linux 7.3.0-rc3 by allocating a manual tag (0x18) on socket A via SIOCMCTPALLOCTAG for peer EID 9 and sending a 4-byte message with MCTP_TAG_PREALLOC from socket B in the same network namespace. On the unfixed kernel, sendto(sock_b) using socket A's preallocated tag succeeds (ret = 4); with this patch applied, sendto(sock_b) fails with -ENOENT (errno = 2) while sendto(sock_a) succeeds (ret = 4). Fixes: 63ed1aab3d40 ("mctp: Add SIOCMCTP{ALLOC,DROP}TAG ioctls for tag control") Suggested-by: Jeremy Kerr Cc: stable@vger.kernel.org Signed-off-by: Hui Peng Link: https://patch.msgid.link/20260921051002.1656692-1-benquike@gmail.com Signed-off-by: Jakub Kicinski --- net/mctp/route.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/mctp/route.c b/net/mctp/route.c index b19c63a5691a..e7c95eeacb48 100644 --- a/net/mctp/route.c +++ b/net/mctp/route.c @@ -825,7 +825,7 @@ static struct mctp_sk_key *mctp_lookup_prealloc_tag(struct mctp_sock *msk, spin_lock_irqsave(&mns->keys_lock, flags); - hlist_for_each_entry(tmp, &mns->keys, hlist) { + hlist_for_each_entry(tmp, &msk->keys, sklist) { if (tmp->net != netid) continue; From fe99bbeee5c5dbd3abc30721a8079ced59649d97 Mon Sep 17 00:00:00 2001 From: Yilin Zhang Date: Thu, 24 Sep 2026 12:49:00 +0800 Subject: [PATCH 522/694] tcp: fix use-after-free of retransmit_skb_hint in tcp_send_synack() When tcp_send_synack() replaces the cloned SYN skb at the head of the retransmit queue with a copy, it frees the original with tcp_rtx_queue_unlink_and_free() and only repairs tp->highest_sack. tp->retransmit_skb_hint keeps pointing at the freed skbuff_fclone_cache object. The dangling hint is read in tcp_verify_retransmit_hint() and used as the root of the rbtree walk in tcp_xmit_retransmit_queue(). An unprivileged TFO client (sendmsg(MSG_FASTOPEN)) can arm the hint with an attacker-supplied ICMP fragmentation-needed message, after which a simultaneous open frees the armed SYN skb: BUG: KASAN: slab-use-after-free in tcp_mark_skb_lost (net/ipv4/tcp_input.c:1316) Read of size 4 at addr ffff88800604d928 by task swapper/1/0 Call Trace: tcp_mark_skb_lost (net/ipv4/tcp_input.c:1316) tcp_simple_retransmit (net/ipv4/tcp_input.c:3158) tcp_v4_err (net/ipv4/tcp_ipv4.c:587) Sync the hint to the copy. Fixes: c31b70c9968f ("tcp: Add logic to check for SYN w/ data in tcp_simple_retransmit") Reported-by: Kimi Security Team Tested-by: Weiming Shi Signed-off-by: Yilin Zhang Reviewed-by: Eric Dumazet Link: https://patch.msgid.link/8a9dff4063a2745653b7e88ceb745d75efa16e68.1790224474.git.yilinzhang@moonshot.ai Signed-off-by: Jakub Kicinski --- net/ipv4/tcp_output.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c index d960e3de7d50..e0c392e29de5 100644 --- a/net/ipv4/tcp_output.c +++ b/net/ipv4/tcp_output.c @@ -3886,6 +3886,7 @@ void tcp_send_active_reset(struct sock *sk, enum sk_rst_reason reason) */ int tcp_send_synack(struct sock *sk) { + struct tcp_sock *tp = tcp_sk(sk); struct sk_buff *skb; skb = tcp_rtx_queue_head(sk); @@ -3903,6 +3904,8 @@ int tcp_send_synack(struct sock *sk) if (!nskb) return -ENOMEM; INIT_LIST_HEAD(&nskb->tcp_tsorted_anchor); + if (skb == tp->retransmit_skb_hint) + tp->retransmit_skb_hint = nskb; tcp_highest_sack_replace(sk, skb, nskb); tcp_rtx_queue_unlink_and_free(skb, sk); __skb_header_release(nskb); From 1765a153d985c231357145e26798f9408db10e42 Mon Sep 17 00:00:00 2001 From: Guopeng Zhang Date: Thu, 24 Sep 2026 17:15:36 +0800 Subject: [PATCH 523/694] cgroup/pids: Restore pids.events notifications in local mode A fork rejected by the pids controller increments the counter reported by pids.events. When local event accounting is selected, however, pids_event() returns after notifying only events_local_file, leaving pids.events pollers asleep. On legacy hierarchies, pids.events.local does not exist. With pids_localevents, pids.events reports the same local counter. In both cases, pids.events changes without generating a notification. This can be reproduced with a pids_localevents mount: mkdir /tmp/test mount -t cgroup2 -o pids_localevents none /tmp/test mkdir /tmp/test/t echo 1 > /tmp/test/t/pids.max cat /tmp/test/t/pids.events # max 0 timeout 3 inotifywait -e modify /tmp/test/t/pids.events & sh -c 'echo $$ > /tmp/test/t/cgroup.procs; (true &)' 2>/dev/null wait cat /tmp/test/t/pids.events # max 1 Without this patch, inotifywait times out without reporting an event. Notify pids.events before returning from the local event path. Fixes: 3f26a885a068 ("cgroup/pids: Add pids.events.local") Cc: stable@vger.kernel.org # v6.11+ Signed-off-by: Guopeng Zhang Signed-off-by: Tejun Heo --- kernel/cgroup/pids.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/kernel/cgroup/pids.c b/kernel/cgroup/pids.c index ecbb839d2acb..78cdc0558d0c 100644 --- a/kernel/cgroup/pids.c +++ b/kernel/cgroup/pids.c @@ -253,6 +253,11 @@ static void pids_event(struct pids_cgroup *pids_forking, } if (!cgroup_subsys_on_dfl(pids_cgrp_subsys) || cgrp_dfl_root.flags & CGRP_ROOT_PIDS_LOCAL_EVENTS) { + /* + * pids.events reports the local counter on legacy hierarchies + * and when pids_localevents is enabled. + */ + cgroup_file_notify(&p->events_file); cgroup_file_notify(&p->events_local_file); return; } From f75f21ef36285e5f56ee0c428bd2909ee81165b9 Mon Sep 17 00:00:00 2001 From: Ming Wang Date: Sun, 20 Sep 2026 15:44:59 +0800 Subject: [PATCH 524/694] net: usb: cdc_mbim: add MeiG Smart SRM821 to ZLP whitelist The MeiG Smart SRM821 5G module (0x2dee:0x4d53) crashes and drops off the USB bus when it receives a Zero Length Packet (ZLP) after sending or receiving an NTB of exactly 16384 bytes (tx_max). According to the MBIM specification, devices do not require a ZLP if the NTB size is exactly dwNtbOutMaxSize. However, the cdc_mbim driver defaults to sending ZLPs for devices not explicitly whitelisted to accommodate non-conformant hardware. This default behavior breaks the strictly conformant MeiG SRM821 module. Add this device to the ZLP conformance whitelist (cdc_mbim_info) so the driver will pad the NTB to avoid sending ZLPs, preventing the device firmware from crashing. Cc: stable@vger.kernel.org Signed-off-by: Ming Wang Link: https://patch.msgid.link/20260920074500.826121-1-wangming01@loongson.cn Signed-off-by: Jakub Kicinski --- drivers/net/usb/cdc_mbim.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/drivers/net/usb/cdc_mbim.c b/drivers/net/usb/cdc_mbim.c index 877fb0ed7d3d..a7010a0664c7 100644 --- a/drivers/net/usb/cdc_mbim.c +++ b/drivers/net/usb/cdc_mbim.c @@ -635,6 +635,11 @@ static const struct usb_device_id mbim_devs[] = { .driver_info = (unsigned long)&cdc_mbim_info, }, + /* MeiG Smart SRM821 ZLP conformance */ + { USB_DEVICE_AND_INTERFACE_INFO(0x2dee, 0x4d53, USB_CLASS_COMM, USB_CDC_SUBCLASS_MBIM, USB_CDC_PROTO_NONE), + .driver_info = (unsigned long)&cdc_mbim_info, + }, + /* Some Huawei devices, ME906s-158 (12d1:15c1) and E3372 * (12d1:157d), are known to fail unless the NDP is placed * after the IP packets. Applying the quirk to all Huawei From a940003f44e7e441c228151dd212642152700ec8 Mon Sep 17 00:00:00 2001 From: Aleksei Sviridkin Date: Mon, 21 Sep 2026 01:20:44 +0300 Subject: [PATCH 525/694] net: phylink: record the PHY only once bringup cannot fail phylink_bringup_phy() stores the PHY in pl->phydev before its last fallible step: on a MAC whose phylink ops implement LPI, phy_eee_rx_clock_stop() can fail with a real MDIO error. The callers unwind with phy_detach(), which knows nothing about pl->phydev, so a pointer to a PHY that is no longer attached outlives the failed connect. What that costs depends on how the caller got here. phylink_connect_phy() goes through phylink_attach_phy(), which refuses to attach while pl->phydev is set, turning a transient MDIO error into a permanent -EBUSY. The SFP path is worse than that: sfp_sm_probe_phy() answers the failure with phy_device_remove() and phy_device_free(), and it assigns sfp->mod_phy only past that error return, so nothing clears pl->phydev and it is left pointing at a freed phy_device that phylink_resolve() and the ethtool helpers go on reading. phylink_fwnode_phy_connect() has no such check, so a later connect overwrites the stale pointer and hides the problem. A disconnect does not: phylink_disconnect_phy() hands that pointer to phy_disconnect(), and the second phy_detach() on the same PHY drops references the first one already released. Found while making a DSA port survive a PHY whose driver arrives after the switch probes: keeping the port across a failed connect and retrying is what makes this window reachable. Publish the pointer after the last call that can fail instead of unwinding it afterwards. Nothing between the two points reads pl->phydev, and the registration that follows cannot fail: phy_request_interrupt() falls back to polling on its own. The PHY-side state keeps the order it had, so no MDIO operation moves relative to another. Fixes: 03abf2a7c654 ("net: phylink: add EEE management") Signed-off-by: Aleksei Sviridkin Link: https://patch.msgid.link/20260920222044.1752860-1-f@lex.la Signed-off-by: Jakub Kicinski --- drivers/net/phy/phylink.c | 20 +++++++++++++++++--- 1 file changed, 17 insertions(+), 3 deletions(-) diff --git a/drivers/net/phy/phylink.c b/drivers/net/phy/phylink.c index a1458da8111b..1bbcf46c8356 100644 --- a/drivers/net/phy/phylink.c +++ b/drivers/net/phy/phylink.c @@ -2129,7 +2129,6 @@ static int phylink_bringup_phy(struct phylink *pl, struct phy_device *phy, mutex_lock(&pl->phydev_mutex); mutex_lock(&phy->lock); mutex_lock(&pl->state_mutex); - pl->phydev = phy; pl->phy_state.interface = interface; pl->phy_state.pause = MLO_PAUSE_NONE; pl->phy_state.speed = SPEED_UNKNOWN; @@ -2196,10 +2195,25 @@ static int phylink_bringup_phy(struct phylink *pl, struct phy_device *phy, ret = 0; } - if (ret == 0 && phy_interrupt_is_valid(phy)) + if (ret) + return ret; + + /* Nothing below can fail, so the PHY can be recorded now. Doing it + * here rather than above keeps a failed bringup from leaving + * pl->phydev pointing at a PHY the caller is about to detach. + */ + mutex_lock(&pl->phydev_mutex); + mutex_lock(&phy->lock); + mutex_lock(&pl->state_mutex); + pl->phydev = phy; + mutex_unlock(&pl->state_mutex); + mutex_unlock(&phy->lock); + mutex_unlock(&pl->phydev_mutex); + + if (phy_interrupt_is_valid(phy)) phy_request_interrupt(phy); - return ret; + return 0; } static int phylink_attach_phy(struct phylink *pl, struct phy_device *phy, From 3173cba1170131816972ed2b6185cb970ba8b747 Mon Sep 17 00:00:00 2001 From: Jiawen Wu Date: Mon, 21 Sep 2026 15:15:49 +0800 Subject: [PATCH 526/694] net: libwx: fix races in Tx timestamp handling wx->ptp_tx_skb is shared between the Tx path, the PTP auxiliary worker and the timestamp cleanup paths. The WX_STATE_PTP_TX_IN_PROGRESS bit prevents multiple Tx paths from submitting timestamp requests, but does not serialize the worker against cleanup. As a result, wx_ptp_clear_tx_timestamp() can free an skb after wx_ptp_tx_hwtstamp_work() has obtained its pointer. The worker may then pass the freed skb to skb_tstamp_tx() and release the same reference again. The cleanup path may also clear the in-progress bit while the worker is still processing the old skb. This allows the Tx path to publish a new skb which the worker can subsequently overwrite with NULL, leaking its reference. Add a dedicated spinlock to protect publication and consumption of the Tx timestamp skb. Detach the skb and clear the in-progress bit while holding the lock, then deliver the timestamp and release the skb after dropping it. Use the same locked cleanup in the quiesce path, but keep the detach there free of register accesses: quiesce runs during PCIe error recovery, where MMIO is not reliable, and it deliberately did not touch the device before. The lock is taken with interrupts disabled, because netpoll can call ndo_start_xmit() with hard interrupts already off. When handling a Tx DMA mapping failure, keep the transmit path reference until after comparing the skb under the lock. This prevents skb address reuse from making the error path mistake a newer timestamp request for the failed one. Fixes: 06e75161b9d4 ("net: wangxun: Add support for PTP clock") Reported-by: Sashiko Link: https://netdev-ai.bots.linux.dev/sashiko/#/patchset/6C7EC12D69217315%2B20260818074721.45536-1-jiawenwu%40trustnetic.com Signed-off-by: Jiawen Wu Link: https://patch.msgid.link/77431AF9A0E369F3+20260921071549.1141804-1-jiawenwu@trustnetic.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/wangxun/libwx/wx_hw.c | 1 + drivers/net/ethernet/wangxun/libwx/wx_lib.c | 47 ++++-- drivers/net/ethernet/wangxun/libwx/wx_ptp.c | 144 ++++++++++++------- drivers/net/ethernet/wangxun/libwx/wx_type.h | 2 + 4 files changed, 133 insertions(+), 61 deletions(-) diff --git a/drivers/net/ethernet/wangxun/libwx/wx_hw.c b/drivers/net/ethernet/wangxun/libwx/wx_hw.c index 122c4952d203..113552586be7 100644 --- a/drivers/net/ethernet/wangxun/libwx/wx_hw.c +++ b/drivers/net/ethernet/wangxun/libwx/wx_hw.c @@ -2518,6 +2518,7 @@ int wx_sw_init(struct wx *wx) } spin_lock_init(&wx->hw_stats_lock); + spin_lock_init(&wx->ptp_tx_lock); mutex_init(&wx->reset_lock); bitmap_zero(wx->state, WX_STATE_NBITS); bitmap_zero(wx->flags, WX_PF_FLAGS_NBITS); diff --git a/drivers/net/ethernet/wangxun/libwx/wx_lib.c b/drivers/net/ethernet/wangxun/libwx/wx_lib.c index ed5aad7857bd..dcbf5811046e 100644 --- a/drivers/net/ethernet/wangxun/libwx/wx_lib.c +++ b/drivers/net/ethernet/wangxun/libwx/wx_lib.c @@ -1200,9 +1200,11 @@ static int wx_tx_map(struct wx_ring *tx_ring, i--; } - dev_kfree_skb_any(first->skb); - first->skb = NULL; - + /* first->skb is released by the caller, which keeps a reference on it + * until the PTP cleanup has compared it against wx->ptp_tx_skb. That + * prevents the address from being reused by a newer request while the + * comparison is pending. + */ tx_ring->next_to_use = i; return -ENOMEM; @@ -1649,9 +1651,11 @@ static netdev_tx_t wx_xmit_frame_ring(struct sk_buff *skb, if (unlikely(skb_shinfo(skb)->tx_flags & SKBTX_HW_TSTAMP) && wx->ptp_clock) { + unsigned long flags; + + spin_lock_irqsave(&wx->ptp_tx_lock, flags); if (wx->tstamp_config.tx_type == HWTSTAMP_TX_ON && - !test_and_set_bit_lock(WX_STATE_PTP_TX_IN_PROGRESS, - wx->state)) { + !test_and_set_bit(WX_STATE_PTP_TX_IN_PROGRESS, wx->state)) { skb_shinfo(skb)->tx_flags |= SKBTX_IN_PROGRESS; tx_flags |= WX_TX_FLAGS_TSTAMP; wx->ptp_tx_skb = skb_get(skb); @@ -1659,6 +1663,7 @@ static netdev_tx_t wx_xmit_frame_ring(struct sk_buff *skb, } else { wx->tx_hwtstamp_skipped++; } + spin_unlock_irqrestore(&wx->ptp_tx_lock, flags); } /* record initial flags and protocol */ @@ -1677,19 +1682,35 @@ static netdev_tx_t wx_xmit_frame_ring(struct sk_buff *skb, wx->atr(tx_ring, first, ptype); if (wx_tx_map(tx_ring, first, hdr_len)) - goto cleanup_tx_tstamp; + goto out_drop; return NETDEV_TX_OK; out_drop: + /* The frame never reached the hardware, so no timestamp will ever be + * reported for it and the request has to be cancelled. The slot is + * shared, though: wx_ptp_clear_tx_timestamp() or wx_ptp_tx_hang() may + * have dropped our request already, and a transmit on another queue + * can have claimed the slot since. Only cancel it while it is still + * ours, otherwise we would free somebody else's skb and release their + * in-progress bit. + */ + if (unlikely(tx_flags & WX_TX_FLAGS_TSTAMP)) { + struct sk_buff *ptp_tx_skb = NULL; + unsigned long flags; + + spin_lock_irqsave(&wx->ptp_tx_lock, flags); + if (wx->ptp_tx_skb == skb) { + ptp_tx_skb = wx->ptp_tx_skb; + wx->ptp_tx_skb = NULL; + clear_bit(WX_STATE_PTP_TX_IN_PROGRESS, wx->state); + wx->tx_hwtstamp_errors++; + } + spin_unlock_irqrestore(&wx->ptp_tx_lock, flags); + + dev_kfree_skb_any(ptp_tx_skb); + } dev_kfree_skb_any(first->skb); first->skb = NULL; -cleanup_tx_tstamp: - if (unlikely(tx_flags & WX_TX_FLAGS_TSTAMP)) { - dev_kfree_skb_any(wx->ptp_tx_skb); - wx->ptp_tx_skb = NULL; - wx->tx_hwtstamp_errors++; - clear_bit_unlock(WX_STATE_PTP_TX_IN_PROGRESS, wx->state); - } return NETDEV_TX_OK; } diff --git a/drivers/net/ethernet/wangxun/libwx/wx_ptp.c b/drivers/net/ethernet/wangxun/libwx/wx_ptp.c index 4708e7f3958f..65b8937f6e94 100644 --- a/drivers/net/ethernet/wangxun/libwx/wx_ptp.c +++ b/drivers/net/ethernet/wangxun/libwx/wx_ptp.c @@ -129,6 +129,34 @@ static int wx_ptp_settime64(struct ptp_clock_info *ptp, return 0; } +/** + * __wx_ptp_detach_tx_skb - detach the skb tracking the Tx timestamp request + * @wx: the private board structure + * + * Detach the skb of the outstanding request and release the in-progress bit, + * so that a new request can be submitted. + * + * This performs no register access. Callers that need a timestamp the hardware + * may have left latched must unlatch it themselves, while the device is known + * to be alive. wx_ptp_quiesce() runs during PCIe error recovery, where MMIO is + * not reliable, and therefore deliberately skips the unlatch. + * + * Context: Expects wx->ptp_tx_lock to be held by the caller. + * Return: the detached skb, or NULL if no request was outstanding. The caller + * owns the returned reference and must release it once the lock is dropped. + */ +static struct sk_buff *__wx_ptp_detach_tx_skb(struct wx *wx) +{ + struct sk_buff *skb = wx->ptp_tx_skb; + + lockdep_assert_held(&wx->ptp_tx_lock); + + wx->ptp_tx_skb = NULL; + clear_bit(WX_STATE_PTP_TX_IN_PROGRESS, wx->state); + + return skb; +} + /** * wx_ptp_clear_tx_timestamp - utility function to clear Tx timestamp state * @wx: the private board structure @@ -139,12 +167,16 @@ static int wx_ptp_settime64(struct ptp_clock_info *ptp, */ static void wx_ptp_clear_tx_timestamp(struct wx *wx) { + struct sk_buff *skb; + unsigned long flags; + + spin_lock_irqsave(&wx->ptp_tx_lock, flags); + /* Unlatch a timestamp the hardware may have left pending. */ rd32ptp(wx, WX_TSC_1588_STMPH); - if (wx->ptp_tx_skb) { - dev_kfree_skb_any(wx->ptp_tx_skb); - wx->ptp_tx_skb = NULL; - } - clear_bit_unlock(WX_STATE_PTP_TX_IN_PROGRESS, wx->state); + skb = __wx_ptp_detach_tx_skb(wx); + spin_unlock_irqrestore(&wx->ptp_tx_lock, flags); + + dev_kfree_skb_any(skb); } /** @@ -175,49 +207,54 @@ static void wx_ptp_convert_to_hwtstamp(struct wx *wx, } /** - * wx_ptp_tx_hwtstamp - utility function which checks for TX time stamp + * wx_ptp_tx_hwtstamp_work - check for a pending Tx time stamp * @wx: the private board struct * - * if the timestamp is valid, we convert it into the timecounter ns - * value, then store that result into the shhwtstamps structure which - * is passed up the network stack + * If a Tx timestamp request is outstanding and the hardware has latched a + * valid value, we convert it into the timecounter ns value, then store that + * result into the shhwtstamps structure which is passed up the network stack. + * + * Return: 0 when there is nothing left to poll for, -1 when the timestamp is + * not available yet and the caller should poll again. */ -static void wx_ptp_tx_hwtstamp(struct wx *wx) -{ - struct skb_shared_hwtstamps shhwtstamps; - struct sk_buff *skb = wx->ptp_tx_skb; - u64 regval = 0; - - regval |= (u64)rd32ptp(wx, WX_TSC_1588_STMPL); - regval |= (u64)rd32ptp(wx, WX_TSC_1588_STMPH) << 32; - - wx_ptp_convert_to_hwtstamp(wx, &shhwtstamps, regval); - - wx->ptp_tx_skb = NULL; - clear_bit_unlock(WX_STATE_PTP_TX_IN_PROGRESS, wx->state); - skb_tstamp_tx(skb, &shhwtstamps); - dev_kfree_skb_any(skb); - wx->tx_hwtstamp_pkts++; -} - static int wx_ptp_tx_hwtstamp_work(struct wx *wx) { + struct skb_shared_hwtstamps shhwtstamps; + unsigned long flags; + struct sk_buff *skb; u32 tsynctxctl; + u64 regval = 0; + + spin_lock_irqsave(&wx->ptp_tx_lock, flags); /* we have to have a valid skb to poll for a timestamp */ if (!wx->ptp_tx_skb) { - wx_ptp_clear_tx_timestamp(wx); + rd32ptp(wx, WX_TSC_1588_STMPH); + __wx_ptp_detach_tx_skb(wx); + spin_unlock_irqrestore(&wx->ptp_tx_lock, flags); return 0; } /* stop polling once we have a valid timestamp */ tsynctxctl = rd32ptp(wx, WX_TSC_1588_CTL); - if (tsynctxctl & WX_TSC_1588_CTL_VALID) { - wx_ptp_tx_hwtstamp(wx); - return 0; + if (!(tsynctxctl & WX_TSC_1588_CTL_VALID)) { + spin_unlock_irqrestore(&wx->ptp_tx_lock, flags); + return -1; } - return -1; + regval |= (u64)rd32ptp(wx, WX_TSC_1588_STMPL); + regval |= (u64)rd32ptp(wx, WX_TSC_1588_STMPH) << 32; + skb = wx->ptp_tx_skb; + wx->ptp_tx_skb = NULL; + clear_bit(WX_STATE_PTP_TX_IN_PROGRESS, wx->state); + spin_unlock_irqrestore(&wx->ptp_tx_lock, flags); + + wx_ptp_convert_to_hwtstamp(wx, &shhwtstamps, regval); + skb_tstamp_tx(skb, &shhwtstamps); + dev_kfree_skb_any(skb); + wx->tx_hwtstamp_pkts++; + + return 0; } /** @@ -296,24 +333,29 @@ static void wx_ptp_rx_hang(struct wx *wx) */ static void wx_ptp_tx_hang(struct wx *wx) { - bool timeout = time_is_before_jiffies(wx->ptp_tx_start + - WX_PTP_TX_TIMEOUT); + struct sk_buff *skb = NULL; + unsigned long flags; - if (!wx->ptp_tx_skb) - return; - - if (!test_bit(WX_STATE_PTP_TX_IN_PROGRESS, wx->state)) - return; + spin_lock_irqsave(&wx->ptp_tx_lock, flags); /* If we haven't received a timestamp within the timeout, it is * reasonable to assume that it will never occur, so we can unlock the * timestamp bit when this occurs. */ - if (timeout) { - wx_ptp_clear_tx_timestamp(wx); - wx->tx_hwtstamp_timeouts++; - dev_warn(&wx->pdev->dev, "clearing Tx timestamp hang\n"); + if (wx->ptp_tx_skb && + test_bit(WX_STATE_PTP_TX_IN_PROGRESS, wx->state) && + time_is_before_jiffies(wx->ptp_tx_start + WX_PTP_TX_TIMEOUT)) { + rd32ptp(wx, WX_TSC_1588_STMPH); + skb = __wx_ptp_detach_tx_skb(wx); } + spin_unlock_irqrestore(&wx->ptp_tx_lock, flags); + + if (!skb) + return; + + dev_kfree_skb_any(skb); + wx->tx_hwtstamp_timeouts++; + dev_warn(&wx->pdev->dev, "clearing Tx timestamp hang\n"); } static long wx_ptp_do_aux_work(struct ptp_clock_info *ptp) @@ -841,6 +883,9 @@ EXPORT_SYMBOL(wx_ptp_stop); void wx_ptp_quiesce(struct wx *wx) { + struct sk_buff *skb; + unsigned long flags; + if (!test_and_clear_bit(WX_STATE_PTP_RUNNING, wx->state)) return; @@ -849,11 +894,14 @@ void wx_ptp_quiesce(struct wx *wx) if (wx->ptp_clock) ptp_cancel_worker_sync(wx->ptp_clock); - if (wx->ptp_tx_skb) { - dev_kfree_skb_any(wx->ptp_tx_skb); - wx->ptp_tx_skb = NULL; - } - clear_bit_unlock(WX_STATE_PTP_TX_IN_PROGRESS, wx->state); + /* Drop a pending Tx timestamp request. Do not touch the registers + * here: quiesce runs during PCIe error recovery, where the device may + * already be gone and MMIO is not reliable. + */ + spin_lock_irqsave(&wx->ptp_tx_lock, flags); + skb = __wx_ptp_detach_tx_skb(wx); + spin_unlock_irqrestore(&wx->ptp_tx_lock, flags); + dev_kfree_skb_any(skb); if (wx->ptp_clock) { ptp_clock_unregister(wx->ptp_clock); diff --git a/drivers/net/ethernet/wangxun/libwx/wx_type.h b/drivers/net/ethernet/wangxun/libwx/wx_type.h index 9454e90258d8..afd980dbb793 100644 --- a/drivers/net/ethernet/wangxun/libwx/wx_type.h +++ b/drivers/net/ethernet/wangxun/libwx/wx_type.h @@ -1430,6 +1430,8 @@ struct wx { unsigned long last_overflow_check; unsigned long last_rx_ptp_check; unsigned long ptp_tx_start; + /* protects ptp_tx_skb, ptp_tx_start and the in-progress state bit */ + spinlock_t ptp_tx_lock; seqlock_t hw_tc_lock; /* seqlock for ptp */ struct cyclecounter hw_cc; struct timecounter hw_tc; From 26b2bd70d22457556e2fa01cbf1192cb1a94d619 Mon Sep 17 00:00:00 2001 From: Ilya Maximets Date: Mon, 21 Sep 2026 16:55:43 +0200 Subject: [PATCH 527/694] net: openvswitch: conntrack: avoid modifying shared unconfirmed ct entry In a case where skb with an unconfirmed ct entry gets cloned, we may end up committing both but with different sets of extensions. The series of events: 1. The first clone wants to commit and runs the helpers wiring up the extension pointer into the expectation list. 2. Then it looses the confirmation keeping the entry unconfirmed. 3. Second clone now wants to commit labels and adds the new extension for that breaking the pointer in the expectation list causing UAF on the destruction path later. While this is possible to trigger, there should be no practical network pipeline where committing both clones without modifications into the same zone is needed. So, let's just reset the entry in case for some reason we got an skb with a shared one during commit. This doesn't affect any known use cases, but avoids any potential problems with sharing and modification of the unconfirmed ct entry. The fixes tag points to the introduction of helpers, since that's the main UAF trigger for the sharing. Fixes: cae3a2627520 ("openvswitch: Allow attaching helpers to ct action") Cc: stable@vger.kernel.org Reported-by: Axel Mierczuk Signed-off-by: Ilya Maximets Reviewed-by: Aaron Conole Link: https://patch.msgid.link/20260921145655.3167436-2-i.maximets@ovn.org Signed-off-by: Jakub Kicinski --- include/net/netfilter/nf_conntrack.h | 5 +++++ net/openvswitch/conntrack.c | 12 ++++++++++++ 2 files changed, 17 insertions(+) diff --git a/include/net/netfilter/nf_conntrack.h b/include/net/netfilter/nf_conntrack.h index bc42dd0e10e6..c39425e54d87 100644 --- a/include/net/netfilter/nf_conntrack.h +++ b/include/net/netfilter/nf_conntrack.h @@ -185,6 +185,11 @@ static inline void nf_ct_put(struct nf_conn *ct) nf_ct_destroy(&ct->ct_general); } +static inline bool nf_ct_shared(const struct nf_conn *ct) +{ + return refcount_read(&ct->ct_general.use) > 1; +} + /* load module; enable/disable conntrack in this namespace */ int nf_ct_netns_get(struct net *net, u8 nfproto); void nf_ct_netns_put(struct net *net, u8 nfproto); diff --git a/net/openvswitch/conntrack.c b/net/openvswitch/conntrack.c index 0f433688e17b..a733029c28dd 100644 --- a/net/openvswitch/conntrack.c +++ b/net/openvswitch/conntrack.c @@ -734,6 +734,18 @@ static int __ovs_ct_lookup(struct net *net, struct sw_flow_key *key, enum ip_conntrack_info ctinfo; struct nf_conn *ct; + /* If the ct entry is not confirmed and shared with some other skb, + * e.g., a cloned one, we can't just modify it with the commit as we + * must not modify the extension set. Reset. + */ + if (cached && info->commit) { + ct = nf_ct_get(skb, &ctinfo); + if (ct && !nf_ct_is_confirmed(ct) && nf_ct_shared(ct)) { + nf_reset_ct(skb); + cached = false; + } + } + if (!cached) { struct nf_hook_state state = { .hook = NF_INET_PRE_ROUTING, From 5e6c14dd42a1c1fe938e573dc6c9098145b2b0c4 Mon Sep 17 00:00:00 2001 From: Ilya Maximets Date: Mon, 21 Sep 2026 16:55:44 +0200 Subject: [PATCH 528/694] net: openvswitch: conntrack: remove 'add_helper' dead code This variable can only become 'true' when the connection is not confirmed, but it is only checked when it is confirmed. So, it can be treated as being always false and just removed. Fixes: 3c1860543fcc ("openvswitch: add nf_ct_is_confirmed check before assigning the helper") Cc: stable@vger.kernel.org Signed-off-by: Ilya Maximets Reviewed-by: Aaron Conole Link: https://patch.msgid.link/20260921145655.3167436-3-i.maximets@ovn.org Signed-off-by: Jakub Kicinski --- net/openvswitch/conntrack.c | 10 ++-------- 1 file changed, 2 insertions(+), 8 deletions(-) diff --git a/net/openvswitch/conntrack.c b/net/openvswitch/conntrack.c index a733029c28dd..c20f096eef40 100644 --- a/net/openvswitch/conntrack.c +++ b/net/openvswitch/conntrack.c @@ -778,8 +778,6 @@ static int __ovs_ct_lookup(struct net *net, struct sw_flow_key *key, ct = nf_ct_get(skb, &ctinfo); if (ct) { - bool add_helper = false; - /* Packets starting a new connection must be NATted before the * helper, so that the helper knows about the NAT. We enforce * this by delaying both NAT and helper calls for unconfirmed @@ -811,7 +809,6 @@ static int __ovs_ct_lookup(struct net *net, struct sw_flow_key *key, GFP_ATOMIC); if (err) return err; - add_helper = true; /* helper installed, add seqadj if NAT is required */ if (info->nat && !nfct_seqadj(ct)) { @@ -821,13 +818,10 @@ static int __ovs_ct_lookup(struct net *net, struct sw_flow_key *key, } /* Call the helper only if: - * - nf_conntrack_in() was executed above ("!cached") or a - * helper was just attached ("add_helper") for a confirmed - * connection, or + * - nf_conntrack_in() was executed above ("!cached"), or * - When committing an unconfirmed connection. */ - if ((nf_ct_is_confirmed(ct) ? !cached || add_helper : - info->commit)) { + if ((nf_ct_is_confirmed(ct) ? !cached : info->commit)) { int err = nf_ct_helper(skb, ct, ctinfo, info->family); err = verdict_to_errno(err); From 1a4151e6be57b098b7a5ebfbde58585e83200cdc Mon Sep 17 00:00:00 2001 From: Ilya Maximets Date: Mon, 21 Sep 2026 16:55:45 +0200 Subject: [PATCH 529/694] net: openvswitch: conntrack: fix helper UAF due to extensions realloc While calling the helpers, a raw pointer to the extensions area is wired into expectations list: -> nf_ct_helper() -> helper->help() -> nf_ct_expect_related_report() -> nf_ct_expect_insert() -> hlist_add_head_rcu(&exp->lnode, &master_help->expectations) In case the connection is not confirmed yet, more extensions can be added afterwards with *_ext_add() calls reallocating the extension space and leaving the now invalid pointer in the expectations list that is later accessed while removing the expectation. Make sure that helpers are called at the end after all the other extensions are already added. Note that the helper rejection now leaves the mark and labels set, but that's not different from how the NAT was handled before or how the mark and the labels were handled on confirmation failure. And there are no atomicity guarantees provided by the API anyway. Fixes: cae3a2627520 ("openvswitch: Allow attaching helpers to ct action") Cc: stable@vger.kernel.org Reported-by: Axel Mierczuk Signed-off-by: Ilya Maximets Reviewed-by: Aaron Conole Link: https://patch.msgid.link/20260921145655.3167436-4-i.maximets@ovn.org Signed-off-by: Jakub Kicinski --- net/openvswitch/conntrack.c | 19 +++++++++++++++---- 1 file changed, 15 insertions(+), 4 deletions(-) diff --git a/net/openvswitch/conntrack.c b/net/openvswitch/conntrack.c index c20f096eef40..d3326edcabf7 100644 --- a/net/openvswitch/conntrack.c +++ b/net/openvswitch/conntrack.c @@ -817,11 +817,14 @@ static int __ovs_ct_lookup(struct net *net, struct sw_flow_key *key, } } - /* Call the helper only if: - * - nf_conntrack_in() was executed above ("!cached"), or - * - When committing an unconfirmed connection. + /* Call the helper only if nf_conntrack_in() was executed + * above ("!cached"). + * + * For unconfirmed connections it will be called later during + * commit as we need to have all the other extensions allocated + * before the call. */ - if ((nf_ct_is_confirmed(ct) ? !cached : info->commit)) { + if (nf_ct_is_confirmed(ct) && !cached) { int err = nf_ct_helper(skb, ct, ctinfo, info->family); err = verdict_to_errno(err); @@ -1025,6 +1028,14 @@ static int ovs_ct_commit(struct net *net, struct sw_flow_key *key, return err; nf_conn_act_ct_ext_add(skb, ct, ctinfo); + + /* Call the helpers now. We couldn't do this before as + * all the extensions must be allocated before the call. + */ + err = nf_ct_helper(skb, ct, ctinfo, info->family); + err = verdict_to_errno(err); + if (err) + return err; } else if (IS_ENABLED(CONFIG_NF_CONNTRACK_LABELS) && labels_nonzero(&info->labels.mask)) { err = ovs_ct_set_labels(ct, key, &info->labels.value, From f85009dfcd65e5969526b0db7a49b5413746e630 Mon Sep 17 00:00:00 2001 From: Ilya Maximets Date: Mon, 21 Sep 2026 16:55:46 +0200 Subject: [PATCH 530/694] net/sched: act_ct: avoid modifying shared unconfirmed ct entry In a case where skb with an unconfirmed ct entry gets cloned, we may end up processing both again but with different sets of extensions. The series of events: 1. The first clone wants to commit and runs the helpers wiring up the extension pointer into the expectation list. 2. Then it looses the confirmation keeping the entry unconfirmed. 3. Second clone now wants to commit labels or run NAT and adds the new extension for that breaking the pointer in the expectation list causing UAF on the destruction path later. While this is possible to trigger, there should be no practical network pipeline where we need to process both clones without modifications in the same zone. So, let's just reset the entry in case for some reason we got an skb with a shared one. This doesn't affect any known use cases, but avoids any potential problems with sharing and modification of the unconfirmed ct entry. Unlike openvswitch module, act_ct allows for NAT without commit. Changing that would be a uAPI break. So, act_ct needs to reset on NAT regardless of the commit flag to avoid reallocation of the extension space. This, however, doesn't really change the picture for sensible networking cases as there should be no need to run the same packet twice (before and after the clone) through conntrack without packet header or zone changes and without commit. The fixes tag points to the introduction of helpers, since that's the main UAF trigger for the sharing. Fixes: a21b06e73191 ("net: sched: add helper support in act_ct") Cc: stable@vger.kernel.org Reported-by: Axel Mierczuk Signed-off-by: Ilya Maximets Reviewed-by: Aaron Conole Reviewed-by: Xin Long Reviewed-by: Jamal Hadi Salim Link: https://patch.msgid.link/20260921145655.3167436-5-i.maximets@ovn.org Signed-off-by: Jakub Kicinski --- net/sched/act_ct.c | 18 ++++++++++++++++-- 1 file changed, 16 insertions(+), 2 deletions(-) diff --git a/net/sched/act_ct.c b/net/sched/act_ct.c index 55f3521edb4c..e72143d36b11 100644 --- a/net/sched/act_ct.c +++ b/net/sched/act_ct.c @@ -979,11 +979,11 @@ TC_INDIRECT_SCOPE int tcf_ct_act(struct sk_buff *skb, const struct tc_action *a, struct tcf_result *res) { struct net *net = dev_net(skb->dev); + bool cached, commit, clear, nat; enum ip_conntrack_info ctinfo; struct tcf_ct *c = to_ct(a); struct nf_conn *tmpl = NULL; struct nf_hook_state state; - bool cached, commit, clear; int nh_ofs, err, retval; struct tcf_ct_params *p; bool add_helper = false; @@ -998,6 +998,7 @@ TC_INDIRECT_SCOPE int tcf_ct_act(struct sk_buff *skb, const struct tc_action *a, retval = p->action; commit = p->ct_action & TCA_CT_ACT_COMMIT; clear = p->ct_action & TCA_CT_ACT_CLEAR; + nat = p->ct_action & TCA_CT_ACT_NAT; tmpl = p->tmpl; tcf_lastuse_update(&c->tcf_tm); @@ -1046,6 +1047,19 @@ TC_INDIRECT_SCOPE int tcf_ct_act(struct sk_buff *skb, const struct tc_action *a, * different zone. */ cached = tcf_ct_skb_nfct_cached(net, skb, p); + + /* If the ct entry is not confirmed and shared with some other skb, + * e.g., a cloned one, we can't just modify it with a commit or nat + * as we must not modify the extension set. Reset. + */ + if (cached && (commit || nat)) { + ct = nf_ct_get(skb, &ctinfo); + if (ct && !nf_ct_is_confirmed(ct) && nf_ct_shared(ct)) { + nf_reset_ct(skb); + cached = false; + } + } + if (!cached) { if (tcf_ct_flow_table_lookup(p, skb, family)) { skip_add = true; @@ -1083,7 +1097,7 @@ TC_INDIRECT_SCOPE int tcf_ct_act(struct sk_buff *skb, const struct tc_action *a, if (err) goto drop; add_helper = true; - if (p->ct_action & TCA_CT_ACT_NAT && !nfct_seqadj(ct)) { + if (nat && !nfct_seqadj(ct)) { if (!nfct_seqadj_ext_add(ct)) goto drop; } From 00df72e39f306e2f7adb68528a5c92109da6a0a9 Mon Sep 17 00:00:00 2001 From: Ilya Maximets Date: Mon, 21 Sep 2026 16:55:47 +0200 Subject: [PATCH 531/694] net/sched: act_ct: remove 'add_helper' dead code This variable can only become 'true' when the connection is not confirmed, but it is only checked when it is confirmed. So, it can be treated as being always false and just removed. Fixes: a21b06e73191 ("net: sched: add helper support in act_ct") Cc: stable@vger.kernel.org Signed-off-by: Ilya Maximets Reviewed-by: Aaron Conole Reviewed-by: Xin Long Reviewed-by: Jamal Hadi Salim Link: https://patch.msgid.link/20260921145655.3167436-6-i.maximets@ovn.org Signed-off-by: Jakub Kicinski --- net/sched/act_ct.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/net/sched/act_ct.c b/net/sched/act_ct.c index e72143d36b11..f62051ec9d57 100644 --- a/net/sched/act_ct.c +++ b/net/sched/act_ct.c @@ -986,7 +986,6 @@ TC_INDIRECT_SCOPE int tcf_ct_act(struct sk_buff *skb, const struct tc_action *a, struct nf_hook_state state; int nh_ofs, err, retval; struct tcf_ct_params *p; - bool add_helper = false; bool skb_is_ours = false; bool skip_add = false; bool defrag = false; @@ -1096,14 +1095,14 @@ TC_INDIRECT_SCOPE int tcf_ct_act(struct sk_buff *skb, const struct tc_action *a, err = __nf_ct_try_assign_helper(ct, p->tmpl, GFP_ATOMIC); if (err) goto drop; - add_helper = true; + if (nat && !nfct_seqadj(ct)) { if (!nfct_seqadj_ext_add(ct)) goto drop; } } - if (nf_ct_is_confirmed(ct) ? ((!cached && !skip_add) || add_helper) : commit) { + if (nf_ct_is_confirmed(ct) ? (!cached && !skip_add) : commit) { err = nf_ct_helper(skb, ct, ctinfo, family); if (err != NF_ACCEPT) goto nf_error; From dad19b59da050cb60d3f7023dac2a042a84bf0bd Mon Sep 17 00:00:00 2001 From: Ilya Maximets Date: Mon, 21 Sep 2026 16:55:48 +0200 Subject: [PATCH 532/694] net/sched: act_ct: fix helper UAF due to extensions realloc While calling the helpers, a raw pointer to the extensions area is wired into expectations list: -> nf_ct_helper() -> helper->help() -> nf_ct_expect_related_report() -> nf_ct_expect_insert() -> hlist_add_head_rcu(&exp->lnode, &master_help->expectations) In case the connection is not confirmed yet, more extensions can be added afterwards with *_ext_add() calls reallocating the extension space and leaving the now invalid pointer in the expectations list that is later accessed while removing the expectation. Make sure that helpers are called at the end after all the other extensions are already added. Note that the helper rejection now leaves the mark and labels set, but that's not different from how the NAT was handled before or how the mark and the labels were handled on confirmation failure. And there are no atomicity guarantees provided by the API anyway. Fixes: a21b06e73191 ("net: sched: add helper support in act_ct") Cc: stable@vger.kernel.org Reported-by: Axel Mierczuk Signed-off-by: Ilya Maximets Reviewed-by: Xin Long Reviewed-by: Jamal Hadi Salim Reviewed-by: Aaron Conole Link: https://patch.msgid.link/20260921145655.3167436-7-i.maximets@ovn.org Signed-off-by: Jakub Kicinski --- net/sched/act_ct.c | 18 ++++++++++++------ 1 file changed, 12 insertions(+), 6 deletions(-) diff --git a/net/sched/act_ct.c b/net/sched/act_ct.c index f62051ec9d57..411e3dd92d07 100644 --- a/net/sched/act_ct.c +++ b/net/sched/act_ct.c @@ -1102,6 +1102,18 @@ TC_INDIRECT_SCOPE int tcf_ct_act(struct sk_buff *skb, const struct tc_action *a, } } + if (commit) { + tcf_ct_act_set_mark(ct, p->mark, p->mark_mask); + tcf_ct_act_set_labels(ct, p->labels, p->labels_mask); + + if (!nf_ct_is_confirmed(ct)) + nf_conn_act_ct_ext_add(skb, ct, ctinfo); + } + + /* Run helpers for the connection if nf_conntrack_in() was executed + * or if we're about to commit. This has to be done after all the + * extensions are already added. + */ if (nf_ct_is_confirmed(ct) ? (!cached && !skip_add) : commit) { err = nf_ct_helper(skb, ct, ctinfo, family); if (err != NF_ACCEPT) @@ -1109,12 +1121,6 @@ TC_INDIRECT_SCOPE int tcf_ct_act(struct sk_buff *skb, const struct tc_action *a, } if (commit) { - tcf_ct_act_set_mark(ct, p->mark, p->mark_mask); - tcf_ct_act_set_labels(ct, p->labels, p->labels_mask); - - if (!nf_ct_is_confirmed(ct)) - nf_conn_act_ct_ext_add(skb, ct, ctinfo); - /* This will take care of sending queued events * even if the connection is already confirmed. */ From 0958ea4355e2e9220ad4e13da3b7d94f365ed34f Mon Sep 17 00:00:00 2001 From: Guangshuo Li Date: Mon, 21 Sep 2026 23:42:01 +0800 Subject: [PATCH 533/694] net: ena: fix PHC cleanup on probe failure ena_probe() initializes the PHC as part of ena_device_init(), but the probe failure path does not destroy it before freeing the PHC private data. The normal removal path calls ena_phc_destroy() through ena_destroy_device() before ena_phc_free(). However, if probe fails after ena_device_init() succeeds, the error path reaches ena_phc_free() without unregistering the PTP clock or destroying the device PHC resources. Call ena_phc_destroy() in the probe error path before freeing the PHC private data. This issue was found by manual code inspection. Cc: stable@vger.kernel.org tags and describe this as a consistency cleanup Fixes: e0ea34158ee8 ("net: ena: Add PHC support in the ENA driver") Cc: stable@vger.kernel.org Signed-off-by: Guangshuo Li Cc: stable Link: https://patch.msgid.link/20260921154202.471662-2-lgs201920130244@gmail.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/amazon/ena/ena_netdev.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/net/ethernet/amazon/ena/ena_netdev.c b/drivers/net/ethernet/amazon/ena/ena_netdev.c index ea89619039d8..5f0864d16dd3 100644 --- a/drivers/net/ethernet/amazon/ena/ena_netdev.c +++ b/drivers/net/ethernet/amazon/ena/ena_netdev.c @@ -4122,6 +4122,7 @@ static int ena_probe(struct pci_dev *pdev, const struct pci_device_id *ent) err_device_destroy: ena_com_delete_host_info(ena_dev); ena_com_admin_destroy(ena_dev); + ena_phc_destroy(adapter); ena_devlink_destroy: ena_devlink_free(devlink); err_metrics_destroy: From 9476b4468862927297c94c440863cd8ed1e7cc83 Mon Sep 17 00:00:00 2001 From: Guangshuo Li Date: Mon, 21 Sep 2026 23:42:02 +0800 Subject: [PATCH 534/694] net: ena: fix MMIO read buffer leak on probe failure ena_device_init() initializes the MMIO read mechanism with ena_com_mmio_reg_read_request_init(), which allocates a coherent DMA buffer for MMIO read responses. The normal removal path releases this buffer through ena_com_mmio_reg_read_request_destroy(). However, if ena_probe() fails after ena_device_init() succeeds, the error path destroys the admin resources and eventually frees ena_dev without destroying the MMIO read request, leaving the coherent DMA buffer allocated. Call ena_com_mmio_reg_read_request_destroy() in the probe error path before releasing the remaining device resources. This issue was found by manual code inspection. Fixes: 1738cd3ed342 ("net: ena: Add a driver for Amazon Elastic Network Adapters (ENA)") Cc: stable@vger.kernel.org Signed-off-by: Guangshuo Li Link: https://patch.msgid.link/20260921154202.471662-3-lgs201920130244@gmail.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/amazon/ena/ena_netdev.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/net/ethernet/amazon/ena/ena_netdev.c b/drivers/net/ethernet/amazon/ena/ena_netdev.c index 5f0864d16dd3..7eb6456ed0d5 100644 --- a/drivers/net/ethernet/amazon/ena/ena_netdev.c +++ b/drivers/net/ethernet/amazon/ena/ena_netdev.c @@ -4123,6 +4123,7 @@ static int ena_probe(struct pci_dev *pdev, const struct pci_device_id *ent) ena_com_delete_host_info(ena_dev); ena_com_admin_destroy(ena_dev); ena_phc_destroy(adapter); + ena_com_mmio_reg_read_request_destroy(ena_dev); ena_devlink_destroy: ena_devlink_free(devlink); err_metrics_destroy: From 1a983a4e14c635c40354be110cd9a1a5c94e01e6 Mon Sep 17 00:00:00 2001 From: Sang-Hoon Choi Date: Tue, 22 Sep 2026 03:15:59 +0900 Subject: [PATCH 535/694] nfp: hold IPsec RX state under the XArray lock nfp_net_ipsec_rx() drops the XArray lock before taking a reference to the xfrm_state it found. The delete path can erase the entry and drop the last state reference in that interval. RX can then try to increment a zero refcount after the state has been queued for destruction. The driver queues firmware invalidation asynchronously; the delete path does not wait for the command to complete or drain pending RX processing. The XFRM garbage collector waits for an RCU grace period before freeing the state. That delays reclamation but does not make acquiring a reference from zero valid. Take the xfrm_state reference before releasing the XArray lock so xa_erase() cannot run between lookup and reference acquisition. Fixes: 57f273adbcd4 ("nfp: add framework to support ipsec offloading") Reported-by: Changyul Lee Signed-off-by: Sang-Hoon Choi Link: https://patch.msgid.link/179001455912.44752.17153022439349797877.idr-bug-92@gmail.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/netronome/nfp/crypto/ipsec.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/netronome/nfp/crypto/ipsec.c b/drivers/net/ethernet/netronome/nfp/crypto/ipsec.c index 9e7c285eaa6b..960d7513aa8d 100644 --- a/drivers/net/ethernet/netronome/nfp/crypto/ipsec.c +++ b/drivers/net/ethernet/netronome/nfp/crypto/ipsec.c @@ -625,11 +625,12 @@ int nfp_net_ipsec_rx(struct nfp_meta_parsed *meta, struct sk_buff *skb) xa_lock(&nn->xa_ipsec); x = xa_load(&nn->xa_ipsec, saidx); + if (x) + xfrm_state_hold(x); xa_unlock(&nn->xa_ipsec); if (!x) return -EINVAL; - xfrm_state_hold(x); sp->xvec[sp->len++] = x; sp->olen++; xo = xfrm_offload(skb); From ba3d1f480c7a3fba963e7867ad6cc557c197acbc Mon Sep 17 00:00:00 2001 From: Allison Henderson Date: Mon, 21 Sep 2026 14:50:27 -0700 Subject: [PATCH 536/694] net/rds: size a connection's path set by the transport it ends up with __rds_conn_create() computes npaths from the caller's transport before it decides whether a connection to one of the host's own addresses is to be handled by the loopback transport instead. That substitution is what an RDS/TCP socket sending to a local address gets, and after it the path init loop still runs for the TCP transport's RDS_MPATH_WORKERS paths and allocates an ordered workqueue for each, while rds_loop_conn_alloc() only ever provides transport data for path 0. rds_conn_destroy() sizes its teardown from c_trans, by then the loopback transport, so it visits path 0 only - and rds_conn_path_destroy() would skip the other paths anyway, since it returns before destroy_workqueue() for a path without transport data. kfree(c_path) then drops the last pointers to seven workqueues. That repeats for every such connection, on every netns teardown or module unload, and every distinct local destination address is a separate connection. Recompute npaths once the transport is final, so that creation and destruction agree on the set of paths. The c_path array stays sized for the caller's transport; the unused entries are freed with it. Fixes: 4716af3897e9 ("net/rds: Give each connection path its own workqueue") Signed-off-by: Allison Henderson Link: https://patch.msgid.link/20260921215027.174657-1-achender@kernel.org Signed-off-by: Jakub Kicinski --- net/rds/connection.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/net/rds/connection.c b/net/rds/connection.c index b6c4beb50eaf..c752a8623cfc 100644 --- a/net/rds/connection.c +++ b/net/rds/connection.c @@ -276,6 +276,12 @@ static struct rds_connection *__rds_conn_create(struct net *net, conn->c_trans = trans; + /* The transport may just have been swapped for loopback; size the + * set of paths - which is also what rds_conn_destroy() tears down + * again - by the transport the connection actually uses. + */ + npaths = (trans->t_mp_capable ? RDS_MPATH_WORKERS : 1); + init_waitqueue_head(&conn->c_hs_waitq); for (i = 0; i < npaths; i++) { __rds_conn_path_init(conn, &conn->c_path[i], From 61cb282fe97b3b0ba32ca09417a693162bf4ae3f Mon Sep 17 00:00:00 2001 From: Sidraya Jayagond Date: Tue, 22 Sep 2026 09:31:49 +0200 Subject: [PATCH 537/694] net/smc: fix UAF on lgr list traversal in smcr_port_err() smcr_port_err() traverses smc_lgr_list.list without holding smc_lgr_list.lock, allowing a concurrent smc_lgr_terminate_sched() to free an lgr while it is still being dereferenced. Hold smc_lgr_list.lock across the traversal. Update smc_ib_gid_check() to call smcr_port_err() after releasing the lock. Fixes: 541afa10c126 ("net/smc: add smcr_port_err() and smcr_link_down() processing") Reviewed-by: Mahanta Jambigi Signed-off-by: Sidraya Jayagond Reviewed-by: Dust Li Link: https://patch.msgid.link/20260922073149.474762-1-sidraya@linux.ibm.com Signed-off-by: Jakub Kicinski --- net/smc/smc_core.c | 2 ++ net/smc/smc_ib.c | 10 ++++++++-- 2 files changed, 10 insertions(+), 2 deletions(-) diff --git a/net/smc/smc_core.c b/net/smc/smc_core.c index 04aedd957543..9974149659c2 100644 --- a/net/smc/smc_core.c +++ b/net/smc/smc_core.c @@ -1849,6 +1849,7 @@ void smcr_port_err(struct smc_ib_device *smcibdev, u8 ibport) struct smc_link_group *lgr, *n; int i; + spin_lock_bh(&smc_lgr_list.lock); list_for_each_entry_safe(lgr, n, &smc_lgr_list.list, list) { if (strncmp(smcibdev->pnetid[ibport - 1], lgr->pnet_id, SMC_MAX_PNETID_LEN)) @@ -1863,6 +1864,7 @@ void smcr_port_err(struct smc_ib_device *smcibdev, u8 ibport) smcr_link_down_cond_sched(lnk); } } + spin_unlock_bh(&smc_lgr_list.lock); } static void smc_link_down_work(struct work_struct *work) diff --git a/net/smc/smc_ib.c b/net/smc/smc_ib.c index 9bb495707445..daaa8a72da90 100644 --- a/net/smc/smc_ib.c +++ b/net/smc/smc_ib.c @@ -333,6 +333,7 @@ static bool smc_ib_check_link_gid(u8 gid[SMC_GID_SIZE], bool smcrv2, static void smc_ib_gid_check(struct smc_ib_device *smcibdev, u8 ibport) { struct smc_link_group *lgr; + bool stale_gid = false; int i; spin_lock_bh(&smc_lgr_list.lock); @@ -348,11 +349,16 @@ static void smc_ib_gid_check(struct smc_ib_device *smcibdev, u8 ibport) continue; if (!smc_ib_check_link_gid(lgr->lnk[i].gid, lgr->smc_version == SMC_V2, - smcibdev, ibport)) - smcr_port_err(smcibdev, ibport); + smcibdev, ibport)) { + stale_gid = true; + goto out; + } } } +out: spin_unlock_bh(&smc_lgr_list.lock); + if (stale_gid) + smcr_port_err(smcibdev, ibport); } static int smc_ib_remember_port_attr(struct smc_ib_device *smcibdev, u8 ibport) From b94773dc4df7026a6f29b2c65e96e88d29cdb576 Mon Sep 17 00:00:00 2001 From: Alexander Sverdlin Date: Tue, 22 Sep 2026 09:52:46 +0200 Subject: [PATCH 538/694] net: phy: intel-xway: workaround 100BASE-TX Link-Up issue MaxLinear GSW12x/GSW14x Ethernet Switch Errata Sheet states: "An issue has been sporadically observed after device power-on on the first link-up attempt in 100BASE-TX mode resulting in either the link-up taking a long time, or failing to link-up altogether... Workaround: After power-on, enable Cable Diagnostic Mode for all ports and disable it..." Implement the proposed workaround unconditionally in the Intel XWAY driver (MaxLinear GSW1xx switches incorporate Intel XWAY PHYs) because the diagnostic bits have the same meaning even in older integral PHYs such as GPY111/PEF7071/PHY11G. So it's not clear how to distinguish the affected newer integrated PHYs, but the workaround should not hurt the older PHYs. Cc: stable@vger.kernel.org Fixes: 22335939ec90 ("net: dsa: add driver for MaxLinear GSW1xx switch family") Signed-off-by: Alexander Sverdlin Reviewed-by: Andrew Lunn Link: https://patch.msgid.link/20260922075251.23386-1-alexander.sverdlin@siemens.com Signed-off-by: Jakub Kicinski --- drivers/net/phy/intel-xway.c | 29 ++++++++++++++++++++++++++++- 1 file changed, 28 insertions(+), 1 deletion(-) diff --git a/drivers/net/phy/intel-xway.c b/drivers/net/phy/intel-xway.c index afbcec711744..3cee31bb931f 100644 --- a/drivers/net/phy/intel-xway.c +++ b/drivers/net/phy/intel-xway.c @@ -16,6 +16,11 @@ #define XWAY_MDIO_ISTAT 0x1A /* interrupt status */ #define XWAY_MDIO_LED 0x1B /* led control */ +#define XWAY_MDIO_GCTRL_TM_MASK GENMASK(15, 13) +#define XWAY_MDIO_GCTRL_TM(mode) FIELD_PREP(XWAY_MDIO_GCTRL_TM_MASK, (mode)) +#define XWAY_MDIO_GCTRL_TM_NOP XWAY_MDIO_GCTRL_TM(0) /* Normal operation */ +#define XWAY_MDIO_GCTRL_TM_CDIAG XWAY_MDIO_GCTRL_TM(6) /* Cable diagnostics */ + #define XWAY_MDIO_ERRCNT_SEL GENMASK(11, 8) #define XWAY_MDIO_ERRCNT_COUNT GENMASK(7, 0) #define XWAY_MDIO_ERRCNT_SEL_RXERR 0 @@ -326,6 +331,28 @@ static int xway_gphy_probe(struct phy_device *phydev) return 0; } +static int xway_11g_int_config_init(struct phy_device *phydev) +{ + int err; + + /* An issue has been sporadically observed after device power-on on the + * first link-up attempt in 100BASE-TX mode resulting in either the + * link-up taking a long time, or failing to link-up altogether. + * + * Workaround: + * After power-on, enable Cable Diagnostic Mode for all ports and + * disable it. + */ + err = phy_modify(phydev, MII_CTRL1000, XWAY_MDIO_GCTRL_TM_MASK, XWAY_MDIO_GCTRL_TM_CDIAG); + if (err) + return err; + err = phy_modify(phydev, MII_CTRL1000, XWAY_MDIO_GCTRL_TM_MASK, XWAY_MDIO_GCTRL_TM_NOP); + if (err) + return err; + + return xway_gphy_config_init(phydev); +} + static int xway_gphy14_config_aneg(struct phy_device *phydev) { int reg, err; @@ -735,7 +762,7 @@ static struct phy_driver xway_gphy[] = { .phy_id_mask = 0xffffffff, .name = "Intel XWAY PHY11G (xRX v1.2 integrated)", /* PHY_GBIT_FEATURES */ - .config_init = xway_gphy_config_init, + .config_init = xway_11g_int_config_init, .probe = xway_gphy_probe, .handle_interrupt = xway_gphy_handle_interrupt, .config_intr = xway_gphy_config_intr, From 8e1937fed6738460554ec123c64839e2445e7d53 Mon Sep 17 00:00:00 2001 From: Ginger Li Date: Tue, 22 Sep 2026 16:09:09 +0800 Subject: [PATCH 539/694] tipc: Fix a data race on mon->peer_cnt in mon_timeout() mon_timeout() evaluates dom_size(mon->peer_cnt) before it takes mon->lock, while mon->peer_cnt is updated under that lock by tipc_mon_add_peer() and tipc_mon_remove_peer(). The value can therefore be stale, and the decision whether the local domain has to be recomputed can be based on an outdated member count. Read mon->peer_cnt inside the write_lock_bh(&mon->lock) protected region. Fixes: 35c55c9877f8 ("tipc: add neighbor monitoring framework") Signed-off-by: Ginger Li Reviewed-by: Tung Nguyen Link: https://patch.msgid.link/20260922080909.21123-1-ginger.jzllee@gmail.com Signed-off-by: Jakub Kicinski --- net/tipc/monitor.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/net/tipc/monitor.c b/net/tipc/monitor.c index a94b9b36a700..1a438e312d66 100644 --- a/net/tipc/monitor.c +++ b/net/tipc/monitor.c @@ -632,9 +632,10 @@ static void mon_timeout(struct timer_list *t) { struct tipc_monitor *mon = timer_container_of(mon, t, timer); struct tipc_peer *self; - int best_member_cnt = dom_size(mon->peer_cnt) - 1; + int best_member_cnt; write_lock_bh(&mon->lock); + best_member_cnt = dom_size(mon->peer_cnt) - 1; self = mon->self; if (self && (best_member_cnt != self->applied)) { mon_update_local_domain(mon); From 56d82862a0a243ac14ba11b6d7b57ddc2d064b95 Mon Sep 17 00:00:00 2001 From: Dairui Zhang Date: Wed, 23 Sep 2026 13:01:01 +0800 Subject: [PATCH 540/694] af_packet: fix integer overflow in prb_calc_retire_blk_tmo() prb_calc_retire_blk_tmo() computes in 32-bit int arithmetic: mbits = (blk_size_in_bytes * 8) / (1024 * 1024); If I'm reading the validation right, tp_block_size is user controlled and packet_set_ring() only rejects values that are <= 0 as int or not page aligned, so a 256MiB block goes right through (and alloc_one_pg_vec_page() even has a vzalloc fallback for it). 0x10000000 * 8 wraps to INT_MIN, and on a NIC reporting 1 Gbps (div == 1) the function ends up returning -2047. The condition is actually (8 * size) mod 2^32 >= 2^31 && div == 1, so the trigger set is [256,512), [768,1024), [1280,1536) and [1792,2048) MiB. Other sizes wrap to non-negative values and faster links divide the unsigned value back below 2^31, which is why this doesn't blow up for everyone. What makes it fatal is what happens next in init_prb_bdqc(): p1->interval_ktime = ms_to_ktime(prb_calc_retire_blk_tmo(...)); hrtimer_start(&p1->retire_blk_timer, p1->interval_ktime, HRTIMER_MODE_REL_SOFT); A negative relative timeout expires immediately. The callback unconditionally returns HRTIMER_RESTART, and hrtimer_forward() turns the negative interval into hrtimer_resolution: if (interval < hrtimer_resolution) interval = hrtimer_resolution; So the SOFT timer re-fires at the maximum rate forever, holding sk_receive_queue.lock each pass. One CPU spins in softirq until the socket is closed. Repeat with more rings and the machine is gone. The overflow itself is ancient - it was introduced together with TPACKET_V3 in f6fb8f100b80 ("af-packet: TPACKET_V3 flexible buffer implementation."). Its effect prior to f7460d2989fa ("net: af_packet: Use hrtimer to do the retire operation", v6.18) was not as clear-cut, though: the return value was stored into an unsigned short retire_blk_tov, so a negative result was truncated, and a 0-jiffy delay loop could be programmed as well. Neither is nearly as detrimental as the immediate maximum-rate spin the hrtimer conversion turned it into. (Unrelated to CVE-2019-20812 - that one was the ethtool failure path returning 0, which now returns DEFAULT_PRB_RETIRE_TOV.) Reproducer, needs CAP_NET_RAW (a --network host container has it by default) and a 1 Gbps NIC (QEMU e1000 works): int fd = socket(AF_PACKET, SOCK_RAW, htons(ETH_P_ALL)); bind(fd, ...); int v = TPACKET_V3; setsockopt(fd, SOL_PACKET, PACKET_VERSION, &v, sizeof(v)); struct tpacket_req3 req = { .tp_block_size = 0x10000000, .tp_block_nr = 1, .tp_frame_size = 2048, .tp_frame_nr = 0x10000000 / 2048, .tp_retire_blk_tov = 0, }; setsockopt(fd, SOL_PACKET, PACKET_RX_RING, &req, sizeof(req)); Compute in 64 bits instead. The operands are already bounded by the existing validation, so nothing else changes. If you'd prefer a different fix, just say so and I'll respin. Fixes: f6fb8f100b80 ("af-packet: TPACKET_V3 flexible buffer implementation.") Cc: stable@vger.kernel.org Signed-off-by: Dairui Zhang Reviewed-by: Willem de Bruijn Link: https://patch.msgid.link/20260923050101.1510064-1-zhangdairui@gmail.com Signed-off-by: Jakub Kicinski --- net/packet/af_packet.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c index 64b501db660a..7c83e01526ed 100644 --- a/net/packet/af_packet.c +++ b/net/packet/af_packet.c @@ -617,7 +617,7 @@ static int prb_calc_retire_blk_tmo(struct packet_sock *po, return DEFAULT_PRB_RETIRE_TOV; div = ecmd.base.speed / 1000; - mbits = (blk_size_in_bytes * 8) / (1024 * 1024); + mbits = (u64)blk_size_in_bytes * 8 / (1024 * 1024); if (div) mbits /= div; From 8db67bb6a1fffa4df68fbbc22e39943aeeff9178 Mon Sep 17 00:00:00 2001 From: Coia Prant Date: Wed, 23 Sep 2026 20:37:13 +0800 Subject: [PATCH 541/694] net: ethernet: stmmac: dwmac-rk: fix bulk clock leak when the PHY clock fails gmac_clk_enable() enables the bulk clocks first and then the optional PHY clock. If clk_prepare_enable() on the PHY clock fails, the function returns without rolling back the bulk clocks, and bsp_priv->clk_enabled stays false, so the later gmac_clk_enable(bsp_priv, false) becomes a no-op and the bulk clock references are leaked. Add the missing clk_bulk_disable_unprepare() on that failure path. Fixes: ea449f7fa0bf ("net: ethernet: stmmac: dwmac-rk: rework optional clock handling") Reviewed-by: Maxime Chevallier Reviewed-by: Heiko Stuebner Acked-by: Lorenzo Bianconi Signed-off-by: Coia Prant Link: https://patch.msgid.link/20260923123713.3137146-1-coiaprant@gmail.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/stmicro/stmmac/dwmac-rk.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac-rk.c b/drivers/net/ethernet/stmicro/stmmac/dwmac-rk.c index 8d7042e68926..72bdbcb5e863 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwmac-rk.c +++ b/drivers/net/ethernet/stmicro/stmmac/dwmac-rk.c @@ -1162,8 +1162,11 @@ static int gmac_clk_enable(struct rk_priv_data *bsp_priv, bool enable) return ret; ret = clk_prepare_enable(bsp_priv->clk_phy); - if (ret) + if (ret) { + clk_bulk_disable_unprepare(bsp_priv->num_clks, + bsp_priv->clks); return ret; + } rk_configure_io_clksel(bsp_priv); rk_ungate_rmii_clock(bsp_priv); From 06e3f54e8b22040ada01a28343badf1990b4dd7d Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 23 Sep 2026 13:03:18 +0000 Subject: [PATCH 542/694] net: flush skb_defer_nodes in dev_cpu_dead() When a CPU goes offline, dev_cpu_dead() drains its softnet queues (completion_queue, output_queue, poll_list, process_queue, and input_pkt_queue), but leaves net_hotdata.skb_defer_nodes untouched. If oldcpu goes offline while holding pending skbs in its skb_defer_nodes lists (e.g. below the sysctl_skb_defer_max >> 1 IPI threshold, or if the IPI races with CPU teardown), those skbs remain stranded until oldcpu is brought back online. If any of these skbs hold page_pool fragments, page_pool_destroy() will stall indefinitely waiting for inflight pages to be returned when a netdev or driver is torn down while oldcpu is offline. Additionally, if smp_call_function_single_async() fails in kick_defer_list_purge() because the target CPU went offline, reset defer_ipi_scheduled to 0 so future IPI kicks are not blocked when the CPU comes back online. Also, if oldcpu was the last online CPU on its NUMA node, drain that node's slot across all CPUs so no skbs deferred from that node remain stranded on idle remote CPUs (or if the node itself is subsequently offlined). Finally, in skb_attempt_defer_free(), re-check cpu_online(cpu) and whether the caller migrated CPUs after llist_add(), flushing the node list if so, to close the preemption TOCTOU race against CPU/node teardown. Fixes: 68822bdf76f1 ("net: generalize skb freeing deferral to per-cpu lists") Fixes: 5628f3fe3b16 ("net: add NUMA awareness to skb_attempt_defer_free()") Closes: https://lore.kernel.org/netdev/20260916003430.3612956-1-kris.pan@intel.com/ Signed-off-by: Eric Dumazet Cc: Kris Pan Link: https://patch.msgid.link/20260923130318.607255-1-edumazet@google.com Signed-off-by: Jakub Kicinski --- net/core/dev.c | 47 +++++++++++++++++++++++++++++++++++------------ net/core/dev.h | 2 ++ net/core/skbuff.c | 12 +++++++++--- 3 files changed, 46 insertions(+), 15 deletions(-) diff --git a/net/core/dev.c b/net/core/dev.c index 0292a16e16c2..f660fccfc0db 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -5376,7 +5376,8 @@ void kick_defer_list_purge(unsigned int cpu) backlog_unlock_irq_restore(sd, flags); } else if (!cmpxchg(&sd->defer_ipi_scheduled, 0, 1)) { - smp_call_function_single_async(cpu, &sd->defer_csd); + if (smp_call_function_single_async(cpu, &sd->defer_csd)) + WRITE_ONCE(sd->defer_ipi_scheduled, 0); } } @@ -6900,25 +6901,35 @@ bool napi_complete_done(struct napi_struct *n, int work_done) } EXPORT_SYMBOL(napi_complete_done); -static void skb_defer_free_flush(void) +static void __skb_defer_free_flush(struct skb_defer_node *sdn, int budget) { struct llist_node *free_list; struct sk_buff *skb, *next; + + if (llist_empty(&sdn->defer_list)) + return; + atomic_long_set(&sdn->defer_count, 0); + free_list = llist_del_all(&sdn->defer_list); + + llist_for_each_entry_safe(skb, next, free_list, ll_node) { + prefetch(next); + napi_consume_skb(skb, budget); + } +} + +void skb_defer_node_flush(struct skb_defer_node *sdn) +{ + __skb_defer_free_flush(sdn, 0); +} + +static void skb_defer_free_flush(void) +{ struct skb_defer_node *sdn; int node; for_each_node(node) { sdn = this_cpu_ptr(net_hotdata.skb_defer_nodes) + node; - - if (llist_empty(&sdn->defer_list)) - continue; - atomic_long_set(&sdn->defer_count, 0); - free_list = llist_del_all(&sdn->defer_list); - - llist_for_each_entry_safe(skb, next, free_list, ll_node) { - prefetch(next); - napi_consume_skb(skb, 1); - } + __skb_defer_free_flush(sdn, 1); } } @@ -12897,6 +12908,7 @@ static int dev_cpu_dead(unsigned int oldcpu) struct sk_buff **list_skb; struct sk_buff *skb; unsigned int cpu; + int node; struct softnet_data *sd, *oldsd, *remsd = NULL; local_irq_disable(); @@ -12957,6 +12969,17 @@ static int dev_cpu_dead(unsigned int oldcpu) rps_input_queue_head_incr(oldsd); } + for_each_node(node) + skb_defer_node_flush(per_cpu_ptr(net_hotdata.skb_defer_nodes, + oldcpu) + node); + node = cpu_to_node(oldcpu); + if (node_possible(node) && + !cpumask_intersects(cpumask_of_node(node), cpu_online_mask)) { + for_each_possible_cpu(cpu) + skb_defer_node_flush(per_cpu_ptr(net_hotdata.skb_defer_nodes, + cpu) + node); + } + return 0; } diff --git a/net/core/dev.h b/net/core/dev.h index b757faead4d1..04fb0e9a571e 100644 --- a/net/core/dev.h +++ b/net/core/dev.h @@ -399,6 +399,8 @@ static inline void napi_assert_will_not_race(const struct napi_struct *napi) WARN_ON(READ_ONCE(napi->list_owner) != -1); } +struct skb_defer_node; +void skb_defer_node_flush(struct skb_defer_node *sdn); void kick_defer_list_purge(unsigned int cpu); int dev_set_hwtstamp_phylib(struct net_device *dev, diff --git a/net/core/skbuff.c b/net/core/skbuff.c index b4edbd06655e..c3042d822afa 100644 --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@ -7360,8 +7360,8 @@ void skb_attempt_defer_free(struct sk_buff *skb) struct skb_defer_node *sdn; unsigned long defer_count; unsigned int defer_max; + int cpu, my_cpu; bool kick; - int cpu; if (static_branch_unlikely(&skb_defer_disable_key)) goto nodefer; @@ -7371,7 +7371,8 @@ void skb_attempt_defer_free(struct sk_buff *skb) goto nodefer; cpu = skb->alloc_cpu; - if (cpu == raw_smp_processor_id() || + my_cpu = raw_smp_processor_id(); + if (cpu == my_cpu || WARN_ON_ONCE(cpu >= nr_cpu_ids) || !cpu_online(cpu)) { nodefer: kfree_skb_napi_cache(skb); @@ -7382,7 +7383,7 @@ nodefer: kfree_skb_napi_cache(skb); DEBUG_NET_WARN_ON_ONCE(skb->destructor); DEBUG_NET_WARN_ON_ONCE(skb_nfct(skb)); - sdn = per_cpu_ptr(net_hotdata.skb_defer_nodes, cpu) + numa_node_id(); + sdn = per_cpu_ptr(net_hotdata.skb_defer_nodes, cpu) + cpu_to_node(my_cpu); defer_max = READ_ONCE(net_hotdata.sysctl_skb_defer_max); defer_count = atomic_long_inc_return(&sdn->defer_count); @@ -7392,6 +7393,11 @@ nodefer: kfree_skb_napi_cache(skb); llist_add(&skb->ll_node, &sdn->defer_list); + if (unlikely(!cpu_online(cpu) || my_cpu != raw_smp_processor_id())) { + skb_defer_node_flush(sdn); + return; + } + /* Send an IPI every time queue reaches half capacity. */ kick = (defer_count - 1) == (defer_max >> 1); From 83769c23fb1879edc916a526ba424285033baf2d Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 23 Sep 2026 14:59:42 +0000 Subject: [PATCH 543/694] gve: DQO: fix header length used by gve_can_send_tso() for UDP GSO gve_can_send_tso() computes how many buffers each segment of a GSO packet would span, and for this it needs the length of the headers that the device replicates in front of every segment. It unconditionally uses skb_tcp_all_headers(), which reads the doff field of the TCP header. SKB_GSO_UDP_L4 packets have no TCP header: tcp_hdrlen() then reads one byte of the UDP payload, and header_len can be anything in [0, 60] instead of the transport offset plus the eight bytes of the UDP header that gve_prep_tso() programs into the TSO context descriptor. A wrong header length shifts all the segment boundaries computed in the loop, so the number of buffers per segment can be over or under estimated. In the first case, GSO is needlessly disabled for this packet by gve_features_check_dqo() and the stack has to segment it. In the second case, the driver hands the device a packet whose segments span more than GVE_TX_MAX_DATA_DESCS buffers. Use the UDP header length for SKB_GSO_UDP_L4 packets, matching what gve_prep_tso() does. Fixes: 014c607f86ab ("gve: add support for UDP GSO for DQO format") Closes: https://lore.kernel.org/netdev/CANn89i+MS4L60sFQ49=-f-mibeveUfcrpVkD5X+Qy6SOnEpd6w@mail.gmail.com/ Signed-off-by: Eric Dumazet Cc: Ankit Garg Cc: Harshitha Ramamurthy Cc: Joshua Washington Cc: Willem de Bruijn Reviewed-by: Ankit Garg Reviewed-by: Harshitha Ramamurthy Link: https://patch.msgid.link/20260923145942.731365-1-edumazet@google.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/google/gve/gve_tx_dqo.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/google/gve/gve_tx_dqo.c b/drivers/net/ethernet/google/gve/gve_tx_dqo.c index 80ab0a449ff5..0f6f7c5dbb2e 100644 --- a/drivers/net/ethernet/google/gve/gve_tx_dqo.c +++ b/drivers/net/ethernet/google/gve/gve_tx_dqo.c @@ -918,13 +918,19 @@ static bool gve_can_send_tso(const struct sk_buff *skb) { const int max_bufs_per_seg = GVE_TX_MAX_DATA_DESCS - 1; const struct skb_shared_info *shinfo = skb_shinfo(skb); - const int header_len = skb_tcp_all_headers(skb); const int gso_size = shinfo->gso_size; int cur_seg_num_bufs; int prev_frag_size; int cur_seg_size; + int header_len; int i; + /* Must match the header length programmed by gve_prep_tso(). */ + if (skb_is_gso_tcp(skb)) + header_len = skb_tcp_all_headers(skb); + else + header_len = skb_transport_offset(skb) + sizeof(struct udphdr); + cur_seg_size = skb_headlen(skb) - header_len; prev_frag_size = skb_headlen(skb); cur_seg_num_bufs = cur_seg_size > 0; From 3b430ea6234087957b0d3cd181e3116722b59819 Mon Sep 17 00:00:00 2001 From: Eddie Phillips Date: Thu, 24 Sep 2026 00:42:51 +0000 Subject: [PATCH 544/694] gve: fix TX drop when GSO MSS is too small for hw The device has a strict requirement that the minimum MSS (gso_size) for TSO/GSO packets must be at least 88 bytes. If a packet below this threshold is pushed to the hardware, it can cause hardware to silently drop the packet, leading to increased latency and retransmissions. Currently, this is validated too late in the transmit pipeline (gve_prep_tso), leading to silent drops. Fix this by moving the validation into the .ndo_features_check callback (gve_features_check_dqo). If we detect a GSO packet with a gso_size smaller than GVE_TX_MIN_TSO_MSS_DQO, we clear the GSO feature flags for this packet. Fixes: a57e5de476be ("gve: DQO: Add TX path") Signed-off-by: Eddie Phillips Signed-off-by: Eric Dumazet Reviewed-by: Harshitha Ramamurthy Link: https://patch.msgid.link/20260924004252.1196328-2-edumazet@google.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/google/gve/gve_tx_dqo.c | 14 +++----------- 1 file changed, 3 insertions(+), 11 deletions(-) diff --git a/drivers/net/ethernet/google/gve/gve_tx_dqo.c b/drivers/net/ethernet/google/gve/gve_tx_dqo.c index 0f6f7c5dbb2e..e5fe17b04798 100644 --- a/drivers/net/ethernet/google/gve/gve_tx_dqo.c +++ b/drivers/net/ethernet/google/gve/gve_tx_dqo.c @@ -577,17 +577,6 @@ static int gve_prep_tso(struct sk_buff *skb) int header_len; int err; - /* Note: HW requires MSS (gso_size) to be <= 9728 and the total length - * of the TSO to be <= 262143. - * - * However, we don't validate these because: - * - Hypervisor enforces a limit of 9K MTU - * - Kernel will not produce a TSO larger than 64k - */ - - if (unlikely(shinfo->gso_size < GVE_TX_MIN_TSO_MSS_DQO)) - return -1; - /* Needed because we will modify header. */ err = skb_cow_head(skb, 0); if (err < 0) @@ -925,6 +914,9 @@ static bool gve_can_send_tso(const struct sk_buff *skb) int header_len; int i; + if (unlikely(gso_size < GVE_TX_MIN_TSO_MSS_DQO)) + return false; + /* Must match the header length programmed by gve_prep_tso(). */ if (skb_is_gso_tcp(skb)) header_len = skb_tcp_all_headers(skb); From 296c83b5ccc808c080865eb20fd7a477b0355bb7 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 24 Sep 2026 00:42:52 +0000 Subject: [PATCH 545/694] gve: DQO: reject TSO packets with an out of range MSS gve_prep_tso() notes that the device requires the MSS to be <= 9728, but does not enforce it, assuming the 9K MTU enforced by the hypervisor and the 64KB limit on TSO sizes are enough. This does not hold for packets that were not generated locally. A guest behind a tap, or any packet socket user, can provide an arbitrary gso_size in virtio_net_hdr. Layer 2 forwarding does not check the MTU for GSO packets (is_skb_forwardable()), and gso_features_check() only bounds skb->len and gso_segs, never gso_size. Such a packet reaches gve_tx_fill_tso_ctx_desc(), which puts gso_size into the mss field of the TSO context descriptor. This field is 14 bits wide, so a gso_size of 16384 is silently turned into an MSS of zero. Drop these packets from gve_prep_tso(), and make sure that gve_features_check_dqo() leaves their GSO bits alone: skb_segment() splits at gso_size regardless of the MTU, so falling back to software segmentation would give the device non TSO packets bigger than the 9728 bytes it supports. Note that the device can still be given oversized non TSO packets when the stack segments in software for other reasons, for instance after TSO has been disabled with ethtool. This is a generic issue, because the MTU check is skipped for GSO packets in the forwarding path, and is addressed separately. Fixes: a57e5de476be ("gve: DQO: Add TX path") Signed-off-by: Eric Dumazet Reviewed-by: Harshitha Ramamurthy Link: https://patch.msgid.link/20260924004252.1196328-3-edumazet@google.com Signed-off-by: Jakub Kicinski --- .../net/ethernet/google/gve/gve_desc_dqo.h | 5 ++++ drivers/net/ethernet/google/gve/gve_tx_dqo.c | 26 ++++++++++++++++++- 2 files changed, 30 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/google/gve/gve_desc_dqo.h b/drivers/net/ethernet/google/gve/gve_desc_dqo.h index f7786b03c744..d2c86c8eeae2 100644 --- a/drivers/net/ethernet/google/gve/gve_desc_dqo.h +++ b/drivers/net/ethernet/google/gve/gve_desc_dqo.h @@ -14,6 +14,11 @@ #define GVE_TX_MAX_HDR_SIZE_DQO 255 #define GVE_TX_MIN_TSO_MSS_DQO 88 +/* HW limit. This also has to fit in the 14 bits of the mss field of + * struct gve_tx_tso_context_desc_dqo. + */ +#define GVE_TX_MAX_TSO_MSS_DQO 9728 + #ifndef __LITTLE_ENDIAN_BITFIELD #error "Only little endian supported" #endif diff --git a/drivers/net/ethernet/google/gve/gve_tx_dqo.c b/drivers/net/ethernet/google/gve/gve_tx_dqo.c index e5fe17b04798..616c1921aebe 100644 --- a/drivers/net/ethernet/google/gve/gve_tx_dqo.c +++ b/drivers/net/ethernet/google/gve/gve_tx_dqo.c @@ -577,6 +577,20 @@ static int gve_prep_tso(struct sk_buff *skb) int header_len; int err; + /* Note: HW requires the total length of the TSO to be <= 262143, + * this is enforced by netif_set_tso_max_size(). + * + * MSS (gso_size) can not be trusted: packets forwarded from a tap or + * injected by a packet socket can carry an arbitrary value, while the + * mss field of the TSO context descriptor is only 14 bits wide. + * + * A too big MSS is dropped here instead of being rejected from + * gve_features_check_dqo(), because software segmentation would + * produce packets larger than the device can send. + */ + if (unlikely(shinfo->gso_size > GVE_TX_MAX_TSO_MSS_DQO)) + return -1; + /* Needed because we will modify header. */ err = skb_cow_head(skb, 0); if (err < 0) @@ -964,7 +978,17 @@ netdev_features_t gve_features_check_dqo(struct sk_buff *skb, struct net_device *dev, netdev_features_t features) { - if (skb_is_gso(skb) && !gve_can_send_tso(skb)) + if (!skb_is_gso(skb)) + return features; + + /* Keep the GSO bits for a too big MSS, so that gve_prep_tso() drops + * the packet: software segmentation would give packets larger than + * the device can send. + */ + if (skb_shinfo(skb)->gso_size > GVE_TX_MAX_TSO_MSS_DQO) + return features; + + if (!gve_can_send_tso(skb)) return features & ~NETIF_F_GSO_MASK; return features; From 72b5b9a28b996e09b8b5b944370c79851bb68f52 Mon Sep 17 00:00:00 2001 From: Zixuan Chai Date: Thu, 24 Sep 2026 09:26:05 +0800 Subject: [PATCH 546/694] llc: reserve device headroom for allocated frames llc_alloc_frame() reserves link-layer headroom using the device type. This is insufficient for stacked Ethernet devices such as VLAN devices, where vlan_dev_hard_header() pushes a VLAN header before the lower device's Ethernet header. An LLC response on such a device can therefore underflow skb headroom in eth_header(). Use LL_RESERVED_SPACE() to account for the device's actual required headroom while preserving the existing LLC device-type check. Fixes: bf9ae5386bca ("llc: use dev_hard_header") Cc: stable@vger.kernel.org Reported-by: VEGA Signed-off-by: Zixuan Chai Signed-off-by: Ren Wei Reviewed-by: Eric Dumazet Link: https://patch.msgid.link/20260924012613.2533934-1-weir@nebusec.ai Signed-off-by: Jakub Kicinski --- net/llc/llc_sap.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/net/llc/llc_sap.c b/net/llc/llc_sap.c index 1bd446a21092..3904a1b4ba84 100644 --- a/net/llc/llc_sap.c +++ b/net/llc/llc_sap.c @@ -19,12 +19,12 @@ #include #include -static int llc_mac_header_len(unsigned short devtype) +static int llc_mac_header_len(struct net_device *dev) { - switch (devtype) { + switch (dev->type) { case ARPHRD_ETHER: case ARPHRD_LOOPBACK: - return sizeof(struct ethhdr); + return LL_RESERVED_SPACE(dev); } return 0; } @@ -45,7 +45,7 @@ struct sk_buff *llc_alloc_frame(struct sock *sk, struct net_device *dev, int hlen = type == LLC_PDU_TYPE_U ? 3 : 4; struct sk_buff *skb; - hlen += llc_mac_header_len(dev->type); + hlen += llc_mac_header_len(dev); skb = alloc_skb(hlen + data_size, GFP_ATOMIC); if (skb) { From 72f9dd522f8d6c5a00be9695c7bb74631eb5069e Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 24 Sep 2026 08:29:48 +0000 Subject: [PATCH 547/694] llc: fix skb UAF and leaks on llc_mac_hdr_init() failure In llc_conn_ac_resend_i_xxx_x_set_0_or_send_rr(), if llc_mac_hdr_init() fails, kfree_skb(skb) is called instead of kfree_skb(nskb). This leaks the newly allocated nskb, reads from the freed skb via LLC_I_GET_NR(pdu), and double-frees skb when llc_conn_state_process() drops its reference. In llc_sap_action_send_xid_r() and llc_sap_action_send_test_r(), nskb is leaked if llc_mac_hdr_init() returns an error. Free nskb in all three error paths. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Closes: https://lore.kernel.org/netdev/179022851638.2160803.1808206741379444999@kernel.org/ Signed-off-by: Eric Dumazet Link: https://patch.msgid.link/20260924082951.1599377-2-edumazet@google.com Signed-off-by: Jakub Kicinski --- net/llc/llc_c_ac.c | 2 +- net/llc/llc_s_ac.c | 4 ++++ 2 files changed, 5 insertions(+), 1 deletion(-) diff --git a/net/llc/llc_c_ac.c b/net/llc/llc_c_ac.c index 724ecd741d4c..1aa7fe28acdd 100644 --- a/net/llc/llc_c_ac.c +++ b/net/llc/llc_c_ac.c @@ -437,7 +437,7 @@ int llc_conn_ac_resend_i_xxx_x_set_0_or_send_rr(struct sock *sk, if (likely(!rc)) llc_conn_send_pdu(sk, nskb); else - kfree_skb(skb); + kfree_skb(nskb); } if (rc) { nr = LLC_I_GET_NR(pdu); diff --git a/net/llc/llc_s_ac.c b/net/llc/llc_s_ac.c index 98deee560373..831998211b52 100644 --- a/net/llc/llc_s_ac.c +++ b/net/llc/llc_s_ac.c @@ -121,6 +121,8 @@ int llc_sap_action_send_xid_r(struct llc_sap *sap, struct sk_buff *skb) rc = llc_mac_hdr_init(nskb, mac_sa, mac_da); if (likely(!rc)) rc = dev_queue_xmit(nskb); + else + kfree_skb(nskb); out: return rc; } @@ -170,6 +172,8 @@ int llc_sap_action_send_test_r(struct llc_sap *sap, struct sk_buff *skb) rc = llc_mac_hdr_init(nskb, mac_sa, mac_da); if (likely(!rc)) rc = dev_queue_xmit(nskb); + else + kfree_skb(nskb); out: return rc; } From ac704ff08e511c87643799c385f55ecd69b85e03 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 24 Sep 2026 08:29:49 +0000 Subject: [PATCH 548/694] bridge: check llc_mac_hdr_init() return value in br_send_bpdu() If llc_mac_hdr_init() fails (for instance if the port device type does not support LLC or dev_hard_header() fails), br_send_bpdu() should drop the skb instead of resetting the mac header to the LLC payload and transmitting a malformed frame. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Closes: https://lore.kernel.org/netdev/179022851638.2160803.1808206741379444999@kernel.org/ Cc: Nikolay Aleksandrov Cc: Ido Schimmel Cc: bridge@lists.linux.dev Signed-off-by: Eric Dumazet Acked-by: Nikolay Aleksandrov Link: https://patch.msgid.link/20260924082951.1599377-3-edumazet@google.com Signed-off-by: Jakub Kicinski --- net/bridge/br_stp_bpdu.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/net/bridge/br_stp_bpdu.c b/net/bridge/br_stp_bpdu.c index 74ec42ba1e7d..21d092f5acbb 100644 --- a/net/bridge/br_stp_bpdu.c +++ b/net/bridge/br_stp_bpdu.c @@ -52,7 +52,10 @@ static void br_send_bpdu(struct net_bridge_port *p, LLC_SAP_BSPAN, LLC_PDU_CMD); llc_pdu_init_as_ui_cmd(skb); - llc_mac_hdr_init(skb, p->dev->dev_addr, p->br->group_addr); + if (llc_mac_hdr_init(skb, p->dev->dev_addr, p->br->group_addr)) { + kfree_skb(skb); + return; + } skb_reset_mac_header(skb); From 907b978e82cb4c1c245fc2985bb27c5d5c88c8f6 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 24 Sep 2026 08:29:50 +0000 Subject: [PATCH 549/694] net/sched: sch_teql: fix shadowed err in __teql_resolve() __teql_resolve() declares an inner 'int err;' inside the 'if (neigh_event_send(n, skb_res) == 0)' block, shadowing the outer 'int err = 0;'. As a result, a negative return from dev_hard_header() is written to the inner variable and __teql_resolve() still returns 0. Remove the shadowed variable and set the outer err to -EINVAL when dev_hard_header() returns a negative error. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Closes: https://lore.kernel.org/netdev/179022851638.2160803.1808206741379444999@kernel.org/ Cc: Jamal Hadi Salim Cc: Jiri Pirko Signed-off-by: Eric Dumazet Link: https://patch.msgid.link/20260924082951.1599377-4-edumazet@google.com Signed-off-by: Jakub Kicinski --- net/sched/sch_teql.c | 7 ++----- 1 file changed, 2 insertions(+), 5 deletions(-) diff --git a/net/sched/sch_teql.c b/net/sched/sch_teql.c index 9e52afc2d980..409ce50cc0db 100644 --- a/net/sched/sch_teql.c +++ b/net/sched/sch_teql.c @@ -265,14 +265,11 @@ __teql_resolve(struct sk_buff *skb, struct sk_buff *skb_res, } if (neigh_event_send(n, skb_res) == 0) { - int err; char haddr[MAX_ADDR_LEN]; neigh_ha_snapshot(haddr, n, dev); - err = dev_hard_header(skb, dev, ntohs(skb_protocol(skb, false)), - haddr, NULL, skb->len); - - if (err < 0) + if (dev_hard_header(skb, dev, ntohs(skb_protocol(skb, false)), + haddr, NULL, skb->len) < 0) err = -EINVAL; } else { err = (skb_res == NULL) ? -EAGAIN : 1; From cd5dd68267c4238795fadaf02b3575ca3f8a6500 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 24 Sep 2026 08:29:51 +0000 Subject: [PATCH 550/694] vlan: ensure sufficient headroom in vlan_dev_hard_header() Callers that only reserve ETH_HLEN or less (such as llc_alloc_frame()), or skbs allocated before dynamic device/headroom changes (e.g. toggling VLAN_FLAG_REORDER_HDR or bonding/team switching slaves), can reach vlan_dev_hard_header() with insufficient headroom and trigger skb_under_panic(). Use skb_cow_head() in vlan_dev_hard_header() when VLAN_FLAG_REORDER_HDR is not set to ensure sufficient headroom for the VLAN header(s) and the underlying device hard header. Use READ_ONCE() to read dev->hard_header_len and dev->needed_headroom as they can be updated concurrently under RTNL (e.g. in vlan_transfer_features()) while vlan_dev_hard_header() runs locklessly on the transmit path. Also avoid LL_RESERVED_SPACE(dev) here so that the extra HH_DATA_MOD alignment padding does not trigger unnecessary pskb_expand_head() reallocations on inner stacked VLAN devices after the outer VLAN header has been pushed. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Reported-by: Zixuan Chai Closes: https://lore.kernel.org/netdev/cover.1789987105.git.petalzu987@gmail.com/ Link: https://lore.kernel.org/netdev/179022851638.2160803.1808206741379444999@kernel.org/ Cc: Hangbin Liu Signed-off-by: Eric Dumazet Link: https://patch.msgid.link/20260924082951.1599377-5-edumazet@google.com Signed-off-by: Jakub Kicinski --- net/8021q/vlan_dev.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/net/8021q/vlan_dev.c b/net/8021q/vlan_dev.c index 2859cbac3f26..c949c6a82945 100644 --- a/net/8021q/vlan_dev.c +++ b/net/8021q/vlan_dev.c @@ -55,6 +55,11 @@ static int vlan_dev_hard_header(struct sk_buff *skb, struct net_device *dev, int rc; if (!(vlan->flags & VLAN_FLAG_REORDER_HDR)) { + unsigned int hlen = READ_ONCE(dev->hard_header_len) + + READ_ONCE(dev->needed_headroom); + + if (skb_cow_head(skb, hlen) < 0) + return -ENOMEM; vhdr = skb_push(skb, VLAN_HLEN); vlan_tci = vlan->vlan_id; From fc6d80eb504458d6416b75a94188b268c95c6533 Mon Sep 17 00:00:00 2001 From: Willem de Bruijn Date: Thu, 24 Sep 2026 11:44:12 -0400 Subject: [PATCH 551/694] tcp: prevent collapsing skbs across boundary in rtx queue tcp_write_collapse_fence() sets TCP_SKB_CB(skb)->eor = 1 on tcp_write_queue_tail(sk) to prevent skbs queued after a switch to device encryption from being collapsed into earlier skbs. The fence is a no-op if all earlier data has already been transmitted when the switch happens: sk->sk_write_queue is empty. The not yet acknowledged earlier skbs wait in sk->tcp_rtx_queue with eor 0. On a subsequent retransmit or SACK shift, tcp_retrans_try_collapse() or tcp_shift_skb_data() can then merge an skb queued after the switch into one queued before it. Both users of the fence are affected: - psp: devices only encrypt skbs with skb->decrypted set. The merged skb keeps decrypted = 0 from the earlier skb, so merged data sent after psp_sock_assoc_set_tx() is retransmitted in cleartext. - tls device offload: the merged skb straddles the start marker set in tls_set_device_offload(). The software fallback (fill_sg_in() returns -EINVAL) and the mlx5, nfp and funeth drivers cannot handle such an skb and drop it. Every retransmit rebuilds the same skb, so the connection stalls. Fix this in two places, for defense in depth: 1. Fall back to tcp_rtx_queue_tail(sk) in tcp_write_collapse_fence() when tcp_write_queue_tail(sk) is NULL. 2. Check !skb_cmp_decrypted(to, from) in tcp_skb_can_collapse(), as tcp_skb_can_collapse_rx() does on receive. skb_shift(), which both collapse paths call, already has a DEBUG_NET_WARN_ON_ONCE() for this condition. Fixes: e8f69799810c ("net/tls: Add generic NIC offload infrastructure") Cc: stable@vger.kernel.org Signed-off-by: Willem de Bruijn Reviewed-by: Eric Dumazet Reviewed-by: Daniel Zahka Link: https://patch.msgid.link/20260924154427.953800-1-willemdebruijn.kernel@gmail.com Signed-off-by: Jakub Kicinski --- include/net/tcp.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/include/net/tcp.h b/include/net/tcp.h index 436495ff2271..4416cdf9bf30 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -1232,9 +1232,9 @@ static inline bool tcp_skb_can_collapse_to(const struct sk_buff *skb) static inline bool tcp_skb_can_collapse(const struct sk_buff *to, const struct sk_buff *from) { - /* skb_cmp_decrypted() not needed, use tcp_write_collapse_fence() */ return likely(tcp_skb_can_collapse_to(to) && mptcp_skb_can_collapse(to, from) && + !skb_cmp_decrypted(to, from) && skb_pure_zcopy_same(to, from) && skb_frags_readable(to) == skb_frags_readable(from)); } @@ -2327,7 +2327,7 @@ static inline void tcp_rtx_queue_unlink_and_free(struct sk_buff *skb, struct soc static inline void tcp_write_collapse_fence(struct sock *sk) { - struct sk_buff *skb = tcp_write_queue_tail(sk); + struct sk_buff *skb = tcp_write_queue_tail(sk) ?: tcp_rtx_queue_tail(sk); if (skb) TCP_SKB_CB(skb)->eor = 1; From 11536ee3d3e0b1bd35b6f3f8df55a6053eb0c71d Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 23 Sep 2026 08:32:03 +0000 Subject: [PATCH 552/694] MAINTAINERS: update Eric Dumazet's email address My @google.com address is going away soon, let's switch to @kernel.org. Signed-off-by: Eric Dumazet Link: https://patch.msgid.link/20260923083203.383923-1-edumazet@google.com Signed-off-by: Jakub Kicinski --- .mailmap | 2 ++ MAINTAINERS | 8 ++++---- 2 files changed, 6 insertions(+), 4 deletions(-) diff --git a/.mailmap b/.mailmap index 389b94a0124e..b8d0dc185714 100644 --- a/.mailmap +++ b/.mailmap @@ -286,6 +286,8 @@ Ed L. Cashin Elliot Berman Enric Balletbo i Serra Enric Balletbo i Serra +Eric Dumazet +Eric Dumazet Erik Kaneda Ethan Carter Edwards Ethan Edwards Eugen Hristev diff --git a/MAINTAINERS b/MAINTAINERS index c52ec2d7d3c1..c915b114c99c 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -18885,7 +18885,7 @@ F: net/core/drop_monitor.c NETWORKING DRIVERS M: Andrew Lunn M: "David S. Miller" -M: Eric Dumazet +M: Eric Dumazet M: Jakub Kicinski M: Paolo Abeni L: netdev@vger.kernel.org @@ -19002,7 +19002,7 @@ K: struct\s+phy_port|phy_port_ NETWORKING [GENERAL] M: "David S. Miller" -M: Eric Dumazet +M: Eric Dumazet M: Jakub Kicinski M: Paolo Abeni R: Simon Horman @@ -19228,7 +19228,7 @@ F: net/ipv6/seg6* F: tools/testing/selftests/net/srv6* NETWORKING [TCP] -M: Eric Dumazet +M: Eric Dumazet M: Neal Cardwell R: Kuniyuki Iwashima L: netdev@vger.kernel.org @@ -19261,7 +19261,7 @@ F: net/tls/ F: tools/testing/selftests/net/tls.c NETWORKING [SOCKETS] -M: Eric Dumazet +M: Eric Dumazet M: Kuniyuki Iwashima M: Paolo Abeni M: Willem de Bruijn From cb97bf3d4f91453b881acaf8e9f0cc47bb40b604 Mon Sep 17 00:00:00 2001 From: Viken Dadhaniya Date: Mon, 21 Sep 2026 17:17:01 +0530 Subject: [PATCH 553/694] i2c: qcom-geni: Fix hardcoded clock index in SE_GENI_CLK_SEL qcom_geni_i2c_conf() writes a hardcoded 0 to SE_GENI_CLK_SEL, which selects an index from the hardware clock performance table. This always picks the first table entry regardless of the actual source clock configuration. On platforms where the matching entry is not at index 0, the wrong source clock divider is active and the I2C bus runs at an incorrect frequency. Use geni_se_clk_freq_match() in geni_i2c_clk_map_idx() to find the performance table index for the source clock (32 MHz or 19.2 MHz). Store the resolved index in a new clk_idx field in geni_i2c_dev and write it to SE_GENI_CLK_SEL instead of the hardcoded 0. Fixes: 37692de5d523 ("i2c: i2c-qcom-geni: Add bus driver for the Qualcomm GENI I2C controller") Signed-off-by: Viken Dadhaniya Cc: # v4.19+ Reviewed-by: Mukesh Kumar Savaliya Signed-off-by: Andi Shyti Link: https://patch.msgid.link/20260921-i2c-fix-se-clk-conf-v2-1-8b5537ceff2d@oss.qualcomm.com --- drivers/i2c/busses/i2c-qcom-geni.c | 38 ++++++++++++++++++++++++++---- 1 file changed, 33 insertions(+), 5 deletions(-) diff --git a/drivers/i2c/busses/i2c-qcom-geni.c b/drivers/i2c/busses/i2c-qcom-geni.c index e9e41a174c20..6a39e21d5ed4 100644 --- a/drivers/i2c/busses/i2c-qcom-geni.c +++ b/drivers/i2c/busses/i2c-qcom-geni.c @@ -82,6 +82,9 @@ enum geni_i2c_err_code { #define XFER_TIMEOUT HZ #define RST_TIMEOUT HZ +#define GENI_SE_CLK_32MHZ (32 * HZ_PER_MHZ) +#define GENI_SE_CLK_19P2MHZ 19200000UL + struct geni_i2c_desc { bool no_dma_support; unsigned int tx_fifo_depth; @@ -127,6 +130,7 @@ struct geni_i2c_dev { spinlock_t lock; u32 clk_freq_out; const struct geni_i2c_clk_fld *clk_fld; + u32 clk_idx; void *dma_buf; size_t xfer_len; dma_addr_t dma_addr; @@ -197,19 +201,44 @@ static const struct geni_i2c_clk_fld geni_i2c_clk_map_32mhz[] = { static int geni_i2c_clk_map_idx(struct geni_i2c_dev *gi2c) { const struct geni_i2c_clk_fld *itr; + unsigned long res_freq; - if (clk_get_rate(gi2c->se.clk) == 32 * HZ_PER_MHZ) + /* + * Frequency counter tables are calibrated for a specific source + * clock frequency and are not valid for any multiple of it + * (e.g. 64 MHz, 128 MHz). + * Use exact=true and verify res_freq matches req_freq literally + * to reject harmonics: a 64 MHz clock that divides evenly to + * 32 MHz would pass exact matching but produce double the intended + * I2C frequency with these counter values. + */ + if (!geni_se_clk_freq_match(&gi2c->se, GENI_SE_CLK_32MHZ, + &gi2c->clk_idx, &res_freq, true) && + res_freq == GENI_SE_CLK_32MHZ) { itr = geni_i2c_clk_map_32mhz; - else + } else if (!geni_se_clk_freq_match(&gi2c->se, GENI_SE_CLK_19P2MHZ, + &gi2c->clk_idx, &res_freq, true) && + res_freq == GENI_SE_CLK_19P2MHZ) { itr = geni_i2c_clk_map_19p2mhz; + } else { + dev_err(gi2c->se.dev, + "Unsupported SE source clock: must be exactly 32 MHz or 19.2 MHz\n"); + return -EINVAL; + } while (itr->clk_freq_out != 0) { if (itr->clk_freq_out == gi2c->clk_freq_out) { gi2c->clk_fld = itr; + dev_dbg(gi2c->se.dev, + "I2C clk selected: freq: %u Hz, clk_idx: %u\n", + gi2c->clk_freq_out, gi2c->clk_idx); return 0; } itr++; } + + dev_err(gi2c->se.dev, "Unsupported I2C output frequency %u Hz\n", gi2c->clk_freq_out); + return -EINVAL; } @@ -219,7 +248,7 @@ static int qcom_geni_i2c_conf(struct geni_se *se, unsigned long freq) const struct geni_i2c_clk_fld *itr = gi2c->clk_fld; u32 val; - writel_relaxed(0, gi2c->se.base + SE_GENI_CLK_SEL); + writel_relaxed(gi2c->clk_idx, gi2c->se.base + SE_GENI_CLK_SEL); val = (itr->clk_div << CLK_DIV_SHFT) | SER_CLK_EN; writel_relaxed(val, gi2c->se.base + GENI_SER_M_CLK_CFG); @@ -1111,8 +1140,7 @@ static int geni_i2c_resources_init(struct geni_se *se) ret = geni_i2c_clk_map_idx(gi2c); if (ret) - return dev_err_probe(gi2c->se.dev, ret, "Invalid clk frequency %d Hz\n", - gi2c->clk_freq_out); + return ret; return geni_icc_set_bw_ab(&gi2c->se, GENI_DEFAULT_BW, GENI_DEFAULT_BW, Bps_to_icc(gi2c->clk_freq_out)); From f050c3e61d2a1aaece3170d459e4ce5fa2486c12 Mon Sep 17 00:00:00 2001 From: Vineet Gupta Date: Wed, 23 Sep 2026 15:01:38 -0700 Subject: [PATCH 554/694] ARC: arch_cmpxchg_relaxed to use size of pointed type not pointer Bradley spotted this snafu when working on two-byte cmpxchg emulation. He originally folded this fix into his series but the fix itself uncovered a pre-existing issue, missing arch_atomic64_cmpxchg_relaxed which just needed to be wired up. Fixes: e188f3330a13 ("ARC: cmpxchg/xchg: rewrite as macros to make type safe") Reported-by: kernel test robot Closes: https://lore.kernel.org/oe-kbuild-all/202609240539.AkulvjVe-lkp@intel.com/ Suggested-by: Bradley Morgan Signed-off-by: Vineet Gupta --- arch/arc/include/asm/atomic64-arcv2.h | 6 ++++++ arch/arc/include/asm/cmpxchg.h | 4 ++-- 2 files changed, 8 insertions(+), 2 deletions(-) diff --git a/arch/arc/include/asm/atomic64-arcv2.h b/arch/arc/include/asm/atomic64-arcv2.h index 73080a664369..1ca4a1047fb3 100644 --- a/arch/arc/include/asm/atomic64-arcv2.h +++ b/arch/arc/include/asm/atomic64-arcv2.h @@ -156,6 +156,12 @@ static inline u64 __arch_cmpxchg64_relaxed(volatile void *ptr, u64 old, u64 new) } #define arch_cmpxchg64_relaxed __arch_cmpxchg64_relaxed +static inline s64 arch_atomic64_cmpxchg_relaxed(atomic64_t *v, s64 old, s64 new) +{ + return arch_cmpxchg64_relaxed(&v->counter, old, new); +} +#define arch_atomic64_cmpxchg_relaxed arch_atomic64_cmpxchg_relaxed + static inline s64 arch_atomic64_xchg(atomic64_t *ptr, s64 new) { s64 prev; diff --git a/arch/arc/include/asm/cmpxchg.h b/arch/arc/include/asm/cmpxchg.h index 76f43db0890f..66b3743c9400 100644 --- a/arch/arc/include/asm/cmpxchg.h +++ b/arch/arc/include/asm/cmpxchg.h @@ -46,7 +46,7 @@ __typeof__(*(ptr)) _n_ = (new); \ __typeof__(*(ptr)) _prev_; \ \ - switch(sizeof((_p_))) { \ + switch (sizeof(*(_p_))) { \ case 1: \ _prev_ = (__typeof__(*(ptr)))cmpxchg_emu_u8((volatile u8 *__force)_p_, (uintptr_t)_o_, (uintptr_t)_n_); \ break; \ @@ -102,7 +102,7 @@ __typeof__(ptr) _p_ = (ptr); \ __typeof__(*(ptr)) _val_ = (val); \ \ - switch(sizeof(*(_p_))) { \ + switch (sizeof(*(_p_))) { \ case 4: \ _val_ = __arch_xchg(_p_, _val_); \ break; \ From 113dcdfadf30ea11fbbdfcd4f6ea87687655cc5b Mon Sep 17 00:00:00 2001 From: Matthias Goergens Date: Fri, 25 Sep 2026 13:23:29 +0800 Subject: [PATCH 555/694] MAINTAINERS: name the libata/linux for-next branch The T: entry for LIBATA SUBSYSTEM (Serial and Parallel ATA drivers) names libata/linux without a branch. The repository's HEAD pointer points to branch master, which has no active development. Active development is on the for-next branch. Name the branch so the entry identifies where development happens. Documentation/process/submitting-patches.rst sends contributors to the T: entry to find the tree to prepare patches against, so a branch-less entry whose HEAD is already in mainline points them to the wrong branch. Reviewed-by: Damien Le Moal Signed-off-by: Matthias Goergens Link: https://lore.kernel.org/r/20260925052329.2683619-1-matthias.goergens@gmail.com Signed-off-by: Niklas Cassel --- MAINTAINERS | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/MAINTAINERS b/MAINTAINERS index 3a19da74d00c..3e7ede95c7cf 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -14911,7 +14911,7 @@ M: Damien Le Moal M: Niklas Cassel L: linux-ide@vger.kernel.org S: Maintained -T: git git://git.kernel.org/pub/scm/linux/kernel/git/libata/linux.git +T: git git://git.kernel.org/pub/scm/linux/kernel/git/libata/linux.git for-next F: Documentation/ABI/testing/sysfs-ata F: Documentation/devicetree/bindings/ata/ F: drivers/ata/ From ae146bc1abdeb4607abf2975b858c053024e8ac1 Mon Sep 17 00:00:00 2001 From: Amir Goldstein Date: Mon, 21 Sep 2026 12:40:13 +0200 Subject: [PATCH 556/694] ovl: fix UAF in ovl_do_mkdir() debug print ovl_do_mkdir() prints the input dentry with %pd after vfs_mkdir(). Since commit fe497f0759e0 ("VFS: change vfs_mkdir() to unlock on failure."), vfs_mkdir() calls end_creating() on the input dentry on failure and may replace it on success, so the post-call %pd can use-after-free the dentry when CONFIG_OVERLAY_FS_DEBUG is enabled. Print the dentry before the call and only the result afterward. Reported-by: syzbot+ced26b784bf977d223dd@syzkaller.appspotmail.com Closes: https://syzkaller.appspot.com/bug?extid=ced26b784bf977d223dd Fixes: fe497f0759e0 ("VFS: change vfs_mkdir() to unlock on failure.") Signed-off-by: Amir Goldstein Link: https://patch.msgid.link/20260921104013.40475-1-amir73il@gmail.com Signed-off-by: Christian Brauner (Amutable) --- fs/overlayfs/overlayfs.h | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/fs/overlayfs/overlayfs.h b/fs/overlayfs/overlayfs.h index e0d8c6152e9f..7f3558372c59 100644 --- a/fs/overlayfs/overlayfs.h +++ b/fs/overlayfs/overlayfs.h @@ -254,8 +254,10 @@ static inline struct dentry *ovl_do_mkdir(struct ovl_fs *ofs, { struct dentry *ret; + /* vfs_mkdir() drops @dentry on failure and may replace it on success */ + pr_debug("mkdir(%pd2, 0%o)\n", dentry, mode); ret = vfs_mkdir(ovl_upper_mnt_idmap(ofs), dir, dentry, mode, NULL); - pr_debug("mkdir(%pd2, 0%o) = %i\n", dentry, mode, PTR_ERR_OR_ZERO(ret)); + pr_debug("...mkdir = %i\n", PTR_ERR_OR_ZERO(ret)); return ret; } From 76d8e697242e7e4e30dd08eff7c278728f9dd8e2 Mon Sep 17 00:00:00 2001 From: Drif Abdelmalek Mohamed Said Date: Fri, 18 Sep 2026 23:42:04 +0100 Subject: [PATCH 557/694] dcache: unpoison the inline name buffer in __d_alloc() syzbot reported: BUG: KMSAN: uninit-value in dentry_string_cmp fs/dcache.c:291 [inline] BUG: KMSAN: uninit-value in dentry_cmp fs/dcache.c:322 [inline] BUG: KMSAN: uninit-value in __d_lookup_rcu+0x37d/0x5e0 fs/dcache.c:2522 dentry_string_cmp fs/dcache.c:291 [inline] dentry_cmp fs/dcache.c:322 [inline] __d_lookup_rcu+0x37d/0x5e0 fs/dcache.c:2522 lookup_fast+0x194/0xa40 fs/namei.c:1854 lookup_fast_for_open fs/namei.c:4545 [inline] open_last_lookups fs/namei.c:4579 [inline] path_openat+0x9ef/0x6540 fs/namei.c:4856 do_file_open+0x2aa/0x680 fs/namei.c:4888 do_sys_openat2+0x17c/0x390 fs/open.c:1395 do_sys_open fs/open.c:1401 [inline] __do_sys_openat fs/open.c:1417 [inline] __se_sys_openat fs/open.c:1412 [inline] __x64_sys_openat+0x240/0x300 fs/open.c:1412 x64_sys_call+0x2445/0x3ea0 arch/x86/include/generated/asm/syscalls_64.h:258 do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline] do_syscall_64+0x15d/0x3c0 arch/x86/entry/syscall_64.c:94 entry_SYSCALL_64_after_hwframe+0x77/0x7f Uninit was stored to memory at: copy_name fs/dcache.c:3031 [inline] __d_move+0xd29/0x21f0 fs/dcache.c:3099 d_move+0x71/0xf0 fs/dcache.c:3147 vfs_rename+0x2619/0x2770 fs/namei.c:6085 filename_renameat2+0xa59/0x1230 fs/namei.c:6188 __do_sys_rename fs/namei.c:6232 [inline] __se_sys_rename+0xc5/0x5c0 fs/namei.c:6228 __x64_sys_rename+0x78/0xb0 fs/namei.c:6228 x64_sys_call+0x329/0x3ea0 arch/x86/include/generated/asm/syscalls_64.h:83 do_syscall_x64 arch/x86/entry/syscall_64.c:63 do_syscall_64+0x15d/0x3c0 arch/x86/entry/syscall_64.c:94 entry_SYSCALL_64_after_hwframe+0x77/0x7f Uninit was created at: slab_post_alloc_hook mm/slub.c:4617 [inline] slab_alloc_node mm/slub.c:4939 [inline] kmem_cache_alloc_lru_noprof+0x376/0x1230 mm/s __d_alloc+0x52/0x9f0 fs/dcache.c:1902 d_alloc+0x57/0x300 fs/dcache.c:1981 lookup_one_qstr_excl+0x19d/0x7a0 fs/namei.c:1806 __start_renaming+0x341/0x850 fs/namei.c:3888 filename_renameat2+0x625/0x1230 fs/namei.c:6163 __do_sys_rename fs/namei.c:6232 [inline] __se_sys_rename+0xc5/0x5c0 fs/namei.c:6228 __x64_sys_rename+0x78/0xb0 fs/namei.c:6228 x64_sys_call+0x329/0x3ea0 arch/x86/include/generated/asm/syscalls_64.h:83 do_syscall_x64 arch/x86/entry/syscall_64.c:63 do_syscall_64+0x15d/0x3c0 arch/x86/entry/syscall_64.c:94 entry_SYSCALL_64_after_hwframe+0x77/0x7f The race is between a concurrent open() and rename() of the same path. __d_alloc() only stores the name itself and its terminating NUL, so the rest of the inline buffer (d_shortname, DNAME_INLINE_LEN bytes) is left uninitialized. copy_name(), called from rename(), copies that buffer as a whole, so the uninitialized tail is propagated into the dentry that is being moved. Meanwhile __d_lookup_rcu(), called from open(), is an optimistic lockless lookup: it checks d_name.hash_len first and leaves the seqcount retry to its caller, so it can end up comparing against a dentry whose name a rename is rewriting in place, using a stale (longer) length. The comparison then runs past the terminating NUL and reads bytes of the uninitialized tail, which KMSAN reports. The read is harmless by design: it stays inside the buffer, the name is still NUL-terminated, and the result is thrown away by the seqcount retry. It is not specific to KMSAN either - with CONFIG_DCACHE_WORD_ACCESS enabled the very same bytes are read by read_word_at_a_time(), which is __no_sanitize_or_inline and therefore invisible to KMSAN. KMSAN builds only see the instrumented byte-at-a-time dentry_string_cmp() because CONFIG_DCACHE_WORD_ACCESS is disabled when KMSAN is enabled on x86: commit 7cf8f44a5a1c ("x86: fs: kmsan: disable CONFIG_DCACHE_WORD_ACCESS") Zeroing the inline buffer would hide the report, but it would add a memset() to a hot allocation path just to initialize bytes that are never used as part of a name. Instead, tell KMSAN the inline buffer is initialized: kmsan_unpoison_memory() compiles to nothing unless CONFIG_KMSAN is set, and doing it at allocation time is enough for every dentry, because copy_name() and swap_names() copy the whole buffer and thus propagate its shadow. Reported-by: syzbot+7ff3adde89dd795ad4c4@syzkaller.appspotmail.com Closes: https://syzkaller.appspot.com/bug?extid=7ff3adde89dd795ad4c4 Signed-off-by: Drif Abdelmalek Mohamed Said Changes in v3: - Annotate for KMSAN instead of zeroing, as suggested in review: the read is harmless, so unpoison the inline buffer in __d_alloc() with kmsan_unpoison_memory() (a no-op unless CONFIG_KMSAN) rather than adding a memset() to the dentry allocation path. - Document why only KMSAN builds report this at all: with CONFIG_DCACHE_WORD_ACCESS the same read goes through read_word_at_a_time(), which KMSAN does not instrument. - Rewrite the commit message; the previous one had several truncated lines. Link: https://patch.msgid.link/20260918224204.3056-1-drifabdelmalekmohamedsaid@gmail.com Reviewed-by: Jan Kara Signed-off-by: Christian Brauner (Amutable) --- fs/dcache.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/fs/dcache.c b/fs/dcache.c index 1b1a81f10da6..a66be85f9d01 100644 --- a/fs/dcache.c +++ b/fs/dcache.c @@ -1916,6 +1916,10 @@ static struct dentry *__d_alloc(struct super_block *sb, const struct qstr *name) * be overwriting an internal NUL character */ dentry->d_shortname.string[DNAME_INLINE_LEN-1] = 0; + + /* Racy __d_lookup_rcu() walk may read past the NUL; harmless */ + kmsan_unpoison_memory(dentry->d_shortname.string, DNAME_INLINE_LEN); + if (unlikely(!name)) { name = &slash_name; dname = dentry->d_shortname.string; From 5bfa9f1a9dcb6ecb607adbc1c0226605c972935b Mon Sep 17 00:00:00 2001 From: Andrea Parri Date: Thu, 24 Sep 2026 11:21:39 +0200 Subject: [PATCH 558/694] kprobes: Fix permanent hang when flushing the kprobe optimizer Writing 0 to /proc/sys/debug/kprobes-optimization while a kprobe is jump-optimized never returns. The writer sleeps in D state forever with kprobe_sysctl_mutex held, so any later read or write of that sysctl hangs as well. For example, with vfs_read+9 as an optimizable address in this build: # cd /sys/kernel/tracing # echo 'p:myprobe vfs_read+9' >> kprobe_events # echo 1 > events/kprobes/myprobe/enable # # wait until /sys/kernel/debug/kprobes/list shows [OPTIMIZED] # echo 0 > /proc/sys/debug/kprobes-optimization INFO: task sh:246 blocked for more than 10 seconds. Call Trace: __schedule+0x1176/0x4f70 schedule+0xdc/0x2c0 schedule_timeout+0x17b/0x260 wait_for_completion+0x173/0x3c0 wait_for_kprobe_optimizer_locked+0xbc/0x130 proc_kprobes_optimization_handler+0x156/0x1b0 proc_sys_call_handler+0x324/0x490 vfs_write+0x52d/0xfe0 ksys_write+0xff/0x200 do_syscall_64+0x106/0x630 entry_SYSCALL_64_after_hwframe+0x77/0x7f ... INFO: task cat:265 is blocked on a mutex likely owned by task sh:246. wait_for_kprobe_optimizer_locked() reinitializes optimizer_completion, asks the optimizer thread to flush and sleeps in wait_for_completion(). The thread drains the (un)optimizing lists, but calls complete() only if completion_done() is true, i.e. if the completion is already done, which never happens while someone waits. disarm_all_kprobes() and kprobe_trace_self_tests_init() wait the same way. Calling complete() unconditionally would not be enough: the waiter drops kprobe_mutex while it sleeps, and nothing else serializes the sysctl handler against the debugfs "enabled" file. A second flusher that still finds the lists non-empty, e.g. because a disabled probe is queued for unoptimizing, reinitializes the completion under the first: sysctl write debugfs "enabled" write unoptimize_all_kprobes() wait_for_kprobe_optimizer_locked() init_completion(c) mutex_unlock(&kprobe_mutex) wait_for_completion(c) disarm_all_kprobes() wait_for_kprobe_optimizer_locked() init_completion(c) // c->wait is reset, the first // waiter is off the queue mutex_unlock(&kprobe_mutex) wait_for_completion(c) kprobe_optimizer() complete(c) // wakes the debugfs writer only where c is &optimizer_completion. Lining up the two writes during an optimizer pass loses the sysctl writer this way. Replace the completion with a counter of optimizer passes, bumped at the end of each pass and signalled with wake_up_var_locked(), both under kprobe_mutex. A flusher samples the count and waits with wait_var_event_mutex(), which drops kprobe_mutex only while sleeping, so a new count means a whole pass ran in the meantime. Nothing is reinitialized, so several flushers can sleep in the wait at once. Link: https://lore.kernel.org/all/20260924092142.199198-1-parri.andrea@gmail.com/ Fixes: 73c12f209462 ("kprobes: Use dedicated kthread for kprobe optimizer") Cc: stable@vger.kernel.org Assisted-by: LLM Signed-off-by: Andrea Parri Signed-off-by: Masami Hiramatsu (Google) --- kernel/kprobes.c | 22 ++++++++++++++-------- 1 file changed, 14 insertions(+), 8 deletions(-) diff --git a/kernel/kprobes.c b/kernel/kprobes.c index 6337da5cab9e..4edd8ca5c657 100644 --- a/kernel/kprobes.c +++ b/kernel/kprobes.c @@ -42,6 +42,7 @@ #include #include #include +#include #include #include @@ -526,7 +527,8 @@ enum { OPTIMIZER_ST_FLUSHING = 2, }; -static DECLARE_COMPLETION(optimizer_completion); +/* Bumped at the end of each kprobe_optimizer() pass, under 'kprobe_mutex' */ +static unsigned long optimizer_passes; #define OPTIMIZE_DELAY 5 @@ -654,9 +656,9 @@ static void kprobe_optimizer(void) do_free_cleaned_kprobes(); } - /* Step 5: Kick optimizer again if needed. But if there is a flush requested, */ - if (completion_done(&optimizer_completion)) - complete(&optimizer_completion); + /* Step 5: Wake up flushers, and kick optimizer again if needed. */ + optimizer_passes++; + wake_up_var_locked(&optimizer_passes, &kprobe_mutex); if (!list_empty(&optimizing_list) || !list_empty(&unoptimizing_list)) kick_kprobe_optimizer(); /*normal kick*/ @@ -708,7 +710,8 @@ static void wait_for_kprobe_optimizer_locked(void) lockdep_assert_held(&kprobe_mutex); while (!list_empty(&optimizing_list) || !list_empty(&unoptimizing_list)) { - init_completion(&optimizer_completion); + unsigned long passes = optimizer_passes; + /* * Set state to OPTIMIZER_ST_FLUSHING and wake up the thread if it's * idle. If it's already kicked, it will see the state change. @@ -717,9 +720,12 @@ static void wait_for_kprobe_optimizer_locked(void) OPTIMIZER_ST_FLUSHING) != OPTIMIZER_ST_FLUSHING) wake_up(&kprobe_optimizer_wait); - mutex_unlock(&kprobe_mutex); - wait_for_completion(&optimizer_completion); - mutex_lock(&kprobe_mutex); + /* + * kprobe_optimizer() holds 'kprobe_mutex' for a whole pass, which + * this drops while sleeping, so a new count means a full pass ran. + */ + wait_var_event_mutex(&optimizer_passes, + optimizer_passes != passes, &kprobe_mutex); } } From aa5e44b29ffe4eaa08cc2237fd65bc2596bc023e Mon Sep 17 00:00:00 2001 From: Hui Peng Date: Sat, 19 Sep 2026 20:48:08 +0000 Subject: [PATCH 559/694] autofs: fix sbi->pipe file reference leak in autofs_kill_sb() When autofs_fill_super() fails before clearing AUTOFS_SBI_CATATONIC (for example, when find_get_pid() fails on an invalid pgrp mount option, or when an fs_context is closed before mounting), deactivate_locked_super() invokes autofs_kill_sb() -> autofs_catatonic_mode(sbi). Because AUTOFS_SBI_CATATONIC is still set in sbi->flags, autofs_catatonic_mode() returns early without calling fput(sbi->pipe), permanently leaking the pipe struct file reference. Explicitly release sbi->pipe in autofs_kill_sb() if it is still non-NULL after autofs_catatonic_mode(). Fixes: ebc921ca9b92 ("autofs: copy autofs4 to autofs") Signed-off-by: Hui Peng Link: https://patch.msgid.link/20260919204808.2812930-1-benquike@gmail.com Signed-off-by: Christian Brauner (Amutable) --- fs/autofs/inode.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/fs/autofs/inode.c b/fs/autofs/inode.c index 6b15a3717ba7..066c16f2ea56 100644 --- a/fs/autofs/inode.c +++ b/fs/autofs/inode.c @@ -51,6 +51,10 @@ void autofs_kill_sb(struct super_block *sb) if (sbi) { /* Free wait queues, close pipe */ autofs_catatonic_mode(sbi); + if (sbi->pipe) { + fput(sbi->pipe); + sbi->pipe = NULL; + } put_pid(sbi->oz_pgrp); } From 1c1a342aceec79528b3ba51f376eca2be5928fe7 Mon Sep 17 00:00:00 2001 From: Sagnik Sasmal Date: Thu, 10 Sep 2026 22:44:05 +0000 Subject: [PATCH 560/694] mtd: spinand: Do not update the QE bit on devices without one Commit be0b86c648bf ("mtd: spinand: Gather all the bus interface steps in one single function") moved quad-enable setup into spinand_configure_chip(). The new code only determines whether quad mode is needed when SPINAND_HAS_QE_BIT is set, but calls spinand_init_quad_enable() unconditionally. This clears configuration register bit 0 on devices without a QE bit. That bit is not universally a QE bit. On the Winbond W25N02KV it is H-DIS, which disables the active-low HOLD function. Clearing H-DIS enables HOLD during single and dual I/O operations. If IO3 is not kept high, the flash can pause a command and ignore clock and data. H-DIS is not restored by the FFh reset command, allowing the incorrect state to survive an SoC warm reboot while the flash remains powered. Before the refactoring, spinand_init_quad_enable() returned without touching the configuration register on devices without SPINAND_HAS_QE_BIT. Restore that behavior by only calling the helper when the flag is set. Return zero explicitly once SSDR configuration completes, as all errors are returned immediately. This avoids returning an uninitialized value when neither optional configuration step runs. The regression was reproduced on a JioRouter JIDU6401 with an MT7986 SoC and a W25N02KV. With Linux 6.18.44, sysupgrade failed and the following warm reboot hung in BL2. With this change applied, both sysupgrade and warm reboot completed successfully. Fixes: be0b86c648bf ("mtd: spinand: Gather all the bus interface steps in one single function") Cc: stable@vger.kernel.org Suggested-by: Miquel Raynal Assisted-by: LLM Signed-off-by: Sagnik Sasmal Signed-off-by: Miquel Raynal --- drivers/mtd/nand/spi/core.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/drivers/mtd/nand/spi/core.c b/drivers/mtd/nand/spi/core.c index 43bbe2629b85..95353777d7f7 100644 --- a/drivers/mtd/nand/spi/core.c +++ b/drivers/mtd/nand/spi/core.c @@ -1813,11 +1813,11 @@ static int spinand_configure_chip(struct spinand_device *spinand) spinand->ssdr_op_templates.write_cache->data.buswidth == 4 || spinand->ssdr_op_templates.update_cache->data.buswidth == 4) quad_enable = true; - } - ret = spinand_init_quad_enable(spinand, quad_enable); - if (ret) - return ret; + ret = spinand_init_quad_enable(spinand, quad_enable); + if (ret) + return ret; + } if (spinand->configure_chip) { ret = spinand->configure_chip(spinand, SSDR); @@ -1825,7 +1825,7 @@ static int spinand_configure_chip(struct spinand_device *spinand) return ret; } - return ret; + return 0; } static int spinand_init_flash(struct spinand_device *spinand) From 35d442ed1f86465e49df3119fb898f985186db13 Mon Sep 17 00:00:00 2001 From: Andrea Parri Date: Tue, 22 Sep 2026 16:55:30 +0200 Subject: [PATCH 561/694] bpf: fs/xattr: don't assume the inode is locked in path_unlink/path_rmdir bpf_lsm_has_d_inode_locked() makes the verifier rewrite bpf_[set|remove]_dentry_xattr() to the _locked variants, which assume that the caller already holds the inode's i_rwsem. The path_unlink and path_rmdir hooks are listed, but security_path_unlink() and security_path_rmdir() run before vfs_unlink()/vfs_rmdir() take the victim inode's i_rwsem, so a sleepable BPF LSM program attached to either hook mutates the victim's xattrs without the lock held. Drop the two path hooks from d_inode_locked_hooks so that the verifier keeps the locking bpf_[set|remove]_dentry_xattr() variants, which take the lock themselves. Fixes: 56467292794b8 ("bpf: fs/xattr: Add BPF kfuncs to set and remove xattrs") Cc: stable@vger.kernel.org Signed-off-by: Andrea Parri Link: https://patch.msgid.link/20260922145530.369775-1-parri.andrea@gmail.com Signed-off-by: Christian Brauner (Amutable) --- fs/bpf_fs_kfuncs.c | 4 ---- 1 file changed, 4 deletions(-) diff --git a/fs/bpf_fs_kfuncs.c b/fs/bpf_fs_kfuncs.c index 6cb877267978..357a379ef92a 100644 --- a/fs/bpf_fs_kfuncs.c +++ b/fs/bpf_fs_kfuncs.c @@ -472,10 +472,6 @@ BTF_ID(func, bpf_lsm_inode_rmdir) BTF_ID(func, bpf_lsm_inode_setattr) BTF_ID(func, bpf_lsm_inode_setxattr) BTF_ID(func, bpf_lsm_inode_unlink) -#ifdef CONFIG_SECURITY_PATH -BTF_ID(func, bpf_lsm_path_unlink) -BTF_ID(func, bpf_lsm_path_rmdir) -#endif /* CONFIG_SECURITY_PATH */ BTF_SET_END(d_inode_locked_hooks) bool bpf_lsm_has_d_inode_locked(const struct bpf_prog *prog) From b78b728e21c32ec4c330b299f657fb1eb02dffc2 Mon Sep 17 00:00:00 2001 From: Hao Ge Date: Wed, 23 Sep 2026 14:37:59 +0800 Subject: [PATCH 562/694] netfs: Fix missing alloc tagging of direct mempool allocations Commit 1d78d56c43ef ("netfs: Fix folio_queue ENOMEM in writeback by adding a mempool") added a mempool for the folio_queues and made the request, subrequest and folio_queue allocations distinguish between writeback and everything else. Writeback is part of memory reclaim and must not fail due to ENOMEM, so it allocates under GFP_NOFS through mempool_alloc(), which may dip into the pool's reserve and, if that runs empty, wait for elements to be returned. The GFP_KERNEL paths, which can return -ENOMEM to their callers, invoke the pool's ->alloc() callback directly instead. The direct call, however, skips the alloc_hooks() wrapper that the mempool_alloc() macro provides. The pool callbacks, mempool_alloc_slab() and mempool_kmalloc(), call kmem_cache_alloc_noprof() and kmalloc_noprof() and rely on current->alloc_tag having been set by the caller. With CONFIG_MEM_ALLOC_PROFILING_DEBUG=y this leads to current->alloc_tag not set WARNING: ./include/linux/alloc_tag.h:161 at __alloc_tagging_slab_alloc_hook alloc_tag was not set WARNING: ./include/linux/alloc_tag.h:166 at __alloc_tagging_slab_free_hook at allocation and free time respectively, as reported when reading files on a CIFS mount. The allocations are also missing from /proc/allocinfo. Wrap the direct ->alloc() invocations in alloc_hooks() with a new mempool_alloc_noreserve() helper in include/linux/mempool.h, next to the other alloc_hooks()-wrapped macros such as mempool_alloc(). The GFP_KERNEL paths keep their failable allocation semantics, they just get tagged now. Fixes: 1d78d56c43ef ("netfs: Fix folio_queue ENOMEM in writeback by adding a mempool") Reported-by: Erhard Furtner Closes: https://lore.kernel.org/all/0b004319-9ef7-437c-a4dd-174d6a9a83db@mailbox.org/ Tested-by: Erhard Furtner Suggested-by: Suren Baghdasaryan Cc: stable@vger.kernel.org Signed-off-by: Hao Ge Link: https://patch.msgid.link/20260923063759.34667-1-hao.ge@linux.dev Acked-by: Vlastimil Babka (SUSE) Signed-off-by: Christian Brauner (Amutable) --- fs/netfs/objects.c | 4 ++-- fs/netfs/rolling_buffer.c | 2 +- include/linux/mempool.h | 7 +++++++ 3 files changed, 10 insertions(+), 3 deletions(-) diff --git a/fs/netfs/objects.c b/fs/netfs/objects.c index 7f6a3e912602..ad549daa9c79 100644 --- a/fs/netfs/objects.c +++ b/fs/netfs/objects.c @@ -34,7 +34,7 @@ struct netfs_io_request *netfs_alloc_request(struct address_space *mapping, rreq = mempool_alloc(mempool, gfp); } else { - rreq = mempool->alloc(gfp, mempool->pool_data); + rreq = mempool_alloc_noreserve(mempool, gfp); if (!rreq) return ERR_PTR(-ENOMEM); } @@ -214,7 +214,7 @@ struct netfs_io_subrequest *netfs_alloc_subrequest(struct netfs_io_request *rreq struct kmem_cache *cache = mempool->pool_data; if (rreq->gfp == GFP_KERNEL) - subreq = mempool->alloc(rreq->gfp, mempool->pool_data); + subreq = mempool_alloc_noreserve(mempool, rreq->gfp); else subreq = mempool_alloc(mempool, rreq->gfp); if (!subreq) diff --git a/fs/netfs/rolling_buffer.c b/fs/netfs/rolling_buffer.c index 424e77a9a109..d30d5ef6d86e 100644 --- a/fs/netfs/rolling_buffer.c +++ b/fs/netfs/rolling_buffer.c @@ -29,7 +29,7 @@ struct folio_queue *netfs_folioq_alloc(unsigned int rreq_id, gfp_t gfp, struct folio_queue *fq; if (gfp == GFP_KERNEL) - fq = netfs_folioq_pool.alloc(gfp, netfs_folioq_pool.pool_data); + fq = mempool_alloc_noreserve(&netfs_folioq_pool, gfp); else fq = mempool_alloc(&netfs_folioq_pool, gfp); if (fq) { diff --git a/include/linux/mempool.h b/include/linux/mempool.h index a0fa6d43e0dc..6da502aef2f7 100644 --- a/include/linux/mempool.h +++ b/include/linux/mempool.h @@ -70,6 +70,13 @@ int mempool_alloc_bulk_noprof(struct mempool *pool, void **elem, #define mempool_alloc_bulk(...) \ alloc_hooks(mempool_alloc_bulk_noprof(__VA_ARGS__)) +/* + * Allocate a new element without dipping into the pool's reserves or + * waiting. Returns NULL on failure. + */ +#define mempool_alloc_noreserve(_pool, _gfp) \ + alloc_hooks((_pool)->alloc(_gfp, (_pool)->pool_data)) + void *mempool_alloc_preallocated(struct mempool *pool) __malloc; void mempool_free(void *element, struct mempool *pool); unsigned int mempool_free_bulk(struct mempool *pool, void **elem, From aa650d87a498844060eecb1d336f4d46f4d1b0b3 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?J=C3=A9r=C3=A9my=20Jean?= Date: Tue, 22 Sep 2026 20:27:35 +0000 Subject: [PATCH 563/694] audit: fix exe mark UAF in kill_rules() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit kill_rules() removes mixed AUDIT_DIR and AUDIT_EXE rules when an audit tree is pruned. It drops entry->rule.exe before removing the rule from the RCU-visible filter lists. After a rule has been installed with AUDIT_ADD_RULE, which requires CAP_AUDIT_CONTROL, removing the watched directory can race with another task that is still evaluating the rule. In that case, fsnotify can free the executable mark before the reader reaches audit_mark_compare(), causing a use-after-free. KASAN reports: BUG: KASAN: slab-use-after-free in audit_mark_compare+0x8d/0xa0 Unlink the published rules from the RCU-visible lists and retain them on tree->rules for cleanup. If any rule has an executable mark, wait for a single RCU grace period before removing the marks and scheduling the entries for freeing. Otherwise, call_rcu() already provides the required deferred freeing without a synchronous wait. Cc: stable@vger.kernel.org Fixes: 34d99af52ad4 ("audit: implement audit by executable") Assisted-by: Codex:gpt-5 Signed-off-by: Jérémy Jean Reviewed-by: Ricardo Robaina Tested-by: Ricardo Robaina Reviewed-by: Bradley Morgan Signed-off-by: Paul Moore --- kernel/audit_tree.c | 36 ++++++++++++++++++++++++++---------- 1 file changed, 26 insertions(+), 10 deletions(-) diff --git a/kernel/audit_tree.c b/kernel/audit_tree.c index 1ed19b775912..f2e81be8265e 100644 --- a/kernel/audit_tree.c +++ b/kernel/audit_tree.c @@ -545,21 +545,37 @@ static void kill_rules(struct audit_context *context, struct audit_tree *tree) { struct audit_krule *rule, *next; struct audit_entry *entry; + bool need_sync = false; + + list_for_each_entry_safe(rule, next, &tree->rules, rlist) { + entry = container_of(rule, struct audit_entry, rule); + + if (rule->tree) { + /* not a half-baked one */ + audit_tree_log_remove_rule(context, rule); + rule->tree = NULL; + list_del_rcu(&entry->list); + list_del(&entry->rule.list); + if (entry->rule.exe) + need_sync = true; + } else { + list_del_init(&rule->rlist); + } + } + + if (list_empty(&tree->rules)) + return; + + if (need_sync) + synchronize_rcu(); list_for_each_entry_safe(rule, next, &tree->rules, rlist) { entry = container_of(rule, struct audit_entry, rule); list_del_init(&rule->rlist); - if (rule->tree) { - /* not a half-baked one */ - audit_tree_log_remove_rule(context, rule); - if (entry->rule.exe) - audit_remove_mark(entry->rule.exe); - rule->tree = NULL; - list_del_rcu(&entry->list); - list_del(&entry->rule.list); - call_rcu(&entry->rcu, audit_free_rule_rcu); - } + if (entry->rule.exe) + audit_remove_mark(entry->rule.exe); + call_rcu(&entry->rcu, audit_free_rule_rcu); } } From 9814077275eca36ebf8d510d2f076d235ff9f51a Mon Sep 17 00:00:00 2001 From: Fan Wu Date: Tue, 22 Sep 2026 20:13:48 -0700 Subject: [PATCH 564/694] ipe: fix use-after-free when auditing a newly loaded policy new_policy() audits the policy after ipe_new_policyfs_node() publishes it and drops the new directory's inode lock. A concurrent delete can free the policy while ipe_audit_policy_load() is still using it. Audit the successful load under that lock. Fixes: f44554b5067b ("audit,ipe: add IPE auditing support") Cc: stable@vger.kernel.org Assisted-by: LLM [FW: remove model name according to latest guideline] Signed-off-by: Fan Wu --- security/ipe/fs.c | 8 +++----- security/ipe/policy_fs.c | 3 +++ 2 files changed, 6 insertions(+), 5 deletions(-) diff --git a/security/ipe/fs.c b/security/ipe/fs.c index 076c111c85c8..847a76afb93d 100644 --- a/security/ipe/fs.c +++ b/security/ipe/fs.c @@ -159,18 +159,16 @@ static ssize_t new_policy(struct file *f, const char __user *data, } rc = ipe_new_policyfs_node(p); - if (rc) - goto out; out: kfree(copy); if (rc < 0) { ipe_free_policy(p); ipe_audit_policy_load(ERR_PTR(rc)); - } else { - ipe_audit_policy_load(p); + return rc; } - return (rc < 0) ? rc : len; + + return len; } static const struct file_operations np_fops = { diff --git a/security/ipe/policy_fs.c b/security/ipe/policy_fs.c index 9d92d8a14b13..a7aeb57483c6 100644 --- a/security/ipe/policy_fs.c +++ b/security/ipe/policy_fs.c @@ -481,6 +481,9 @@ int ipe_new_policyfs_node(struct ipe_policy *p) inode_lock(root); p->policyfs = policyfs; root->i_private = p; + /* Only audit signed policies from userspace */ + if (p->pkcs7) + ipe_audit_policy_load(p); inode_unlock(root); return 0; From 2776e9c28513a1c855a94792b292cbcc533418c8 Mon Sep 17 00:00:00 2001 From: Fan Wu Date: Tue, 22 Sep 2026 20:13:49 -0700 Subject: [PATCH 565/694] ipe: protect the dm-verity root hash with RCU ipe_bdev_setintegrity() frees the old root hash when dm-verity publishes a new one on ->preresume, while policy evaluation can still be dereferencing it. Protect the root hash with RCU. The evaluation path already runs under rcu_read_lock(). Fixes: e155858dd995 ("ipe: add support for dm-verity as a trust provider") Cc: stable@vger.kernel.org Assisted-by: LLM [FW: remove model name according to latest guideline] Signed-off-by: Fan Wu --- security/ipe/eval.c | 12 ++++++++---- security/ipe/eval.h | 2 +- security/ipe/hooks.c | 22 +++++++++++++++++----- 3 files changed, 26 insertions(+), 10 deletions(-) diff --git a/security/ipe/eval.c b/security/ipe/eval.c index 21439c5be336..4a7be96c84f0 100644 --- a/security/ipe/eval.c +++ b/security/ipe/eval.c @@ -134,10 +134,14 @@ static bool evaluate_boot_verified(const struct ipe_eval_ctx *const ctx) static bool evaluate_dmv_roothash(const struct ipe_eval_ctx *const ctx, struct ipe_prop *p) { - return !!ctx->ipe_bdev && - !!ctx->ipe_bdev->root_hash && - ipe_digest_eval(p->value, - ctx->ipe_bdev->root_hash); + const struct digest_info *root_hash; + + if (!ctx->ipe_bdev) + return false; + + root_hash = rcu_dereference(ctx->ipe_bdev->root_hash); + + return root_hash && ipe_digest_eval(p->value, root_hash); } #else static bool evaluate_dmv_roothash(const struct ipe_eval_ctx *const ctx, diff --git a/security/ipe/eval.h b/security/ipe/eval.h index fef65a36468c..b26d1147d070 100644 --- a/security/ipe/eval.h +++ b/security/ipe/eval.h @@ -27,7 +27,7 @@ struct ipe_bdev { #ifdef CONFIG_IPE_PROP_DM_VERITY_SIGNATURE bool dm_verity_signed; #endif /* CONFIG_IPE_PROP_DM_VERITY_SIGNATURE */ - struct digest_info *root_hash; + struct digest_info __rcu *root_hash; }; #endif /* CONFIG_IPE_PROP_DM_VERITY */ diff --git a/security/ipe/hooks.c b/security/ipe/hooks.c index 0ae54a880405..d878b52ceec8 100644 --- a/security/ipe/hooks.c +++ b/security/ipe/hooks.c @@ -9,6 +9,7 @@ #include #include #include +#include #include "ipe.h" #include "hooks.h" @@ -232,7 +233,20 @@ void ipe_bdev_free_security(struct block_device *bdev) { struct ipe_bdev *blob = ipe_bdev(bdev); - ipe_digest_free(blob->root_hash); + ipe_digest_free(rcu_access_pointer(blob->root_hash)); +} + +static void ipe_set_dmverity_roothash(struct ipe_bdev *blob, + struct digest_info *info) +{ + struct digest_info *old; + + /* Protected by device-mapper's md->suspend_lock */ + old = rcu_replace_pointer(blob->root_hash, info, true); + if (old) { + synchronize_rcu(); + ipe_digest_free(old); + } } #ifdef CONFIG_IPE_PROP_DM_VERITY_SIGNATURE @@ -280,8 +294,7 @@ int ipe_bdev_setintegrity(struct block_device *bdev, enum lsm_integrity_type typ return -EINVAL; if (!value) { - ipe_digest_free(blob->root_hash); - blob->root_hash = NULL; + ipe_set_dmverity_roothash(blob, NULL); return 0; } @@ -301,8 +314,7 @@ int ipe_bdev_setintegrity(struct block_device *bdev, enum lsm_integrity_type typ info->digest_len = digest->digest_len; - ipe_digest_free(blob->root_hash); - blob->root_hash = info; + ipe_set_dmverity_roothash(blob, info); return 0; err: From aadf655e8ea9a8ffa24bb8d69f504572e8d18181 Mon Sep 17 00:00:00 2001 From: Binbin Deng <18983559317@163.com> Date: Tue, 22 Sep 2026 21:19:19 +0800 Subject: [PATCH 566/694] net/mctp: publish the mctp_dev only after it is initialised mctp_add_dev() links the new mctp_dev into dev->mctp_ptr before mdev->dev is assigned, so a reader that picks the device up in that window dereferences a net_device pointer that is still NULL: KASAN: null-ptr-deref in range [0x00000000000000b0-0x00000000000000b7] Call Trace: ? __pfx_mctp_sendmsg (./include/linux/sockptr.h:49) ? __pfx_mctp_dst_output (net/mctp/route.c:41) ? selinux_socket_sendmsg (security/selinux/hooks.c:5278) ____sys_sendmsg (net/socket.c:775 (discriminator 1) net/socket.c:790 (discriminator 1) net/socket.c:2684 (discriminator 1)) ? __pfx_____sys_sendmsg (net/socket.c:1131) ? __pfx_copy_msghdr_from_user (net/socket.c:2590) ? update_cfs_rq_load_avg (kernel/sched/fair.c:5478) ___sys_sendmsg (net/socket.c:2738) ? __pfx____sys_sendmsg (net/socket.c:2625) ? perf_event_task_tick (./include/linux/rcupdate.h:838) ? sched_tick (kernel/sched/core.c:5818) ? clockevents_program_event (kernel/time/clockevents.c:372) ? fdget (./include/linux/rcupdate.h:873 fs/file.c:1100) __sys_sendmsg (net/socket.c:2770) ? __pfx___sys_sendmsg (net/socket.c:2751) do_syscall_64 (arch/x86/entry/syscall_64.c:63) entry_SYSCALL_64_after_hwframe (arch/x86/entry/entry_64.S:121) Fix by publishing only once the object is fully initialised: __mctp_dev_get() runs under rcu_read_lock() only and hands this object to readers as soon as mctp_ptr is assigned, so they must never observe mdev->dev == NULL. Fixes: 583be982d934 ("mctp: Add device handling and netlink interface") Signed-off-by: Binbin Deng <18983559317@163.com> Reviewed-by: Eric Dumazet Link: https://patch.msgid.link/729830b8.130.1a0c945579d.Coremail.18983559317@163.com Signed-off-by: Jakub Kicinski --- net/mctp/device.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/net/mctp/device.c b/net/mctp/device.c index 822120e860c8..9f76ff092559 100644 --- a/net/mctp/device.c +++ b/net/mctp/device.c @@ -344,13 +344,13 @@ static struct mctp_dev *mctp_add_dev(struct net_device *dev) mdev->net = mctp_default_net(dev_net(dev)); - /* associate to net_device */ refcount_set(&mdev->refs, 1); - rcu_assign_pointer(dev->mctp_ptr, mdev); - dev_hold(dev); mdev->dev = dev; + /* Associate to net_device, publish the mdev. */ + rcu_assign_pointer(dev->mctp_ptr, mdev); + return mdev; } From 0bb8eab29b55045281a4963d4557f2776cdf8cfa Mon Sep 17 00:00:00 2001 From: Jamal Hadi Salim Date: Thu, 24 Sep 2026 04:32:49 -0400 Subject: [PATCH 567/694] net/sched: cls_api: reclaim an empty proto on the error path Two racing tc filter add requests on the same chain/prio of an unlocked classifier both run change() on the shared proto and both can fail: the winner's tcf_chain_tp_delete_empty() attempt gives up because the loser's handle is still in the idr, and the loser's error path drops only its own reference without a second reclamation attempt. The empty proto stays linked in the chain, holding the chain reference, a block reference and the classifier module reference until the chain or block is torn down. Reclaim the proto on the error path of any failed request that holds a proto reference. The reclamation is emptiness-gated: tcf_chain_tp_delete_empty() unlinks the proto only when delete_empty() admits it is empty, so a live shared proto is never unlinked. A proto the request created is reclaimed unconditionally - it is the only owner, so marking it for deletion is safe even without a delete_empty callback. Classifiers without one (the check marks the proto unconditionally) are rtnl-serialized, so the raced window this guard closes cannot arise for them. This is a follow-up to commit d4e359b3608a ("net/sched: cls_api: fix teardown of an adopted proto on insert-race loss"), which stopped the loser of the insert race from unlinking the winner's live proto but left the empty-proto residual in place. Conditions to recreate: - CONFIG_NET_CLS_FLOWER=y; veth pair - tc qdisc add dev veth0 ingress - two concurrent `tc filter add dev veth0 ingress protocol ip pref 1 flower skip_sw ... action drop` (both fail in fl_hw_replace_filter after publishing their handle in the idr); repeat in a loop - an empty flower tp stays linked after both requests fail; visible as a bare `filter protocol ip pref 1 flower chain 0` header in `tc filter show` with no filter entries - CAP_NET_ADMIN (namespace-local via unshare -Urn suffices) Fixes: 8b64678e0af8 ("net: sched: refactor tp insert/delete for concurrent execution") Reported-by: Sashiko (nipa) Closes: https://netdev-ai.bots.linux.dev/sashiko/#/patchset/20260805134049.927864-1-victor@mojatatu.com Signed-off-by: Jamal Hadi Salim Reviewed-by: Simon Horman Link: https://patch.msgid.link/QDISC-JCOT.v1.20260910090924@mojatatu.com Signed-off-by: Jakub Kicinski --- net/sched/cls_api.c | 15 ++++++++++++++- 1 file changed, 14 insertions(+), 1 deletion(-) diff --git a/net/sched/cls_api.c b/net/sched/cls_api.c index c47d2ee13641..a9f54988561f 100644 --- a/net/sched/cls_api.c +++ b/net/sched/cls_api.c @@ -2463,7 +2463,20 @@ static int tc_new_tfilter(struct sk_buff *skb, struct nlmsghdr *n, } errout: - if (err && tp_state == TP_CREATED) + if (err && !IS_ERR_OR_NULL(tp) && + (tp_state == TP_CREATED || tp->ops->delete_empty)) + /* + * The request is dropping its reference to tp. If it was + * the last user (the idr is empty now), reclaim the proto. + * A tp this request created is reclaimed unconditionally: + * it is the only owner, so marking it for deletion is + * safe. Otherwise only classifiers with a delete_empty + * callback are reclaimed -- the callback admits an empty + * proto only, so a live shared proto is never unlinked. + * Classifiers without one (deleting is set + * unconditionally) are rtnl-serialized, so the raced + * window this guard closes cannot arise for them. + */ tcf_chain_tp_delete_empty(chain, tp, rtnl_held, NULL); errout_tp: if (chain) { From 0e7fe2b0ca45343b53345edc174898b448ccdaa4 Mon Sep 17 00:00:00 2001 From: Nicolai Buchwitz Date: Thu, 24 Sep 2026 12:19:21 +0200 Subject: [PATCH 568/694] net: bcmgenet: allocate RX buffers as page fragments Since the page_pool conversion every RX buffer is a whole page and the skb truesize is the page, although the hardware writes at most 2 KiB of it. On 64 KiB pages a packet therefore counts 65792 bytes against the socket buffer where it used to count 2752. As a result a UDP socket with the default buffer starts to drop after three packets, and the RX rings pin 16 MiB for 512 KiB of buffers. Fix this and allocate the buffers as page fragments, so the truesize is what a packet occupies. 4 KiB pages stay one buffer per page. Sync each buffer in the refill path, as page_pool can only sync a whole page on recycle. On 4 KiB pages that is twice what the hardware wrote. Fixes: 7bc054c2d4ed ("net: bcmgenet: convert RX path to page_pool") Reported-by: Karl Mehltretter Closes: https://lore.kernel.org/all/20260924065839.56793-1-kmehltretter@gmail.com/ Signed-off-by: Nicolai Buchwitz Reviewed-by: Simon Horman Link: https://patch.msgid.link/20260924101922.2675127-1-nb@tipi-net.de Signed-off-by: Jakub Kicinski --- .../net/ethernet/broadcom/genet/bcmgenet.c | 29 ++++++++++++++----- .../net/ethernet/broadcom/genet/bcmgenet.h | 2 ++ 2 files changed, 23 insertions(+), 8 deletions(-) diff --git a/drivers/net/ethernet/broadcom/genet/bcmgenet.c b/drivers/net/ethernet/broadcom/genet/bcmgenet.c index 21668e41b696..f725d26e6020 100644 --- a/drivers/net/ethernet/broadcom/genet/bcmgenet.c +++ b/drivers/net/ethernet/broadcom/genet/bcmgenet.c @@ -57,6 +57,9 @@ */ #define GENET_RSB_PAD (sizeof(struct status_64) + 2) +/* RX buffer plus the skb_shared_info napi_build_skb() places behind it */ +#define GENET_RX_BUF_SIZE SKB_HEAD_ALIGN(RX_BUF_LENGTH) + /* Tx/Rx DMA register offset, skip 256 descriptors */ #define WORDS_PER_BD(p) (p->hw_params->words_per_bd) #define DMA_DESC_SIZE (WORDS_PER_BD(priv) * sizeof(u32)) @@ -2254,11 +2257,12 @@ static int bcmgenet_rx_refill(struct bcmgenet_rx_ring *ring, struct enet_cb *cb) { struct bcmgenet_priv *priv = ring->priv; + unsigned int size = GENET_RX_BUF_SIZE; + unsigned int offset; dma_addr_t mapping; struct page *page; - page = page_pool_alloc_pages(ring->page_pool, - GFP_ATOMIC); + page = page_pool_dev_alloc(ring->page_pool, &offset, &size); if (!page) { priv->mib.alloc_rx_buff_failed++; netif_err(priv, rx_err, priv->dev, @@ -2267,9 +2271,13 @@ static int bcmgenet_rx_refill(struct bcmgenet_rx_ring *ring, } /* page_pool handles DMA mapping via PP_FLAG_DMA_MAP */ - mapping = page_pool_get_dma_addr(page); + mapping = page_pool_get_dma_addr(page) + offset; + dma_sync_single_for_device(&priv->pdev->dev, mapping, RX_BUF_LENGTH, + DMA_FROM_DEVICE); cb->rx_page = page; + cb->rx_offset = offset; + cb->rx_size = size; dmadesc_set_addr(priv, cb->bd_addr, mapping); return 0; @@ -2323,6 +2331,7 @@ static unsigned int bcmgenet_desc_rx(struct bcmgenet_rx_ring *ring, while ((rxpktprocessed < rxpkttoprocess) && (rxpktprocessed < budget)) { + unsigned int rx_offset, rx_size; struct status_64 *status; struct page *rx_page; void *hard_start; @@ -2332,6 +2341,8 @@ static unsigned int bcmgenet_desc_rx(struct bcmgenet_rx_ring *ring, /* Save the received page before refilling */ rx_page = cb->rx_page; + rx_offset = cb->rx_offset; + rx_size = cb->rx_size; if (bcmgenet_rx_refill(ring, cb)) { BCMGENET_STATS64_INC(stats, dropped); @@ -2341,10 +2352,10 @@ static unsigned int bcmgenet_desc_rx(struct bcmgenet_rx_ring *ring, /* Sync the full buffer; the HW may have written anywhere * up to RX_BUF_LENGTH. */ - page_pool_dma_sync_for_cpu(ring->page_pool, rx_page, 0, + page_pool_dma_sync_for_cpu(ring->page_pool, rx_page, rx_offset, RX_BUF_LENGTH); - hard_start = page_address(rx_page); + hard_start = page_address(rx_page) + rx_offset; status = (struct status_64 *)hard_start; dma_length_status = status->length_status; @@ -2410,7 +2421,7 @@ static unsigned int bcmgenet_desc_rx(struct bcmgenet_rx_ring *ring, /* Build SKB from the page - data starts at hard_start, * frame begins after RSB(64) + pad(2) = 66 bytes. */ - skb = napi_build_skb(hard_start, PAGE_SIZE); + skb = napi_build_skb(hard_start, rx_size); if (unlikely(!skb)) { BCMGENET_STATS64_INC(stats, dropped); page_pool_put_full_page(ring->page_pool, rx_page, @@ -2762,14 +2773,16 @@ static void bcmgenet_init_tx_ring(struct bcmgenet_priv *priv, static int bcmgenet_rx_ring_create_pool(struct bcmgenet_priv *priv, struct bcmgenet_rx_ring *ring) { + /* Buffers share a page. bcmgenet_rx_refill() syncs each one for the + * device, PP_FLAG_DMA_SYNC_DEV would sync the whole page. + */ struct page_pool_params pp_params = { .order = 0, - .flags = PP_FLAG_DMA_MAP | PP_FLAG_DMA_SYNC_DEV, + .flags = PP_FLAG_DMA_MAP, .pool_size = ring->size, .nid = NUMA_NO_NODE, .dev = &priv->pdev->dev, .dma_dir = DMA_FROM_DEVICE, - .max_len = RX_BUF_LENGTH, }; int err; diff --git a/drivers/net/ethernet/broadcom/genet/bcmgenet.h b/drivers/net/ethernet/broadcom/genet/bcmgenet.h index 22a958ba9902..86f2aed20dbe 100644 --- a/drivers/net/ethernet/broadcom/genet/bcmgenet.h +++ b/drivers/net/ethernet/broadcom/genet/bcmgenet.h @@ -470,6 +470,8 @@ struct bcmgenet_rx_stats64 { struct enet_cb { struct sk_buff *skb; struct page *rx_page; + unsigned int rx_offset; + unsigned int rx_size; void __iomem *bd_addr; DEFINE_DMA_UNMAP_ADDR(dma_addr); DEFINE_DMA_UNMAP_LEN(dma_len); From a7bfaba4823e3c165bb2004c74eff7c096672bc7 Mon Sep 17 00:00:00 2001 From: Qishuai Liu Date: Wed, 23 Sep 2026 13:20:48 +0000 Subject: [PATCH 569/694] ipv6: fix prefix route expiry in modify_prefix_route() modify_prefix_route() is given the lifetime in clock_t relative to now, but fib6_set_expires() wants an absolute jiffies value. So when a permanent address is changed to a finite valid_lft, the prefix route ends up already expired and GC removes it. Steps to reproduce: ip link add dummy9 type dummy ip link set dummy9 up ip -6 addr add 2001:db8:9::1/64 dev dummy9 ip -6 addr change 2001:db8:9::1/64 dev dummy9 valid_lft 3600 preferred_lft 3600 ip -6 route show dev dummy9 # expires is negative Fixes: 8308f3ff1753 ("net/ipv6: Add support for specifying metric of connected routes") Signed-off-by: Qishuai Liu Reviewed-by: Hangbin Liu Reviewed-by: Ido Schimmel Link: https://patch.msgid.link/20260923132048.3272878-1-lqs@lqs.me Signed-off-by: Jakub Kicinski --- net/ipv6/addrconf.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c index 9d89be7e0544..c90ee6dd7446 100644 --- a/net/ipv6/addrconf.c +++ b/net/ipv6/addrconf.c @@ -4874,7 +4874,7 @@ static int modify_prefix_route(struct net *net, struct inet6_ifaddr *ifp, fib6_clean_expires(f6i); fib6_may_remove_gc_list(net, f6i); } else { - fib6_set_expires(f6i, expires); + fib6_set_expires(f6i, jiffies + clock_t_to_jiffies(expires)); fib6_add_gc_list(f6i); } From 12c1f6e03f944e399bd2c88441dca5dc702b95a5 Mon Sep 17 00:00:00 2001 From: Sean Christopherson Date: Wed, 23 Sep 2026 09:37:20 -0700 Subject: [PATCH 570/694] KVM: SEV: Free have_run_cpus during VM destruction even if VM is no longer SEV Unconditionally free SEV's "have run CPUs" cpumask in the VM destroy path, i.e. even for what appear to be non-SEV VMs, as an SEV VM becomes a non-SEV VM if its state is intra-host migrated. Alternatively, the mask could be freed in sev_migrate_from() when "converting" the source VM, but that gets annoying because ideally KVM would nullify the mask to guard against UAF, and nullifying the mask would need be conditioned on CPUMASK_OFFSTACK=y. Freeing the mask during sev_migrate_from() is also not robust against other KVM bugs, though that's kind of a moot point since any such bugs would show up even if sev->active is never set. I.e. KVM must get that side of things correct. But, that's not a great reason to add more code just to make things marginally less robust. Fixes: 6f38f8c57464 ("KVM: SVM: Flush cache only on CPUs running SEV guest") Cc: stable@vger.kernel.org Reported-by: Stefan Teodorescu Signed-off-by: Sean Christopherson Message-ID: <20260923163721.1584779-2-seanjc@google.com> Signed-off-by: Paolo Bonzini --- arch/x86/kvm/svm/sev.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/arch/x86/kvm/svm/sev.c b/arch/x86/kvm/svm/sev.c index 5705723f1f41..cdc1c04f60da 100644 --- a/arch/x86/kvm/svm/sev.c +++ b/arch/x86/kvm/svm/sev.c @@ -2980,13 +2980,17 @@ void sev_vm_destroy(struct kvm *kvm) struct list_head *head = &sev->regions_list; struct list_head *pos, *q; + /* + * Free the mask even if the VM is not *currently* an SEV VM, as it may + * have been an SEV VM prior to intra-host migration. + */ + free_cpumask_var(sev->have_run_cpus); + if (!sev_guest(kvm)) return; WARN_ON(!list_empty(&sev->mirror_vms)); - free_cpumask_var(sev->have_run_cpus); - /* * If this is a mirror VM, remove it from the owner's list of a mirrors * and skip ASID cleanup (the ASID is tied to the lifetime of the owner). From 93de2a6a4b91b72607136dd656edf03fb399d27f Mon Sep 17 00:00:00 2001 From: Sean Christopherson Date: Wed, 23 Sep 2026 09:37:21 -0700 Subject: [PATCH 571/694] KVM: SEV: Do cache maintenance on the source VM during intra-host migration Manually perform cache maintenance on the source VM during intra-host migration to ensure no stale data is left in CPU caches after the VM is destroyed. Because the source VM is "converted" to a non-SEV VM, KVM's memory reclaim flows won't trigger cache maintenance, e.g. when all guest memory is reclaimed in response to detaching from the mmu_notifier. Note, relying on the destination VM to do cache maintenance isn't an option as KVM doesn't require identical guest memory configurations, i.e. the source VM may have access to memory that the destination VM does not. Enforcing equivalent memory configurations is infeasible, as it would require a *deep* comparison of memslots, e.g. to verify that not only are the memslot identical, but what the memslots point at is also identical. Fixes: b56639318bb2 ("KVM: SEV: Add support for SEV intra host migration") Cc: stable@vger.kernel.org Reported-by: Stefan Teodorescu Signed-off-by: Sean Christopherson Message-ID: <20260923163721.1584779-3-seanjc@google.com> Signed-off-by: Paolo Bonzini --- arch/x86/kvm/svm/sev.c | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/arch/x86/kvm/svm/sev.c b/arch/x86/kvm/svm/sev.c index cdc1c04f60da..63eb2155a774 100644 --- a/arch/x86/kvm/svm/sev.c +++ b/arch/x86/kvm/svm/sev.c @@ -2048,6 +2048,12 @@ static void sev_migrate_from(struct kvm *dst_kvm, struct kvm *src_kvm) src->pages_locked = 0; src->es_active = false; + /* + * Do cache maintenance on the source VM as it is no longer an SEV VM, + * i.e. memory reclaim flows won't trigger cache maintenance on the VM. + */ + sev_writeback_caches(src_kvm); + list_cut_before(&dst->regions_list, &src->regions_list, &src->regions_list); mutex_lock(&sev_mirror_lock); @@ -2187,6 +2193,10 @@ int sev_vm_move_enc_context_from(struct kvm *kvm, unsigned int source_fd) * the set of CPUs from the source. If a CPU was used to run a vCPU in * the source VM but is never used for the destination VM, then the CPU * can only have cached memory that was accessible to the source VM. + * Furthermore, KVM *must* perform cache maintenance on the source VM, + * as the source VM may have access to memory that the destination VM + * does not, i.e. KVM could skip flushes if memory is reclaimed from + * the old VM but not the new VM. */ if (!zalloc_cpumask_var(&dst_sev->have_run_cpus, GFP_KERNEL_ACCOUNT)) { ret = -ENOMEM; From 8264c0d73f07f6c3e6768d5a41b56c621f912a37 Mon Sep 17 00:00:00 2001 From: Sean Christopherson Date: Mon, 21 Sep 2026 10:44:39 -0700 Subject: [PATCH 572/694] KVM: Reject attempts to lock all vCPUs if vCPU creation is in-progress Reject locking of all vCPUs if vCPU creation is in-progress, i.e. if the number of "created" vCPUs doesn't match the number of "onlined" vCPUs. It's simply not possible to guarantee that KVM has truly locked all vCPUs if one or more vCPUs are actively being created. Holding kvm->lock does prevent in-flight vCPUs from being fully onlined, but it's infeasible for common KVM to know whether or not that provides sufficient protection. In practice, this is likely a minor bug fix for the ARM and RISC-V usage of kvm_trylock_all_vcpus(), and a glorified nop for everything else. E.g. ARM's kvm_timer_vcpu_init() can race kvm_vm_ioctl_set_counter_offset() with respect to observing KVM_ARCH_FLAG_VM_COUNTER_OFFSET. Opportunistically drop x86's existing manual checks on vCPU creation being in-progress as all of x86's checks immediately precede or follow locking of all vCPUs. Leave arm64 and RISC-V alone for the moment, as their checks aren't as obviously redundant/equivalent. Signed-off-by: Sean Christopherson Tested-by: Jean-Christophe Guillain Tested-by: Naveen N Rao (AMD) Message-ID: <20260921174445.911676-2-seanjc@google.com> Signed-off-by: Paolo Bonzini --- arch/x86/kvm/svm/sev.c | 10 ---------- arch/x86/kvm/vmx/tdx.c | 5 ----- virt/kvm/kvm_main.c | 6 ++++++ 3 files changed, 6 insertions(+), 15 deletions(-) diff --git a/arch/x86/kvm/svm/sev.c b/arch/x86/kvm/svm/sev.c index 63eb2155a774..6aa86a78711e 100644 --- a/arch/x86/kvm/svm/sev.c +++ b/arch/x86/kvm/svm/sev.c @@ -1125,9 +1125,6 @@ static int sev_launch_update_vmsa(struct kvm *kvm, struct kvm_sev_cmd *argp) if (!sev_es_guest(kvm)) return -ENOTTY; - if (kvm_is_vcpu_creation_in_progress(kvm)) - return -EBUSY; - ret = kvm_lock_all_vcpus(kvm); if (ret) return ret; @@ -2121,10 +2118,6 @@ static int sev_check_source_vcpus(struct kvm *dst, struct kvm *src) struct kvm_vcpu *src_vcpu; unsigned long i; - if (kvm_is_vcpu_creation_in_progress(src) || - kvm_is_vcpu_creation_in_progress(dst)) - return -EBUSY; - if (!sev_es_guest(src)) return 0; @@ -2520,9 +2513,6 @@ static int snp_launch_update_vmsa(struct kvm *kvm, struct kvm_sev_cmd *argp) unsigned long i; int ret; - if (kvm_is_vcpu_creation_in_progress(kvm)) - return -EBUSY; - ret = kvm_lock_all_vcpus(kvm); if (ret) return ret; diff --git a/arch/x86/kvm/vmx/tdx.c b/arch/x86/kvm/vmx/tdx.c index 6c842e9191a5..00850a3a77d6 100644 --- a/arch/x86/kvm/vmx/tdx.c +++ b/arch/x86/kvm/vmx/tdx.c @@ -2728,11 +2728,6 @@ static tdx_vm_state_guard_t tdx_acquire_vm_state_locks(struct kvm *kvm) mutex_lock(&kvm->lock); - if (kvm->created_vcpus != atomic_read(&kvm->online_vcpus)) { - r = -EBUSY; - goto out_err; - } - r = kvm_lock_all_vcpus(kvm); if (r) goto out_err; diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index 85f42289748d..aad8cf00a1c8 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -1363,6 +1363,9 @@ int kvm_trylock_all_vcpus(struct kvm *kvm) lockdep_assert_held(&kvm->lock); + if (kvm_is_vcpu_creation_in_progress(kvm)) + return -EBUSY; + kvm_for_each_vcpu(i, vcpu, kvm) if (!mutex_trylock_nest_lock(&vcpu->mutex, &kvm->lock)) goto out_unlock; @@ -1386,6 +1389,9 @@ int kvm_lock_all_vcpus(struct kvm *kvm) lockdep_assert_held(&kvm->lock); + if (kvm_is_vcpu_creation_in_progress(kvm)) + return -EBUSY; + kvm_for_each_vcpu(i, vcpu, kvm) { r = mutex_lock_killable_nest_lock(&vcpu->mutex, &kvm->lock); if (r) From e3995b05e95d03953954dd60ba803ebfa7755253 Mon Sep 17 00:00:00 2001 From: Sean Christopherson Date: Mon, 21 Sep 2026 10:44:40 -0700 Subject: [PATCH 573/694] KVM: arm64: vgic: Rely on vCPU creation check in "trylock all vCPUs" Now that KVM's APIs for locking all vCPUs return -EBUSY if vCPU creation is in-progress, drop the manual check for the same from vGIC creation, and update the comments accordingly. Note, while KVM arm64 guards many vGIC operations with its arch-specific config_lock, holding kvm->lock is sufficient to guarantee a stable result for "is vCPU creation in-progress". So, no functional change intended. Note #2, the open coded check in vgic_init() is racy when called without kvm->lock held, e.g. via vgic_lazy_init(). I.e. that check needs to stay open coded to avoid triggering a lockdep assert. Whether or not the race is "fine" is a problem for a different day. Signed-off-by: Sean Christopherson Tested-by: Jean-Christophe Guillain Tested-by: Naveen N Rao (AMD) Message-ID: <20260921174445.911676-3-seanjc@google.com> Signed-off-by: Paolo Bonzini --- arch/arm64/kvm/vgic/vgic-init.c | 12 ++++-------- 1 file changed, 4 insertions(+), 8 deletions(-) diff --git a/arch/arm64/kvm/vgic/vgic-init.c b/arch/arm64/kvm/vgic/vgic-init.c index 4012df6002ea..a58575df36e9 100644 --- a/arch/arm64/kvm/vgic/vgic-init.c +++ b/arch/arm64/kvm/vgic/vgic-init.c @@ -97,6 +97,9 @@ int kvm_vgic_create(struct kvm *kvm, u32 type) /* * - Acquiring the vCPU mutex for every *online* vCPU to prevent * concurrent vCPU ioctls for vCPUs already visible to userspace. + * This also ensures KVM isn't in the middle of creating a vCPU, + * i.e. that there are no vCPUs that have been created but aren't + * yet fully online. */ ret = -EBUSY; if (kvm_trylock_all_vcpus(kvm)) @@ -105,18 +108,11 @@ int kvm_vgic_create(struct kvm *kvm, u32 type) /* * - Taking the config_lock which protects VGIC data structures such * as the per-vCPU arrays of private IRQs (SGIs, PPIs). - */ - mutex_lock(&kvm->arch.config_lock); - - /* - * - Bailing on the entire thing if a vCPU is in the middle of creation, - * dropped the kvm->lock, but hasn't reached kvm_arch_vcpu_create(). * * The whole combination of this guarantees that no vCPU can get into * KVM with a VGIC configuration inconsistent with the VM's VGIC. */ - if (kvm->created_vcpus != atomic_read(&kvm->online_vcpus)) - goto out_unlock; + mutex_lock(&kvm->arch.config_lock); if (irqchip_in_kernel(kvm)) { ret = -EEXIST; From 10f4082b754f18ed82f94d47f0e959ef80038895 Mon Sep 17 00:00:00 2001 From: Sean Christopherson Date: Mon, 21 Sep 2026 10:44:41 -0700 Subject: [PATCH 574/694] KVM: RISC-V: Use kvm_is_vcpu_creation_in_progress() instead of open-coded equivalent Use kvm_is_vcpu_creation_in_progress() instead of an open-coded equivalent during AIA initialization. Unlike similar vGIC code in arm64, the relevant RISC-V code runs under kvm->lock, i.e. can use the standard API without hitting lockdep false positive. No functional change intended. Signed-off-by: Sean Christopherson Tested-by: Jean-Christophe Guillain Tested-by: Naveen N Rao (AMD) Message-ID: <20260921174445.911676-4-seanjc@google.com> Signed-off-by: Paolo Bonzini --- arch/riscv/kvm/aia_device.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/riscv/kvm/aia_device.c b/arch/riscv/kvm/aia_device.c index efc7c0bcfba9..97833a04268a 100644 --- a/arch/riscv/kvm/aia_device.c +++ b/arch/riscv/kvm/aia_device.c @@ -237,7 +237,7 @@ static int aia_init(struct kvm *kvm) return -EBUSY; /* We might be in the middle of creating a VCPU? */ - if (kvm->created_vcpus != atomic_read(&kvm->online_vcpus)) + if (kvm_is_vcpu_creation_in_progress(kvm)) return -EBUSY; /* Number of sources should be less than or equals number of IDs */ From 6418b715d5a24fa952e8c4baa587dd812dd4765b Mon Sep 17 00:00:00 2001 From: Sean Christopherson Date: Mon, 21 Sep 2026 10:44:42 -0700 Subject: [PATCH 575/694] KVM: Protect all of kvm_vm_ioctl_create_vcpu() with kvm->lock When creating a vCPU, don't drop kvm->lock to when doing the bulk of actual vCPU creation, as allowing multiple vCPUs to be created in parallel adds significant complexity in KVM (as evidenced by the many related bugs), and all known VMMs fully serialize vCPU creation. Remove all manually locking of kvm->lock from kvm_arch_vcpu_{post,}create() for obvious reasons. For many years, "everyone" has assumed that dropping kvm->lock was done for performance reasons optimization, e.g. to allow userspace to create all vCPUs concurrently for latency purposes. But as above, no known VMM does that. Looking at the history of this code, before commit 11ec28047118 ("KVM: Convert vm lock to a mutex"), kvm->lock was a spinlock. I.e. KVM *had* to drop kvm->lock when doing the bulk of vCPU creation, otherwise KVM couldn't do normal memory allocations. When kvm->lock got turned into a mutex for unrelated reasons, no one took advantage updated of the change to simplify vCPU creation. And 19 years later, everyone just assumed that KVM continued to deal with the complexity for performance reasons. Furthermore, naively parallelizing vCPU creation in userspace is likely a net negative due to the overheads of task creation. Unless a VMM carefully avoids the extra overhead related to parallelization, e.g. spawns each vCPU's thread before creating the vCPU, creating vCPUs concurrently is a net *negative* up until about ~64 vCPUs, after which the times are a wash. The absolute speed of light _is_ faster if KVM doesn't hold kvm-lock, but at vCPU counts of ~16 or less, it's probably in the noise when considering total VM creation time, as the added latency is less than 1ms up until 16 or so vCPUs. On top of all that, KVM has had a *lot* of fatal bugs (most often found by syzkaller) related to vCPUs being created while trying to do per-VM operations (basically, see every flow that locks all vCPUs). I.e. the parallel vCPU creation "support" is actively harmful as the only "use case" is for misbehaving userspace to exploit KVM bugs. Serializing vCPU creation will allow reverting commit 97d65b544f48 ("KVM: Check for duplicate vcpu_id as early as possible"), which had "minor" math error: the worst case scenario isn't "256 bytes per VM", it's "256 unsigned longs per VM", i.e. 2048 bytes per VM, which doubles the size of each VM and pushes several architectures into order-1 allocations. Signed-off-by: Sean Christopherson Tested-by: Jean-Christophe Guillain Tested-by: Naveen N Rao (AMD) Message-ID: <20260921174445.911676-5-seanjc@google.com> Signed-off-by: Paolo Bonzini --- arch/powerpc/kvm/book3s_hv.c | 2 -- arch/s390/kvm/s390/s390.c | 5 +---- virt/kvm/kvm_main.c | 22 +++++----------------- 3 files changed, 6 insertions(+), 23 deletions(-) diff --git a/arch/powerpc/kvm/book3s_hv.c b/arch/powerpc/kvm/book3s_hv.c index dbac3573b2c8..de05d721edc4 100644 --- a/arch/powerpc/kvm/book3s_hv.c +++ b/arch/powerpc/kvm/book3s_hv.c @@ -3058,7 +3058,6 @@ static int kvmppc_core_vcpu_create_hv(struct kvm_vcpu *vcpu) init_waitqueue_head(&vcpu->arch.cpu_run); - mutex_lock(&kvm->lock); vcore = NULL; err = -EINVAL; if (cpu_has_feature(CPU_FTR_ARCH_300)) { @@ -3091,7 +3090,6 @@ static int kvmppc_core_vcpu_create_hv(struct kvm_vcpu *vcpu) mutex_unlock(&kvm->arch.mmu_setup_lock); } } - mutex_unlock(&kvm->lock); if (!vcore) return err; diff --git a/arch/s390/kvm/s390/s390.c b/arch/s390/kvm/s390/s390.c index efd77b9cd36c..e4912db67adc 100644 --- a/arch/s390/kvm/s390/s390.c +++ b/arch/s390/kvm/s390/s390.c @@ -3579,12 +3579,11 @@ void kvm_arch_vcpu_put(struct kvm_vcpu *vcpu) void kvm_arch_vcpu_postcreate(struct kvm_vcpu *vcpu) { - mutex_lock(&vcpu->kvm->lock); preempt_disable(); vcpu->arch.sie_block->epoch = vcpu->kvm->arch.epoch; vcpu->arch.sie_block->epdx = vcpu->kvm->arch.epdx; preempt_enable(); - mutex_unlock(&vcpu->kvm->lock); + if (!kvm_is_ucontrol(vcpu->kvm)) { vcpu->arch.gmap = vcpu->kvm->arch.gmap; sca_add_vcpu(vcpu); @@ -3757,13 +3756,11 @@ static int kvm_s390_vcpu_setup(struct kvm_vcpu *vcpu) kvm_s390_vcpu_pci_setup(vcpu); - mutex_lock(&vcpu->kvm->lock); if (kvm_s390_pv_is_protected(vcpu->kvm)) { rc = kvm_s390_pv_create_cpu(vcpu, &uvrc, &uvrrc); if (rc) kvm_s390_vcpu_unsetup_cmma(vcpu); } - mutex_unlock(&vcpu->kvm->lock); return rc; } diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index aad8cf00a1c8..cae0a9cdc33e 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -4188,6 +4188,8 @@ static int kvm_vm_ioctl_create_vcpu(struct kvm *kvm, unsigned long id) struct kvm_vcpu *vcpu; struct page *page; + guard(mutex)(&kvm->lock); + /* * KVM tracks vCPU IDs as 'int', be kind to userspace and reject * too-large values instead of silently truncating. @@ -4200,26 +4202,18 @@ static int kvm_vm_ioctl_create_vcpu(struct kvm *kvm, unsigned long id) if (id >= KVM_MAX_VCPU_IDS) return -EINVAL; - mutex_lock(&kvm->lock); - if (kvm->created_vcpus >= kvm->max_vcpus) { - mutex_unlock(&kvm->lock); + if (kvm->created_vcpus >= kvm->max_vcpus) return -EINVAL; - } - if (test_bit(id, kvm->vcpu_ids)) { - mutex_unlock(&kvm->lock); + if (test_bit(id, kvm->vcpu_ids)) return -EEXIST; - } r = kvm_arch_vcpu_precreate(kvm, id); - if (r) { - mutex_unlock(&kvm->lock); + if (r) return r; - } kvm->created_vcpus++; __set_bit(id, kvm->vcpu_ids); - mutex_unlock(&kvm->lock); vcpu = kmem_cache_zalloc(kvm_vcpu_cache, GFP_KERNEL_ACCOUNT); if (!vcpu) { @@ -4250,8 +4244,6 @@ static int kvm_vm_ioctl_create_vcpu(struct kvm *kvm, unsigned long id) goto arch_vcpu_destroy; } - mutex_lock(&kvm->lock); - if (WARN_ON_ONCE(kvm_get_vcpu_by_id(kvm, id))) { r = -EEXIST; goto unlock_vcpu_destroy; @@ -4290,7 +4282,6 @@ static int kvm_vm_ioctl_create_vcpu(struct kvm *kvm, unsigned long id) atomic_inc(&kvm->online_vcpus); mutex_unlock(&vcpu->mutex); - mutex_unlock(&kvm->lock); kvm_arch_vcpu_postcreate(vcpu); kvm_create_vcpu_debugfs(vcpu); return r; @@ -4301,7 +4292,6 @@ static int kvm_vm_ioctl_create_vcpu(struct kvm *kvm, unsigned long id) xa_erase(&kvm->vcpu_array, vcpu->vcpu_idx); unlock_vcpu_destroy: vcpu->vcpu_idx = -1; - mutex_unlock(&kvm->lock); kvm_dirty_ring_free(&vcpu->dirty_ring); arch_vcpu_destroy: kvm_arch_vcpu_destroy(vcpu); @@ -4310,10 +4300,8 @@ static int kvm_vm_ioctl_create_vcpu(struct kvm *kvm, unsigned long id) vcpu_free: kmem_cache_free(kvm_vcpu_cache, vcpu); vcpu_decrement: - mutex_lock(&kvm->lock); kvm->created_vcpus--; __clear_bit(id, kvm->vcpu_ids); - mutex_unlock(&kvm->lock); return r; } From c5866d42f7116ab609e3ab6b4b699282e8287e3f Mon Sep 17 00:00:00 2001 From: Sean Christopherson Date: Mon, 21 Sep 2026 10:44:43 -0700 Subject: [PATCH 576/694] KVM: Move check for existing vCPU ID to the top of vCPU creation Now that kvm->lock is held for the entirety of vCPU creation, check for a conflicting vCPU ID at the begnning of vCPU creation, before the arch precreate() hook is invoked. This will allow reverting commit 97d65b544f48 ("KVM: Check for duplicate vcpu_id as early as possible"). For now, keep the redundant vcpu_ids tracking as a sanity check. No functional change intended (absent KVM bugs, checking vcpu_ids and walking kvm_get_vcpu_by_id() should yield the same result). Signed-off-by: Sean Christopherson Tested-by: Jean-Christophe Guillain Tested-by: Naveen N Rao (AMD) Message-ID: <20260921174445.911676-6-seanjc@google.com> Signed-off-by: Paolo Bonzini --- virt/kvm/kvm_main.c | 10 ++++------ 1 file changed, 4 insertions(+), 6 deletions(-) diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index cae0a9cdc33e..0e5b639559ba 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -4205,7 +4205,10 @@ static int kvm_vm_ioctl_create_vcpu(struct kvm *kvm, unsigned long id) if (kvm->created_vcpus >= kvm->max_vcpus) return -EINVAL; - if (test_bit(id, kvm->vcpu_ids)) + if (kvm_get_vcpu_by_id(kvm, id)) + return -EEXIST; + + if (WARN_ON_ONCE(test_bit(id, kvm->vcpu_ids))) return -EEXIST; r = kvm_arch_vcpu_precreate(kvm, id); @@ -4244,11 +4247,6 @@ static int kvm_vm_ioctl_create_vcpu(struct kvm *kvm, unsigned long id) goto arch_vcpu_destroy; } - if (WARN_ON_ONCE(kvm_get_vcpu_by_id(kvm, id))) { - r = -EEXIST; - goto unlock_vcpu_destroy; - } - /* * Set the vCPU's index *before* the vCPU is reachable by other tasks. * Unwind the index back to -1 on failure so that KVM can use the index From 700f8d24436c45cb3cc8956f35c90a1b6a6487c2 Mon Sep 17 00:00:00 2001 From: Sean Christopherson Date: Mon, 21 Sep 2026 10:44:44 -0700 Subject: [PATCH 577/694] Revert "KVM: Check for duplicate vcpu_id as early as possible" MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Now that KVM uses kvm_get_vcpu_by_id() to check for an existing vCPU ID before doing any meaningful work, which was made possible by holding kvm->lock for the entirety of vCPU creation, revert the now-redundant "early" vCPU ID tracking. The claims about the impact of kvm->vcpu_ids on the memory footprint were a wee bit wrong: the worst case scenario isn't 256 bytes per VM, it's 256 "unsigned longs" per VM, i.e. 2048 bytes per VM. Increasing the size of "struct kvm" by 2048 nearly doubled the total size on many architectures, and tripped x86's KVM_SANITY_CHECK_VM_STRUCT_SIZE, which was added to detect this *exact* scenario, where a single change significantly increased the size of "struct kvm". I.e. attempting to build KVM with CONFIG_DEBUG_KERNEL=n fails on x86 (the build failures got missed because all build bots apparently test only CONFIG_DEBUG_KERNEL=y kernels, and maintainers' test flows were similarly lacking). This reverts commit 97d65b544f48b2ee49f6aea32145e3e7969955dc. Fixes: 97d65b544f48 ("KVM: Check for duplicate vcpu_id as early as possible") Reported-by: Jean-Christophe Guillain Closes: https://lore.kernel.org/all/56a4bc35ee605588b7cc36c8e45c12b5f3b506cb.camel@guillain.net Reported-by: Paweł S Closes: https://lore.kernel.org/all/CABD%3DWFOS4j4hDv%2BpW-eEM9HAM2q2GY_iYdAG%2BqvYcUEinUrcQQ@mail.gmail.com Tested-by: Jean-Christophe Guillain Signed-off-by: Sean Christopherson Tested-by: Naveen N Rao (AMD) Message-ID: <20260921174445.911676-7-seanjc@google.com> Signed-off-by: Paolo Bonzini --- include/linux/kvm_host.h | 1 - virt/kvm/kvm_main.c | 5 ----- 2 files changed, 6 deletions(-) diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h index cf7fe835c4ad..c844def6dd95 100644 --- a/include/linux/kvm_host.h +++ b/include/linux/kvm_host.h @@ -791,7 +791,6 @@ struct kvm { /* The current active memslot set for each address space */ struct kvm_memslots __rcu *memslots[KVM_MAX_NR_ADDRESS_SPACES]; struct xarray vcpu_array; - DECLARE_BITMAP(vcpu_ids, KVM_MAX_VCPU_IDS); /* * Protected by slots_lock, but can be read outside if an * incorrect answer is acceptable. diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index 0e5b639559ba..02bab6923e03 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -4208,15 +4208,11 @@ static int kvm_vm_ioctl_create_vcpu(struct kvm *kvm, unsigned long id) if (kvm_get_vcpu_by_id(kvm, id)) return -EEXIST; - if (WARN_ON_ONCE(test_bit(id, kvm->vcpu_ids))) - return -EEXIST; - r = kvm_arch_vcpu_precreate(kvm, id); if (r) return r; kvm->created_vcpus++; - __set_bit(id, kvm->vcpu_ids); vcpu = kmem_cache_zalloc(kvm_vcpu_cache, GFP_KERNEL_ACCOUNT); if (!vcpu) { @@ -4299,7 +4295,6 @@ static int kvm_vm_ioctl_create_vcpu(struct kvm *kvm, unsigned long id) kmem_cache_free(kvm_vcpu_cache, vcpu); vcpu_decrement: kvm->created_vcpus--; - __clear_bit(id, kvm->vcpu_ids); return r; } From 41e005e2c5126ccb906d6374b3b97fcb223f7f8f Mon Sep 17 00:00:00 2001 From: Sean Christopherson Date: Mon, 21 Sep 2026 10:44:45 -0700 Subject: [PATCH 578/694] KVM: WARN if vCPU creation is in-progress when locking all vCPUs Now that KVM holds kvm->lock for the entirety of vCPU creation, from when created_vcpus is incremented until the new vCPU is fully onlined, WARN if the impossible happens and KVM somehow sees a discrepancy between the number of vCPUs "created" and "onlined". Signed-off-by: Sean Christopherson Tested-by: Jean-Christophe Guillain Tested-by: Naveen N Rao (AMD) Message-ID: <20260921174445.911676-8-seanjc@google.com> Signed-off-by: Paolo Bonzini --- virt/kvm/kvm_main.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index 02bab6923e03..5a782849521c 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -1363,7 +1363,7 @@ int kvm_trylock_all_vcpus(struct kvm *kvm) lockdep_assert_held(&kvm->lock); - if (kvm_is_vcpu_creation_in_progress(kvm)) + if (WARN_ON_ONCE(kvm_is_vcpu_creation_in_progress(kvm))) return -EBUSY; kvm_for_each_vcpu(i, vcpu, kvm) @@ -1389,7 +1389,7 @@ int kvm_lock_all_vcpus(struct kvm *kvm) lockdep_assert_held(&kvm->lock); - if (kvm_is_vcpu_creation_in_progress(kvm)) + if (WARN_ON_ONCE(kvm_is_vcpu_creation_in_progress(kvm))) return -EBUSY; kvm_for_each_vcpu(i, vcpu, kvm) { From b6249c1d5665098c8bf12f8d3e4712bf04c775bc Mon Sep 17 00:00:00 2001 From: Sean Christopherson Date: Fri, 25 Sep 2026 16:42:56 -0700 Subject: [PATCH 579/694] KVM: x86/mmu: Bail from shadow walks if the root is invalid or a dummy When walking shadow page tables, immediately terminate the walk if the root is a "dummy" root, i.e. a root whose top-level page table is backed by the zero page, but otherwise doesn't exist. If memslot creation races with a stage-2 page fault (EPT violation or #NPF) from L2, then if the stars align, KVM will attempt to walk shadow page tables using the zero page and hit a NULL pointer deref. BUG: kernel NULL pointer dereference, address: 0000000000000021 #PF: supervisor read access in kernel mode #PF: error_code(0x0000) - not-present page PGD 0 P4D 0 Oops: Oops: 0000 [#1] SMP CPU: 30 UID: 1000 PID: 941 Comm: qemu Not tainted 7.2.0-rc2-1b731e5ded48-next-vm #1741 PREEMPTLAZY Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 0.0.0 02/06/2015 RIP: 0010:__kvm_mmu_invalidate_addr+0xea/0x210 [kvm] Call Trace: kvm_mmu_invalidate_addr+0x92/0xe0 [kvm] __kvm_inject_emulated_page_fault+0x67/0x80 [kvm] ept_page_fault+0x160/0x850 [kvm] kvm_mmu_do_page_fault+0x102/0x1f0 [kvm] kvm_mmu_page_fault+0x8e/0x6b0 [kvm] vmx_handle_exit+0x163/0x640 [kvm_intel] kvm_arch_vcpu_ioctl_run+0x960/0x2120 [kvm] kvm_vcpu_ioctl+0x2c7/0x970 [kvm] __x64_sys_ioctl+0x90/0xd0 do_syscall_64+0x67/0x5f0 entry_SYSCALL_64_after_hwframe+0x4b/0x53 RIP: 0033:0x7f606d0b53bb Opportunistically harden the shadow walks against fully invalid roots, but WARN, as all callers are expected/required to pre-check for a valid root. Don't WARN in the dummy root case as the whole point of using a dummy root is to provide a root that's valid enough to enter the guest, i.e. it should Just Work for all flows except those that *need* to know about dummy roots. Alternatively, KVM could allocate a dedicated page and associated shadow page structure for the dummy root, which is very tempting as it such an approach should be more resilient against unexpected behavior. But that would be a much larger and thus riskier change than simply terminating walks of dummy roots. Fixes: 0e3223d8d00a ("KVM: x86/mmu: Use dummy root, backed by zero page, for !visible guest roots") Reported-by: Gabriel Schneider Cc: stable@vger.kernel.org Signed-off-by: Sean Christopherson Message-ID: <20260925234256.2384816-1-seanjc@google.com> Signed-off-by: Paolo Bonzini --- arch/x86/kvm/mmu/mmu.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/arch/x86/kvm/mmu/mmu.c b/arch/x86/kvm/mmu/mmu.c index 8e62476e477b..a0d608e3fceb 100644 --- a/arch/x86/kvm/mmu/mmu.c +++ b/arch/x86/kvm/mmu/mmu.c @@ -2512,6 +2512,12 @@ static void shadow_walk_init_using_root(struct kvm_shadow_walk_iterator *iterato iterator->addr = addr; iterator->shadow_addr = root; + + if (WARN_ON_ONCE(!VALID_PAGE(root)) || kvm_mmu_is_dummy_root(root)) { + iterator->level = 0; + return; + } + iterator->level = vcpu->arch.mmu->root_role.level; if (iterator->level >= PT64_ROOT_4LEVEL && From 60d93c27859fb3ec5b5a689e24de303f166477d0 Mon Sep 17 00:00:00 2001 From: Sean Christopherson Date: Fri, 4 Sep 2026 10:06:39 -0700 Subject: [PATCH 580/694] KVM: SVM: Preserve TLB control (i.e. pending TLB flush) on failed VMRUN Don't reset the VMCB's TLB control back to "do nothing" on a failed VMRUN, as empirical testing shows that the CPU performs the requested TLB flush if and only if VMRUN is successful, i.e. clearing TLB control on a failed VMRUN effectively drops a TLB flush. Explicitly track the need to flush all ASIDs on a per-CPU basis, as the ASID reuse condition is tied to the pCPU, not to the vCPU. As a bonus, this also obviates the need to avoid clobbering FLUSH_ALL_ASID with TLB_CONTROL_FLUSH_ASID, e.g. in svm_flush_tlb_asid(). Deliberately don't bother saving/restoring the "old" tlb_ctl on failure, in quotes because it's not exactly the old tlb_ctl, it's the tlb_ctl from after pre_svm_run(), but before updating tlb_ctl for flush_all_asids. If VMRUN fails and TLB_CONTROL_FLUSH_ALL_ASID is forced, then the next successful run of the VMCB *may* unnecessarily flush all ASIDs, which strictly speaking could result in noisy neighbor issues. However, the fact that new_asid() is already guest-triggerable, because of KVM's flawed behavior of clearing the ASID on emulated INIT, means that a guest can already trigger a flush of all ASIDs at roughly the same rate. And once KVM stops clobbering the ASID on emulated INIT, *or* assigns a static ASID to each vCPU, this flaw goes away. Fixes: 38e5e92fe8c0 ("KVM: SVM: Implement Flush-By-Asid feature") Cc: stable@vger.kernel.org Reported-by: Stefan Teodorescu Suggested-by: Yosry Ahmed Cc: Tom Lendacky Cc: Jim Mattson Signed-off-by: Sean Christopherson Message-ID: <20260904170642.3291466-2-seanjc@google.com> Signed-off-by: Paolo Bonzini --- arch/x86/kvm/svm/svm.c | 13 ++++++++++--- arch/x86/kvm/svm/svm.h | 1 + 2 files changed, 11 insertions(+), 3 deletions(-) diff --git a/arch/x86/kvm/svm/svm.c b/arch/x86/kvm/svm/svm.c index 7d59d301e1e5..afbaaaab84ed 100644 --- a/arch/x86/kvm/svm/svm.c +++ b/arch/x86/kvm/svm/svm.c @@ -1902,8 +1902,7 @@ static void new_asid(struct vcpu_svm *svm, struct svm_cpu_data *sd) if (sd->next_asid > sd->max_asid) { ++sd->asid_generation; sd->next_asid = sd->min_asid; - svm->vmcb->control.tlb_ctl = TLB_CONTROL_FLUSH_ALL_ASID; - vmcb_mark_dirty(svm->vmcb, VMCB_ASID); + sd->flush_all_asids = true; } svm->current_vmcb->asid_generation = sd->asid_generation; @@ -4528,6 +4527,11 @@ static __no_kcsan fastpath_t svm_vcpu_run(struct kvm_vcpu *vcpu, u64 run_flags) svm->vmcb->control.asid = svm->asid; vmcb_mark_dirty(svm->vmcb, VMCB_ASID); } + if (this_cpu_ptr(&svm_data)->flush_all_asids) { + svm->vmcb->control.tlb_ctl = TLB_CONTROL_FLUSH_ALL_ASID; + vmcb_mark_dirty(svm->vmcb, VMCB_ASID); + } + svm->vmcb->save.cr2 = vcpu->arch.cr2; if (guest_cpu_cap_has(vcpu, X86_FEATURE_ERAPS) && @@ -4618,7 +4622,10 @@ static __no_kcsan fastpath_t svm_vcpu_run(struct kvm_vcpu *vcpu, u64 run_flags) vcpu->arch.nested_run_pending = 0; } - svm->vmcb->control.tlb_ctl = TLB_CONTROL_DO_NOTHING; + if (!svm_is_vmrun_failure(svm->vmcb->control.exit_code)) { + this_cpu_ptr(&svm_data)->flush_all_asids = false; + svm->vmcb->control.tlb_ctl = TLB_CONTROL_DO_NOTHING; + } /* * Unconditionally mask off the CLEAR_RAP bit, the AND is just as cheap diff --git a/arch/x86/kvm/svm/svm.h b/arch/x86/kvm/svm/svm.h index e958943b8162..84f19026d3e8 100644 --- a/arch/x86/kvm/svm/svm.h +++ b/arch/x86/kvm/svm/svm.h @@ -376,6 +376,7 @@ struct svm_cpu_data { u32 next_asid; u32 min_asid; + bool flush_all_asids; bool bp_spec_reduce_set; struct vmcb *save_area; From e27fc915c8f575ab5630e27d97465c2077d4b9d1 Mon Sep 17 00:00:00 2001 From: Sean Christopherson Date: Fri, 4 Sep 2026 10:06:40 -0700 Subject: [PATCH 581/694] KVM: SVM: Update control fields on #VMEXIT if and only if VMRUN succeeded Leave control.erap_ctl and control.clean as-is in the VMCS if VMRUN fails, because as per AMD: there's no explicit architectural guarantee about the behavior in the presence of VMRUN failures. So the best thing to do would be to assume that if VMRUN fails, the actions requested in the control fields may not have been performed. Cc: stable@vger.kernel.org Signed-off-by: Sean Christopherson Message-ID: <20260904170642.3291466-3-seanjc@google.com> Signed-off-by: Paolo Bonzini --- arch/x86/kvm/svm/svm.c | 18 +++++++++--------- 1 file changed, 9 insertions(+), 9 deletions(-) diff --git a/arch/x86/kvm/svm/svm.c b/arch/x86/kvm/svm/svm.c index afbaaaab84ed..b63e7c69aa1c 100644 --- a/arch/x86/kvm/svm/svm.c +++ b/arch/x86/kvm/svm/svm.c @@ -4625,17 +4625,17 @@ static __no_kcsan fastpath_t svm_vcpu_run(struct kvm_vcpu *vcpu, u64 run_flags) if (!svm_is_vmrun_failure(svm->vmcb->control.exit_code)) { this_cpu_ptr(&svm_data)->flush_all_asids = false; svm->vmcb->control.tlb_ctl = TLB_CONTROL_DO_NOTHING; + + /* + * Unconditionally mask off the CLEAR_RAP bit, the AND is just + * as cheap as the TEST+Jcc to avoid it. + */ + if (cpu_feature_enabled(X86_FEATURE_ERAPS)) + svm->vmcb->control.erap_ctl &= ~ERAP_CONTROL_CLEAR_RAP; + + vmcb_mark_all_clean(svm->vmcb); } - /* - * Unconditionally mask off the CLEAR_RAP bit, the AND is just as cheap - * as the TEST+Jcc to avoid it. - */ - if (cpu_feature_enabled(X86_FEATURE_ERAPS)) - svm->vmcb->control.erap_ctl &= ~ERAP_CONTROL_CLEAR_RAP; - - vmcb_mark_all_clean(svm->vmcb); - /* if exit due to PF check for async PF */ if (svm->vmcb->control.exit_code == SVM_EXIT_EXCP_BASE + PF_VECTOR) vcpu->arch.apf.host_apf_flags = From 64c5b15af71ba3c26b2088bdbb0bbab3dbe58e2c Mon Sep 17 00:00:00 2001 From: Sean Christopherson Date: Fri, 4 Sep 2026 10:06:41 -0700 Subject: [PATCH 582/694] KVM: SVM: Don't mark ASID fields as dirty when setting control.tlb_ctl Don't mark the ASID as dirty in the VMCB when requesting a TLB flush via control.tlb_ctl. Per "15.15.3 VMCB Clean Field" of the July 2026, Revision 3.45 version of the APM: The following are explicitly not cached and not represented by Clean bits: * TLB_Control Fixes: 7e8e6eed75e2 ("KVM: SVM: Move asid to vcpu_svm") Suggested-by: Yosry Ahmed Signed-off-by: Sean Christopherson Message-ID: <20260904170642.3291466-4-seanjc@google.com> Signed-off-by: Paolo Bonzini --- arch/x86/kvm/svm/sev.c | 1 - arch/x86/kvm/svm/svm.c | 4 +--- 2 files changed, 1 insertion(+), 4 deletions(-) diff --git a/arch/x86/kvm/svm/sev.c b/arch/x86/kvm/svm/sev.c index 6aa86a78711e..3448d56520c6 100644 --- a/arch/x86/kvm/svm/sev.c +++ b/arch/x86/kvm/svm/sev.c @@ -3620,7 +3620,6 @@ int pre_sev_run(struct vcpu_svm *svm, int cpu) sd->sev_vmcbs[asid] = svm->vmcb; svm->vmcb->control.tlb_ctl = TLB_CONTROL_FLUSH_ASID; - vmcb_mark_dirty(svm->vmcb, VMCB_ASID); return 0; } diff --git a/arch/x86/kvm/svm/svm.c b/arch/x86/kvm/svm/svm.c index b63e7c69aa1c..830ace75e986 100644 --- a/arch/x86/kvm/svm/svm.c +++ b/arch/x86/kvm/svm/svm.c @@ -4527,10 +4527,8 @@ static __no_kcsan fastpath_t svm_vcpu_run(struct kvm_vcpu *vcpu, u64 run_flags) svm->vmcb->control.asid = svm->asid; vmcb_mark_dirty(svm->vmcb, VMCB_ASID); } - if (this_cpu_ptr(&svm_data)->flush_all_asids) { + if (this_cpu_ptr(&svm_data)->flush_all_asids) svm->vmcb->control.tlb_ctl = TLB_CONTROL_FLUSH_ALL_ASID; - vmcb_mark_dirty(svm->vmcb, VMCB_ASID); - } svm->vmcb->save.cr2 = vcpu->arch.cr2; From 73392706d0e0f3aff299321de41d73abe3f53d9d Mon Sep 17 00:00:00 2001 From: Sean Christopherson Date: Fri, 4 Sep 2026 10:06:42 -0700 Subject: [PATCH 583/694] KVM: SVM: Sync guest's PERF_CNTR_GLOBAL_CTL from h/w only on successful VMRUN Don't (re)read PERF_CNTR_GLOBAL_CTL from hardware on a failed VMRUN, as the purpose of the read is to synchronize KVM's cache with any writes done by the guest, and the guest can't possibly have modified the MSR if it never got a chance to run. Signed-off-by: Sean Christopherson Message-ID: <20260904170642.3291466-5-seanjc@google.com> Signed-off-by: Paolo Bonzini --- arch/x86/kvm/svm/svm.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/arch/x86/kvm/svm/svm.c b/arch/x86/kvm/svm/svm.c index 830ace75e986..f4f488328ea4 100644 --- a/arch/x86/kvm/svm/svm.c +++ b/arch/x86/kvm/svm/svm.c @@ -4632,6 +4632,10 @@ static __no_kcsan fastpath_t svm_vcpu_run(struct kvm_vcpu *vcpu, u64 run_flags) svm->vmcb->control.erap_ctl &= ~ERAP_CONTROL_CLEAR_RAP; vmcb_mark_all_clean(svm->vmcb); + + if (!msr_write_intercepted(svm, MSR_AMD64_PERF_CNTR_GLOBAL_CTL)) + rdmsrq(MSR_AMD64_PERF_CNTR_GLOBAL_CTL, + vcpu_to_pmu(vcpu)->global_ctrl); } /* if exit due to PF check for async PF */ @@ -4641,9 +4645,6 @@ static __no_kcsan fastpath_t svm_vcpu_run(struct kvm_vcpu *vcpu, u64 run_flags) kvm_clear_available_registers(vcpu, SVM_REGS_LAZY_LOAD_SET); - if (!msr_write_intercepted(svm, MSR_AMD64_PERF_CNTR_GLOBAL_CTL)) - rdmsrq(MSR_AMD64_PERF_CNTR_GLOBAL_CTL, vcpu_to_pmu(vcpu)->global_ctrl); - trace_kvm_exit(vcpu, KVM_ISA_SVM); svm_complete_interrupts(vcpu); From 9e06bbb9ade7cdf28d4a3b5b14e5812881c8df85 Mon Sep 17 00:00:00 2001 From: Sean Christopherson Date: Wed, 26 Aug 2026 12:58:33 -0700 Subject: [PATCH 584/694] KVM: SVM: Use the active VMCB's MSR bitmap when checking if MSR is intercepted Use the MSR permission bitmap of the active VMCB instead of assuming that KVM is always using vmcb02's bitmap when L2 is active, as KVM uses msrpm02 if and only if L1 wants to intercept MSR accesses, i.e. if and only if KVM needs to merge msprm01 with msrpm12. Don't bother tracking the virtual address of the bitmap that's being used, as __va() is cheap on x86, and caching the virtual address would introduce yet another source of potentially stale information. Fixes: b2ac58f90540 ("KVM/SVM: Allow direct access to MSR_IA32_SPEC_CTRL") Cc: stable@vger.kernel.org Reported-by: Stefan Teodorescu Signed-off-by: Sean Christopherson Message-ID: <20260826195833.844526-1-seanjc@google.com> Signed-off-by: Paolo Bonzini --- arch/x86/kvm/svm/svm.c | 11 +---------- 1 file changed, 1 insertion(+), 10 deletions(-) diff --git a/arch/x86/kvm/svm/svm.c b/arch/x86/kvm/svm/svm.c index f4f488328ea4..fe0cba4731b2 100644 --- a/arch/x86/kvm/svm/svm.c +++ b/arch/x86/kvm/svm/svm.c @@ -673,16 +673,7 @@ static void clr_dr_intercepts(struct vcpu_svm *svm) static bool msr_write_intercepted(struct vcpu_svm *svm, u32 msr) { - /* - * For non-nested case: - * If the L01 MSR bitmap does not intercept the MSR, then we need to - * save it. - * - * For nested case: - * If the L02 MSR bitmap does not intercept the MSR, then we need to - * save it. - */ - void *msrpm = is_guest_mode(&svm->vcpu) ? svm->nested.msrpm : svm->msrpm; + void *msrpm = __va(__sme_clr(svm->vmcb->control.msrpm_base_pa)); return svm_test_msr_bitmap_write(msrpm, msr); } From 2abb25273c82061dd4670690c81c319c03740bf6 Mon Sep 17 00:00:00 2001 From: Sean Christopherson Date: Thu, 13 Aug 2026 14:44:03 -0700 Subject: [PATCH 585/694] KVM: nVMX: Force MSR bitmap refresh if runtime eVMCS controls are modified Force a refresh of the vmcs02 MSR bitmap during nested VM-Enter if the runtime eVMCS controls (pin, primary, secondary, etc.) are being updated. If L1 isn't intercepting TPR writes, runs L2 with TPR virtualization, and then runs the same L2 with TPR virtualization disabled, KVM will fail to refresh msr_bitmap02 and leave TPR in passthrough mode even though TPR virtualization is disabled. I.e. failure to refresh the bitmap lets L2 (or L1 by proxy) read and write L0's TPR. Fixes: 502d2bf5f2fd ("KVM: nVMX: Implement Enlightened MSR Bitmap feature") Cc: stable@vger.kernel.org Reviewed-by: Vitaly Kuznetsov Signed-off-by: Sean Christopherson Signed-off-by: Paolo Bonzini --- arch/x86/kvm/vmx/nested.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/arch/x86/kvm/vmx/nested.c b/arch/x86/kvm/vmx/nested.c index 40c1a5f6fa8a..b25216862740 100644 --- a/arch/x86/kvm/vmx/nested.c +++ b/arch/x86/kvm/vmx/nested.c @@ -1754,6 +1754,9 @@ static void copy_vmcs12_to_shadow(struct vcpu_vmx *vmx) static void copy_enlightened_to_vmcs12(struct vcpu_vmx *vmx, u32 hv_clean_fields) { #ifdef CONFIG_KVM_HYPERV + const u64 runtime_controls = HV_VMX_ENLIGHTENED_CLEAN_FIELD_CONTROL_GRP1 | + HV_VMX_ENLIGHTENED_CLEAN_FIELD_CONTROL_GRP2 | + HV_VMX_ENLIGHTENED_CLEAN_FIELD_CONTROL_PROC; struct vmcs12 *vmcs12 = vmx->nested.cached_vmcs12; struct hv_enlightened_vmcs *evmcs = nested_vmx_evmcs(vmx); struct kvm_vcpu_hv *hv_vcpu = to_hv_vcpu(&vmx->vcpu); @@ -1762,6 +1765,9 @@ static void copy_enlightened_to_vmcs12(struct vcpu_vmx *vmx, u32 hv_clean_fields vmcs12->tpr_threshold = evmcs->tpr_threshold; vmcs12->guest_rip = evmcs->guest_rip; + if ((hv_clean_fields & runtime_controls) != runtime_controls) + vmx->nested.force_msr_bitmap_recalc = true; + if (unlikely(!(hv_clean_fields & HV_VMX_ENLIGHTENED_CLEAN_FIELD_ENLIGHTENMENTSCONTROL))) { hv_vcpu->nested.pa_page_gpa = evmcs->partition_assist_page; From 8916c7a87455370b25523d5b1b9015b5244d18c8 Mon Sep 17 00:00:00 2001 From: Sean Christopherson Date: Fri, 10 Jul 2026 09:20:52 -0700 Subject: [PATCH 586/694] KVM: selftests: Add x2APIC MSR test for inhibiting APICv while nested Add a selftest to verify that KVM intercepts x2APIC MSR accesses for L1 after APICv is inhibited while L2 is active. This is a regression test for an AVIC bug where KVM would skip updating x2APIC MSR intercepts while L2 is active, thus giving L1 access to a wide swath of L0's x2APIC surface. Signed-off-by: Sean Christopherson Message-ID: <20260710162052.2188574-3-seanjc@google.com> Signed-off-by: Paolo Bonzini --- tools/testing/selftests/kvm/Makefile.kvm | 1 + .../selftests/kvm/x86/nested_x2apic_test.c | 117 ++++++++++++++++++ 2 files changed, 118 insertions(+) create mode 100644 tools/testing/selftests/kvm/x86/nested_x2apic_test.c diff --git a/tools/testing/selftests/kvm/Makefile.kvm b/tools/testing/selftests/kvm/Makefile.kvm index 6a1482e3a286..752f81b070fa 100644 --- a/tools/testing/selftests/kvm/Makefile.kvm +++ b/tools/testing/selftests/kvm/Makefile.kvm @@ -103,6 +103,7 @@ TEST_GEN_PROGS_x86 += x86/nested_tdp_fault_test TEST_GEN_PROGS_x86 += x86/nested_tsc_adjust_test TEST_GEN_PROGS_x86 += x86/nested_tsc_scaling_test TEST_GEN_PROGS_x86 += x86/nested_vmsave_vmload_test +TEST_GEN_PROGS_x86 += x86/nested_x2apic_test TEST_GEN_PROGS_x86 += x86/platform_info_test TEST_GEN_PROGS_x86 += x86/pmu_counters_test TEST_GEN_PROGS_x86 += x86/pmu_event_filter_test diff --git a/tools/testing/selftests/kvm/x86/nested_x2apic_test.c b/tools/testing/selftests/kvm/x86/nested_x2apic_test.c new file mode 100644 index 000000000000..e77b4c347272 --- /dev/null +++ b/tools/testing/selftests/kvm/x86/nested_x2apic_test.c @@ -0,0 +1,117 @@ +// SPDX-License-Identifier: GPL-2.0-only +#include "test_util.h" +#include "kvm_util.h" +#include "processor.h" +#include "vmx.h" +#include "svm_util.h" + +/* + * Use the kernel's posted interrupt vectors to minimize the risk of crashing + * the host if KVM is buggy. Note, the vectors aren't set in stone, ideally + * these will be kept up-to-date if the kernel vectors change, but it's "fine" + * if they are stale. + */ +#define POSTED_INTR_VECTOR 0xf2 +#define POSTED_INTR_WAKEUP_VECTOR 0xf1 +#define POSTED_INTR_NESTED_VECTOR 0xf0 + +static volatile unsigned int nr_irqs; + +static void guest_irq_handler(struct ex_regs *regs) +{ + nr_irqs++; + x2apic_write_reg(APIC_EOI, 0); +} + +static void l2_guest_code(void) +{ + wrmsr(MSR_IA32_APICBASE, rdmsr(MSR_IA32_APICBASE) & GENMASK_ULL(11, 0)); + asm volatile("cpuid" ::: "eax", "ebx", "ecx", "edx"); +} + +static void l1_svm_code(struct svm_test_data *svm) +{ + struct vmcb_control_area *ctrl = &svm->vmcb->control; + + generic_svm_setup(svm, l2_guest_code); + ctrl->intercept |= BIT_ULL(INTERCEPT_CPUID) | BIT_ULL(INTERCEPT_MSR_PROT); + + run_guest(svm->vmcb, svm->vmcb_gpa); + GUEST_ASSERT_EQ(ctrl->exit_code, SVM_EXIT_CPUID); + + stgi(); +} + +static void l1_vmx_code(struct vmx_pages *vmx) +{ + u64 control; + + GUEST_ASSERT_EQ(prepare_for_vmx_operation(vmx), true); + GUEST_ASSERT_EQ(load_vmcs(vmx), true); + + prepare_vmcs(vmx, NULL); + GUEST_ASSERT_EQ(vmwrite(GUEST_RIP, (unsigned long)l2_guest_code), 0); + + control = vmreadz(CPU_BASED_VM_EXEC_CONTROL); + control |= CPU_BASED_USE_MSR_BITMAPS; + GUEST_ASSERT_EQ(vmwrite(CPU_BASED_VM_EXEC_CONTROL, control), 0); + + GUEST_ASSERT(!vmlaunch()); + GUEST_ASSERT_EQ(vmreadz(VM_EXIT_REASON), EXIT_REASON_CPUID); +} + +static void l1_guest_code(void *test_data) +{ + x2apic_enable(); + + if (this_cpu_has(X86_FEATURE_SVM)) + l1_svm_code(test_data); + else + l1_vmx_code(test_data); + + sti_nop(); + + x2apic_write_reg(APIC_ICR, APIC_DEST_SELF | APIC_INT_ASSERT | POSTED_INTR_VECTOR); + x2apic_write_reg(APIC_ICR, APIC_DEST_SELF | APIC_INT_ASSERT | POSTED_INTR_WAKEUP_VECTOR); + x2apic_write_reg(APIC_ICR, APIC_DEST_SELF | APIC_INT_ASSERT | POSTED_INTR_NESTED_VECTOR); + GUEST_ASSERT_EQ(nr_irqs, 3); + GUEST_DONE(); +} + +int main(int argc, char *argv[]) +{ + gva_t nested_test_data_gva; + struct kvm_vcpu *vcpu; + struct kvm_vm *vm; + struct ucall uc; + + TEST_REQUIRE(kvm_cpu_has(X86_FEATURE_SVM) || kvm_cpu_has(X86_FEATURE_VMX)); + + vm = vm_create_with_one_vcpu(&vcpu, l1_guest_code); + vm_install_exception_handler(vm, POSTED_INTR_VECTOR, guest_irq_handler); + vm_install_exception_handler(vm, POSTED_INTR_WAKEUP_VECTOR, guest_irq_handler); + vm_install_exception_handler(vm, POSTED_INTR_NESTED_VECTOR, guest_irq_handler); + + if (kvm_cpu_has(X86_FEATURE_SVM)) + vcpu_alloc_svm(vm, &nested_test_data_gva); + else + vcpu_alloc_vmx(vm, &nested_test_data_gva); + + vcpu_args_set(vcpu, 1, nested_test_data_gva); + + vcpu_run(vcpu); + + TEST_ASSERT_KVM_EXIT_REASON(vcpu, KVM_EXIT_IO); + + switch (get_ucall(vcpu, &uc)) { + case UCALL_DONE: + break; + case UCALL_ABORT: + REPORT_GUEST_ASSERT(uc); + break; + default: + TEST_FAIL("Expected DONE, got unexpected ucall %lu", uc.cmd); + } + + kvm_vm_free(vm); +} From 0c19bc0ef11d8ea7f929a3ec3ce549cfe0781859 Mon Sep 17 00:00:00 2001 From: Sean Christopherson Date: Thu, 13 Aug 2026 15:36:07 -0700 Subject: [PATCH 587/694] KVM: selftests: Run the nested x2APIC with and without APICv being inhibited in L2 Signed-off-by: Sean Christopherson Message-ID: <20260813223610.2043560-4-seanjc@google.com> Signed-off-by: Paolo Bonzini --- .../selftests/kvm/x86/nested_x2apic_test.c | 25 ++++++++++++++++--- 1 file changed, 21 insertions(+), 4 deletions(-) diff --git a/tools/testing/selftests/kvm/x86/nested_x2apic_test.c b/tools/testing/selftests/kvm/x86/nested_x2apic_test.c index e77b4c347272..a1072bf499ee 100644 --- a/tools/testing/selftests/kvm/x86/nested_x2apic_test.c +++ b/tools/testing/selftests/kvm/x86/nested_x2apic_test.c @@ -15,6 +15,7 @@ #define POSTED_INTR_WAKEUP_VECTOR 0xf1 #define POSTED_INTR_NESTED_VECTOR 0xf0 +static bool inhibit_apicv; static volatile unsigned int nr_irqs; static void guest_irq_handler(struct ex_regs *regs) @@ -25,7 +26,8 @@ static void guest_irq_handler(struct ex_regs *regs) static void l2_guest_code(void) { - wrmsr(MSR_IA32_APICBASE, rdmsr(MSR_IA32_APICBASE) & GENMASK_ULL(11, 0)); + if (inhibit_apicv) + wrmsr(MSR_IA32_APICBASE, rdmsr(MSR_IA32_APICBASE) & GENMASK_ULL(11, 0)); asm volatile("cpuid" ::: "eax", "ebx", "ecx", "edx"); } @@ -78,20 +80,20 @@ static void l1_guest_code(void *test_data) GUEST_DONE(); } -int main(int argc, char *argv[]) +static void __test_x2apic_intercepts(void) { gva_t nested_test_data_gva; struct kvm_vcpu *vcpu; struct kvm_vm *vm; struct ucall uc; - TEST_REQUIRE(kvm_cpu_has(X86_FEATURE_SVM) || kvm_cpu_has(X86_FEATURE_VMX)); - vm = vm_create_with_one_vcpu(&vcpu, l1_guest_code); vm_install_exception_handler(vm, POSTED_INTR_VECTOR, guest_irq_handler); vm_install_exception_handler(vm, POSTED_INTR_WAKEUP_VECTOR, guest_irq_handler); vm_install_exception_handler(vm, POSTED_INTR_NESTED_VECTOR, guest_irq_handler); + sync_global_to_guest(vm, inhibit_apicv); + if (kvm_cpu_has(X86_FEATURE_SVM)) vcpu_alloc_svm(vm, &nested_test_data_gva); else @@ -115,3 +117,18 @@ int main(int argc, char *argv[]) kvm_vm_free(vm); } + +#define test_x2apic_intercepts(inhibit_apic_setting) \ +do { \ + inhibit_apic_setting; \ + \ + __test_x2apic_intercepts(); \ +} while (0) + +int main(int argc, char *argv[]) +{ + TEST_REQUIRE(kvm_cpu_has(X86_FEATURE_SVM) || kvm_cpu_has(X86_FEATURE_VMX)); + + test_x2apic_intercepts(inhibit_apicv = true); + test_x2apic_intercepts(inhibit_apicv = false); +} From 2f00ba93cc285092cd3b7ffebbc9abb6004f55e7 Mon Sep 17 00:00:00 2001 From: Sean Christopherson Date: Thu, 13 Aug 2026 15:36:08 -0700 Subject: [PATCH 588/694] KVM: selftests: Verify that L0's TPR doesn't get clobbered Signed-off-by: Sean Christopherson Message-ID: <20260813223610.2043560-5-seanjc@google.com> Signed-off-by: Paolo Bonzini --- .../selftests/kvm/x86/nested_x2apic_test.c | 16 ++++++++++++++++ 1 file changed, 16 insertions(+) diff --git a/tools/testing/selftests/kvm/x86/nested_x2apic_test.c b/tools/testing/selftests/kvm/x86/nested_x2apic_test.c index a1072bf499ee..3b59ba3e3342 100644 --- a/tools/testing/selftests/kvm/x86/nested_x2apic_test.c +++ b/tools/testing/selftests/kvm/x86/nested_x2apic_test.c @@ -28,6 +28,10 @@ static void l2_guest_code(void) { if (inhibit_apicv) wrmsr(MSR_IA32_APICBASE, rdmsr(MSR_IA32_APICBASE) & GENMASK_ULL(11, 0)); + + x2apic_write_reg(APIC_TASKPRI, 0xf0); + GUEST_ASSERT_EQ(x2apic_read_reg(APIC_TASKPRI), 0xf0); + asm volatile("cpuid" ::: "eax", "ebx", "ecx", "edx"); } @@ -73,10 +77,22 @@ static void l1_guest_code(void *test_data) sti_nop(); + GUEST_ASSERT_EQ(x2apic_read_reg(APIC_TASKPRI), 0xf0); + x2apic_write_reg(APIC_ICR, APIC_DEST_SELF | APIC_INT_ASSERT | POSTED_INTR_VECTOR); + GUEST_ASSERT_EQ(nr_irqs, 0); + + x2apic_write_reg(APIC_TASKPRI, 0xff); + GUEST_ASSERT_EQ(x2apic_read_reg(APIC_TASKPRI), 0xff); x2apic_write_reg(APIC_ICR, APIC_DEST_SELF | APIC_INT_ASSERT | POSTED_INTR_WAKEUP_VECTOR); + GUEST_ASSERT_EQ(nr_irqs, 0); + + x2apic_write_reg(APIC_TASKPRI, 0); + GUEST_ASSERT_EQ(x2apic_read_reg(APIC_TASKPRI), 0); + x2apic_write_reg(APIC_ICR, APIC_DEST_SELF | APIC_INT_ASSERT | POSTED_INTR_NESTED_VECTOR); GUEST_ASSERT_EQ(nr_irqs, 3); + GUEST_DONE(); } From 31c88350b7dd1522792f726f79607f31bb55c50f Mon Sep 17 00:00:00 2001 From: Hui Peng Date: Thu, 24 Sep 2026 04:27:27 +0000 Subject: [PATCH 589/694] cgroup/cpuset: Return PERR_NOCPUS in remote_partition_enable() on subpartitions_cpus conflict When a remote partition is created underneath an existing local partition via a non-partition (PRS_MEMBER) intermediate cgroup, update_prstate() sees parent->partition_root_state == PRS_MEMBER and calls remote_partition_enable(). Commit 86888c7bd117 ("cgroup/cpuset: Add warnings to catch inconsistency in exclusive CPUs") replaced the cpumask_intersects(tmp->new_cpus, subpartitions_cpus) error check in remote_partition_enable() with WARN_ON_ONCE(). As a result, remote_partition_enable() emits a warning and proceeds to enable the remote partition on CPUs that are already owned by the ancestor local partition in subpartitions_cpus. This can be reproduced on Linux 7.3.0-rc3 with: mkdir -p /tmp/cg1 mount -t cgroup2 none /tmp/cg1 echo "+cpuset" > /tmp/cg1/cgroup.subtree_control mkdir /tmp/cg1/A echo 1 > /tmp/cg1/A/cpuset.cpus echo 1 > /tmp/cg1/A/cpuset.cpus.exclusive echo root > /tmp/cg1/A/cpuset.cpus.partition echo "+cpuset" > /tmp/cg1/A/cgroup.subtree_control mkdir /tmp/cg1/A/B echo 1 > /tmp/cg1/A/B/cpuset.cpus echo 1 > /tmp/cg1/A/B/cpuset.cpus.exclusive echo "+cpuset" > /tmp/cg1/A/B/cgroup.subtree_control mkdir /tmp/cg1/A/B/D echo 1 > /tmp/cg1/A/B/D/cpuset.cpus echo 1 > /tmp/cg1/A/B/D/cpuset.cpus.exclusive echo root > /tmp/cg1/A/B/D/cpuset.cpus.partition which triggers: WARNING: kernel/cgroup/cpuset.c:1594 at remote_partition_enable+0x1c1/0x300 and leaves both /tmp/cg1/A and /tmp/cg1/A/B/D as active root partitions claiming exclusive CPU 1. Fix this by returning PERR_NOCPUS when tmp->new_cpus intersects subpartitions_cpus in remote_partition_enable(), matching the error code used by remote_cpus_update() for the same subpartitions_cpus conflict, and add a regression test case to tools/testing/selftests/cgroup/test_cpuset_prs.sh. Tested in QEMU on Linux 7.3.0-rc3 using the reproducer above and tools/testing/selftests/cgroup/test_cpuset_prs.sh. Fixes: 86888c7bd117 ("cgroup/cpuset: Add warnings to catch inconsistency in exclusive CPUs") Suggested-by: Guopeng Zhang Cc: stable@vger.kernel.org Assisted-by: LLM Signed-off-by: Hui Peng Reviewed-by: Waiman Long Signed-off-by: Tejun Heo --- kernel/cgroup/cpuset.c | 3 ++- tools/testing/selftests/cgroup/test_cpuset_prs.sh | 2 ++ 2 files changed, 4 insertions(+), 1 deletion(-) diff --git a/kernel/cgroup/cpuset.c b/kernel/cgroup/cpuset.c index 8f24171b6055..1fcec89a28b9 100644 --- a/kernel/cgroup/cpuset.c +++ b/kernel/cgroup/cpuset.c @@ -1591,10 +1591,11 @@ static int remote_partition_enable(struct cpuset *cs, int new_prs, * above it or remote partition root underneath it is not allowed. */ compute_excpus(cs, tmp->new_cpus); - WARN_ON_ONCE(cpumask_intersects(tmp->new_cpus, subpartitions_cpus)); if (!cpumask_intersects(tmp->new_cpus, cpu_active_mask) || cpumask_subset(top_cpuset.effective_cpus, tmp->new_cpus)) return PERR_INVCPUS; + if (cpumask_intersects(tmp->new_cpus, subpartitions_cpus)) + return PERR_NOCPUS; if (((new_prs == PRS_ISOLATED) && !isolated_cpus_can_update(tmp->new_cpus, NULL)) || prstate_housekeeping_conflict(new_prs, tmp->new_cpus)) diff --git a/tools/testing/selftests/cgroup/test_cpuset_prs.sh b/tools/testing/selftests/cgroup/test_cpuset_prs.sh index 131d8b4551ef..7efd5e645767 100755 --- a/tools/testing/selftests/cgroup/test_cpuset_prs.sh +++ b/tools/testing/selftests/cgroup/test_cpuset_prs.sh @@ -298,6 +298,8 @@ TEST_MATRIX=( " C0-4:X2-4 C1-4:X2-4:P2 C2-4:X4:P1 \ . . . X1 . 0 A1:0-1|A2:2-4|A3:2-4 \ A1:P0|A2:P2|A3:P-1 2-4" + " CX1-3:P1 CX1-3 CX1-3 . . . P1 . 0 A1:1-3|A2:1-3|A3:1-3 \ + A1:P1|A2:P0|A3:P-1" # Remote partition offline tests " C0-3 C1-3 C2-3 . X2-3 X2-3 X2-3:P2:O2=0 . 0 A1:0-1|A2:1|A3:3 A1:P0|A3:P2 2-3" From 94480606a677deb68d5622cf0ded88626514b3f2 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Sat, 26 Sep 2026 09:23:02 -1000 Subject: [PATCH 590/694] sched_ext: Add a size argument to scx_bpf_cid_topo() so struct scx_cid_topo can grow scx_bpf_cid_topo() copies struct scx_cid_topo into a buffer the BPF program sized from its own vmlinux.h while the verifier sizes the write from the running kernel's BTF. The struct may grow and each growth then breaks every scheduler built against the older layout, rejected at load or written past its buffer. This is the usual hole for a struct handed to BPF, closed elsewhere with a size argument, and it was missed here. Take the buffer size, copy the smaller of it and the kernel's struct and set the rest to -1. Accesses to the copy are CO-RE relocated, so the struct can grow by appending fields, which its comment now states. The kfunc changes in place: the cid interface is still being finalized and no released scheduler uses the current form. Fixes: e9b55af47edf ("sched_ext: Add topological CPU IDs (cids)") Cc: stable@vger.kernel.org # v7.2+ Signed-off-by: Tejun Heo Reviewed-by: Andrea Righi --- kernel/sched/ext/cid.c | 26 +++++++++++++++--------- kernel/sched/ext/types.h | 4 ++++ tools/sched_ext/include/scx/common.bpf.h | 2 +- 3 files changed, 21 insertions(+), 11 deletions(-) diff --git a/kernel/sched/ext/cid.c b/kernel/sched/ext/cid.c index 39f88deb94bc..d4d5f537832a 100644 --- a/kernel/sched/ext/cid.c +++ b/kernel/sched/ext/cid.c @@ -912,30 +912,36 @@ bool scx_cmask_empty(const struct scx_cmask *m) /** * scx_bpf_cid_topo - Copy out per-cid topology info * @cid: cid to look up - * @out__uninit: where to copy the topology info; fully written by this call + * @out: where to copy the topology info + * @out__sz: size of @out, the program's sizeof(struct scx_cid_topo) * @aux: implicit BPF argument to access bpf_prog_aux hidden from BPF progs * - * Fill @out__uninit with the topology info for @cid. Trigger scx_error() if - * @cid is out of range. If @cid is valid but in the no-topo section, all fields - * are set to -1. All fields are also set to -1 when no cid tables have been - * published yet, which a program may observe while racing the root enable. + * Fill @out with the topology info for @cid. Trigger scx_error() if @cid is out + * of range. If @cid is valid but in the no-topo section, all fields are set to + * -1. All fields are also set to -1 when no cid tables have been published yet, + * which a program may observe while racing the root enable. + * + * The program's struct may be older or newer than the kernel's. The smaller of + * @out__sz and the kernel's size is copied and the rest of @out is set to -1. */ -__bpf_kfunc void scx_bpf_cid_topo(s32 cid, struct scx_cid_topo *out__uninit, +__bpf_kfunc void scx_bpf_cid_topo(s32 cid, struct scx_cid_topo *out, size_t out__sz, const struct bpf_prog_aux *aux) { + size_t len = min(out__sz, sizeof(*out)); struct scx_cid_topo *topo; struct scx_sched *sch; + /* the error cases and fields the kernel lacks read as -1 */ + memset(out, 0xff, out__sz); + guard(rcu)(); sch = scx_prog_sched(aux); topo = rcu_dereference(scx_cid_topo); - if (unlikely(!sch) || !cid_valid(sch, cid) || unlikely(!topo)) { - *out__uninit = SCX_CID_TOPO_NEG; + if (unlikely(!sch) || !cid_valid(sch, cid) || unlikely(!topo)) return; - } - *out__uninit = topo[cid]; + memcpy(out, &topo[cid], len); } __bpf_kfunc_end_defs(); diff --git a/kernel/sched/ext/types.h b/kernel/sched/ext/types.h index 943d8d429a2c..139176cf9fc6 100644 --- a/kernel/sched/ext/types.h +++ b/kernel/sched/ext/types.h @@ -70,6 +70,10 @@ enum scx_consts { * smaller shards if the LLC exceeds the target size. No-topo cids are packed * into their own max-sized shards. * + * New fields are appended, never inserted: scx_bpf_cid_topo() copies this + * struct out sized by the program's own layout, and an older program's copy + * must stay a prefix of the kernel's. + * * @core_cid: first cid of this cid's core (smt-sibling group) * @core_idx: global index of that core, in [0, nr_cores_at_init) * @llc_cid: first cid of this cid's LLC diff --git a/tools/sched_ext/include/scx/common.bpf.h b/tools/sched_ext/include/scx/common.bpf.h index 2ddb01a059fd..22f24ebef8a9 100644 --- a/tools/sched_ext/include/scx/common.bpf.h +++ b/tools/sched_ext/include/scx/common.bpf.h @@ -106,7 +106,7 @@ u64 scx_bpf_now(void) __ksym __weak; void scx_bpf_events(struct scx_event_stats *events, size_t events__sz) __ksym __weak; s32 scx_bpf_cpu_to_cid(s32 cpu) __ksym __weak; s32 scx_bpf_cid_to_cpu(s32 cid) __ksym __weak; -void scx_bpf_cid_topo(s32 cid, struct scx_cid_topo *out) __ksym __weak; +void scx_bpf_cid_topo(s32 cid, struct scx_cid_topo *out, size_t out__sz) __ksym __weak; void scx_bpf_kick_cid(s32 cid, u64 flags) __ksym __weak; s32 scx_bpf_task_cid(const struct task_struct *p) __ksym __weak; s32 scx_bpf_this_cid(void) __ksym __weak; From c79cf15eb35a4c703bf18f8c33c47e80dd660f8e Mon Sep 17 00:00:00 2001 From: Hao Ge Date: Thu, 27 Aug 2026 11:05:03 +0800 Subject: [PATCH 591/694] module: fix lost error code from codetag_load_module() If codetag_load_module() fails, err is not set to reflect the failure and load_module() returns 0 after the module has been torn down. Also, if the module is a livepatch, mod->klp_info allocated by copy_module_elf() leaks on this error path. Free it via a new livepatch_cleanup label. Link: https://lore.kernel.org/20260827030503.49171-1-hao.ge@linux.dev Fixes: 044d2aee6c57 ("alloc_tag: handle module codetag load errors as module load failures") Signed-off-by: Hao Ge Signed-off-by: Andrew Morton Reported-by: Sashiko Suggested-by: Petr Pavlu Reviewed-by: Bradley Morgan Cc: Aaron Tomlin Cc: Luis Chamberalin Cc: Sami Tolvanen Cc: Suren Baghdasaryan Cc: --- kernel/module/main.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/kernel/module/main.c b/kernel/module/main.c index d0e1e0bd2ad0..c1b34dc1e89a 100644 --- a/kernel/module/main.c +++ b/kernel/module/main.c @@ -3581,8 +3581,9 @@ static int load_module(struct load_info *info, const char __user *uargs, goto sysfs_cleanup; } - if (codetag_load_module(mod)) - goto sysfs_cleanup; + err = codetag_load_module(mod); + if (err) + goto livepatch_cleanup; /* Get rid of temporary copy. */ free_copy(info, flags); @@ -3592,6 +3593,9 @@ static int load_module(struct load_info *info, const char __user *uargs, return do_init_module(mod); + livepatch_cleanup: + if (is_livepatch_module(mod)) + free_module_elf(mod); sysfs_cleanup: mod_sysfs_teardown(mod); coming_cleanup: From 72c3d682c165b836b54a3d07a01f72e39c47022c Mon Sep 17 00:00:00 2001 From: Baolin Wang Date: Mon, 14 Sep 2026 10:12:41 +0800 Subject: [PATCH 592/694] mm: shmem: ignore sysfs configs for shmem forced collapse According to Documentation/mm/transhuge.rst, MADV_COLLAPSE is expected to ignore any THP or mTHP interface settings. However, after commit 26c7d8413aaf ("mm: thp: support "THPeligible" semantics for mTHP with anonymous shmem"), performing MADV_COLLAPSE on shmem will depend on /sys/.../hugepages-2048kB/shmem_enabled being set to "inherit" (although that is the default), which can cause MADV_COLLAPSE to fail unexpectedly. This could cause a userspace-visible performance regression. Fix this by returning the result of shmem_huge_global_enabled() directly when MADV_COLLAPSE is requested, allowing PMD-order collapse while ignoring shmem THP/mTHP settings. Link: https://lore.kernel.org/063f655b4d6c4234f3aa27ed6ecab10283ecb880.1789351825.git.baolin.wang@linux.alibaba.com Fixes: 26c7d8413aaf ("mm: thp: support "THPeligible" semantics for mTHP with anonymous shmem") Signed-off-by: Baolin Wang Signed-off-by: Andrew Morton Reported-by: Kiryl Shutsemau (Meta) Closes: https://lore.kernel.org/all/20260908125105.1510704-7-kirill@shutemov.name/ Reviewed-by: Kiryl Shutsemau (Meta) Acked-by: Zi Yan Reviewed-by: Lance Yang Reviewed-by: Barry Song Acked-by: Qi Zheng Acked-by: David Hildenbrand (Arm) Reviewed-by: Lorenzo Stoakes (ARM) Cc: Dev Jain Cc: Hugh Dickins Cc: Liam R. Howlett Cc: Ryan Roberts Cc: --- mm/shmem.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/mm/shmem.c b/mm/shmem.c index 897fa2b61346..848316eaa7f4 100644 --- a/mm/shmem.c +++ b/mm/shmem.c @@ -1847,8 +1847,11 @@ unsigned long shmem_allowable_huge_orders(struct inode *inode, global_orders = shmem_huge_global_enabled(inode, index, write_end, shmem_huge_force, vma, vm_flags); - /* Tmpfs huge pages allocation */ - if (!vma || !vma_is_anon_shmem(vma)) + /* + * Tmpfs huge pages allocation or forced collapse ignores + * sysfs configs. + */ + if (!vma || !vma_is_anon_shmem(vma) || shmem_huge_force) return global_orders; /* From 347c6ed8cf7768883252e106994de62c7947f0a0 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Tue, 15 Sep 2026 22:23:47 +0200 Subject: [PATCH 593/694] alloc_tag: avoid implicit padding in uapi The implied padding causes a harmless warning when testing the uapi headers with -Wpadded that could in theory indicate incompatibilities or data leaks: ./usr/include/linux/alloc_tag.h:41:1: error: padding struct size to alignment boundary with 7 bytes [-Werror=padded] The code here is fine, but it's better to make the padding explicit and avoid the warning here. Link: https://lore.kernel.org/20260916065830.1619425-1-arnd@kernel.org Link: https://lore.kernel.org/20260915202404.3568029-1-arnd@kernel.org Fixes: 1d581ab2348c ("alloc_tag: add ioctl to /proc/allocinfo") Signed-off-by: Arnd Bergmann Signed-off-by: Andrew Morton Acked-by: Suren Baghdasaryan Acked-by: SJ Park Acked-by: Hao Ge --- include/uapi/linux/alloc_tag.h | 1 + 1 file changed, 1 insertion(+) diff --git a/include/uapi/linux/alloc_tag.h b/include/uapi/linux/alloc_tag.h index e3ad94444864..7d4618bea043 100644 --- a/include/uapi/linux/alloc_tag.h +++ b/include/uapi/linux/alloc_tag.h @@ -38,6 +38,7 @@ struct allocinfo_counter { __u64 bytes; __u64 calls; __u8 accurate; + __u8 pad[7]; } __attribute__((aligned(8))); struct allocinfo_tag_data { From 52ae167ce16609c7e6fffef588b3c2c26de2db19 Mon Sep 17 00:00:00 2001 From: SJ Park Date: Wed, 16 Sep 2026 06:50:19 -0700 Subject: [PATCH 594/694] mm/damon/core: don't skip damos_adjust_quota() while esz is not zero DAMOS could unexpectedly stop working when a user disables quota using the online parameters commit feature. Fix it by correcting a wrong quota unset check in damos_adjust_quota(). DAMON users could disable all quotas by unsetting time and size quotas, and removing all quota goals. The intention of disabling quotas would be making DAMOS run at full speed. When such quota disabled setup is detected, damos_adjust_quota() skips all its work. The skipped works include effective size quota (damos_quota->esz) updates and charged quota amount (damos_quota->charged_sz) resets. The intention is to avoid doing unnecessary work when quotas are disabled. However, users could do the setup while effective size quota is non-zero, by doing the disabling with the online DAMON parameters commit feature. In this case, because the effective size quota exists, DAMOS will keep working with the quota until it is fully charged. After the effective quota is fully charged, the charged quota amount (damos_quota->charged_sz) cannot be reset because damos_adjust_quota() skips it. Then, DAMOS stops working until the quota is newly set or DAMON is entirely restarted. The problem happens because damos_adjust_quota() assumes the user setup for disabling quota immediately disabled it. In reality, the quota is still working until the effective size quota is also updated to zero. Other logic for catching that uses damos_quota_is_set(), which understands the fact and therefore checks the effective size quota in addition to the user setup. Fix the issue by using damos_quota_is_set() in damos_adjust_quota() to determine if its works should be skipped. The user impact is a non-deterministic and unexpected DAMOS stop behavior. That is, users would disable quotas using the online parameters commit feature, expecting DAMOS will run at full speed. However, depending on the timing, the setup can be updated while the effective size quota is non-zero. Due to the above mentioned internal mechanism, DAMOS stops working instead of running at full speed. It is unexpected behavior. It is also non-deterministic because sometimes the setup is done when the effective size quota is zero, depending on the timing. It doesn't cause critical issues like crashes or leaks. Users can simply set a reasonable quota again, or restart DAMON. But definitely it is an unexpected and non-deterministic behavior that makes it difficult to reliably use. Also investigating the root cause of the behavior would be quite difficult. Link: https://lore.kernel.org/20260916135020.86483-1-sj@kernel.org Fixes: da87878010e5 ("mm/damon/sysfs: support online inputs update") Signed-off-by: SJ Park Signed-off-by: Andrew Morton Reviewed-by: Kunwu Chan Cc: # 5.19.x --- mm/damon/core.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/mm/damon/core.c b/mm/damon/core.c index 1764620903f2..ab3c4d754964 100644 --- a/mm/damon/core.c +++ b/mm/damon/core.c @@ -3147,7 +3147,7 @@ static void damos_adjust_quota(struct damon_ctx *c, struct damos *s) unsigned long cumulated_sz, cached_esz; unsigned int score, max_score = 0; - if (!quota->ms && !quota->sz && list_empty("a->goals)) + if (!damos_quota_is_set(quota)) return; /* First charge window */ From 22ab0647764c38924669673634ae333578d77768 Mon Sep 17 00:00:00 2001 From: Andrey Ryabinin Date: Wed, 16 Sep 2026 19:51:13 +0200 Subject: [PATCH 595/694] kasan: unpoison task stack below watermark only in generic mode CPU resume and BPF exception handling can discard stack frames without running their compiler-generated epilogues. Generic KASAN needs kasan_unpoison_task_stack_below() to clear the redzones left behind by those frames before the stack is reused. CONFIG_KASAN_STACK also enables this helper for SW_TAGS. The helper derives the stack base from an untagged stack pointer, so kasan_unpoison() writes KASAN_TAG_KERNEL (0xff) into the shadow for [base, watermark). For a vmapped task stack, vm_area->addr still carries the original random allocation tag. This creates a tag mismatch at the stack base even if that memory has never held an instrumented stack object. When the task exits and its stack is not cached, thread_stack_free_rcu() passes vm_area->addr to vfree(). In RCU callback context this reaches vfree_atomic(), whose llist_add() writes to the allocation base through the tagged pointer and triggers a false KASAN invalid-access report: BUG: KASAN: invalid-access in vfree_atomic+0x90/0x150 Write of size 8 at addr c2ffffc0a8f70000 by task rcuop/7/75 Pointer tag: [c2], memory tag: [ff] Call trace: __hwasan_store8_noabort+0xe8/0xf8 vfree_atomic+0x90/0x150 vfree+0x220/0x298 thread_stack_free_rcu+0x3c/0x4c rcu_do_batch+0x308/0xaf0 rcu_nocb_cb_kthread+0x33c/0x708 The deferred-free path started using the tagged vm_area->addr in commit 449e0b4ed5a1 ("fork: clean-up naming of vm_stack/vm_struct variables in vmap stacks code"). Previously it freed the untagged pointer derived from tsk->stack, so the write was never tag-checked. SW_TAGS does not need the blanket shadow reset to make subsequent stack accesses valid: untagged kernel pointers have the match-all 0xff tag, and the compiler initializes the shadow tags of instrumented stack objects before use. Return early unless CONFIG_KASAN_GENERIC is enabled. Keep the mode check in the common helper so it covers both resume and bpf_throw(). Link: https://lore.kernel.org/20260916175113.1327454-1-ryabinin.a.a@gmail.com Fixes: 449e0b4ed5a1 ("fork: clean-up naming of vm_stack/vm_struct variables in vmap stacks code") Signed-off-by: Andrey Ryabinin Signed-off-by: Andrew Morton Reported-by: Shaobo Huang Closes: https://lore.kernel.org/all/20260806123020.90869-1-huangshaobo3@xiaomi.com/ Assisted-by: LLM Cc: Alexander Potapenko Cc: Andrey Konovalov Cc: David Hildenbrand Cc: Dmitry Vyukov Cc: Lorenzo Stoakes Cc: Vincenzo Frascino Cc: --- mm/kasan/common.c | 13 ++++++++++++- 1 file changed, 12 insertions(+), 1 deletion(-) diff --git a/mm/kasan/common.c b/mm/kasan/common.c index b7d05c2a6d93..7d21c4db68d2 100644 --- a/mm/kasan/common.c +++ b/mm/kasan/common.c @@ -114,12 +114,23 @@ void kasan_unpoison_task_stack(struct task_struct *task) /* Unpoison the stack for the current task beyond a watermark sp value. */ asmlinkage void kasan_unpoison_task_stack_below(const void *watermark) { + void *base; + + /* + * Only the generic mode needs this. In the tag-based mode the + * compiler fully initializes the shadow of stack variables on + * function entry, so stale tags left below the watermark are + * harmless. + */ + if (!IS_ENABLED(CONFIG_KASAN_GENERIC)) + return; + /* * Calculate the task stack base address. Avoid using 'current' * because this function is called by early resume code which hasn't * yet set up the percpu register (%gs). */ - void *base = (void *)((unsigned long)watermark & ~(THREAD_SIZE - 1)); + base = (void *)((unsigned long)watermark & ~(THREAD_SIZE - 1)); kasan_unpoison(base, watermark - base, false); } From 711d886fe76e28854814759f51ee175e279ef8c7 Mon Sep 17 00:00:00 2001 From: "David Hildenbrand (Arm)" Date: Fri, 18 Sep 2026 18:34:24 +0200 Subject: [PATCH 596/694] MAINTAINERS: split up MEMORY MANAGEMENT - MEMORY POLICY AND MIGRATION Patch series "MAINTAINERS: rework MEMORY MANAGEMENT - MEMORY POLICY". Let's rework MEMORY MANAGEMENT - MEMORY POLICY AND MIGRATION. As I can use some maintenance help in that area, add Gregory as a new co-maintainer for the split out MEMORY MANAGEMENT - NUMA PLACEMENT section. This patch (of 3): Let's split it up into MIGRATION and NUMA PLACEMENT. The latter is a better fitting description for mempolicy.c. Make a best guess about which pieces existing reviewers are interested in. Move the split sections to keep alphabetical order. Link: https://lore.kernel.org/20260918-maintainers-mempolicy-v1-1-9a94cac6d135@kernel.org Signed-off-by: David Hildenbrand (Arm) Signed-off-by: Andrew Morton Acked-by: Lorenzo Stoakes (ARM) Acked-by: Zi Yan Reviewed-by: Joshua Hahn Reviewed-by: SJ Park Acked-by: Vlastimil Babka (SUSE) Cc: Alistair Popple Cc: Byungchul Park Cc: Gregory Price Cc: "Huang, Ying" Cc: Liam R. Howlett Cc: Matthew Brost Cc: Michal Hocko Cc: Mike Rapoport Cc: Rakie Kim Cc: Suren Baghdasaryan --- MAINTAINERS | 62 ++++++++++++++++++++++++++++++++--------------------- 1 file changed, 37 insertions(+), 25 deletions(-) diff --git a/MAINTAINERS b/MAINTAINERS index 140eafcbbd78..3f70a28df4b1 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -17208,31 +17208,6 @@ F: include/trace/events/ksm.h F: mm/ksm.c F: mm/mm_slot.h -MEMORY MANAGEMENT - MEMORY POLICY AND MIGRATION -M: Andrew Morton -M: David Hildenbrand -R: Zi Yan -R: Matthew Brost -R: Joshua Hahn -R: Rakie Kim -R: Byungchul Park -R: Gregory Price -R: Ying Huang -R: Alistair Popple -L: linux-mm@kvack.org -S: Maintained -W: http://www.linux-mm.org -T: git git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm -F: Documentation/ABI/testing/sysfs-kernel-mm-mempolicy -F: Documentation/ABI/testing/sysfs-kernel-mm-mempolicy-weighted-interleave -F: include/linux/mempolicy.h -F: include/uapi/linux/mempolicy.h -F: include/linux/migrate.h -F: include/linux/migrate_mode.h -F: mm/mempolicy.c -F: mm/migrate.c -F: mm/migrate_device.c - MEMORY MANAGEMENT - MGLRU (MULTI-GEN LRU) M: Andrew Morton R: Kairui Song @@ -17255,6 +17230,25 @@ F: include/linux/mmzone.h F: mm/vmscan.c F: mm/workingset.c +MEMORY MANAGEMENT - MIGRATION +M: Andrew Morton +M: David Hildenbrand +R: Zi Yan +R: Matthew Brost +R: Joshua Hahn +R: Byungchul Park +R: Gregory Price +R: Ying Huang +R: Alistair Popple +L: linux-mm@kvack.org +S: Maintained +W: http://www.linux-mm.org +T: git git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm +F: include/linux/migrate.h +F: include/linux/migrate_mode.h +F: mm/migrate.c +F: mm/migrate_device.c + MEMORY MANAGEMENT - MISC M: Andrew Morton M: David Hildenbrand @@ -17307,6 +17301,24 @@ F: mm/numa.c F: mm/numa_emulation.c F: mm/numa_memblks.c +MEMORY MANAGEMENT - NUMA PLACEMENT +M: Andrew Morton +M: David Hildenbrand +R: Zi Yan +R: Joshua Hahn +R: Rakie Kim +R: Gregory Price +R: Ying Huang +L: linux-mm@kvack.org +S: Maintained +W: http://www.linux-mm.org +T: git git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm +F: Documentation/ABI/testing/sysfs-kernel-mm-mempolicy +F: Documentation/ABI/testing/sysfs-kernel-mm-mempolicy-weighted-interleave +F: include/linux/mempolicy.h +F: include/uapi/linux/mempolicy.h +F: mm/mempolicy.c + MEMORY MANAGEMENT - OOM KILLER M: Michal Hocko R: David Rientjes From 4050a73a90f456887dc35d5bbfa4f4790eb5c2cf Mon Sep 17 00:00:00 2001 From: "David Hildenbrand (Arm)" Date: Fri, 18 Sep 2026 18:34:25 +0200 Subject: [PATCH 597/694] MAINTAINERS: move memory tiering under MEMORY MANAGEMENT - NUMA PLACEMENT NUMA PLACEMENT is a better place for memory tiering. My best guess is that existing MISC reviewers are not that interested in memory tiering, so don't carry any over. Link: https://lore.kernel.org/20260918-maintainers-mempolicy-v1-2-9a94cac6d135@kernel.org Signed-off-by: David Hildenbrand (Arm) Signed-off-by: Andrew Morton Acked-by: Lorenzo Stoakes (ARM) Acked-by: Zi Yan Reviewed-by: Joshua Hahn Reviewed-byt: SJ Park Acked-by: Vlastimil Babka (SUSE) Cc: Alistair Popple Cc: Byungchul Park Cc: Gregory Price Cc: "Huang, Ying" Cc: Liam R. Howlett Cc: Matthew Brost Cc: Michal Hocko Cc: Mike Rapoport Cc: Rakie Kim Cc: Suren Baghdasaryan --- MAINTAINERS | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/MAINTAINERS b/MAINTAINERS index 3f70a28df4b1..d69f112e5abb 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -17264,7 +17264,6 @@ W: http://www.linux-mm.org T: git git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm F: Documentation/ABI/testing/sysfs-kernel-mm F: Documentation/ABI/testing/sysfs-kernel-mm-cma -F: Documentation/ABI/testing/sysfs-kernel-mm-memory-tiers F: Documentation/ABI/testing/sysfs-kernel-mm-numa F: Documentation/admin-guide/mm/ F: Documentation/mm/ @@ -17272,7 +17271,6 @@ F: drivers/char/mem.c F: include/linux/cma.h F: include/linux/dmapool.h F: include/linux/ioremap.h -F: include/linux/memory-tiers.h F: include/linux/page_idle.h F: mm/backing-dev.c F: mm/cma.c @@ -17284,7 +17282,6 @@ F: mm/early_ioremap.c F: mm/fadvise.c F: mm/ioremap.c F: mm/mapping_dirty_helpers.c -F: mm/memory-tiers.c F: mm/page_idle.c F: mm/pgalloc-track.h F: mm/process_vm_access.c @@ -17313,10 +17310,13 @@ L: linux-mm@kvack.org S: Maintained W: http://www.linux-mm.org T: git git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm +F: Documentation/ABI/testing/sysfs-kernel-mm-memory-tiers F: Documentation/ABI/testing/sysfs-kernel-mm-mempolicy F: Documentation/ABI/testing/sysfs-kernel-mm-mempolicy-weighted-interleave +F: include/linux/memory-tiers.h F: include/linux/mempolicy.h F: include/uapi/linux/mempolicy.h +F: mm/memory-tiers.c F: mm/mempolicy.c MEMORY MANAGEMENT - OOM KILLER From b3f0c1a4e41f356c3126a676ad55c0f6881e7514 Mon Sep 17 00:00:00 2001 From: "David Hildenbrand (Arm)" Date: Fri, 18 Sep 2026 18:34:26 +0200 Subject: [PATCH 598/694] MAINTAINERS: make Gregory a co-maintainer of MEMORY MANAGEMENT - NUMA PLACEMENT Gregory is extremely familiar with NUMA placement handling and volunteer to help co-maintain the numa placement bits. So add him as a co-maintainer. Link: https://lore.kernel.org/20260918-maintainers-mempolicy-v1-3-9a94cac6d135@kernel.org Signed-off-by: David Hildenbrand (Arm) Signed-off-by: Andrew Morton Acked-by: Lorenzo Stoakes (ARM) Acked-by: Gregory Price (Meta) Acked-by: Zi Yan Reviewed-by: Joshua Hahn Acked-by: SJ Park Acked-by: Vlastimil Babka (SUSE) Cc: Alistair Popple Cc: Byungchul Park Cc: "Huang, Ying" Cc: Liam R. Howlett Cc: Matthew Brost Cc: Michal Hocko Cc: Mike Rapoport Cc: Rakie Kim Cc: Suren Baghdasaryan --- MAINTAINERS | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/MAINTAINERS b/MAINTAINERS index d69f112e5abb..9fcda07a32a0 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -17301,10 +17301,10 @@ F: mm/numa_memblks.c MEMORY MANAGEMENT - NUMA PLACEMENT M: Andrew Morton M: David Hildenbrand +M: Gregory Price R: Zi Yan R: Joshua Hahn R: Rakie Kim -R: Gregory Price R: Ying Huang L: linux-mm@kvack.org S: Maintained From 6cc8f549f6888e36036ab14d84a8ec46b4a393c0 Mon Sep 17 00:00:00 2001 From: Joseph Qi Date: Wed, 23 Sep 2026 08:39:20 +0800 Subject: [PATCH 599/694] MAINTAINERS: add Heming Zhao as ocfs2 reviewer Heming Zhao has contributed ocfs2 a lot recent years, both as a author and reviewer. So add him as ocfs2 reviewer. Link: https://lore.kernel.org/20260923003920.2382398-1-joseph.qi@linux.alibaba.com Signed-off-by: Joseph Qi Signed-off-by: Andrew Morton Acked-by: Mark Fasheh Cc: Heming Zhao Cc: Joel Becker --- MAINTAINERS | 1 + 1 file changed, 1 insertion(+) diff --git a/MAINTAINERS b/MAINTAINERS index 9fcda07a32a0..23de9fc4c2a7 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -20568,6 +20568,7 @@ ORACLE CLUSTER FILESYSTEM 2 (OCFS2) M: Mark Fasheh M: Joel Becker M: Joseph Qi +R: Heming Zhao L: ocfs2-devel@lists.linux.dev S: Supported W: http://ocfs2.wiki.kernel.org From 976d5e0ddac885b0219252f004f6c561c344df9d Mon Sep 17 00:00:00 2001 From: "Lorenzo Stoakes (ARM)" Date: Wed, 23 Sep 2026 18:45:41 +0100 Subject: [PATCH 600/694] mm/vma: predicate setting mmap_prepare VMA fields on new vma alloc It only makes sense to manipulate VMA fields if a new VMA was allocated, rather than merged. VMA merging does not compare vm_ops or vm_private_data, so a merged VMA keeps its own, which is also what the legacy f_op->mmap path does since it never touches an existing VMA. Currently, these fields will get overwritten by whatever state is established in the mmap_prepare hook, and if the VMA was merged, vm_ops->mapped will not have been called, so this could destructively clear existing state without replacing it with anything valid. There is an implicit requirement that vm_private_data and vm_ops are fungible across VMAs which means that losing the 'new' state is fine. However in this case the 'old' state is being overwritten by potentially invalid 'new' state, so this must be rectified. Additionally constify have_mmap_prepare while here. All existing in-tree users either derive state from the file or are unmergeable due to VMA flags, so this has no direct impact. Link: https://lore.kernel.org/20260923-fix-mmap-prepare-overwrite-v1-1-3b3f1bfcdf5e@kernel.org Fixes: c84bf6dd2b83 ("mm: introduce new .mmap_prepare() file callback") Signed-off-by: Lorenzo Stoakes (ARM) Signed-off-by: Andrew Morton Reviewed-by: Suren Baghdasaryan Reviewed-by: Gregory Price (Meta) Acked-by: Zi Yan Acked-by: Vlastimil Babka (SUSE) Cc: Liam R. Howlett Cc: Jann Horn Cc: Pedro Falcato Cc: --- mm/vma.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/mm/vma.c b/mm/vma.c index 9f0a0acf694a..6cde67883fb0 100644 --- a/mm/vma.c +++ b/mm/vma.c @@ -2849,7 +2849,7 @@ static unsigned long __mmap_region(struct file *file, unsigned long addr, { struct mm_struct *mm = current->mm; struct vm_area_struct *vma = NULL; - bool have_mmap_prepare = file && file->f_op->mmap_prepare; + const bool have_mmap_prepare = file && file->f_op->mmap_prepare; VMA_ITERATOR(vmi, mm, addr); const pgoff_t anon_pgoff = addr >> PAGE_SHIFT; MMAP_STATE(map, mm, &vmi, addr, len, pgoff, anon_pgoff, vma_flags, file); @@ -2892,7 +2892,7 @@ static unsigned long __mmap_region(struct file *file, unsigned long addr, allocated_new = true; } - if (have_mmap_prepare) + if (have_mmap_prepare && allocated_new) set_vma_user_defined_fields(vma, &map); __mmap_complete(&map, vma); From db6365ced4d5855e321f772b240c0e473bcfcdd5 Mon Sep 17 00:00:00 2001 From: Pavankumar Kondeti Date: Fri, 25 Sep 2026 15:25:12 +0530 Subject: [PATCH 601/694] workqueue: Fix NULL current_pwq deref in flush dependency check check_flush_dependency() uses current_wq_worker() to determine whether the caller is a workqueue worker and then dereferences worker->current_pwq to test whether the current workqueue is WQ_MEM_RECLAIM. current_wq_worker() only means that %current has PF_WQ_WORKER set. A kworker can reach check_flush_dependency() while it is not executing a work item. One such path is worker_thread() acting as the pool manager, where create_worker() does GFP_KERNEL allocation and the allocation path invokes the OOM notifier. In that state worker->current_pwq is NULL because current_pwq is set only by process_one_work() and cleared again after the work function returns. [ 416.760634][ T375] Call trace: [ 416.760638][ T375] check_flush_dependency+0x80/0x120 (P) [ 416.760648][ T375] __flush_work+0x98/0x224 [ 416.760657][ T375] flush_work+0x30/0x44 [ 416.760665][ T375] ... [ 416.760710][ T375] blocking_notifier_call_chain+0x58/0xa0 [ 416.760719][ T375] out_of_memory+0xb4/0x458 [ 416.760730][ T375] __alloc_pages_may_oom+0x11c/0x1a8 [ 416.760739][ T375] __alloc_pages_slowpath+0x314/0x46c [ 416.760746][ T375] __alloc_frozen_pages_noprof+0x110/0x1a4 [ 416.760753][ T375] new_slab+0x12c/0x484 [ 416.760759][ T375] ___slab_alloc+0x7a8/0xc7c [ 416.760765][ T375] __slab_alloc+0x74/0xd8 [ 416.760772][ T375] __kmalloc_cache_node_noprof+0x2ac/0x304 [ 416.760779][ T375] alloc_worker+0x28/0x60 [ 416.760785][ T375] create_worker+0x4c/0x20c [ 416.760790][ T375] worker_thread+0xe8/0x2b8 [ 416.760796][ T375] kthread+0x1a8/0x200 [ 416.760805][ T375] ret_from_fork+0x10/0x20 Guard the WQ_MEM_RECLAIM-worker warning with worker->current_pwq. If the kworker is not currently executing a work item, there is no current workqueue to diagnose with that warning. The PF_MEMALLOC warning is left unchanged so explicit reclaim context flushing a !WQ_MEM_RECLAIM target is still reported. Fixes: fca839c00a12 ("workqueue: warn if memory reclaim tries to flush !WQ_MEM_RECLAIM workqueue") Cc: stable@vger.kernel.org Assisted-by: LLM Signed-off-by: Pavankumar Kondeti Signed-off-by: Tejun Heo --- kernel/workqueue.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/kernel/workqueue.c b/kernel/workqueue.c index b8bec1689b7a..4a73d305d88a 100644 --- a/kernel/workqueue.c +++ b/kernel/workqueue.c @@ -3918,7 +3918,7 @@ static void check_flush_dependency(struct workqueue_struct *target_wq, WARN_ONCE(current->flags & PF_MEMALLOC, "workqueue: PF_MEMALLOC task %d(%s) is flushing !WQ_MEM_RECLAIM %s:%ps", current->pid, current->comm, target_wq->name, target_func); - WARN_ONCE(worker && ((worker->current_pwq->wq->flags & + WARN_ONCE(worker && worker->current_pwq && ((worker->current_pwq->wq->flags & (WQ_MEM_RECLAIM | __WQ_LEGACY)) == WQ_MEM_RECLAIM), "workqueue: WQ_MEM_RECLAIM %s:%ps is flushing !WQ_MEM_RECLAIM %s:%ps", worker->current_pwq->wq->name, worker->current_func, From a442c8a89fb9c531f9d6ab86120c31decd778d51 Mon Sep 17 00:00:00 2001 From: Florian Westphal Date: Fri, 4 Sep 2026 20:03:07 +0200 Subject: [PATCH 602/694] netfilter: ipset: do not update comments from kernel-side adds 'Fixes' commit stopped calling ip_set_init_comment() for hash types from kernel-side-adds (xtables .. -j SET). ip_set_init_comment() says: "The kadt functions don't use the comment extensions in any way." But bitmap set type calls the function from kadt cb too. While this appears to be safe (serialized via the set spinlock), it seems better to not call the init function either, least of all to keep behaviour consistent. ip_set_list calls ip_set_init_comment() only from uadt cb, it can be kept as-is. This was triggered by yet another LLM review, hinting that the existing rcu_dereference_protected() cannot be downgraded to only check if the nfnl mutex is held. Fixes: f30415929be8 ("netfilter: ipset: do not update comments from kernel-side hash adds") Signed-off-by: Florian Westphal Signed-off-by: Pablo Neira Ayuso --- net/netfilter/ipset/ip_set_bitmap_gen.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/netfilter/ipset/ip_set_bitmap_gen.h b/net/netfilter/ipset/ip_set_bitmap_gen.h index d6a7e6604542..ae376fa3e7a3 100644 --- a/net/netfilter/ipset/ip_set_bitmap_gen.h +++ b/net/netfilter/ipset/ip_set_bitmap_gen.h @@ -159,7 +159,7 @@ mtype_add(struct ip_set *set, void *value, const struct ip_set_ext *ext, if (SET_WITH_COUNTER(set)) ip_set_init_counter(ext_counter(x, set), ext); - if (SET_WITH_COMMENT(set)) + if (SET_WITH_COMMENT(set) && !ext->target) ip_set_init_comment(set, ext_comment(x, set), ext); if (SET_WITH_SKBINFO(set)) ip_set_init_skbinfo(ext_skbinfo(x, set), ext); From 72d3fcf802c45d00b300f25b848a93c3a2bd7c7e Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Sun, 27 Sep 2026 13:55:01 -0700 Subject: [PATCH 603/694] Linux 7.3-rc5 --- Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/Makefile b/Makefile index 751a08643bf8..6b8b812f23b9 100644 --- a/Makefile +++ b/Makefile @@ -2,7 +2,7 @@ VERSION = 7 PATCHLEVEL = 3 SUBLEVEL = 0 -EXTRAVERSION = -rc4 +EXTRAVERSION = -rc5 NAME = Baby Opossum Posse # *DOCUMENTATION* From 4febc02cd02f021f2b6d0768332be643ef14a12d Mon Sep 17 00:00:00 2001 From: Felix Fietkau Date: Sat, 26 Sep 2026 14:02:16 +0200 Subject: [PATCH 604/694] wifi: mac80211: set info->band for 802.3 encap offload frames Unlike the 802.11 TX paths, ieee80211_8023_xmit() does not set info->band from the channel context on non-MLD interfaces. It stays at 0, so ieee80211_get_tx_rates() uses the wrong band for these frames. Set it the same way as ieee80211_build_hdr(): drop the frame if there is no channel context, and leave the band at 0 on MLD interfaces. Reported-by: Andrea Pesaresi Fixes: 50ff477a8639 ("mac80211: add 802.11 encapsulation offloading support") Cc: stable@vger.kernel.org Signed-off-by: Felix Fietkau Link: https://patch.msgid.link/20260926120216.4054712-1-nbd@nbd.name Signed-off-by: Johannes Berg --- net/mac80211/tx.c | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/net/mac80211/tx.c b/net/mac80211/tx.c index c8217f6d146e..b88acb18e04d 100644 --- a/net/mac80211/tx.c +++ b/net/mac80211/tx.c @@ -4735,6 +4735,7 @@ static void ieee80211_8023_xmit(struct ieee80211_sub_if_data *sdata, { struct ieee80211_tx_info *info; struct ieee80211_local *local = sdata->local; + struct ieee80211_chanctx_conf *chanctx_conf; struct tid_ampdu_tx *tid_tx = NULL; struct sk_buff *seg, *next; unsigned int skbs = 0, len = 0; @@ -4784,6 +4785,16 @@ static void ieee80211_8023_xmit(struct ieee80211_sub_if_data *sdata, sdata = container_of(sdata->bss, struct ieee80211_sub_if_data, u.ap); + /* MLD transmissions must not rely on the band */ + if (!ieee80211_vif_is_mld(&sdata->vif)) { + chanctx_conf = rcu_dereference(sdata->vif.bss_conf.chanctx_conf); + if (unlikely(!chanctx_conf)) { + kfree_skb_list(skb); + return; + } + info->band = chanctx_conf->def.chan->band; + } + info->flags |= IEEE80211_TX_CTL_HW_80211_ENCAP; info->control.vif = &sdata->vif; From 56e236759f0a8dc910c2ae7cd3c161f298ee5794 Mon Sep 17 00:00:00 2001 From: Ruide Cao Date: Wed, 23 Sep 2026 02:02:23 +0800 Subject: [PATCH 605/694] wifi: mac80211: reject invalid 320 MHz CSA bandwidth The HT/VHT channel definition validator can receive a 320 MHz bandwidth indication from a received CSA frame on a non-6 GHz link. It warns for this unsupported width but continues with an uninitialized vht_operation.chan_width, which is then read by ieee80211_chandef_vht_oper(). With panic_on_warn enabled, this lets a received frame panic the kernel. Reject the channel definition before entering the VHT operation conversion. This preserves the existing CSA fallback while avoiding both the warning and the uninitialized read. Fixes: 21c3f8f95554 ("wifi: mac80211: refactor STA CSA parsing flows") Cc: stable@vger.kernel.org Reported-by: Vega Assisted-by: LLM Signed-off-by: Ruide Cao Signed-off-by: Ren Wei Link: https://patch.msgid.link/bd68e360e06135c750397cb6be003a01a834a179.1787222001.git.vega.cover-letter@nebusec.ai Signed-off-by: Johannes Berg --- net/mac80211/spectmgmt.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/net/mac80211/spectmgmt.c b/net/mac80211/spectmgmt.c index 880f4625775d..f5e4970869b6 100644 --- a/net/mac80211/spectmgmt.c +++ b/net/mac80211/spectmgmt.c @@ -111,8 +111,8 @@ validate_chandef_by_ht_vht_oper(struct ieee80211_sub_if_data *sdata, switch (chan_width) { case NL80211_CHAN_WIDTH_320: - WARN_ON(1); - break; + chandef->chan = NULL; + return; case NL80211_CHAN_WIDTH_160: vht_oper.chan_width = IEEE80211_VHT_CHANWIDTH_80MHZ; vht_oper.center_freq_seg1_idx = vht_oper.center_freq_seg0_idx; From 6f63e919fe1e335b8abcb3a28bfd4804a98d875a Mon Sep 17 00:00:00 2001 From: "Cen Zhang (Microsoft)" Date: Thu, 24 Sep 2026 22:44:08 -0400 Subject: [PATCH 606/694] wifi: mac80211: fix slab-out-of-bounds read in ieee80211_monitor_select_queue() ieee80211_monitor_select_queue() validates the skb length using skb->len before dereferencing pointers into skb->data to access the 802.11 header. However, skb->len includes data in both the linear head buffer and non-linear fragments/pages. When AF_PACKET sends a packet large enough to become non-linear, the 802.11 header at skb->data + len_rthdr may extend past the linear head buffer even though skb->len appears sufficient. This leads to a slab-out-of-bounds read when accessing hdr->frame_control, as the kernel reads beyond the allocated skb head buffer: BUG: KASAN: slab-out-of-bounds in ieee80211_monitor_select_queue+0x1ed/0x220 syzbot has hit the same issues. Fix this by checking skb_headlen(skb) (which gives the length of the linear data region) instead of skb->len before dereferencing skb->data pointers. This ensures the required bytes are actually present in the linear portion of the skb. Apply the same fix to ieee80211_validate_radiotap_len() which has the same class of bug: it uses skb->len to validate accesses to skb->data, and to the corresponding checks in ieee80211_monitor_start_xmit(): for drivers advertising NETIF_F_SG the skb is not linearized before ndo_start_xmit, so the 802.11 header can be read past the linear buffer there as well. Fixes: cf0277e714a0 ("mac80211: fix skb buffering issue") Fixes: 9b8a74e3482f ("[MAC80211]: Improve sanity checks on injected packets") Reported-by: AutonomousCodeSecurity@microsoft.com Reported-by: syzbot+610e40369bc02181bad0@syzkaller.appspotmail.com Closes: https://syzkaller.appspot.com/bug?extid=610e40369bc02181bad0 Reported-by: syzbot+878643e0580bc580f883@syzkaller.appspotmail.com Closes: https://syzkaller.appspot.com/bug?extid=878643e0580bc580f883 Assisted-by: GitHub-Copilot:claude-opus-4.6 Signed-off-by: Cen Zhang (Microsoft) Reviewed-by: Francis Perron Link: https://patch.msgid.link/20260925024408.32143-1-cenzhang@linux.microsoft.com Signed-off-by: Johannes Berg --- net/mac80211/iface.c | 4 ++-- net/mac80211/tx.c | 10 +++++----- 2 files changed, 7 insertions(+), 7 deletions(-) diff --git a/net/mac80211/iface.c b/net/mac80211/iface.c index 889c32fd8de1..88942317fec8 100644 --- a/net/mac80211/iface.c +++ b/net/mac80211/iface.c @@ -998,8 +998,8 @@ static u16 ieee80211_monitor_select_queue(struct net_device *dev, len_rthdr = ieee80211_get_radiotap_len(skb->data); hdr = (struct ieee80211_hdr *)(skb->data + len_rthdr); - if (skb->len < len_rthdr + 2 || - skb->len < len_rthdr + ieee80211_hdrlen(hdr->frame_control)) + if (skb_headlen(skb) < len_rthdr + 2 || + skb_headlen(skb) < len_rthdr + ieee80211_hdrlen(hdr->frame_control)) return 0; /* doesn't matter, frame will be dropped */ return ieee80211_select_queue_80211(sdata, skb, hdr); diff --git a/net/mac80211/tx.c b/net/mac80211/tx.c index b88acb18e04d..d51f810dd2c5 100644 --- a/net/mac80211/tx.c +++ b/net/mac80211/tx.c @@ -2091,7 +2091,7 @@ static bool ieee80211_validate_radiotap_len(struct sk_buff *skb) (struct ieee80211_radiotap_header *)skb->data; /* check for not even having the fixed radiotap header part */ - if (unlikely(skb->len < sizeof(struct ieee80211_radiotap_header))) + if (unlikely(skb_headlen(skb) < sizeof(struct ieee80211_radiotap_header))) return false; /* too short to be possibly valid */ /* is it a header version we can trust to find length from? */ @@ -2099,7 +2099,7 @@ static bool ieee80211_validate_radiotap_len(struct sk_buff *skb) return false; /* only version 0 is supported */ /* does the skb contain enough to deliver on the alleged length? */ - if (unlikely(skb->len < ieee80211_get_radiotap_len(skb->data))) + if (unlikely(skb_headlen(skb) < ieee80211_get_radiotap_len(skb->data))) return false; /* skb too short for claimed rt header extent */ return true; @@ -2388,13 +2388,13 @@ netdev_tx_t ieee80211_monitor_start_xmit(struct sk_buff *skb, skb_set_network_header(skb, len_rthdr); skb_set_transport_header(skb, len_rthdr); - if (skb->len < len_rthdr + 2) + if (skb_headlen(skb) < len_rthdr + 2) goto fail; hdr = (struct ieee80211_hdr *)(skb->data + len_rthdr); hdrlen = ieee80211_hdrlen(hdr->frame_control); - if (skb->len < len_rthdr + hdrlen) + if (skb_headlen(skb) < len_rthdr + hdrlen) goto fail; /* @@ -2402,7 +2402,7 @@ netdev_tx_t ieee80211_monitor_start_xmit(struct sk_buff *skb, * carrying a rfc1042 header */ if (ieee80211_is_data(hdr->frame_control) && - skb->len >= len_rthdr + hdrlen + sizeof(rfc1042_header) + 2) { + skb_headlen(skb) >= len_rthdr + hdrlen + sizeof(rfc1042_header) + 2) { u8 *payload = (u8 *)hdr + hdrlen; if (ether_addr_equal(payload, rfc1042_header)) From 024e05f73a4c1263d031614bc36700ec135eecb4 Mon Sep 17 00:00:00 2001 From: Chengfeng Ye Date: Sun, 27 Sep 2026 19:04:50 +0800 Subject: [PATCH 607/694] Bluetooth: hci_conn: Lock parent access during enhanced SCO setup Bluetooth: hci_conn: Lock parent access during enhanced SCO setup hci_enhanced_setup_sync() runs on the request workqueue without the hci_dev_lock held by its caller when setup was queued. Its CVSD capability check and find_next_esco_param() dereference conn->parent while the receive workqueue can unlink and release that parent. The following interleaving can cause a use-after-free: hci_enhanced_setup_sync() hci_disconn_complete_evt() load conn->parent hci_dev_lock() hci_conn_del(ACL parent) unlink SCO child drop link's parent reference clear child->parent release ACL parent bt_link_release() kfree(parent) hci_dev_unlock() read parent->features[0][3] The reference held for the queued SCO child does not keep its ACL parent alive after unlinking. Commit 42de40abe25d ("Bluetooth: hci_conn: fix the SCO setup context lifetime") protects the child stored in the queued context, but leaves these parent accesses unprotected. With a 40 ms diagnostic delay after loading conn->parent, an instrumented kernel based on fd179f8a05be, which already contains 42de40abe25d, reported: BUG: KASAN: slab-use-after-free in hci_enhanced_setup_sync+0xda5/0xdf0 Read of size 1 at addr ffff888102204047 by task kworker/u17:0/93 Call Trace: hci_enhanced_setup_sync+0xda5/0xdf0 hci_cmd_sync_work+0x13c/0x290 process_one_work+0x6b4/0x10e0 Allocated by task 92: __hci_conn_add+0x304/0x1df0 hci_connect_acl+0x349/0x3e0 hci_connect_sco+0x3b/0x9a0 sco_sock_connect+0x475/0xca0 Freed by task 94: kfree+0x121/0x3c0 bt_link_release+0x79/0xa0 device_release+0xc8/0x240 kobject_put+0x14d/0x280 hci_conn_del+0x524/0xe30 hci_disconn_complete_evt+0x403/0x8c0 hci_event_packet+0x71b/0xb20 hci_rx_work+0x293/0x730 The accessed address is 71 bytes into the freed ACL parent, at its features[0][3] byte; the queued SCO child is a different object. The diagnostic preserves the loaded parent across the delay, matching the unmodified compiled capability check, and does not change the parent references or teardown path. Hold hci_dev_lock() across the codec switch, including every call to find_next_esco_param(), and release it on all selection errors. Keep configure_datapath_sync() outside the critical section because it waits for HCI events. Preserve parameter selection and existing return values. Fixes: e07a06b4eb41 ("Bluetooth: Convert SCO configure_datapath to hci_sync") Cc: stable@vger.kernel.org Assisted-by: GPT-6-Astra Signed-off-by: Chengfeng Ye Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/hci_conn.c | 18 +++++++++++++----- 1 file changed, 13 insertions(+), 5 deletions(-) diff --git a/net/bluetooth/hci_conn.c b/net/bluetooth/hci_conn.c index 96195d2fd10f..cf44452e0766 100644 --- a/net/bluetooth/hci_conn.c +++ b/net/bluetooth/hci_conn.c @@ -302,11 +302,13 @@ static int hci_enhanced_setup_sync(struct hci_dev *hdev, void *data) cp.tx_bandwidth = cpu_to_le32(0x00001f40); cp.rx_bandwidth = cpu_to_le32(0x00001f40); + hci_dev_lock(hdev); + switch (conn->codec.id) { case BT_CODEC_MSBC: if (!find_next_esco_param(conn, esco_param_msbc, ARRAY_SIZE(esco_param_msbc))) - return -EINVAL; + goto unlock; param = &esco_param_msbc[conn->attempt - 1]; cp.tx_coding_format.id = 0x05; @@ -332,7 +334,7 @@ static int hci_enhanced_setup_sync(struct hci_dev *hdev, void *data) case BT_CODEC_TRANSPARENT: if (!find_next_esco_param(conn, esco_param_msbc, ARRAY_SIZE(esco_param_msbc))) - return -EINVAL; + goto unlock; param = &esco_param_msbc[conn->attempt - 1]; cp.tx_coding_format.id = 0x03; @@ -359,11 +361,11 @@ static int hci_enhanced_setup_sync(struct hci_dev *hdev, void *data) if (conn->parent && lmp_esco_capable(conn->parent)) { if (!find_next_esco_param(conn, esco_param_cvsd, ARRAY_SIZE(esco_param_cvsd))) - return -EINVAL; + goto unlock; param = &esco_param_cvsd[conn->attempt - 1]; } else { if (conn->attempt > ARRAY_SIZE(sco_param_cvsd)) - return -EINVAL; + goto unlock; param = &sco_param_cvsd[conn->attempt - 1]; } cp.tx_coding_format.id = 2; @@ -386,9 +388,11 @@ static int hci_enhanced_setup_sync(struct hci_dev *hdev, void *data) cp.out_transport_unit_size = 16; break; default: - return -EINVAL; + goto unlock; } + hci_dev_unlock(hdev); + cp.retrans_effort = param->retrans_effort; cp.pkt_type = __cpu_to_le16(param->pkt_type); cp.max_latency = __cpu_to_le16(param->max_latency); @@ -397,6 +401,10 @@ static int hci_enhanced_setup_sync(struct hci_dev *hdev, void *data) return -EIO; return 0; + +unlock: + hci_dev_unlock(hdev); + return -EINVAL; } static bool hci_setup_sync_conn(struct hci_conn *conn, __u16 handle) From 1a572db43c057566771b485598ad2e66cb390821 Mon Sep 17 00:00:00 2001 From: Chengfeng Ye Date: Sun, 27 Sep 2026 01:29:35 +0800 Subject: [PATCH 608/694] Bluetooth: hci_core: Serialize fragmented ISO packet queueing The fragmented path in hci_queue_iso() calls __skb_queue_tail() without holding conn->data_q.lock. The socket lock serializes senders, but the transmit worker can concurrently remove packets from this queue using skb_dequeue(). After an insertion saves the old tail pointer, hci_sched_iso() can dequeue that packet and pass it to the driver. The driver can free the packet before the insertion resumes and writes to the old tail's next pointer, causing a use-after-free. Concurrent updates can also corrupt the queue. KASAN reported: BUG: KASAN: slab-use-after-free in hci_send_iso+0xc7d/0xe10 Call Trace: hci_send_iso+0xc7d/0xe10 iso_sock_sendmsg+0x710/0x900 __sys_sendto+0x34a/0x3a0 Allocated by task 86: __alloc_skb+0xdd/0x820 alloc_skb_with_frags+0x7e/0x770 sock_alloc_send_pskb+0x67a/0x820 bt_skb_sendmsg.constprop.0+0xc0/0x6c0 iso_sock_sendmsg+0x44e/0x900 Freed by task 90: kmem_cache_free+0xcb/0x3d0 vhci_read+0x33f/0x4d0 vfs_read+0x177/0xa20 Hold the queue lock across the entire fragment batch, as hci_queue_acl() does, to serialize insertion against dequeue while preserving fragment ordering. Fixes: 26afbd826ee3 ("Bluetooth: Add initial implementation of CIS connections") Cc: stable@vger.kernel.org Signed-off-by: Chengfeng Ye Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/hci_core.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/net/bluetooth/hci_core.c b/net/bluetooth/hci_core.c index bbe391525530..5d1cdb1e7d8b 100644 --- a/net/bluetooth/hci_core.c +++ b/net/bluetooth/hci_core.c @@ -3327,6 +3327,8 @@ static void hci_queue_iso(struct hci_conn *conn, struct sk_buff_head *queue, skb_shinfo(skb)->frag_list = NULL; + spin_lock_bh(&queue->lock); + __skb_queue_tail(queue, skb); do { @@ -3341,6 +3343,8 @@ static void hci_queue_iso(struct hci_conn *conn, struct sk_buff_head *queue, __skb_queue_tail(queue, skb); } while (list); + + spin_unlock_bh(&queue->lock); } bt_dev_dbg(hdev, "hcon %p queued %d", conn, skb_queue_len(queue)); From b1f24c1ab52345e810c42a1a81286cb5a7c92252 Mon Sep 17 00:00:00 2001 From: Nguyen Ngoc Thang Date: Sun, 27 Sep 2026 22:18:02 +0700 Subject: [PATCH 609/694] Bluetooth: hci_sync: don't drain cmd_sync backlog on unregister hci_unregister_dev() disables cmd_work and cmd_timer, then calls hci_cmd_sync_clear(), whose cancel_work_sync() waits for hci_cmd_sync_work() to return. That worker dequeues and runs every entry on cmd_sync_work_list. With cmd_work disabled nothing reaches the controller, so each queued HCI command waits the full HCI_CMD_TIMEOUT (2s). The backlog has no bound. Userspace can keep queuing MGMT commands such as MGMT_OP_GET_CLOCK_INFO against a controller that doesn't answer. Closing /dev/vhci then blocks in vhci_release() for backlog * 2s: INFO: task syz-executor:5749 blocked for more than 143 seconds. cancel_work_sync hci_cmd_sync_clear hci_unregister_dev vhci_release In a local reproduction the backlog held more than 5000 entries, which comes to hours of hang. Stop the worker from taking new entries once HCI_UNREGISTER is set, and wake any request still waiting with -ENODEV before cancelling the work. The entries left over are destroyed with -ECANCELED by the existing sweep in hci_cmd_sync_clear(). They stay on the list until the work has stopped, so hci_cmd_sync_dequeue() and friends still see them. A callback already running may still issue another command, which delays unregister by at most one timeout per remaining command, not by the whole backlog. Fixes: 008ee9eb8a11 ("Bluetooth: hci_sync: Fix not processing all entries on cmd_sync_work") Reported-by: syzbot+217e3f1283cafe80586e@syzkaller.appspotmail.com Closes: https://syzkaller.appspot.com/bug?extid=217e3f1283cafe80586e Signed-off-by: Nguyen Ngoc Thang Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/hci_sync.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/net/bluetooth/hci_sync.c b/net/bluetooth/hci_sync.c index 74e2b04c84b2..e41fc1427d34 100644 --- a/net/bluetooth/hci_sync.c +++ b/net/bluetooth/hci_sync.c @@ -312,6 +312,10 @@ static void hci_cmd_sync_work(struct work_struct *work) while (1) { struct hci_cmd_sync_work_entry *entry; + /* Leave the backlog to hci_cmd_sync_clear() */ + if (hci_dev_test_flag(hdev, HCI_UNREGISTER)) + break; + mutex_lock(&hdev->cmd_sync_work_lock); entry = list_first_entry_or_null(&hdev->cmd_sync_work_list, struct hci_cmd_sync_work_entry, @@ -658,6 +662,8 @@ void hci_cmd_sync_clear(struct hci_dev *hdev) { struct hci_cmd_sync_work_entry *entry, *tmp; + /* cmd_work is disabled, the pending request can only time out */ + hci_cmd_sync_cancel_sync(hdev, ENODEV); cancel_work_sync(&hdev->cmd_sync_work); cancel_work_sync(&hdev->reenable_adv_work); From 8ed67535fdff40ca652bcd4ac7da68d11d01bd34 Mon Sep 17 00:00:00 2001 From: Chengfeng Ye Date: Sun, 27 Sep 2026 01:28:31 +0800 Subject: [PATCH 610/694] Bluetooth: hci_sync: Fix inquiry cache use-after-free The sync command worker holds hdev->req_lock, but inquiry-cache updates and flushes use hdev->lock. Both hci_acl_create_conn_sync() and hci_stop_discovery_sync() look up entries and read their fields without taking hdev->lock. After either lookup returns, a concurrent HCIINQUIRY ioctl can acquire hdev->lock and flush the cache, freeing the entry. The worker then reads the freed entry while preparing a create-connection or remote-name-cancel command. The list traversal also races with cache updates and removal. KASAN reported these accesses: BUG: KASAN: slab-use-after-free in hci_acl_create_conn_sync+0x5f1/0x650 Workqueue: hci0 hci_cmd_sync_work Call Trace: hci_acl_create_conn_sync+0x5f1/0x650 hci_cmd_sync_work+0x13c/0x290 Allocated by task 90: hci_inquiry_cache_update+0x3e6/0x7d0 hci_inquiry_result_evt+0x3cb/0x560 Freed by task 93: hci_inquiry_cache_flush+0x111/0x2b0 hci_inquiry+0x2f2/0x780 hci_sock_ioctl+0x269/0x5f0 BUG: KASAN: slab-use-after-free in hci_stop_discovery_sync+0x3b1/0x3c0 Workqueue: hci0 hci_cmd_sync_work Call Trace: hci_stop_discovery_sync+0x3b1/0x3c0 hci_cmd_sync_work+0x173/0x300 Allocated by task 86: hci_inquiry_cache_update+0x483/0x940 hci_inquiry_result_evt+0x3cb/0x560 Freed by task 91: hci_inquiry_cache_flush+0x13e/0x2f0 hci_inquiry+0x2f2/0x780 hci_sock_ioctl+0x269/0x5f0 Hold hdev->lock across each lookup and all reads from its result. Copy the remote address before unlocking so discovery cancellation does not retain a cache entry pointer. Release the lock before sending synchronous HCI commands, since their completion handlers may need the same lock. Fixes: cf75ad8b41d2 ("Bluetooth: hci_sync: Convert MGMT_SET_POWERED") Fixes: 45340097ce6e ("Bluetooth: hci_conn: Only do ACL connections sequentially") Cc: stable@vger.kernel.org Signed-off-by: Chengfeng Ye Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/hci_sync.c | 13 +++++++++++-- 1 file changed, 11 insertions(+), 2 deletions(-) diff --git a/net/bluetooth/hci_sync.c b/net/bluetooth/hci_sync.c index e41fc1427d34..c01c8b58d9e8 100644 --- a/net/bluetooth/hci_sync.c +++ b/net/bluetooth/hci_sync.c @@ -5770,6 +5770,7 @@ int hci_stop_discovery_sync(struct hci_dev *hdev) { struct discovery_state *d = &hdev->discovery; struct inquiry_entry *e; + bdaddr_t addr; int err; bt_dev_dbg(hdev, "state %u", hdev->discovery.state); @@ -5805,15 +5806,21 @@ int hci_stop_discovery_sync(struct hci_dev *hdev) return 0; if (d->state == DISCOVERY_RESOLVING || d->state == DISCOVERY_STOPPING) { + hci_dev_lock(hdev); e = hci_inquiry_cache_lookup_resolve(hdev, BDADDR_ANY, NAME_PENDING); - if (!e) + if (!e) { + hci_dev_unlock(hdev); return 0; + } + + bacpy(&addr, &e->data.bdaddr); + hci_dev_unlock(hdev); /* Ignore cancel errors since it should interfere with stopping * of the discovery. */ - hci_remote_name_cancel_sync(hdev, &e->data.bdaddr); + hci_remote_name_cancel_sync(hdev, &addr); } return 0; @@ -7242,6 +7249,7 @@ static int hci_acl_create_conn_sync(struct hci_dev *hdev, void *data) bacpy(&cp.bdaddr, &conn->dst); cp.pscan_rep_mode = 0x02; + hci_dev_lock(hdev); ie = hci_inquiry_cache_lookup(hdev, &conn->dst); if (ie) { if (inquiry_entry_age(ie) <= INQUIRY_ENTRY_AGE_MAX) { @@ -7253,6 +7261,7 @@ static int hci_acl_create_conn_sync(struct hci_dev *hdev, void *data) memcpy(conn->dev_class, ie->data.dev_class, 3); } + hci_dev_unlock(hdev); cp.pkt_type = cpu_to_le16(conn->pkt_type); if (lmp_rswitch_capable(hdev) && !(hdev->link_mode & HCI_LM_MASTER)) From 2d696c1ed3468dcb62578a3911e687110d463520 Mon Sep 17 00:00:00 2001 From: Chengfeng Ye Date: Sun, 27 Sep 2026 01:27:23 +0800 Subject: [PATCH 611/694] Bluetooth: RFCOMM: Fix initial port reference race For RFCOMM_RELEASE_ONHUP devices, rfcomm_tty_install() and __rfcomm_release_dev() both use RFCOMM_TTY_OWNED to decide who drops the initial tty_port reference. The release path tests the bit separately from the install path setting it, so both can drop that reference. The synchronous hangup does not prevent this race: port->tty is not assigned until tty_port_open(), after installation. The following interleaving is possible: 1. Install and release each obtain a reference with rfcomm_dev_get(). 2. Release observes RFCOMM_TTY_OWNED clear. 3. Install sets RFCOMM_TTY_OWNED and drops the initial reference. 4. Release drops the same initial reference again. 5. TTY cleanup drops its reference and frees the device. 6. Release performs its final tty_port_put() on the freed port. KASAN reported: BUG: KASAN: slab-use-after-free in tty_port_put+0x22/0x190 Write of size 4 at addr ffff8881001d3d5c by task poc/92 Call Trace: tty_port_put+0x22/0x190 rfcomm_dev_ioctl+0x1d4/0x1930 sock_do_ioctl+0x110/0x260 sock_ioctl+0x380/0x590 __x64_sys_ioctl+0x134/0x1c0 Allocated by task 88: rfcomm_dev_ioctl+0x8f6/0x1930 sock_do_ioctl+0x110/0x260 sock_ioctl+0x380/0x590 __x64_sys_ioctl+0x134/0x1c0 Freed by task 65: kfree+0x131/0x3c0 rfcomm_dev_destruct+0x23b/0x2f0 release_one_tty+0xc1/0x370 process_one_work+0x661/0x1090 Use test_and_set_bit() in both paths so that only the caller that changes RFCOMM_TTY_OWNED from clear to set drops the initial reference. Each path keeps its own lookup reference until release or TTY cleanup, preserving the existing callback ordering and error handling. Fixes: 80ea73378af4 ("Bluetooth: Fix unreleased rfcomm_dev reference") Cc: stable@vger.kernel.org Assisted-by: LLM Signed-off-by: Chengfeng Ye Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/rfcomm/tty.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) diff --git a/net/bluetooth/rfcomm/tty.c b/net/bluetooth/rfcomm/tty.c index b2c1060394e6..dc3cdf614def 100644 --- a/net/bluetooth/rfcomm/tty.c +++ b/net/bluetooth/rfcomm/tty.c @@ -462,7 +462,7 @@ static int __rfcomm_release_dev(void __user *arg) /* Shut down TTY synchronously before freeing rfcomm_dev */ tty_port_tty_vhangup(&dev->port); - if (!test_bit(RFCOMM_TTY_OWNED, &dev->status)) + if (!test_and_set_bit(RFCOMM_TTY_OWNED, &dev->status)) tty_port_put(&dev->port); tty_port_put(&dev->port); @@ -724,10 +724,9 @@ static int rfcomm_tty_install(struct tty_driver *driver, struct tty_struct *tty) * when the last process closes the tty. The behaviour is expected by * userspace. */ - if (test_bit(RFCOMM_RELEASE_ONHUP, &dev->flags)) { - set_bit(RFCOMM_TTY_OWNED, &dev->status); + if (test_bit(RFCOMM_RELEASE_ONHUP, &dev->flags) && + !test_and_set_bit(RFCOMM_TTY_OWNED, &dev->status)) tty_port_put(&dev->port); - } return 0; } From 81a2345f1984f0b36d3226571bc196316a01b648 Mon Sep 17 00:00:00 2001 From: Chengfeng Ye Date: Sun, 27 Sep 2026 14:42:50 +0800 Subject: [PATCH 612/694] Bluetooth: Serialize SMP remote OOB data access build_pairing_cmd() looks up remote OOB data and copies its contents without holding hdev->lock, which serializes the list's writers. After SMP finds an entry, a concurrent management Remove Remote OOB Data command can unlink and free it before SMP reads its present flag or copies its random and confirmation values. Removal can also invalidate an entry while the lookup is still traversing the list. KASAN reported: BUG: KASAN: slab-use-after-free in build_pairing_cmd+0x948/0x9b0 Call Trace: build_pairing_cmd+0x948/0x9b0 smp_recv_cb+0x459f/0x8110 l2cap_recv_frame+0xf14/0x9190 l2cap_recv_acldata+0xa64/0xd40 hci_rx_work+0x4ca/0x730 Allocated by task 87: hci_add_remote_oob_data+0x11d/0x530 add_remote_oob_data+0x282/0x400 hci_sock_sendmsg+0x1033/0x1ea0 Freed by task 93: hci_remote_oob_data_clear+0x108/0x1c0 remove_remote_oob_data+0x198/0x220 hci_sock_sendmsg+0x1033/0x1ea0 Taking hdev->lock in build_pairing_cmd() would recurse for callers that already hold it and invert the device-to-L2CAP lock order on the receive path. Add a per-device remote_oob_lock instead, held across the SMP lookup and copies and by the add, remove and clear helpers. Cover initialization and in-place updates as well, so SMP cannot read partially initialized or updated OOB values. Release the mutex on allocation failure, preserving the existing error return. The new critical sections acquire no device, connection or channel locks. Writers retain their existing hdev->lock protection, which continues to serialize the other readers without changing their locking or behavior. Link: https://lore.kernel.org/r/00660cd3-7d71-13a4-f617-229e6defb701@gmail.com Fixes: 02b05bd8b0a6 ("Bluetooth: Set SMP OOB flag if OOB data is available") Cc: stable@vger.kernel.org Signed-off-by: Chengfeng Ye Signed-off-by: Luiz Augusto von Dentz --- include/net/bluetooth/hci_core.h | 1 + net/bluetooth/hci_core.c | 12 +++++++++++- net/bluetooth/smp.c | 2 ++ 3 files changed, 14 insertions(+), 1 deletion(-) diff --git a/include/net/bluetooth/hci_core.h b/include/net/bluetooth/hci_core.h index ca77ad0d9393..fa2a367731b5 100644 --- a/include/net/bluetooth/hci_core.h +++ b/include/net/bluetooth/hci_core.h @@ -562,6 +562,7 @@ struct hci_dev { struct list_head link_keys; struct list_head long_term_keys; struct list_head identity_resolving_keys; + struct mutex remote_oob_lock; struct list_head remote_oob_data; struct list_head le_accept_list; struct list_head le_resolv_list; diff --git a/net/bluetooth/hci_core.c b/net/bluetooth/hci_core.c index 5d1cdb1e7d8b..bc662e24a8e4 100644 --- a/net/bluetooth/hci_core.c +++ b/net/bluetooth/hci_core.c @@ -1489,8 +1489,10 @@ int hci_remove_remote_oob_data(struct hci_dev *hdev, bdaddr_t *bdaddr, BT_DBG("%s removing %pMR (%u)", hdev->name, bdaddr, bdaddr_type); + mutex_lock(&hdev->remote_oob_lock); list_del(&data->list); kfree(data); + mutex_unlock(&hdev->remote_oob_lock); return 0; } @@ -1499,10 +1501,12 @@ void hci_remote_oob_data_clear(struct hci_dev *hdev) { struct oob_data *data, *n; + mutex_lock(&hdev->remote_oob_lock); list_for_each_entry_safe(data, n, &hdev->remote_oob_data, list) { list_del(&data->list); kfree(data); } + mutex_unlock(&hdev->remote_oob_lock); } int hci_add_remote_oob_data(struct hci_dev *hdev, bdaddr_t *bdaddr, @@ -1511,11 +1515,14 @@ int hci_add_remote_oob_data(struct hci_dev *hdev, bdaddr_t *bdaddr, { struct oob_data *data; + mutex_lock(&hdev->remote_oob_lock); data = hci_find_remote_oob_data(hdev, bdaddr, bdaddr_type); if (!data) { data = kmalloc_obj(*data); - if (!data) + if (!data) { + mutex_unlock(&hdev->remote_oob_lock); return -ENOMEM; + } bacpy(&data->bdaddr, bdaddr); data->bdaddr_type = bdaddr_type; @@ -1548,6 +1555,8 @@ int hci_add_remote_oob_data(struct hci_dev *hdev, bdaddr_t *bdaddr, BT_DBG("%s for %pMR", hdev->name, bdaddr); + mutex_unlock(&hdev->remote_oob_lock); + return 0; } @@ -2485,6 +2494,7 @@ struct hci_dev *hci_alloc_dev_priv(int sizeof_priv) mutex_init(&hdev->lock); mutex_init(&hdev->req_lock); mutex_init(&hdev->mgmt_pending_lock); + mutex_init(&hdev->remote_oob_lock); ida_init(&hdev->unset_handle_ida); diff --git a/net/bluetooth/smp.c b/net/bluetooth/smp.c index d23f9d0729c4..2df303f6a38f 100644 --- a/net/bluetooth/smp.c +++ b/net/bluetooth/smp.c @@ -662,6 +662,7 @@ static void build_pairing_cmd(struct l2cap_conn *conn, else bdaddr_type = BDADDR_LE_RANDOM; + mutex_lock(&hdev->remote_oob_lock); oob_data = hci_find_remote_oob_data(hdev, &hcon->dst, bdaddr_type); if (oob_data && oob_data->present) { @@ -672,6 +673,7 @@ static void build_pairing_cmd(struct l2cap_conn *conn, SMP_DBG("OOB Remote Confirmation: %16phN", smp->pcnf); SMP_DBG("OOB Remote Random: %16phN", smp->rr); } + mutex_unlock(&hdev->remote_oob_lock); } else { authreq &= ~SMP_AUTH_SC; From f4845ab191a4f686a5da481fbba90ed73bb693b0 Mon Sep 17 00:00:00 2001 From: Chengfeng Ye Date: Sun, 27 Sep 2026 01:04:02 +0800 Subject: [PATCH 613/694] Bluetooth: hci_core: Serialize ACL scheduling with channel deletion hci_chan_sent() selects a channel under RCU but drops the read lock before accessing chan->conn and returning the channel. The ACL and LE schedulers then use its packet queue and update its transmit counters without any protection against channel deletion. After TX selects a channel and releases RCU, a disconnect command timeout on the separate request workqueue can run hci_conn_failed() and l2cap_conn_del(). hci_chan_del() can then unlink the channel, complete synchronize_rcu(), purge its queue and free it before TX resumes. This causes use-after-free both in hci_chan_sent() and in its callers. KASAN reported: BUG: KASAN: slab-use-after-free in hci_chan_sent+0x892/0x9b0 Workqueue: hci0 hci_tx_work Call Trace: hci_chan_sent+0x892/0x9b0 hci_tx_work+0x5e6/0xb70 Allocated by task 91: hci_chan_create+0xe3/0x350 l2cap_conn_add.part.0+0x12/0xa30 l2cap_chan_connect+0x110d/0x1b60 l2cap_sock_connect+0x310/0x530 Freed by task 99: hci_chan_del+0x11f/0x170 l2cap_conn_del+0x4f1/0x800 l2cap_connect_cfm+0x88c/0xd30 hci_conn_failed+0x150/0x250 hci_abort_conn_sync+0x3e3/0x800 hci_cmd_sync_run+0x7e/0xc0 hci_abort_conn+0x105/0x1f0 disconnect_sync+0x157/0x290 hci_cmd_sync_work+0x13c/0x290 Hold the existing device mutex across channel selection and transmission in both schedulers to serialize them with channel teardown. Keep timeout handling outside the critical sections because hci_link_tx_to() acquires the same mutex. This also permits the transmit path to sleep, unlike extending the RCU read-side critical section across packet submission. Fixes: 3eff45eaf817 ("Bluetooth: convert tx_task to workqueue") Cc: stable@vger.kernel.org Assisted-by: LLM Signed-off-by: Chengfeng Ye Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/hci_core.c | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/net/bluetooth/hci_core.c b/net/bluetooth/hci_core.c index bc662e24a8e4..a92ab4db4676 100644 --- a/net/bluetooth/hci_core.c +++ b/net/bluetooth/hci_core.c @@ -3675,6 +3675,8 @@ static void hci_sched_acl_pkt(struct hci_dev *hdev) __check_timeout(hdev, cnt, ACL_LINK); + hci_dev_lock(hdev); + while (hdev->acl_cnt && (chan = hci_chan_sent(hdev, ACL_LINK, "e))) { u32 priority = (skb_peek(&chan->data_q))->priority; @@ -3706,6 +3708,8 @@ static void hci_sched_acl_pkt(struct hci_dev *hdev) if (cnt != hdev->acl_cnt) hci_prio_recalculate(hdev, ACL_LINK); + + hci_dev_unlock(hdev); } static void hci_sched_acl(struct hci_dev *hdev) @@ -3734,6 +3738,8 @@ static void hci_sched_le(struct hci_dev *hdev) __check_timeout(hdev, *cnt, LE_LINK); + hci_dev_lock(hdev); + tmp = *cnt; while (*cnt && (chan = hci_chan_sent(hdev, LE_LINK, "e))) { u32 priority = (skb_peek(&chan->data_q))->priority; @@ -3762,6 +3768,8 @@ static void hci_sched_le(struct hci_dev *hdev) if (*cnt != tmp) hci_prio_recalculate(hdev, LE_LINK); + + hci_dev_unlock(hdev); } /* Schedule iso */ From 848a7a91c7f03675d3bd8db6f7721cbf9cb50e21 Mon Sep 17 00:00:00 2001 From: Chengfeng Ye Date: Sun, 27 Sep 2026 01:04:03 +0800 Subject: [PATCH 614/694] Bluetooth: hci_core: Serialize SCO and ISO scheduling with teardown hci_low_sent() selects a connection under RCU but drops the read lock before calculating its quota and returning it to the scheduler. The connection is then used without lifetime protection by hci_sched_iso() and hci_sched_sco(). After the TX worker drops the RCU read lock, hci_abort_conn_sync() on hdev->req_workqueue can remove the connection from the hash, complete synchronize_rcu(), purge its queues and release it. The TX worker on hdev->workqueue can then access the freed connection in hci_quote_sent() or while dequeuing packets and updating conn->sent. KASAN reported: BUG: KASAN: slab-use-after-free in hci_low_sent+0x730/0x840 Workqueue: hci0 hci_tx_work Call Trace: hci_low_sent+0x730/0x840 hci_sched_iso+0x25e/0x4d0 hci_tx_work+0x239/0xcb0 Allocated by task 93: __hci_conn_add+0x16f/0x1b40 hci_bind_bis+0x782/0x17b0 hci_connect_bis+0xa0/0x510 iso_sock_connect+0x589/0x1050 Freed by task 88: kfree+0x131/0x3c0 device_release+0xc8/0x240 kobject_put+0x14d/0x280 hci_conn_del+0x55a/0xe80 hci_disconnect_sync+0x156/0x180 hci_abort_conn_sync+0x3e7/0x940 hci_cmd_sync_work+0x13c/0x290 Hold hci_dev_lock() across connection selection and transmission in both SCO and ISO scheduling, serializing them with connection teardown. This also prevents queuing completion timestamps after the connection queues have been purged. Extending RCU across transmission would be unsafe because the transmit path can sleep. Keep the ISO timeout check outside the mutex since hci_link_tx_to() takes it itself. The preceding channel fix already holds this mutex in the ACL and LE schedulers, which call the SCO scheduler between packets. Move the SCO body to __hci_sched_sco(), assert that its caller holds the mutex, and use it directly from these locked paths. Keep a locking hci_sched_sco() wrapper for the direct calls from hci_tx_work(). This avoids recursively acquiring the device mutex while preserving the scheduling order. Remove the obsolete claim that connection removal disables TX. Fixes: bf4c63252490 ("Bluetooth: convert conn hash to RCU") Cc: stable@vger.kernel.org Assisted-by: LLM Signed-off-by: Chengfeng Ye Signed-off-by: Luiz Augusto von Dentz --- net/bluetooth/hci_core.c | 26 ++++++++++++++++++-------- 1 file changed, 18 insertions(+), 8 deletions(-) diff --git a/net/bluetooth/hci_core.c b/net/bluetooth/hci_core.c index a92ab4db4676..2076689eb302 100644 --- a/net/bluetooth/hci_core.c +++ b/net/bluetooth/hci_core.c @@ -3418,9 +3418,6 @@ static struct hci_conn *hci_low_sent(struct hci_dev *hdev, __u8 type, struct hci_conn *conn = NULL, *c; unsigned int num = 0, min = ~0; - /* We don't have to lock device here. Connections are always - * added and removed with TX task disabled. */ - rcu_read_lock(); list_for_each_entry_rcu(c, &h->list, list) { @@ -3625,13 +3622,15 @@ static void __check_timeout(struct hci_dev *hdev, unsigned int cnt, u8 type) } /* Schedule SCO */ -static void hci_sched_sco(struct hci_dev *hdev, __u8 type) +static void __hci_sched_sco(struct hci_dev *hdev, __u8 type) { struct hci_conn *conn; struct sk_buff *skb; int quote, *cnt; unsigned int pkts = hdev->sco_pkts; + lockdep_assert_held(&hdev->lock); + bt_dev_dbg(hdev, "type %u", type); if (!hci_conn_num(hdev, type) || !pkts) @@ -3666,6 +3665,13 @@ static void hci_sched_sco(struct hci_dev *hdev, __u8 type) queue_work(hdev->workqueue, &hdev->tx_work); } +static void hci_sched_sco(struct hci_dev *hdev, __u8 type) +{ + hci_dev_lock(hdev); + __hci_sched_sco(hdev, type); + hci_dev_unlock(hdev); +} + static void hci_sched_acl_pkt(struct hci_dev *hdev) { unsigned int cnt = hdev->acl_cnt; @@ -3701,8 +3707,8 @@ static void hci_sched_acl_pkt(struct hci_dev *hdev) chan->conn->sent++; /* Send pending SCO packets right away */ - hci_sched_sco(hdev, SCO_LINK); - hci_sched_sco(hdev, ESCO_LINK); + __hci_sched_sco(hdev, SCO_LINK); + __hci_sched_sco(hdev, ESCO_LINK); } } @@ -3761,8 +3767,8 @@ static void hci_sched_le(struct hci_dev *hdev) chan->conn->sent++; /* Send pending SCO packets right away */ - hci_sched_sco(hdev, SCO_LINK); - hci_sched_sco(hdev, ESCO_LINK); + __hci_sched_sco(hdev, SCO_LINK); + __hci_sched_sco(hdev, ESCO_LINK); } } @@ -3788,6 +3794,8 @@ static void hci_sched_iso(struct hci_dev *hdev, __u8 type) __check_timeout(hdev, *cnt, type); + hci_dev_lock(hdev); + while (*cnt && (conn = hci_low_sent(hdev, type, "e))) { while (quote-- && (skb = skb_dequeue(&conn->data_q))) { BT_DBG("skb %p len %d", skb, skb->len); @@ -3801,6 +3809,8 @@ static void hci_sched_iso(struct hci_dev *hdev, __u8 type) (*cnt)--; } } + + hci_dev_unlock(hdev); } static void hci_tx_work(struct work_struct *work) From b630929bcc4c867c9abc8cd5a0ebee3bb5ce5dcd Mon Sep 17 00:00:00 2001 From: Sean Christopherson Date: Sat, 26 Sep 2026 02:04:03 -0400 Subject: [PATCH 615/694] KVM: selftests: Extend nested x2APIC test to validate disabling x2APIC virt Signed-off-by: Sean Christopherson Message-ID: <20260813223610.2043560-6-seanjc@google.com> Signed-off-by: Paolo Bonzini --- .../selftests/kvm/x86/nested_x2apic_test.c | 88 ++++++++++++++----- 1 file changed, 68 insertions(+), 20 deletions(-) diff --git a/tools/testing/selftests/kvm/x86/nested_x2apic_test.c b/tools/testing/selftests/kvm/x86/nested_x2apic_test.c index 3b59ba3e3342..964937aefe39 100644 --- a/tools/testing/selftests/kvm/x86/nested_x2apic_test.c +++ b/tools/testing/selftests/kvm/x86/nested_x2apic_test.c @@ -29,10 +29,12 @@ static void l2_guest_code(void) if (inhibit_apicv) wrmsr(MSR_IA32_APICBASE, rdmsr(MSR_IA32_APICBASE) & GENMASK_ULL(11, 0)); - x2apic_write_reg(APIC_TASKPRI, 0xf0); - GUEST_ASSERT_EQ(x2apic_read_reg(APIC_TASKPRI), 0xf0); + for (;;) { + x2apic_write_reg(APIC_TASKPRI, 0xf0); + GUEST_ASSERT_EQ(x2apic_read_reg(APIC_TASKPRI), 0xf0); - asm volatile("cpuid" ::: "eax", "ebx", "ecx", "edx"); + asm volatile("cpuid" ::: "eax", "ebx", "ecx", "edx"); + } } static void l1_svm_code(struct svm_test_data *svm) @@ -46,6 +48,7 @@ static void l1_svm_code(struct svm_test_data *svm) GUEST_ASSERT_EQ(ctrl->exit_code, SVM_EXIT_CPUID); stgi(); + x2apic_write_reg(APIC_TASKPRI, 0); } static void l1_vmx_code(struct vmx_pages *vmx) @@ -58,22 +61,50 @@ static void l1_vmx_code(struct vmx_pages *vmx) prepare_vmcs(vmx, NULL); GUEST_ASSERT_EQ(vmwrite(GUEST_RIP, (unsigned long)l2_guest_code), 0); + control = vmreadz(PIN_BASED_VM_EXEC_CONTROL); + control |= PIN_BASED_EXT_INTR_MASK; + vmwrite(PIN_BASED_VM_EXEC_CONTROL, control); + control = vmreadz(CPU_BASED_VM_EXEC_CONTROL); - control |= CPU_BASED_USE_MSR_BITMAPS; + control |= CPU_BASED_USE_MSR_BITMAPS | CPU_BASED_TPR_SHADOW; GUEST_ASSERT_EQ(vmwrite(CPU_BASED_VM_EXEC_CONTROL, control), 0); + control = vmreadz(SECONDARY_VM_EXEC_CONTROL); + control |= SECONDARY_EXEC_VIRTUALIZE_X2APIC_MODE | + SECONDARY_EXEC_APIC_REGISTER_VIRT | + SECONDARY_EXEC_VIRTUAL_INTR_DELIVERY; + control &= (rdmsr(MSR_IA32_VMX_PROCBASED_CTLS2) >> 32); + GUEST_ASSERT_EQ(vmwrite(SECONDARY_VM_EXEC_CONTROL, control), 0); + GUEST_ASSERT(!vmlaunch()); GUEST_ASSERT_EQ(vmreadz(VM_EXIT_REASON), EXIT_REASON_CPUID); + GUEST_ASSERT_EQ(vmwrite(GUEST_RIP, + vmreadz(GUEST_RIP) + vmreadz(VM_EXIT_INSTRUCTION_LEN)), 0); } -static void l1_guest_code(void *test_data) +static void l1_vmx_code_part2(void) { - x2apic_enable(); + u64 control; - if (this_cpu_has(X86_FEATURE_SVM)) - l1_svm_code(test_data); - else - l1_vmx_code(test_data); + control = vmreadz(CPU_BASED_VM_EXEC_CONTROL); + control &= ~CPU_BASED_TPR_SHADOW; + GUEST_ASSERT_EQ(vmwrite(CPU_BASED_VM_EXEC_CONTROL, control), 0); + + control = vmreadz(SECONDARY_VM_EXEC_CONTROL); + control &= ~(SECONDARY_EXEC_VIRTUALIZE_X2APIC_MODE | + SECONDARY_EXEC_APIC_REGISTER_VIRT | + SECONDARY_EXEC_VIRTUAL_INTR_DELIVERY); + GUEST_ASSERT_EQ(vmwrite(SECONDARY_VM_EXEC_CONTROL, control), 0); + + GUEST_ASSERT(!vmresume()); + GUEST_ASSERT_EQ(vmreadz(VM_EXIT_REASON), EXIT_REASON_CPUID); + GUEST_ASSERT_EQ(vmwrite(GUEST_RIP, + vmreadz(GUEST_RIP) + vmreadz(VM_EXIT_INSTRUCTION_LEN)), 0); +} + +static void l1_test_x2apic_intercepts(void) +{ + GUEST_ASSERT_EQ(nr_irqs, 0); sti_nop(); @@ -93,16 +124,40 @@ static void l1_guest_code(void *test_data) x2apic_write_reg(APIC_ICR, APIC_DEST_SELF | APIC_INT_ASSERT | POSTED_INTR_NESTED_VECTOR); GUEST_ASSERT_EQ(nr_irqs, 3); + nr_irqs = 0; +} + +static void l1_guest_code(void *test_data) +{ + x2apic_enable(); + + if (this_cpu_has(X86_FEATURE_SVM)) + l1_svm_code(test_data); + else + l1_vmx_code(test_data); + + GUEST_ASSERT_EQ(x2apic_read_reg(APIC_TASKPRI), 0); + x2apic_write_reg(APIC_TASKPRI, 0xf0); + + l1_test_x2apic_intercepts(); + + if (this_cpu_has(X86_FEATURE_VMX)) { + l1_vmx_code_part2(); + l1_test_x2apic_intercepts(); + } + GUEST_DONE(); } -static void __test_x2apic_intercepts(void) +static void __test_x2apic_intercepts(bool with_inhibit_apicv) { gva_t nested_test_data_gva; struct kvm_vcpu *vcpu; struct kvm_vm *vm; struct ucall uc; + inhibit_apicv = with_inhibit_apicv; + vm = vm_create_with_one_vcpu(&vcpu, l1_guest_code); vm_install_exception_handler(vm, POSTED_INTR_VECTOR, guest_irq_handler); vm_install_exception_handler(vm, POSTED_INTR_WAKEUP_VECTOR, guest_irq_handler); @@ -134,17 +189,10 @@ static void __test_x2apic_intercepts(void) kvm_vm_free(vm); } -#define test_x2apic_intercepts(inhibit_apic_setting) \ -do { \ - inhibit_apic_setting; \ - \ - __test_x2apic_intercepts(); \ -} while (0) - int main(int argc, char *argv[]) { TEST_REQUIRE(kvm_cpu_has(X86_FEATURE_SVM) || kvm_cpu_has(X86_FEATURE_VMX)); - test_x2apic_intercepts(inhibit_apicv = true); - test_x2apic_intercepts(inhibit_apicv = false); + test_x2apic_intercepts(true); + test_x2apic_intercepts(false); } From b0bc51910f0391fea09c0a1d32352fec2acb2686 Mon Sep 17 00:00:00 2001 From: Sean Christopherson Date: Thu, 13 Aug 2026 15:36:10 -0700 Subject: [PATCH 616/694] KVM: selftests: Extend nested x2APIC test to validate using eVMCS for vmcs12 Signed-off-by: Sean Christopherson Message-ID: <20260813223610.2043560-7-seanjc@google.com> Signed-off-by: Paolo Bonzini --- .../selftests/kvm/x86/nested_x2apic_test.c | 42 +++++++++++++++---- 1 file changed, 33 insertions(+), 9 deletions(-) diff --git a/tools/testing/selftests/kvm/x86/nested_x2apic_test.c b/tools/testing/selftests/kvm/x86/nested_x2apic_test.c index 964937aefe39..ce204ce29a9c 100644 --- a/tools/testing/selftests/kvm/x86/nested_x2apic_test.c +++ b/tools/testing/selftests/kvm/x86/nested_x2apic_test.c @@ -51,12 +51,24 @@ static void l1_svm_code(struct svm_test_data *svm) x2apic_write_reg(APIC_TASKPRI, 0); } -static void l1_vmx_code(struct vmx_pages *vmx) +static void l1_vmx_code(struct vmx_pages *vmx, struct hyperv_test_pages *hv_pages) { u64 control; + if (hv_pages) { + wrmsr(HV_X64_MSR_GUEST_OS_ID, HYPERV_LINUX_OS_ID); + enable_vp_assist(hv_pages->vp_assist_gpa, hv_pages->vp_assist); + evmcs_enable(); + } + GUEST_ASSERT_EQ(prepare_for_vmx_operation(vmx), true); - GUEST_ASSERT_EQ(load_vmcs(vmx), true); + + if (hv_pages) { + GUEST_ASSERT(load_evmcs(hv_pages)); + current_evmcs->hv_enlightenments_control.msr_bitmap = 1; + } else { + GUEST_ASSERT(load_vmcs(vmx)); + } prepare_vmcs(vmx, NULL); GUEST_ASSERT_EQ(vmwrite(GUEST_RIP, (unsigned long)l2_guest_code), 0); @@ -127,14 +139,14 @@ static void l1_test_x2apic_intercepts(void) nr_irqs = 0; } -static void l1_guest_code(void *test_data) +static void l1_guest_code(void *test_data, void *hv_pages) { x2apic_enable(); if (this_cpu_has(X86_FEATURE_SVM)) l1_svm_code(test_data); else - l1_vmx_code(test_data); + l1_vmx_code(test_data, hv_pages); GUEST_ASSERT_EQ(x2apic_read_reg(APIC_TASKPRI), 0); x2apic_write_reg(APIC_TASKPRI, 0xf0); @@ -149,9 +161,9 @@ static void l1_guest_code(void *test_data) GUEST_DONE(); } -static void __test_x2apic_intercepts(bool with_inhibit_apicv) +static void test_x2apic_intercepts(bool with_inhibit_apicv, bool use_evmcs) { - gva_t nested_test_data_gva; + gva_t nested_test_data_gva, hv_pages_gva = 0; struct kvm_vcpu *vcpu; struct kvm_vm *vm; struct ucall uc; @@ -170,7 +182,14 @@ static void __test_x2apic_intercepts(bool with_inhibit_apicv) else vcpu_alloc_vmx(vm, &nested_test_data_gva); - vcpu_args_set(vcpu, 1, nested_test_data_gva); + if (use_evmcs) { + vcpu_set_hv_cpuid(vcpu); + vcpu_enable_evmcs(vcpu); + + vcpu_alloc_hyperv_test_pages(vm, &hv_pages_gva); + } + + vcpu_args_set(vcpu, 2, nested_test_data_gva, hv_pages_gva); vcpu_run(vcpu); @@ -193,6 +212,11 @@ int main(int argc, char *argv[]) { TEST_REQUIRE(kvm_cpu_has(X86_FEATURE_SVM) || kvm_cpu_has(X86_FEATURE_VMX)); - test_x2apic_intercepts(true); - test_x2apic_intercepts(false); + test_x2apic_intercepts(true, false); + test_x2apic_intercepts(false, false); + + if (kvm_has_cap(KVM_CAP_HYPERV_ENLIGHTENED_VMCS)) { + test_x2apic_intercepts(true, true); + test_x2apic_intercepts(false, true); + } } From 8abbc76120a74bfba1851bd97528099d715e45c6 Mon Sep 17 00:00:00 2001 From: Sean Christopherson Date: Mon, 28 Sep 2026 08:46:43 -0700 Subject: [PATCH 617/694] KVM: SEV: Nullify "have run CPUs" mask pointer when freeing it Nullify have_run_cpus when freeing the mask, particularly in the error path of __sev_guest_init(), so that KVM doesn't have to subtly use sev->active to track whether or not the mask has been freed. As pointed out by Sashiko, blindly freeing the mask in sev_vm_destroy() results in a double-free if the mask is freed if __sev_guest_init() fails. Throw the logic in a helper as nullifying the pointer is frustratingly difficult and weird due to have_run_cpus being a single-entry array when CPUMASK_OFFSTACK=n. Deliberately don't use CPUMASK_VAR_NULL, as it's not directly assignable when the cpumask is on-stack, e.g. requires using a local variable and a memcpy(), which is beyond ridiculous. Furthermore, while clearing the on-stack bitmask is an unnecessary and arguably unwanted side effect, KVM absolutely relies on '0' being the "null" value given that the struct is zero-allocated. Opportunistically add an alloc() helper to pair with free(); there are just enough call sites to make doing so worthwhile. Fixes: 12c1f6e03f94 ("KVM: SEV: Free have_run_cpus during VM destruction even if VM is no longer SEV") Cc: stable@vger.kernel.org Reported-by: Sashiko Bot Closes: https://lore.kernel.org/all/20260923165349.CAAF01F000FF@smtp.kernel.org Signed-off-by: Sean Christopherson Message-ID: <20260928154644.2559454-2-seanjc@google.com> Signed-off-by: Paolo Bonzini --- arch/x86/kvm/svm/sev.c | 33 ++++++++++++++++++++++----------- 1 file changed, 22 insertions(+), 11 deletions(-) diff --git a/arch/x86/kvm/svm/sev.c b/arch/x86/kvm/svm/sev.c index 3448d56520c6..d3a2e6a51efc 100644 --- a/arch/x86/kvm/svm/sev.c +++ b/arch/x86/kvm/svm/sev.c @@ -488,6 +488,20 @@ static void snp_guest_req_cleanup(struct kvm *kvm) sev->guest_resp_buf = NULL; } +static int sev_alloc_have_run_cpus(struct kvm_sev_info *sev) +{ + if (!zalloc_cpumask_var(&sev->have_run_cpus, GFP_KERNEL_ACCOUNT)) + return -ENOMEM; + + return 0; +} + +static void sev_free_have_run_cpus(struct kvm_sev_info *sev) +{ + free_cpumask_var(sev->have_run_cpus); + memset(&sev->have_run_cpus, 0, sizeof(sev->have_run_cpus)); +} + static int __sev_guest_init(struct kvm *kvm, struct kvm_sev_cmd *argp, struct kvm_sev_init *data, unsigned long vm_type) @@ -545,10 +559,9 @@ static int __sev_guest_init(struct kvm *kvm, struct kvm_sev_cmd *argp, if (ret) goto e_free_asid; - if (!zalloc_cpumask_var(&sev->have_run_cpus, GFP_KERNEL_ACCOUNT)) { - ret = -ENOMEM; + ret = sev_alloc_have_run_cpus(sev); + if (ret) goto e_free_asid; - } /* This needs to happen after SEV/SNP firmware initialization. */ if (snp_active) { @@ -566,7 +579,7 @@ static int __sev_guest_init(struct kvm *kvm, struct kvm_sev_cmd *argp, return 0; e_free: - free_cpumask_var(sev->have_run_cpus); + sev_free_have_run_cpus(sev); e_free_asid: argp->error = init_args.error; sev_asid_free(sev); @@ -2191,10 +2204,9 @@ int sev_vm_move_enc_context_from(struct kvm *kvm, unsigned int source_fd) * does not, i.e. KVM could skip flushes if memory is reclaimed from * the old VM but not the new VM. */ - if (!zalloc_cpumask_var(&dst_sev->have_run_cpus, GFP_KERNEL_ACCOUNT)) { - ret = -ENOMEM; + ret = sev_alloc_have_run_cpus(dst_sev); + if (ret) goto out_source_vcpu; - } sev_migrate_from(kvm, source_kvm); kvm_vm_dead(source_kvm); @@ -2888,10 +2900,9 @@ int sev_vm_copy_enc_context_from(struct kvm *kvm, unsigned int source_fd) } mirror_sev = to_kvm_sev_info(kvm); - if (!zalloc_cpumask_var(&mirror_sev->have_run_cpus, GFP_KERNEL_ACCOUNT)) { - ret = -ENOMEM; + ret = sev_alloc_have_run_cpus(mirror_sev); + if (ret) goto e_unlock; - } /* * The mirror kvm holds an enc_context_owner ref so its asid can't @@ -2984,7 +2995,7 @@ void sev_vm_destroy(struct kvm *kvm) * Free the mask even if the VM is not *currently* an SEV VM, as it may * have been an SEV VM prior to intra-host migration. */ - free_cpumask_var(sev->have_run_cpus); + sev_free_have_run_cpus(sev); if (!sev_guest(kvm)) return; From 973ea70393e885e540f714904e51bc6cac80e3d7 Mon Sep 17 00:00:00 2001 From: Sean Christopherson Date: Mon, 28 Sep 2026 08:46:44 -0700 Subject: [PATCH 618/694] KVM: SEV: Do cache maintenance on the source VM *before* clearing SEV state Explicitly flush caches after intra-host migration before clearing "SEV active" on the source VM, as doing cache maintenance afterwards creates a tiny window where memory reclaim could return memory to the host without performing a cache flush, e.g. as pointed out by Sashiko: CPU1 in sev_migrate_from(): src->active = false; CPU2 running concurrent unmap: Since active is false, the automatic cache flush in sev_guest_memory_reclaimed is skipped. The host frees and reallocates the page. CPU1 in sev_migrate_from(): sev_writeback_caches(src_kvm); Executes a hardware cache flush (wbnoinvd), which writes the guest's old dirty ciphertext over the new page owner's data. Fixes: 93de2a6a4b91 ("KVM: SEV: Do cache maintenance on the source VM during intra-host migration") Cc: stable@vger.kernel.org Reported-by: Sashiko Bot Closes: https://lore.kernel.org/all/20260923165304.1662E1F000FF@smtp.kernel.org Signed-off-by: Sean Christopherson Message-ID: <20260928154644.2559454-3-seanjc@google.com> Signed-off-by: Paolo Bonzini --- arch/x86/kvm/svm/sev.c | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) diff --git a/arch/x86/kvm/svm/sev.c b/arch/x86/kvm/svm/sev.c index d3a2e6a51efc..0c1ebb16cec6 100644 --- a/arch/x86/kvm/svm/sev.c +++ b/arch/x86/kvm/svm/sev.c @@ -2045,6 +2045,13 @@ static void sev_migrate_from(struct kvm *dst_kvm, struct kvm *src_kvm) struct kvm_sev_info *mirror; unsigned long i; + /* + * Do cache maintenance on the source VM *before* clearing "SEV active", + * as memory reclaim flows won't trigger cache maintenance on the VM + * once it's no longer an SEV VM. + */ + sev_writeback_caches(src_kvm); + dst->active = true; dst->asid = src->asid; dst->handle = src->handle; @@ -2058,12 +2065,6 @@ static void sev_migrate_from(struct kvm *dst_kvm, struct kvm *src_kvm) src->pages_locked = 0; src->es_active = false; - /* - * Do cache maintenance on the source VM as it is no longer an SEV VM, - * i.e. memory reclaim flows won't trigger cache maintenance on the VM. - */ - sev_writeback_caches(src_kvm); - list_cut_before(&dst->regions_list, &src->regions_list, &src->regions_list); mutex_lock(&sev_mirror_lock); From 6b2462f07cb74cd0a5cb868b36c09e9f51f56d3f Mon Sep 17 00:00:00 2001 From: Christian Melendez Date: Tue, 22 Sep 2026 23:14:39 +0200 Subject: [PATCH 619/694] MAINTAINERS: add missing entry for b53.rst Documentation/networking/dsa/b53.rst is not covered by the BROADCOM B53/SF2 ETHERNET SWITCH DRIVER entry. Signed-off-by: Christian Melendez Link: https://patch.msgid.link/20260922211439.369917-2-chrismelnu@gmail.com Signed-off-by: Jakub Kicinski --- MAINTAINERS | 1 + 1 file changed, 1 insertion(+) diff --git a/MAINTAINERS b/MAINTAINERS index c915b114c99c..7d334b504bc3 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -5143,6 +5143,7 @@ L: netdev@vger.kernel.org L: openwrt-devel@lists.openwrt.org (subscribers-only) S: Supported F: Documentation/devicetree/bindings/net/dsa/brcm,b53.yaml +F: Documentation/networking/dsa/b53.rst F: drivers/net/dsa/b53/* F: drivers/net/dsa/bcm_sf2* F: include/linux/dsa/brcm.h From eb7b84af0ec5c71b4920e999611b466b3784496c Mon Sep 17 00:00:00 2001 From: Florian Fainelli Date: Tue, 22 Sep 2026 16:24:31 -0700 Subject: [PATCH 620/694] net: systemport: Fix buffer overflow in bcm_sysport_get_stats() When running on SYSTEMPORT Lite, certain statistics are unsupported and skipped during bcm_sysport_get_stats(). The variable 'j' tracks the compacted index into the destination data buffer, whereas 'i' iterates over all elements in bcm_sysport_gstrings_stats. Because the buffer allocated by ethtool is sized only according to bcm_sysport_get_sset_count(), storing values at data[i] instead of data[j] writes past the allocated array bounds, leading to memory corruption. Fix this by writing to data[j] instead of data[i]. Fixes: 10377ba7673d ("net: systemport: Support 64bit statistics") Reviewed-by: Nicolai Buchwitz Signed-off-by: Florian Fainelli Link: https://patch.msgid.link/20260922232440.598918-2-florian.fainelli@broadcom.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/broadcom/bcmsysport.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/broadcom/bcmsysport.c b/drivers/net/ethernet/broadcom/bcmsysport.c index 4d06c6ba6641..db627cd15fb7 100644 --- a/drivers/net/ethernet/broadcom/bcmsysport.c +++ b/drivers/net/ethernet/broadcom/bcmsysport.c @@ -482,10 +482,10 @@ static void bcm_sysport_get_stats(struct net_device *dev, s->type == BCM_SYSPORT_STAT_NETDEV64) { do { start = u64_stats_fetch_begin(syncp); - data[i] = *(u64 *)p; + data[j] = *(u64 *)p; } while (u64_stats_fetch_retry(syncp, start)); } else - data[i] = *(u32 *)p; + data[j] = *(u32 *)p; j++; } From 242935adfeb096f09b29fb2cfcf12f9eb2911fc6 Mon Sep 17 00:00:00 2001 From: Florian Fainelli Date: Tue, 22 Sep 2026 16:24:35 -0700 Subject: [PATCH 621/694] net: systemport: Fix RUNT MIB counter register offset calculation In UniMAC hardware, there is a 0xC byte gap between the RX MIB counters and the TX MIB counters, and a second 0xC byte gap between the TX MIB counters and the RX RUNT MIB counters. In bcm_sysport_update_mib_counters(), 'offset' was only set to UMAC_MIB_STAT_OFFSET (0xC) for all non-RX counters, omitting the second 0xC gap for BCM_SYSPORT_STAT_RUNT counters. As a result, all 4 RUNT MIB counters were read from unmapped gap register space. Fix this by setting offset to 2 * UMAC_MIB_STAT_OFFSET (0x18) when reading BCM_SYSPORT_STAT_RUNT counters. Fixes: 80105befdb4b ("net: systemport: add Broadcom SYSTEMPORT Ethernet MAC driver") Reviewed-by: Nicolai Buchwitz Signed-off-by: Florian Fainelli Link: https://patch.msgid.link/20260922232440.598918-6-florian.fainelli@broadcom.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/broadcom/bcmsysport.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/broadcom/bcmsysport.c b/drivers/net/ethernet/broadcom/bcmsysport.c index db627cd15fb7..78909f516823 100644 --- a/drivers/net/ethernet/broadcom/bcmsysport.c +++ b/drivers/net/ethernet/broadcom/bcmsysport.c @@ -391,8 +391,10 @@ static void bcm_sysport_update_mib_counters(struct bcm_sysport_priv *priv) if (priv->is_lite) continue; - if (s->type != BCM_SYSPORT_STAT_MIB_RX) + if (s->type == BCM_SYSPORT_STAT_MIB_TX) offset = UMAC_MIB_STAT_OFFSET; + else if (s->type == BCM_SYSPORT_STAT_RUNT) + offset = 2 * UMAC_MIB_STAT_OFFSET; val = umac_readl(priv, UMAC_MIB_START + j + offset); break; case BCM_SYSPORT_STAT_RXCHK: From 04c824940d52871c49866a4ac67b504146f7cf35 Mon Sep 17 00:00:00 2001 From: Anirudh Virdi Date: Thu, 24 Sep 2026 12:12:44 +0530 Subject: [PATCH 622/694] net/mlx5: Fix slab-out-of-bounds when handling team device events The mlx5_handle_changeupper_event() and mlx5_handle_changeinfodata_event() functions check for LAG masters (which includes both bonding and team devices) but only call bonding-specific APIs. When processing team device events, calling bond_slave_get_rcu() and bond_is_slave_inactive() on team_port structures causes KASAN to detect an out-of-bounds memory access since team_port is smaller than bond slave. Fix this by wrapping the bond-specific API calls with a check for bonding devices. This allows the function to still process LAG events for both bonding and teams but only calls bond-specific functions when dealing with actual bonding devices. For mlx5_handle_changeinfodata_event(), keep the explicit bond check as mlx5 doesn't handle state information for team ports anyway. Tested with Mellanox ConnectX-5 on Linux 7.2.0-rc6: - Bonding: PASS (no regressions) - Team device: PASS (no KASAN errors) Fixes: 54493a08e21f ("net/mlx5: Lag, record inactive state of bond device") Suggested-by: Mark Bloch Signed-off-by: Anirudh Virdi Reviewed-by: Mark Bloch Link: https://patch.msgid.link/20260924064244.94045-1-avirdi@redhat.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/mellanox/mlx5/core/lag/lag.c | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lag/lag.c b/drivers/net/ethernet/mellanox/mlx5/core/lag/lag.c index dd14cdc378de..3b34bec559e0 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/lag/lag.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/lag/lag.c @@ -1943,9 +1943,11 @@ static int mlx5_handle_changeupper_event(struct mlx5_lag *ldev, } } if (i < MLX5_MAX_PORTS) { - slave = bond_slave_get_rcu(ndev_tmp); - if (slave) - has_inactive |= bond_is_slave_inactive(slave); + if (netif_is_bond_master(upper)) { + slave = bond_slave_get_rcu(ndev_tmp); + if (slave) + has_inactive |= bond_is_slave_inactive(slave); + } bond_status |= (1 << idx); } @@ -2029,7 +2031,7 @@ static int mlx5_handle_changeinfodata_event(struct mlx5_lag *ldev, bool has_inactive = 0; int idx; - if (!netif_is_lag_master(ndev)) + if (!netif_is_bond_master(ndev)) return 0; rcu_read_lock(); From 056994563935e049093f100efaad7e63ad113814 Mon Sep 17 00:00:00 2001 From: Geetha sowjanya Date: Tue, 22 Sep 2026 15:51:48 +0530 Subject: [PATCH 623/694] octeontx2-pf: Fix aura BPID assignment when CONFIG_DCB is enabled Previously, BPID assignment under CONFIG_DCB assumed `queue_to_pfc_map` was always initialized. For SDP VFs this leads to invalid memory access as it was not initialized. This patch adds a NULL check for `queue_to_pfc_map` before dereferencing it. Also, simplifies the logic by always assigning a default BPID first, then conditionally overriding it if CONFIG_DCB is enabled and the map exists. Fixes: 184fb40f731b ("octeontx2-pf: Avoid adding dcbnl_ops for LBK and SDP vf") Signed-off-by: Geetha sowjanya Signed-off-by: Subrat Pandey Reviewed-by: Simon Horman Link: https://patch.msgid.link/20260922102149.2078168-2-subratp@marvell.com Signed-off-by: Jakub Kicinski --- .../net/ethernet/marvell/octeontx2/nic/cn20k.c | 11 +---------- .../ethernet/marvell/octeontx2/nic/otx2_common.c | 16 ++++++---------- .../ethernet/marvell/octeontx2/nic/otx2_common.h | 9 +++++++++ 3 files changed, 16 insertions(+), 20 deletions(-) diff --git a/drivers/net/ethernet/marvell/octeontx2/nic/cn20k.c b/drivers/net/ethernet/marvell/octeontx2/nic/cn20k.c index 8e41431c7f9c..e92b77789f08 100644 --- a/drivers/net/ethernet/marvell/octeontx2/nic/cn20k.c +++ b/drivers/net/ethernet/marvell/octeontx2/nic/cn20k.c @@ -242,15 +242,6 @@ int cn20k_register_pfvf_mbox_intr(struct otx2_nic *pf, int numvfs) #define RQ_BP_LVL_AURA (255 - ((85 * 256) / 100)) /* BP when 85% is full */ -static u8 cn20k_aura_bpid_idx(struct otx2_nic *pfvf, int aura_id) -{ -#ifdef CONFIG_DCB - return pfvf->queue_to_pfc_map[aura_id]; -#else - return 0; -#endif -} - static int cn20k_tc_get_entry_index(struct otx2_flow_config *flow_cfg, struct otx2_tc_flow *node) { @@ -580,7 +571,7 @@ static int cn20k_aura_aq_init(struct otx2_nic *pfvf, int aura_id, if (pfvf->nix_blkaddr == BLKADDR_NIX1) aq->aura.bp_ena = 1; - bpid_idx = cn20k_aura_bpid_idx(pfvf, aura_id); + bpid_idx = otx2_get_bpid_idx(pfvf, aura_id); aq->aura.bpid = pfvf->bpid[bpid_idx]; /* Set backpressure level for RQ's Aura */ diff --git a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_common.c b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_common.c index 175992188c18..0f3f405d9f1b 100644 --- a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_common.c +++ b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_common.c @@ -1055,6 +1055,7 @@ int otx2_cq_init(struct otx2_nic *pfvf, u16 qidx) struct nix_aq_enq_req *aq; struct otx2_cq_queue *cq; struct otx2_pool *pool; + u8 bpid_idx; cq = &qset->cq[qidx]; cq->cq_idx = qidx; @@ -1132,11 +1133,8 @@ int otx2_cq_init(struct otx2_nic *pfvf, u16 qidx) if (!is_otx2_lbkvf(pfvf->pdev)) { /* Enable receive CQ backpressure */ aq->cq.bp_ena = 1; -#ifdef CONFIG_DCB - aq->cq.bpid = pfvf->bpid[pfvf->queue_to_pfc_map[qidx]]; -#else - aq->cq.bpid = pfvf->bpid[0]; -#endif + bpid_idx = otx2_get_bpid_idx(pfvf, qidx); + aq->cq.bpid = pfvf->bpid[bpid_idx]; /* Set backpressure level is same as cq pass level */ aq->cq.bp = RQ_PASS_LVL_CQ(pfvf->hw.rq_skid, qset->rqe_cnt); @@ -1378,6 +1376,7 @@ int otx2_aura_aq_init(struct otx2_nic *pfvf, int aura_id, { struct npa_aq_enq_req *aq; struct otx2_pool *pool; + u8 bpid_idx; int err; pool = &pfvf->qset.pool[pool_id]; @@ -1433,11 +1432,8 @@ int otx2_aura_aq_init(struct otx2_nic *pfvf, int aura_id, */ if (pfvf->nix_blkaddr == BLKADDR_NIX1) aq->aura.bp_ena = 1; -#ifdef CONFIG_DCB - aq->aura.nix0_bpid = pfvf->bpid[pfvf->queue_to_pfc_map[aura_id]]; -#else - aq->aura.nix0_bpid = pfvf->bpid[0]; -#endif + bpid_idx = otx2_get_bpid_idx(pfvf, aura_id); + aq->aura.nix0_bpid = pfvf->bpid[bpid_idx]; /* Set backpressure level for RQ's Aura */ aq->aura.bp = RQ_BP_LVL_AURA; diff --git a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_common.h b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_common.h index eecee612b7b2..44cd56a0c2c4 100644 --- a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_common.h +++ b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_common.h @@ -1029,6 +1029,15 @@ static inline int otx2_tc_flower_rule_cnt(struct otx2_nic *pfvf) return pfvf->flow_cfg->nr_flows; } +static inline u8 otx2_get_bpid_idx(struct otx2_nic *pfvf, int qidx) +{ +#ifdef CONFIG_DCB + if (pfvf->queue_to_pfc_map) + return pfvf->queue_to_pfc_map[qidx]; +#endif + return 0; +} + /* MSI-X APIs */ void otx2_free_cints(struct otx2_nic *pfvf, int n); void otx2_set_cints_affinity(struct otx2_nic *pfvf); From cdef07f3002f708a345392ee28d0d9bfac8383a1 Mon Sep 17 00:00:00 2001 From: Baha Mesleh Date: Tue, 22 Sep 2026 15:51:49 +0530 Subject: [PATCH 624/694] octeontx2-af: mcs: Fix SC resource cleanup loop The SC resource cleanup and statistics loops were incorrectly iterating over secy.max instead of sc.max, so the highest SC id could be left allocated across teardown or FLR. Use sc.max as the loop bound. Fixes: cfc14181d497 ("octeontx2-af: cn10k: mcs: Manage the MCS block hardware resources") Signed-off-by: Baha Mesleh Signed-off-by: Subrat Pandey Reviewed-by: Simon Horman Link: https://patch.msgid.link/20260922102149.2078168-3-subratp@marvell.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/marvell/octeontx2/af/mcs.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/marvell/octeontx2/af/mcs.c b/drivers/net/ethernet/marvell/octeontx2/af/mcs.c index 211c10aa5880..30c672d526a0 100644 --- a/drivers/net/ethernet/marvell/octeontx2/af/mcs.c +++ b/drivers/net/ethernet/marvell/octeontx2/af/mcs.c @@ -314,7 +314,7 @@ int mcs_clear_all_stats(struct mcs *mcs, u16 pcifunc, int dir) } /* Clear SC stats */ - for (id = 0; id < map->secy.max; id++) { + for (id = 0; id < map->sc.max; id++) { if (map->sc2pf_map[id] != pcifunc) continue; mcs_clear_stats(mcs, MCS_SC_STATS, id, dir); @@ -738,7 +738,7 @@ int mcs_free_all_rsrc(struct mcs *mcs, int dir, u16 pcifunc) } /* free sc entries */ - for (id = 0; id < map->secy.max; id++) { + for (id = 0; id < map->sc.max; id++) { if (map->sc2pf_map[id] != pcifunc) continue; mcs_free_rsrc(&map->sc, map->sc2pf_map, id, pcifunc); From 512ccd3d0e91e791fb37442aa0a2599aca19783d Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?J=C3=A9r=C3=A9my=20Jean?= Date: Thu, 24 Sep 2026 20:21:05 +0000 Subject: [PATCH 625/694] tipc: prevent GCM nonce reuse on peer key changes MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit TIPC can encrypt traffic between nodes using a different transmit key for each node. In this mode, the AES-GCM nonce for a packet sent to a known peer consists of a 32-bit prefix (a per-key salt XOR the peer's address) followed by a 64-bit counter. That counter is stored in the peer's RX crypto object. When the peer reports a change in which key it uses to receive packets, TIPC resets this counter. The sender can still be using the same TX key and salt, so subsequent packets reuse earlier nonces. This nonce reuse breaks confidentiality and exposes GCM's authentication key. This makes forgeries trivial: an attacker can exploit CTR malleability to alter captured ciphertexts and use the recovered authentication key to compute a valid tag for the modified ciphertext, under the same key and nonce. Use the TX key's existing aead->seqno counter instead. All encryptions using that key object share the same atomic counter, so concurrent encryptions get distinct nonce counter values. The counter survives key activation and peer reconnection, and peer key-status reports cannot reset it. This prevents those transitions from causing nonce reuse while the same TX key remains installed. The nonce format is unchanged, and receivers do not require consecutive counter values, so sharing the counter across peers remains compatible with existing receivers. A pre-existing check still invokes key revocation in the unlikely event that the counter wraps to zero. Fixes: fc1b6d6de220 ("tipc: introduce TIPC encryption & authentication") Signed-off-by: Jérémy Jean Reviewed-by: Tung Nguyen Link: https://patch.msgid.link/20260924202105.3722778-1-Jeremy.Jean@oss.cyber.gouv.fr Signed-off-by: Jakub Kicinski --- net/tipc/crypto.c | 20 +++++--------------- 1 file changed, 5 insertions(+), 15 deletions(-) diff --git a/net/tipc/crypto.c b/net/tipc/crypto.c index 16f1ed1f6b1b..4409bdb70fe9 100644 --- a/net/tipc/crypto.c +++ b/net/tipc/crypto.c @@ -144,7 +144,7 @@ struct tipc_tfm { * @rcu: struct rcu_head * @key: the aead key * @gen: the key's generation - * @seqno: the key seqno (cluster scope) + * @seqno: the per-key TX nonce counter * @refcnt: the key reference counter */ struct tipc_aead { @@ -190,7 +190,6 @@ struct tipc_crypto_stats { * @rekeying_intv: rekeying interval (in minutes) * @stats: the crypto statistics * @name: the crypto name - * @sndnxt: the per-peer sndnxt (TX) * @timer1: general timer 1 (jiffies) * @timer2: general timer 2 (jiffies) * @working: the crypto is working or not @@ -219,7 +218,6 @@ struct tipc_crypto { struct tipc_crypto_stats __percpu *stats; char name[48]; - atomic64_t sndnxt ____cacheline_aligned; unsigned long timer1; unsigned long timer2; union { @@ -1051,14 +1049,11 @@ static int tipc_ehdr_build(struct net *net, struct tipc_aead *aead, WARN_ON(skb_headroom(skb) < ehsz); ehdr = (struct tipc_ehdr *)skb_push(skb, ehsz); - /* Obtain a seqno first: - * Use the key seqno (= cluster wise) if dest is unknown or we're in - * cluster key mode, otherwise it's better for a per-peer seqno! + /* + * Keep the nonce unique for the lifetime of the TX key, + * including key state changes and peer reconnection. */ - if (!__rx || aead->mode == CLUSTER_KEY) - seqno = atomic64_inc_return(&aead->seqno); - else - seqno = atomic64_inc_return(&__rx->sndnxt); + seqno = atomic64_inc_return(&aead->seqno); /* Revoke the key if seqno is wrapped around */ if (unlikely(!seqno)) @@ -1237,7 +1232,6 @@ void tipc_crypto_key_flush(struct tipc_crypto *c) tipc_crypto_key_set_state(c, 0, 0, 0); for (k = KEY_MIN; k <= KEY_MAX; k++) tipc_crypto_key_detach(c->aead[k], &c->lock); - atomic64_set(&c->sndnxt, 0); spin_unlock_bh(&c->lock); } @@ -1384,8 +1378,6 @@ static struct tipc_aead *tipc_crypto_key_pick_tx(struct tipc_crypto *tx, * It also considers if peer has no key, then we need to make own master key * (if any) taking over i.e. starting grace period and also trigger key * distributing process. - * - * The "per-peer" sndnxt is also reset when the peer key has switched. */ static void tipc_crypto_key_synch(struct tipc_crypto *rx, struct sk_buff *skb) { @@ -1436,7 +1428,6 @@ static void tipc_crypto_key_synch(struct tipc_crypto *rx, struct sk_buff *skb) if (cur) tipc_aead_users_dec(tx->aead[cur], 0); - atomic64_set(&rx->sndnxt, 0); /* Mark the point TX key users changed */ tx->timer1 = jiffies; @@ -1501,7 +1492,6 @@ int tipc_crypto_start(struct tipc_crypto **crypto, struct net *net, tipc_crypto_key_set_state(c, 0, 0, 0); atomic_set(&c->key_distr, 0); atomic_set(&c->peer_rx_active, 0); - atomic64_set(&c->sndnxt, 0); c->timer1 = jiffies; c->timer2 = jiffies; c->rekeying_intv = TIPC_REKEYING_INTV_DEF; From bdea4980182e3badac74f04df6c07492b91c569e Mon Sep 17 00:00:00 2001 From: Mina Almasry Date: Fri, 25 Sep 2026 14:41:11 +0000 Subject: [PATCH 626/694] net: page_pool: fix use-after-free in page_pool_recycle_ring_bulk() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit With CONFIG_PAGE_POOL_STATS=y, page_pool_recycle_ring_bulk() updates the recycle stats after dropping the producer lock. If it just recycled the last inflight netmems of a pool being destroyed, page_pool_release() can pass its producer-lock barrier and free the pool before recycle_stat_add() runs. Commit fcc680a647ba7 ("page_pool: allow mixing PPs within one bulk") moved this stat update after the unlock. Commit 271683bb2cf32 ("page_pool: Fix use-after-free in page_pool_recycle_in_ring") later added the barrier, but only fixed page_pool_recycle_in_ring(). Move the update back under the lock. Fixes: fcc680a647ba7 ("page_pool: allow mixing PPs within one bulk") Link: https://lore.kernel.org/r/179028939171.2160803.706522228583664641@kernel.org Cc: Kaifeng Wang Cc: Dong Chenchen Signed-off-by: Mina Almasry Reviewed-by: Simon Horman Reviewed-by: Toke Høiland-Jørgensen Link: https://patch.msgid.link/20260925144127.1445667-1-almasrymina@google.com Signed-off-by: Jakub Kicinski --- net/core/page_pool.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/core/page_pool.c b/net/core/page_pool.c index 08d7f35cf608..d7c88c0b67e5 100644 --- a/net/core/page_pool.c +++ b/net/core/page_pool.c @@ -951,8 +951,8 @@ static void page_pool_recycle_ring_bulk(struct page_pool *pool, } } - page_pool_producer_unlock(pool, in_softirq); recycle_stat_add(pool, ring, i); + page_pool_producer_unlock(pool, in_softirq); /* Hopefully all pages were returned into ptr_ring */ if (likely(i == bulk_len)) From 88095728511e0df9c89528944f5a1762ad298983 Mon Sep 17 00:00:00 2001 From: Lorenzo Bianconi Date: Wed, 23 Sep 2026 11:14:28 +0200 Subject: [PATCH 627/694] net: stmmac: fix rx Scatter-Gather support When a received frame is larger than dma_buf_sz, the DMA scatters it across multiple RX descriptors (rx Scatter-Gather). The secondary RX buffer (sec_page) was only allocated and programmed when split-header (SPH) was active, so for regular frames buffer2 was neither allocated nor backed by a valid mapping. As soon as an incoming frame overflowed buffer1, the DMA wrote the overflow into the unmapped secondary-buffer address, triggering an SMMU translation fault on IOMMU-based platforms: arm-smmu 15000000.iommu: Unhandled context fault: fsr=0x402, iova=0x00000000, fsynr=0x7f0011, cbfrsynra=0x1c90, cb=11 arm-smmu 15000000.iommu: FSR = 00000402 [Format=2 TF], SID=0x1c90 arm-smmu 15000000.iommu: FSYNR0 = 007f0011 [S1CBNDX=127 WNR PLVL=1] Enable scatter-gather for non-SPH frames on cores that can program an independent secondary RX buffer (GMAC4/XGMAC): allocate and mark buffer2 as valid in stmmac_init_rx_buffers() and stmmac_rx_refill(), and account for it in the buffer length computation. Legacy cores have no set_sec_addr op, so they keep buffer2 disabled. Since buffer2 is handed to the DMA at page offset 0, the page pool sync window is widened to cover both buffers in every mode: offset is set to 0 and max_len to dma_buf_sz + stmmac_rx_offset(). The FCS can straddle the buffer1/buffer2 boundary and a descriptor boundary, so it is now stripped from the tail of the assembled frame with pskb_trim() instead of from a single buffer, avoiding an unsigned underflow for frames that overflow a buffer by 1..3 bytes. For single-buffer frames the XDP program must not see the FCS, so it is removed from the XDP buffer before the program runs. Native XDP currently only supports single-buffer (linear) frames. An oversized frame accepted by the MAC (jumbo enabled) is received via buffer2; the XDP program only sees buffer1, so on a TX/REDIRECT verdict the frame is forwarded truncated. XDP multi-buffer support to handle this case is planned as a follow-up. AF_XDP zero-copy RX is not covered by this change: a ZC queue still programs buffer2 at DMA address 0 (and XGMAC has no buffer2-valid bit), so an oversized frame overflowing buffer1 can still trigger the same SMMU translation fault. Handling is planned as a follow-up. Fixes: 88ebe2cf7f3f ("net: stmmac: Rework stmmac_rx()") Signed-off-by: Lorenzo Bianconi Link: https://patch.msgid.link/20260923-stmmac-rx-sg-fix-v3-1-ed26fea7180d@oss.qualcomm.com Signed-off-by: Jakub Kicinski --- .../net/ethernet/stmicro/stmmac/stmmac_main.c | 83 +++++++++++-------- 1 file changed, 49 insertions(+), 34 deletions(-) diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c index 3f34d491c959..ec62fa7418f4 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c @@ -1629,6 +1629,16 @@ static void stmmac_clear_descriptors(struct stmmac_priv *priv, stmmac_clear_tx_descriptors(priv, dma_conf, queue); } +static bool stmmac_rx_check_buf2_cap(struct stmmac_priv *priv) +{ + /* Only cores that can program an independent secondary RX buffer + * (used for scatter-gather overflow or split-header payload) back + * buffer2. Legacy cores have no set_sec_addr op, so buffer2 is + * never handed to the hardware there. + */ + return priv->hw->desc && priv->hw->desc->set_sec_addr; +} + /** * stmmac_init_rx_buffers - init the RX descriptor buffer. * @priv: driver private structure @@ -1659,16 +1669,13 @@ static int stmmac_init_rx_buffers(struct stmmac_priv *priv, buf->page_offset = stmmac_rx_offset(priv); } - if (priv->sph_active && !buf->sec_page) { + if (stmmac_rx_check_buf2_cap(priv) && !buf->sec_page) { buf->sec_page = page_pool_alloc_pages(rx_q->page_pool, gfp); if (!buf->sec_page) return -ENOMEM; buf->sec_addr = page_pool_get_dma_addr(buf->sec_page); stmmac_set_desc_sec_addr(priv, p, buf->sec_addr, true); - } else { - buf->sec_page = NULL; - stmmac_set_desc_sec_addr(priv, p, buf->sec_addr, false); } buf->addr = page_pool_get_dma_addr(buf->page) + buf->page_offset; @@ -2256,13 +2263,8 @@ static int __alloc_dma_rx_desc_resources(struct stmmac_priv *priv, pp_params.nid = dev_to_node(priv->device); pp_params.dev = priv->device; pp_params.dma_dir = xdp_prog ? DMA_BIDIRECTIONAL : DMA_FROM_DEVICE; - pp_params.offset = stmmac_rx_offset(priv); - pp_params.max_len = dma_conf->dma_buf_sz; - - if (priv->sph_active) { - pp_params.offset = 0; - pp_params.max_len += stmmac_rx_offset(priv); - } + pp_params.offset = 0; + pp_params.max_len = dma_conf->dma_buf_sz + stmmac_rx_offset(priv); rx_q->page_pool = page_pool_create(&pp_params); if (IS_ERR(rx_q->page_pool)) { @@ -5099,7 +5101,7 @@ static inline void stmmac_rx_refill(struct stmmac_priv *priv, u32 queue) break; } - if (priv->sph_active && !buf->sec_page) { + if (stmmac_rx_check_buf2_cap(priv) && !buf->sec_page) { buf->sec_page = page_pool_alloc_pages(rx_q->page_pool, gfp); if (!buf->sec_page) break; @@ -5110,10 +5112,8 @@ static inline void stmmac_rx_refill(struct stmmac_priv *priv, u32 queue) buf->addr = page_pool_get_dma_addr(buf->page) + buf->page_offset; stmmac_set_desc_addr(priv, p, buf->addr); - if (priv->sph_active) - stmmac_set_desc_sec_addr(priv, p, buf->sec_addr, true); - else - stmmac_set_desc_sec_addr(priv, p, buf->sec_addr, false); + stmmac_set_desc_sec_addr(priv, p, buf->sec_addr, + stmmac_rx_check_buf2_cap(priv)); stmmac_refill_desc3(priv, rx_q, p); rx_q->rx_count_frames++; @@ -5144,7 +5144,9 @@ static unsigned int stmmac_rx_buf1_len(struct stmmac_priv *priv, unsigned int plen = 0, hlen = 0; int coe = priv->hw->rx_csum; - /* Not first descriptor, buffer is always zero */ + /* Not first descriptor, SPH enabled: buffer1 only carries the + * split header of the first descriptor, so it is zero here. + */ if (priv->sph_active && len) return 0; @@ -5155,14 +5157,16 @@ static unsigned int stmmac_rx_buf1_len(struct stmmac_priv *priv, return hlen; } - /* First descriptor, not last descriptor and not split header */ + /* Not last descriptor and not split header: buffer1 is fully filled */ if (status & rx_not_ls) return priv->dma_conf.dma_buf_sz; plen = stmmac_get_rx_frame_len(priv, p, coe); - /* First descriptor and last descriptor and not split header */ - return min_t(unsigned int, priv->dma_conf.dma_buf_sz, plen); + /* Last descriptor and not split header: buffer1 holds the remaining + * bytes of the frame, up to dma_buf_sz + */ + return min_t(unsigned int, priv->dma_conf.dma_buf_sz, plen - len); } static unsigned int stmmac_rx_buf2_len(struct stmmac_priv *priv, @@ -5172,8 +5176,7 @@ static unsigned int stmmac_rx_buf2_len(struct stmmac_priv *priv, int coe = priv->hw->rx_csum; unsigned int plen = 0; - /* Not split header, buffer is not available */ - if (!priv->sph_active) + if (!stmmac_rx_check_buf2_cap(priv)) return 0; /* For GMAC4, when split header is enabled, in some rare cases, the @@ -5190,14 +5193,15 @@ static unsigned int stmmac_rx_buf2_len(struct stmmac_priv *priv, * Thus 'plen - len' always gives the correct length of buf2. */ - /* Not GMAC4 and not last descriptor */ - if (priv->plat->core_type != DWMAC_CORE_GMAC4 && (status & rx_not_ls)) + /* Not GMAC4, or non-SPH and not last descriptor */ + if ((priv->plat->core_type != DWMAC_CORE_GMAC4 || !priv->sph_active) && + (status & rx_not_ls)) return priv->dma_conf.dma_buf_sz; /* GMAC4 or last descriptor */ plen = stmmac_get_rx_frame_len(priv, p, coe); - return plen - len; + return plen > len ? plen - len : 0; } static int stmmac_xdp_xmit_xdpf(struct stmmac_priv *priv, int queue, @@ -5720,6 +5724,7 @@ static int stmmac_rx(struct stmmac_priv *priv, int limit, u32 queue) unsigned int desc_size; struct sk_buff *skb = NULL; struct stmmac_xdp_buff ctx; + bool fcs_stripped = false; int xdp_status = 0; int bufsz; @@ -5809,24 +5814,25 @@ static int stmmac_rx(struct stmmac_priv *priv, int limit, u32 queue) len += buf2_len; /* ACS is disabled; strip manually. */ - if (likely(!(status & rx_not_ls))) { - if (buf2_len) { - buf2_len -= ETH_FCS_LEN; - len -= ETH_FCS_LEN; - } else if (buf1_len) { - buf1_len -= ETH_FCS_LEN; - len -= ETH_FCS_LEN; - } - } + if (likely(!(status & rx_not_ls))) + len -= ETH_FCS_LEN; if (!skb) { unsigned int pre_len, sync_len; + /* Each frame starts here: reset the FCS handling */ + fcs_stripped = false; + dma_sync_single_for_cpu(priv->device, buf->addr, buf1_len, dma_dir); net_prefetch(page_address(buf->page) + buf->page_offset); + if (stmmac_xdp_is_enabled(priv) && !buf2_len) { + buf1_len -= ETH_FCS_LEN; + fcs_stripped = true; + } + xdp_init_buff(&ctx.xdp, bufsz, &rx_q->xdp_rxq); xdp_prepare_buff(&ctx.xdp, page_address(buf->page), buf->page_offset, buf1_len, true); @@ -5927,6 +5933,15 @@ static int stmmac_rx(struct stmmac_priv *priv, int limit, u32 queue) /* Got entire packet into SKB. Finish it. */ + /* Remove FCS if needed */ + if (!fcs_stripped && pskb_trim(skb, len)) { + dev_kfree_skb_any(skb); + skb = NULL; + rx_dropped++; + count++; + continue; + } + stmmac_get_rx_hwtstamp(priv, p, np, skb); if (priv->hw->hw_vlan_en) From 8ae3f4067ba4ae3a15581392f6dac422e3d3def0 Mon Sep 17 00:00:00 2001 From: Andrea Mayer Date: Fri, 25 Sep 2026 15:38:07 +0200 Subject: [PATCH 628/694] seg6: ensure packet data is writable before modifying SRH and IPv6 DA advance_nextseg() modifies the SRH Segments Left field and the IPv6 destination address without ensuring the packet data is writable. seg6_next_csid_advance_arg() has the same problem when it advances the NEXT-C-SID argument in the destination address. The skb may be cloned, for example by an AF_PACKET socket receiving on the ingress device. advance_nextseg() and seg6_next_csid_advance_arg() then write into the packet data shared with the clone. A read from that socket can return the modified packet instead of the received one. The simplified path below shows this for advance_nextseg(): __netif_receive_skb_one_core __netif_receive_skb_core deliver_skb [orig: users=2, cloned=0] packet_rcv skb_clone clone queued to the AF_PACKET socket consume_skb(orig) [orig: users=1, cloned=1] ipv6_rcv ip6_rcv_core skb_share_check: no-op [orig: users=1] [...] input_action_end_core advance_nextseg writes into the data shared with the clone Call skb_ensure_writable() in advance_nextseg() and in seg6_next_csid_advance_arg() before they modify the packet data. skb_ensure_writable() may reallocate skb->head, which invalidates the pointers into the packet data taken before the call. advance_nextseg() now returns a valid SRH pointer, or NULL if skb_ensure_writable() fails. seg6_next_csid_advance_arg() takes the pointer to the destination address from the skb after skb_ensure_writable(). On failure, the callers drop the packet with SKB_DROP_REASON_NOMEM. Fixes: 140f04c33bbc ("ipv6: sr: implement several seg6local actions") Fixes: 848f3c0d4769 ("seg6: add NEXT-C-SID support for SRv6 End behavior") Signed-off-by: Andrea Mayer Link: https://patch.msgid.link/20260925133807.32-1-andrea.mayer@uniroma2.it Signed-off-by: Jakub Kicinski --- net/ipv6/seg6_local.c | 140 +++++++++++++++++++++++++++++++++++------- 1 file changed, 117 insertions(+), 23 deletions(-) diff --git a/net/ipv6/seg6_local.c b/net/ipv6/seg6_local.c index d1070aec7b72..53f7b2452575 100644 --- a/net/ipv6/seg6_local.c +++ b/net/ipv6/seg6_local.c @@ -278,13 +278,43 @@ static bool decap_and_validate(struct sk_buff *skb, int proto) return true; } -static void advance_nextseg(struct ipv6_sr_hdr *srh, struct in6_addr *daddr) +/* advance the SRH to the next segment and set the IPv6 DA accordingly. + * skb pointers may change: after this call, the caller must evaluate again + * any pointer into the packet data. + * + * This function returns: + * - the SRH on success; + * - NULL when the skb cannot be made writable. In this case, the function + * sets the drop reason to SKB_DROP_REASON_NOMEM. + */ +static struct ipv6_sr_hdr *advance_nextseg(struct sk_buff *skb, + struct ipv6_sr_hdr *srh, + enum skb_drop_reason *reason) { struct in6_addr *addr; + int srhoff; + int wlen; + + srhoff = (unsigned char *)srh - skb->data; + /* we write only the Segment Left field and the IPv6 DA. The segment + * list is only read, and seg6_get_srh() already pulled it into the + * linear area. + */ + wlen = srhoff + sizeof(*srh); + + if (unlikely(skb_ensure_writable(skb, wlen))) { + *reason = SKB_DROP_REASON_NOMEM; + return NULL; + } + + /* skb_ensure_writable() may change skb pointers; evaluate srh again */ + srh = (struct ipv6_sr_hdr *)(skb->data + srhoff); srh->segments_left--; addr = srh->segments + srh->segments_left; - *daddr = *addr; + ipv6_hdr(skb)->daddr = *addr; + + return srh; } static int @@ -382,12 +412,26 @@ static bool seg6_next_csid_is_arg_zero(const struct in6_addr *addr, return true; } -/* assume that DA.Argument length > 0 */ -static void seg6_next_csid_advance_arg(struct in6_addr *addr, - const struct seg6_flavors_info *finfo) +/* assume that DA.Argument length > 0. + * skb pointers may change: after this call, the caller must evaluate again + * any pointer into the packet data. + * + * This function returns: + * - SKB_NOT_DROPPED_YET on success; + * - SKB_DROP_REASON_NOMEM when the skb cannot be made writable. + */ +static enum skb_drop_reason +seg6_next_csid_advance_arg(struct sk_buff *skb, + const struct seg6_flavors_info *finfo) { __u8 fnc_octects = seg6_flv_lcnode_func_octects(finfo); __u8 blk_octects = seg6_flv_lcblock_octects(finfo); + struct in6_addr *addr; + + if (unlikely(skb_ensure_writable(skb, sizeof(struct ipv6hdr)))) + return SKB_DROP_REASON_NOMEM; + + addr = &ipv6_hdr(skb)->daddr; /* advance DA.Argument */ memmove(&addr->s6_addr[blk_octects], @@ -395,6 +439,8 @@ static void seg6_next_csid_advance_arg(struct in6_addr *addr, 16 - blk_octects - fnc_octects); memset(&addr->s6_addr[16 - fnc_octects], 0x00, fnc_octects); + + return SKB_NOT_DROPPED_YET; } static int input_action_end_finish(struct sk_buff *skb, @@ -408,31 +454,42 @@ static int input_action_end_finish(struct sk_buff *skb, static int input_action_end_core(struct sk_buff *skb, struct seg6_local_lwt *slwt) { + enum skb_drop_reason reason = SKB_DROP_REASON_NOT_SPECIFIED; struct ipv6_sr_hdr *srh; + int err = -EINVAL; srh = get_and_validate_srh(skb); if (!srh) goto drop; - advance_nextseg(srh, &ipv6_hdr(skb)->daddr); + srh = advance_nextseg(skb, srh, &reason); + if (!srh) { + err = -ENOMEM; + goto drop; + } return input_action_end_finish(skb, slwt); drop: - kfree_skb(skb); - return -EINVAL; + kfree_skb_reason(skb, reason); + return err; } static int end_next_csid_core(struct sk_buff *skb, struct seg6_local_lwt *slwt) { const struct seg6_flavors_info *finfo = &slwt->flv_info; struct in6_addr *daddr = &ipv6_hdr(skb)->daddr; + enum skb_drop_reason reason; if (seg6_next_csid_is_arg_zero(daddr, finfo)) return input_action_end_core(skb, slwt); /* update DA */ - seg6_next_csid_advance_arg(daddr, finfo); + reason = seg6_next_csid_advance_arg(skb, finfo); + if (reason) { + kfree_skb_reason(skb, reason); + return -ENOMEM; + } return input_action_end_finish(skb, slwt); } @@ -448,19 +505,25 @@ static int input_action_end_x_finish(struct sk_buff *skb, static int input_action_end_x_core(struct sk_buff *skb, struct seg6_local_lwt *slwt) { + enum skb_drop_reason reason = SKB_DROP_REASON_NOT_SPECIFIED; struct ipv6_sr_hdr *srh; + int err = -EINVAL; srh = get_and_validate_srh(skb); if (!srh) goto drop; - advance_nextseg(srh, &ipv6_hdr(skb)->daddr); + srh = advance_nextseg(skb, srh, &reason); + if (!srh) { + err = -ENOMEM; + goto drop; + } return input_action_end_x_finish(skb, slwt); drop: - kfree_skb(skb); - return -EINVAL; + kfree_skb_reason(skb, reason); + return err; } static int end_x_next_csid_core(struct sk_buff *skb, @@ -468,12 +531,17 @@ static int end_x_next_csid_core(struct sk_buff *skb, { const struct seg6_flavors_info *finfo = &slwt->flv_info; struct in6_addr *daddr = &ipv6_hdr(skb)->daddr; + enum skb_drop_reason reason; if (seg6_next_csid_is_arg_zero(daddr, finfo)) return input_action_end_x_core(skb, slwt); /* update DA */ - seg6_next_csid_advance_arg(daddr, finfo); + reason = seg6_next_csid_advance_arg(skb, finfo); + if (reason) { + kfree_skb_reason(skb, reason); + return -ENOMEM; + } return input_action_end_x_finish(skb, slwt); } @@ -760,10 +828,12 @@ static bool seg6_pop_srh(struct sk_buff *skb, int srhoff) */ static int end_flv8986_core(struct sk_buff *skb, struct seg6_local_lwt *slwt) { + enum skb_drop_reason reason = SKB_DROP_REASON_NOT_SPECIFIED; const struct seg6_flavors_info *finfo = &slwt->flv_info; enum seg6_local_flv_action action; enum seg6_local_pktinfo pinfo; struct ipv6_sr_hdr *srh; + int err = -EINVAL; __u32 flvmask; int srhoff; @@ -787,10 +857,18 @@ static int end_flv8986_core(struct sk_buff *skb, struct seg6_local_lwt *slwt) switch (action) { case SEG6_LOCAL_FLV_ACT_END: /* process the packet as the "standard" End behavior */ - advance_nextseg(srh, &ipv6_hdr(skb)->daddr); + srh = advance_nextseg(skb, srh, &reason); + if (!srh) { + err = -ENOMEM; + goto drop; + } break; case SEG6_LOCAL_FLV_ACT_PSP: - advance_nextseg(srh, &ipv6_hdr(skb)->daddr); + srh = advance_nextseg(skb, srh, &reason); + if (!srh) { + err = -ENOMEM; + goto drop; + } if (unlikely(!seg6_pop_srh(skb, srhoff))) goto drop; @@ -807,8 +885,8 @@ static int end_flv8986_core(struct sk_buff *skb, struct seg6_local_lwt *slwt) return input_action_end_finish(skb, slwt); drop: - kfree_skb(skb); - return -EINVAL; + kfree_skb_reason(skb, reason); + return err; } /* regular endpoint function */ @@ -847,21 +925,27 @@ static int input_action_end_x(struct sk_buff *skb, struct seg6_local_lwt *slwt) static int input_action_end_t(struct sk_buff *skb, struct seg6_local_lwt *slwt) { + enum skb_drop_reason reason = SKB_DROP_REASON_NOT_SPECIFIED; struct ipv6_sr_hdr *srh; + int err = -EINVAL; srh = get_and_validate_srh(skb); if (!srh) goto drop; - advance_nextseg(srh, &ipv6_hdr(skb)->daddr); + srh = advance_nextseg(skb, srh, &reason); + if (!srh) { + err = -ENOMEM; + goto drop; + } seg6_lookup_nexthop(skb, NULL, slwt->table); return dst_input(skb); drop: - kfree_skb(skb); - return -EINVAL; + kfree_skb_reason(skb, reason); + return err; } /* decapsulate and forward inner L2 frame on specified interface */ @@ -1375,6 +1459,7 @@ static int input_action_end_b6(struct sk_buff *skb, struct seg6_local_lwt *slwt) static int input_action_end_b6_encap(struct sk_buff *skb, struct seg6_local_lwt *slwt) { + enum skb_drop_reason reason = SKB_DROP_REASON_NOT_SPECIFIED; struct ipv6_sr_hdr *srh; int err = -EINVAL; @@ -1382,7 +1467,11 @@ static int input_action_end_b6_encap(struct sk_buff *skb, if (!srh) goto drop; - advance_nextseg(srh, &ipv6_hdr(skb)->daddr); + srh = advance_nextseg(skb, srh, &reason); + if (!srh) { + err = -ENOMEM; + goto drop; + } skb_reset_inner_headers(skb); skb->encapsulation = 1; @@ -1398,7 +1487,7 @@ static int input_action_end_b6_encap(struct sk_buff *skb, return dst_input(skb); drop: - kfree_skb(skb); + kfree_skb_reason(skb, reason); return err; } @@ -1434,6 +1523,7 @@ static int input_action_end_bpf(struct sk_buff *skb, struct seg6_local_lwt *slwt) { struct seg6_bpf_srh_state *srh_state; + enum skb_drop_reason reason; struct ipv6_sr_hdr *srh; int ret; @@ -1442,7 +1532,11 @@ static int input_action_end_bpf(struct sk_buff *skb, kfree_skb(skb); return -EINVAL; } - advance_nextseg(srh, &ipv6_hdr(skb)->daddr); + srh = advance_nextseg(skb, srh, &reason); + if (!srh) { + kfree_skb_reason(skb, reason); + return -ENOMEM; + } /* The access to the per-CPU buffer srh_state is protected by running * always in softirq context (with disabled BH). On PREEMPT_RT the From 036778a4fad19b54b73ae229a1901665aa8750a2 Mon Sep 17 00:00:00 2001 From: Nicolai Buchwitz Date: Fri, 25 Sep 2026 15:52:36 +0200 Subject: [PATCH 629/694] net: ethtool: let tsconfig reach a PHY-only timestamp provider TSCONFIG_GET and TSCONFIG_SET reject a device that implements neither hwtstamp NDO, even when its PHY can serve the request. The ioctls they meant to replace handle it, so the two interfaces disagree on the same hardware and user space has to pick one. Accept the default timestamping PHY and an already installed provider on both sides. On the set side the test moves into ethnl_set_tsconfig() as the validate callback runs without rtnl. On the get side it stays ahead of ethnl_ops_begin(), so a device that can serve nothing keeps failing with EOPNOTSUPP and a dump still skips it rather than stopping there. A netdev provider needs ndo_hwtstamp_set to be programmed at all, so don't pick that source without it, and test for ndo_hwtstamp_get before calling it. Fixes: 6e9e2eed4f39 ("net: ethtool: Add support for tsconfig command to get/set hwtstamp config") Signed-off-by: Nicolai Buchwitz Reviewed-by: Maxime Chevallier Reviewed-by: Kory Maincent Link: https://patch.msgid.link/20260925135237.3432266-3-nb@tipi-net.de Signed-off-by: Jakub Kicinski --- net/core/dev_ioctl.c | 3 ++- net/ethtool/tsconfig.c | 24 ++++++++++-------------- 2 files changed, 12 insertions(+), 15 deletions(-) diff --git a/net/core/dev_ioctl.c b/net/core/dev_ioctl.c index 164643140a52..f6029f60c1dc 100644 --- a/net/core/dev_ioctl.c +++ b/net/core/dev_ioctl.c @@ -267,7 +267,8 @@ int dev_get_hwtstamp_phylib(struct net_device *dev, hwprov->phydev) return phy_hwtstamp_get(hwprov->phydev, cfg); - if (hwprov->source == HWTSTAMP_SOURCE_NETDEV) + if (hwprov->source == HWTSTAMP_SOURCE_NETDEV && + dev->netdev_ops->ndo_hwtstamp_get) return dev->netdev_ops->ndo_hwtstamp_get(dev, cfg); return -EOPNOTSUPP; diff --git a/net/ethtool/tsconfig.c b/net/ethtool/tsconfig.c index 6be3aa5d4bc1..2db0e7ba8b9f 100644 --- a/net/ethtool/tsconfig.c +++ b/net/ethtool/tsconfig.c @@ -1,6 +1,7 @@ // SPDX-License-Identifier: GPL-2.0-only #include +#include #include #include @@ -42,7 +43,9 @@ static int tsconfig_prepare_data(const struct ethnl_req_info *req_base, struct kernel_hwtstamp_config cfg = {}; int ret; - if (!dev->netdev_ops->ndo_hwtstamp_get) + if (!dev->netdev_ops->ndo_hwtstamp_get && + !phy_is_default_hwtstamp(dev->phydev) && + !netdev_ops_lock_dereference(dev->hwprov, dev)) return -EOPNOTSUPP; ret = ethnl_ops_begin(dev); @@ -248,17 +251,6 @@ static int tsconfig_send_reply(struct net_device *dev, struct genl_info *info) return ret; } -static int ethnl_set_tsconfig_validate(struct ethnl_req_info *req_base, - struct genl_info *info) -{ - const struct net_device_ops *ops = req_base->dev->netdev_ops; - - if (!ops->ndo_hwtstamp_set || !ops->ndo_hwtstamp_get) - return -EOPNOTSUPP; - - return 1; -} - static struct hwtstamp_provider * tsconfig_set_hwprov_from_desc(struct net_device *dev, struct genl_info *info, @@ -272,7 +264,7 @@ tsconfig_set_hwprov_from_desc(struct net_device *dev, int ret; ret = ethtool_net_get_ts_info_by_phc(dev, &ts_info, hwprov_desc); - if (!ret) { + if (!ret && dev->netdev_ops->ndo_hwtstamp_set) { /* Found */ source = HWTSTAMP_SOURCE_NETDEV; } else { @@ -313,6 +305,11 @@ static int ethnl_set_tsconfig(struct ethnl_req_info *req_base, if (!netif_device_present(dev)) return -ENODEV; + if (!dev->netdev_ops->ndo_hwtstamp_set && + !phy_is_default_hwtstamp(dev->phydev) && + !netdev_ops_lock_dereference(dev->hwprov, dev)) + return -EOPNOTSUPP; + if (tb[ETHTOOL_A_TSCONFIG_HWTSTAMP_PROVIDER]) { struct hwtstamp_provider_desc __hwprov_desc = {.index = -1}; struct hwtstamp_provider *__hwprov; @@ -459,6 +456,5 @@ const struct ethnl_request_ops ethnl_tsconfig_request_ops = { .reply_size = tsconfig_reply_size, .fill_reply = tsconfig_fill_reply, - .set_validate = ethnl_set_tsconfig_validate, .set = ethnl_set_tsconfig, }; From 37e02c42a00be692c06343e11779aadc45f45971 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Fri, 25 Sep 2026 13:52:43 +0000 Subject: [PATCH 630/694] net: restrict SO_RESERVE_MEM to TCP sockets and cap max value Commit 2bb2f5fb21b0 ("net: add new socket option SO_RESERVE_MEM") and commit d00c8ee31729 ("net: fix possible NULL deref in sock_reserve_memory") only checked sk_has_account(sk), which is true for both TCP and UDP sockets. However, SO_RESERVE_MEM and sk_unused_reserved_mem() are currently only supported by TCP: - On UDP sockets, sk->sk_forward_alloc is protected by sk->sk_receive_queue.lock, whereas sock_reserve_memory() and sock_release_reserved_memory() only acquire lock_sock(sk). Concurrent UDP packet reception/release and setsockopt(SO_RESERVE_MEM) corrupt sk_forward_alloc and memcg accounting. - udp_rmem_release() reclaims excess sk_forward_alloc without accounting for sk_unused_reserved_mem(sk). Restrict sock_reserve_memory() to TCP sockets (sk_is_tcp(sk)) for now. Supporting SO_RESERVE_MEM for UDP (acquiring sk_receive_queue.lock and honoring sk_unused_reserved_mem() in udp_rmem_release()) can be done in a future net-next series if needed. In addition, reject val > SZ_1G with -EINVAL in sk_setsockopt(SO_RESERVE_MEM). Without an upper bound, values near INT_MAX cause sk_mem_pages(delta) and (pages << PAGE_SHIFT) to overflow 32-bit signed int, corrupting sk->sk_forward_alloc and sk->sk_reserved_mem. Using a page-aligned cap (SZ_1G) ensures that the page-rounded sk->sk_reserved_mem reported by getsockopt(SO_RESERVE_MEM) can always be passed back to setsockopt(SO_RESERVE_MEM). Fixes: 2bb2f5fb21b0 ("net: add new socket option SO_RESERVE_MEM") Reported-by: Cai Xinchen Closes: https://lore.kernel.org/netdev/5a88421d-10ef-4fca-9acb-85a27a3c1173@huawei.com/ Signed-off-by: Eric Dumazet Reviewed-by: Wei Wang Link: https://patch.msgid.link/20260925135244.3715196-2-edumazet@google.com Signed-off-by: Jakub Kicinski --- net/core/sock.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/net/core/sock.c b/net/core/sock.c index d23333bb4f3f..e8551df8330f 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -1034,7 +1034,7 @@ static int sock_reserve_memory(struct sock *sk, int bytes) bool charged; int pages; - if (!mem_cgroup_sk_enabled(sk) || !sk_has_account(sk)) + if (!mem_cgroup_sk_enabled(sk) || !sk_is_tcp(sk)) return -EOPNOTSUPP; if (!bytes) @@ -1661,7 +1661,7 @@ int sk_setsockopt(struct sock *sk, int level, int optname, { int delta; - if (val < 0) { + if (val < 0 || val > SZ_1G) { ret = -EINVAL; break; } From e955f14fa8467d70aaeea5830b4d1773d5b25500 Mon Sep 17 00:00:00 2001 From: Zhan Xusheng Date: Mon, 28 Sep 2026 10:02:04 +0800 Subject: [PATCH 631/694] sysctl: Negate before converting in the int read path proc_int_k2u_conv_kop() reports the sign through *negp and the magnitude through *u_ptr, but for a negative value it hands the sign-extended int to the converter and negates the result: *u_ptr = k_ptr_op ? -k_ptr_op((ulong)val) : -(ulong)val; With div_hz() and CONFIG_HZ=1000 a stored -1000 becomes (ulong)-1000 / 1000 == 18446744073709550, and negating that wraps: # echo -1 > /proc/sys/net/ipv4/tcp_fin_timeout # cat /proc/sys/net/ipv4/tcp_fin_timeout -18428297329635842066 The magnitude is HZ dependent but the wrap is not: the negation happens after the division at every CONFIG_HZ. Take the magnitude first and convert that, which is what the open-coded version did before commit 2dc164a48e6f ("sysctl: Create converter functions with two new macros") folded it into a macro. The k_ptr_op == NULL branch was already correct. All three int converters that pass a k_ptr_op are affected: proc_dointvec_jiffies(), proc_dointvec_userhz_jiffies() and proc_dointvec_ms_jiffies(). Fixes: 2dc164a48e6f ("sysctl: Create converter functions with two new macros") Cc: stable@vger.kernel.org Reviewed-by: Bradley Morgan Signed-off-by: Zhan Xusheng Signed-off-by: Joel Granados --- kernel/sysctl.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/kernel/sysctl.c b/kernel/sysctl.c index f7b75985d542..38597f26b34c 100644 --- a/kernel/sysctl.c +++ b/kernel/sysctl.c @@ -483,7 +483,7 @@ int proc_int_k2u_conv_kop(ulong *u_ptr, const int *k_ptr, bool *negp, if (val < 0) { *negp = true; - *u_ptr = k_ptr_op ? -k_ptr_op((ulong)val) : -(ulong)val; + *u_ptr = k_ptr_op ? k_ptr_op(-(ulong)val) : -(ulong)val; } else { *negp = false; *u_ptr = k_ptr_op ? k_ptr_op((ulong)val) : (ulong) val; From fdd6553e1e53b0421d89c7469c8a36d2eadb1115 Mon Sep 17 00:00:00 2001 From: Zhan Xusheng Date: Mon, 28 Sep 2026 10:02:05 +0800 Subject: [PATCH 632/694] time/jiffies: Saturate in mult_hz() instead of wrapping Return ULONG_MAX for values that don't fit an unsigned long. proc_int_u2k_conv_uop() now correctly rejects a result above INT_MAX. This is the erroneous behaviour that is being fixed. The input has to exceed ULONG_MAX / HZ for the product to wrap, so the value below is specific to CONFIG_HZ=1000: # echo 18446744073709552 > /proc/sys/net/ipv4/tcp_keepalive_time # cat /proc/sys/net/ipv4/tcp_keepalive_time 0 That value is now rejected with an error. The original bound ("*u_ptr > INT_MAX / HZ") was removed in commit 2dc164a48e6f ("sysctl: Create converter functions with two new macros"). Fixes: 2dc164a48e6f ("sysctl: Create converter functions with two new macros") Cc: stable@vger.kernel.org Signed-off-by: Zhan Xusheng Signed-off-by: Joel Granados --- kernel/time/jiffies.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/kernel/time/jiffies.c b/kernel/time/jiffies.c index 80c354811538..9b3487d40cd6 100644 --- a/kernel/time/jiffies.c +++ b/kernel/time/jiffies.c @@ -101,6 +101,8 @@ void __init register_refined_jiffies(long cycles_per_second) #ifdef CONFIG_SYSCTL static ulong mult_hz(const ulong val) { + if (val >= ULONG_MAX / HZ) + return ULONG_MAX; return val * HZ; } From 0f17d2e3dd9cda3ee6d94033d5e5489af168920d Mon Sep 17 00:00:00 2001 From: Xuanqiang Luo Date: Thu, 24 Sep 2026 13:52:52 +0800 Subject: [PATCH 633/694] pfcp: fix socket lifetime on netdevice registration failure pfcp_newlink() creates the UDP socket before register_netdevice(). If registration fails after pfcp_dev_init() succeeds, the core calls pfcp_dev_uninit(), which releases the socket and clears pfcp->sk. The newlink error path then releases it again, causing a NULL pointer dereference. This was observed with failslab fault injection: FAULT_INJECTION: forcing a failure. kobject: kobject_add_internal failed for pfcp0 (error: -12 parent: net) BUG: KASAN: null-ptr-deref in udp_tunnel_sock_release+0x1c/0x50 Read of size 8 at addr 0000000000000120 by task ip/1037 Call trace: show_stack+0x18/0x24 (C) dump_stack_lvl+0x78/0x90 print_report+0x468/0x5cc kasan_report+0xa4/0xf0 __asan_load8+0x7c/0xd0 udp_tunnel_sock_release+0x1c/0x50 pfcp_newlink+0x128/0x184 rtnl_newlink+0x848/0xe44 rtnetlink_rcv_msg+0x468/0x514 netlink_rcv_skb+0xc0/0x1f0 rtnetlink_rcv+0x18/0x24 netlink_unicast+0x4b8/0x558 netlink_sendmsg+0x2b8/0x584 ... Return from pfcp_del_sock() if pfcp->sk is NULL. Fixes: 76c8764ef36a5 ("pfcp: add PFCP module") Cc: stable@vger.kernel.org Signed-off-by: Xuanqiang Luo Reviewed-by: Simon Horman Link: https://patch.msgid.link/20260924055252.33488-1-xuanqiang.luo@linux.dev Signed-off-by: Paolo Abeni --- drivers/net/pfcp.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/net/pfcp.c b/drivers/net/pfcp.c index e1cca779d2ec..3ecd43cafdf0 100644 --- a/drivers/net/pfcp.c +++ b/drivers/net/pfcp.c @@ -105,6 +105,9 @@ static int pfcp_encap_recv(struct sock *sk, struct sk_buff *skb) static void pfcp_del_sock(struct pfcp_dev *pfcp) { + if (!pfcp->sk) + return; + udp_tunnel_sock_release(pfcp->sk); pfcp->sk = NULL; } From 006c8b834aed8b02bd8fb9fb05e9e7ead37d1763 Mon Sep 17 00:00:00 2001 From: Zihan Xi Date: Thu, 24 Sep 2026 05:15:21 +0000 Subject: [PATCH 634/694] net: gso: limit recursive IP-in-IP segmentation IP-in-IP GSO can re-enter inet_gso_segment() or ipv6_gso_segment() for each nested IP header. encap_level tracks header bytes, not callback depth, so a deep chain can exhaust the kernel stack. Making inet_gso_segment() stackable introduced unbounded IPv4 nesting; IPIP GSO/TSO later made the path reachable. The IPv6 stackable path was introduced separately and uses the same guard. Count IPv4 and IPv6 GSO handler entries in skb_gso_cb, initialized once per top-level GSO operation and preserved across GRE/UDP context changes. Use the existing IP_TUNNEL_RECURSION_LIMIT for both handlers. The first five entries pass, and the sixth returns -EINVAL before dispatching another GSO callback. Fixes: 3347c9602955 ("ipv4: gso: make inet_gso_segment() stackable") Cc: stable@vger.kernel.org Reported-by: Vega Closes: https://lore.kernel.org/all/cover.1790157745.git.zihanx@nebusec.ai/ Assisted-by: LLM Co-developed-by: Luxing Yin Signed-off-by: Luxing Yin Signed-off-by: Zihan Xi Reviewed-by: Willem de Bruijn Link: https://patch.msgid.link/20260924051521.32568-2-zihanx@nebusec.ai Signed-off-by: Paolo Abeni --- include/net/gso.h | 8 ++++++++ net/core/gso.c | 1 + net/ipv4/af_inet.c | 3 +++ net/ipv6/ip6_offload.c | 4 ++++ 4 files changed, 16 insertions(+) diff --git a/include/net/gso.h b/include/net/gso.h index 29975440cad5..0749230d414e 100644 --- a/include/net/gso.h +++ b/include/net/gso.h @@ -19,10 +19,18 @@ struct skb_gso_cb { int encap_level; __wsum csum; __u16 csum_start; + /* Number of IPv4/IPv6 GSO handler entries for this packet. */ + u8 recursion_counter; }; #define SKB_GSO_CB_OFFSET 32 #define SKB_GSO_CB(skb) ((struct skb_gso_cb *)((skb)->cb + SKB_GSO_CB_OFFSET)) +static inline bool gso_recursion_inc_test(struct sk_buff *skb, + unsigned int limit) +{ + return ++SKB_GSO_CB(skb)->recursion_counter > limit; +} + static inline int skb_tnl_header_len(const struct sk_buff *inner_skb) { return (skb_mac_header(inner_skb) - inner_skb->head) - diff --git a/net/core/gso.c b/net/core/gso.c index bcd156372f4d..e96ef6350064 100644 --- a/net/core/gso.c +++ b/net/core/gso.c @@ -117,6 +117,7 @@ struct sk_buff *__skb_gso_segment(struct sk_buff *skb, SKB_GSO_CB(skb)->mac_offset = skb_headroom(skb); SKB_GSO_CB(skb)->encap_level = 0; + SKB_GSO_CB(skb)->recursion_counter = 0; skb_reset_mac_header(skb); skb_reset_mac_len(skb); diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c index 32d006c1a8ee..4ce38c99fef9 100644 --- a/net/ipv4/af_inet.c +++ b/net/ipv4/af_inet.c @@ -1375,6 +1375,9 @@ struct sk_buff *inet_gso_segment(struct sk_buff *skb, int id; skb_reset_network_header(skb); + if (unlikely(gso_recursion_inc_test(skb, + IP_TUNNEL_RECURSION_LIMIT))) + goto out; nhoff = skb_network_header(skb) - skb_mac_header(skb); if (unlikely(!pskb_may_pull(skb, sizeof(*iph)))) goto out; diff --git a/net/ipv6/ip6_offload.c b/net/ipv6/ip6_offload.c index 78f50c93c536..700f1861d8eb 100644 --- a/net/ipv6/ip6_offload.c +++ b/net/ipv6/ip6_offload.c @@ -17,6 +17,7 @@ #include #include #include +#include #include "ip6_offload.h" #include "tcpv6_offload.c" @@ -104,6 +105,9 @@ static struct sk_buff *ipv6_gso_segment(struct sk_buff *skb, bool gso_partial; skb_reset_network_header(skb); + if (unlikely(gso_recursion_inc_test(skb, + IP_TUNNEL_RECURSION_LIMIT))) + goto out; nhoff = skb_network_header(skb) - skb_mac_header(skb); if (unlikely(!pskb_may_pull(skb, sizeof(*ipv6h)))) goto out; From f8932b2c0c9ae6fb577bcc05a2080c8498f91c0d Mon Sep 17 00:00:00 2001 From: Hannu Varjoranta Date: Fri, 25 Sep 2026 14:52:37 +0300 Subject: [PATCH 635/694] gve: DQO: accept TSO packets with non-protocol gso_type bits Since commit 014c607f86ab ("gve: add support for UDP GSO for DQO format"), gve_prep_tso() matches shinfo->gso_type exactly. gso_type is a bitmask, though: SKB_GSO_DODGY is set on every GSO packet that comes from tun/tap, packet sockets or other untrusted sources, and SKB_GSO_TCP_ECN and SKB_GSO_TCP_FIXEDID can be set as well. Such packets hit the default case, gve_tx_add_skb_dqo() fails and gve_try_tx_skb() drops them, counting them in tx_dropped. These packets do reach the driver: tcp_gso_segment() passes DODGY skbs through unsegmented to devices that support TSO, after recomputing gso_segs, so drivers must tolerate the bit. This breaks virtual machines behind a tap on GCE instances using the DQO queue formats. Every TSO packet forwarded from a guest (gso_type SKB_GSO_TCPV4 | SKB_GSO_DODGY) is dropped, and guest uploads slow to a crawl of retransmissions or stall entirely, while the host's own TSO traffic (gso_type SKB_GSO_TCPV4) is unaffected. On an n4 instance (DQO-QPL) with a Cloud Hypervisor guest, a 64 MB upload from the guest went from a 60 s timeout at ~0.9 MB/s to 0.19 s with this change, with tx_dropped no longer increasing. Host TCP with ECN is hit as well: gve advertises NETIF_F_TSO_ECN, so a TSO packet carrying CWR has SKB_GSO_TCP_ECN set and is dropped too. Restore the bitmask test that commit 1b9f75634441 ("gve: ignore nonrelevant GSO type bits when processing TSO headers") introduced for the same problem, keeping the UDP GSO support. While here, reload shinfo after skb_cow_head(). If the head was cloned, pskb_expand_head() moves skb_shared_info to the new head, and the pointer cached at function entry can then refer to memory that another clone frees. Fixes: 014c607f86ab ("gve: add support for UDP GSO for DQO format") Signed-off-by: Hannu Varjoranta Reviewed-by: Eric Dumazet Reviewed-by: Ankit Garg Reviewed-by: Harshitha Ramamurthy Link: https://patch.msgid.link/20260925115237.56465-1-hannu@varjosoft.com Signed-off-by: Paolo Abeni --- drivers/net/ethernet/google/gve/gve_tx_dqo.c | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) diff --git a/drivers/net/ethernet/google/gve/gve_tx_dqo.c b/drivers/net/ethernet/google/gve/gve_tx_dqo.c index 616c1921aebe..ad99cbb77e29 100644 --- a/drivers/net/ethernet/google/gve/gve_tx_dqo.c +++ b/drivers/net/ethernet/google/gve/gve_tx_dqo.c @@ -595,25 +595,25 @@ static int gve_prep_tso(struct sk_buff *skb) err = skb_cow_head(skb, 0); if (err < 0) return err; + shinfo = skb_shinfo(skb); l4_start = skb_transport_offset(skb); paylen = skb->len - l4_start; - switch (shinfo->gso_type) { - case SKB_GSO_TCPV4: - case SKB_GSO_TCPV6: + /* gso_type is a bitmask: SKB_GSO_DODGY, SKB_GSO_TCP_ECN and + * SKB_GSO_TCP_FIXEDID may be set alongside the protocol bit. + */ + if (shinfo->gso_type & (SKB_GSO_TCPV4 | SKB_GSO_TCPV6)) { tcp = tcp_hdr(skb); csum_replace_by_diff(&tcp->check, (__force __wsum)htonl(paylen)); header_len = skb_tcp_all_headers(skb); - break; - case SKB_GSO_UDP_L4: + } else if (shinfo->gso_type & SKB_GSO_UDP_L4) { udp = udp_hdr(skb); csum_replace_by_diff(&udp->check, (__force __wsum)htonl(paylen)); header_len = sizeof(struct udphdr) + l4_start; - break; - default: + } else { return -EINVAL; } From f341dc4d934a69ffbe6c21e3609ee725790188f1 Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Mon, 21 Sep 2026 21:12:00 +0200 Subject: [PATCH 636/694] cpufreq: intel_pstate: Fix max_freq fallback in cpufreq_update_pressure() After commit d2d5c129d07e ("cpufreq: Make cpufreq_update_pressure() fall back to cpuinfo.max_freq"), cpufreq pressure appears in the CPU load balancer unexpectedly in some cases in which it was not present before, leading to confusion and uncertainty. Clearly, the scheduler assumes that cpufreq pressure will not be set unless the capacity reference frequency of the CPU is known, and the commit mentioned above violates that assumption. However, in some cases the capacity reference frequency of the CPU is in fact known even though arch_scale_freq_ref() returns 0 and in those cases it should be possible to set cpufreq pressure as appropriate. For this purpose, introduce a new cpufreq driver callback returning the CPU capacity reference frequency, .scale_freq_ref(), and make cpufreq_update_pressure() invoke it, if present, instead of falling back to cpuinfo.max_freq unconditionally. Add that callback to the intel_pstate driver and make it return 0 unless the scale-invariant capacity of the given CPU has been explicitly set, in which cases its reference frequency is always cpuinfo.max_freq. Fixes: d2d5c129d07e ("cpufreq: Make cpufreq_update_pressure() fall back to cpuinfo.max_freq") Reported-by: Jianyong Wu Closes: https://lore.kernel.org/linux-pm/20260915065747.1671965-1-wujianyong@hygon.cn/ Tested-by: Jianyong Wu Signed-off-by: Rafael J. Wysocki Tested-by: Ricardo Neri # Intel hybrid parts Tested-by: Chen Yu [ rjw: Add READ_ONCE() around a capacity_perf read ] Link: https://patch.msgid.link/12975163.O9o76ZdvQC@rafael.j.wysocki Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/cpufreq.c | 4 ++-- drivers/cpufreq/intel_pstate.c | 10 ++++++++++ include/linux/cpufreq.h | 3 +++ 3 files changed, 15 insertions(+), 2 deletions(-) diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c index 96515880b4ac..54dde8419bdc 100644 --- a/drivers/cpufreq/cpufreq.c +++ b/drivers/cpufreq/cpufreq.c @@ -2590,8 +2590,8 @@ static void cpufreq_update_pressure(struct cpufreq_policy *policy) cpu = cpumask_first(policy->related_cpus); max_freq = arch_scale_freq_ref(cpu); - if (!max_freq) - max_freq = policy->cpuinfo.max_freq; + if (!max_freq && cpufreq_driver->scale_freq_ref) + max_freq = cpufreq_driver->scale_freq_ref(policy); capped_freq = policy->max; diff --git a/drivers/cpufreq/intel_pstate.c b/drivers/cpufreq/intel_pstate.c index ceb340f7a110..17158cb25dd8 100644 --- a/drivers/cpufreq/intel_pstate.c +++ b/drivers/cpufreq/intel_pstate.c @@ -1135,6 +1135,14 @@ static bool hybrid_clear_max_perf_cpu(void) return ret; } +static unsigned int intel_pstate_scale_freq_ref(struct cpufreq_policy *policy) +{ + if (READ_ONCE(all_cpu_data[policy->cpu]->capacity_perf)) + return policy->cpuinfo.max_freq; + + return 0; +} + static void intel_pstate_update_freq_limits(struct cpudata *cpu) { int scaling = cpu->pstate.scaling; @@ -3088,6 +3096,7 @@ static struct cpufreq_driver intel_pstate = { .offline = intel_pstate_cpu_offline, .online = intel_pstate_cpu_online, .update_limits = intel_pstate_update_limits, + .scale_freq_ref = intel_pstate_scale_freq_ref, .name = "intel_pstate", }; @@ -3411,6 +3420,7 @@ static struct cpufreq_driver intel_cpufreq = { .suspend = intel_cpufreq_suspend, .resume = intel_pstate_resume, .update_limits = intel_pstate_update_limits, + .scale_freq_ref = intel_pstate_scale_freq_ref, .name = "intel_cpufreq", }; diff --git a/include/linux/cpufreq.h b/include/linux/cpufreq.h index 35ce665edfd8..d3d0d9d02aa4 100644 --- a/include/linux/cpufreq.h +++ b/include/linux/cpufreq.h @@ -420,6 +420,9 @@ struct cpufreq_driver { /* Will be called after the driver is fully initialized */ void (*ready)(struct cpufreq_policy *policy); + /* Return the capacity reference frequency for policy. */ + unsigned int (*scale_freq_ref)(struct cpufreq_policy *policy); + struct freq_attr **attr; /* platform specific boost support code */ From 0923198be4ae714b46350f742cb19ccd75bfeafa Mon Sep 17 00:00:00 2001 From: Sagar Taunk Date: Fri, 25 Sep 2026 08:55:04 +0000 Subject: [PATCH 637/694] rust: net: netlink: validate attribute length before casting to `c_int` `put()` trusted an unchecked `as` cast from `usize` to `c_int`. When the length exceeds `i32::MAX` that cast wraps around to a negative value. This ultimately resulted in a kernel panic when the reinterpreted value via `__nla_reserve()` and `skb_put()` became enormous. Validate payload and header both fit together in a `u16`, rejecting any payload that wouldn't leave room for `NLA_HDRLEN`. Fixes: 5eaa5fbb6e6c ("rust: netlink: add raw netlink abstraction") Reviewed-by: Alice Ryhl Reviewed-by: Alexandre Courbot Reviewed-by: Carlos Llamas Reviewed-by: Gary Guo Signed-off-by: Sagar Taunk Link: https://patch.msgid.link/20260925085449.22553-1-sagartaunk@proton.me Signed-off-by: Paolo Abeni --- rust/kernel/net/netlink.rs | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/rust/kernel/net/netlink.rs b/rust/kernel/net/netlink.rs index 22ef3dde36fa..2f8733a694a9 100644 --- a/rust/kernel/net/netlink.rs +++ b/rust/kernel/net/netlink.rs @@ -11,6 +11,7 @@ use kernel::{ alloc::{self, AllocError}, error::to_result, + num::casts::u16_as_usize, prelude::*, transmute::AsBytes, types::Opaque, @@ -86,9 +87,17 @@ fn put(&mut self, attrtype: c_int, value: &T) -> Result where T: ?Sized + AsBytes, { + // `nla_len` is a 16-bit field that encodes the total attribute length + // (header + payload). Subtracting the header size from `u16::MAX` gives + // the largest payload that still fits within that field. + const MAX_PAYLOAD_LEN: usize = u16_as_usize(u16::MAX) - size_of::(); + let skb = self.skb.skb.as_ptr(); let len = size_of_val(value); let ptr = core::ptr::from_ref(value).cast::(); + if len > MAX_PAYLOAD_LEN { + return Err(EMSGSIZE); + } // SAFETY: `skb` is valid by `NetlinkSkBuff` type invariants, and the provided value is // readable and initialized for its `size_of` bytes. to_result(unsafe { bindings::nla_put(skb, attrtype, len as c_int, ptr) }) From d031465acc362866f27e4f8f79cfecf1037fe2e5 Mon Sep 17 00:00:00 2001 From: Jamal Hadi Salim Date: Sat, 26 Sep 2026 14:03:27 -0400 Subject: [PATCH 638/694] net/sched: fq_codel: match the no-drop threshold to the packet size commit d9ebd8f9aa8b ("net/sched: fq_codel: clamp default quantum and mtu") clamped both q->quantum and q->cparams.mtu to [256, FQ_CODEL_QUANTUM_MAX]. The two fields mean different things: quantum is a DRR credit that wants the 256 floor, but cparams.mtu is the CoDel no-drop threshold (codel_impl.h "*backlog <= params->mtu"). On a link whose maximum transmitted packet size is below 256, the floor extends CoDel's minimum-backlog exemption beyond one packet and delays drop or mark eligibility by several small packets. Split the clamp. quantum keeps [256, FQ_CODEL_QUANTUM_MAX]; cparams.mtu tracks psched_mtu() (the device MTU plus its hard-header length) with only the upper bound that guards the original overflow (psched_mtu() wrapping to ~2 GiB on a huge-MTU device). Conditions to recreate the bug: attach an fq_codel qdisc on a link whose MTU plus hard_header_len is below 256 (e.g. a CAN interface). At that MTU the no-drop threshold must equal the device MTU plus its hard-header length; before this patch it was forced to 256. Basic Testing done: with dev->mtu=100 and hard_header_len=14, a return probe on fq_codel_init() observed cparams.mtu change from 256 to 114 Fixes: d9ebd8f9aa8b ("net/sched: fq_codel: clamp default quantum and mtu") Reported-by: Sashiko (nipa) Link: https://netdev-ai.bots.linux.dev/sashiko/#/patchset/20260819143136.57350-1-jhs@mojatatu.com Signed-off-by: Jamal Hadi Salim Link: https://patch.msgid.link/QDISC-34MS.v1.20260925165535@mojatatu.com Signed-off-by: Paolo Abeni --- net/sched/sch_fq_codel.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/net/sched/sch_fq_codel.c b/net/sched/sch_fq_codel.c index 969b2510b0b8..e6c87a32950f 100644 --- a/net/sched/sch_fq_codel.c +++ b/net/sched/sch_fq_codel.c @@ -517,8 +517,8 @@ static int fq_codel_init(struct Qdisc *sch, struct nlattr *opt, q->flows_cnt = 1024; q->memory_limit = 32 << 20; /* 32 MBytes */ q->drop_batch_size = 64; - mtu = clamp_t(u32, psched_mtu(qdisc_dev(sch)), 256, FQ_CODEL_QUANTUM_MAX); - q->quantum = mtu; + mtu = min_t(u32, psched_mtu(qdisc_dev(sch)), FQ_CODEL_QUANTUM_MAX); + q->quantum = clamp_t(u32, mtu, 256, FQ_CODEL_QUANTUM_MAX); INIT_LIST_HEAD(&q->new_flows); INIT_LIST_HEAD(&q->old_flows); codel_params_init(&q->cparams); From 54518e0e827f4ca9229ae657022c60bf60f5c1bf Mon Sep 17 00:00:00 2001 From: Jamal Hadi Salim Date: Sat, 26 Sep 2026 14:03:28 -0400 Subject: [PATCH 639/694] net/sched: sch_codel: match the no-drop threshold to the packet size commit 6439461f1618 ("net/sched: sch_codel: clamp default mtu to avoid disabling CoDel") clamped q->params.mtu to [256, 1 << 20]. The value is the CoDel no-drop threshold (codel_impl.h "*backlog <= params->mtu"), so on a link whose maximum transmitted packet size is below 256 the floor extends CoDel's minimum-backlog exemption beyond one packet and delays drop or mark eligibility by several small packets. Keep the upper bound that guards the original overflow (psched_mtu() wrapping to ~2 GiB on a huge-MTU device) but drop the 256 floor, so the threshold tracks the real device packet size. Conditions to recreate the bug: attach a codel qdisc on a link whose MTU plus hard_header_len is below 256 (e.g. a CAN interface). At that MTU the no-drop threshold must equal the device MTU plus its hard-header length; before this patch it was forced to 256. Requires CAP_NET_ADMIN in a user namespace. Fixes: 6439461f1618 ("net/sched: sch_codel: clamp default mtu to avoid disabling CoDel") Reported-by: Sashiko (nipa) Link: https://netdev-ai.bots.linux.dev/sashiko/#/patchset/20260819143136.57350-1-jhs@mojatatu.com Signed-off-by: Jamal Hadi Salim Link: https://patch.msgid.link/QDISC-34MS.v1.20260925165535@mojatatu.com.2 Signed-off-by: Paolo Abeni --- net/sched/sch_codel.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/sched/sch_codel.c b/net/sched/sch_codel.c index 6aa5829d6961..a1ba2e355f2d 100644 --- a/net/sched/sch_codel.c +++ b/net/sched/sch_codel.c @@ -205,7 +205,7 @@ static int codel_init(struct Qdisc *sch, struct nlattr *opt, codel_params_init(&q->params); codel_vars_init(&q->vars); codel_stats_init(&q->stats); - q->params.mtu = clamp_t(u32, psched_mtu(qdisc_dev(sch)), 256, 1 << 20); + q->params.mtu = min_t(u32, psched_mtu(qdisc_dev(sch)), 1 << 20); if (opt) { int err = codel_change(sch, opt, extack); From e079ef0b1c40c7a4a1be0c1ac5dbff95685c3beb Mon Sep 17 00:00:00 2001 From: Marc Zyngier Date: Tue, 29 Sep 2026 10:35:44 +0100 Subject: [PATCH 640/694] KVM: arm64: vgic: Allow last_lr_irq to be NULL when LRs are not overflowing last_lr_irq is always populated when there is any interrupt populated in the AP list. Not only this is not necessary (it is only useful when we completely fill the LRs), but this is in the way of further fixes. Make sure last_lr_irq is kept to NULL when we LRs are not completely full. Fixes: 6da5e537f5afe ("KVM: arm64: vgic: Pick EOIcount deactivations from AP-list tail") Reviewed-by: Fuad Tabba Tested-by: Fuad Tabba Signed-off-by: Marc Zyngier Cc: stable@vger.kernel.org Link: https://patch.msgid.link/20260929093548.3598547-4-maz@kernel.org Signed-off-by: Oliver Upton --- arch/arm64/kvm/vgic/vgic-v2.c | 6 ++++-- arch/arm64/kvm/vgic/vgic-v3.c | 6 ++++-- arch/arm64/kvm/vgic/vgic.c | 8 +++----- 3 files changed, 11 insertions(+), 9 deletions(-) diff --git a/arch/arm64/kvm/vgic/vgic-v2.c b/arch/arm64/kvm/vgic/vgic-v2.c index 7182f63fc938..70cc53ba810a 100644 --- a/arch/arm64/kvm/vgic/vgic-v2.c +++ b/arch/arm64/kvm/vgic/vgic-v2.c @@ -122,6 +122,10 @@ void vgic_v2_fold_lr_state(struct kvm_vcpu *vcpu) for (int lr = 0; lr < vgic_cpu->vgic_v2.used_lrs; lr++) vgic_v2_fold_lr(vcpu, cpuif->vgic_lr[lr]); + cpuif->used_lrs = 0; + if (!irq) + return; + /* See the GICv3 equivalent for the EOIcount handling rationale */ list_for_each_entry_continue(irq, &vgic_cpu->ap_list_head, ap_list) { u32 lr; @@ -144,8 +148,6 @@ void vgic_v2_fold_lr_state(struct kvm_vcpu *vcpu) vgic_v2_fold_lr(vcpu, lr); eoicount--; } - - cpuif->used_lrs = 0; } void vgic_v2_deactivate(struct kvm_vcpu *vcpu, u32 val) diff --git a/arch/arm64/kvm/vgic/vgic-v3.c b/arch/arm64/kvm/vgic/vgic-v3.c index 726e20a1da6e..346bacb3198f 100644 --- a/arch/arm64/kvm/vgic/vgic-v3.c +++ b/arch/arm64/kvm/vgic/vgic-v3.c @@ -155,6 +155,10 @@ void vgic_v3_fold_lr_state(struct kvm_vcpu *vcpu) for (int lr = 0; lr < cpuif->used_lrs; lr++) vgic_v3_fold_lr(vcpu, cpuif->vgic_lr[lr]); + cpuif->used_lrs = 0; + if (!irq) + return; + /* * EOIMode=0: use EOIcount to emulate deactivation. We are * guaranteed to deactivate in reverse order of the activation, so @@ -188,8 +192,6 @@ void vgic_v3_fold_lr_state(struct kvm_vcpu *vcpu) vgic_v3_fold_lr(vcpu, lr); eoicount--; } - - cpuif->used_lrs = 0; } void vgic_v3_deactivate(struct kvm_vcpu *vcpu, u64 val) diff --git a/arch/arm64/kvm/vgic/vgic.c b/arch/arm64/kvm/vgic/vgic.c index b25303d9919f..425503e0c582 100644 --- a/arch/arm64/kvm/vgic/vgic.c +++ b/arch/arm64/kvm/vgic/vgic.c @@ -866,9 +866,6 @@ static void vgic_fold_state(struct kvm_vcpu *vcpu) return; } - if (!*host_data_ptr(last_lr_irq)) - return; - if (kvm_vgic_global_state.type == VGIC_V2) vgic_v2_fold_lr_state(vcpu); else @@ -1021,11 +1018,12 @@ static void vgic_flush_lr_state(struct kvm_vcpu *vcpu) scoped_guard(raw_spinlock, &irq->irq_lock) { if (likely(vgic_target_oracle(irq) == vcpu)) { vgic_populate_lr(vcpu, irq, count++); - *host_data_ptr(last_lr_irq) = irq; + if (count == kvm_vgic_global_state.nr_lr) + *host_data_ptr(last_lr_irq) = irq; } } - if (count == kvm_vgic_global_state.nr_lr) + if (*host_data_ptr(last_lr_irq)) break; } From 19bc5d79ece613f242b5ec03921a3b465321bfed Mon Sep 17 00:00:00 2001 From: Marc Zyngier Date: Tue, 29 Sep 2026 10:35:45 +0100 Subject: [PATCH 641/694] KVM: arm64: vgic: Take a refcount on IRQs referenced by last_lr_irq Referencing the last interrupt inserted in an LR is rather fragile, as this interrupt can vanish if a concurrently unmapped LPI. Solve this by bumping up the refcount on the interrupt when populating last_lr_irq, and drop it at vgic_prune_ap_list() time, when LPIs are being reclaimed. Fixes: 6da5e537f5afe ("KVM: arm64: vgic: Pick EOIcount deactivations from AP-list tail") Reported-by: Yuchao Zhang Reviewed-by: Fuad Tabba Tested-by: Fuad Tabba Signed-off-by: Marc Zyngier Cc: stable@vger.kernel.org Link: https://patch.msgid.link/20260929093548.3598547-5-maz@kernel.org Signed-off-by: Oliver Upton --- arch/arm64/kvm/vgic/vgic.c | 15 ++++++++++++++- 1 file changed, 14 insertions(+), 1 deletion(-) diff --git a/arch/arm64/kvm/vgic/vgic.c b/arch/arm64/kvm/vgic/vgic.c index 425503e0c582..5cf5a1ef86cd 100644 --- a/arch/arm64/kvm/vgic/vgic.c +++ b/arch/arm64/kvm/vgic/vgic.c @@ -853,6 +853,17 @@ static void vgic_prune_ap_list(struct kvm_vcpu *vcpu) goto retry; } + /* + * Fix the last_lr_irq refcount which was obtained while + * populating the LRs. This can also result in the LPI being + * deleted. + */ + irq = *host_data_ptr(last_lr_irq); + if (irq) { + deleted_lpis |= vgic_put_irq_norelease(vcpu->kvm, irq); + *host_data_ptr(last_lr_irq) = NULL; + } + raw_spin_unlock(&vgic_cpu->ap_list_lock); if (unlikely(deleted_lpis)) @@ -1018,8 +1029,10 @@ static void vgic_flush_lr_state(struct kvm_vcpu *vcpu) scoped_guard(raw_spinlock, &irq->irq_lock) { if (likely(vgic_target_oracle(irq) == vcpu)) { vgic_populate_lr(vcpu, irq, count++); - if (count == kvm_vgic_global_state.nr_lr) + if (count == kvm_vgic_global_state.nr_lr) { + vgic_get_irq_ref(irq); *host_data_ptr(last_lr_irq) = irq; + } } } From 751f4641560be8568dcc5af400c0e55a283abf51 Mon Sep 17 00:00:00 2001 From: Marc Zyngier Date: Tue, 29 Sep 2026 10:35:47 +0100 Subject: [PATCH 642/694] KVM: arm64: vgic-its: Fix MOVALL handling of source redistributor The MOVALL command moves the pending state of all LPIs targeting a given redistributor to another one. However, we seem to have lost the filtering on the source RD, which means we move all LPIs to the target. Not quite what the spec mandates. Hack update_affinity() to take an optional source vcpu that is used as a filter when non-NULL, restoring the filtering that was performed by vgic_copy_lpi_list() back in the days. Fixes: 11f4f8f3e6e06 ("KVM: arm64: vgic-its: Walk LPI xarray in vgic_its_cmd_handle_movall()") Signed-off-by: Marc Zyngier Cc: stable@vger.kernel.org Link: https://patch.msgid.link/20260929093548.3598547-7-maz@kernel.org Signed-off-by: Oliver Upton --- arch/arm64/kvm/vgic/vgic-its.c | 12 ++++++++---- 1 file changed, 8 insertions(+), 4 deletions(-) diff --git a/arch/arm64/kvm/vgic/vgic-its.c b/arch/arm64/kvm/vgic/vgic-its.c index 0904ae850c35..c10a11373e10 100644 --- a/arch/arm64/kvm/vgic/vgic-its.c +++ b/arch/arm64/kvm/vgic/vgic-its.c @@ -319,12 +319,16 @@ static int update_lpi_config(struct kvm *kvm, struct vgic_irq *irq, return ret; } -static int update_affinity(struct vgic_irq *irq, struct kvm_vcpu *vcpu) +static int update_affinity(struct vgic_irq *irq, + struct kvm_vcpu *from_vcpu, struct kvm_vcpu *vcpu) { struct its_vlpi_map map; int ret; guard(raw_spinlock_irqsave)(&irq->irq_lock); + if (from_vcpu && irq->target_vcpu != from_vcpu) + return 0; + irq->target_vcpu = vcpu; if (!irq->hw) @@ -362,7 +366,7 @@ static void update_affinity_ite(struct kvm *kvm, struct its_ite *ite) return; vcpu = collection_to_vcpu(kvm, ite->collection); - update_affinity(ite->irq, vcpu); + update_affinity(ite->irq, NULL, vcpu); } /* @@ -856,7 +860,7 @@ static int vgic_its_cmd_handle_movi(struct kvm *kvm, struct vgic_its *its, vgic_its_invalidate_cache(its); - return update_affinity(ite->irq, vcpu); + return update_affinity(ite->irq, NULL, vcpu); } static bool __is_visible_gfn_locked(struct vgic_its *its, gpa_t gpa) @@ -1383,7 +1387,7 @@ static int vgic_its_cmd_handle_movall(struct kvm *kvm, struct vgic_its *its, if (!irq) continue; - update_affinity(irq, vcpu2); + update_affinity(irq, vcpu1, vcpu2); vgic_put_irq(kvm, irq); } From 4bdd2b3a708c1feb701cb8b62feb8ad2fd77bb07 Mon Sep 17 00:00:00 2001 From: Fuad Tabba Date: Tue, 29 Sep 2026 10:00:28 +0100 Subject: [PATCH 643/694] KVM: arm64: Apply the fine-grained UNDEFs without FEAT_FGT FGUs don't actually require hardware support, so don't gate fgt trap config insertion on FGT. This is the only action required to allow FGUs always, as kvm_calculate_traps() already computes the FGU bits with or without FEAT_FGT. This also fixes a spurious WARN when a TLBI OS is run on a guest in vEL1 without FEAT_TLBIOS on non FEAT_FGT hardware. In this configuration the course-grained HCR_EL2.TTLBOS trap reaches the handler handle_tlbi_el1(), which expects an EL1 TLBI only from vEL2. With FGUs enabled the UNDEF will be injected without needing the specific handler. Fixes: f5a5a406b4b8b ("KVM: arm64: Propagate and handle Fine-Grained UNDEF bits") Cc: stable@vger.kernel.org Signed-off-by: Fuad Tabba Reviewed-by: Wei-Lin Chang Link: https://patch.msgid.link/20260929090031.3829185-2-fuad.tabba@linux.dev [oliver: Take Wei-Lin's suggested changelog] Signed-off-by: Oliver Upton --- arch/arm64/kvm/emulate-nested.c | 3 --- 1 file changed, 3 deletions(-) diff --git a/arch/arm64/kvm/emulate-nested.c b/arch/arm64/kvm/emulate-nested.c index 3806ff0920fe..f59d89d04eca 100644 --- a/arch/arm64/kvm/emulate-nested.c +++ b/arch/arm64/kvm/emulate-nested.c @@ -2386,9 +2386,6 @@ int __init populate_nv_trap_config(void) print_nv_trap_error(fgt, "FGT bit is reserved", ret); } - if (!cpus_have_final_cap(ARM64_HAS_FGT)) - continue; - prev = xa_store(&sr_forward_xa, enc, xa_mk_value(tc.val), GFP_KERNEL); From 80ae56184b57176b908cd6a7ececf274a7e4b3c7 Mon Sep 17 00:00:00 2001 From: Fuad Tabba Date: Tue, 29 Sep 2026 10:00:29 +0100 Subject: [PATCH 644/694] KVM: arm64: Clear HCR_EL2.RW for 32-bit non-protected vCPUs pKVM keeps its own copy of each vCPU's HCR_EL2 at EL2, initialised with RW set unconditionally. Nothing stops userspace from creating a non-protected AArch32 VM, and with RW set, entering one of its vCPUs is an illegal exception return: KVM_RUN fails with KVM_EXIT_FAIL_ENTRY. Clear RW for a vCPU that is AArch32 at EL1, when the CPU has AArch32 EL1. On a CPU without it, RW stays set and the entry still fails, rather than EL2 switching *32_EL2 registers that are UNDEFINED there. The host already rejects such a vCPU, but EL2 takes the vCPU's features from the host and doesn't rely on that. Fixes: b56680de9c648 ("KVM: arm64: Initialize trap register values in hyp in pKVM") Cc: stable@vger.kernel.org Signed-off-by: Fuad Tabba Link: https://patch.msgid.link/20260929090031.3829185-3-fuad.tabba@linux.dev Signed-off-by: Oliver Upton --- arch/arm64/kvm/hyp/nvhe/pkvm.c | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/arch/arm64/kvm/hyp/nvhe/pkvm.c b/arch/arm64/kvm/hyp/nvhe/pkvm.c index bb3e0dc0676e..cd09d8888202 100644 --- a/arch/arm64/kvm/hyp/nvhe/pkvm.c +++ b/arch/arm64/kvm/hyp/nvhe/pkvm.c @@ -54,6 +54,15 @@ static void pkvm_vcpu_reset_hcr(struct kvm_vcpu *vcpu) else vcpu->arch.hcr_el2 |= HCR_TID2; + /* + * Without AArch32 EL1, leave RW set and let the entry fail with an + * illegal exception return: the *32_EL2 registers EL2 would otherwise + * switch are UNDEFINED there. + */ + if (vcpu_has_feature(vcpu, KVM_ARM_VCPU_EL1_32BIT) && + cpus_have_final_cap(ARM64_HAS_32BIT_EL1)) + vcpu->arch.hcr_el2 &= ~HCR_EL2_RW; + if (vcpu_has_ptrauth(vcpu)) vcpu->arch.hcr_el2 |= (HCR_API | HCR_APK); From 04447b95c62bef2b3591486c66244c49591a3dbd Mon Sep 17 00:00:00 2001 From: Fuad Tabba Date: Tue, 29 Sep 2026 10:00:30 +0100 Subject: [PATCH 645/694] KVM: arm64: Use the host's HCR_EL2 for non-protected VMs in pKVM For each vCPU, pKVM's EL2 builds its own HCR_EL2 and takes only TWI, TWE and VSE from the host's value on each entry. For a non-protected VM it has fallen behind the host's: on a CPU with MTE the guest can read GMID_EL1, on one with FEAT_EVT2 but no FGT it can execute a TLBI OS its ID registers hide, an interrupt injected without a vGIC (VI, VF) never arrives, and set/way emulation loses the TVM trap it relies on. For a non-protected VM, also take the bits the host varies with the VM's configuration or at runtime: VI, VF, TVM, TID2, TID4, TID5 and TTLBOS. They only pend interrupts for, or add traps to, a VM the host controls. The traps exit to the host, which handles them as it does without pKVM. The bits that change what EL2 does on entry and exit (E2H, RW, API/APK) or depend only on the CPU (TEA, TERR, FWB) stay EL2's. A protected VM still takes only TWI, TWE and VSE. EL2 now sets TID2 or TID4 only for a protected VM, and never sets ATA, as the host rejects KVM_CAP_ARM_MTE in pKVM. Fixes: b56680de9c648 ("KVM: arm64: Initialize trap register values in hyp in pKVM") Signed-off-by: Fuad Tabba Link: https://patch.msgid.link/20260929090031.3829185-4-fuad.tabba@linux.dev Signed-off-by: Oliver Upton --- arch/arm64/kvm/hyp/include/nvhe/pkvm.h | 9 +++++++++ arch/arm64/kvm/hyp/nvhe/hyp-main.c | 7 ++++--- arch/arm64/kvm/hyp/nvhe/pkvm.c | 18 ++++++++---------- 3 files changed, 21 insertions(+), 13 deletions(-) diff --git a/arch/arm64/kvm/hyp/include/nvhe/pkvm.h b/arch/arm64/kvm/hyp/include/nvhe/pkvm.h index c904647d2f76..5c050f21066a 100644 --- a/arch/arm64/kvm/hyp/include/nvhe/pkvm.h +++ b/arch/arm64/kvm/hyp/include/nvhe/pkvm.h @@ -12,6 +12,15 @@ #include #include +/* + * HCR_EL2 bits EL2 takes from the host on each entry, per VM type. The rest + * are EL2's own and nothing the host sets there reaches the guest. + */ +#define PKVM_HCR_EL2_HOST_PVM (HCR_EL2_TWI | HCR_EL2_TWE | HCR_EL2_VSE) +#define PKVM_HCR_EL2_HOST_NPVM (PKVM_HCR_EL2_HOST_PVM | HCR_EL2_VI | HCR_EL2_VF | \ + HCR_EL2_TVM | HCR_EL2_TID2 | HCR_EL2_TID4 | \ + HCR_EL2_TID5 | HCR_EL2_TTLBOS) + /* * Holds the relevant data for maintaining the vcpu state completely at hyp. */ diff --git a/arch/arm64/kvm/hyp/nvhe/hyp-main.c b/arch/arm64/kvm/hyp/nvhe/hyp-main.c index 9a3b92e626ad..ac64a036b0a9 100644 --- a/arch/arm64/kvm/hyp/nvhe/hyp-main.c +++ b/arch/arm64/kvm/hyp/nvhe/hyp-main.c @@ -216,6 +216,7 @@ static void sync_debug_state(struct pkvm_hyp_vcpu *hyp_vcpu) static void flush_hyp_vcpu(struct pkvm_hyp_vcpu *hyp_vcpu) { struct kvm_vcpu *host_vcpu = hyp_vcpu->host_vcpu; + u64 host_hcr_mask = PKVM_HCR_EL2_HOST_PVM; fpsimd_sve_flush(); flush_debug_state(hyp_vcpu); @@ -228,6 +229,7 @@ static void flush_hyp_vcpu(struct pkvm_hyp_vcpu *hyp_vcpu) if (!pkvm_hyp_vcpu_is_protected(hyp_vcpu)) { if (vcpu_get_flag(host_vcpu, PKVM_HOST_STATE_DIRTY)) flush_hyp_vcpu_state(hyp_vcpu); + host_hcr_mask = PKVM_HCR_EL2_HOST_NPVM; } else { hyp_vcpu->vcpu.arch.ctxt = host_vcpu->arch.ctxt; } @@ -241,9 +243,8 @@ static void flush_hyp_vcpu(struct pkvm_hyp_vcpu *hyp_vcpu) * trap-control bit, so it must flow to the hyp vCPU alongside TWI/TWE * for the vSError to be delivered. sync_hyp_vcpu() reflects it back. */ - hyp_vcpu->vcpu.arch.hcr_el2 &= ~(HCR_TWI | HCR_TWE | HCR_VSE); - hyp_vcpu->vcpu.arch.hcr_el2 |= READ_ONCE(host_vcpu->arch.hcr_el2) & - (HCR_TWI | HCR_TWE | HCR_VSE); + hyp_vcpu->vcpu.arch.hcr_el2 &= ~host_hcr_mask; + hyp_vcpu->vcpu.arch.hcr_el2 |= READ_ONCE(host_vcpu->arch.hcr_el2) & host_hcr_mask; hyp_vcpu->vcpu.arch.iflags = host_vcpu->arch.iflags; diff --git a/arch/arm64/kvm/hyp/nvhe/pkvm.c b/arch/arm64/kvm/hyp/nvhe/pkvm.c index cd09d8888202..6290c4b62659 100644 --- a/arch/arm64/kvm/hyp/nvhe/pkvm.c +++ b/arch/arm64/kvm/hyp/nvhe/pkvm.c @@ -30,6 +30,7 @@ unsigned int kvm_host_sve_max_vl; */ static DEFINE_PER_CPU(struct pkvm_hyp_vcpu *, loaded_hyp_vcpu); +/* The PKVM_HCR_EL2_HOST_{PVM,NPVM} bits of this value come from the host on each entry. */ static void pkvm_vcpu_reset_hcr(struct kvm_vcpu *vcpu) { vcpu->arch.hcr_el2 = HCR_GUEST_FLAGS; @@ -47,13 +48,6 @@ static void pkvm_vcpu_reset_hcr(struct kvm_vcpu *vcpu) if (cpus_have_final_cap(ARM64_HAS_STAGE2_FWB)) vcpu->arch.hcr_el2 |= HCR_FWB; - if (cpus_have_final_cap(ARM64_HAS_EVT) && - !cpus_have_final_cap(ARM64_MISMATCHED_CACHE_TYPE) && - kvm_read_vm_id_reg(vcpu->kvm, SYS_CTR_EL0) == read_cpuid(CTR_EL0)) - vcpu->arch.hcr_el2 |= HCR_TID4; - else - vcpu->arch.hcr_el2 |= HCR_TID2; - /* * Without AArch32 EL1, leave RW set and let the entry fail with an * illegal exception return: the *32_EL2 registers EL2 would otherwise @@ -65,9 +59,6 @@ static void pkvm_vcpu_reset_hcr(struct kvm_vcpu *vcpu) if (vcpu_has_ptrauth(vcpu)) vcpu->arch.hcr_el2 |= (HCR_API | HCR_APK); - - if (kvm_has_mte(vcpu->kvm)) - vcpu->arch.hcr_el2 |= HCR_ATA; } static void pvm_init_traps_hcr(struct kvm_vcpu *vcpu) @@ -85,6 +76,13 @@ static void pvm_init_traps_hcr(struct kvm_vcpu *vcpu) */ val |= HCR_TACR | HCR_TIDCP | HCR_TID3 | HCR_TID1; + if (cpus_have_final_cap(ARM64_HAS_EVT) && + !cpus_have_final_cap(ARM64_MISMATCHED_CACHE_TYPE) && + kvm_read_vm_id_reg(kvm, SYS_CTR_EL0) == read_cpuid(CTR_EL0)) + val |= HCR_EL2_TID4; + else + val |= HCR_EL2_TID2; + if (!kvm_has_feat(kvm, ID_AA64PFR0_EL1, RAS, IMP)) { val |= HCR_TERR | HCR_TEA; val &= ~(HCR_FIEN); From afb4334fb52c7b416bbc9c4b166f6806509f734b Mon Sep 17 00:00:00 2001 From: Fuad Tabba Date: Tue, 29 Sep 2026 10:00:31 +0100 Subject: [PATCH 646/694] KVM: arm64: selftests: Check a feature hidden in an ID register is UNDEF Userspace can hide a feature from a guest by clearing its field in a writable ID register, and KVM then makes the feature's instructions UNDEFINED in the guest by trapping or disabling them. No selftest checks that. Add a test that runs the instruction of each of TLBI OS, MOPS, TCR2_EL1 and FPMR once with its field as advertised and once with it cleared, and expects an UNDEF only when cleared. A feature the vCPU doesn't advertise is skipped, as is hidden TLBI OS on a CPU with neither FGT nor FEAT_EVT2, where KVM can't trap it. Signed-off-by: Fuad Tabba Link: https://patch.msgid.link/20260929090031.3829185-5-fuad.tabba@linux.dev Signed-off-by: Oliver Upton --- tools/testing/selftests/kvm/Makefile.kvm | 1 + .../selftests/kvm/arm64/hidden_features.c | 184 ++++++++++++++++++ 2 files changed, 185 insertions(+) create mode 100644 tools/testing/selftests/kvm/arm64/hidden_features.c diff --git a/tools/testing/selftests/kvm/Makefile.kvm b/tools/testing/selftests/kvm/Makefile.kvm index 6a1482e3a286..51b98f47044f 100644 --- a/tools/testing/selftests/kvm/Makefile.kvm +++ b/tools/testing/selftests/kvm/Makefile.kvm @@ -195,6 +195,7 @@ TEST_GEN_PROGS_arm64 += arm64/vgic_v5 TEST_GEN_PROGS_arm64 += arm64/vpmu_counter_access TEST_GEN_PROGS_arm64 += arm64/no-vgic TEST_GEN_PROGS_arm64 += arm64/idreg-idst +TEST_GEN_PROGS_arm64 += arm64/hidden_features TEST_GEN_PROGS_arm64 += arm64/kvm-uuid TEST_GEN_PROGS_arm64 += access_tracking_perf_test TEST_GEN_PROGS_arm64 += arch_timer diff --git a/tools/testing/selftests/kvm/arm64/hidden_features.c b/tools/testing/selftests/kvm/arm64/hidden_features.c new file mode 100644 index 000000000000..194d746e7605 --- /dev/null +++ b/tools/testing/selftests/kvm/arm64/hidden_features.c @@ -0,0 +1,184 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * hidden_features - Check that a feature's instruction runs in the guest when + * its ID register field is advertised, and is UNDEFINED when userspace clears + * the field. + * + * Copyright (c) 2026 Google LLC + * Author: Fuad Tabba + */ +#include "kvm_util.h" +#include "processor.h" +#include "test_util.h" + +static volatile bool undef; + +static void guest_tlbi_os(void) +{ + /* tlbi vmalle1os */ + asm volatile("sys #0, c8, c1, #0\n\tdsb ish\n\tisb" ::: "memory"); +} + +static void guest_mops(void) +{ + register u64 *d asm("x0"); + register u64 n asm("x1"); + register u64 s asm("x2"); + u64 buf[8]; + + d = buf; + n = sizeof(buf); + s = 0; + /* setp [x0]!, x1!, x2; setm; sete */ + asm volatile(".inst 0x19c20420\n\t.inst 0x19c24420\n\t.inst 0x19c28420" + : "+r"(d), "+r"(n) : "r"(s) : "cc", "memory"); +} + +static void guest_tcr2(void) +{ + read_sysreg_s(SYS_TCR2_EL1); +} + +static void guest_fpmr(void) +{ + read_sysreg_s(SYS_FPMR); +} + +struct feature { + const char *name; + u64 id_reg; + u64 mask; + u8 shift; + u64 min; + void (*insn)(void); + bool (*trappable)(struct kvm_vcpu *vcpu); +}; + +/* Without FGT, KVM traps a hidden TLBI OS only through HCR_EL2.TTLBOS (FEAT_EVT2). */ +static bool tlbi_os_trappable(struct kvm_vcpu *vcpu) +{ + u64 mmfr0 = vcpu_get_reg(vcpu, KVM_ARM64_SYS_REG(SYS_ID_AA64MMFR0_EL1)); + u64 mmfr2 = vcpu_get_reg(vcpu, KVM_ARM64_SYS_REG(SYS_ID_AA64MMFR2_EL1)); + + return SYS_FIELD_GET(ID_AA64MMFR0_EL1, FGT, mmfr0) >= ID_AA64MMFR0_EL1_FGT_IMP || + SYS_FIELD_GET(ID_AA64MMFR2_EL1, EVT, mmfr2) >= ID_AA64MMFR2_EL1_EVT_TTLBxS; +} + +#define FEATURE(n, reg, field, min_val, fn, trap) \ +{ \ + .name = n, \ + .id_reg = SYS_##reg, \ + .mask = reg##_##field##_MASK, \ + .shift = reg##_##field##_SHIFT, \ + .min = reg##_##field##_##min_val, \ + .insn = fn, \ + .trappable = trap, \ +} + +static const struct feature features[] = { + FEATURE("TLBI OS", ID_AA64ISAR0_EL1, TLB, OS, guest_tlbi_os, tlbi_os_trappable), + FEATURE("MOPS", ID_AA64ISAR2_EL1, MOPS, IMP, guest_mops, NULL), + FEATURE("TCR2_EL1", ID_AA64MMFR3_EL1, TCRX, IMP, guest_tcr2, NULL), + FEATURE("FPMR", ID_AA64PFR2_EL1, FPMR, IMP, guest_fpmr, NULL), +}; + +static void guest_code(const struct feature *feat) +{ + undef = false; + feat->insn(); + GUEST_SYNC(undef); + GUEST_DONE(); +} + +static void guest_undef_handler(struct ex_regs *regs) +{ + undef = true; + regs->pc += 4; +} + +static bool run(const struct feature *feat, bool hide) +{ + struct kvm_vcpu *vcpu; + struct kvm_vm *vm; + struct ucall uc; + bool got = false; + u64 val; + + vm = vm_create_with_one_vcpu(&vcpu, (void *)guest_code); + vm_init_descriptor_tables(vm); + vcpu_init_descriptor_tables(vcpu); + vm_install_sync_handler(vm, VECTOR_SYNC_CURRENT, ESR_ELx_EC_UNKNOWN, guest_undef_handler); + vcpu_args_set(vcpu, 1, feat); + + if (hide) { + val = vcpu_get_reg(vcpu, KVM_ARM64_SYS_REG(feat->id_reg)); + vcpu_set_reg(vcpu, KVM_ARM64_SYS_REG(feat->id_reg), val & ~feat->mask); + } + + for (;;) { + vcpu_run(vcpu); + switch (get_ucall(vcpu, &uc)) { + case UCALL_SYNC: + got = uc.args[1]; + break; + case UCALL_ABORT: + REPORT_GUEST_ASSERT(uc); + break; + case UCALL_DONE: + kvm_vm_free(vm); + return got; + default: + TEST_FAIL("Unknown ucall %lu", uc.cmd); + } + } +} + +static void probe_feature(const struct feature *feat, bool *present, bool *trappable) +{ + struct kvm_vcpu *vcpu; + struct kvm_vm *vm; + u64 val; + + vm = vm_create_with_one_vcpu(&vcpu, NULL); + val = vcpu_get_reg(vcpu, KVM_ARM64_SYS_REG(feat->id_reg)); + *present = ((val & feat->mask) >> feat->shift) >= feat->min; + *trappable = !feat->trappable || feat->trappable(vcpu); + kvm_vm_free(vm); +} + +int main(void) +{ + const struct feature *feat; + bool present, trappable; + int i; + + test_disable_default_vgic(); + + ksft_print_header(); + ksft_set_plan(ARRAY_SIZE(features) * 2); + + for (i = 0; i < ARRAY_SIZE(features); i++) { + feat = &features[i]; + + probe_feature(feat, &present, &trappable); + if (!present) { + ksft_test_result_skip("%s advertised, not supported\n", feat->name); + ksft_test_result_skip("%s hidden, not supported\n", feat->name); + continue; + } + + if (run(feat, false)) + ksft_test_result_fail("%s advertised, UNDEF\n", feat->name); + else + ksft_test_result_pass("%s advertised\n", feat->name); + + if (!trappable) + ksft_test_result_skip("%s hidden, not trappable\n", feat->name); + else if (run(feat, true)) + ksft_test_result_pass("%s hidden\n", feat->name); + else + ksft_test_result_fail("%s hidden, no UNDEF\n", feat->name); + } + + ksft_finished(); +} From 1a70f0b71826fb764e8490dda55bb1b853e0d1de Mon Sep 17 00:00:00 2001 From: Jiangshan Yi Date: Tue, 1 Sep 2026 17:30:49 +0300 Subject: [PATCH 647/694] tpm: Fix heap buffer overflow in tpm_transmit_cmd() struct tpm_buf now uses an embedded flexible array for data, which is 6 bytes shorter than TPM_BUFSIZE due to the struct header. But tpm_transmit_cmd() still passes PAGE_SIZE to tpm_transmit(), and after clamping to TPM_BUFSIZE in tpm_try_transmit(), chip->ops->recv() can write past the end of data[] by 6 bytes. Pass buf->capacity to tpm_transmit() instead. Fixes: 3d9e043dab0a ("tpm-buf: Memory-safe allocations") Signed-off-by: Jiangshan Yi Link: https://lore.kernel.org/r/20260901062850.379870-1-yijiangshan@kylinos.cn Reviewed-by: Jarkko Sakkinen Signed-off-by: Jarkko Sakkinen --- drivers/char/tpm/tpm-interface.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/char/tpm/tpm-interface.c b/drivers/char/tpm/tpm-interface.c index f745a098908b..1ccdbde98b69 100644 --- a/drivers/char/tpm/tpm-interface.c +++ b/drivers/char/tpm/tpm-interface.c @@ -268,7 +268,7 @@ ssize_t tpm_transmit_cmd(struct tpm_chip *chip, struct tpm_buf *buf, int err; ssize_t len; - len = tpm_transmit(chip, buf->data, PAGE_SIZE); + len = tpm_transmit(chip, buf->data, buf->capacity); if (len < 0) return len; From 1e2f9a611b334ce6ed1dc4a9e1587c7a11d2be30 Mon Sep 17 00:00:00 2001 From: Jiangshan Yi Date: Wed, 2 Sep 2026 15:48:39 +0800 Subject: [PATCH 648/694] tpm: Fix auth session leak in tpm2_get_random() error path When tpm_buf_fill_hmac_session() fails in tpm2_get_random(), the error path returns without calling tpm2_end_auth_session(), leaking the auth session. All other error paths in the function handle this correctly. Fixes: 3d9e043dab0a ("tpm-buf: Memory-safe allocations") Cc: stable@vger.kernel.orga # v6.12+ Signed-off-by: Jiangshan Yi Link: https://lore.kernel.org/r/20260902074839.417419-1-yijiangshan@kylinos.cn Reviewed-by: Jarkko Sakkinen Signed-off-by: Jarkko Sakkinen --- drivers/char/tpm/tpm2-cmd.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/char/tpm/tpm2-cmd.c b/drivers/char/tpm/tpm2-cmd.c index 48cec39995fe..96e0e06bbbf0 100644 --- a/drivers/char/tpm/tpm2-cmd.c +++ b/drivers/char/tpm/tpm2-cmd.c @@ -281,8 +281,10 @@ int tpm2_get_random(struct tpm_chip *chip, u8 *dest, size_t max) } tpm_buf_append_u16(buf, num_bytes); err = tpm_buf_fill_hmac_session(chip, buf); - if (err) + if (err) { + tpm2_end_auth_session(chip); return err; + } err = tpm_transmit_cmd(chip, buf, offsetof(struct tpm2_get_random_out, From 9b522400bf5c082feb9a0037a053ea822a2c7e4d Mon Sep 17 00:00:00 2001 From: Jiangshan Yi Date: Thu, 3 Sep 2026 11:58:37 +0800 Subject: [PATCH 649/694] tpm: fix off-by-four bounds check in tpm2_get_random() When the response carries the TPM2_ST_SESSIONS tag, tpm2_get_random() skips the 4-byte parameter size field before locating the random data, but the bounds check still validates the response length against TPM_HEADER_SIZE. A truncated response can pass the check and make memcpy() read up to 4 bytes past the response end, so stale buffer contents end up in the caller's random bytes. Fix this by checking the response length against 'offset', which already includes the skipped parameter size field. Cc: stable@vger.kernel.org # v6.12+ Fixes: 1b6d7f9eb150 ("tpm: add session encryption protection to tpm2_get_random()") Reported-by: Sashiko Closes: https://sashiko.dev/#/patchset/20260902074839.417419-1-yijiangshan%40kylinos.cn Signed-off-by: Jiangshan Yi Reviewed-by: Jarkko Sakkinen Link: https://lore.kernel.org/r/20260903035837.219284-1-yijiangshan@kylinos.cn Signed-off-by: Jarkko Sakkinen --- drivers/char/tpm/tpm2-cmd.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/char/tpm/tpm2-cmd.c b/drivers/char/tpm/tpm2-cmd.c index 96e0e06bbbf0..ae22295df798 100644 --- a/drivers/char/tpm/tpm2-cmd.c +++ b/drivers/char/tpm/tpm2-cmd.c @@ -307,7 +307,7 @@ int tpm2_get_random(struct tpm_chip *chip, u8 *dest, size_t max) out = (struct tpm2_get_random_out *)&buf->data[offset]; recd = min_t(u32, be16_to_cpu(out->size), num_bytes); if (tpm_buf_length(buf) < - TPM_HEADER_SIZE + + offset + offsetof(struct tpm2_get_random_out, buffer) + recd) { tpm2_end_auth_session(chip); From 015fb29a748342a186f37b602ac70017098c8251 Mon Sep 17 00:00:00 2001 From: Surendran Kanagaraj Date: Fri, 18 Sep 2026 00:07:31 +0000 Subject: [PATCH 650/694] tpm: Disable TPM on null key name mismatch The null key name check exists to protect against TPM reset attacks, so a mismatch should stop the device from serving further requests. Currently it does not disable the chip when it finds a mismatch. The mismatch is logged: tpm tpm0: null key integrity check failed but the chip keeps serving commands: / # tpm2_getcap -c properties-fixed TPM_PT_FAMILY_INDICATOR: as UINT32: 0x08322e3000 as string: "2.0" ... tpm2_load_null() where the null key name check is run sets the chip as disabled only if the rc is non zero. When the mismatch is seen, rc is zero at that point and it returns success. The other issue is that the caller expects the null key handle to be populated when the function returns 0 which it does here without writing the handle and proceeds assuming the null key handle is valid. During the test, I noticed that tpm2_start_auth_session() uses the uninitialized stack value as the key handle since tpm2_load_null() returns 0 despite the integrity failure and proceeds with TPM2_CC_START_AUTH_SESS with this value as salt key handle. Set rc to -ENODEV on the mismatch. The error path then disables the chip and returns the correct code to the caller. Tested in QEMU with swtpm and CONFIG_TCG_TPM2_HMAC=y by making TPM2_CC_CONTEXT_LOAD fail with TPM2_RC_INTEGRITY and changing the name of the re-created null key. The chip is now disabled on the mismatch. Fixes: cc7d8594342a ("tpm: Rollback tpm2_load_null()") Fixes: 423893fcbe7e ("tpm: Disable TPM on tpm2_create_primary() failure") Cc: stable@vger.kernel.org # v6.12+ Assisted-by: LLM Signed-off-by: Surendran Kanagaraj Link: https://lore.kernel.org/r/20260918000731.48657-1-surenkj@amazon.com Reviewed-by: Gunnar Kudrjavets Reviewed-by: Jarkko Sakkinen Signed-off-by: Jarkko Sakkinen --- drivers/char/tpm/tpm2-sessions.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/char/tpm/tpm2-sessions.c b/drivers/char/tpm/tpm2-sessions.c index cf8f1fd6790b..ca1e2bf424e1 100644 --- a/drivers/char/tpm/tpm2-sessions.c +++ b/drivers/char/tpm/tpm2-sessions.c @@ -975,6 +975,7 @@ static int tpm2_load_null(struct tpm_chip *chip, u32 *null_key) /* Deduce from the name change TPM interference: */ dev_err(&chip->dev, "null key integrity check failed\n"); tpm2_flush_context(chip, tmp_null_key); + rc = -ENODEV; err: if (rc) { From 226154ea87952ad7141ce2dcf52a6c2edd71b8fe Mon Sep 17 00:00:00 2001 From: Mohamad Raizudeen Date: Sat, 26 Sep 2026 20:29:03 +0530 Subject: [PATCH 651/694] crypto: x86/aes-gcm - fix always true check for last AAD segment In gcm_process_assoc(), a segment that is not the last one must have its length rounded down to a multiple of 16 bytes, as required by the assembly. The check for a non-last segment is `if (unlikely(assoclen)) /* Not the last segment yet? */` where assoclen is the number of AAD bytes remaining after the current segment. Since the conversion to the new scatterwalk API, assoclen is decremented at the end of the loop body rather than the beginning, so it still includes the current segment when the check executes, making the check always true. As a result the last segment was rounded down as well, causing some avoidable extra work: an additional memcpy into the temporary buffer and an additional call into the assembly after the loop. The GCM authentication tag is unaffected either way, so the self-tests pass and this went unnoticed. It is purely an efficiency issue rather than a correctness one. Fix this by moving the assoclen decrement back to the beginning of the loop body. Fixes: e9787deff49e ("crypto: x86/aes-gcm - use the new scatterwalk functions") Cc: stable@vger.kernel.org Suggested-by: Eric Biggers Signed-off-by: Mohamad Raizudeen Link: https://patch.msgid.link/20260926145903.6061-1-raizudeen.kerneldev@gmail.com Signed-off-by: Eric Biggers --- arch/x86/crypto/aesni-intel_glue.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/x86/crypto/aesni-intel_glue.c b/arch/x86/crypto/aesni-intel_glue.c index f522fff9231e..1903ed8bbe3b 100644 --- a/arch/x86/crypto/aesni-intel_glue.c +++ b/arch/x86/crypto/aesni-intel_glue.c @@ -1293,6 +1293,7 @@ static void gcm_process_assoc(const struct aes_gcm_key *key, u8 ghash_acc[16], unsigned int len; const u8 *src = walk.addr; + assoclen -= orig_len_this_step; if (unlikely(pos)) { len = min(len_this_step, 16 - pos); memcpy(&buf[pos], src, len); @@ -1320,7 +1321,6 @@ static void gcm_process_assoc(const struct aes_gcm_key *key, u8 ghash_acc[16], kernel_fpu_end(); kernel_fpu_begin(); } - assoclen -= orig_len_this_step; } if (unlikely(pos)) aes_gcm_aad_update(key, ghash_acc, buf, pos, flags); From 2cf81fc9505b148b1b70068303ea6ea4b1212403 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ilkka=20Lappetel=C3=A4inen?= Date: Fri, 25 Sep 2026 15:48:29 +0300 Subject: [PATCH 652/694] net/packet: prevent TX_RING byte count overflow MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit tpacket_snd() drains every SEND_REQUEST frame in a TX ring and adds each valid packet length to the signed int len_sum. Userspace can recycle completed frames concurrently, so a single blocking send call is not bounded by the ring size. After enough successful transmissions len_sum wraps into the negative range. In particular, 65536 65535-byte frames followed by a 65007-byte frame produce 0xfffffdef, or -EIOCBQUEUED. sock_sendmsg_nosec() treats that as an impossible return and triggers a BUG. On systems configured to panic on oops, a CAP_NET_RAW holder can panic the host kernel. Stop before the next frame would make the byte count exceed INT_MAX. The positive short count leaves that frame pending for a later send call. A two-thread TPACKET_V2 reproducer triggered the BUG on an arm64 build of Linux 7.3-rc4. With this change, the same reproducer returned the expected positive short count without an oops. Fixes: 69e3c75f4d54 ("net: TX_RING and packet mmap") Cc: stable@vger.kernel.org Signed-off-by: Ilkka Lappeteläinen Reviewed-by: Willem de Bruijn Link: https://patch.msgid.link/20260925124829.6595-1-ilkka.lappetelainen@gmail.com Signed-off-by: Jakub Kicinski --- net/packet/af_packet.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c index 7c83e01526ed..8865c1a42667 100644 --- a/net/packet/af_packet.c +++ b/net/packet/af_packet.c @@ -2874,6 +2874,9 @@ static int tpacket_snd(struct packet_sock *po, struct msghdr *msg) vnet_hdr.hdr_len); has_vnet_hdr = true; } + /* Prevent the byte count from wrapping into a negative errno */ + if (unlikely(tp_len > INT_MAX - len_sum)) + break; copylen = max_t(int, copylen, hard_header_len); skb = sock_alloc_send_skb(&po->sk, hlen + tlen + sizeof(struct sockaddr_ll) + From febec6a7076499d16b37f556724a942633a270a6 Mon Sep 17 00:00:00 2001 From: Alireza Asgari Date: Thu, 24 Sep 2026 13:00:24 +0000 Subject: [PATCH 653/694] net: fix checksum offsets in skb_splice_from_iter() skb_splice_from_iter() appends multiple page fragments before updating skb->len. However, skb_splice_csum_page() uses that unchanged length as the offset for every fragment checksum. If bytes already appended during the call have an odd total length, a later fragment's checksum is combined with the wrong parity. For example, prime a UDP socket with sendto(MSG_MORE) and splice two distinct pipe buffers containing "abc" and "DEFGH". Uncorking reports success, but the receiver discards the packet for a bad checksum. This also affects IPv6 and fragments beginning near a page boundary. Pass the initial skb length plus the bytes already spliced to the checksum helper. Keep the existing final length update and partial-progress error handling unchanged. CHECKSUM_PARTIAL does not use this helper and remains unaffected. Fixes: 2e910b95329c ("net: Add a function to splice pages into an skbuff for MSG_SPLICE_PAGES") Cc: stable@vger.kernel.org Signed-off-by: Alireza Asgari Link: https://patch.msgid.link/20260924-fix-udp-splice-checksum-v1-1-fe61d65447a8@asgari.net Signed-off-by: Jakub Kicinski --- net/core/skbuff.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/net/core/skbuff.c b/net/core/skbuff.c index c3042d822afa..4aea06d5167d 100644 --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@ -7409,7 +7409,8 @@ nodefer: kfree_skb_napi_cache(skb); } static void skb_splice_csum_page(struct sk_buff *skb, struct page *page, - size_t offset, size_t len) + size_t offset, size_t len, + unsigned int csum_offset) { const char *kaddr; __wsum csum; @@ -7417,7 +7418,7 @@ static void skb_splice_csum_page(struct sk_buff *skb, struct page *page, kaddr = kmap_local_page(page); csum = csum_partial(kaddr + offset, len, 0); kunmap_local(kaddr); - skb->csum = csum_block_add(skb->csum, csum, skb->len); + skb->csum = csum_block_add(skb->csum, csum, csum_offset); } /** @@ -7477,7 +7478,8 @@ ssize_t skb_splice_from_iter(struct sk_buff *skb, struct iov_iter *iter, } if (skb->ip_summed == CHECKSUM_NONE) - skb_splice_csum_page(skb, page, off, part); + skb_splice_csum_page(skb, page, off, part, + skb->len + spliced); off = 0; spliced += part; From 1221246d3bd08d8b5228777e1a9b3361ca946e64 Mon Sep 17 00:00:00 2001 From: Alireza Asgari Date: Thu, 24 Sep 2026 13:00:25 +0000 Subject: [PATCH 654/694] selftests: net: cover UDP splice checksum fragment alignment Exercise software UDP checksums when splice() transfers multiple pipe fragments into a corked datagram. Use separate source pipes so short writes cannot merge, and use odd fragment lengths to expose checksum-position errors between fragments. Cover IPv4, IPv6 and IPv4-mapped destinations, connected and unconnected sockets, resident prefixes, even and odd fragment lengths, and a fragment starting near a page boundary. Ordinary sends provide a control. Check the complete payload and reject extra datagrams after a successful send. Keep receive waits bounded and skip unavailable address families or pipe capacities. The test uses only loopback sockets with ephemeral ports and has no dependency on the application that exposed the regression. Signed-off-by: Alireza Asgari Link: https://patch.msgid.link/20260924-fix-udp-splice-checksum-v1-2-fe61d65447a8@asgari.net Signed-off-by: Jakub Kicinski --- tools/testing/selftests/net/.gitignore | 1 + tools/testing/selftests/net/Makefile | 1 + .../selftests/net/udp_splice_checksum.c | 378 ++++++++++++++++++ 3 files changed, 380 insertions(+) create mode 100644 tools/testing/selftests/net/udp_splice_checksum.c diff --git a/tools/testing/selftests/net/.gitignore b/tools/testing/selftests/net/.gitignore index c9f46031ac73..dacd36ed8455 100644 --- a/tools/testing/selftests/net/.gitignore +++ b/tools/testing/selftests/net/.gitignore @@ -55,6 +55,7 @@ tools tun txring_overwrite txtimestamp +udp_splice_checksum udpgso udpgso_bench_rx udpgso_bench_tx diff --git a/tools/testing/selftests/net/Makefile b/tools/testing/selftests/net/Makefile index 3ee3378f8b26..590b33e16d9f 100644 --- a/tools/testing/selftests/net/Makefile +++ b/tools/testing/selftests/net/Makefile @@ -199,6 +199,7 @@ TEST_GEN_PROGS := \ tap \ tcp_port_share \ tls \ + udp_splice_checksum \ # end of TEST_GEN_PROGS TEST_FILES := \ diff --git a/tools/testing/selftests/net/udp_splice_checksum.c b/tools/testing/selftests/net/udp_splice_checksum.c new file mode 100644 index 000000000000..dc61444e6481 --- /dev/null +++ b/tools/testing/selftests/net/udp_splice_checksum.c @@ -0,0 +1,378 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Exercise UDP software checksums over multiple spliced pipe fragments. + * All traffic stays on loopback; no privileges or external services are needed. + * Exit status: 0 = pass, 1 = failure, 4 = all cases skipped. + */ +#define _GNU_SOURCE + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include "kselftest.h" + +#define TIMEOUT_MS 200 + +struct test_case { + const char *name; + const char *body; + const char *prefix; + size_t fragments[3]; + bool ordinary; + bool page_tail; +}; + +static const struct test_case cases[] = { + { "ordinary-control", "abcDEFGH", "", { 3, 5 }, true, false }, + { "uniform-odd-fragments", "BBBBBBBB", "", { 3, 5 }, false, false }, + { "even-fragments", "abcdEFGH", "", { 4, 4 }, false, false }, + { "odd-fragments", "abcDEFGH", "", { 3, 5 }, false, false }, + { "odd-prefix-odd-fragments", "abcDEFGH", "P", { 3, 5 }, false, false }, + { "even-prefix-odd-fragments", "abcDEFGH", "PQ", { 3, 5 }, false, false }, + { "page-tail-odd-fragments", "abcDEFGH", "", { 3, 5 }, false, true }, + { "three-fragments", "abcDEFGhijkl", "", { 3, 4, 5 }, false, false }, + { "odd-prefix-three-fragments", + "abcDEFGhijkl", + "P", + { 3, 4, 5 }, + false, + false }, +}; + +static const char *const families[] = { "IPv4", "IPv6", "IPv4-mapped" }; + +static int error(const char *what) +{ + ksft_perror(what); + return 1; +} + +static bool unavailable(void) +{ + return errno == EAFNOSUPPORT || errno == EPROTONOSUPPORT || + errno == EADDRNOTAVAIL; +} + +static long long now_ms(void) +{ + struct timespec ts; + + if (clock_gettime(CLOCK_MONOTONIC, &ts)) + ksft_exit_fail_perror("clock_gettime"); + return (long long)ts.tv_sec * 1000 + ts.tv_nsec / 1000000; +} + +static int receive_one(int fd, const char *expected, size_t length) +{ + struct pollfd pfd = { .fd = fd, .events = POLLIN }; + long long deadline = now_ms() + TIMEOUT_MS; + int remaining, ret; + char received[64]; + ssize_t count; + + for (;;) { + remaining = deadline - now_ms(); + if (remaining < 0) + remaining = 0; + ret = poll(&pfd, 1, remaining); + if (ret < 0) { + if (errno == EINTR) + continue; + return error("poll"); + } + if (!ret) { + ksft_print_msg("no datagram within %d ms after send\n", + TIMEOUT_MS); + return 1; + } + count = recv(fd, received, sizeof(received), + MSG_DONTWAIT | MSG_TRUNC); + if (count < 0) { + if (errno == EINTR || errno == EAGAIN) + continue; + return error("recv"); + } + break; + } + if (count != (ssize_t)length || memcmp(received, expected, length)) { + ksft_print_msg("received %zd bytes; expected %zu matching\n", + count, length); + return 1; + } + count = recv(fd, received, sizeof(received), MSG_DONTWAIT | MSG_TRUNC); + if (count >= 0) { + ksft_print_msg("unexpected extra datagram (%zd bytes)\n", + count); + return 1; + } + if (errno != EAGAIN) + return error("extra-datagram check"); + return 0; +} + +/* Separate source pipes prevent adjacent small writes merging into one buffer. */ +static int append_fragment(int dest, const char *data, size_t length) +{ + int source[2], ret = 1; + + if (pipe2(source, O_NONBLOCK | O_CLOEXEC)) + return error("source pipe"); + if (write(source[1], data, length) != (ssize_t)length) { + error("source write"); + goto out; + } + if (splice(source[0], NULL, dest, NULL, length, SPLICE_F_NONBLOCK) != + (ssize_t)length) { + error("pipe-to-pipe splice"); + goto out; + } + ret = 0; +out: + close(source[0]); + close(source[1]); + return ret; +} + +static int pipe_capacity(int fd, unsigned int slots, const char **skip_reason) +{ + long required = slots * sysconf(_SC_PAGESIZE); + int capacity = fcntl(fd, F_GETPIPE_SZ); + + if (capacity < 0) + return error("query pipe capacity"); + if (capacity >= required || + fcntl(fd, F_SETPIPE_SZ, required) >= required) + return 0; + ksft_print_msg("cannot reserve %u pipe slots: %s\n", slots, + strerror(errno)); + *skip_reason = "required pipe capacity unavailable"; + return KSFT_SKIP; +} + +/* Leave three bytes at a page's end, followed by five bytes on the next page. */ +static int append_page_tail(int dest, const char *body, + const char **skip_reason) +{ + long page = sysconf(_SC_PAGESIZE); + int source[2], ret = 1; + char *data; + + if (page < 8) { + ksft_print_msg("invalid page size %ld\n", page); + return 1; + } + data = malloc(page + 5); + if (!data) + return error("malloc"); + memset(data, 'x', page - 3); + memcpy(data + page - 3, body, 8); + if (pipe2(source, O_NONBLOCK | O_CLOEXEC)) { + error("page-tail source pipe"); + goto free_data; + } + ret = pipe_capacity(source[1], 2, skip_reason); + if (ret) + goto out; + ret = 1; + if (write(source[1], data, page + 5) != page + 5) { + error("page-tail write"); + goto out; + } + if (read(source[0], data, page - 3) != page - 3) { + error("page-tail discard"); + goto out; + } + if (splice(source[0], NULL, dest, NULL, 8, SPLICE_F_NONBLOCK) != 8) { + error("page-tail pipe-to-pipe splice"); + goto out; + } + ret = 0; +out: + close(source[0]); + close(source[1]); +free_data: + free(data); + return ret; +} + +static int run_case(unsigned int family, bool connected, + const struct test_case *test, const char **skip_reason) +{ + size_t body_length = strlen(test->body), offset = 0; + int tx_family = family == 0 ? AF_INET : AF_INET6; + int rx_family = family == 1 ? AF_INET6 : AF_INET; + int rx = -1, tx = -1, pipefd[2] = { -1, -1 }; + size_t prefix_length = strlen(test->prefix); + struct sockaddr_storage address = { 0 }; + socklen_t address_length; + struct sockaddr_in6 *v6; + struct sockaddr_in *v4; + int ret = 1, zero = 0; + char expected[64]; + unsigned int i; + ssize_t count; + + v4 = (struct sockaddr_in *)&address; + v6 = (struct sockaddr_in6 *)&address; + *skip_reason = "address family unavailable"; + memcpy(expected, test->prefix, prefix_length); + memcpy(expected + prefix_length, test->body, body_length); + rx = socket(rx_family, SOCK_DGRAM | SOCK_CLOEXEC, 0); + if (rx < 0) { + ret = unavailable() ? KSFT_SKIP : error("receive socket"); + goto out; + } + if (rx_family == AF_INET) { + v4->sin_family = AF_INET; + v4->sin_addr.s_addr = htonl(INADDR_LOOPBACK); + address_length = sizeof(*v4); + } else { + v6->sin6_family = AF_INET6; + v6->sin6_addr = in6addr_loopback; + address_length = sizeof(*v6); + } + if (bind(rx, (struct sockaddr *)&address, address_length)) { + ret = unavailable() ? KSFT_SKIP : error("bind loopback"); + goto out; + } + if (getsockname(rx, (struct sockaddr *)&address, &address_length)) { + error("getsockname"); + goto out; + } + if (family == 2) { + unsigned short port = v4->sin_port; + + memset(&address, 0, sizeof(address)); + v6->sin6_family = AF_INET6; + v6->sin6_port = port; + inet_pton(AF_INET6, "::ffff:127.0.0.1", &v6->sin6_addr); + address_length = sizeof(*v6); + } + tx = socket(tx_family, SOCK_DGRAM | SOCK_CLOEXEC, 0); + if (tx < 0) { + ret = unavailable() ? KSFT_SKIP : error("send socket"); + goto out; + } + if (family == 2 && + setsockopt(tx, IPPROTO_IPV6, IPV6_V6ONLY, &zero, sizeof(zero))) { + error("disable IPV6_V6ONLY"); + goto out; + } + if (connected && + connect(tx, (struct sockaddr *)&address, address_length)) { + error("connect"); + goto out; + } + if (test->ordinary) { + count = sendto(tx, expected, prefix_length + body_length, 0, + (struct sockaddr *)&address, address_length); + if (count != (ssize_t)(prefix_length + body_length)) { + error("ordinary sendto"); + goto out; + } + } else { + if (pipe2(pipefd, O_NONBLOCK | O_CLOEXEC)) { + error("assembled pipe"); + goto out; + } + ret = pipe_capacity(pipefd[1], test->fragments[2] ? 3 : 2, + skip_reason); + if (ret) + goto out; + ret = 1; + if (test->page_tail) { + ret = append_page_tail(pipefd[1], test->body, + skip_reason); + if (ret) + goto out; + ret = 1; + } else { + for (i = 0; i < ARRAY_SIZE(test->fragments); i++) { + if (!test->fragments[i]) + break; + if (append_fragment(pipefd[1], + test->body + offset, + test->fragments[i])) + goto out; + offset += test->fragments[i]; + } + } + /* Even an empty prefix establishes the destination and UDP cork. */ + count = sendto(tx, test->prefix, prefix_length, MSG_MORE, + (struct sockaddr *)&address, address_length); + if (count != (ssize_t)prefix_length) { + error("prime sendto(MSG_MORE)"); + goto out; + } + count = splice(pipefd[0], NULL, tx, NULL, body_length, + SPLICE_F_NONBLOCK | SPLICE_F_MORE); + if (count != (ssize_t)body_length) { + ksft_print_msg("socket splice: %zd, expected %zu: %s\n", + count, body_length, + count < 0 ? strerror(errno) : + "short transfer"); + goto out; + } + if (sendto(tx, "", 0, 0, (struct sockaddr *)&address, + address_length)) { + error("commit sendto"); + goto out; + } + } + ret = receive_one(rx, expected, prefix_length + body_length); +out: + if (pipefd[0] >= 0) { + close(pipefd[0]); + close(pipefd[1]); + } + if (tx >= 0) + close(tx); + if (rx >= 0) + close(rx); + return ret; +} + +int main(void) +{ + unsigned int family, connected, i; + const char *connection, *skip; + int ret; + + ksft_print_header(); + ksft_set_plan(ARRAY_SIZE(families) * 2 * ARRAY_SIZE(cases)); + for (family = 0; family < ARRAY_SIZE(families); family++) { + for (connected = 0; connected < 2; connected++) { + connection = connected ? "connected" : "unconnected"; + for (i = 0; i < ARRAY_SIZE(cases); i++) { + ret = run_case(family, connected, &cases[i], + &skip); + if (ret == KSFT_SKIP) + ksft_test_result_skip("%s %s %s: %s\n", + families[family], + connection, + cases[i].name, + skip); + else + ksft_test_result(!ret, "%s %s %s\n", + families[family], + connection, + cases[i].name); + } + } + } + if (!ksft_get_pass_cnt() && !ksft_get_fail_cnt()) { + /* ksft_exit_skip() would add a result beyond the plan. */ + ksft_print_cnts(); + return KSFT_SKIP; + } + ksft_finished(); +} From 8b9993437429c66d11430ae61ffa5937a6abb4ae Mon Sep 17 00:00:00 2001 From: Yoshihisa Yamamoto Date: Thu, 24 Sep 2026 10:36:57 +0000 Subject: [PATCH 655/694] net: pcs: rzn1-miic: Clear MIIC protection state before unprotect sequence On RZ/T2H and RZ/N1D MIIC, writing 0x0000 to the MIIC protection register clears the protection state machine. The initial protection state cannot be assumed to be constant, as it may be influenced by previous activity before the driver takes ownership of the hardware. Clear the protection state before issuing the unprotect sequence so that register access always starts from a known state, regardless of any previous activity. Fixes: 7dc54d3b8d91 ("net: pcs: add Renesas MII converter driver") Signed-off-by: Yoshihisa Yamamoto Reviewed-by: Lad Prabhakar Link: https://patch.msgid.link/TYCPR01MB74815955A10817B64BD6FB93A4812@TYCPR01MB7481.jpnprd01.prod.outlook.com Signed-off-by: Jakub Kicinski --- drivers/net/pcs/pcs-rzn1-miic.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/net/pcs/pcs-rzn1-miic.c b/drivers/net/pcs/pcs-rzn1-miic.c index cb74861e823c..3aa513467060 100644 --- a/drivers/net/pcs/pcs-rzn1-miic.c +++ b/drivers/net/pcs/pcs-rzn1-miic.c @@ -270,6 +270,9 @@ static struct miic_port *phylink_pcs_to_miic_port(struct phylink_pcs *pcs) static void miic_unlock_regs(struct miic *miic) { + /* Clear protection state */ + writel(0x0000, miic->base + MIIC_PRCMD); + /* Unprotect register writes */ writel(0x00A5, miic->base + MIIC_PRCMD); writel(0x0001, miic->base + MIIC_PRCMD); From 19419faf58dfe5bcc8a9e60e622976c5f8e3ddc1 Mon Sep 17 00:00:00 2001 From: Nicolai Buchwitz Date: Mon, 28 Sep 2026 12:56:41 +0200 Subject: [PATCH 656/694] net: bcmgenet: fix NULL dereference in set_coalesce before first open The Rx ring priv and index are only set in bcmgenet_init_rx_ring(), which runs at open. Setting the Rx coalescing parameters on an interface that was never opened, e.g. "ethtool -C eth0 rx-usecs 50", dereferences a NULL ring->priv. The oops happens with RTNL held, so networking stays stuck until a reboot: Unable to handle kernel paging request at virtual address 0000000000002788 pc : bcmgenet_set_rx_coalesce.isra.0+0x10/0x88 lr : bcmgenet_set_coalesce+0xe8/0x180 Call trace: bcmgenet_set_rx_coalesce.isra.0+0x10/0x88 (P) __ethnl_set_coalesce.isra.0+0x490/0x570 ethnl_set_coalesce+0x48/0xc0 ethnl_default_set_doit+0xf4/0x230 genl_family_rcv_msg_doit+0xe8/0x160 genl_rcv_msg+0x220/0x2a0 netlink_rcv_skb+0x68/0x140 genl_rcv+0x40/0x60 netlink_unicast+0x338/0x3c0 netlink_sendmsg+0x19c/0x3f8 __sock_sendmsg+0x64/0xc0 __sys_sendto+0x124/0x198 __arm64_sys_sendto+0x30/0x48 Set priv and index at probe. Writing the registers while the interface is down is harmless because open resets the MAC and bcmgenet_init_rx_coalesce() reapplies the stored values. It was observed on a Raspberry Pi CM4, but should apply to all other genet devices. Fixes: 9f4ca05827a2 ("net: bcmgenet: Add support for adaptive RX coalescing") Signed-off-by: Nicolai Buchwitz Reviewed-by: Florian Fainelli Link: https://patch.msgid.link/20260928105642.856210-1-nb@tipi-net.de Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/broadcom/genet/bcmgenet.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/broadcom/genet/bcmgenet.c b/drivers/net/ethernet/broadcom/genet/bcmgenet.c index f725d26e6020..e8908916558b 100644 --- a/drivers/net/ethernet/broadcom/genet/bcmgenet.c +++ b/drivers/net/ethernet/broadcom/genet/bcmgenet.c @@ -4151,8 +4151,11 @@ static int bcmgenet_probe(struct platform_device *pdev) netif_set_real_num_rx_queues(priv->dev, priv->hw_params->rx_queues + 1); /* Set default coalescing parameters */ - for (i = 0; i <= priv->hw_params->rx_queues; i++) + for (i = 0; i <= priv->hw_params->rx_queues; i++) { + priv->rx_rings[i].priv = priv; + priv->rx_rings[i].index = i; priv->rx_rings[i].rx_max_coalesced_frames = 1; + } /* Initialize u64 stats seq counter for 32bit machines */ for (i = 0; i <= GENET_MAX_MQ_CNT; i++) { From ea4d4b5dddb5121e64f56fd8e0720bd8b447b63d Mon Sep 17 00:00:00 2001 From: Jamal Hadi Salim Date: Mon, 28 Sep 2026 08:46:16 -0400 Subject: [PATCH 657/694] net: cap skb->queue_mapping when the tx queue is picked skbedit can set skb->queue_mapping and raise the per-CPU skip_txqueue flag so __dev_queue_xmit() honours the mapping. __dev_queue_xmit() cleared the flag before sch_handle_egress() and only read it afterwards, so the flag was not confined to the xmit that set it: a nested xmit (mirred redirect or mirror, or a drop after skbedit) could set the flag and the outer xmit would consume it for an skb that never went through skbedit. A forwarded packet still carries the ingress NIC's rx_queue + 1 in skb->queue_mapping, so the outer device then indexes its tx queue state with that stale value. Taprio's child array q->qdiscs[] is sized to the device's queue count, so taprio_enqueue() indexes past its allocation and dereferences the result as a struct Qdisc *. We (ab)use the skb->nf_skip_egress which means "skip netfilter egress for this packet" to tag to "am I in tc egress?". Despite the overload I dont see it as a conflict since the marker is set only around the single sch_handle_egress() call and ingress path is guarded by tc_at_ingress. I will send a followup(net-next) patch once this hits net-next to rename the skb->nf_skip_egress bit/flag to skb->skip_egress Arm the flag only from the egress classifier that can use it: raise skip_txqueue from tcf_skbedit_act() only when it runs inside sch_handle_egress(), thanks to skb->nf_skip_egress. An egress qdisc classifier runs in q->enqueue(), after the tx queue has been picked, so a mapping it sets cannot affect the current packet; arming the flag there only pollutes it for a later xmit. Then own the flag for the xmit frame the egress hook runs in: save the incoming value and clear it just before sch_handle_egress(), and restore it after the hook - on the consumed (drop) path, or, in the same call that reads it, on the surviving path. The save and the restores stay inside the egress_needed_key static branch, so a packet pays for them only when egress hooks are active (2f1e85b1aee4). Store the value netdev_cap_txqueue() selected back into skb->queue_mapping in netdev_tx_queue_mapping(), as netdev_core_pick_tx() already does, so the skip_txqueue path never hands a later reader on the xmit path a mapping the device cannot serve. A store made still later in the same frame, by a tc BPF program attached to a transmit qdisc, is outside this path and is not re-capped; a separate followup will resolve that path. netdev_xmit_skip_txqueue() returns the previous flag value so the save-and-clear is one call, and a no-op stub is provided when CONFIG_NET_EGRESS is disabled. skb->nf_skip_egress is compiled under CONFIG_NET_EGRESS rather than CONFIG_NETFILTER_SKIP_EGRESS, so skb_at_tc_egress() is valid whenever the egress path is built. A local user in a network namespace can redirect a packet from a device with more TX queues to one with fewer after setting a mapping valid only on the larger device. That reaches these reads and, under KASAN, faults with "slab-out-of-bounds in taprio_enqueue". Conditions to recreate the bug: the report's own trigger is a local user with CAP_NET_ADMIN in a network namespace, so no eBPF program is needed. With CONFIG_NET_SCH_TAPRIO=y, CONFIG_NET_ACT_SKBEDIT=y, CONFIG_NET_ACT_MIRRED=y, CONFIG_NET_CLS_MATCHALL=y, CONFIG_NET_SCH_PRIO=y and KASAN enabled, create qa (3 queues), qb (2 queues) and qc (1 queue) as dummy devices; put a taprio root on qb (num_tc 1, queues 2@0) and clsact on all three; then add an egress matchall filter on every device. On qa: "action skbedit queue_mapping 2 pipe action mirred egress redirect dev qb". On qb: "action mirred egress mirror dev qc". On qc: "action skbedit queue_mapping 0 pipe". Send one packet out qa. qc's skbedit sets the flag while qb's outer xmit is in flight; without the fix qb consumes it and reads its two-entry taprio child array with the forwarded packet's stale mapping. A qc whose skbedit is instead installed in a transmit-qdisc classifier (a matchall filter on the qc root qdisc) reaches the same code path the same way without the fix. Testing: on a KASAN build with panic_on_warn=1 the unfixed kernel panics with "BUG: KASAN: slab-out-of-bounds in taprio_enqueue", a read 0 bytes past a 16-byte taprio_init() allocation, for the clsact-setter and the transmit-qdisc-classifier reproducers and for a clsact skbedit-then-tc-BPF store; the fixed kernel runs all three with no report, and the BPF store variant additionally shows the expected "selects TX queue" clamp notice from the write-back. Fixes: 2f1e85b1aee4 ("net: sched: use queue_mapping to pick tx queue") Reported-by: Zero Day Initiative Link: https://lore.kernel.org/netdev/CANn89iLwYx8nCVf0pCEk_MmEiyC6kQaMwCQT9WkQVeeNzNQHqQ@mail.gmail.com/ Link: https://lore.kernel.org/netdev/179008581937.2160803.7117814290574262942@kernel.org/ Link: https://lore.kernel.org/netdev/179033713973.2160803.4914570693994398206@kernel.org/ Link: https://lore.kernel.org/netdev/20260925180407.63647514@kernel.org/ Link: https://lore.kernel.org/netdev/CANn89i+k-mZKDQVtvws_MEXeuMTAdaCcOXFZE-RfhcGTu90sjA@mail.gmail.com/ Suggested-by: Eric Dumazet Suggested-by: Jakub Kicinski Tested-by: hybris Signed-off-by: Jamal Hadi Salim Reviewed-by: Eric Dumazet Link: https://patch.msgid.link/QDISC-9R8V.v4.20260928081529@mojatatu.com Signed-off-by: Jakub Kicinski --- include/linux/netfilter_netdev.h | 2 +- include/linux/rtnetlink.h | 7 ++++++- include/linux/skbuff.h | 2 +- include/net/sch_generic.h | 9 +++++++++ net/core/dev.c | 31 ++++++++++++++++++++++++------- net/sched/act_skbedit.c | 5 ++--- 6 files changed, 43 insertions(+), 13 deletions(-) diff --git a/include/linux/netfilter_netdev.h b/include/linux/netfilter_netdev.h index 3175073a66ba..2a854a0bbd4b 100644 --- a/include/linux/netfilter_netdev.h +++ b/include/linux/netfilter_netdev.h @@ -133,7 +133,7 @@ static inline struct sk_buff *nf_hook_egress(struct sk_buff *skb, int *rc, static inline void nf_skip_egress(struct sk_buff *skb, bool skip) { -#ifdef CONFIG_NETFILTER_SKIP_EGRESS +#ifdef CONFIG_NET_EGRESS skb->nf_skip_egress = skip; #endif } diff --git a/include/linux/rtnetlink.h b/include/linux/rtnetlink.h index 95729339e7a5..a54ec40d095c 100644 --- a/include/linux/rtnetlink.h +++ b/include/linux/rtnetlink.h @@ -186,7 +186,12 @@ void net_dec_ingress_queue(void); #ifdef CONFIG_NET_EGRESS void net_inc_egress_queue(void); void net_dec_egress_queue(void); -void netdev_xmit_skip_txqueue(bool skip); +bool netdev_xmit_skip_txqueue(bool skip); +#else +static inline bool netdev_xmit_skip_txqueue(bool skip) +{ + return false; +} #endif void rtnetlink_init(void); diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index 84308498a3a8..a3ff380d43f0 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -1020,7 +1020,7 @@ struct sk_buff { #ifdef CONFIG_NET_REDIRECT __u8 from_ingress:1; #endif -#ifdef CONFIG_NETFILTER_SKIP_EGRESS +#ifdef CONFIG_NET_EGRESS __u8 nf_skip_egress:1; #endif #ifdef CONFIG_SKB_DECRYPTED diff --git a/include/net/sch_generic.h b/include/net/sch_generic.h index f35bd06a6bad..1acaadb3e2cf 100644 --- a/include/net/sch_generic.h +++ b/include/net/sch_generic.h @@ -810,6 +810,15 @@ static inline bool skb_at_tc_ingress(const struct sk_buff *skb) #endif } +static inline bool skb_at_tc_egress(const struct sk_buff *skb) +{ +#ifdef CONFIG_NET_EGRESS + return skb->nf_skip_egress && !skb_at_tc_ingress(skb); +#else + return false; +#endif +} + static inline bool skb_skip_tc_classify(struct sk_buff *skb) { #ifdef CONFIG_NET_CLS_ACT diff --git a/net/core/dev.c b/net/core/dev.c index f660fccfc0db..8638c994bbcb 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -4404,9 +4404,14 @@ EXPORT_SYMBOL(dev_loopback_xmit); static struct netdev_queue * netdev_tx_queue_mapping(struct net_device *dev, struct sk_buff *skb) { - int qm = skb_get_queue_mapping(skb); + int queue = skb_get_queue_mapping(skb); + int capped; - return netdev_get_tx_queue(dev, netdev_cap_txqueue(dev, qm)); + capped = netdev_cap_txqueue(dev, queue); + if (unlikely(capped != queue)) + skb_set_queue_mapping(skb, capped); + + return netdev_get_tx_queue(dev, capped); } #ifndef CONFIG_PREEMPT_RT @@ -4415,9 +4420,13 @@ static bool netdev_xmit_txqueue_skipped(void) return __this_cpu_read(softnet_data.xmit.skip_txqueue); } -void netdev_xmit_skip_txqueue(bool skip) +bool netdev_xmit_skip_txqueue(bool skip) { + bool prev = netdev_xmit_txqueue_skipped(); + __this_cpu_write(softnet_data.xmit.skip_txqueue, skip); + + return prev; } EXPORT_SYMBOL_GPL(netdev_xmit_skip_txqueue); @@ -4427,9 +4436,13 @@ static bool netdev_xmit_txqueue_skipped(void) return current->net_xmit.skip_txqueue; } -void netdev_xmit_skip_txqueue(bool skip) +bool netdev_xmit_skip_txqueue(bool skip) { + bool prev = netdev_xmit_txqueue_skipped(); + current->net_xmit.skip_txqueue = skip; + + return prev; } EXPORT_SYMBOL_GPL(netdev_xmit_skip_txqueue); #endif @@ -4848,21 +4861,25 @@ int __dev_queue_xmit(struct sk_buff *skb, struct net_device *sb_dev) tcx_set_ingress(skb, false); #ifdef CONFIG_NET_EGRESS if (static_branch_unlikely(&egress_needed_key)) { + bool skip_txq; + if (nf_hook_egress_active()) { skb = nf_hook_egress(skb, &rc, dev); if (!skb) goto out; } - netdev_xmit_skip_txqueue(false); + skip_txq = netdev_xmit_skip_txqueue(false); nf_skip_egress(skb, true); skb = sch_handle_egress(skb, &rc, dev); - if (!skb) + if (!skb) { + netdev_xmit_skip_txqueue(skip_txq); goto out; + } nf_skip_egress(skb, false); - if (netdev_xmit_txqueue_skipped()) + if (netdev_xmit_skip_txqueue(skip_txq)) txq = netdev_tx_queue_mapping(dev, skb); } #endif diff --git a/net/sched/act_skbedit.c b/net/sched/act_skbedit.c index bfec6b668410..26258b353cde 100644 --- a/net/sched/act_skbedit.c +++ b/net/sched/act_skbedit.c @@ -72,9 +72,8 @@ TC_INDIRECT_SCOPE int tcf_skbedit_act(struct sk_buff *skb, } if (params->flags & SKBEDIT_F_QUEUE_MAPPING && skb->dev->real_num_tx_queues > params->queue_mapping) { -#ifdef CONFIG_NET_EGRESS - netdev_xmit_skip_txqueue(true); -#endif + if (skb_at_tc_egress(skb)) + netdev_xmit_skip_txqueue(true); skb_set_queue_mapping(skb, tcf_skbedit_hash(params, skb)); } if (params->flags & SKBEDIT_F_MARK) { From 5f0764cb1c81a9712bce605d38efe6d844a70b29 Mon Sep 17 00:00:00 2001 From: Willem de Bruijn Date: Sat, 26 Sep 2026 10:04:54 -0400 Subject: [PATCH 658/694] net: extend IPv6 exthdr detection of tunneled packets Commit c4336a07eb6b ("net: correctly handle tunneled traffic on IPV6_CSUM GSO fallback") split skb_gso_has_extension_hdr() into mutually exclusive branches on skb->encapsulation. This did not yet address all paths: 1. With skb->encapsulation set, the outer header is not checked. IPv6 tunnels such as ip6_gre and ip6_tunnel add an outer Destination Options header by default (encap_limit). Their GSO packets skip software GSO, then skb_csum_hwoffload_help() sees the outer extension header and calls skb_checksum_help() on the GSO skb, which warns and drops it. 2. Tunnels over IPv6 without an outer transport header, such as ip6_tunnel, leave skb->transport_header at the inner transport header. skb_network_header_len() then spans the outer IPv6, tunnel and inner IP headers, a false positive. 3. UDP tunnels without an inner network header, such as SCTP-in-UDP or PSP, have no inner IPv6 header to check. Decide on the outer header alone. 4. Directly dereferencing inner_ip_hdr(skb)->version without skb_header_pointer() is unsafe. Instead, check ipv6_ext_hdr(nexthdr) on the outer IPv6 header and, if set, on the inner IPv6 header. Read the headers with skb_header_pointer(). Keep the skb_network_header_len() check when !skb->encapsulation to also catch encapsulated packets without skb->encapsulation (e.g., virtio). Use the same helper in skb_csum_hwoffload_help(). Its open coded test has the false positive of (2) and ignores the inner header. Background: checksum offload of tunneled packets invariants: Non-GSO skb: - If the inner packet is CHECKSUM_PARTIAL, Local Checksum Offload computes the outer checksum in software and the device offloads only the inner L4 checksum. - If the inner packet is CHECKSUM_NONE (e.g., SCTP-in-UDP, ESP-in-UDP, Remote Checksum Offload), the device offloads the outer UDP or GRE checksum instead. GSO skb: - A device with NETIF_F_GSO_UDP_TUNNEL_CSUM or NETIF_F_GSO_GRE_CSUM computes both inner and outer checksums per segment. The outer checksum is seeded with only the pseudo-header checksum. A NETIF_F_IPV6_CSUM device must parse through the outer headers to reach the inner ones. Fixes: c4336a07eb6b ("net: correctly handle tunneled traffic on IPV6_CSUM GSO fallback") Cc: stable@vger.kernel.org Signed-off-by: Willem de Bruijn Link: https://patch.msgid.link/20260926140506.2335137-1-willemdebruijn.kernel@gmail.com Signed-off-by: Jakub Kicinski --- net/core/dev.c | 47 +++++++++++++++++++++++++++++++---------------- 1 file changed, 31 insertions(+), 16 deletions(-) diff --git a/net/core/dev.c b/net/core/dev.c index 8638c994bbcb..18dc88990510 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -3818,20 +3818,36 @@ static netdev_features_t dflt_features_check(struct sk_buff *skb, return vlan_features_check(skb, features); } -static bool skb_gso_has_extension_hdr(const struct sk_buff *skb) +static bool __skb_has_ipv6_ext_hdr(const struct sk_buff *skb, int nhoff) { - if (!skb->encapsulation) - return ((skb_shinfo(skb)->gso_type & SKB_GSO_TCPV6 || - (skb_shinfo(skb)->gso_type & SKB_GSO_UDP_L4 && - vlan_get_protocol(skb) == htons(ETH_P_IPV6))) && - skb_transport_header_was_set(skb) && - skb_network_header_len(skb) != sizeof(struct ipv6hdr)); - else - return (!skb_inner_network_header_was_set(skb) || - ((skb_shinfo(skb)->gso_type & SKB_GSO_TCPV6 || - (skb_shinfo(skb)->gso_type & SKB_GSO_UDP_L4 && - inner_ip_hdr(skb)->version == 6)) && - skb_inner_network_header_len(skb) != sizeof(struct ipv6hdr))); + const struct ipv6hdr *ip6h; + struct ipv6hdr _ip6h; + + ip6h = skb_header_pointer(skb, nhoff, sizeof(_ip6h), &_ip6h); + return ip6h && ip6h->version == 6 && ipv6_ext_hdr(ip6h->nexthdr); +} + +static bool skb_has_ipv6_extension_hdr(const struct sk_buff *skb) +{ + if (vlan_get_protocol(skb) == htons(ETH_P_IPV6)) { + if (__skb_has_ipv6_ext_hdr(skb, skb_network_offset(skb))) + return true; + + /* Catch tunnels without skb->encapsulation (e.g., virtio). */ + if (!skb->encapsulation && + skb_transport_header_was_set(skb) && + skb_network_header_len(skb) != sizeof(struct ipv6hdr)) + return true; + } + + /* Tunnels without an inner network header, such as SCTP-in-UDP or + * PSP, have no inner IP header and thus no inner extension header. + */ + if (skb->encapsulation && skb_inner_network_header_was_set(skb) && + __skb_has_ipv6_ext_hdr(skb, skb_inner_network_offset(skb))) + return true; + + return false; } static netdev_features_t gso_features_check(const struct sk_buff *skb, @@ -3886,7 +3902,7 @@ static netdev_features_t gso_features_check(const struct sk_buff *skb, * so neither does TSO that depends on it. */ if (features & NETIF_F_IPV6_CSUM && - skb_gso_has_extension_hdr(skb)) + skb_has_ipv6_extension_hdr(skb)) features &= ~(NETIF_F_IPV6_CSUM | NETIF_F_TSO6 | NETIF_F_GSO_UDP_L4); return features; @@ -3988,8 +4004,7 @@ int skb_csum_hwoffload_help(struct sk_buff *skb, return 0; if (features & (NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM)) { - if (vlan_get_protocol(skb) == htons(ETH_P_IPV6) && - skb_network_header_len(skb) != sizeof(struct ipv6hdr)) + if (skb_has_ipv6_extension_hdr(skb)) goto sw_checksum; switch (skb->csum_offset) { From 6f0c2c4f5e7143eba75153ecc8af3a03f7b6a98b Mon Sep 17 00:00:00 2001 From: Sandeep Haemoon Date: Thu, 24 Sep 2026 23:51:37 +0300 Subject: [PATCH 659/694] net: ethernet: mtk_eth_soc: allocate dummy netdev sooner The RX rings and the shared NAPI are carried by an internal "dummy" netdev (eth->dummy_dev). mtk_probe() creates that dummy device, and adds the shared tx/rx NAPI to it, only after the MAC netdevs have been registered. That order has been there since the MT7623 support was added and leaves a window in which the netdevs are visible to the network core before the shared NAPI and the dummy device exist. If anything brings the first netdev up during that window (e.g. netifd opening the WAN link the instant the interface is registered), the first-open path runs without them. On NETSYS v2 and later SoCs mtk_rx_alloc() -> __xdp_rxq_info_reg() then dereferences eth->dummy_dev == NULL: WARNING: CPU: ... Missing net_device from driver in net/core/xdp.c and mtk_open() fails with -ENODEV. On NETSYS v1 SoCs (MT7621/MT7622/MT7623/MT7629) there is no page pool, so the same open instead calls napi_enable() on a napi_struct that netif_napi_add() has not initialised yet and dereferences a NULL napi->dev. Allocate the dummy netdev and add the shared tx/rx NAPI to it before the register_netdev() loop. Both the dummy-allocation failure and a register_netdev() failure now unwind through mtk_unreg_dev(), which unregisters the net_device notifiers and the netdevs registered so far, stopping a concurrently-opened netdev (ndo_stop) and its NAPI before the dummy carrier is freed. A netdev that opened during the window can also have queued the frame-engine reset worker (via the tx watchdog or phylink), so the probe error path now cancels eth->pending_work before freeing the shared state, as mtk_cleanup() already does on remove. Fixes: 656e705243fd ("net-next: mediatek: add support for MT7623 ethernet") Signed-off-by: Sandeep Haemoon Reviewed-by: Simon Horman Link: https://patch.msgid.link/patch-e52c13380b0768fec63af04c80e4b487@loljews.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/mediatek/mtk_eth_soc.c | 38 ++++++++++++++------- 1 file changed, 26 insertions(+), 12 deletions(-) diff --git a/drivers/net/ethernet/mediatek/mtk_eth_soc.c b/drivers/net/ethernet/mediatek/mtk_eth_soc.c index 2ea5dfe85539..ad0fdc4133fb 100644 --- a/drivers/net/ethernet/mediatek/mtk_eth_soc.c +++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.c @@ -5341,6 +5341,22 @@ static int mtk_probe(struct platform_device *pdev) } } + /* we run 2 devices on the same DMA ring so we need a dummy device + * for NAPI to work. Allocate it before registering the netdevs so a + * concurrent ndo_open (e.g. netifd bringing the first netdev up the + * instant it is registered) never observes eth->dummy_dev == NULL in + * mtk_dma_init() -> mtk_rx_alloc() -> __xdp_rxq_info_reg() (net/core/xdp.c + * "Missing net_device from driver") and fails the first open. + */ + eth->dummy_dev = alloc_netdev_dummy(0); + if (!eth->dummy_dev) { + err = -ENOMEM; + dev_err(eth->dev, "failed to allocated dummy device\n"); + goto err_unreg_netdev; + } + netif_napi_add(eth->dummy_dev, ð->tx_napi, mtk_napi_tx); + netif_napi_add(eth->dummy_dev, ð->rx_napi, mtk_napi_rx); + for (i = 0; i < MTK_MAX_DEVS; i++) { if (!eth->netdev[i]) continue; @@ -5355,18 +5371,6 @@ static int mtk_probe(struct platform_device *pdev) eth->netdev[i]->base_addr, eth->irq[MTK_FE_IRQ_SHARED]); } - /* we run 2 devices on the same DMA ring so we need a dummy device - * for NAPI to work - */ - eth->dummy_dev = alloc_netdev_dummy(0); - if (!eth->dummy_dev) { - err = -ENOMEM; - dev_err(eth->dev, "failed to allocated dummy device\n"); - goto err_unreg_netdev; - } - netif_napi_add(eth->dummy_dev, ð->tx_napi, mtk_napi_tx); - netif_napi_add(eth->dummy_dev, ð->rx_napi, mtk_napi_rx); - platform_set_drvdata(pdev, eth); schedule_delayed_work(ð->reset.monitor_work, MTK_DMA_MONITOR_TIMEOUT); @@ -5376,6 +5380,16 @@ static int mtk_probe(struct platform_device *pdev) err_unreg_netdev: mtk_unreg_dev(eth); err_deinit_ppe: + /* A netdev opened while probe was still registering can have queued + * the frame-engine reset worker, which takes rtnl and drives the + * hardware; cancel it before anything below is torn down. + */ + cancel_work_sync(ð->pending_work); + if (eth->dummy_dev) { + netif_napi_del(ð->tx_napi); + netif_napi_del(ð->rx_napi); + free_netdev(eth->dummy_dev); + } mtk_ppe_deinit(eth); mtk_mdio_cleanup(eth); err_free_dev: From 20f3599d85b416fd292a199364cb1248dcd9c780 Mon Sep 17 00:00:00 2001 From: Yongzhao Chen Date: Sun, 27 Sep 2026 17:51:36 +0200 Subject: [PATCH 660/694] net: phy: qcom: at803x: Fix IPQ5018 short-cable DAC values When "qcom,dac-preset-short-cable" is set, ipq5018_config_init() programs the MDAC (MMD1 0x8100) and EDAC (debug 0x4380) fields. Both fields occupy bits 15:8 (IPQ5018_PHY_DAC_MASK), but the value 0x10 is passed unshifted as the set argument of phy_modify_mmd() and at803x_debug_reg_mask(). Neither helper shifts or masks that argument, so both fields are cleared to 0x00 instead of being set to 0x10, and bit 4 of the low byte, which is outside the field, is set. Use FIELD_PREP() to place the value in the field. This matches the vendor SDK, which clears bits 15:8 and ORs in the value shifted left by 8. On a Redmi AX5400 board, where the IPQ5018 internal PHY connects to a QCA8337 switch PHY without a cable, MDAC and EDAC read 0x6868 and 0x7800 before the write. With this change they read back 0x1068 and 0x1000, with the low byte preserved. Without it, the same writes would leave 0x0078 and 0x0010. No in-tree DTS sets this property yet, but it is documented in qca,ar803x.yaml and used by several IPQ5018 boards in OpenWrt. Fixes: d46502279a11 ("net: phy: qcom: at803x: Add Qualcomm IPQ5018 Internal PHY support") Signed-off-by: Yongzhao Chen Reviewed-by: Andrew Lunn Link: https://patch.msgid.link/20260927155136.2489-1-yongzhao.derek@gmail.com Signed-off-by: Jakub Kicinski --- drivers/net/phy/qcom/at803x.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/drivers/net/phy/qcom/at803x.c b/drivers/net/phy/qcom/at803x.c index 6872dbf77856..cacbadf1f484 100644 --- a/drivers/net/phy/qcom/at803x.c +++ b/drivers/net/phy/qcom/at803x.c @@ -1051,11 +1051,15 @@ static int ipq5018_config_init(struct phy_device *phydev) if (priv->set_short_cable_dac) { /* setting MDAC (Multi-level Digital-to-Analog Converter) in MMD1 */ phy_modify_mmd(phydev, MDIO_MMD_PMAPMD, IPQ5018_PHY_MMD1_MDAC, - IPQ5018_PHY_DAC_MASK, IPQ5018_PHY_MMD1_MDAC_VAL); + IPQ5018_PHY_DAC_MASK, + FIELD_PREP(IPQ5018_PHY_DAC_MASK, + IPQ5018_PHY_MMD1_MDAC_VAL)); /* setting EDAC (Error-detection and Correction) in debug register */ at803x_debug_reg_mask(phydev, IPQ5018_PHY_DEBUG_EDAC, - IPQ5018_PHY_DAC_MASK, IPQ5018_PHY_DEBUG_EDAC_VAL); + IPQ5018_PHY_DAC_MASK, + FIELD_PREP(IPQ5018_PHY_DAC_MASK, + IPQ5018_PHY_DEBUG_EDAC_VAL)); } return 0; From 0d2c49915bc3aea4521cda5e4d42f24f63b94585 Mon Sep 17 00:00:00 2001 From: Quentin Freimanis Date: Sun, 27 Sep 2026 21:45:42 -0700 Subject: [PATCH 661/694] net: sparx5: skip ptp deinit if init was skipped sparx5_ptp_init() returns early on the base lan969x variants because they do not have the SPX5_FEATURE_PTP flag. It also returns early when no "ptp" interrupt is described. Unbinding the driver then causes a NULL pointer dereference when cleaning up uninitialized tx_skbs queues: Unable to handle kernel NULL pointer dereference at virtual address 0000000000000008 Call trace: skb_queue_purge_reason+0x68/0x120 (P) sparx5_ptp_deinit+0x64/0xc0 mchp_sparx5_remove+0x38/0x70 platform_remove+0x20/0x30 device_remove+0x4c/0x80 Before per-port tx_skbs were added, a NULL dereference would have happened when calling ptp_clock_unregister() on the never-registered PTP clocks. Fix by returning early in sparx5_ptp_deinit() if PTP is not used. Fixes: 0933bd04047c ("net: sparx5: Add support for ptp clocks") Signed-off-by: Quentin Freimanis Reviewed-by: Daniel Machon Link: https://patch.msgid.link/20260928044542.29287-1-quentin@q-lab.dev Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/microchip/sparx5/sparx5_ptp.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/net/ethernet/microchip/sparx5/sparx5_ptp.c b/drivers/net/ethernet/microchip/sparx5/sparx5_ptp.c index a16ec8136d6d..5953e04ef14e 100644 --- a/drivers/net/ethernet/microchip/sparx5/sparx5_ptp.c +++ b/drivers/net/ethernet/microchip/sparx5/sparx5_ptp.c @@ -673,6 +673,9 @@ void sparx5_ptp_deinit(struct sparx5 *sparx5) struct sparx5_port *port; int i; + if (!sparx5->ptp) + return; + if (sparx5->ptp_irq >= 0) { disable_irq(sparx5->ptp_irq); sparx5->ptp_irq = -ENXIO; From 1b92780ae4fc3be64f05d97ed8833c445e73b0cf Mon Sep 17 00:00:00 2001 From: Jeff Jo Date: Thu, 24 Sep 2026 15:44:58 -0700 Subject: [PATCH 662/694] tcp: refresh TS.Recent for accepted old ACKs A TCP packet can carry new data while acknowledging traffic in the opposite direction. With overlapping traffic in both directions, a delayed packet's acknowledgment can be older than one Linux has already accepted, even when that packet fills a gap in the received data. Linux accepts the data, but tcp_ack() takes the old_ack path and skips updating TS.Recent, the timestamp saved for outgoing acknowledgments. The reply therefore echoes an older timestamp. If the sender uses this echo to measure round-trip time after a long idle period, its estimate includes the idle time and can reduce its sending rate. Update TS.Recent in old_ack using tcp_replace_ts_recent(), before SACK processing can trigger a transmission. This reuses the existing timestamp and sequence checks, including PAWS protection against old duplicate packets. ACK validation already rejects old ACKs in SYN_RECV before this path, so no additional state check is needed. Echoing the timestamp of the packet that fills the receive gap follows RFC 7323 section 4.3. In a socket reproduction with 300 seconds idle, controlled reordering and retransmission to exercise timestamp-based RTT sampling, the sender's smoothed round-trip time was 37.5 seconds without the fix and 15.5 ms with it. Fixes: 12fb3dd9dc3c ("tcp: call tcp_replace_ts_recent() from tcp_ack()") Assisted-by: LLM sparse Signed-off-by: Jeff Jo Reviewed-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv4/tcp_input.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index 92bc60716f33..99baf14afdfd 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -4465,6 +4465,12 @@ static int tcp_ack(struct sock *sk, const struct sk_buff *skb, int flag) return 1; old_ack: + /* An old ACK can carry new data. Update TS.Recent before SACK + * processing can trigger a retransmission. + */ + if (flag & FLAG_UPDATE_TS_RECENT) + tcp_replace_ts_recent(tp, TCP_SKB_CB(skb)->seq); + /* If data was SACKed, tag it and see if we should send more data. * If data was DSACKed, see if we can undo a cwnd reduction. */ From 81fe707ae58e67e5a029a9778c4b0c1b478bcb97 Mon Sep 17 00:00:00 2001 From: Jeff Jo Date: Thu, 24 Sep 2026 15:44:59 -0700 Subject: [PATCH 663/694] selftests: net: check timestamp echo after an old ACK Add a regression test for a gap-filling packet whose acknowledgment has become old. Linux first sends data to the peer. Deliver two peer data packets out of order: the later packet acknowledges Linux's data, while the delayed packet still carries the earlier acknowledgment. Require the ACK that closes the receive gap to echo the delayed packet's timestamp, 301000. Without the fix, Linux accepts the data but still echoes the previously saved timestamp, 1000. Also check that the application can read all 34 bytes. Use a large jump in peer timestamps to represent the idle interval, with no real wait, loss or retransmission. The test directly checks the outgoing timestamp echo. It requires packetdrill's merged TSecr verification fix (linked below); older tools incorrectly pass on an unfixed kernel. Use the existing packetdrill selftest runner for IPv4, IPv6 and IPv4-mapped IPv6. Link: https://github.com/google/packetdrill/commit/83f72d3f9085d0e26eb4d206fe4d7cfab5b6d872 Assisted-by: LLM Signed-off-by: Jeff Jo Reviewed-by: Eric Dumazet Signed-off-by: David S. Miller --- .../net/packetdrill/tcp_old_ack_ts.pkt | 22 +++++++++++++++++++ 1 file changed, 22 insertions(+) create mode 100644 tools/testing/selftests/net/packetdrill/tcp_old_ack_ts.pkt diff --git a/tools/testing/selftests/net/packetdrill/tcp_old_ack_ts.pkt b/tools/testing/selftests/net/packetdrill/tcp_old_ack_ts.pkt new file mode 100644 index 000000000000..f072e5e4c449 --- /dev/null +++ b/tools/testing/selftests/net/packetdrill/tcp_old_ack_ts.pkt @@ -0,0 +1,22 @@ +// SPDX-License-Identifier: GPL-2.0 +// An out-of-order segment advances SND.UNA. The gap filler still carries +// its original, older ACK. The reply must echo the gap filler's timestamp. +--tcp_ts_tick_usecs=1000 + +`./defaults.sh` + +0.000 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3 ++0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0 ++0 bind(3, ..., ...) = 0 ++0 listen(3, 1) = 0 +0.100 < S 0:0(0) win 20000 ++0 > S. 0:0(0) ack 1 +0.200 < . 1:1(0) ack 1 win 20000 ++0 accept(3, ..., ...) = 4 +0.300 write(4, ..., 17) = 17 ++0 > P. 1:18(17) ack 1 +0.310 < P. 18:35(17) ack 18 win 20000 ++0 > . 18:18(0) ack 1 +0.320 < P. 1:18(17) ack 1 win 20000 ++0 > . 18:18(0) ack 35 ++0 read(4, ..., 34) = 34 From 747928b5d4ff8c8de55138a54c13980e4bb88578 Mon Sep 17 00:00:00 2001 From: Aohan Mei Date: Thu, 10 Sep 2026 16:33:26 +0800 Subject: [PATCH 664/694] netfilter: nft_flow_offload: drop flowtable reference on init error path nft_flow_offload_init() bumps the flowtable use count with nft_use_inc() before calling nf_ct_netns_get(). When the latter fails, the error is returned as-is and the reference is leaked. The upper layers do not balance it either: nf_tables_newexpr() clears expr->ops when the expression init callback fails, so the nft_expr_more() iteration in nft_rule_expr_deactivate() and nf_tables_rule_destroy() stops right before the failed expression and its ->destroy callback, which would drop the reference, never runs. Each failed rule addition therefore leaks one flowtable reference and the flowtable can no longer be removed: NFT_MSG_DELFLOWTABLE keeps reporting -EBUSY even though no rule references it. Save the nf_ct_netns_get() return value and undo the nft_use_inc() when it fails, restoring the inc/dec pairing within nft_flow_offload_init() itself. Fixes: a3c90f7a2323 ("netfilter: nf_tables: flow offload expression") Reported-by: TencentOS Corvus AI Cc: stable@vger.kernel.org Assisted-by: CodeBuddy:Kimi-K3 Signed-off-by: Aohan Mei Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nft_flow_offload.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/net/netfilter/nft_flow_offload.c b/net/netfilter/nft_flow_offload.c index 32b4281038dd..d3c5651dd699 100644 --- a/net/netfilter/nft_flow_offload.c +++ b/net/netfilter/nft_flow_offload.c @@ -160,6 +160,7 @@ static int nft_flow_offload_init(const struct nft_ctx *ctx, struct nft_flow_offload *priv = nft_expr_priv(expr); u8 genmask = nft_genmask_next(ctx->net); struct nft_flowtable *flowtable; + int err; if (!tb[NFTA_FLOW_TABLE_NAME]) return -EINVAL; @@ -174,7 +175,11 @@ static int nft_flow_offload_init(const struct nft_ctx *ctx, priv->flowtable = flowtable; - return nf_ct_netns_get(ctx->net, ctx->family); + err = nf_ct_netns_get(ctx->net, ctx->family); + if (err < 0) + nft_use_dec(&flowtable->use); + + return err; } static void nft_flow_offload_deactivate(const struct nft_ctx *ctx, From bdac17779f46817f60103baf09f0370546c3f023 Mon Sep 17 00:00:00 2001 From: Julian Anastasov Date: Thu, 10 Sep 2026 13:08:31 +0300 Subject: [PATCH 665/694] ipvs: fix missing counter decrement in lblc LBLC may delete cache entries for destinations that are removed or overloaded and replace them with available ones. But ip_vs_lblc_new() forgets to decrement the tbl->entries counter after calling ip_vs_lblc_del(). This can lead to increased shrinking of the cache with every new garbage collection. Fixes: 2f3d771a35fe ("ipvs: do not use dest after ip_vs_dest_put in LBLC") Link: https://sashiko.dev/#/patchset/0bdd5abe9968ded7ca2b9cb6844ba83d94cc8d53.1787318053.git.zhilinz%40nebusec.ai Signed-off-by: Julian Anastasov Signed-off-by: Pablo Neira Ayuso --- net/netfilter/ipvs/ip_vs_lblc.c | 1 + 1 file changed, 1 insertion(+) diff --git a/net/netfilter/ipvs/ip_vs_lblc.c b/net/netfilter/ipvs/ip_vs_lblc.c index 693bcc82ccb7..a2b574904f13 100644 --- a/net/netfilter/ipvs/ip_vs_lblc.c +++ b/net/netfilter/ipvs/ip_vs_lblc.c @@ -203,6 +203,7 @@ ip_vs_lblc_new(struct ip_vs_lblc_table *tbl, const union nf_inet_addr *daddr, if (en->dest == dest) return en; ip_vs_lblc_del(en); + atomic_dec(&tbl->entries); } en = kmalloc_obj(*en, GFP_ATOMIC); if (!en) From 2a3c3de660f96865f303fd25f1285f2ffbfd521d Mon Sep 17 00:00:00 2001 From: Zhiling Zou Date: Thu, 10 Sep 2026 13:08:32 +0300 Subject: [PATCH 666/694] ipvs: bound LBLCR and LBLC cache growth ip_vs_lblcr_new() and ip_vs_lblc_new() create cache entries for every previously unseen destination address. The table max_size only tells the periodic collector to reclaim entries after the cache has already exceeded the limit. It does not reclaim entries that the attacker continues to use. Reject new cache entries once either table reaches max_size * 3 / 2. The extra headroom lets the periodic collector catch up while the existing scheduler fallback continues to use the selected destination when cache creation fails. New traffic therefore stays serviceable without growing the tables further. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Cc: stable@vger.kernel.org Reported-by: Vega Suggested-by: Julian Anastasov Signed-off-by: Zhiling Zou Acked-by: Julian Anastasov Signed-off-by: Pablo Neira Ayuso --- net/netfilter/ipvs/ip_vs_lblc.c | 3 +++ net/netfilter/ipvs/ip_vs_lblcr.c | 3 +++ 2 files changed, 6 insertions(+) diff --git a/net/netfilter/ipvs/ip_vs_lblc.c b/net/netfilter/ipvs/ip_vs_lblc.c index a2b574904f13..55af77a56929 100644 --- a/net/netfilter/ipvs/ip_vs_lblc.c +++ b/net/netfilter/ipvs/ip_vs_lblc.c @@ -205,6 +205,9 @@ ip_vs_lblc_new(struct ip_vs_lblc_table *tbl, const union nf_inet_addr *daddr, ip_vs_lblc_del(en); atomic_dec(&tbl->entries); } + if (atomic_read(&tbl->entries) >= tbl->max_size * 3 / 2) + return NULL; + en = kmalloc_obj(*en, GFP_ATOMIC); if (!en) return NULL; diff --git a/net/netfilter/ipvs/ip_vs_lblcr.c b/net/netfilter/ipvs/ip_vs_lblcr.c index f53f05ceea36..858393b1d2d1 100644 --- a/net/netfilter/ipvs/ip_vs_lblcr.c +++ b/net/netfilter/ipvs/ip_vs_lblcr.c @@ -363,6 +363,9 @@ ip_vs_lblcr_new(struct ip_vs_lblcr_table *tbl, const union nf_inet_addr *daddr, en = ip_vs_lblcr_get(af, tbl, daddr); if (!en) { + if (atomic_read(&tbl->entries) >= tbl->max_size * 3 / 2) + return NULL; + en = kmalloc_obj(*en, GFP_ATOMIC); if (!en) return NULL; From 616cf5c06934364ab1002b420ebe975850b4b208 Mon Sep 17 00:00:00 2001 From: Julian Anastasov Date: Sat, 19 Sep 2026 21:07:32 +0300 Subject: [PATCH 667/694] ipvs: do not create invisible templates The IP_VS_CONN_F_ONE_PACKET flag was implemented for normal connections. When conn template inherits this flag from dest->conn_flags it will not be hashed. As result, we will create new template for every new normal connection. Fix it to allow one template to be used by many normal connections. Fixes: 26ec037f9841 ("IPVS: one-packet scheduling") Link: https://netdev-ai.bots.linux.dev/sashiko/#/patchset/20260916231652.127456-1-pablo%40netfilter.org Signed-off-by: Julian Anastasov Signed-off-by: Pablo Neira Ayuso --- net/netfilter/ipvs/ip_vs_conn.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/net/netfilter/ipvs/ip_vs_conn.c b/net/netfilter/ipvs/ip_vs_conn.c index 6fa3e1dc534c..cb009208826f 100644 --- a/net/netfilter/ipvs/ip_vs_conn.c +++ b/net/netfilter/ipvs/ip_vs_conn.c @@ -1102,6 +1102,9 @@ ip_vs_bind_dest(struct ip_vs_conn *cp, struct ip_vs_dest *dest) if (cp->protocol != IPPROTO_UDP) conn_flags &= ~IP_VS_CONN_F_ONE_PACKET; flags = cp->flags; + /* Only visible templates can control multiple connections */ + if (flags & IP_VS_CONN_F_TEMPLATE) + conn_flags &= ~IP_VS_CONN_F_ONE_PACKET; /* Bind with the destination and its corresponding transmitter */ if (flags & IP_VS_CONN_F_SYNC) { /* Synced conns are hashed, so they can not get this flag */ From f96e91748f6304668fe647e686d199d7207d36f1 Mon Sep 17 00:00:00 2001 From: Julian Anastasov Date: Sat, 19 Sep 2026 21:07:33 +0300 Subject: [PATCH 668/694] ipvs: filter some flags received in the backup server While the IPVS SYNC protocol is not secure by design we can still protect the backup server from messages that can wreak havoc. This commit addresses problems from received connection flags or their combinations. We now drop messages as follows: 1. the NO_CPORT+TEMPLATE combination allows lookups for normal connections to hit template which can break in many ways. While the master does not sync connections with NO_CPORT flag, i.e. before they are established, we still accept NO_CPORT without TEMPLATE. 2. ONE_PACKET: it is not sent by master, so we do not expect it in backup. Before now it was ignored by IP_VS_CONN_F_BACKUP_MASK for protocol v1 while protocol v0 created connections that are not hashed and dropped immediately. Better to apply the IP_VS_CONN_F_BACKUP_MASK also to the flags from v0 messages for consistency with v1. Fixes: 87375ab47cd0 ("[IPVS]: ip_vs_ftp breaks connections using persistence") Signed-off-by: Julian Anastasov Signed-off-by: Pablo Neira Ayuso --- net/netfilter/ipvs/ip_vs_sync.c | 33 ++++++++++++++++++++++++++++++--- 1 file changed, 30 insertions(+), 3 deletions(-) diff --git a/net/netfilter/ipvs/ip_vs_sync.c b/net/netfilter/ipvs/ip_vs_sync.c index 5383aeafb0ae..69dc28153ec1 100644 --- a/net/netfilter/ipvs/ip_vs_sync.c +++ b/net/netfilter/ipvs/ip_vs_sync.c @@ -949,6 +949,21 @@ static void ip_vs_proc_conn(struct netns_ipvs *ipvs, struct ip_vs_conn_param *pa ip_vs_conn_put(cp); } +/* Check for incompatible flags */ +static bool ip_vs_sync_validate_flags(u32 flags) +{ + /* We do not expect NO_CPORT, especially to allow lookups + * to hit templates + */ + if (flags & IP_VS_CONN_F_NO_CPORT) { + if (flags & IP_VS_CONN_F_TEMPLATE) + return false; + } + if (flags & IP_VS_CONN_F_ONE_PACKET) + return false; + return true; +} + /* * Process received multicast message for Version 0 */ @@ -972,8 +987,7 @@ static void ip_vs_process_message_v0(struct netns_ipvs *ipvs, const char *buffer return; } s = (struct ip_vs_sync_conn_v0 *) p; - flags = ntohs(s->flags) | IP_VS_CONN_F_SYNC; - flags &= ~IP_VS_CONN_F_HASHED; + flags = ntohs(s->flags); if (flags & IP_VS_CONN_F_SEQ_MASK) { opt = (struct ip_vs_sync_conn_options *)&s[1]; p += FULL_CONN_SIZE; @@ -986,6 +1000,13 @@ static void ip_vs_process_message_v0(struct netns_ipvs *ipvs, const char *buffer p += SIMPLE_CONN_SIZE; } + if (!ip_vs_sync_validate_flags(flags)) { + IP_VS_DBG(2, "BACKUP v0, Invalid flags 0x%X\n", flags); + continue; + } + flags &= IP_VS_CONN_F_BACKUP_MASK; + flags |= IP_VS_CONN_F_SYNC; + state = ntohs(s->state); if (!(flags & IP_VS_CONN_F_TEMPLATE)) { pp = ip_vs_proto_get(s->protocol); @@ -1141,7 +1162,13 @@ static inline int ip_vs_proc_sync_conn(struct netns_ipvs *ipvs, __u8 *p, __u8 *m } /* Get flags and Mask off unsupported */ - flags = ntohl(s->v4.flags) & IP_VS_CONN_F_BACKUP_MASK; + flags = ntohl(s->v4.flags); + if (!ip_vs_sync_validate_flags(flags)) { + IP_VS_DBG(3, "BACKUP, Invalid flags 0x%X\n", flags); + retc = 25; + goto out; + } + flags &= IP_VS_CONN_F_BACKUP_MASK; flags |= IP_VS_CONN_F_SYNC; state = ntohs(s->v4.state); From 86b6471e690c67d4c5a1892993f34e080f1eae9d Mon Sep 17 00:00:00 2001 From: Weiming Shi Date: Mon, 21 Sep 2026 10:08:22 +0800 Subject: [PATCH 669/694] netfilter: nft_set_rbtree: skip transaction elements during GC Since nft_set_commit_update() runs set commit callbacks before processing NEWSETELEM transactions, nft_rbtree_gc_scan() can observe elements added by the transaction being committed. The scan records an interval end in rbe_end without checking the element's transaction state. A later, unrelated expired start then moves both elements to the expired list. The synchronous GC queue can free the new end element before the transaction subsequently activates it, causing a use-after-free. Only consider elements that are fully active in both generations. This keeps transaction-state elements out of the GC scan and preserves interval pairing across skipped elements. KASAN reports: BUG: KASAN: slab-use-after-free in nft_setelem_activate nft_setelem_activate net/netfilter/nf_tables_api.c:7047 nf_tables_commit net/netfilter/nf_tables_api.c:11137 Allocated by task 130: nft_set_elem_init net/netfilter/nf_tables_api.c:6794 nft_add_set_elem net/netfilter/nf_tables_api.c:7523 Freed by task 130: nft_trans_gc_trans_free net/netfilter/nf_tables_api.c:10506 rcu_core kernel/rcu/tree.c:2919 Fixes: 1e3b9e1c77fe ("netfilter: nf_tables: call set ops .commit when building new ruleset blob") Reported-by: Assisted-by: LLM Signed-off-by: Weiming Shi Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nft_set_rbtree.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/net/netfilter/nft_set_rbtree.c b/net/netfilter/nft_set_rbtree.c index 9894832281c4..12431b55752f 100644 --- a/net/netfilter/nft_set_rbtree.c +++ b/net/netfilter/nft_set_rbtree.c @@ -900,6 +900,8 @@ static void nft_rbtree_gc_scan(struct nft_set *set) next = rb_next(node); rbe = rb_entry(node, struct nft_rbtree_elem, node); + if (!nft_set_elem_active(&rbe->ext, NFT_GENMASK_ANY)) + continue; /* elements are reversed in the rbtree for historical reasons, * from highest to lowest value, that is why end element is From 16d464013ec2267b00a83f4bfbb97b3ecc63bfa7 Mon Sep 17 00:00:00 2001 From: Fernando Fernandez Mancera Date: Tue, 22 Sep 2026 21:18:52 +0200 Subject: [PATCH 670/694] netfilter: bpf: reject invalid NAT manipulation types As bpf_ct_set_nat_info() is not validating the NAT manipulation type a wrong value can be passed directly to nf_nat_setup_info(). This triggers the WARN_ON() at nf_nat_setup_info() and if panic_on_warn isn't set, then IPS_SRC_NAT_DONE is set without adding nat_bysource and conntrack cleanup tries to unlink an uninitialized hlist node. Fix this by checking that NAT manipulation type is correct before calling nf_nat_setup_info(). In addition, if the WARN_ON is hit, return NF_DROP instead of continuing with the processing to avoid similar situations in the future. Reported-by: VEGA Fixes: 0fabd2aa199f ("net: netfilter: add bpf_ct_set_nat_info kfunc helper") Signed-off-by: Fernando Fernandez Mancera Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nf_nat_bpf.c | 3 +++ net/netfilter/nf_nat_core.c | 5 +++-- 2 files changed, 6 insertions(+), 2 deletions(-) diff --git a/net/netfilter/nf_nat_bpf.c b/net/netfilter/nf_nat_bpf.c index f9dd85ccea01..7572b58c448b 100644 --- a/net/netfilter/nf_nat_bpf.c +++ b/net/netfilter/nf_nat_bpf.c @@ -39,6 +39,9 @@ __bpf_kfunc int bpf_ct_set_nat_info(struct nf_conn___init *nfct, if (proto != NFPROTO_IPV4 && proto != NFPROTO_IPV6) return -EINVAL; + if (manip != NF_NAT_MANIP_SRC && manip != NF_NAT_MANIP_DST) + return -EINVAL; + memset(&range, 0, sizeof(struct nf_nat_range2)); range.flags = NF_NAT_RANGE_MAP_IPS; range.min_addr = *addr; diff --git a/net/netfilter/nf_nat_core.c b/net/netfilter/nf_nat_core.c index a4858c2b2d65..cc8e1e81006d 100644 --- a/net/netfilter/nf_nat_core.c +++ b/net/netfilter/nf_nat_core.c @@ -767,8 +767,9 @@ nf_nat_setup_info(struct nf_conn *ct, if (nf_ct_is_confirmed(ct)) return NF_ACCEPT; - WARN_ON(maniptype != NF_NAT_MANIP_SRC && - maniptype != NF_NAT_MANIP_DST); + if (WARN_ON(maniptype != NF_NAT_MANIP_SRC && + maniptype != NF_NAT_MANIP_DST)) + return NF_DROP; if (WARN_ON(nf_nat_initialized(ct, maniptype))) return NF_DROP; From 7c549fb7eecd01fdba7c0d12c01da876ef8a3214 Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Wed, 23 Sep 2026 10:53:02 +0200 Subject: [PATCH 671/694] netfilter: flowtable: generalize pending status bit Rename NF_FLOW_HW_PENDING to NF_FLOW_PENDING and use it to inhibit the flowtable GC worker until pending hw offload work has been completed. Apparently, nf_flow_offload_stats() can schedule work to retrieve stats while the flow is being removed by GC. And this bit can also be used in a follow up patch to disable GC until the flow has been fully added in both directions. Revert the reordering done in commit d644b23afe1e ("netfilter: flowtable: publish HW_DEAD after worker is done") to prevent a race between GC and hw offload handler. Fixes: 2c8897953f3b ("netfilter: flowtable: Add pending bit for offload work") Signed-off-by: Pablo Neira Ayuso --- include/net/netfilter/nf_flow_table.h | 2 +- net/netfilter/nf_flow_table_core.c | 7 ++++++- net/netfilter/nf_flow_table_offload.c | 14 +++++--------- net/sched/act_ct.c | 2 +- 4 files changed, 13 insertions(+), 12 deletions(-) diff --git a/include/net/netfilter/nf_flow_table.h b/include/net/netfilter/nf_flow_table.h index f2e2771f188f..5b611efaa3cd 100644 --- a/include/net/netfilter/nf_flow_table.h +++ b/include/net/netfilter/nf_flow_table.h @@ -183,10 +183,10 @@ enum nf_flow_flags { NF_FLOW_DNAT, NF_FLOW_CLOSING, NF_FLOW_TEARDOWN, + NF_FLOW_PENDING, NF_FLOW_HW, NF_FLOW_HW_DYING, NF_FLOW_HW_DEAD, - NF_FLOW_HW_PENDING, NF_FLOW_HW_BIDIRECTIONAL, NF_FLOW_HW_ESTABLISHED, }; diff --git a/net/netfilter/nf_flow_table_core.c b/net/netfilter/nf_flow_table_core.c index 934c6151f558..36bbc7be2f74 100644 --- a/net/netfilter/nf_flow_table_core.c +++ b/net/netfilter/nf_flow_table_core.c @@ -575,7 +575,12 @@ static void nf_flow_table_extend_ct_timeout(struct nf_conn *ct) static void nf_flow_offload_gc_step(struct nf_flowtable *flow_table, struct flow_offload *flow, void *data) { - bool teardown = test_bit(NF_FLOW_TEARDOWN, &flow->flags); + bool teardown; + + if (test_bit(NF_FLOW_PENDING, &flow->flags)) + return; + + teardown = test_bit(NF_FLOW_TEARDOWN, &flow->flags); if (nf_flow_has_expired(flow) || nf_ct_is_dying(flow->ct) || diff --git a/net/netfilter/nf_flow_table_offload.c b/net/netfilter/nf_flow_table_offload.c index 6757fd89c1f1..4365859220e6 100644 --- a/net/netfilter/nf_flow_table_offload.c +++ b/net/netfilter/nf_flow_table_offload.c @@ -995,6 +995,7 @@ static void flow_offload_work_del(struct flow_offload_work *offload) flow_offload_tuple_del(offload, FLOW_OFFLOAD_DIR_ORIGINAL); if (test_bit(NF_FLOW_HW_BIDIRECTIONAL, &offload->flow->flags)) flow_offload_tuple_del(offload, FLOW_OFFLOAD_DIR_REPLY); + set_bit(NF_FLOW_HW_DEAD, &offload->flow->flags); } static void flow_offload_tuple_stats(struct flow_offload_work *offload, @@ -1056,13 +1057,8 @@ static void flow_offload_work_handler(struct work_struct *work) default: WARN_ON_ONCE(1); } - - clear_bit(NF_FLOW_HW_PENDING, &offload->flow->flags); - if (offload->cmd == FLOW_CLS_DESTROY) { - /* Publish after the worker's last flow access. */ - smp_mb__before_atomic(); - set_bit(NF_FLOW_HW_DEAD, &offload->flow->flags); - } + smp_mb__before_atomic(); + clear_bit(NF_FLOW_PENDING, &offload->flow->flags); kfree(offload); } @@ -1089,12 +1085,12 @@ nf_flow_offload_work_alloc(struct nf_flowtable *flowtable, { struct flow_offload_work *offload; - if (test_and_set_bit(NF_FLOW_HW_PENDING, &flow->flags)) + if (test_and_set_bit(NF_FLOW_PENDING, &flow->flags)) return NULL; offload = kmalloc_obj(struct flow_offload_work, GFP_ATOMIC); if (!offload) { - clear_bit(NF_FLOW_HW_PENDING, &flow->flags); + clear_bit(NF_FLOW_PENDING, &flow->flags); return NULL; } diff --git a/net/sched/act_ct.c b/net/sched/act_ct.c index 55f3521edb4c..626c9a5af0ef 100644 --- a/net/sched/act_ct.c +++ b/net/sched/act_ct.c @@ -289,7 +289,7 @@ static bool tcf_ct_flow_is_outdated(const struct flow_offload *flow) { return test_bit(IPS_SEEN_REPLY_BIT, &flow->ct->status) && test_bit(IPS_HW_OFFLOAD_BIT, &flow->ct->status) && - !test_bit(NF_FLOW_HW_PENDING, &flow->flags) && + !test_bit(NF_FLOW_PENDING, &flow->flags) && !test_bit(NF_FLOW_HW_ESTABLISHED, &flow->flags); } From 3ae37eafd36694bb2d3227f60ac98fbf602470a2 Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Wed, 23 Sep 2026 20:46:18 +0200 Subject: [PATCH 672/694] netfilter: flowtable: restore ieee80211 forward path Before commit 871df5007eda ("netfilter: flowtable: bail out if forward path cannot be discovered"), there was a fallback to set up a forward path in case .ndo_fill_forward_path fails or DEV_PATH_MTK_WDMA was used. Such fallback was used by commit d787a3e38f01 ("mac80211: add support for .ndo_fill_forward_path"). One possibility is to handle DEV_PATH_MTK_WDMA from the flowtable forward path discovery. However, this is only used internally by drivers to retrieve mtk_wdma information to set up hardware offload. Felix decided to use the .fill_forward_path interface for this purpose due to the lack of a better interface at that time. Add a new DEV_PATH_IEEE80211 path which is offered if the new ieee80211 flag is set on in the struct net_device_path_ctx to restore the flowtable with a ieee80211 netdevice. Handle this new DEV_PATH_IEEE80211 path just like DEV_PATH_ETHERNET and DEV_PATH_DSA, ie. this is the last netdevice in the stack. This new ieee80211 flag is implicitly unset for mtk_ppe and airoha which call dev_fill_forward_path() to retrieve a DEV_PATH_MTK_WDMA path. Fixes: 871df5007eda ("netfilter: flowtable: bail out if forward path cannot be discovered") Signed-off-by: Pablo Neira Ayuso --- include/linux/netdevice.h | 3 +++ net/mac80211/iface.c | 7 +++++++ net/netfilter/nf_flow_table_path.c | 3 +++ 3 files changed, 13 insertions(+) diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index 87cafc932e9e..3cff2174dc03 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -887,6 +887,7 @@ enum net_device_path_type { DEV_PATH_DSA, DEV_PATH_MTK_WDMA, DEV_PATH_TUN, + DEV_PATH_IEEE80211, }; struct net_device_path { @@ -953,6 +954,8 @@ struct net_device_path_ctx { u16 id; __be16 proto; } vlan[NET_DEVICE_PATH_VLAN_MAX]; + + bool ieee80211; }; enum tc_setup_type { diff --git a/net/mac80211/iface.c b/net/mac80211/iface.c index 889c32fd8de1..c5584435fde9 100644 --- a/net/mac80211/iface.c +++ b/net/mac80211/iface.c @@ -1023,6 +1023,13 @@ static int ieee80211_netdev_fill_forward_path(struct net_device_path_ctx *ctx, struct sta_info *sta; int ret = -ENOENT; + if (ctx->ieee80211) { + path->type = DEV_PATH_IEEE80211; + path->dev = ctx->dev; + ctx->dev = NULL; + return 0; + } + sdata = IEEE80211_DEV_TO_SUB_IF(ctx->dev); local = sdata->local; diff --git a/net/netfilter/nf_flow_table_path.c b/net/netfilter/nf_flow_table_path.c index 1e55644f2edb..d90013685bf1 100644 --- a/net/netfilter/nf_flow_table_path.c +++ b/net/netfilter/nf_flow_table_path.c @@ -53,6 +53,7 @@ static int nft_dev_fill_forward_path(const struct dst_entry *dst_cache, struct net_device_path_ctx ctx = { .dev = dev, .ether_type = ether_type, + .ieee80211 = true, }; struct neighbour *n; u8 nud_state; @@ -114,6 +115,7 @@ static int nft_dev_path_info(struct net_device_path_stack *stack, path = &stack->path[i]; switch (path->type) { case DEV_PATH_ETHERNET: + case DEV_PATH_IEEE80211: case DEV_PATH_DSA: case DEV_PATH_VLAN: case DEV_PATH_PPPOE: @@ -123,6 +125,7 @@ static int nft_dev_path_info(struct net_device_path_stack *stack, memcpy(info->h_source, path->dev->dev_addr, ETH_ALEN); if (path->type == DEV_PATH_ETHERNET || + path->type == DEV_PATH_IEEE80211 || path->type == DEV_PATH_DSA) break; From 99b43ede9e355ba35244cc9470bf1819774ce39d Mon Sep 17 00:00:00 2001 From: Harshit Mogalapalli Date: Thu, 24 Sep 2026 23:39:36 -0700 Subject: [PATCH 673/694] net: microchip: vcap: stop scanning after deleting key field The only call outside the KUnit tests is in sparx5_tc_add_rule_copy(), which passes a fixed key list with no duplicate entries. Therefore, the potential double-free is not currently reachable in tree. However, vcap_filter_rule_keys() is exported and does not require its key list entries to be unique. If a future or external caller supplies the same key at another position that the inner loop visits, it calls list_del() and kfree() a second time on the freed field, leading to a potential double-free. Break the inner loop after removing a matching field. The outer safe list traversal continues to process the remaining rule fields. Fixes: 465a38a269e9 ("net: microchip: sparx5: Support for copying and modifying rules in the API") Reported-by: Dan Carpenter Closes: https://lore.kernel.org/all/ahs-lCAXXikkaHky@stanley.mountain/ Signed-off-by: Harshit Mogalapalli Reviewed-by: Daniel Machon Signed-off-by: David S. Miller --- drivers/net/ethernet/microchip/vcap/vcap_api.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/ethernet/microchip/vcap/vcap_api.c b/drivers/net/ethernet/microchip/vcap/vcap_api.c index 788c0728d763..0f8d98f2094d 100644 --- a/drivers/net/ethernet/microchip/vcap/vcap_api.c +++ b/drivers/net/ethernet/microchip/vcap/vcap_api.c @@ -3527,8 +3527,8 @@ int vcap_filter_rule_keys(struct vcap_rule *rule, if (key == keylist[idx]) { list_del(&ckf->ctrl.list); kfree(ckf); - idx++; err = 0; + break; } } } From 572af6872e520c7102e58362ef7cd7c2ed4342be Mon Sep 17 00:00:00 2001 From: Eric Biggers Date: Tue, 29 Sep 2026 15:27:52 -0700 Subject: [PATCH 674/694] crypto: aes - Fix undesired override of some optimized AES modes The new library APIs for AES encryption modes were wired up to the traditional crypto API via crypto/aes.c. However, for now the kernel is still in a transitional state where various architectures still have architecture-optimized implementations of AES modes in arch/*/crypto/, wired up to the traditional crypto API only. Because of that, the crypto/aes.c algorithms were given a cra_priority of only 110 to prevent them from overriding arch/*/crypto/ in the traditional crypto API. However, because of how the traditional crypto API works, the cra_priority trick doesn't work in cases where the relevant algorithm isn't directly implemented by arch/*/crypto/ but rather is provided by a template instance using other code in arch/*/crypto/. For example, x86 doesn't have its own "ccm(aes)" but rather relies on the "ccm" template constructing it from the x86-optimized "ctr(aes)". The existence of the library-based "ccm(aes)" prevents that, even though its priority is lower than what the template would produce. Thus, "ccm(aes)" ends up using the slower single-block AES code. Therefore, skip wiring up the relevant library-based code to the traditional crypto API on architectures where this problem can occur, as determined by what exists in arch/*/crypto/ for each architecture. This is ugly, but it's also temporary: these conditions will go away as architecture-optimized implementations of AES modes are migrated into the library. But until then, we need to prevent performance regressions by ensuring that the optimized code continues to be used. Fixes: 20df21a482aa ("crypto: aes - Add CBC and CBC-CTS support using library") Fixes: 8ca62072faa1 ("crypto: aes - Add GCM support using library") Fixes: f70ad727d1d6 ("crypto: aes - Add CCM support using library") Fixes: 94efa0c9fb36 ("crypto: aes - Add XTS support using library") Closes: https://github.com/sparclinux/issues/issues/106 Acked-by: Ard Biesheuvel Tested-by: Stian Halseth Link: https://patch.msgid.link/20260929222752.36427-1-ebiggers@kernel.org Signed-off-by: Eric Biggers --- crypto/aes.c | 51 +++++++++++++++++++++++++++++++++++++++++++++++---- 1 file changed, 47 insertions(+), 4 deletions(-) diff --git a/crypto/aes.c b/crypto/aes.c index 94791f481e98..51eee78396ea 100644 --- a/crypto/aes.c +++ b/crypto/aes.c @@ -637,7 +637,18 @@ static struct skcipher_alg skcipher_algs[] = { .decrypt = crypto_aes_cbc_decrypt, }, #endif -#if IS_ENABLED(CONFIG_CRYPTO_CTS) + /* + * Don't register library-based "cts(cbc(aes))" on architectures where + * it might block a "better" implementation from being instantiated via + * the "cts" template. These exclusions are temporary and will go away + * as the arch-optimized AES code is migrated into the library. + */ +#if IS_ENABLED(CONFIG_CRYPTO_CTS) && \ + !(IS_ENABLED(CONFIG_ARM) || \ + IS_ENABLED(CONFIG_ARM64) || \ + IS_ENABLED(CONFIG_PPC) || \ + IS_ENABLED(CONFIG_S390) || \ + IS_ENABLED(CONFIG_SPARC)) { .base.cra_name = "cts(cbc(aes))", .base.cra_driver_name = "cts-cbc-aes-lib", @@ -687,7 +698,13 @@ static struct skcipher_alg skcipher_algs[] = { .decrypt = crypto_aes_xctr_crypt, }, #endif -#if IS_ENABLED(CONFIG_CRYPTO_XTS) + /* + * Don't register library-based "xts(aes)" on architectures where it + * might block a "better" implementation from being instantiated via the + * "xts" template. This exclusion is temporary and will go away when + * the library AES-XTS is optimized for SPARC. + */ +#if IS_ENABLED(CONFIG_CRYPTO_XTS) && !IS_ENABLED(CONFIG_SPARC) { .base.cra_name = "xts(aes)", .base.cra_driver_name = "xts-aes-lib", @@ -980,7 +997,20 @@ static __maybe_unused int crypto_aes_ccm_decrypt(struct aead_request *req) } static struct aead_alg aead_algs[] = { -#if IS_ENABLED(CONFIG_CRYPTO_GCM) + /* + * Don't register library-based "gcm(aes)" and "rfc4106(gcm(aes))" on + * architectures where they might block a "better" implementation from + * being instantiated via the "gcm" and "rfc4106" templates. These + * exclusions are temporary and will go away as the arch-optimized AES + * code is migrated into the library. + */ +#if IS_ENABLED(CONFIG_CRYPTO_GCM) && \ + !(IS_ENABLED(CONFIG_ARM) || \ + IS_ENABLED(CONFIG_ARM64) || \ + IS_ENABLED(CONFIG_PPC) || \ + IS_ENABLED(CONFIG_RISCV) || \ + IS_ENABLED(CONFIG_S390) || \ + IS_ENABLED(CONFIG_SPARC)) { .base.cra_name = "gcm(aes)", .base.cra_driver_name = "gcm-aes-lib", @@ -1012,7 +1042,20 @@ static struct aead_alg aead_algs[] = { .chunksize = AES_BLOCK_SIZE, }, #endif /* CONFIG_CRYPTO_GCM */ -#if IS_ENABLED(CONFIG_CRYPTO_CCM) + /* + * Don't register library-based "ccm(aes)" on architectures where it + * might block a "better" implementation from being instantiated via the + * "ccm" template. These exclusions are temporary and will go away as + * the arch-optimized AES code is migrated into the library. + */ +#if IS_ENABLED(CONFIG_CRYPTO_CCM) && \ + !(IS_ENABLED(CONFIG_ARM) || \ + IS_ENABLED(CONFIG_ARM64) || \ + IS_ENABLED(CONFIG_PPC) || \ + IS_ENABLED(CONFIG_RISCV) || \ + IS_ENABLED(CONFIG_S390) || \ + IS_ENABLED(CONFIG_SPARC) || \ + IS_ENABLED(CONFIG_X86)) { .base.cra_name = "ccm(aes)", .base.cra_driver_name = "ccm-aes-lib", From 77f21a59e8cf7efcc1c4b3e9176e8b99993a7ae4 Mon Sep 17 00:00:00 2001 From: Yuho Choi Date: Wed, 23 Sep 2026 21:13:50 -0400 Subject: [PATCH 675/694] of: Put coreboot node after compatibility check of_find_compatible_node() returns a referenced device node, but EXCLUDED_DEFAULT_CELLS_PLATFORMS used the result only as a boolean and discarded the reference. This leaked a reference each time the address-cells or size-cells warning condition was evaluated on a coreboot system. Use a helper that checks for the coreboot node with a scoped reference, so the reference is dropped once the check is done. Fixes: 8600058ba28a7 ("of: Add coreboot firmware to excluded default cells list") Signed-off-by: Yuho Choi Link: https://patch.msgid.link/20260924011358.734722-1-oss.patchbox@gmail.com Signed-off-by: Rob Herring (Arm) --- drivers/of/base.c | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-) diff --git a/drivers/of/base.c b/drivers/of/base.c index 378703dbc11f..57313055d9ed 100644 --- a/drivers/of/base.c +++ b/drivers/of/base.c @@ -95,9 +95,17 @@ static bool __of_node_is_type(const struct device_node *np, const char *type) return !strcmp(match, type); } +static bool of_coreboot_present(void) +{ + struct device_node *np __free(device_node) = + of_find_compatible_node(NULL, NULL, "coreboot"); + + return np; +} + #define EXCLUDED_DEFAULT_CELLS_PLATFORMS ( \ IS_ENABLED(CONFIG_SPARC) || \ - of_find_compatible_node(NULL, NULL, "coreboot") \ + of_coreboot_present() \ ) int of_bus_n_addr_cells(struct device_node *np) From 30724547b221e0e670ddfef140fc7d816b2aaec6 Mon Sep 17 00:00:00 2001 From: Wentao Liang Date: Thu, 17 Sep 2026 12:42:47 +0000 Subject: [PATCH 676/694] of/irq: Fix remaining refcount leaks in of_irq_init() Successfully initialized controllers are moved to intc_parent_list, but the final cleanup only frees the descriptors. The device node references taken with of_node_get() and the interrupt parent references are therefore leaked on the normal return path. Drop both before kfree(), as the intc_desc_list cleanup already does. Fixes: 8363ccb917c6 ("of/irq: add missing of_node_put") Cc: stable@vger.kernel.org Signed-off-by: Wentao Liang Link: https://patch.msgid.link/20260917124247.2151674-1-vulab@iscas.ac.cn Signed-off-by: Rob Herring (Arm) --- drivers/of/irq.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/of/irq.c b/drivers/of/irq.c index ec035367c950..ef1ed9743907 100644 --- a/drivers/of/irq.c +++ b/drivers/of/irq.c @@ -754,6 +754,8 @@ void __init of_irq_init(const struct of_device_id *matches) list_for_each_entry_safe(desc, temp_desc, &intc_parent_list, list) { list_del(&desc->list); + of_node_put(desc->interrupt_parent); + of_node_put(desc->dev); kfree(desc); } err: From 5642b1648d5c4561d8ed7a410d0a7017f706604d Mon Sep 17 00:00:00 2001 From: Quentin Freimanis Date: Sun, 27 Sep 2026 21:50:12 -0700 Subject: [PATCH 677/694] net: sparx5: make ports inherit the switch base mac address type When the switch uses a random base MAC address it does not mark each port's address as random: $ dmesg | grep "MAC addr" sparx5-switch e00c0000.switch: MAC addr was not set, use random MAC $ cat /sys/class/net/eth0/addr_assign_type 0 0 is NET_ADDR_PERM, so userspace thinks this is a stable address. With a random MAC reported as NET_ADDR_PERM, systemd's MACAddressPolicy=persistent [1] leaves it unchanged, so a DHCP client gets a new IP address every boot and a static reservation can't be used. Reporting it as NET_ADDR_RANDOM makes systemd replace it with a stable address derived from the interface name and machine-id. Fixes: f3cad2611a77 ("net: sparx5: add hostmode with phylink support") Signed-off-by: Quentin Freimanis Reviewed-by: Simon Horman Reviewed-by: Daniel Machon Link: https://patch.msgid.link/20260928045012.30178-1-quentin@q-lab.dev Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/microchip/sparx5/sparx5_main.c | 1 + drivers/net/ethernet/microchip/sparx5/sparx5_main.h | 1 + drivers/net/ethernet/microchip/sparx5/sparx5_netdev.c | 1 + 3 files changed, 3 insertions(+) diff --git a/drivers/net/ethernet/microchip/sparx5/sparx5_main.c b/drivers/net/ethernet/microchip/sparx5/sparx5_main.c index dad713e9ddd5..b530fb6d3d0c 100644 --- a/drivers/net/ethernet/microchip/sparx5/sparx5_main.c +++ b/drivers/net/ethernet/microchip/sparx5/sparx5_main.c @@ -898,6 +898,7 @@ static int mchp_sparx5_probe(struct platform_device *pdev) dev_info(sparx5->dev, "MAC addr was not set, use random MAC\n"); eth_random_addr(sparx5->base_mac); sparx5->base_mac[5] = 0; + sparx5->base_mac_assign_type = NET_ADDR_RANDOM; } sparx5->fdma_irq = platform_get_irq_byname(sparx5->pdev, "fdma"); diff --git a/drivers/net/ethernet/microchip/sparx5/sparx5_main.h b/drivers/net/ethernet/microchip/sparx5/sparx5_main.h index eb57b86fbe22..0f7d8440f614 100644 --- a/drivers/net/ethernet/microchip/sparx5/sparx5_main.h +++ b/drivers/net/ethernet/microchip/sparx5/sparx5_main.h @@ -392,6 +392,7 @@ struct sparx5 { struct notifier_block switchdev_blocking_nb; /* Switch state */ u8 base_mac[ETH_ALEN]; + u8 base_mac_assign_type; /* Associated bridge device (when bridged) */ struct net_device *hw_bridge_dev; /* Bridged interfaces */ diff --git a/drivers/net/ethernet/microchip/sparx5/sparx5_netdev.c b/drivers/net/ethernet/microchip/sparx5/sparx5_netdev.c index 4dedf5d92641..b3725cf2dd44 100644 --- a/drivers/net/ethernet/microchip/sparx5/sparx5_netdev.c +++ b/drivers/net/ethernet/microchip/sparx5/sparx5_netdev.c @@ -294,6 +294,7 @@ struct net_device *sparx5_create_netdev(struct sparx5 *sparx5, u32 portno) ndev->ethtool_ops = &sparx5_ethtool_ops; eth_hw_addr_gen(ndev, sparx5->base_mac, portno + 1); + ndev->addr_assign_type = sparx5->base_mac_assign_type; return ndev; } From 4f1da630d13de0370e2f6361f961f1c8b384af1c Mon Sep 17 00:00:00 2001 From: Jun Yang Date: Sat, 26 Sep 2026 17:55:15 +0800 Subject: [PATCH 678/694] sctp: check RCV_SHUTDOWN after the sendmsg connect wait sctp_wait_for_connect() drops the socket lock while it sleeps. An out-of-the-blue ABORT can then be processed from the socket backlog and unlink the association. If a concurrent shutdown(fd, SHUT_RD) sets RCV_SHUTDOWN, the waiter breaks with err == 0 before checking asoc->base.dead. Its final sctp_association_put() can then free the association, leaving sctp_sendmsg_to_asoc() to continue with a dangling pointer. Check RCV_SHUTDOWN along with the wait error in sctp_sendmsg_to_asoc() before using the association again. The check only accesses the socket, so it needs no additional association reference. Return the existing -ESRCH so that sctp_sendmsg() skips freeing a new association that may already have been destroyed. Keep sctp_wait_for_connect() unchanged to preserve its behavior for the connect() caller. Fixes: 668c9beb9020 ("sctp: implement assign_number for sctp_stream_interleave") Cc: stable@vger.kernel.org Reported-by: TencentOS Corvus AI Signed-off-by: Jun Yang Acked-by: Xin Long Link: https://patch.msgid.link/20260926095606.68601-1-juny24602@gmail.com Signed-off-by: Jakub Kicinski --- net/sctp/socket.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/sctp/socket.c b/net/sctp/socket.c index c7b9e325ec1c..4652fd90d9a6 100644 --- a/net/sctp/socket.c +++ b/net/sctp/socket.c @@ -1844,7 +1844,7 @@ static int sctp_sendmsg_to_asoc(struct sctp_association *asoc, if (asoc->ep->intl_enable) { timeo = sock_sndtimeo(sk, 0); err = sctp_wait_for_connect(asoc, &timeo); - if (err) { + if (err || sk->sk_shutdown & RCV_SHUTDOWN) { err = -ESRCH; goto err; } From 9e792901418d814295674d1f53d6b615fb5c15c8 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Jari=20Sepp=C3=A4l=C3=A4?= Date: Mon, 28 Sep 2026 11:47:11 +0300 Subject: [PATCH 679/694] octeontx2-pf: Fix RSS indirection table size MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The conversion to the dedicated RSS context operations replaced the pointer to struct otx2_rss_ctx with an inline u32 ind_tbl[] array. However, otx2_rss_init() still uses sizeof(*rss->ind_tbl) to set rss_size. This now yields the size of one u32 (4), rather than the 256 entries in the indirection table. The RSS initialization and hardware programming loops use rss_size, so only four entries are initialized and programmed despite the NIX LF being allocated a 256-entry table. On an OCTEON CN102 with eight RX queues, the table contained 0, 1, 2, 3 followed by zeros. Traffic was concentrated on queue 0 and ethtool -X equal 8 did not correct the table. Use ARRAY_SIZE() to count the indirection table entries. On the same hardware, this restores the full table repeating queue numbers 0 through 7, and bidirectional multi-flow traffic increments multiple RX queues. Runtime testing on an Asterfusion ET2500 (OCTEON CN102 A0). Fixes: 62e01d8c4170 ("eth: otx2: migrate to the *_rxfh_context ops") Cc: stable@vger.kernel.org Closes: https://lore.kernel.org/netdev/361df563-e29e-43bf-8806-24bdfefed33d@nocloud.org/ Signed-off-by: Jari Seppälä Reviewed-by: Ratheesh Kannoth Link: https://patch.msgid.link/20260928084711.2462571-1-jase@nocloud.org Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/marvell/octeontx2/nic/otx2_common.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_common.c b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_common.c index 0f3f405d9f1b..b11389a841d1 100644 --- a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_common.c +++ b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_common.c @@ -394,7 +394,7 @@ int otx2_rss_init(struct otx2_nic *pfvf) struct otx2_rss_info *rss = &pfvf->hw.rss_info; int idx, ret = 0; - rss->rss_size = sizeof(*rss->ind_tbl); + rss->rss_size = ARRAY_SIZE(rss->ind_tbl); /* Init RSS key if it is not setup already */ if (!rss->enable) From 3cdeaef1754ab8155cdd828c958ada3d36e9ad9a Mon Sep 17 00:00:00 2001 From: Oleksij Rempel Date: Fri, 25 Sep 2026 09:10:40 +0200 Subject: [PATCH 680/694] r8169: disable EEE on RTL8168h/8111h Force EEE off on the RTL8168h/8111h (RTL_GIGA_MAC_VER_46) at PHY connect with phy_disable_eee(). Commit 202fef9bbbf5 ("net: phy: realtek: fix EEE advertisement write on the internal PHY MMD path") made EEE actually advertise on the generic Realtek PHY; the write had been a no-op before. On the RTL8168h that un-masks a latent defect: once EEE negotiates, RX silently stalls after ~7-20 minutes. The carrier stays up, no counter or dmesg moves, and only "ip link set down/up" recovers it; disabling EEE keeps the link stable. Root-causing the RTL8168h LPI/RX path needs hardware not available now, so disable EEE for this version. Use phy_disable_eee() rather than dropping the version from rtl_supports_eee(): the latter also skips rtl_enable_tx_lpi(), whose disable branch clears the MAC TX-LPI bits (ERI 0x1b0[1:0]) on link up. rtl_hw_start_8168h_1() does not clear them (the RTL8402/RTL8106e init does), so a warm reboot from an EEE-active state could otherwise leave TX-LPI asserted while the PHY no longer negotiates EEE. Keeping the version EEE-capable preserves that clear path. This also covers the RTL8168M, which shares RTL_GIGA_MAC_VER_46. Reported-by: Dennis Piecha Closes: https://lore.kernel.org/all/353419280.954808.1790290283530@mail.yahoo.com/ Fixes: 202fef9bbbf5 ("net: phy: realtek: fix EEE advertisement write on the internal PHY MMD path") Cc: stable@vger.kernel.org Signed-off-by: Oleksij Rempel Link: https://patch.msgid.link/20260925071040.2137178-1-o.rempel@pengutronix.de Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/realtek/r8169_main.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/drivers/net/ethernet/realtek/r8169_main.c b/drivers/net/ethernet/realtek/r8169_main.c index ec4fc21fa21f..ee1bc9bd0447 100644 --- a/drivers/net/ethernet/realtek/r8169_main.c +++ b/drivers/net/ethernet/realtek/r8169_main.c @@ -5484,6 +5484,12 @@ static int r8169_mdio_register(struct rtl8169_private *tp) phy_disable_eee_mode(tp->phydev, ETHTOOL_LINK_MODE_2500baseT_Full_BIT); + /* Keep EEE off but the version stays EEE-capable so that link-up + * still clears the MAC TX-LPI bits. + */ + if (tp->mac_version == RTL_GIGA_MAC_VER_46) + phy_disable_eee(tp->phydev); + /* PHY will be woken up in rtl_open() */ phy_suspend(tp->phydev); From 7b97273cbebe9dd4fddc457f0c4f2a124d0e3996 Mon Sep 17 00:00:00 2001 From: Prathamesh Deshpande Date: Sat, 26 Sep 2026 17:53:57 +0100 Subject: [PATCH 681/694] net/mlx5e: Fix AF_XDP TX timestamp teardown NULL dereference During XSK TX queue teardown, outstanding descriptors are completed without a CQE. If one requested a TX timestamp, the completion path passes that NULL CQE to mlx5e_xsk_fill_timestamp(), which dereferences it. Return zero when no CQE is available, indicating that teardown did not produce a TX timestamp. Fixes: ec706a860eba ("net/mlx5e: Implement AF_XDP TX timestamp and checksum offload") Signed-off-by: Prathamesh Deshpande Reviewed-by: Tariq Toukan Link: https://patch.msgid.link/20260926165402.5902-1-prathameshdeshpande7@gmail.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c b/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c index 77ea51bfbaae..2cca20e1b9b7 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c @@ -293,6 +293,9 @@ static u64 mlx5e_xsk_fill_timestamp(void *_priv) struct mlx5e_xsk_tx_complete *priv = _priv; u64 ts; + if (!priv->cqe) + return 0; + ts = get_cqe_ts(priv->cqe); if (mlx5_is_real_time_rq(priv->cq->mdev) || mlx5_is_real_time_sq(priv->cq->mdev)) From 8ec454dcd00ce3375119111c6394964455b36d0d Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 28 Sep 2026 19:45:24 +0000 Subject: [PATCH 682/694] ipv6: sr: use skb_get_hash_net() in seg6_make_flowlabel() Since commit d58e468b1112 ("flow_dissector: implements flow dissector BPF hook") __skb_flow_dissect() needs a net pointer, either from skb->dev, skb->sk, or since commit 3cbf4ffba5ee ("net: plumb network namespace into __skb_flow_dissect") a caller provided pointer. syzbot was able to reach seg6_make_flowlabel() with an skb having neither skb->dev nor skb->sk set: a TIPC UDP bearer sends a discovery message through an IPv4 route using seg6 encap, while net.ipv6.seg6_flowlabel is set to 1. seg6_make_flowlabel() already has a net pointer, use skb_get_hash_net(). WARNING: net/core/flow_dissector.c:1131 at __skb_flow_dissect+0x910/0x5368 net/core/flow_dissector.c:1126, CPU#0: syz.0.17/4930 Call trace: __skb_flow_dissect+0x910/0x5368 net/core/flow_dissector.c:1126 (P) __skb_get_hash_net+0xe0/0x29c net/core/flow_dissector.c:1903 skb_get_hash include/linux/skbuff.h:1663 [inline] seg6_make_flowlabel+0xcc/0x1ec net/ipv6/seg6_iptunnel.c:132 __seg6_do_srh_encap+0x320/0xbf4 net/ipv6/seg6_iptunnel.c:161 seg6_do_srh+0x4b4/0xa44 net/ipv6/seg6_iptunnel.c:431 seg6_output_core+0x164/0x688 net/ipv6/seg6_iptunnel.c:681 seg6_output+0x44/0x1ac net/ipv6/seg6_iptunnel.c:744 lwtunnel_output+0x3d0/0x664 net/core/lwtunnel.c:356 dst_output include/net/dst.h:470 [inline] ip_local_out+0x110/0x148 net/ipv4/ip_output.c:131 iptunnel_xmit+0x50c/0xd38 net/ipv4/ip_tunnel_core.c:97 udp_tunnel_xmit_skb+0x220/0x348 net/ipv4/udp_tunnel_core.c:187 tipc_udp_xmit+0x75c/0x9c0 net/tipc/udp_media.c:202 tipc_udp_send_msg+0x214/0x374 net/tipc/udp_media.c:274 tipc_bearer_xmit_skb+0x260/0x3b0 net/tipc/bearer.c:576 tipc_enable_bearer net/tipc/bearer.c:366 [inline] __tipc_nl_bearer_enable+0xc90/0xfb0 net/tipc/bearer.c:1048 tipc_nl_bearer_enable+0x2c/0x48 net/tipc/bearer.c:1057 genl_family_rcv_msg_doit+0x1e4/0x2d4 net/netlink/genetlink.c:1114 Fixes: d58e468b1112 ("flow_dissector: implements flow dissector BPF hook") Reported-by: syzbot+9408fbe0e6452a12e9ab@syzkaller.appspotmail.com Closes: https://lore.kernel.org/netdev/6abac2eb.3654fce1.1bec97.0000.GAE@google.com/ Signed-off-by: Eric Dumazet Reviewed-by: Xuanqiang Luo Reviewed-by: Ido Schimmel Link: https://patch.msgid.link/20260928194524.3617299-1-edumazet@kernel.org Signed-off-by: Jakub Kicinski --- net/ipv6/seg6_iptunnel.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/ipv6/seg6_iptunnel.c b/net/ipv6/seg6_iptunnel.c index 61c6a27bf202..7cbe0685f8bb 100644 --- a/net/ipv6/seg6_iptunnel.c +++ b/net/ipv6/seg6_iptunnel.c @@ -129,7 +129,7 @@ static __be32 seg6_make_flowlabel(struct net *net, struct sk_buff *skb, u32 hash; if (do_flowlabel > 0) { - hash = skb_get_hash(skb); + hash = skb_get_hash_net(net, skb); hash = rol32(hash, 16); flowlabel = (__force __be32)hash & IPV6_FLOWLABEL_MASK; } else if (!do_flowlabel && skb->protocol == htons(ETH_P_IPV6)) { From 8f1c2a10500a84a621ff13073f96deea0753ed42 Mon Sep 17 00:00:00 2001 From: Lorenzo Bianconi Date: Tue, 29 Sep 2026 15:43:03 +0200 Subject: [PATCH 683/694] net: mvneta: clear XDP pfmemalloc flag between frames MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit mvneta_swbm_add_rx_fragment() sets XDP_FLAGS_FRAGS_PF_MEMALLOC on the xdp_buff when a fragment page is a pfmemalloc one (page under memory pressure). The xdp_buff is reused for the next frame, but only the XDP_FLAGS_HAS_FRAGS bit was cleared at frame start, so the pfmemalloc bit leaked from one frame into the following ones. mvneta_swbm_build_skb() propagates the flag to skb->pfmemalloc through xdp_update_skb_frags_info(), so the skb of a subsequent fragmented frame could be wrongly marked as pfmemalloc even if none of its pages are under pressure. Clear all the xdp_buff flags in mvneta_swbm_rx_frame(), which is invoked for each new frame, instead of just the XDP_FLAGS_HAS_FRAGS bit. Fixes: ed7a58cb40bd ("net: marvell: rely on xdp_update_skb_shared_info utility routine") Reviewed-by: Simon Horman Signed-off-by: Lorenzo Bianconi Reviewed-by: Toke Høiland-Jørgensen Link: https://patch.msgid.link/20260929-mvneta-xdp-clear-frag-fix-v4-1-1e63b25eeed8@oss.qualcomm.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/marvell/mvneta.c | 2 +- include/net/xdp.h | 5 +++++ 2 files changed, 6 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/marvell/mvneta.c b/drivers/net/ethernet/marvell/mvneta.c index 543e566425c1..4f479b6f731e 100644 --- a/drivers/net/ethernet/marvell/mvneta.c +++ b/drivers/net/ethernet/marvell/mvneta.c @@ -2340,7 +2340,7 @@ mvneta_swbm_rx_frame(struct mvneta_port *pp, /* Prefetch header */ prefetch(data); - xdp_buff_clear_frags_flag(xdp); + xdp_reinit_buff(xdp); xdp_prepare_buff(xdp, data, pp->rx_offset_correction + MVNETA_MH_SIZE, data_len, true); } diff --git a/include/net/xdp.h b/include/net/xdp.h index aa742f413c35..07231adfb5f8 100644 --- a/include/net/xdp.h +++ b/include/net/xdp.h @@ -106,6 +106,11 @@ struct xdp_buff { }; }; +static __always_inline void xdp_reinit_buff(struct xdp_buff *xdp) +{ + xdp->flags = 0; +} + static __always_inline bool xdp_buff_has_frags(const struct xdp_buff *xdp) { return !!(xdp->flags & XDP_FLAGS_HAS_FRAGS); From 7375d38364a9aa66fb31716bcefef38aecad75d8 Mon Sep 17 00:00:00 2001 From: Reinhard Speyerer Date: Fri, 25 Sep 2026 15:14:32 +0200 Subject: [PATCH 684/694] net: usb: qmi_wwan: add Rolling Wireless RN947R Add Rolling Wireless RN947R 0x9300 composition: DIAG + ADB + NMEA + MODEM + RMNET + QDSS + ADPL T: Bus=01 Lev=01 Prnt=01 Port=02 Cnt=02 Dev#= 6 Spd=480 MxCh= 0 D: Ver= 2.10 Cls=ef(misc ) Sub=02 Prot=01 MxPS=64 #Cfgs= 1 P: Vendor=33f8 ProdID=9300 Rev= 0.00 S: Manufacturer=Rolling Wireless S: Product=RN947R S: SerialNumber=xxxxxxxxxxx C:* #Ifs= 7 Cfg#= 1 Atr=a0 MxPwr=500mA I:* If#= 0 Alt= 0 #EPs= 2 Cls=ff(vend.) Sub=ff Prot=30 Driver=option E: Ad=01(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=81(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms I:* If#= 1 Alt= 0 #EPs= 2 Cls=ff(vend.) Sub=42 Prot=01 Driver=usbfs E: Ad=02(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=82(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms I:* If#= 2 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=ff Prot=40 Driver=option E: Ad=84(I) Atr=03(Int.) MxPS= 10 Ivl=32ms E: Ad=83(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=03(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms I:* If#= 3 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=ff Prot=40 Driver=option E: Ad=86(I) Atr=03(Int.) MxPS= 10 Ivl=32ms E: Ad=85(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=04(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms I:* If#= 8 Alt= 0 #EPs= 3 Cls=ff(vend.) Sub=ff Prot=50 Driver=qmi_wwan E: Ad=88(I) Atr=03(Int.) MxPS= 8 Ivl=32ms E: Ad=87(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms E: Ad=05(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms I:* If#=21 Alt= 0 #EPs= 1 Cls=ff(vend.) Sub=ff Prot=70 Driver=(none) E: Ad=89(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms I:* If#=22 Alt= 0 #EPs= 1 Cls=ff(vend.) Sub=ff Prot=80 Driver=(none) E: Ad=8a(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms Signed-off-by: Reinhard Speyerer Link: https://patch.msgid.link/arZzuOOvLUZtdOlv@arcor.de Signed-off-by: Jakub Kicinski --- drivers/net/usb/qmi_wwan.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/net/usb/qmi_wwan.c b/drivers/net/usb/qmi_wwan.c index 0e2ab567dd40..a44d1b9756a1 100644 --- a/drivers/net/usb/qmi_wwan.c +++ b/drivers/net/usb/qmi_wwan.c @@ -1455,6 +1455,7 @@ static const struct usb_device_id products[] = { {QMI_QUIRK_SET_DTR(0x1546, 0x1312, 4)}, /* u-blox LARA-R6 01B */ {QMI_QUIRK_SET_DTR(0x1546, 0x1342, 4)}, /* u-blox LARA-L6 */ {QMI_QUIRK_SET_DTR(0x33f8, 0x0104, 4)}, /* Rolling RW101 RMNET */ + {QMI_QUIRK_SET_DTR(0x33f8, 0x9300, 8)}, /* Rolling RN947R RMNET */ {QMI_FIXED_INTF(0x2dee, 0x4d22, 5)}, /* MeiG Smart SRM825L */ {QMI_QUIRK_SET_DTR(0x2dee, 0x4d63, 0)}, /* MeiG SRM813Q w/ Modem(PPP) */ {QMI_QUIRK_SET_DTR(0x2dee, 0x4d64, 0)}, /* MeiG SRM813Q w/ NMEA */ From 6e0022b5ae3dc5b833af0fcc1578dc20a1d6bc71 Mon Sep 17 00:00:00 2001 From: Bartosz Golaszewski Date: Wed, 23 Sep 2026 17:14:35 +0200 Subject: [PATCH 685/694] net: phy: aquantia: fix system interface type not updated in forced mode aqr_gen1_read_status() decodes the MDIO_PHYXS_VEND_IF_STATUS register to determine which SerDes interface the PHY is currently using on its system side and stores the result in phydev->interface. phylink relies on this value to configure the MAC. The autoneg == AUTONEG_DISABLE check is not correct: MDIO_PHYXS_VEND_IF_STATUS is set by the PHY firmware based on the negotiated link speed, not based on whether autoneg was used to reach it. When the link comes up at 1G in forced mode, the register correctly reads SGMII, but the early return prevents phydev->interface from being updated. It stays at whatever value it held before (typically 2500BASE-X from the initial autoneg run), so phylink configures the MAC for the wrong interface and the link cannot come up. Remove the autoneg guard so that the system interface type is always decoded when the link is up. Cc: stable@vger.kernel.org Fixes: 110a2432c520 ("net: phy: aquantia: add downshift support") Signed-off-by: Bartosz Golaszewski Link: https://patch.msgid.link/20260923-qcom-sa8255p-emac-v15-1-e82f33720737@oss.qualcomm.com Signed-off-by: Paolo Abeni --- drivers/net/phy/aquantia/aquantia_main.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/net/phy/aquantia/aquantia_main.c b/drivers/net/phy/aquantia/aquantia_main.c index e95d391a6dc2..096407bebd52 100644 --- a/drivers/net/phy/aquantia/aquantia_main.c +++ b/drivers/net/phy/aquantia/aquantia_main.c @@ -544,7 +544,7 @@ static int aqr_gen1_read_status(struct phy_device *phydev) if (ret) return ret; - if (!phydev->link || phydev->autoneg == AUTONEG_DISABLE) + if (!phydev->link) return 0; /* The status register is not immediately correct on line side link up. @@ -594,6 +594,9 @@ static int aqr_gen1_read_status(struct phy_device *phydev) phydev->interface = aqr_translate_interface(phydev, interface); + if (phydev->autoneg == AUTONEG_DISABLE) + return 0; + /* Read rate from vendor register */ return aqr_gen1_read_rate(phydev); } From 71cc2c67fb8f8d5aa8154eb482e9846d2214f11b Mon Sep 17 00:00:00 2001 From: Mostafa Saleh Date: Tue, 29 Sep 2026 20:02:09 +0000 Subject: [PATCH 686/694] KVM: arm64: Use stage-1 leaf size for VM_PFNMAP When commit 2aa53d68cee6 ("KVM: arm64: Try stage2 block mapping for host device MMIO") added VM_PFNMAP support to get_vma_page_shift(), stage-1 page tables did not support huge PFNMAP (as in VFIO-PCI vfio_pci_mmap_huge_fault()) and transparent_hugepage_adjust() was unsafe for MMIO as it dereferenced struct page. Since commit 6011cf68c885 ("KVM: arm64: Walk userspace page tables to compute the THP mapping size"), transparent_hugepage_adjust() instead walks the host stage-1 page tables via get_user_mapping_size() without touching struct page. Meanwhile, commit 3e509c9b03f9 ("mm/arm64: support large pfn mappings") enabled stage-1 huge PFNMAP. So. we can drop the VMA-based VM_PFNMAP size calculation in get_vma_page_shift() and let transparent_hugepage_adjust() derive the stage-2 mapping size directly from the populated stage-1 leaf for non-cacheable mappings as well. Assisted-by: LLM Cc: stable@vger.kernel.org Fixes: 2aa53d68cee6 ("KVM: arm64: Try stage2 block mapping for host device MMIO") Reviewed-by: Marc Zyngier Signed-off-by: Mostafa Saleh Signed-off-by: Oliver Upton --- arch/arm64/kvm/mmu.c | 47 ++++++-------------------------------------- 1 file changed, 6 insertions(+), 41 deletions(-) diff --git a/arch/arm64/kvm/mmu.c b/arch/arm64/kvm/mmu.c index 2d44cd6a5aed..bf6d6526639f 100644 --- a/arch/arm64/kvm/mmu.c +++ b/arch/arm64/kvm/mmu.c @@ -1468,32 +1468,11 @@ transparent_hugepage_adjust(struct kvm *kvm, struct kvm_memory_slot *memslot, return PAGE_SIZE; } -static int get_vma_page_shift(struct vm_area_struct *vma, unsigned long hva) +static int get_vma_page_shift(struct vm_area_struct *vma) { - unsigned long pa; - - if (is_vm_hugetlb_page(vma) && !(vma->vm_flags & VM_PFNMAP)) + if (is_vm_hugetlb_page(vma)) return huge_page_shift(hstate_vma(vma)); - if (!(vma->vm_flags & VM_PFNMAP)) - return PAGE_SHIFT; - - VM_BUG_ON(is_vm_hugetlb_page(vma)); - - pa = (vma->vm_pgoff << PAGE_SHIFT) + (hva - vma->vm_start); - -#ifndef __PAGETABLE_PMD_FOLDED - if ((hva & (PUD_SIZE - 1)) == (pa & (PUD_SIZE - 1)) && - ALIGN_DOWN(hva, PUD_SIZE) >= vma->vm_start && - ALIGN(hva, PUD_SIZE) <= vma->vm_end) - return PUD_SHIFT; -#endif - - if ((hva & (PMD_SIZE - 1)) == (pa & (PMD_SIZE - 1)) && - ALIGN_DOWN(hva, PMD_SIZE) >= vma->vm_start && - ALIGN(hva, PMD_SIZE) <= vma->vm_end) - return PMD_SHIFT; - return PAGE_SHIFT; } @@ -1794,7 +1773,7 @@ static short kvm_s2_resolve_vma_size(const struct kvm_s2_fault_desc *s2fd, vma_shift = PAGE_SHIFT; } else { s2vi->max_map_size = PUD_SIZE; - vma_shift = get_vma_page_shift(vma, s2fd->hva); + vma_shift = get_vma_page_shift(vma); } switch (vma_shift) { @@ -1952,16 +1931,6 @@ static int kvm_s2_fault_pin_pfn(const struct kvm_s2_fault_desc *s2fd, return -EFAULT; } } else { - /* - * If the page was identified as device early by looking at - * the VMA flags, vma_pagesize is already representing the - * largest quantity we can map. If instead it was mapped - * via __kvm_faultin_pfn(), vma_pagesize is set to PAGE_SIZE - * and must not be upgraded. - * - * In both cases, we don't let transparent_hugepage_adjust() - * change things at the last minute. - */ s2vi->map_non_cacheable = true; } @@ -2051,10 +2020,10 @@ static int kvm_s2_fault_map(const struct kvm_s2_fault_desc *s2fd, /* * If we are not forced to use page mapping, check if we are - * backed by a THP and thus use block mapping if possible. + * backed by a huge stage-1 mapping and thus use block mapping if + * possible. */ - if (mapping_size == PAGE_SIZE && - !(s2vi->max_map_size == PAGE_SIZE || s2vi->map_non_cacheable)) { + if (mapping_size == PAGE_SIZE && s2vi->max_map_size != PAGE_SIZE) { if (perm_fault_granule > PAGE_SIZE) { mapping_size = perm_fault_granule; } else { @@ -2135,10 +2104,6 @@ static int user_mem_abort(const struct kvm_s2_fault_desc *s2fd) return ret; } - /* - * Let's check if we will get back a huge page backed by hugetlbfs, or - * get block mapping for device MMIO region. - */ ret = kvm_s2_fault_pin_pfn(s2fd, &s2vi); if (ret != 1) return ret; From 93e5f7f6189af76a2b2e2dfe647838a7b9f99d21 Mon Sep 17 00:00:00 2001 From: Jiale Yao Date: Thu, 24 Sep 2026 22:29:11 +0800 Subject: [PATCH 687/694] HID: roccat: isku: reject short button reports The raw event callback runs before HID core validates and zero-pads the report. It passes the buffer to helpers which inspect data[0] and, for a button report, read the complete struct isku_report_button. A truncated report can therefore cause an out-of-bounds read. Require enough data for the button report before calling either helper. Commit 47669bec44fe ("HID: asus: refactor the two workqueues and init sequence") added the same kind of raw-event length validation to hid-asus. Fixes: d41c2a7011df ("HID: roccat: Add support for Isku keyboard") Cc: stable@vger.kernel.org Signed-off-by: Jiale Yao Signed-off-by: Benjamin Tissoires --- drivers/hid/hid-roccat-isku.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/hid/hid-roccat-isku.c b/drivers/hid/hid-roccat-isku.c index 93a49c93ae8c..ef1ec9ccaa82 100644 --- a/drivers/hid/hid-roccat-isku.c +++ b/drivers/hid/hid-roccat-isku.c @@ -411,6 +411,9 @@ static int isku_raw_event(struct hid_device *hdev, if (isku == NULL) return 0; + if (size < sizeof(struct isku_report_button)) + return 0; + isku_keep_values_up_to_date(isku, data); if (isku->roccat_claimed) From d879e17a8b4c572c0b53b4e42a36849484e8f00c Mon Sep 17 00:00:00 2001 From: Jiale Yao Date: Thu, 24 Sep 2026 22:29:12 +0800 Subject: [PATCH 688/694] HID: roccat: koneplus: reject short button reports The raw event callback runs before HID core validates and zero-pads the report. It passes the buffer to helpers which inspect data[0] and, for a button report, read the complete struct koneplus_mouse_report_button. A truncated report can therefore cause an out-of-bounds read. Require enough data for the button report before calling either helper. Commit 47669bec44fe ("HID: asus: refactor the two workqueues and init sequence") added the same kind of raw-event length validation to hid-asus. Fixes: 47dbdbffe15b ("HID: roccat: Add support for Roccat Kone[+] v2") Cc: stable@vger.kernel.org Signed-off-by: Jiale Yao Signed-off-by: Benjamin Tissoires --- drivers/hid/hid-roccat-koneplus.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/hid/hid-roccat-koneplus.c b/drivers/hid/hid-roccat-koneplus.c index f80a60539a96..df39868d7e6a 100644 --- a/drivers/hid/hid-roccat-koneplus.c +++ b/drivers/hid/hid-roccat-koneplus.c @@ -523,6 +523,9 @@ static int koneplus_raw_event(struct hid_device *hdev, if (koneplus == NULL) return 0; + if (size < sizeof(struct koneplus_mouse_report_button)) + return 0; + koneplus_keep_values_up_to_date(koneplus, data); if (koneplus->roccat_claimed) From 1590b7c9bea0abe7a5d25dbfe06287c48c238a6f Mon Sep 17 00:00:00 2001 From: Jiale Yao Date: Thu, 24 Sep 2026 22:29:13 +0800 Subject: [PATCH 689/694] HID: roccat: konepure: reject short button reports The raw event callback runs before HID core validates and zero-pads the report. It reads the report number without checking the received length and passes matching data to roccat_report_event(), which copies a complete struct konepure_mouse_report_button. A truncated report can therefore cause an out-of-bounds read and expose adjacent data through the Roccat character device. Require a complete button report before inspecting or forwarding it. Commit 47669bec44fe ("HID: asus: refactor the two workqueues and init sequence") added the same kind of raw-event length validation to hid-asus. Fixes: 8936aa31cd5f ("HID: roccat: add support for Roccat Kone Pure gaming mouse") Cc: stable@vger.kernel.org Signed-off-by: Jiale Yao Signed-off-by: Benjamin Tissoires --- drivers/hid/hid-roccat-konepure.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/hid/hid-roccat-konepure.c b/drivers/hid/hid-roccat-konepure.c index 7f753dfc2a10..529466c33bf8 100644 --- a/drivers/hid/hid-roccat-konepure.c +++ b/drivers/hid/hid-roccat-konepure.c @@ -181,6 +181,9 @@ static int konepure_raw_event(struct hid_device *hdev, != USB_INTERFACE_PROTOCOL_MOUSE) return 0; + if (size < sizeof(struct konepure_mouse_report_button)) + return 0; + if (data[0] != KONEPURE_MOUSE_REPORT_NUMBER_BUTTON) return 0; From e95917a606f486e46f5136fc1b72190742a091d6 Mon Sep 17 00:00:00 2001 From: Jiale Yao Date: Thu, 24 Sep 2026 22:29:14 +0800 Subject: [PATCH 690/694] HID: roccat: kovaplus: reject short button reports The raw event callback runs before HID core validates and zero-pads the report. It passes the buffer to helpers which inspect data[0] and read a complete struct kovaplus_mouse_report_button for button reports. A truncated report can therefore cause an out-of-bounds read. Require enough data for the button report before calling either helper. Commit 47669bec44fe ("HID: asus: refactor the two workqueues and init sequence") added the same kind of raw-event length validation to hid-asus. Fixes: 0e70f97f257e ("HID: roccat: Add support for Kova[+] mouse") Cc: stable@vger.kernel.org Signed-off-by: Jiale Yao Signed-off-by: Benjamin Tissoires --- drivers/hid/hid-roccat-kovaplus.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/hid/hid-roccat-kovaplus.c b/drivers/hid/hid-roccat-kovaplus.c index 9ec42c218ef9..26832b279313 100644 --- a/drivers/hid/hid-roccat-kovaplus.c +++ b/drivers/hid/hid-roccat-kovaplus.c @@ -614,6 +614,9 @@ static int kovaplus_raw_event(struct hid_device *hdev, if (kovaplus == NULL) return 0; + if (size < sizeof(struct kovaplus_mouse_report_button)) + return 0; + kovaplus_keep_values_up_to_date(kovaplus, data); if (kovaplus->roccat_claimed) From 493268f459bb6aeeecbc77f2204c11ad74d05213 Mon Sep 17 00:00:00 2001 From: Jiale Yao Date: Thu, 24 Sep 2026 22:29:15 +0800 Subject: [PATCH 691/694] HID: roccat: pyra: reject short button reports The raw event callback runs before HID core validates and zero-pads the report. It passes the buffer to helpers which inspect data[0] and read a complete struct pyra_mouse_event_button for button reports. A truncated report can therefore cause an out-of-bounds read. Require enough data for the button report before calling either helper. Commit 47669bec44fe ("HID: asus: refactor the two workqueues and init sequence") added the same kind of raw-event length validation to hid-asus. Fixes: cb7cf3da0daa ("HID: roccat: add driver for Roccat Pyra mouse") Cc: stable@vger.kernel.org Signed-off-by: Jiale Yao Signed-off-by: Benjamin Tissoires --- drivers/hid/hid-roccat-pyra.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/hid/hid-roccat-pyra.c b/drivers/hid/hid-roccat-pyra.c index 0d515995bb9d..44c2b416f2db 100644 --- a/drivers/hid/hid-roccat-pyra.c +++ b/drivers/hid/hid-roccat-pyra.c @@ -557,6 +557,9 @@ static int pyra_raw_event(struct hid_device *hdev, struct hid_report *report, if (pyra == NULL) return 0; + if (size < sizeof(struct pyra_mouse_event_button)) + return 0; + pyra_keep_values_up_to_date(pyra, data); if (pyra->roccat_claimed) From 02f3607d0beb82e76481568bf631235e6a319b47 Mon Sep 17 00:00:00 2001 From: Jiale Yao Date: Thu, 24 Sep 2026 22:29:16 +0800 Subject: [PATCH 692/694] HID: roccat: ryos: reject short special reports The raw event callback runs before HID core validates and zero-pads the report. It reads the report number without checking the received length and passes matching data to roccat_report_event(), which copies a complete struct ryos_report_special. A truncated report can therefore cause an out-of-bounds read and expose adjacent data through the Roccat character device. Require a complete special report before inspecting or forwarding it. Commit 47669bec44fe ("HID: asus: refactor the two workqueues and init sequence") added the same kind of raw-event length validation to hid-asus. Fixes: 6f3a19360545 ("HID: roccat: add support for Ryos MK keyboards") Cc: stable@vger.kernel.org Signed-off-by: Jiale Yao Signed-off-by: Benjamin Tissoires --- drivers/hid/hid-roccat-ryos.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/hid/hid-roccat-ryos.c b/drivers/hid/hid-roccat-ryos.c index db83f42457da..85df7955b87c 100644 --- a/drivers/hid/hid-roccat-ryos.c +++ b/drivers/hid/hid-roccat-ryos.c @@ -189,6 +189,9 @@ static int ryos_raw_event(struct hid_device *hdev, != RYOS_USB_INTERFACE_PROTOCOL) return 0; + if (size < sizeof(struct ryos_report_special)) + return 0; + if (data[0] != RYOS_REPORT_NUMBER_SPECIAL) return 0; From 76c51fdc825a0486f8d88ff896d39acec8f377b4 Mon Sep 17 00:00:00 2001 From: Jiale Yao Date: Thu, 24 Sep 2026 22:29:17 +0800 Subject: [PATCH 693/694] HID: roccat: savu: reject short special reports The raw event callback runs before HID core validates and zero-pads the report. It passes the buffer to a helper which inspects data[0] and, for a special report, reads the complete struct savu_mouse_report_special. A truncated report can therefore cause an out-of-bounds read. Require enough data for the special report before calling the helper. Commit 47669bec44fe ("HID: asus: refactor the two workqueues and init sequence") added the same kind of raw-event length validation to hid-asus. Fixes: 6a2a6390cf09 ("HID: roccat: add support for Roccat Savu") Cc: stable@vger.kernel.org Signed-off-by: Jiale Yao Signed-off-by: Benjamin Tissoires --- drivers/hid/hid-roccat-savu.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/hid/hid-roccat-savu.c b/drivers/hid/hid-roccat-savu.c index 679136933560..cfd86267e35b 100644 --- a/drivers/hid/hid-roccat-savu.c +++ b/drivers/hid/hid-roccat-savu.c @@ -182,6 +182,9 @@ static int savu_raw_event(struct hid_device *hdev, if (savu == NULL) return 0; + if (size < sizeof(struct savu_mouse_report_special)) + return 0; + if (savu->roccat_claimed) savu_report_to_chrdev(savu, data); From b0a42af46363b65d6e5018d971189c4882d689c3 Mon Sep 17 00:00:00 2001 From: Dmitry Antipov Date: Tue, 29 Sep 2026 09:57:42 +0300 Subject: [PATCH 694/694] HID: roccat: fix device access in roccat_release() In roccat_release(), access the device using file-specific reader data rather than global array, thus preventing the case when original device was disconnected and a new device has connected using the same slot indexed by minor number. Reported-by: Sashiko Link: https://sashiko.dev/#/patchset/20260902094551.200587-1-dmantipov@yandex.ru?part=2 Signed-off-by: Dmitry Antipov Signed-off-by: Benjamin Tissoires --- drivers/hid/hid-roccat.c | 13 ++++--------- 1 file changed, 4 insertions(+), 9 deletions(-) diff --git a/drivers/hid/hid-roccat.c b/drivers/hid/hid-roccat.c index 5deb6da8d4f7..b3e69c79c4b4 100644 --- a/drivers/hid/hid-roccat.c +++ b/drivers/hid/hid-roccat.c @@ -211,24 +211,19 @@ static int roccat_open(struct inode *inode, struct file *file) static int roccat_release(struct inode *inode, struct file *file) { - unsigned int minor = iminor(inode); struct roccat_reader *reader = file->private_data; - struct roccat_device *device; + struct roccat_device *device = reader->device; - mutex_lock(&devices_lock); - - device = devices[minor]; - if (!device) { - mutex_unlock(&devices_lock); - pr_emerg("roccat device with minor %d doesn't exist\n", minor); + if (WARN_ON(!device)) return -ENODEV; - } mutex_lock(&device->readers_lock); list_del(&reader->node); mutex_unlock(&device->readers_lock); kfree(reader); + mutex_lock(&devices_lock); + if (!--device->open) { /* removing last reader */ if (device->exist) {