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 f6fb8f100b ("af-packet: TPACKET_V3 flexible buffer
implementation."). Its effect prior to f7460d2989 ("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: f6fb8f100b ("af-packet: TPACKET_V3 flexible buffer implementation.")
Cc: stable@vger.kernel.org
Signed-off-by: Dairui Zhang <zhangdairui@gmail.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Link: https://patch.msgid.link/20260923050101.1510064-1-zhangdairui@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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: 35c55c9877 ("tipc: add neighbor monitoring framework")
Signed-off-by: Ginger Li <ginger.jzllee@gmail.com>
Reviewed-by: Tung Nguyen <tung.quang.nguyen@est.tech>
Link: https://patch.msgid.link/20260922080909.21123-1-ginger.jzllee@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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: 541afa10c1 ("net/smc: add smcr_port_err() and smcr_link_down() processing")
Reviewed-by: Mahanta Jambigi <mjambigi@linux.ibm.com>
Signed-off-by: Sidraya Jayagond <sidraya@linux.ibm.com>
Reviewed-by: Dust Li <dust.li@linux.alibaba.com>
Link: https://patch.msgid.link/20260922073149.474762-1-sidraya@linux.ibm.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
__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: 4716af3897 ("net/rds: Give each connection path its own workqueue")
Signed-off-by: Allison Henderson <achender@kernel.org>
Link: https://patch.msgid.link/20260921215027.174657-1-achender@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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: a21b06e731 ("net: sched: add helper support in act_ct")
Cc: stable@vger.kernel.org
Reported-by: Axel Mierczuk <axel.mierczuk@1password.com>
Signed-off-by: Ilya Maximets <i.maximets@ovn.org>
Reviewed-by: Xin Long <lucien.xin@gmail.com>
Reviewed-by: Jamal Hadi Salim <jhs@mojatatu.com>
Reviewed-by: Aaron Conole <aconole@redhat.com>
Link: https://patch.msgid.link/20260921145655.3167436-7-i.maximets@ovn.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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: a21b06e731 ("net: sched: add helper support in act_ct")
Cc: stable@vger.kernel.org
Signed-off-by: Ilya Maximets <i.maximets@ovn.org>
Reviewed-by: Aaron Conole <aconole@redhat.com>
Reviewed-by: Xin Long <lucien.xin@gmail.com>
Reviewed-by: Jamal Hadi Salim <jhs@mojatatu.com>
Link: https://patch.msgid.link/20260921145655.3167436-6-i.maximets@ovn.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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: a21b06e731 ("net: sched: add helper support in act_ct")
Cc: stable@vger.kernel.org
Reported-by: Axel Mierczuk <axel.mierczuk@1password.com>
Signed-off-by: Ilya Maximets <i.maximets@ovn.org>
Reviewed-by: Aaron Conole <aconole@redhat.com>
Reviewed-by: Xin Long <lucien.xin@gmail.com>
Reviewed-by: Jamal Hadi Salim <jhs@mojatatu.com>
Link: https://patch.msgid.link/20260921145655.3167436-5-i.maximets@ovn.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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: cae3a26275 ("openvswitch: Allow attaching helpers to ct action")
Cc: stable@vger.kernel.org
Reported-by: Axel Mierczuk <axel.mierczuk@1password.com>
Signed-off-by: Ilya Maximets <i.maximets@ovn.org>
Reviewed-by: Aaron Conole <aconole@redhat.com>
Link: https://patch.msgid.link/20260921145655.3167436-4-i.maximets@ovn.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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: 3c1860543f ("openvswitch: add nf_ct_is_confirmed check before assigning the helper")
Cc: stable@vger.kernel.org
Signed-off-by: Ilya Maximets <i.maximets@ovn.org>
Reviewed-by: Aaron Conole <aconole@redhat.com>
Link: https://patch.msgid.link/20260921145655.3167436-3-i.maximets@ovn.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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: cae3a26275 ("openvswitch: Allow attaching helpers to ct action")
Cc: stable@vger.kernel.org
Reported-by: Axel Mierczuk <axel.mierczuk@1password.com>
Signed-off-by: Ilya Maximets <i.maximets@ovn.org>
Reviewed-by: Aaron Conole <aconole@redhat.com>
Link: https://patch.msgid.link/20260921145655.3167436-2-i.maximets@ovn.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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: c31b70c996 ("tcp: Add logic to check for SYN w/ data in tcp_simple_retransmit")
Reported-by: Kimi Security Team <bug-report@moonshot.ai>
Tested-by: Weiming Shi <shiweiming@moonshot.ai>
Signed-off-by: Yilin Zhang <yilinzhang@moonshot.ai>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/8a9dff4063a2745653b7e88ceb745d75efa16e68.1790224474.git.yilinzhang@moonshot.ai
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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: 63ed1aab3d ("mctp: Add SIOCMCTP{ALLOC,DROP}TAG ioctls for tag control")
Suggested-by: Jeremy Kerr <jk@codeconstruct.com.au>
Cc: stable@vger.kernel.org
Signed-off-by: Hui Peng <benquike@gmail.com>
Link: https://patch.msgid.link/20260921051002.1656692-1-benquike@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
-----BEGIN PGP SIGNATURE-----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=ZxKg
-----END PGP SIGNATURE-----
Merge tag 'bpf-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf
Pull bpf fixes from Alexei Starovoitov:
- Fix bpf_skb_change_tail() to drop the checksum offload instead of
rejecting the trim of CHECKSUM_PARTIAL skbs (Daniel Borkmann)
- Add KF_PERFMON kfunc flag and require CAP_PERFMON for kfuncs that
read arbitrary memory and for untrusted read-only memory reads
(Daniel Borkmann)
- Clear scalar delta on narrowing stack spill (Daniel Borkmann)
- Set up the frame pointer for the exception callback in arm64 JIT, and
zero-fill other CPUs when BPF_F_CPU update creates a per-cpu hash
element (Donggeun Yoo)
- Various fixes (Emil Tsalapatis):
- Fix bounds check underflow for skb-backed dynptrs
- Fix rx_queue_mapping context access code generation in bpf_sock
- Reject packet pointer arguments to subprogs that may mutate the
packet
- Reject ALU instructions that see arena and non-arena operands on
different code paths
- Fix copied_seq double-counting on sockmap self-redirect
(Geliang Tang)
- Fix divide-by-zero in btf_struct_walk() on a flexible array of
zero-sized elements, fix out-of-bounds read of rtt_min in sock_ops
(Jiayuan Chen)
- Fix bpf_sock_destroy() out-of-bounds read of sk_protocol on TIME_WAIT
and request socks, and sleeping under RCU when destroying a listener
with pending children (Jiayuan Chen)
- Fix JEQ/JNE with immediate operand in MIPS32 JIT and missing zero
extension of BSWAP 16/32 in MIPS64 JIT (Johan Almbladh)
- Avoid soft lockup in htab lookup[_and_delete] batch operations on
large maps (Jose Fernandez)
- Various fixes (Kumar Kartikeya Dwivedi):
- Verify global subprogs in each sleepability context they are
called from
- Make post-verification instruction rewrites killable
- Preserve packet pointer displacement in regsafe()
- Apply CO-RE relocations before subprogram validation, restrict
CO-RE poisoning to relocatable instructions, and reject truncated
ldimm64 CO-RE relocations in libbpf
- Assign lock identity to callback map values
- Compare stack frames in regs_exact()
- Bound ownership depth through local kptrs and graph roots
- Fix u32 overflow in map batch operations when the map size exceeds
4GB (Masoud Aghasi)
- Fix UAF in bpf memalloc due to concurrent consumption of ttrace lists
in alloc_bulk() (Pu Lehui)
- Allow gotox as the terminal instruction of a program or a subprogram
(Siddharth Chintamaneni)
- Disallow bpf_skb_pull_data() for LWT_SEG6LOCAL, skip unsettled links
in link iterator, and reject dev-bound-only programs on other devices
(Weiming Shi)
- Reject non-negative stack offsets in stack_slot_obj_get_spi()
(Xu Yunxiang)
- Check params size before reading reserved fields in
bpf_crypto_ctx_create() (Yuqi Xu)
- Reject max_entries > INT_MAX in sock_map_alloc() (Zhao Gongyi)
- Use a 32-bit compare in xsk_map_gen_lookup() (Zhiling Zou)
- Use kvfree() in xdp_test_run_teardown() (Zhixing Chen)
* tag 'bpf-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf: (58 commits)
selftests/bpf: Test per-cpu initialization of a BPF_F_CPU created element
bpf: Zero-fill other CPUs when BPF_F_CPU creates a per-cpu hash element
bpf: Fix BSWAP 32 and 16 on MIPS64
bpf: Fix immediate JMP JEQ/JNE on MIPS32
bpf: Reject dev-bound-only programs on other devices
bpf, sockmap: Reject max_entries > INT_MAX in sock_map_alloc
selftests/bpf: Test for mixed arena/nonarena code paths
bpf: Prevent variable arena/non-arena register contents
selftests/bpf: Test rejection of pkt args to mutating subprogs
bpf: Reject pkt arguments in mutating subprogs
selftests/bpf: Add selftests for rx_queue_mapping context access
bpf: Fix bpf_sock context code generation
selftests/bpf: Test dynptr slices past end of skb
bpf: Fix bounds check for skb-backed dynptrs
selftests/bpf: Reject iterator destruction through fp+0
bpf: Reject non-negative offsets in stack_slot_obj_get_spi()
bpf: Check params size before reading reserved fields
selftests/bpf: Check local object ownership depth
bpf: Bound ownership depth through local kptrs and graph roots
selftests/bpf: Cover frame changes in bounded loops
...
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: 084e2f6566 ("Support to encoding decoding skb mark on IFE action")
Fixes: 200e10f469 ("Support to encoding decoding skb prio on IFE action")
Fixes: 408fbc22ef ("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 <fangxy@xiaopeng.com>
Link: https://patch.msgid.link/20260921125441.81459-1-fangxy@xiaopeng.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
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: 1659185fb4 ("RDS: IB: Support Fastreg MR (FRMR) memory registration mode")
Cc: stable@vger.kernel.org
Signed-off-by: Dongliang Qin <cccccccccccc777777@gmail.com>
Reviewed-by: Allison Henderson <achender@kernel.org>
Link: https://patch.msgid.link/20260922031546.3874605-1-cccccccccccc777777@gmail.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
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: 130c0f47fd ("ipconfig: send host-name in DHCP requests")
Cc: stable@vger.kernel.org
Reported-by: Vega <vega@nebusec.ai>
Assisted-by: LLM
Signed-off-by: Yuqi Xu <xuyuqiabc@gmail.com>
Reviewed-by: Ren Wei <weir@nebusec.ai>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/7808dfbfa2162dfd0b19f59aff5742d6e0db2abb.1789798023.git.xuyuqiabc@gmail.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Mostly security fixes.
fix use-after-free in nfc_get_local_general_bytes
llcp: Fix list corruption / refcount desync in nfc_llcp_recv_dm()
llcp: Fix race condition in accept_queue lifecycle
llcp: drop truncated I/RR/RNR PDUs in nfc_llcp_recv_hdlc()
llcp: fix -ENOMEM on connect with zero-length service name
llcp: fix WKS SAP hijacking via prefix match in nfc_llcp_wks_sap()
llcp: fix sdreq TLV list leak on parse/alloc/send failure
llcp: fix slab-out-of-bounds reads when logging service names
selftests/nci: Fix out-of-bounds store on thread join
selftests: nci: Correct pthread_create return value check
selftests: nci: Fix uninitialized family ID on missing attribute
virtual_ncidev: Add missing ioctl compat handler
nfcmrvl: validate helper command length before pull
pn533: fix OOB read in pn533_acr122_is_rx_frame_valid()
port100: reject frames whose declared length exceeds the received data
st21nfca: validate ISO15693 inventory length
st21nfca: validate received frame size
trf7970a: power down on startup RX gain failure
Signed-off-by: David Heidelberg <david@ixit.cz>
-----BEGIN PGP SIGNATURE-----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=lO2R
-----END PGP SIGNATURE-----
Merge tag 'nfc-7.3-rc5' of https://codeberg.org/linux-nfc/linux
David Heidelberg says:
====================
NFC fixes for net 7.3-rc5
* tag 'nfc-7.3-rc5' of https://codeberg.org/linux-nfc/linux:
nfc: pn533: fix OOB read in pn533_acr122_is_rx_frame_valid()
nfc: llcp: fix slab-out-of-bounds reads when logging service names
nfc: llcp: fix WKS SAP hijacking via prefix match in nfc_llcp_wks_sap()
nfc: llcp: fix -ENOMEM on connect with zero-length service name
nfc: st21nfca: validate ISO15693 inventory length
nfc: fix use-after-free in nfc_get_local_general_bytes
nfc: trf7970a: power down on startup RX gain failure
nfc: llcp: fix sdreq TLV list leak on parse/alloc/send failure
nfc: llcp: drop truncated I/RR/RNR PDUs in nfc_llcp_recv_hdlc()
nfc: virtual_ncidev: Add missing ioctl compat handler
selftests/nci: Fix out-of-bounds store on thread join
selftests: nci: Fix uninitialized family ID on missing attribute
nfc: llcp: Fix race condition in accept_queue lifecycle
selftests: nci: Correct pthread_create return value check
nfc: port100: reject frames whose declared length exceeds the received data
nfc: llcp: Fix list corruption / refcount desync in nfc_llcp_recv_dm()
nfc: st21nfca: validate received frame size
nfc: nfcmrvl: validate helper command length before pull
====================
Link: https://patch.msgid.link/adeaccc1-cc04-4bb9-a28a-61a61d75ba14@ixit.cz
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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: 184c449f91 ("net: Add support for XPS with QoS via traffic classes")
Signed-off-by: Norbert Szetei <norbert@doyensec.com>
Link: https://patch.msgid.link/162DD16F-54C6-444A-9E09-0B8CB3D591F2@doyensec.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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: 2e15ea390e ("ip_gre: Add support to collect tunnel metadata.")
Signed-off-by: Xuanqiang Luo <luoxuanqiang@kylinos.cn>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Hangbin Liu <liuhangbin@kylinos.cn>
Link: https://patch.msgid.link/20260921031859.9283-1-xuanqiang.luo@linux.dev
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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: 36fbf1e52b ("net: rtnetlink: add linkprop commands to add and delete alternative ifnames")
Cc: stable@vger.kernel.org
Reported-by: Vega <vega@nebusec.ai>
Signed-off-by: Zijie Huang <milkory@outlook.com>
Signed-off-by: Ren Wei <weir@nebusec.ai>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Link: https://patch.msgid.link/fabf02a70787d17299e4b3153eadffaf20d154b3.1789910973.git.milkory@outlook.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Commit 7a9bc9e3f4 ("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: 23461551c0 ("fou: Support for foo-over-udp RX path")
Fixes: 7a9bc9e3f4 ("fou: Don't allow 0 for FOU_ATTR_IPPROTO.")
Cc: stable@vger.kernel.org
Signed-off-by: Hui Peng <benquike@gmail.com>
Reviewed-by: Hangbin Liu <liuhangbin@kylinos.cn>
Link: https://patch.msgid.link/20260921045920.1613098-1-benquike@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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: 915d7e5e59 ("ipv6: sr: add code base for control plane support of SR-IPv6")
Cc: stable@vger.kernel.org
Signed-off-by: Hui Peng <benquike@gmail.com>
Reviewed-by: Hangbin Liu <liuhangbin@kylinos.cn>
Reviewed-by: Justin Iurman <justin.iurman@gmail.com>
Reviewed-by: Andrea Mayer <andrea.mayer@uniroma2.it>
Link: https://patch.msgid.link/20260921044025.1535982-1-benquike@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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: 1431fb31ec ("xen-netback: fix fragment detection in checksum setup")
Suggested-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Shihuang Liu <shlomojune6@gmail.com>
Link: https://patch.msgid.link/20260919133604.50948-1-shlomojune6@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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: a6acb535af ("bridge: mdb: Add MDB bulk deletion support")
Cc: stable@vger.kernel.org
Signed-off-by: Fourie Zhang <fouriezhang@tencent.com>
Acked-by: Nikolay Aleksandrov <razor@blackwall.org>
Link: https://patch.msgid.link/20260920110852.60293-1-fouriezhang@tencent.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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: 4e76e75d6a ("net sched actions: calculate add/delete event message size")
Reported-by: Sashiko <sashiko-bot@kernel.org>
Link: https://netdev-ai.bots.linux.dev/sashiko/#/patchset/20260824153903.4143642-1-victor@mojatatu.com
Tested-by: hybris <hybris@mojatatu.ai>
Co-developed-by: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: Victor Nogueira <victor@mojatatu.com>
Link: https://patch.msgid.link/QDISC-3BLH.v1.20260914203033@mojatatu.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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: d9b8d8e19b ("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 <omermetekaya0@gmail.com>
Link: https://patch.msgid.link/20260908161952.731468-1-omermetekaya0@gmail.com
Signed-off-by: David Heidelberg <david@ixit.cz>
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: d646960f79 ("NFC: Initial LLCP support")
Signed-off-by: Ömer Mete Kaya <omermetekaya0@gmail.com>
Link: https://patch.msgid.link/20260909121437.33744-1-omermetekaya0@gmail.com
Signed-off-by: David Heidelberg <david@ixit.cz>
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: d646960f79 ("NFC: Initial LLCP support")
Signed-off-by: Ömer Mete Kaya <omermetekaya0@gmail.com>
Link: https://patch.msgid.link/20260909122029.34081-1-omermetekaya0@gmail.com
Signed-off-by: David Heidelberg <david@ixit.cz>
Commit 6709d4b7bc ("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: 6709d4b7bc ("net: nfc: Fix use-after-free caused by nfc_llcp_find_local")
Cc: stable@vger.kernel.org
Reported-by: Vega <vega@nebusec.ai>
Assisted-by: LLM
Signed-off-by: Luxiao Xu <rakukuip@gmail.com>
Signed-off-by: Ren Wei <weir@nebusec.ai>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/3cbaac3bee23f8ff3a3284ed32d347696eb1d208.1788841683.git.rakukuip@gmail.com
Signed-off-by: David Heidelberg <david@ixit.cz>
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: d9b8d8e19b ("NFC: llcp: Service Name Lookup netlink interface")
Assisted-by: Claude:claude-opus-4
Signed-off-by: Cong Nguyen <congnt264@gmail.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20260914121129.2098606-1-congnt264@gmail.com
Signed-off-by: David Heidelberg <david@ixit.cz>
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 95674f506c ("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: d646960f79 ("NFC: Initial LLCP support")
Cc: stable@vger.kernel.org
Assisted-by: LLM
Signed-off-by: Aamir Ahmed <elb12345@hotmail.co.uk>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/AS8P251MB0001789BBF04B72745C7D96BC8BA2@AS8P251MB0001.EURP251.PROD.OUTLOOK.COM
Signed-off-by: David Heidelberg <david@ixit.cz>
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: 50b78b2a65 ("NFC: Fix sleeping in atomic when releasing socket")
Signed-off-by: Lee Jones <lee@kernel.org>
Link: https://patch.msgid.link/20260902123033.1169067-1-lee@kernel.org
Signed-off-by: David Heidelberg <david@ixit.cz>
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 b493ea2765 ("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: a69f32af86 ("NFC: Socket linked list")
Signed-off-by: Aldo Ariel Panzardo <qwe.aldo@gmail.com>
Link: https://patch.msgid.link/20260716232657.203145-1-qwe.aldo@gmail.com
Signed-off-by: David Heidelberg <david@ixit.cz>
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: dcda165706 ("Bluetooth: hci_core: Fix build warnings")
Reported-by: Changyul Lee <lcy8047@gmail.com>
Assisted-by: LLM
Signed-off-by: Sang-Hoon Choi <csh0052@gmail.com>
Reported-by: Changyul Lee <lcy8047@gmail.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
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: 1da177e4c3 ("Linux-2.6.12-rc2")
Cc: stable@vger.kernel.org
Signed-off-by: Linmao Li <lilinmao@kylinos.cn>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
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 b04202d606 ("net/sched: replace strncpy with strscpy") replaced
it with strscpy(), which does not pad, creating this condition.
Zero-initialising the struct closes it.
Fixes: b04202d606 ("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 <bernard.ladenthin@gmail.com>
Acked-by: Jamal Hadi Salim <jhs@mojatatu.com>
Link: https://patch.msgid.link/20260918133953.12494-1-bernard.ladenthin@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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 <almasrymina@google.com>
Acked-by: Stanislav Fomichev <sdf@fomichev.me>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Link: https://patch.msgid.link/20260921195545.493253-1-almasrymina@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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: 1da177e4c3 ("Linux-2.6.12-rc2")
Reported-by: Tangxin Xie <xietangxin@h-partners.com>
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Link: https://patch.msgid.link/c31b5e3ee2b7274e804f5eba2f21e2412e7eef7a.1790013825.git.lucien.xin@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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 992cc9f94c ("net/packet: defer vmalloc TX_RING
free until skbs finish").
Reported-by: Katherine Leaver <kleaver@janestreet.com>
Reported-by: Bjoern Doebel <doebel@amazon.de>
Closes: https://lore.kernel.org/netdev/20260909085542.3370986-1-doebel@amazon.de/
Fixes: 5cd8d46ea1 ("packet: copy user buffers before orphan or clone")
Cc: stable@vger.kernel.org
Signed-off-by: Willem de Bruijn <willemb@google.com>
Link: https://patch.msgid.link/20260919004748.1463985-3-willemdebruijn.kernel@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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 03a9d10ecf ("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: 1da177e4c3 ("Linux-2.6.12-rc2")
Suggested-by: Xin Long <lucien.xin@gmail.com>
Reported-by: TencentOS Corvus AI <corvus@tencent.com>
Cc: stable@vger.kernel.org
Signed-off-by: Aohan Mei <henrymei@tencent.com>
Acked-by: Xin Long <lucien.xin@gmail.com>
Link: https://patch.msgid.link/20260921093707.1432184-1-ljp1205831794@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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: 0d2c4f9640 ("bpf: Eliminate rlimit-based memory accounting for sockmap and sockhash maps")
Signed-off-by: Zhao Gongyi <zhaogongyi@BYTEDANCE.COM>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://patch.msgid.link/20260917121016.48171-1-zhaogongyi@bytedance.com
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: c3c16f2ea6 ("bpf: Add rx_queue_mapping to bpf_sock")
Reported-by: Nicholas Carlini <nicholas@carlini.com>
Suggested-by: Nicholas Carlini <nicholas@carlini.com>
Signed-off-by: Emil Tsalapatis <emil@etsalapatis.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Reviewed-by: Jiayuan Chen <jiayuan.chen@linux.dev>
Link: https://patch.msgid.link/20260922172028.6269-4-emil@etsalapatis.com
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: 26588edbef ("genetlink: support split policies in ctrl_dumppolicy_put_op()")
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Link: https://patch.msgid.link/20260918222949.4190284-1-kuba@kernel.org
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
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 6996a2d2d0 ("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 644f9108f3 ("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: 78c91ae2c6 ("ipv4/udp: Add 4-tuple hash for connected socket")
Assisted-by: LLM
Signed-off-by: Shardul Bankar <shardul.b@mpiricsoftware.com>
Reviewed-by: Kuniyuki Iwashima <kuniyu@google.com>
Link: https://patch.msgid.link/20260917-udp_hash4_fix_v1-v1-2-718891af0d7a@mpiricsoftware.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
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 644f9108f3 ("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 78c91ae2c6 ("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: 644f9108f3 ("udp: Make rehash4 independent in udp_lib_rehash()")
Assisted-by: LLM
Signed-off-by: Shardul Bankar <shardul.b@mpiricsoftware.com>
Reviewed-by: Kuniyuki Iwashima <kuniyu@google.com>
Link: https://patch.msgid.link/20260917-udp_hash4_fix_v1-v1-1-718891af0d7a@mpiricsoftware.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
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: 97ffe75791 ("mac80211: factor out station lookup from ieee80211_build_hdr()")
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Link: https://patch.msgid.link/20260916094833.164219-1-nbd@nbd.name
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
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: 0afd425b1b ("cfg80211: fix duplicated scan entries after channel switch")
Reported-by: Yilin Zhang <yilinzhang@moonshot.ai>
Closes: https://lore.kernel.org/linux-wireless/20260901080427.1983015-1-yilinzhang@moonshot.ai/
Assisted-by: Codex:gpt-6-astra
Signed-off-by: Zhao Li <enderaoelyther@gmail.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
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: 5023b14cf4 ("mac80211: support profile split between elements")
Cc: stable@kernel.org
Assisted-by: LLM sparse kasan
[reorder variable declaration]
Signed-off-by: Zhao Li <enderaoelyther@gmail.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
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: a87f736d94 ("mac80211: use RCU for RX aggregation")
Cc: stable@kernel.org # 5.15+
Assisted-by: LLM sparse kasan
Signed-off-by: Zhao Li <enderaoelyther@gmail.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
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: a5f69d94d8 ("mac80211: Get rid of search loop for rate group index")
Fixes: 9208247d74 ("mac80211: minstrel_ht: add basic support for VHT rates <= 3SS@80MHz")
Reported-by: Vega <vega@nebusec.ai>
Assisted-by: LLM
Signed-off-by: Yuqi Xu <xuyuqiabc@gmail.com>
Reviewed-by: Ren Wei <weir@nebusec.ai>
Link: https://patch.msgid.link/607e985d71fbbe63ebfc182a37f882f04f980f7f.1789958876.git.xuyuqiabc@gmail.com
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Core:
- hci_conn: fix CIS hold ownership on reuse
- hci_sock: reject out-of-range OCF values
- hci_sock: validate event length before filtering
- L2CAP: validate frame length before control and FCS access
- RFCOMM: fix NULL dereference of dlc->session in RFCOMM_CONNINFO
- RFCOMM: Reject short EA=0 frames in rfcomm_recv_frame()
- ISO: release unused CIS holds after channel attach
- ISO: balance the parent hold in hci_bind_bis()
- SMP: reject Security Request over BR/EDR
- MGMT: fix race in read_unconf_index_list()
- MGMT: Dequeue pending mesh_send_sync entries on cancel
- BNEP: fix out-of-bounds reads on short RX/TX frames and control fallthrough
Drivers:
- btintel_pcie: validate device-supplied DMA indices
- btnxpuart: Fix skb leak in nxp_process_fw_dump()
-----BEGIN PGP SIGNATURE-----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=cLTc
-----END PGP SIGNATURE-----
Merge tag 'for-net-2026-09-21' of git://git.kernel.org/pub/scm/linux/kernel/git/bluetooth/bluetooth
Luiz Augusto von Dentz says:
====================
bluetooth pull request for net:
Core:
- hci_conn: fix CIS hold ownership on reuse
- hci_sock: reject out-of-range OCF values
- hci_sock: validate event length before filtering
- L2CAP: validate frame length before control and FCS access
- RFCOMM: fix NULL dereference of dlc->session in RFCOMM_CONNINFO
- RFCOMM: Reject short EA=0 frames in rfcomm_recv_frame()
- ISO: release unused CIS holds after channel attach
- ISO: balance the parent hold in hci_bind_bis()
- SMP: reject Security Request over BR/EDR
- MGMT: fix race in read_unconf_index_list()
- MGMT: Dequeue pending mesh_send_sync entries on cancel
- BNEP: fix out-of-bounds reads on short RX/TX frames and control fallthrough
Drivers:
- btintel_pcie: validate device-supplied DMA indices
- btnxpuart: Fix skb leak in nxp_process_fw_dump()
* tag 'for-net-2026-09-21' of git://git.kernel.org/pub/scm/linux/kernel/git/bluetooth/bluetooth:
Bluetooth: RFCOMM: Reject short EA=0 frames in rfcomm_recv_frame()
Bluetooth: RFCOMM: fix NULL dereference of dlc->session in RFCOMM_CONNINFO
Bluetooth: btintel_pcie: validate device-supplied DMA indices
Bluetooth: bnep: fix out-of-bounds reads on short RX/TX frames and control fallthrough
Bluetooth: mgmt: fix race in read_unconf_index_list()
Bluetooth: L2CAP: validate frame length before control and FCS access
Bluetooth: ISO: balance the parent hold in hci_bind_bis()
Bluetooth: hci_sock: validate event length before filtering
Bluetooth: hci_sock: reject out-of-range OCF values
Bluetooth: ISO: release unused CIS holds after channel attach
Bluetooth: hci_conn: fix CIS hold ownership on reuse
Bluetooth: mgmt: Dequeue pending mesh_send_sync entries on cancel
Bluetooth: btnxpuart: Fix skb leak in nxp_process_fw_dump()
Bluetooth: SMP: reject Security Request over BR/EDR
====================
Link: https://patch.msgid.link/20260921135807.3459373-1-luiz.dentz@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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
a4c2fd7f78 ("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 98aa546af5
("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 9a6f0c4d57 ("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: 98aa546af5 ("inet: remove (struct uncached_list)->quarantine")
Signed-off-by: Kuniyuki Iwashima <kuniyu@google.com>
Reviewed-by: Hangbin Liu <liuhangbin@kylinos.cn>
Reviewed-by: Xuanqiang Luo <luoxuanqiang@kylinos.cn>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20260920191558.2990636-1-kuniyu@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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: 5062245a5a ("net: remove legacy way to get/set HW timestamp config")
Signed-off-by: Nicolai Buchwitz <nb@tipi-net.de>
Reviewed-by: Kory Maincent <kory.maincent@bootlin.com>
Link: https://patch.msgid.link/20260918095540.34286-1-nb@tipi-net.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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: cc5c073a69 ("ipv6: Move exception bucket to fib6_nh")
Signed-off-by: Kuniyuki Iwashima <kuniyu@google.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Link: https://patch.msgid.link/20260918082209.2853582-1-kuniyu@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
-----BEGIN PGP SIGNATURE-----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=RfSQ
-----END PGP SIGNATURE-----
Merge tag 'nf-26-09-18' of git://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf
Pablo Neira Ayuso says:
====================
Netfilter/IPVS fixes for net
The following patchset contains Netfilter/IPVS fixes for net, they are:
1) Set on HW_DEAD after HW_PENDING is cleared in the flowtable offload
to ensure GC does not zap it, from Jérémy Jean.
2) Hold the nfnetlink_queue mutex while removing the queue instance
from the netlink notifier that handles NETLINK_URELEASE to fix a
possible race with the UNBIND command. From Florian Westphal.
3) Reject route with NULL rt6i_idev in ip6t_rpfilter. From Weiming Shi.
4) Reject rtinfo->addrnr set to zero from ip6t_rt .checkentry path.
This also fortifies the datapath loop as per Florian's request.
From Luxiao Xu.
5) Fix checksuming in nft_synproxy for IPv6, from Karl Mehltretter.
6) Revalidate ihl before calling icmp_send() in IPVS,
from Julian Anastasov.
7) Fix suspicious RCU usage splat in ctnetlink with expectations.
8) Check for expired catchall elements in the insert and deactivate
path. From Aohan Mei.
* tag 'nf-26-09-18' of git://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf:
netfilter: nf_tables: skip expired catchall elements on insert and delete
netfilter: ctnetlink: fix suspicious RCU usage in expect_iter_name
ipvs: revalidate ihl before icmp_send
netfilter: nft_synproxy: use the family-aware checksum helper
netfilter: ip6t_rt: fix zero-address non-strict match out-of-bounds read
netfilter: ip6t_rpfilter: reject routes without inet6_dev
netfilter: nfnetlink_queue: hold nfnl mutex in event notifier
netfilter: flowtable: publish HW_DEAD after worker is done
====================
Link: https://patch.msgid.link/20260918112844.194503-1-pablo@netfilter.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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: 264637941c ("wifi: cfg80211: Add Support to Set RTS Threshold for each Radio")
Signed-off-by: Stanislaw Gruszka <stf_xl@wp.pl>
Link: https://patch.msgid.link/20260919092454.50266-1-stf_xl@wp.pl
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
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: 39404feee6 ("mac80211: FILS AEAD protection for station mode association frames")
Cc: stable@vger.kernel.org
Reported-by: Vega <vega@nebusec.ai>
Assisted-by: LLM
Signed-off-by: Yuqi Xu <xuyuqiabc@gmail.com>
Reviewed-by: Ren Wei <weir@nebusec.ai>
Link: https://patch.msgid.link/325c048982d637a3af80d74ac5b36e9369b93f1e.1789751126.git.xuyuqiabc@gmail.com
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
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: 24f7580e85 ("minstrel_ht: fixed rate mode through debugfs")
Cc: stable@vger.kernel.org
Reported-by: Vega <vega@nebusec.ai>
Assisted-by: LLM
Signed-off-by: Yuqi Xu <xuyuqiabc@gmail.com>
Reviewed-by: Ren Wei <weir@nebusec.ai>
Link: https://patch.msgid.link/288e6f6d30dc82ff40da502755e6666982f6b735.1789798000.git.xuyuqiabc@gmail.com
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
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: d5edb9ae8d ("wifi: mac80211: mesh fast xmit support")
Cc: stable@vger.kernel.org
Reported-by: Vega <vega@nebusec.ai>
Assisted-by: LLM
Co-developed-by: Luxing Yin <root@tr0jan.top>
Signed-off-by: Luxing Yin <root@tr0jan.top>
Signed-off-by: Zihan Xi <zihanx@nebusec.ai>
Link: https://patch.msgid.link/1daa7a98199fe6965c2da7c42717073a6fd39a0b.1789615568.git.zihanx@nebusec.ai
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
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 69f1322368 ("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: 69f1322368 ("mac80211: shrink struct ieee80211_fragment_entry")
Signed-off-by: Yuchao Zhang <ndaugoing@gmail.com>
Link: https://patch.msgid.link/20260918034004.85078-2-ndaugoing@gmail.com
[assign a separate drop reason]
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
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 d34ba2168a
("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: d34ba2168a ("mac80211: don't delay station destruction")
Cc: stable@vger.kernel.org
Assisted-by: LLM sparse smatch coccinelle kasan
Signed-off-by: Zhao Li <enderaoelyther@gmail.com>
Link: https://patch.msgid.link/20260916113735.31029-1-enderaoelyther@gmail.com
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
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: ece1a2e7e8 ("mac80211: Remove mesh paths when an interface is removed")
Cc: stable@vger.kernel.org
Assisted-by: LLM sparse kasan
Signed-off-by: Zhao Li <enderaoelyther@gmail.com>
Link: https://patch.msgid.link/20260916113649.28315-3-enderaoelyther@gmail.com
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
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: ae76eef027 ("mac80211: return new mpath from mesh_path_add()")
Fixes: 60854fd945 ("mac80211: mesh: convert path table to rhashtable")
Cc: stable@vger.kernel.org
Assisted-by: LLM sparse kasan
Signed-off-by: Zhao Li <enderaoelyther@gmail.com>
Link: https://patch.msgid.link/20260916113649.28315-2-enderaoelyther@gmail.com
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
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: b230e5bf50 ("Bluetooth: RFCOMM: validate skb length in rfcomm_recv_frame")
Assisted-by: LLM
Signed-off-by: Hui Peng <benquike@gmail.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
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:
<TASK>
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
</TASK>
0x10 is the offset of sock in struct rfcomm_session;
rfcomm_sock_getsockopt_old() is inlined into rfcomm_sock_getsockopt().
Commit 43a556b2fd ("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: bb23c0ab82 ("Bluetooth: Add support for deferring RFCOMM connection setup")
Cc: stable@vger.kernel.org
Assisted-by: LLM
Signed-off-by: Hui Peng <benquike@gmail.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
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: 1da177e4c3 ("Linux-2.6.12-rc2")
Assisted-by: LLM
Signed-off-by: Hui Peng <benquike@gmail.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
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: 5bcae31d9c ("mac80211: implement multi-vif in-place reservations")
Cc: stable@vger.kernel.org
Reported-by: Vega <vega@nebusec.ai>
Signed-off-by: Zhiling Zou <zhilinz@nebusec.ai>
Link: https://patch.msgid.link/b9b582b0db3814f641648b8886bb50f9d68f93ac.1785730815.git.zhilinz@nebusec.ai
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
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: 5bcae31d9c ("mac80211: implement multi-vif in-place reservations")
Cc: stable@vger.kernel.org
Reported-by: Vega <vega@nebusec.ai>
Signed-off-by: Zhiling Zou <zhilinz@nebusec.ai>
Link: https://patch.msgid.link/4b6c4b8648ab1920c38a40be32c0b1a2797b4ac6.1785726007.git.zhilinz@nebusec.ai
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Since the MHI HELLO exchange was relocated, it is sent only at device
registration. During a suspend-resume cycle, the firmware in WiFi
cards such as WCN7850 indefinitely waits for another HELLO,
triggering:
ath12k_wifi7_pci 0004:01:00.0: timeout while waiting for restart complete
ath12k_wifi7_pci 0004:01:00.0: failed to resume core: -110
Fix this by triggering the handshake from resume_early in the MHI
transport.
Validated on Qualcomm X1E-801800 on Lenovo Slim 7x across 10
suspend-resume cycles.
Fixes: 544d85de4d ("net: qrtr: Send HELLO message on endpoint register")
Signed-off-by: Daniel J Blueman <daniel@quora.org>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@oss.qualcomm.com>
Reported-by: Jeff Johnson <jeff.johnson@oss.qualcomm.com>
Link: https://lore.kernel.org/all/6257c447-788d-4362-851e-0d552bcf7c56@oss.qualcomm.com/
Tested-by: Jeff Johnson <jeff.johnson@oss.qualcomm.com>
Reported-by: Vlastimil Babka (SUSE) <vbabka@suse.com>
Link: https://lore.kernel.org/all/ab1491bb-cca5-4145-ac7d-31c966abf7b4@suse.com/
Tested-by: Vlastimil Babka (SUSE) <vbabka@kernel.org>
Reported-by: Takashi Iwai <tiwai@suse.de>
Link: https://lore.kernel.org/all/87a4plsg4w.wl-tiwai@suse.de/
Tested-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Thorsten Leemhuis <linux@leemhuis.info>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
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 729c4896ab ("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: a2f7922713 ("net_sched: sch_hfsc: fix classification loops")
Reported-by: Sashiko (gemini + nipa) <sashiko-bot@kernel.org>
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 <victor@mojatatu.com>
Tested-by: hybris <hybris@mojatatu.ai>
Signed-off-by: Jamal Hadi Salim <jhs@mojatatu.com>
Link: https://patch.msgid.link/QDISC-CTUU.v3.20260916184908@mojatatu.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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: 0190c1d452 ("net: sched: atomically check-allocate action")
Cc: stable@vger.kernel.org
Reported-by: Xiang Mei <xmei5@asu.edu>
Signed-off-by: Weiming Shi <bestswngs@gmail.com>
Link: https://patch.msgid.link/20260914065123.4109709-2-bestswngs@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Commit fa8fca8871 ("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: fa8fca8871 ("ipv4: validate IPV4_DEVCONF attributes properly")
Signed-off-by: Quentin Armitage <quentin@armitage.org.uk>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Link: https://patch.msgid.link/20260915213320.1527029-2-quentin@armitage.org.uk
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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: 4c47af4d5e ("net: sctp: rework multihoming retransmission path selection to rfc4960")
Signed-off-by: Yiqi Sun <sunyiqixm@gmail.com>
Acked-by: Xin Long <lucien.xin@gmail.com>
Link: https://patch.msgid.link/20260915095017.942213-1-sunyiqixm@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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: f994752b11 ("net: ethtool: optionally skip rtnl_lock on IOCTL path")
Signed-off-by: Alexander Duyck <alexanderduyck@fb.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/178942019771.7700.338431553546884773.stgit@ahduyck-xeon-server.home.arpa
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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: f8f626754e ("ipv6: Move ipv6_find_hdr() out of Netfilter code.")
Suggested-by: Ilya Maximets <i.maximets@ovn.org>
Suggested-by: Eric Dumazet <edumazet@google.com>
Cc: stable@vger.kernel.org
Signed-off-by: Norbert Szetei <norbert@doyensec.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Ilya Maximets <i.maximets@ovn.org>
Link: https://patch.msgid.link/8F80BA1A-DDFD-432D-9075-242A3435FEB5@doyensec.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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: 64ff70b80f ("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 <ngocthang2710.1999@gmail.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20260915150816.36487-1-ngocthang2710.1999@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Commit 48a5fe3877 ("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: 48a5fe3877 ("tipc: fix bc_ackers underflow on duplicate GRP_ACK_MSG")
Fixes: 2f487712b8 ("tipc: guarantee that group broadcast doesn't bypass group unicast")
Reported-by: James Burton <jamesburton@meta.com>
Cc: stable@vger.kernel.org
Signed-off-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20260913044233.193927-1-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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 <corvus@tencent.com>
Cc: stable@vger.kernel.org
Fixes: aaa31047a6 ("netfilter: nftables: add catch-all set element support")
Assisted-by: CodeBuddy:Kimi-K3
Signed-off-by: Aohan Mei <henrymei@tencent.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
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:
<TASK>
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: f017941060 ("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 <namangulati@google.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
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 7fcc2fe39f ("net:
icmp: avoid invalid transport header access in icmp_send tracepoint")
Fixes: f2edb9f770 ("ipvs: implement passive PMTUD for IPIP packets")
Link: https://sashiko.dev/#/patchset/20260806105211.34622-1-ja%40ssi.bg
Signed-off-by: Julian Anastasov <ja@ssi.bg>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
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: ad49d86e07 ("netfilter: nf_tables: Add synproxy support")
Assisted-by: LLM
Signed-off-by: Karl Mehltretter <kmehltretter@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
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: 1da177e4c3 ("Linux-2.6.12-rc2")
Cc: stable@vger.kernel.org
Reported-by: Vega <vega@nebusec.ai>
Suggested-by: Florian Westphal <fw@strlen.de>
Assisted-by: LLM
Signed-off-by: Luxiao Xu <rakukuip@gmail.com>
Signed-off-by: Ren Wei <weir@nebusec.ai>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
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: e26f9a480f ("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 <fw@strlen.de>
Assisted-by: Claude:gpt-5
Cc: stable@vger.kernel.org
Signed-off-by: Weiming Shi <bestswngs@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
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: 7af4cc3fa1 ("[NETFILTER]: Add "nfnetlink_queue" netfilter queue handler over nfnetlink")
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
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: 2c8897953f ("netfilter: flowtable: Add pending bit for offload work")
Assisted-by: Codex:gpt-5
Signed-off-by: Jérémy Jean <Jeremy.Jean@oss.cyber.gouv.fr>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
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 195374d893 ("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 436c3b66ec ("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: 32607a332c ("ipv4: prefer multipath nexthop that matches source address")
Signed-off-by: Linkui Xiao <xiaolinkui@kylinos.cn>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20260916125316.988044-1-xiaolinkui@126.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
__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: eafb64f40c ("vsock: add netns to vsock core")
Cc: stable@vger.kernel.org
Reviewed-by: Luigi Leonardi <leonardi@redhat.com>
Signed-off-by: Aldo Ariel Panzardo <qwe.aldo@gmail.com>
Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>
Reviewed-by: Bobby Eshleman <bobbyeshleman@meta.com>
Link: https://patch.msgid.link/20260915173050.3176344-1-qwe.aldo@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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: 73af53d820 ("net: sched: cls_u32: Fix u32's systematic failure to free IDR entries for hnodes.")
Reported-by: Sashiko (gemini + nipa) <sashiko-bot@kernel.org>
Closes: https://sashiko.dev/#/patchset/20260822222049.114526-1-jhs@mojatatu.com
Reviewed-by: Victor Nogueira <victor@mojatatu.com>
Tested-by: hybris <hybris@mojatatu.ai>
Signed-off-by: Jamal Hadi Salim <jhs@mojatatu.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/QDISC-LQFE.v1.20260911041746.1@mojatatu.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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: e65650f291 ("bpf: Implement map_gen_lookup() callback for XSKMAP")
Reported-by: Vega <vega@nebusec.ai>
Signed-off-by: Zhiling Zou <zhilinz@nebusec.ai>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Reviewed-by: Emil Tsalapatis <emil@etsalapatis.com>
Link: https://patch.msgid.link/7d2cb8e8dfaa9eb8fdff85156987a60960787dc3.1789056660.git.zhilinz@nebusec.ai
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
If __subflow_push_pending() errors out we should avoid updating the
copied byte counters, to avoid mismatch push call later on.
Fixes: 0fa1b3783a ("mptcp: use get_send wrapper")
Cc: stable@vger.kernel.org
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Reviewed-by: Matthieu Baerts (NGI0) <matttbe@kernel.org>
Signed-off-by: Matthieu Baerts (NGI0) <matttbe@kernel.org>
Link: https://patch.msgid.link/20260917-net-mptcp-misc-fixes-7-3-rc4-v2-3-0cf5c72667c8@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The mptcp scheduler may race with subflow sockets state change: data
transmission on the selected socket may fail and a later release could
try to use mss_now reset to 0 for a divide operation.
Address the issue by explicitly checking for the critical scenario.
Fixes: c886d70286 ("mptcp: do not queue data on closed subflows")
Cc: stable@vger.kernel.org
Reported-by: Shardul Bankar <shardul.b@mpiricsoftware.com>
Reported-by: Xinyang Ge <xinyang@anthropic.com>
Closes: https://lore.kernel.org/20260525194828.1137119-1-shardul.b@mpiricsoftware.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Reviewed-by: Matthieu Baerts (NGI0) <matttbe@kernel.org>
Signed-off-by: Matthieu Baerts (NGI0) <matttbe@kernel.org>
Link: https://patch.msgid.link/20260917-net-mptcp-misc-fixes-7-3-rc4-v2-2-0cf5c72667c8@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Once in a blue moon, the mptcp receive path can recursively call
mptcp_data_ready() via state change under unlucky error conditions, and
then try to hold the data lock again.
Break the recursion loop explicitly checking for the exceptional
condition.
Add a new flag instead of using an existing one like 'closing', to exit
early in subflow_state_change(), and explicitly flush the RX queue at
reset time.
This avoids unneeded processing to check for available data -- calling
get_mapping_status() and more on a dying subflow -- but also in error
reporting and worker scheduling.
Note that we must consume the currently peeked skb before invoking
mptcp_dss_corruption to avoid consuming it again after the eventual
reset has freed it.
Fixes: e32d262c89 ("mptcp: handle consistently DSS corruption")
Cc: stable@vger.kernel.org
Reported-by: Xinyang Ge <xinyang@anthropic.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Reviewed-by: Matthieu Baerts (NGI0) <matttbe@kernel.org>
Signed-off-by: Matthieu Baerts (NGI0) <matttbe@kernel.org>
Link: https://patch.msgid.link/20260917-net-mptcp-misc-fixes-7-3-rc4-v2-1-0cf5c72667c8@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
pskb_carve_inside_header() and pskb_carve_inside_nonlinear() remove
the first bytes of a packet and reallocate skb->head.
All the headers that were present before the operation are gone,
but both functions call skb_headers_offset_update(skb, 0), which
is a no-op : skb->mac_header, skb->network_header,
skb->transport_header and skb->csum_start keep their old values and
now describe bytes which are no longer there.
Both helpers size the new head from the old skb_end_offset(), so the
stale offsets still land inside the new allocation. They point past
skb_tail_pointer() though, to bytes that were never initialized.
pskb_carve_inside_nonlinear() is the worst case, because it leaves a
zombie skb with an empty linear part (skb->data ==
skb_tail_pointer(skb), skb_headlen(skb) == 0), while
skb_mac_header_was_set() is still true and skb->mac_header is way
ahead of skb->data.
The only user of pskb_extract() is rds_tcp_data_recv(), and the
carved skb is queued on tinc->ti_skb_list. When the RDS incoming
message is released, rds_tcp_inc_free() calls skb_queue_purge(),
which frees the skbs with SKB_DROP_REASON_QUEUE_PURGE. This is
visible from drop_monitor, which then tries to pull back to the
(bogus) mac header :
skbuff: __skb_pull(len=234)
skb len=6968 data_len=6968 headroom=0 headlen=0 tailroom=0
end-tail=384 mac=(234,14) mac_len=14 net=(248,40) trans=288
shinfo(txflags=0 nr_frags=1 gso(size=1428 type=16 segs=5))
csum(0x100120 start=288 offset=16 ip_summed=3 complete_sw=0 valid=1 level=0)
hash(0x7b446c6c sw=0 l4=1) proto=0x86dd pkttype=0 iif=60
kernel BUG at ./include/linux/skbuff.h:2847!
Add skb_carve_reset_headers() to mark the mac and transport headers
as not set, reset the network header, clear skb->mac_len, and drop
a now meaningless CHECKSUM_PARTIAL (csum_start no longer describes
anything).
Invalidate the inner offsets as well. Unlike mac_header and
transport_header they have no "unset" sentinel, so a leftover
non-zero value still looks like a real header. Zero
skb->inner_mac_header, skb->inner_network_header,
skb->inner_transport_header, skb->inner_protocol and
skb->encapsulation, so that all the header state is invalidated in
one place.
v2: fixed an inaccurate changelog. The stale offsets stay inside the
new skb->head, which is never smaller than the old one, they
simply point past skb_tail_pointer() to bytes that are gone.
Thanks to Xuanqiang Luo for insisting on this.
Also invalidate the inner header state, as suggested by the
netdev AI review :
https://netdev-ai.bots.linux.dev/sashiko/#/patchset/20260911114922.621937-1-edumazet%40google.com
Fixes: 6fa01ccd88 ("skbuff: Add pskb_extract() helper function")
Reported-by: syzbot+586af68eb819833c2d91@syzkaller.appspotmail.com
Closes: https://lore.kernel.org/netdev/6aa3e9d3.f2639fcc.29487d.0028.GAE@google.com/
Cc: Xuanqiang Luo <xuanqiang.luo@linux.dev>
Cc: Allison Henderson <achender@kernel.org>
Cc: rds-devel@oss.oracle.com
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Xuanqiang Luo <luoxuanqiang@kylinos.cn>
Link: https://patch.msgid.link/20260915130423.3956471-1-edumazet@google.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Exclude old ACKs before SND.UNA from the tcp fast path
as well as ACKs after SND.NXT.
Such ACKs will fall through to the slow path, where tcp_ack()
performs the appropriate validation and challenge ACK handling
according to RFC5961 and Commit 3d501dd326 ("tcp: do not
accept ACK of bytes we never sent").
This prevents old ACKs from being accepted
or modifying connection state as part of the fast path before
appropriate ACK validation is applied.
In particular, this prevents payload carried by a segment with
an excessively old ACK from advancing RCV.NXT before the ACK
is rejected.
Fixes: 31770e34e4 ("tcp: Revert "tcp: remove header prediction"")
Reported-by: Amit Klein <amit.klein@mail.huji.ac.il>
Reported-by: Tamir Shahar <tamir.shahar1@mail.huji.ac.il>
Reported-by: Inbal Schussheim <inbal.lipshtat@mail.huji.ac.il>
Suggested-by: Eric Dumazet <edumazet@google.com>
Cc: stable@vger.kernel.org
Signed-off-by: Inbal Schussheim <inbal.lipshtat@mail.huji.ac.il>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20260914090408.1435080-2-inbal.lipshtat@mail.huji.ac.il
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
PSP conflicts with TLS ULP in its usage of both skb->decrypted and
sk->sk_validate_xmit_skb().
Make PSP mutually exclusive with TLS ULP, the only other user of either
of these. As other users of skb->decrypted come along, they can be added
to sk_has_decrypt_user(). It would make sense to also assert that
sk->sk_validate_xmit_skb() is also NULL in both of these setup paths for
similar future proofing, but the PSP listener/sk_clone() path is still
broken and it could be seen as a regression to not allow rx assoc to run
on a child of a listener socket with PSP tx assoc state.
Include all TCP ULPs in the sk_has_decrypt_user() check, even though TLS
is the only one that conflicts with PSP via the decrypted bit. This is
intentional because PSP was not designed to be used with ULPs. It is
best to close off surface area that may make bugs reachable, until
someone wishes to design and test an actual user of PSP with ULPs.
Fixes: 6b46ca260e ("net: psp: add socket security association code")
Signed-off-by: Daniel Zahka <daniel.zahka@gmail.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Link: https://patch.msgid.link/20260915-psp-ktls-fix-v2-1-0eedc3b148ec@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
sk->sk_flags must only be changed while holding the socket lock,
because sock_set_flag() and sock_reset_flag() use non atomic
operations (__set_bit() and __clear_bit()).
sock_gettstamp() is one of the last places where a bit of sk->sk_flags
is changed from a syscall without owning the socket lock, through
sock_enable_timestamp(sk, SOCK_TIMESTAMP).
sk_set_memalloc() and sk_clear_memalloc() also change sk->sk_flags
without the socket lock, but their callers (nbd, iscsi_tcp, nvme-tcp,
sunrpc, wireguard) need a careful audit, this will be addressed in a
separate patch.
Jungwoo Lee and Wongi Lee reported an UDP socket use-after-free
caused by this bug: a SIOCGSTAMPNS_NEW ioctl racing with bind()
can cancel the SOCK_RCU_FREE bit that udp_lib_get_port() just set,
because both threads perform a read-modify-write on the same word.
CPU 0 (bind) CPU 1 (SIOCGSTAMPNS_NEW)
-------------------------------- ----------------------------
read sk_flags = F read sk_flags = F
compute F | BIT(SOCK_RCU_FREE) compute F | BIT(SOCK_TIMESTAMP)
store F | BIT(SOCK_RCU_FREE)
sk_add_node_rcu(sk, ...)
store F | BIT(SOCK_TIMESTAMP)
After the lost update, SOCK_RCU_FREE is clear while the socket is
visible to lockless UDP receive lookups. sk_destruct() then frees
the socket immediately instead of waiting for a RCU grace period,
while the receive path still holds a reference-less pointer to it:
BUG: KASAN: slab-use-after-free in ipv4_pktinfo_prepare+0x30/0x410
Read of size 8 at addr ffff888008806610 by task exploit/207
CPU: 0 UID: 1000 PID: 207 Comm: exploit Not tainted 6.12.95+ #1
ipv4_pktinfo_prepare+0x30/0x410
udp_queue_rcv_one_skb+0x51c/0x1180
udp_unicast_rcv_skb+0x109/0x350
ip_protocol_deliver_rcu+0x14b/0x310
ip_local_deliver_finish+0x29d/0x390
ip_local_deliver+0x24d/0x2a0
Only grab the socket lock when SOCK_TIMESTAMP has to be set,
to keep the common case lockless.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Reported-by: Jungwoo Lee <jwlee2217@gmail.com>
Reported-by: Wongi Lee <qw3rtyp0@gmail.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20260915043055.3441600-1-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
ipip_fill_forward_path() and ip6_tnl_fill_forward_path() look up the
route to the tunnel's remote endpoint and set ctx->dev to its device,
which is the tunnel itself when that route resolves back to the tunnel.
dev_fill_forward_path() then makes no progress and trips
WARN_ON_ONCE(last_dev == ctx->dev) as soon as a flowtable tries to
offload a flow through the tunnel. That routing loop is a configuration
any CAP_NET_ADMIN user can set up, and ip_tunnel_xmit() and
ip6_tnl_xmit() already treat it as a tx error, so remove the warning and
just fail the walk, as commit 008e7a7c29 ("net: remove WARN_ON_ONCE
when accessing forward path array") did for the path stack overflow.
Fixes: ab427db178 ("netfilter: flowtable: Add IPIP rx sw acceleration")
Fixes: d98103575d ("netfilter: flowtable: Add IP6IP6 rx sw acceleration")
Closes: https://lore.kernel.org/all/CA+0ovCgaRvbd0Udj70b2xxG8Cx3CaCpNhnf1V4RWQuDveZYZhA@mail.gmail.com/
Suggested-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Farhad Alemi <farhad.alemi@berkeley.edu>
Reviewed-by: Xuanqiang Luo <luoxuanqiang@kylinos.cn>
Link: https://patch.msgid.link/CA+0ovCgKDOk+Bg6Gh5Lwx94u_jJjQ30-vY1JcY2BYfhnWJJbPA@mail.gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
tpacket_req3.tp_sizeof_priv is an unsigned int, and packet_set_ring()
validates the full value against the block size. init_prb_bdqc() then
stores it in the unsigned short blk_sizeof_priv field.
Commit 2b6867c2ce ("net/packet: fix overflow in check for priv area
size") fixed the validation arithmetic, but an accepted value above
USHRT_MAX still narrows when it is stored.
For a 131072-byte block, tp_sizeof_priv=65536 is valid. The narrowing
makes offset_to_first_pkt 48 instead of 65584, so packet records can be
placed in the private area that userspace asked the kernel to preserve.
blk_sizeof_priv is internal state, so widen it to hold the validated
UAPI value.
Fixes: f6fb8f100b ("af-packet: TPACKET_V3 flexible buffer implementation.")
Cc: stable@vger.kernel.org
Signed-off-by: Mark Amirkan <markdamirkan@gmail.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Link: https://patch.msgid.link/20260913-b4-send-packet-private-v1-1-925eab2cd388@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Commit 5813022985 ("mptcp: error out earlier on disconnect") made
mptcp_recvmsg() stop when sk_wait_data() returns an error. The error is
stored in err, but the function then jumps to a path which returns
copied. When no data was copied, recvmsg() therefore returns zero and
reports a false EOF.
Store the result in copied, which is the value returned by the function.
This also keeps the usual partial-read result when data was copied before
the error.
A recvmsg() blocked in one thread reproduces the issue when another
thread disconnects the same MPTCP socket with connect(AF_UNSPEC).
Before this change recvmsg() returns zero; afterwards it returns -EPIPE.
Fixes: 5813022985 ("mptcp: error out earlier on disconnect")
Cc: stable@vger.kernel.org
Signed-off-by: Mark Amirkan <markdamirkan@gmail.com>
Reviewed-by: Matthieu Baerts (NGI0) <matttbe@kernel.org>
Link: https://patch.msgid.link/20260913-b4-send-mptcp-recv-error-v1-1-4eaa3684a8b8@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Commit 61fad6816f ("net/packet: tpacket_rcv: avoid a producer race
condition") added rx_owner_map and made tpacket_rcv() claim a V1 or V2
ring slot before converting the virtio-net header. If the conversion
fails, the drop path leaves the slot claimed.
With a one-frame TPACKET_V2 ring, an unsupported UDP GSO packet leaves
the only slot unavailable, so the ring also drops the next valid packet.
Clear the ownership bit on this error path. TPACKET_V3 already clears
its block state here.
Fixes: 61fad6816f ("net/packet: tpacket_rcv: avoid a producer race condition")
Cc: stable@vger.kernel.org
Signed-off-by: Mark Amirkan <markdamirkan@gmail.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Link: https://patch.msgid.link/20260913-b4-send-packet-vnet-v1-1-5545ffb528ae@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
- mac80211: S1G TIM bitmap fix
- ath12k: remove undocumented DT ABI implementation
- various firmware API and over-the-air hardening changes
- fixes for most cfg80211/mac80211 syzbot reports
-----BEGIN PGP SIGNATURE-----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=/zWX
-----END PGP SIGNATURE-----
Merge tag 'wireless-2026-09-16' of https://git.kernel.org/pub/scm/linux/kernel/git/wireless/wireless
Johannes Berg says:
====================
Many fixes:
- mac80211: S1G TIM bitmap fix
- ath12k: remove undocumented DT ABI implementation
- various firmware API and over-the-air hardening changes
- fixes for most cfg80211/mac80211 syzbot reports
* tag 'wireless-2026-09-16' of https://git.kernel.org/pub/scm/linux/kernel/git/wireless/wireless: (67 commits)
wifi: brcmsmac: fix UAF in brcms_free_timer()
wifi: brcmfmac: fix lost 802.1x TX completion wakeup
wifi: ath11k: cleanup arsta in ath11k_mac_peer_cleanup_all()
wifi: wcn36xx: Fix potential use-after-free in TX ack timer teardown
wifi: ath12k: ahb: Revert undocumented ABI and dead code
wifi: mac80211: refuse to make a monitor active when it has no queue
wifi: libipw: reject TKIP frames without a full MIC
wifi: virt_wifi: don't transfer operstate before register
wifi: cfg80211: check if AP has been started or joined a mesh before adding new station
wifi: cfg80211: move link_id validation earlier in nl80211_new_station()
wifi: cfg80211: do not support direct add of station to AP_VLAN interfaces
wifi: cfg80211: verify if AP_VLAN belongs to the correct AP
wifi: mac80211: set up the TX info early to fix failure paths
wifi: mac80211: mesh: release the channel if start fails
wifi: mac80211: mesh: reset the CSA state when leaving
wifi: mac80211: add HE 6 GHz capability in the scan elems len
wifi: mac80211: don't access the TSF of a down interface
wifi: mac80211: don't RCU-dereference the mesh CSA settings we just set
wifi: mac80211: don't allow link changes when iface is down
wifi: mac80211: require a peer station for TDLS setup confirm
...
====================
Link: https://patch.msgid.link/20260916083642.110609-3-johannes@sipsolutions.net
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
-----BEGIN PGP SIGNATURE-----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=skHa
-----END PGP SIGNATURE-----
Merge tag 'ipsec-2026-09-16' of git://git.kernel.org/pub/scm/linux/kernel/git/klassert/ipsec
Steffen Klassert says:
====================
pull request (net): ipsec 2026-09-16
1) xfrm: iptfs: fix stack OOB read in iptfs_skb_reset_frag_walk()
Add the up-front nr_frags guard iptfs_skb_add_frags() already has,
so an out-of-range offset can't walk past the on-stack frags[] array.
2) xfrm: serialize state GC with device state flush
Serialize xfrm_state destruction against the deferred-device pass
with a dedicated mutex, since the device GC list doesn't hold a state
reference and the two paths could free the same state.
3) xfrm: add missing RCU read lock in xfrm_send_migrate_state()
Hold the RCU read lock around xfrm_nlmsg_multicast() so the
rcu_dereference() of net->xfrm.nlsk doesn't warn.
4) xfrm: iptfs: fix runt reassembly panic from short inner tot_len
Require the runt length to cover at least the minimum IP header,
so a tot_len in [6, 19] (IPv4) can't write past the declared length
and trip skb_over_panic().
5) ipv6: xfrm: use full sockets in local error paths
Use skb_to_full_sk() in xfrm6_local_rxpmtu() and xfrm6_local_error()
and bail out without a full socket, so a TCP_NEW_SYN_RECV request_sock
isn't miscast as a full inet/IPv6 socket.
6) xfrm: fix compat ALLOCSPI request use-after-free
Drop the redundant alloc_compat() in xfrm_alloc_userspi() so the
compat translator no longer reads past the payload and publishes a
child a multicast clone can still see after xfrm_user_rcv_msg() frees.
7) xfrm: add missing rcu_read_lock(), skb_dst_force() and dev_hold() for xfrm_trans_reinject()
Force the dst before queuing, hold dev across the workqueue deferral,
and take rcu_read_lock() around the finish() loop, so transport-mode
reinjection doesn't deref non-refcounted dst/dev under workqueue.
8) xfrm: use hlist_del_init_rcu for state_cache and state_cache_input
Switch to hlist_del_init_rcu() so a second __xfrm_state_delete() is
a no-op instead of writing through LIST_POISON2, closing the UAFs.
9) esp: downgrade zerocopy managed frags before mutating skb frags
Call skb_zcopy_downgrade_managed() before ESP rewrites the skb frag
array, so per-frag unrefs in esp_ssg_unref() and skb_release_data()
stay balanced for ubuf-owned managed frags.
10) xfrm: hold net_device reference under RCU in bundle creation
Read dst->dev via dst_dev_rcu() and keep RCU active through
xfrm_fill_dst(), so a concurrent RTM_DELLINK can't free dev
under bundle creation.
11) xfrm: save input state data before secpath resets
Save the state protocol on the stack while it's still valid and
use the saved address family for transport_finish(), so post-reset
dereferences (VTI, XFRM if, MAX_DEPTH error) can't UAF the state.
12) net: xfrm: reject unrepresentable espintcp transport headers
Use the careful transport-header helper and drop the skb through
the XFRM error path when the offset can't be represented, instead
of silently truncating it.
* tag 'ipsec-2026-09-16' of git://git.kernel.org/pub/scm/linux/kernel/git/klassert/ipsec:
net: xfrm: reject unrepresentable espintcp transport headers
xfrm: save input state data before secpath resets
xfrm: hold net_device reference under RCU in bundle creation
esp: downgrade zerocopy managed frags before mutating skb frags
xfrm: use hlist_del_init_rcu for state_cache and state_cache_input
xfrm: add missing rcu_read_lock(), skb_dst_force() and dev_hold() for xfrm_trans_reinject()
xfrm: fix compat ALLOCSPI request use-after-free
ipv6: xfrm: use full sockets in local error paths
xfrm: iptfs: fix runt reassembly panic from short inner tot_len
xfrm: add missing RCU read lock in xfrm_send_migrate_state()
xfrm: serialize state GC with device state flush
xfrm: iptfs: fix stack OOB read in iptfs_skb_reset_frag_walk()
====================
Link: https://patch.msgid.link/20260916101938.118628-1-steffen.klassert@secunet.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
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: 73d1df2a7a ("Bluetooth: Add support for Read Unconfigured Index List command")
Cc: stable@vger.kernel.org
Signed-off-by: Aldo Ariel Panzardo <qwe.aldo@gmail.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
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: 1c2acffb76 ("Bluetooth: Add initial support for ERTM packets transfers")
Fixes: fcc203c30d ("Bluetooth: Add support for FCS option to L2CAP")
Cc: stable@vger.kernel.org
Signed-off-by: Aldo Ariel Panzardo <qwe.aldo@gmail.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
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: fa224d0c09 ("Bluetooth: ISO: Reassociate a socket with an active BIS")
Cc: stable@vger.kernel.org
Signed-off-by: Aldo Ariel Panzardo <qwe.aldo@gmail.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
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: 1da177e4c3 ("Linux-2.6.12-rc2")
Cc: stable@vger.kernel.org
Signed-off-by: Aldo Ariel Panzardo <qwe.aldo@gmail.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
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: 1da177e4c3 ("Linux-2.6.12-rc2")
Cc: stable@vger.kernel.org
Signed-off-by: Aldo Ariel Panzardo <qwe.aldo@gmail.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
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: 69997d50ec ("Bluetooth: ISO: handle bound CIS cleanup via hci_conn")
Cc: stable@vger.kernel.org
Signed-off-by: Aldo Ariel Panzardo <qwe.aldo@gmail.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Commit 69997d50ec ("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: 69997d50ec ("Bluetooth: ISO: handle bound CIS cleanup via hci_conn")
Cc: stable@vger.kernel.org
Signed-off-by: Aldo Ariel Panzardo <qwe.aldo@gmail.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
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: b338d91703 ("Bluetooth: Implement support for Mesh")
Signed-off-by: Lee Jones <lee@kernel.org>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
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: b5ae344d4c ("Bluetooth: Add full SMP BR/EDR support")
Assisted-by: LLM
Signed-off-by: Christiano Amora <christiano.amora@gmail.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
netlink_realloc_groups() uses krealloc() under netlink_table_grab().
Whenever NLGRPSZ(groups) lands in a different kmalloc bucket, the old
bitmap is freed immediately.
Two readers of nlk->groups / nlk->ngroups do not hold the netlink
table lock:
1) sk_diag_dump_groups(). Hashed (bound) sockets are dumped from the
rhashtable walk in __netlink_diag_dump(), which only holds RCU.
Only the mc_list part of the dump takes nl_table_lock.
2) netlink_native_seq_show() (/proc/net/netlink), whose walk has been
lockless since commit 21e4902aea ("netlink: Lockless lookup with
RCU grace period in socket release").
Both can read a freed buffer, and sk_diag_dump_groups() can also read
past the end of the old (smaller) buffer if it happens to load the old
@groups pointer together with the new @ngroups value, copying the
result into a NETLINK_DIAG_GROUPS attribute.
This is the same class of bug that commit f773608026 ("netlink:
access nlk groups safely in netlink bind and getname") fixed for bind()
and getname(); these two readers were missed. Simply grabbing the table
lock in sk_diag_dump_groups() is not an option, because it is also
called with nl_table_lock already held from the mc_list section of the
dump.
Make the lockless readers safe instead:
- Allocate a new bitmap and free the old one after an RCU grace period,
instead of relying on the implicit kfree() done by krealloc().
- Publish @groups before @ngroups, both with release semantics, and have
the lockless readers load @ngroups first. A reader can then never pair
the new (bigger) size with the old (smaller) buffer, and a reader
picking up the new pointer while still seeing the old size is
guaranteed to see the initialized bitmap.
netlink_realloc_groups() is called from process context (bind() and
setsockopt()), so kfree_rcu_mightsleep() can be used, once the table
has been released.
Fixes: 21e4902aea ("netlink: Lockless lookup with RCU grace period in socket release")
Fixes: ad20207432 ("netlink: Use rhashtable walk interface in diag dump")
Reported-by: James Burton <jamesburton@meta.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20260911160804.917099-1-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Allowing switchdev to prevent vlan deletion and error out in __vlan_del
could cause multiple different issues - inconsistent state, memory leaks
when flushing, NULL pointer dereference on bridge error when flushing.
It doesn't make sense to allow it to stop __vlan_del, so log the error
and continue with software vlan deletion. This is also consistent with
8021q behaviour.
Suggested-by: Ido Schimmel <idosch@nvidia.com>
Fixes: bf361ad381 ("net: bridge: check __vlan_vid_del for error")
Fixes: 5454f5c28e ("net: bridge: vlan: check for errors from __vlan_del in __vlan_flush")
Fixes: 2594e9064a ("bridge: vlan: add per-vlan struct and move to rhashtables")
Fixes: 9c86ce2c1a ("net: bridge: Notify about bridge VLANs")
Signed-off-by: Nikolay Aleksandrov <razor@blackwall.org>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Link: https://patch.msgid.link/20260914105258.3436918-1-razor@blackwall.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
ovs_ct_get_conn_labels() adds the labels extension when a conntrack
entry does not have one. Confirmed conntracks can be read locklessly,
so adding an extension may reallocate and free the extension block
while another CPU accesses it.
Only add the extension for unconfirmed conntracks. A confirmed
conntrack without labels now fails the caller's label operation instead
of reallocating its extension storage.
Fixes: c2ac667358 ("openvswitch: Allow matching on conntrack label")
Cc: stable@vger.kernel.org
Reported-by: Vega <vega@nebusec.ai>
Signed-off-by: Zhiling Zou <zhilinz@nebusec.ai>
Reviewed-by: Ilya Maximets <i.maximets@ovn.org>
Reviewed-by: Aaron Conole <aconole@redhat.com>
Link: https://patch.msgid.link/372fbb062b40ae6723684f55484be86ff0064f8e.1789218015.git.zhilinz@nebusec.ai
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
In reset_per_cpu_data(), al is computed as:
al = sizeof(struct net_dm_alert_msg);
al += dm_hit_limit * sizeof(struct net_dm_drop_point);
al += sizeof(struct nlattr);
skb = genlmsg_new(al, GFP_KERNEL);
...
nla = nla_reserve(skb, NLA_UNSPEC, sizeof(struct net_dm_alert_msg));
...
msg = nla_data(nla);
memset(msg, 0, al);
Because al includes sizeof(struct nlattr) (the 4-byte attribute header),
genlmsg_new() allocates al bytes of tailroom starting at nla.
However, msg points to nla_data(nla), which is located
sizeof(struct nlattr) bytes past nla. Calling memset(msg, 0, al)
therefore writes al bytes starting from msg, exceeding the allocated
buffer by sizeof(struct nlattr) (4 bytes) and corrupting
skb_shared_info.
Fix this by letting al represent only the payload length, allocating
the skb with genlmsg_new(nla_total_size(al), GFP_KERNEL), and zeroing
al bytes from msg.
Fixes: 683703a26e ("drop_monitor: Update netlink protocol to include netlink attribute header in alert message")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Hangbin Liu <liuhangbin@kylinos.cn>
Link: https://patch.msgid.link/20260910204612.3762015-5-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
syzbot reported a preemption warning in sk_skb_reason_drop():
BUG: using smp_processor_id() in preemptible [00000000] code: syz.0.17/5917
caller is net_dm_packet_trace_kfree_skb_hit+0x119/0x350 net/core/drop_monitor.c:519
In net_dm_packet_trace_kfree_skb_hit(), data = this_cpu_ptr(&dm_cpu_data)
is evaluated before spin_lock_irqsave(&data->drop_queue.lock, flags).
When kfree_skb() is called from preemptible context (e.g. process context
during close() on /dev/net/tun), preemption is enabled, triggering the
CONFIG_DEBUG_PREEMPT warning in smp_processor_id().
The same pattern exists in net_dm_hw_trap_summary_probe() and
net_dm_hw_trap_packet_probe() for dm_hw_cpu_data.
This is a false positive because each per-cpu structure is protected
by its own spinlock. If the task migrates to another CPU right after
reading the per-cpu pointer, the lock still safely synchronizes
access to that queue.
Use raw_cpu_ptr() instead of this_cpu_ptr() to silence
CONFIG_DEBUG_PREEMPT without disturbing interrupt state or breaking
PREEMPT_RT locking semantics.
Fixes: ca30707dee ("drop_monitor: Add packet alert mode")
Fixes: 5855357cd4 ("drop_monitor: Prepare probe functions for devlink tracepoint")
Reported-by: syzbot+dc57fd6722deb17e92af@syzkaller.appspotmail.com
Closes: https://lore.kernel.org/netdev/6aa316b2.f81106d8.2ab401.0014.GAE@google.com/
Signed-off-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20260910204612.3762015-4-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
In drop_monitor teardown paths (net_dm_trace_off_set(),
net_dm_hw_monitor_stop(), and error unwind paths in net_dm_trace_on_set()
and net_dm_hw_monitor_start()), per-CPU timers are stopped using
timer_delete_sync() followed by cancel_work_sync().
However, there is a circular dependency between send_timer and
dm_alert_work:
1) sched_send_work() (timer callback) schedules dm_alert_work.
2) send_dm_alert() / net_dm_hw_summary_work() calls reset_per_cpu_data()
or net_dm_hw_reset_per_cpu_data().
3) If memory allocation fails under memory pressure in the reset
function, it re-arms the timer via mod_timer(&data->send_timer, ...).
If dm_alert_work is running concurrently while timer_delete_sync()
executes on another CPU, an allocation failure in the worker will
re-arm the timer after timer_delete_sync() has already returned.
Once cancel_work_sync() completes and module_put() is called, the timer
remains active in the timer wheel. If the module is then unloaded, the
timer will fire and execute sched_send_work() in freed memory,
triggering a kernel panic / use-after-free.
Switch from timer_delete_sync() to timer_shutdown_sync(). This guarantees
that any in-flight timer handler has finished and prevents subsequent
re-arming attempts from running workers from succeeding. When monitoring
is restarted later, timer_setup() is invoked, which cleanly
re-initializes the timer.
Fixes: 9398e9c0b1 ("drop_monitor: Perform cleanup upon probe registration failure")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20260910204612.3762015-3-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
If register_trace_napi_poll() fails in net_dm_trace_on_set(),
unregister_trace_kfree_skb() is called to roll back the kfree_skb
tracepoint registration.
However, tracepoint_synchronize_unregister() is omitted before calling
cancel_work_sync() and module_put(). An in-flight probe executing
concurrently on another CPU could call schedule_work() after
cancel_work_sync() has already returned, leaving a pending work item
scheduled after the module reference is dropped. If the module is then
unloaded, executing the work item triggers a kernel panic.
Add tracepoint_synchronize_unregister() after unregister_trace_kfree_skb()
in the error path, matching net_dm_trace_off_set() and
net_dm_hw_probe_unregister().
Fixes: 7c747838a5 ("drop_monitor: Split tracing enable / disable to different functions")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Hangbin Liu <liuhangbin@kylinos.cn>
Link: https://patch.msgid.link/20260910204612.3762015-2-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The following command triggers a kernel panic:
ip link add d0 type dummy; ip link set d0 up
ip route add 10.30.0.0/16 \
encap ip id 300 geneve_opts 4660:66:11223344 dev d0
memcpy: detected buffer overflow: 4 byte write of buffer size 0
kernel BUG at lib/string_helpers.c:1044!
...
ip_tun_parse_opts.part.0.cold+0x10/0x10
ip_tun_build_state+0x116/0x2a0
On kernels built with GCC 15+ and `CONFIG_FORTIFY_SOURCE`, the fortified
`memcpy()` got 0 sized destination with request of 4 bytes length:
static int ip_tun_parse_opts_geneve(...)
{
...
attr = tb[LWTUNNEL_IP_OPT_GENEVE_DATA];
data_len = nla_len(attr); /* == 4 */
struct geneve_opt *opt = ip_tunnel_info_opts(info) + opts_len;
memcpy(opt->opt_data, nla_data(attr), data_len);
/* ^^^^^^^^^^^^^ 0 since options_len is assigned afterwards */
Fixed by initializing the counter before the options are referenced.
Matching what `tunnel_key_opts_set()` already does.
Fixes: bb5e62f2d5 ("net: Add options as a flexible array to struct ip_tunnel_info")
Cc: stable@vger.kernel.org
Signed-off-by: Gris Ge <cnfourt@gmail.com>
Reviewed-by: Hangbin Liu <liuhangbin@kylinos.cn>
Reviewed-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Link: https://patch.msgid.link/20260913090851.468216-1-cnfourt@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
In pppoatm_send(), LLC encapsulation checks whether there is sufficient
headroom for the 4-byte LLC header, but does not ensure that the skb header
is writable.
Normal transmit packets passing through ppp_start_xmit() have their header
unshared via skb_cow_head(). However, packets can also reach pppoatm_send()
via PPP channel bridging (PPPIOCBRIDGECHAN) without going through
ppp_start_xmit().
Use skb_cow_head() to ensure both sufficient headroom and a writable
header before pushing the LLC header.
While at it:
- Call pskb_may_pull(skb, 1) before inspecting skb->data[0] to prevent
out-of-bounds reads on zero-length or non-linear frames (e.g. from
bridging).
- Defer SC_COMP_PROT protocol compression until after pppoatm_may_send()
succeeds. This eliminates the temporary skb allocation on admission failure
and completely removes the fragile "undo" heuristic at the nospace label,
avoiding any risk of reading uninitialized headroom or performing an
unbalanced skb_push().
Fixes: 4cf476ced4 ("ppp: add PPPIOCBRIDGECHAN and PPPIOCUNBRIDGECHAN ioctls")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20260912233048.3977192-1-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
A batched RTM_DELACTION request takes a temporary reference on each
action before attempting any deletion. tcf_action_delete() clears
each processed slot and drops its temporary reference before attempting
the deletion. If deletion fails, tca_action_gd() calls
tcf_action_put_many() to release the remaining references, but its
tcf_act_for_each_action() iterator stops at the first NULL slot.
When a batch stops at an action bound to a filter, this leaks a
reference on each subsequent action. A later delete of an unbound
action can then return success without removing it from the IDR.
Walk the full array in tcf_action_put_many() and skip NULL slots to
release the references held on the unprocessed actions.
Fixes: a0e947c9cc ("net/sched: act_api: avoid non-contiguous action array")
Cc: stable@vger.kernel.org
Signed-off-by: Xuanqiang Luo <luoxuanqiang@kylinos.cn>
Link: https://patch.msgid.link/20260910093413.34509-2-xuanqiang.luo@linux.dev
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Commit bfdb01283e ("af_unix: Assign a unique index to SCC.")
changed Tarjan's algorithm to update lowlink with lowlink,
which is called lowpoint (unix_vertex.scc_index).
unix_vertex_dead() assumes all vertices in an SCC share the same
lowpoint, but this is not always true if an SCC has two or more
back edges, depending on the order of DFS.
For example, the graph below has two back edges from B to A
and from C to B.
A --> B --> C
^ | ^ |
`----' `----'
If DFS walks through A -> B -> C -> B (-> C -> B) -> A (-> B -> A),
each index and scc_index will be updated as follows.
A --> B --> C C = (3, 3) (index, scc_index)
B = (2, 2)
A = (1, 1)
A ... B ... C C = (3, 2)<-.
^ | B = (2, 2) -'
`----' A = (1, 1)
A ... B ... C C = (3, 2)
^ | . . B = (2, 1)<-.
`----' .... A = (1, 1) -'
Then, unix_vertex_dead() thinks that B is passed to another
SCC with scc_index 2, and the SCC is not garbage-collected.
This does not happen if DFS walks in a different order below
or starts from B.
1 3
A --> B --> C
^ | ^ |
`----' `----'
2 4
Let's unify scc_index across the SCC when finalising it.
Note that updating v->index was previously done in unix_scc_dead(),
when called from __unix_walk_scc(), just to save one loop. Since
__unix_walk_scc() now iterates over the SCC anyway, the update is
moved back to __unix_walk_scc() and 'fast' argument is dropped.
Fixes: 4090fa373f ("af_unix: Replace garbage collection algorithm.")
Reported-by: James Burton <jamesburton@meta.com>
Signed-off-by: Kuniyuki Iwashima <kuniyu@google.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20260912030852.1467872-2-kuniyu@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
rfcomm_sock_cleanup_listen() closes unaccepted child sockets through
rfcomm_sock_close(), which takes the child socket lock before
rfcomm_dlc_close() acquires rfcomm_mutex. The RFCOMM worker takes these
locks in reverse order while handling connections and DLC state changes,
so lockdep reports a possible deadlock.
Close dequeued children without taking their socket lock. The accept queue
owns a reference to each child, and bt_accept_dequeue() locks the child
while unlinking it and clearing its parent pointer.
Dropping the child lock makes it important to prevent a concurrent
rfcomm_connect_ind() from enqueueing a new child after cleanup observes an
empty queue. Set a listening socket to BT_CLOSED while its lock is still
held, before dropping the lock and draining the queue. The state check in
rfcomm_connect_ind() then rejects new children once cleanup starts.
Reported-by: syzbot+0cece8fa7d83523f47a3@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=0cece8fa7d83523f47a3
Fixes: b7ce436a5d ("Bluetooth: switch to lock_sock in RFCOMM")
Signed-off-by: Juan Perdomo <jcperdomo100@gmail.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
hci_connect_le() swaps the caller's identity address for the peer's
cached RPA when one is known, and stamps the matching
ADDR_LE_DEV_RANDOM on the local dst_type. On the conn-reuse path only
the address is copied into the connection:
if (conn) {
bacpy(&conn->dst, dst);
so conn->dst ends up holding an RPA while conn->dst_type still names the
identity it was resolved from, and hci_le_create_conn_sync() puts that
pair on air unchanged. An RPA declared as a public address is not
something any peer can answer.
Measured on a CYW43438 against a peer advertising an RPA the host holds
the IRK for, connecting to the identity address over a raw L2CAP socket.
The first attempt creates the connection, the second takes the reuse
path:
LE Create Connection 3C:78:95:78:37:C3 type public
LE Create Connection 5B:75:A2:26:D6:18 type public
LE Connection Complete: Unknown Connection Identifier (0x02)
The second address is the peer's RPA. btmon annotates it with an OUI
lookup rather than "(Resolvable)" precisely because the command declares
it public; the same bit pattern annotates as resolvable once the type is
right.
The mistyped pair is also why nothing downstream repairs it.
hci_bdaddr_is_rpa() tests the type before the address, so an RPA carrying
a public type is not recognised as one, and hci_find_irk_by_addr() then
searches for an identity address that does not match it either.
Copy the type along with the address.
The assignment used to be unconditional just below this block and covered
both paths; it moved into hci_conn_add_unset(), which the reuse path does
not go through.
Cc: stable@vger.kernel.org
Fixes: 14b06c3a88 ("Bluetooth: HCI: Always use the identity address when initializing a connection")
Assisted-by: Claude:claude-opus-5
Signed-off-by: Radek Podgorny <radek@podgorny.cz>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
A BIS connection is matched to its parent socket by looking for a
socket in BT_LISTEN state with the same BIG handle:
iso_conn_ready()
if (test_bit(HCI_CONN_BIG_SYNC, &hcon->flags))
parent = iso_get_sock(hdev, &hcon->src, &hcon->dst,
BT_LISTEN, iso_match_big_hcon, hcon);
The socket was only moved to BT_LISTEN after iso_conn_big_sync()
returned, while the LE BIG Create Sync command has already been queued
by then. If the BIG sync is established before the state is updated,
which is easy to hit with an emulated controller as the command may
complete in a few hundred microseconds, no parent is found and the BIS
connections are never notified to the listening socket.
The user space is then left waiting for connections that never arrive,
e.g. bluetoothd never completes a MediaTransport1.Acquire of a
Broadcast Sink transport.
Move the socket to BT_LISTEN before requesting the BIG sync, so the
state is visible by the time the command is queued, and restore the
previous state if the request could not be started. Since the socket is
briefly visible as a listening socket, child sockets may have been
queued in the meantime, so drain the accept queue before restoring the
state: the cleanup paths of BT_CONNECT2/BT_CONNECTED don't do it and the
children would be left with a dangling parent pointer.
Fixes: fbdc4bc472 ("Bluetooth: ISO: Use defer setup to separate PA sync and BIG sync")
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
iso_get_sock() returns the parent socket with a reference held, which is
dropped by sock_put() once the child socket has been set up. The error
path taken when iso_sock_alloc() fails only calls release_sock() and
returns, leaking the reference and thus the parent socket itself.
Drop the reference on that path as well.
Fixes: fa224d0c09 ("Bluetooth: ISO: Reassociate a socket with an active BIS")
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
hci_dev_close_sync() clears hdev->local_codecs after releasing hdev->lock.
Codec list additions and both traversals in sco_sock_getsockopt() use that
lock, but the close path does not. A close and BT_CODEC query can therefore
interleave as follows:
hci_dev_close_sync() sco_sock_getsockopt()
hci_dev_lock()
fetch codec entry
hci_codec_list_clear()
kfree(entry)
read entry->id
The reader then accesses an entry which the close path has freed. KASAN
reported:
BUG: KASAN: slab-use-after-free in sco_sock_getsockopt+0xfa0/0xfe0
Read of size 1 at addr ffff8881001c3450
Call Trace:
sco_sock_getsockopt+0xfa0/0xfe0
do_sock_getsockopt+0x537/0x7b0
__sys_getsockopt+0xf2/0x170
Allocated by task 92:
hci_codec_list_add.isra.0+0x2c/0x440
hci_read_codec_capabilities+0x224/0x590
hci_read_supported_codecs+0x2c2/0x640
Freed by task 92:
kfree+0x131/0x3c0
hci_codec_list_clear+0xd8/0x160
hci_dev_close_sync+0x92a/0xfa0
Take hdev->lock around the clear operation at its existing point in the
close path. This makes the clear wait for active readers and prevents a new
traversal until the list is empty without changing teardown ordering.
Fixes: b938790e70 ("Bluetooth: hci_codec: Fix leaking content of local_codecs")
Cc: stable@vger.kernel.org
Signed-off-by: Chengfeng Ye <nicoyip.dev@gmail.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
The Read Local Supported Codecs parsers consume the variable-sized
standard codec array before parsing the vendor codec count. Although the
initial reply-size check includes a vendor count byte in the fixed layout,
it does not guarantee that the byte remains after the standard codec array.
If a controller reply ends immediately after that array, calculating the
vendor codec array size reads vnd_codecs->num beyond the skb data. Use
skb_pull_data() to validate and consume each codec header before using its
count in both command variants.
Fixes: 8961987f3f ("Bluetooth: Enumerate local supported codec and cache details")
Fixes: 9ae664028a ("Bluetooth: Add support for Read Local Supported Codecs V2")
Cc: stable@vger.kernel.org
Suggested-by: Luiz Augusto von Dentz <luiz.dentz@gmail.com>
Signed-off-by: Laxman Acharya Padhya <acharyalaxman8848@gmail.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
An identity address only reaches a peer that is advertising an RPA if the
controller resolves it on our behalf. Where it cannot, the host has to put
the peer's on-air address on air itself.
hci_connect_le() still swaps the caller's identity address for the peer's
cached RPA before creating the connection, but __hci_conn_add() resolves
the RPA back to the identity address when it stores it, so the identity is
what goes out. Storing the identity is right when the controller
translates it on the way to the radio; without LL Privacy, or with this
peer absent from the resolving list, nothing does.
A peer advertising an RPA cannot answer its identity address, so the
attempt burns a full create-connection timeout. That is not merely a slow
connect: a controller without extended scanning cannot scan while it is
initiating, so every dead attempt also takes the scanner off the air for
the whole timeout.
Measured on a CYW43438, which reports neither LL Privacy nor extended
advertising (LE features 3f 00 00 08 00 00 00 00), against a peer
advertising a resolvable private address the host holds the IRK for, with
the connection requested on the peer's identity address:
before: LE Create Connection to the identity address, public type
1.61s -> 22.07s, then LE Create Connection Cancel
LE Connection Complete: Unknown Connection Identifier (0x02)
after: LE Create Connection to the peer's RPA, random type
LE Connection Complete: Success
Advertising reports reaching the host per second, same window, same five
unrelated devices on the adapter:
before 1s:2 [nothing from 2s through 21s] 22s:5 23s:3
after 0s:11 1s:5 2s:2 3s:5 4s:3 5s:4 ... 21s:2 22s:1 23s:2
One dead connect costs twenty seconds of scanning for every device on the
adapter, not just the one being dialled.
Keep the RPA in conn->dst unless the controller will translate the
identity address: address resolution enabled and the peer's identity
actually programmed into the resolving list. Testing ll_privacy_capable()
alone would not be enough: it reports the feature bit, not whether
resolution is switched on and not whether this peer is in the list.
Resolution is cleared with the other volatile flags on power-off and
switched off again while suspend pauses scanning, and a peer's IRK is only
programmed along the accept list path, so a direct-connect target, a peer
without HCI_CONN_FLAG_ADDRESS_RESOLUTION, and one that did not fit in a
full list are all absent from it.
With the peer programmed, the identity address stays in conn->dst and the
controller translates it: measured on an Intel controller, the host dials
the identity and LE Enhanced Connection Complete reports Resolved Public
with the peer's RPA in the separate peer resolvable private address field.
With the peer absent from the list the same setup dials the RPA itself.
Everything downstream already copes with an RPA in conn->dst: it is what
every outgoing LE connection stored before 14b06c3a88, the connection
complete event names the address that was dialled, and
le_conn_complete_evt() resolves it back to the identity once the link is
up. ISO links keep the unconditional conversion: they are created from an
existing ACL or a periodic sync and never dial this address themselves.
Keeping the RPA is only right while the peer is still using it, which is
why the preceding patch drops the cached RPA as soon as the peer is seen
advertising its identity address. Without that, a peer that turns privacy
off would be dialled on the address it abandoned rather than the one it
is answering on.
Fixes: 14b06c3a88 ("Bluetooth: HCI: Always use the identity address when initializing a connection")
Assisted-by: Claude:claude-opus-5
Assisted-by: Claude:claude-fable-5
Signed-off-by: Radek Podgorny <radek@podgorny.cz>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
hci_devcd_handle_pkt_init() arms dump_timeout and coredump producers
queue dump_rx without holding an hdev reference. Unregister leaves both
works live, so disconnecting during an active dump lets them access hdev
after hci_release_dev() frees it.
Shut down coredump processing during unregister. Close the producer gate
under dump_q.lock before disabling both works, then free the active buffer
and queued packets under hci_dev_lock. Serializing the gate with enqueue
prevents controller-specific workers from adding packets after the final
purge.
Fixes: 9695ef876f ("Bluetooth: Add support for hci devcoredump")
Reported-by: syzbot+b170dbf55520ebf5969a@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=b170dbf55520ebf5969a
Reported-by: Aby Sam Ross <abysamross@gmail.com>
Link: https://lore.kernel.org/r/20260322210849.68743-1-abysamross@gmail.com
Suggested-by: Aby Sam Ross <abysamross@gmail.com>
Reported-by: Tristan Madani <tristan@talencesecurity.com>
Link: https://lore.kernel.org/r/20260814231248.3096377-1-tristmd@gmail.com
Reported-by: Xiang Mei <xmei5@asu.edu>
Assisted-by: OpenAI Codex:gpt-5
Signed-off-by: Weiming Shi <bestswngs@gmail.com>
Reported-by: Xiang Mei <xmei5@asu.edu>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
hci_send_acl(), hci_send_sco() and hci_send_iso() queue hdev->tx_work
unconditionally. They can run from the L2CAP/SCO/ISO socket send path
while hci_dev_close_sync() is draining hdev->workqueue (HCIDEVDOWN
racing with a socket write). Since that queue_work() is not chained
work from the tx_work worker itself, __queue_work() sees the queue
marked __WQ_DRAINING, warns "cannot queue %ps on wq %s", and drops
the work:
WARNING: CPU: 1 PID: 5985 at kernel/workqueue.c:2352 __queue_work
Call Trace:
queue_work_on
l2cap_chan_send
l2cap_sock_sendmsg
...
hci_dev_close_sync() already sets HCI_CMD_DRAIN_WORKQUEUE before
draining, but only hci_cmd_work() and handle_cmd_cnt_and_timer()
check it before queuing. Route the tx_work producers through the
same guard via a shared hci_sched_tx() helper.
Fixes: 525daaea45 ("Bluetooth: hci_sync: Set HCI_CMD_DRAIN_WORKQUEUE during device close")
Reported-by: syzbot+b6919040d9958e2fc1ae@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=b6919040d9958e2fc1ae
Signed-off-by: ThangNN99 <ngocthang2710.1999@gmail.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
eir_get_service_data() reads a 16-bit UUID from the service data using
get_unaligned_le16() without first checking that the data is long enough
to hold a UUID16 (2 bytes). If a malformed EIR entry has a service data
field with only 1 byte of payload (field_len=2), eir_get_data() returns
dlen=1. The subsequent get_unaligned_le16() then reads 1 byte past the
field boundary.
Additionally, if the corrupted UUID happens to match, the length
calculation "dlen - 2" underflows to SIZE_MAX since dlen is size_t.
Current callers either pass NULL for the length parameter or bounds-check
the returned length, but future callers may not.
Add a check that dlen >= sizeof(u16) and skip fields that are too short
to contain a valid UUID16.
Fixes: 8f9ae5b3ae ("Bluetooth: eir: Add helpers for managing service data")
Cc: stable@vger.kernel.org
Signed-off-by: Aamir Ahmed <elb12345@hotmail.co.uk>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
We can hit a division by zero crash in tcp_rcvbuf_grow()
and tcp_rcv_space_adjust():
divide error: 0000 [#1] PREEMPT SMP
RIP: 0010:tcp_rcvbuf_grow+0x187/0x450 net/ipv4/tcp_input.c:939
...
grow = div_u64(((u64)rcvwin << 1) * (newval - oldval), oldval);
The division uses oldval = tp->rcvq_space.space as divisor.
When tp->rcvq_space.space is zero, this leads to a divide-by-zero
exception.
tp->rcvq_space.space is initialized in tcp_init_buffer_space():
tp->rcvq_space.space = min3(tp->rcv_ssthresh, tp->rcv_wnd,
(u32)TCP_INIT_CWND * tp->advmss);
If tcp_rmem[1] is configured to very small values (such as 1),
sk->sk_rcvbuf is initialized to 1. Then tcp_full_space(sk), which
computes (sk->sk_rcvbuf * scaling_ratio) >> 8, truncates to 0.
This sets tp->window_clamp = 0, tp->rcv_ssthresh = 0, and
tp->rcvq_space.space = 0. Later, when data arrives and DRS is invoked,
tcp_rcvbuf_grow() divides by oldval == 0.
Back in 2015, commit b1cb59cf2e ("net: sysctl_net_core: check SNDBUF
and RCVBUF for min length") ensured that net.core.rmem_default and
net.core.rmem_max cannot be set below SOCK_MIN_RCVBUF. Similarly,
SO_RCVBUF setsockopt enforces max_t(int, val * 2, SOCK_MIN_RCVBUF).
However, net.ipv4.tcp_rmem still had .extra1 = SYSCTL_ONE, allowing
arbitrarily small values.
Because SOCK_MIN_RCVBUF depends on sizeof(struct sk_buff) and cacheline
alignment, its value varies across architectures and configuration options.
Using a fixed constant of 4096 ensures a predictable, architecture-
independent lower bound that is safely above SOCK_MIN_RCVBUF everywhere
and matches the documented 4K default.
Fix this by setting tcp_rmem.extra1 to 4096 and updating the documentation.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20260912144848.3448026-1-edumazet@google.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
hhf_change() stores TCA_HHF_HH_FLOWS_LIMIT with no upper bound. A huge
hh_flows_limit lets each new heavy-hitter flow pass the
hh_flows_current_cnt check in alloc_new_hh() and forces a fixed-size
kzalloc(GFP_ATOMIC) per flow under spoofed traffic, for unbounded memory
growth.
Bound the attribute with NLA_POLICY_MAX() at 2*HH_FLOWS_CNT (the
hhf_init() default) and report the rejected value via extack. The
deprecated nested parse is kept: legacy tc does not set NLA_F_NESTED on
TCA_OPTIONS. Configs relying on hh_limit above the default were relying
on unbounded, unsafe behaviour and are not supported going forward.
hhf_init() also ran hhf_change() before setting the default
hh_flows_limit, so a user-supplied hh_limit at add time was clobbered
back to 2048. Set the default before hhf_change() so the configured
value sticks.
This is a follow-up to commit eb56a495f5 ("net/sched: hhf: clamp
quantum in change and init paths"), which bounded the quantum of the
same qdisc; the hh_flows_limit bound is the remaining unbounded knob of
that series' scope.
Conditions to recreate the bug: CAP_NET_ADMIN in a user namespace;
tc qdisc change dev X root hhf hh_limit 4294967295 succeeds and the
value is echoed by tc qdisc show, unbounding heavy-hitter flow
allocations; also tc qdisc add dev X root hhf hh_limit 500 stores 2048
instead of 500.
Fixes: 10239edf86 ("net-qdisc-hhf: Heavy-Hitter Filter (HHF) qdisc")
Cc: stable@vger.kernel.org
Reported-by: Sashiko (gemini) <sashiko-bot@kernel.org>
Closes: https://sashiko.dev/#/patchset/20260822195509.112717-1-jhs@mojatatu.com
Reviewed-by: Victor Nogueira <victor@mojatatu.com>
Tested-by: hybris <hybris@mojatatu.ai>
Signed-off-by: Jamal Hadi Salim <jhs@mojatatu.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/QDISC-B855.v1.20260911153152@mojatatu.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
This is a follow-up of one of sashiko's pre-existing bug reports.
br_mst_set_state() calls switchdev_port_attr_set() for nonzero MSTIs
while holding rcu_read_lock() which invokes the blocking switchdev
notifier chain and may sleep. Nonzero MSTI changes come from netlink
with rtnl held. Move the switchdev call before entering the rcu section and
assert that rtnl is held.
The call cannot be deferred because netlink needs its error and extack.
Also DSA reads the old bridge MST state during the callback and checks it.
A deferred callback will be late and will see the updated state.
Fixes: 3a7c1661ae ("net: bridge: mst: fix vlan use-after-free")
Signed-off-by: Nikolay Aleksandrov <razor@blackwall.org>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Link: https://patch.msgid.link/20260911105021.1385934-1-razor@blackwall.org
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
When the forward output route cannot be used in icmp_route_lookup(),
it enters the "reverse path" and calls ip_route_input() on fl4_dec.daddr,
the original packet's source address.
ip_route_input() only returns an error for truly invalid packets. For
unreachable addresses it will succeed and return an input route whose
dst.output is set to ip_rt_bug(). The existing check only rejects
RTN_LOCAL routes, so the RTN_UNREACHABLE route types can still be returned
and later used for output, syzkaller triggering a WARN_ON_ONCE()
in ip_rt_bug() as bellow:
------------[ cut here ]------------
WARNING: net/ipv4/route.c:1273 at ip_rt_bug+0x14/0x20
RIP: 0010:ip_rt_bug+0x14/0x20
Call Trace:
ip_push_pending_frames+0xfa/0x100
__icmp_send+0x905/0xf10
ip_options_compile+0xc0/0xd0
ip_rcv_finish_core+0x321/0xae0
ip_rcv+0x1de/0x260
__netif_receive_skb_one_core+0x11a/0x130
netif_receive_skb+0x7b/0x260
tun_get_user+0x11bf/0x1c10
------------[ cut here ]------------
Reject input route that is RTN_UNREACHABLE to fix it. The net warning
is only printed for RTN_LOCAL, as RTN_UNREACHABLE is not the result of
a race condition.
Fixes: 8b7817f3a9 ("[IPSEC]: Add ICMP host relookup support")
Suggested-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Jiayuan Chen <jiayuan.chen@linux.dev>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: Dong Chenchen <dongchenchen2@huawei.com>
Link: https://patch.msgid.link/20260910140042.1880242-1-dongchenchen2@huawei.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
When an SRv6 packet arrives on an interface enslaved to a VRF,
vrf_ip6_rcv() sets IP6SKB_L3SLAVE in IP6CB, but decap_and_validate()
has never set IPSKB_L3SLAVE in IPCB. The bit stayed clear in the
common case, and with CONFIG_IPV6_MIP6 the leftover frag_max_size of
a reassembled outer packet could even set it, with no VRF involved.
Commit 44930446dd ("ipv6: seg6: clear IPv4 control block on IPIP
decapsulation") then made the unreliable bit reliably clear.
The effect of the missing flag is visible with End.DX4 when a
delivery to a local address of the node reaches the socket lookup.
For example, a UDP socket bound to the enslaved ingress interface
does not receive any of the decapsulated packets, while an unbound
socket outside the VRF does.
This contradicts Documentation/networking/vrf.rst: by default the
scope of an unbound UDP or TCP socket is limited to the default VRF.
Set IPSKB_L3SLAVE for IPv4 in decap_and_validate(), which already does
the same for IPv6. The socket lookup then matches the decapsulated
packet like any other packet received on that enslaved interface. Such
a packet matches an unbound UDP or TCP socket only when
udp_l3mdev_accept or tcp_l3mdev_accept is set.
Fixes: 891ef8dd2a ("ipv6: sr: implement additional seg6local actions")
Signed-off-by: Andrea Mayer <andrea.mayer@uniroma2.it>
Reviewed-by: David Ahern <dsahern@kernel.org>
Reviewed-by: Hangbin Liu <liuhangbin@kylinos.cn>
Link: https://patch.msgid.link/20260913194421.31-1-andrea.mayer@uniroma2.it
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
rds_ib_cm_connect_complete() runs from the RDMA-CM event handler with
conn->c_cm_lock held. When the peer negotiates a protocol version
older than RDS_PROTOCOL_COMPAT_VERSION, the handler calls
rds_conn_destroy(), which is only safe in the rmmod path: it
synchronously tears the connection down and flush_work()es the
shutdown work cp_down_w.
That shutdown work (rds_conn_shutdown()) needs cp_cm_lock, which is
the very lock the event handler still holds, so the flush never
completes: the two workers wait on each other and the RDS connection
workqueues stall for good.
All other RDMA-CM failure paths (REJECTED, CONNECT_ERROR,
DISCONNECTED) use rds_conn_drop(), which marks the connection
RDS_CONN_ERROR and schedules the shutdown work asynchronously. Use
it here as well.
Fixes: f147dd9eca ("RDS/IB: Disallow connections less than RDS 3.1")
Reported-by: TencentOS Corvus AI <corvus@tencent.com>
Cc: stable@vger.kernel.org
Reviewed-by: Allison Henderson <achender@kernel.org>
Signed-off-by: Aohan Mei <henrymei@tencent.com>
Link: https://patch.msgid.link/20260911073436.3542080-1-ljp1205831794@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
-----BEGIN PGP SIGNATURE-----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=H3Kp
-----END PGP SIGNATURE-----
Merge tag 'nf-26-09-11' of git://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf
Pablo Neira Ayuso says:
====================
Netfilter fixes for net
1) Fix KMSAN reports an uninit-value in nf_nat_setup_info() for netmap,
from Theodor Arsenij Larionov Trichkine.
2) Restrict deletion of netdevice in basechain and flowtable to exact
matching only, from Fernando F. Mancera.
3) Fix nf_nat_register_fn() error path allowing for a memleak.
4) Hold reference on ct until flow is released to address, otherwise
access to release ct->ext or different ct due to typesafe RCU
semantics.
* tag 'nf-26-09-11' of git://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf:
netfilter: flowtable: hold reference on ct until flow is released
netfilter: nf_nat: unregister and release hooks on error
netfilter: nf_tables: fix device name and prefix match in hook lookup
netfilter: nft_nat: fully initialise new_addr in netmap setup
====================
Link: https://patch.msgid.link/20260913205447.1889203-1-pablo@netfilter.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
A monitor interface only gets a TXQ if it's created active, and one can't
be added later. Setting the flag on a down interface is still allowed, so
the driver is handed a monitor with no queue. ath9k dereferences it:
BUG: kernel NULL pointer dereference, address: 0000000000000066
RIP: 0010:ath_tx_node_init+0x49/0x170 [ath9k]
ath9k_add_interface+0x10c/0x140 [ath9k]
drv_add_interface+0x54/0x250 [mac80211]
ieee80211_do_open+0x32f/0x800 [mac80211]
Reached with CAP_NET_ADMIN by "iw dev X set monitor active" followed by
"ip link set X up". RTNL is held, so netlink operations block behind it.
Refuse the flag when there is no queue to give.
Fixes: 79af1f8661 ("mac80211: avoid allocating TXQs that won't be used")
Cc: stable@vger.kernel.org
Signed-off-by: Devin Wittmayer <lucid_duck@justthetip.ca>
Link: https://patch.msgid.link/20260904200338.10829-1-lucid_duck@justthetip.ca
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Adding a new station to AP makes only sense when the AP has been started
(nl80211_start_ap()) or joined a mesh (__cfg80211_join_mesh()).
Check if AP is up and beaconing on the link or joined the mesh, when
adding new station. Return error if this isn't the case.
Note that libertas devices need special handling since they do not
implement join_mesh() and the decision must be made on channel
definition.
Reported-by: syzbot+9bdc0c5998ab45b05030@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=9bdc0c5998ab45b05030
Signed-off-by: Slawomir Stepien <sst@poczta.fm>
Link: https://patch.msgid.link/20260910080418.725741-3-sst@poczta.fm
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
I do not see a reason why this check is so low in the function. Move it
up right next to param fetch.
This new position is more beneficial for AP/Link state check that will
be added in upcoming commit.
Signed-off-by: Slawomir Stepien <sst@poczta.fm>
Link: https://patch.msgid.link/20260910080418.725741-2-sst@poczta.fm
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Prevent userspace from adding stations directly to AP_VLAN type
interfaces. Userspace should first add the station to the base interface
(AP type) and then can use CMD_SET_STATION to move it to AP_VLAN. The
other way is by using NL80211_ATTR_STA_VLAN.
Without this path, we cannot check if the AP has been started before
adding the station - wdev for AP_VLAN does not store information about
the base AP interface.
Signed-off-by: Slawomir Stepien <sst@poczta.fm>
Link: https://patch.msgid.link/20260910080418.725741-1-sst@poczta.fm
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
The get_vlan() only checks if NL80211_ATTR_STA_VLAN target is an
AP/AP_VLAN/P2P_GO interface on the same wiphy. It has no notion of which
specific AP a given AP_VLAN belongs to.
Fix that by comparing the ethernet addresses of the two net devices.
Given VLAN A' must have the same address as AP A. Otherwise, return
error code.
Signed-off-by: Slawomir Stepien <sst@poczta.fm>
Reported-by: Johannes Berg <johannes@sipsolutions.net>
Link: https://lore.kernel.org/all/22e7ddfc50d7a6a16c437b876dab5fe223799610.camel@sipsolutions.net/
Link: https://patch.msgid.link/20260914081350.83484-1-sst@poczta.fm
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
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: e5c6de5fa0 ("bpf, sockmap: Incorrectly handling copied_seq")
Suggested-by: Jakub Sitnicki <jakub@cloudflare.com>
Suggested-by: Jiayuan Chen <jiayuan.chen@linux.dev>
Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
Reviewed-by: Emil Tsalapatis <emil@etsalapatis.com>
Reviewed-by: Jiayuan Chen <jiayuan.chen@linux.dev>
Link: https://lore.kernel.org/r/1a8e797a1b26e2f695aaac22ac644c2862f63466.1788858299.git.tanggeliang@kylinos.cn
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
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 <running910@gmail.com>
Reviewed-by: Emil Tsalapatis <emil@etsalapatis.com>
Link: https://lore.kernel.org/r/20260903104358.29228-1-running910@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
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: 44f0e43037 ("bpf: Add support for reading sk_state and more")
Reported-by: VEGA <vega@nebusec.ai>
Signed-off-by: Jiayuan Chen <jiayuan.chen@linux.dev>
Reviewed-by: Emil Tsalapatis <emil@etsalapatis.com>
Link: https://lore.kernel.org/r/20260903100921.113374-1-jiayuan.chen@linux.dev
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
neightbl_dump_info() calls neightbl_fill_info() in each loop
to render the default parms.
If there are many devices and neightbl_fill_param_info() failed,
neightbl_fill_info() is called again when the dump resumes:
# ynl --family rt-neigh --dump getneightbl --output-json |
jq '.[] | {name: .name, ifindex: .parms.ifindex}'
...
{
"name": "ndisc_cache",
"ifindex": null
}
...
{
"name": "ndisc_cache",
"ifindex": 6
}
{
"name": "ndisc_cache",
"ifindex": null
}
{
"name": "ndisc_cache",
"ifindex": 5
}
Let's skip neightbl_fill_info() if it is already called in
neightbl_dump_info().
Note that we cannot use !neigh_skip instead of !default_skip
because default_skip == 1 && neigh_skip == 0 could be true
if the first neightbl_fill_param_info() fails.
Also, nidx must be cleared at the end of each table loop;
otherwise, if neightbl_fill_info() for a subsequent table
fails, the leftover nidx from the previous table would be
saved in cb->args[1], resulting in erroneously skipping parms
of the subsequent table in the next dump.
Fixes: c7fb64db00 ("[NETLINK]: Neighbour table configuration and statistics via rtnetlink")
Signed-off-by: Kuniyuki Iwashima <kuniyu@google.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Link: https://patch.msgid.link/20260909233143.2401847-5-kuniyu@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
