siw_get_hdr() can receive an extended DDP/RDMAP header across more than
one TCP callback. The first callback may receive most of the header,
while the next one still limits the copy to hdrlen - MIN_DDP_HDR instead
of the number of missing bytes. This makes the destination move past the
end of the header and overwrite the receive state, including
fpdu_part_rcvd. A later callback can then use a negative fpdu_part_rcvd
value as a copy offset, which creates an OOB write.
Use the number of header bytes already received when calculating the
next copy length.
Fixes: 754209850d ("RDMA/siw: Always consume all skbuf data in sk_data_ready() upcall.")
Signed-off-by: Jérémy Jean <Jeremy.Jean@oss.cyber.gouv.fr>
Link: https://patch.msgid.link/20260908085520.1746329-1-Jeremy.Jean@oss.cyber.gouv.fr
Assisted-by: Codex:gpt-6
Acked-by: Bernard Metzler <bernard.metzler@linux.dev>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
rxe_get_mcg() publishes a newly allocated multicast group in
rxe->mcg_tree before programming the backing Ethernet multicast address
with rxe_mcast_add(), which runs outside mcg_lock. A local userspace
RDMA client reaches this path with ATTACH_MCAST on a UD QP; if
rxe_mcast_add() then returns an error (for example -ENODEV when the
backing netdev has been removed, or a propagated dev_mc_add() error),
the unwind frees the published group without removing it from the tree.
A later lookup of the same MGID dereferences the freed struct rxe_mcg
from __rxe_lookup_mcg().
Fix this by keeping the new mcg private until rxe_mcast_add() succeeds.
Split the tree publication into __rxe_publish_mcg(), call rxe_mcast_add()
before taking the tree reference, and free the still-private mcg on
failure. Because the group is never visible in mcg_tree until the
multicast address is programmed, no concurrent caller can look it up or
attach a QP to a group that is about to be torn down, so the error path
needs no conditional unwind. If another caller publishes the same MGID
while the address is being programmed, the post-add re-check under
mcg_lock finds the winner; this caller then drops its private object and
balances its own rxe_mcast_add() with rxe_mcast_del() before returning
the winner.
Reproduced by forcing the rxe_mcast_add() error return under KASAN:
without the change the next attach to the same MGID reports a
slab-use-after-free in __rxe_lookup_mcg(); with it the forced failure
returns cleanly. A no-injection attach/detach regression, including a
two-QP shared join/leave and re-attach, stays KASAN- and leak-clean.
Fixes: a926a903b7 ("RDMA/rxe: Do not call dev_mc_add/del() under a spinlock")
Signed-off-by: Michael Bommarito <michael.bommarito@gmail.com>
Link: https://patch.msgid.link/20260617022728.2770116-1-michael.bommarito@gmail.com
Reviewed-by: Zhu Yanjun <yanjun.zhu@linux.dev>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Commit 0b261d7c1c ("RDMA/rxe: Break endless pagefault loop for RO
pages") dropped the access permission test from rxe_check_pagefault()
and left only HMM_PFN_VALID. A page faulted in read-only, for example
a page-cache folio behind a PROT_READ file mapping, then satisfies the
check and ODP write operations (RDMA WRITE, RDMA READ response, SEND
payload, atomics) modify it through kmap without ever breaking CoW.
An unprivileged user can register an ODP MR over such a mapping and
have incoming RDMA traffic overwrite the page cache of a file it only
holds O_RDONLY, including /etc/passwd or setuid binaries. This is the
same primitive class as Dirty COW and CVE-2022-2590.
mlx5 has the missing invariant: its ODP path sets the device write bit
only for pfns that carry HMM_PFN_WRITE. Restore it in rxe by requiring
HMM_PFN_WRITE in rxe_check_pagefault() for every operation except
RXE_PAGEFAULT_RDONLY. A write to a non-writable VMA now fails the one
fault attempt with -EPERM from hmm_vma_fault() instead of re-faulting
forever. For a writable VMA the fault breaks CoW and the write lands
in the private page.
Keep pmem flushes on the read-only check. arch_wb_cache_pmem() never
modifies memory, and the FLUSH access bits do not make the umem
writable, so classifying flushes as writes would make every flush
against a flush-only MR fail.
Fixes: 0b261d7c1c ("RDMA/rxe: Break endless pagefault loop for RO pages")
Signed-off-by: Weiming Shi <bestswngs@gmail.com>
Link: https://patch.msgid.link/20260726111533.1037819-1-bestswngs@gmail.com
Reviewed-by: Zhu Yanjun <yanjun.zhu@linux.dev>
Tested-by: Hongqiang Luo <wanbafv@gmail.com>
Tested-by: Xinyu Ma <mmmxny@gmail.com>
Tested-by: Zhanbo Ye <cainyzb@gmail.com>
Reported-by: Weiming Shi <bestswngs@gmail.com>
Reported-by: Shaomin Chen <eeesssooo020@gmail.com>
Reported-by: Rui Ding <threonine42@gmail.com>
Reported-by: Miao Zhao <muel@nova.gal>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
mr_check_range() validates that [iova, iova+length) falls within the
registered MR range using wraparound-prone arithmetic:
if (iova < mr->ibmr.iova ||
iova + length > mr->ibmr.iova + mr->ibmr.length)
A remote peer can craft an RDMA-Write/Read RETH so that iova + length
wraps to 0 (e.g. iova=0xfffffffffffffff8, length=8), bypassing the
check. rxe_mr_iova_to_index() then computes a huge index (int idx, only
guarded by WARN_ON) and rxe_mr_copy_xarray() dereferences
mr->page_info[huge], causing an out-of-bounds read/write and a kernel
oops that is triggerable by an unauthenticated remote peer.
Rewrite the check in overflow-safe form; the first two clauses guarantee
that the subsequent subtractions do not underflow:
if (iova < mr->ibmr.iova ||
length > mr->ibmr.length ||
iova - mr->ibmr.iova > mr->ibmr.length - length)
With the fix, mr_check_range() returns -EINVAL for the crafted iova and
the responder reports REMOTE_ACCESS_ERROR instead of triggering the OOB.
Fixes: 8700e3e7c4 ("Soft RoCE driver")
Signed-off-by: Gang Yan <yangang@kylinos.cn>
Link: https://patch.msgid.link/20260814093740.292954-1-gang.yan@linux.dev
Reviewed-by: Zhu Yanjun <yanjun.zhu@linux.dev>
Reviewed-by: Shukai Ni <shukai.ni@kuleuven.be>
Tested-by: Shukai Ni <shukai.ni@kuleuven.be>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
rxe_rereg_user_mr() reassigns mr->ibmr.pd first and only then
validates the IB_MR_REREG_ACCESS argument:
if (flags & IB_MR_REREG_PD) {
rxe_put(old_pd);
rxe_get(pd);
mr->ibmr.pd = ibpd;
}
if (flags & IB_MR_REREG_ACCESS) {
if (access & ~RXE_ACCESS_SUPPORTED_MR)
return ERR_PTR(-EOPNOTSUPP);
mr->access = access;
}
Both flags pass the entry check because RXE_MR_REREG_SUPPORTED is
IB_MR_REREG_PD | IB_MR_REREG_ACCESS, so a caller can reach the access
check with mr->ibmr.pd already reassigned.
mr->ibmr.pd is owned by the core, which adjusts pd->usecnt only on the
success path: ib_uverbs_rereg_mr() jumps to put_new_uobj on a driver error
without undoing the reassignment, so mr->pd == new_pd while the usecnts
still charge the MR to orig_pd. ib_dereg_mr_user() then decrements
new_pd, whose count can reach zero while a memory window still references
it; uverbs_free_pd() frees the PD on that count alone and rxe_mw_cleanup()
writes to freed memory:
BUG: KASAN: slab-use-after-free in __rxe_put+0x31/0xa0
Write of size 4 at addr ffff8881301dd690 by task rxe_poc/591
__rxe_put+0x31/0xa0
rxe_mw_cleanup+0x42/0x200
__rxe_cleanup+0x115/0x370
rxe_dealloc_mw+0x4c/0x80
Allocated by task 591:
ib_uverbs_alloc_pd+0x258/0x540
Freed by task 591:
ib_dealloc_pd_user+0x174/0x210
uverbs_free_pd+0x8d/0xc0
ib_uverbs_dealloc_pd+0x18e/0x1d0
Validate the access flags before mutating any state so the callback either
applies every requested change or none.
Fixes: 544c7f62cf ("RDMA/rxe: Implement rereg_user_mr")
Signed-off-by: Norbert Szetei <norbert@doyensec.com>
Link: https://patch.msgid.link/46E1D5C0-24BE-4D01-BDB3-634FE09B22C5@doyensec.com
Reviewed-by: Zhu Yanjun <yanjun.zhu@linux.dev>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
We need to clear cep before release state_lock as siw_qp_llp_close and
siw_qp_modify->siw_qp_llp_close did.
Otherwise if siw_qp_modify() fails in siw_accept(), the QP's state_lock
is released before the error path cleanup. A concurrent ibv_modify_qp()
transitioning the QP to ERROR can race in this window:
siw_accept() ibv_modify_qp(ERROR)
---------------------- ----------------------
siw_qp_modify() fails
up_write(&qp->state_lock)
down_write(&qp->state_lock)
nextstate_from_idle():
if (qp->cep)
siw_cep_put(qp->cep) <- frees cep
qp->cep = NULL
goto error
cep->qp = NULL <- UAF
Clear qp->cep and drop the association reference taken by siw_cep_get(),
all under the write lock held from the initial down_write(&qp->state_lock).
Thread B therefore sees qp->cep == NULL, skips its own put, and cannot free
the cep before siw_accept() is done with it.
Fixes: 6c52fdc244 ("rdma/siw: connection management")
Reported-by: Shuangpeng Bai <shuangpeng.kernel@gmail.com>
Link: https://lore.kernel.org/linux-rdma/d6fbe475-a5c2-f975-99b0-a0bd6b6d10e8@linux.dev/T/#m5876c1ff2de8686a9a1173b8f1aa0ff5363a785c
Signed-off-by: Guoqing Jiang <guoqing.jiang@linux.dev>
Link: https://patch.msgid.link/20260827125553.12831-1-guoqing.jiang@linux.dev
Acked-by: Bernard Metzler <bernard.metzler@linux.dev>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Quite alot of buf fixes again:
- Assorted locking, bounds-checking, cleanup, and error-path fixes across
UCMA/CMA, bng_re, bnxt_re, cxgb4, EFA, ERDMA, HFI1, HNS, ionic, iRDMA,
mlx4/mlx5, RXE, SIW, SRP/SRPT, and iSER target.
- netlink report for max # of supported resources
- get_zeroed_page()/etc removal
- Robust udata for ionic
- Allow unique RDMA device names per network namespace
- Completion counters and v2 admit queue support for EFA
- UC QP support for MANA
- Completion timestamps for ionic
- Harden uverbs data validation and resource lifetime handling, fixing several core use-after-free conditions.
- bnxt_re toggle-page ownership and lifetime bug fixes
- dmabuf SRQ support for mlx5
-----BEGIN PGP SIGNATURE-----
iHUEABYKAB0WIQRRRCHOFoQz/8F5bUaFwuHvBreFYQUCaoeO2wAKCRCFwuHvBreF
YcyJAQCn8JeqTuayLchCARJX+9fazTaJES9zj41i3M8a7BfxCAEA2b266g0S660m
7eQ761NiIb1iklSl9rmt8vM22EMcngM=
=qcwp
-----END PGP SIGNATURE-----
Merge tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma
Pull RDMA updates from Jason Gunthorpe:
"About the normal size, still a lot of AI bug fixes and so on, but some
interesting new functionality too:
- Assorted locking, bounds-checking, cleanup, and error-path fixes
across UCMA/CMA, bng_re, bnxt_re, cxgb4, EFA, ERDMA, HFI1, HNS,
ionic, iRDMA, mlx4/mlx5, RXE, SIW, SRP/SRPT, and iSER target.
- netlink report for max # of supported resources
- get_zeroed_page()/etc removal
- Robust udata for ionic
- Allow unique RDMA device names per network namespace
- Completion counters and v2 admit queue support for EFA
- UC QP support for MANA
- Completion timestamps for ionic
- Harden uverbs data validation and resource lifetime handling,
fixing several core use-after-free conditions.
- bnxt_re toggle-page ownership and lifetime bug fixes
- dmabuf SRQ support for mlx5"
* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma: (160 commits)
RDMA/ucma: Allow path records to exactly fit the output buffer
RDMA/uverbs: Guard legacy bundles without method_elm
RDMA/efa: Add support for 128B admin v2 SQ entry
RDMA/efa: Generalize the admin SQ
RDMA/efa: Decouple admin command payload from admin header
RDMA/rxe: Fix OOB in free_rd_atomic_resources()
RDMA/cma: Fix WARNING in res_to_rt
RDMA/cxgb4: Free debugfs on registration failure
RDMA/cxgb4: Cancel reg_work before freeing device on remove
RDMA/ucma: Lock the handler in ucma_set_ib_path()
RDMA/ucma: Lock the handler in ucma_write_cm_event()
RDMA/erdma: restrict the driver to little-endian systems
RDMA/ionic: Embed counter driver data in rdma_counter allocation
RDMA/ionic: Cap eq_count to the eth driver's interrupt vector budget
RDMA/siw: Fix use-after-free in siw_accept()
IB/isert: post the full-feature receive buffers after session registration
IB/isert: delay the final Login Response until the session is registered
RDMA/srp: fix heap information leak on a truncated SRP_CRED_REQ
RDMA/erdma: Hold QP references for AE and CM processing
RDMA/erdma: Hold CQ references when processing EQ events
...
These did not seem worth sending as a dedicated rc PR during the last week
of the cycle.
* ko-rdma/for-rc:
RDMA/ipoib: Drain RCU callbacks during module teardown
RDMA/mlx5: Drain RCU callbacks during module teardown
RDMA/core: Wait for RCU callbacks before unloading ib_core
RDMA/irdma: Prevent overflows in memory contiguity checks
RDMA/siw: publish QP after initialization
RDMA/hns: Fix potential integer overflow in mhop hem cleanup
RDMA/core: Fix memory leak in __ib_create_cq() on invalid cqe
RDMA/mana_ib: initialize err for empty send WR lists
RDMA/erdma: initialize ret for empty receive WR lists
RDMA/irdma: Prevent user-triggered null deref on QP create
RDMA/irdma: Prevent rereg_mr for non-mem regions
RDMA/cma: Fix hardware address comparison length in netevent callback
RDMa/mlx5: Avoid frame overflow warning
IB/mad: Drop unmatched RMPP responses before reassembly
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
siw_accept() looks up the QP supplied by userspace. If that QP is
already in RTS, the function jumps to error cleanup before associating
the incoming CEP with it.
The cleanup tests whether qp->cep is non-NULL and assumes the current
call installed the association. However, qp->cep can point to the CEP
of an existing connection. The cleanup then drops a reference from the
incoming cep, not qp->cep. Once the incoming endpoint loses its
remaining references, this can free it before the subsequent cep->qp
store, causing a use-after-free. It also clears the existing QP
association.
Only release the association reference when qp->cep is the incoming
CEP. This preserves an existing association and avoids accessing the
freed endpoint.
Fixes: 6c52fdc244 ("rdma/siw: connection management")
Signed-off-by: Shuangpeng Bai <shuangpeng.kernel@gmail.com>
Link: https://patch.msgid.link/20260801213632.1086548-1-shuangpeng.kernel@gmail.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
rxe_odp_mr_init_user() stores &umem_odp->umem in mr->umem before
calling rxe_odp_init_pages(). If rxe_odp_init_pages() fails,
rxe_odp_mr_init_user() releases umem_odp and returns an error.
rxe_reg_user_mr() then unwinds the error through rxe_cleanup(),
rxe_mr_cleanup(), ib_umem_release(mr->umem). There is an
IS_ERR_OR_NULL(umem) check at the start of ib_umem_release().
But since mr->umem is NOT reset to NULL in the error handling
path of rxe_odp_mr_init_user(), the check passes and it reads
already-freed fields like umem->is_dmabuf, causing UAF.
Fix the UAF by clearing mr->umem after releasing the failed
ODP umem so the MR cleanup path does not release it again.
Fixes: d03fb5c659 ("RDMA/rxe: Allow registering MRs for On-Demand Paging")
Assisted-by: Codex:gpt-5.6-sol
Signed-off-by: Peiyang He <peiyang_he@smail.nju.edu.cn>
Link: https://patch.msgid.link/70CB6DBCB19624C7+20260727050659.1543627-1-peiyang_he@smail.nju.edu.cn
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Implement an empty disassociate_ucontext() callback so the RDMA core
can move rxe devices between net namespaces. The core requires this
callback to reset user contexts without waiting for userspace.
rxe needs no teardown here: its user-mapped queues live in
reference-counted vmalloc memory (see rxe_mmap.c) that stays valid
while userspace holds the mappings.
Signed-off-by: Jiri Pirko <jiri@nvidia.com>
Link: https://patch.msgid.link/20260716132316.1495242-14-jiri@resnulli.us
Reviewed-by: Yanjun Zhu <yanjun.zhu@linux.dev>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Prepare queue mappings for asynchronous ucontext disassociation during
device disable. Rely on the VMA page references to preserve mapped
memory until the final unmap.
Signed-off-by: Jiri Pirko <jiri@nvidia.com>
Link: https://patch.msgid.link/20260716132316.1495242-13-jiri@resnulli.us
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Alice Mikityanska says:
====================
BIG TCP for UDP tunnels
This series is a follow-up to "BIG TCP without HBH in IPv6", and it adds
support for BIG TCP IPv4/IPv6 workloads in vxlan and geneve. Now that
IPv6 BIG TCP doesn't require stripping the HBH in all various
combinations in tunneled traffic, adding BIG TCP becomes feasible.
Patch 01 adds accessors for the length field in the UDP header, as
suggested by Paolo in review. The usage of udp_set_len is then added in
the following patches that start using length=0 in BIG TCP UDP packets.
Patches 02-04 close the gaps that prevent BIG TCP packets from going
through UDP tunnel code.
Patch 05 validates packets in udp_gro_receive to exclude packets with
length=0 from GRO aggregation.
Patch 06 is for proper formatting in tcpdump (set UDP len to 0 rather
than a trimmed value on overflow).
Patches 07-08 bump up tso_max_size for VXLAN and GENEVE.
Patch 09 adds selftests.
====================$
Link: https://patch.msgid.link/20260710134242.216538-1-alice.kernel@fastmail.im
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Since BIG TCP for UDP tunnels will start using len=0 in the UDP header
as an indicator of a GSO packet bigger than 65535 bytes, this commit
introduces the following getter and setters to use tree-wide, in order
to explicitly mark places where len=0 may be expected, and handle them
properly:
1. udp_set_len() sets uh->len to its real value if it's not bigger than
65535, and to 0 otherwise: to be used in GSO context with aggregated
packets.
2. udp_set_len_short() is to be used when the length is known to fit 16
bits. It WARNs when the caller tries to assign a bigger value if
CONFIG_DEBUG_NET=y.
3. udp_get_len_short() returns len in host byte order: to be used on the
RX side to deal with non-aggregated packets, or to access the raw value
of the len field.
4. udp_get_len() decodes uh->len set by udp_set_len(). It checks whether
the packet is GSO to guard from malformed packets.
At the moment udp_set_len() is not used, a following commit will start
using it after enabling len>65535 for GSO.
Raw uh->len (in network byte order) is still accessed in a few places
for checksum calculation purposes, and to decode len=0 in udpv6_rcv for
jumbograms. udp_rcv and udpv6_rcv will be addressed by the commit that
starts using udp_set_len() to set UDP len=0 for BIG TCP packets in UDP
tunnels.
Signed-off-by: Alice Mikityanska <alice@isovalent.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Acked-by: Jason A. Donenfeld <Jason@zx2c4.com>
Link: https://patch.msgid.link/20260710134242.216538-2-alice.kernel@fastmail.im
Reviewed-by: Nikolay Aleksandrov <razor@blackwall.org>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
The query_device callbacks that neither accept driver-specific input nor
return a driver-specific response open-code the empty udata handling as
ib_is_udata_in_empty() on entry and ib_respond_empty_udata() on exit.
ib_no_udata_io() already combines both steps, so replace the entry check
with it and simply return 0 on success.
Unlike the create and destroy flows, query_device owns no uobject or HW
resource - the extended path fills a stack ib_device_attr that the core
discards on error - so clearing the empty response buffer on entry rather
than on exit is a mechanical change with no functional difference.
Link: https://patch.msgid.link/20260714-convert-to-noio-udata-v1-1-f1f6b6c7c988@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
For a user QP, qp->sq.queue is a ring the application writes directly,
so rxe_post_send() takes the is_user branch and only schedules send_task
without validating the WQE. rxe_requester() consumes it in place via
req_next_wqe() and calls copy_data(), which indexes
&wqe->dma.sge[cur_sge] with the attacker-controlled num_sge/cur_sge.
Only the kernel path bounds num_sge (validate_send_wr()); the user WQE
is never checked, so a local unprivileged user can post a WQE with an
out-of-range cur_sge or oversized num_sge and force an out-of-bounds
read of the per-WQE sge array in copy_data() (vmalloc OOB read, local
DoS).
Bound num_sge to qp->sq.max_sge in rxe_requester() before use, the way
get_srq_wqe() already guards SRQ entries, and bound cur_sge only when
the WQE carries payload (dma.resid): copy_data() returns early on a
zero-length copy before touching dma->sge[], so a zero-payload WQE --
the only kind a max_sge == 0 QP can post -- stays valid.
Reproduced under KASAN; the vmalloc-out-of-bounds in copy_data() is gone.
Fixes: 8700e3e7c4 ("Soft RoCE driver")
Reviewed-by: Zhu Yanjun <yanjun.zhu@linux.dev>
Signed-off-by: Ibrahim Hashimov <security@auditcode.ai>
Link: https://patch.msgid.link/20260712122149.78142-1-security@auditcode.ai
Assisted-by: AuditCode-AI:2026.07
Signed-off-by: Leon Romanovsky <leon@kernel.org>
rxe_qp_from_attr() handles IB_QP_MAX_DEST_RD_ATOMIC outside the
IB_QP_STATE path, so it holds no state_lock and runs while the responder
task rxe_receiver() (recv_task on rxe_wq) is live. A modify_qp() setting
only that attribute calls free_rd_atomic_resources() then
alloc_rd_atomic_resources(), swapping qp->resp.resources[] while
rxe_prepare_res()/find_resource() walk it; free_rd_atomic_resources()
also leaves the cached pointer qp->resp.res dangling. A local
unprivileged user can race the free/realloc into a use-after-free in
rxe_receiver() (local DoS).
Drain recv_task around the swap with rxe_disable_task()/rxe_enable_task(),
as rxe_qp_reset() already does when tearing this array down, re-enabling
only after alloc_rd_atomic_resources() succeeds so the responder never
resumes against a NULL qp->resp.resources on the ENOMEM path. Also clear
qp->resp.res in free_rd_atomic_resources(), like the rxe_resp.c
completion paths.
Reproduced under KASAN; the slab-use-after-free in rxe_receiver() is gone.
Fixes: 8700e3e7c4 ("Soft RoCE driver")
Reviewed-by: Zhu Yanjun <yanjun.zhu@linux.dev>
Signed-off-by: Ibrahim Hashimov <security@auditcode.ai>
Link: https://patch.msgid.link/20260712121720.78001-1-security@auditcode.ai
Assisted-by: AuditCode-AI:2026.07
Signed-off-by: Leon Romanovsky <leon@kernel.org>
rxe advertises ODP but not IB_ODP_SUPPORT_IMPLICIT. The reg_user_mr path
still contained a dead branch that checked the implicit capability and
could never succeed.
Return -EOPNOTSUPP for the implicit ODP address range up front so the
intent is obvious and the unreachable code is gone.
Signed-off-by: xiongweimin <xiongweimin@kylinos.cn>
Cc: linux-rdma@vger.kernel.org
Cc: Jason Gunthorpe <jgg@nvidia.com>
Link: https://patch.msgid.link/20260713010439.331054-1-15927021679@163.com
Reviewed-by: Zhu Yanjun <yanjun.zhu@linux.dev>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
get_map_page() allocates bitmap pages using get_zeroed_page().
The bitmaps can be allocated with kmalloc() as there's nothing special
about them to go directly to the page allocator.
kmalloc() provides a better API that does not require ugly casts and
kfree() does not need to know the size of the freed object.
Performance difference between kmalloc() and __get_free_pages() is not
measurable as both allocators take an object/page from a per-CPU list for
fast path allocations.
For the slow path the performance is anyway determined by the amount of
reclaim involved rather than by what allocator is used.
Replace use of get_zeroed_page() with kzalloc() and free_page() with
kfree().
Link: https://lore.kernel.org/all/635405e4-9423-4a25-a6e7-e03c8ea0bcbe@redhat.com
Signed-off-by: Mike Rapoport (Microsoft) <rppt@kernel.org>
Link: https://patch.msgid.link/20260713-b4-rdma-v2-5-65d2a1a5180c@kernel.org
Signed-off-by: Leon Romanovsky <leon@kernel.org>
The core always hands the driver's query_device() callback a zeroed
ib_device_attr. There are only two callers of the op and both clear the
structure before invoking it: setup_device() memsets &device->attrs, and
ib_uverbs_ex_query_device() passes an on-stack structure initialized to {}.
The open-coded memset(props, 0, sizeof(*props)) at the top of the driver
callbacks is therefore redundant. Remove it from all drivers.
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
rvt_alloc_device() deallocates the IB device when its port array cannot
be allocated but then returns the pointer to the released allocation.
Callers treat any non-NULL value as valid and dereference it, resulting
in a use-after-free.
Return NULL immediately after deallocation so callers can propagate the
allocation failure.
Fixes: ff6acd6951 ("IB/rdmavt: Add device structure allocation")
Link: https://patch.msgid.link/20260708-clean-init-one-hfi1-v1-1-b9e9641268a5@nvidia.com
Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
When do_complete() finds the QP in the error state it returns
RESPST_CHK_RESOURCE. Before commit 49dc9c1f0c ("RDMA/rxe: Cleanup
reset state handling in rxe_resp.c") this was the flush loop:
check_resource() had an error-state branch that fetched each remaining
recv WQE and completed it with IB_WC_WR_FLUSH_ERR, without touching
the current packet. That commit removed the error-state branch from
check_resource() (draining is now done at rxe_receiver() entry) but
kept the do_complete() error-state return.
As a result, when a QP moves to the error state while a packet is
being completed - e.g. an rdma_cm disconnect racing with receive
processing - the responder state machine loops back into the request
processing chain with the already-completed packet still in hand:
check_resource() fetches a fresh recv WQE, execute()/send_data_in()
copies the same packet payload again, do_complete() posts another
IB_WC_SUCCESS CQE (qp->resp.status is still 0), and control returns
to the error-state check. The loop re-executes the same packet once
per posted recv WQE (observed: ~1000 duplicate IB_WC_SUCCESS
completions of one SEND, one per ~8us, matching the RQ occupancy)
until the RQ is exhausted, after which qp->resp.wqe is NULL and
send_data_in() dereferences it:
BUG: kernel NULL pointer dereference, address: 0000000000000014
Workqueue: rxe_wq do_work
RIP: copy_data+0x29/0x1f0
Call Trace:
send_data_in+0x25/0x50
rxe_receiver+0xf36/0x1dd0
The duplicate completions are indistinguishable from real receives to
the ULP. During an rds stress test, the message was accepted as new and
delivered the same datagram to user space hundreds of times, corrupting
the stream; any ULP that relies on RC exactly-once delivery is affected.
A live packet reaching the error-state check in do_complete() has
been executed and completed exactly once and must be consumed, not
re-processed. Return RESPST_CLEANUP for it (dequeue and free); keep
returning RESPST_CHK_RESOURCE for the pkt == NULL case.
Fixes: 49dc9c1f0c ("RDMA/rxe: Cleanup reset state handling in rxe_resp.c")
Assisted-by: Claude-Code:claude-fable-5
Signed-off-by: Allison Henderson <achender@kernel.org>
Link: https://patch.msgid.link/20260711165419.13486-1-achender@kernel.org
Reviewed-by: Zhu Yanjun <yanjun.zhu@linux.dev>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
The capability counter fields in struct ib_device_attr are declared
as signed int, but these values are inherently non-negative. Drivers
maintain their cached caps as u32 and assign them directly into these
int fields; if a cap exceeds INT_MAX the implicit narrowing yields a
negative value visible to the IB core.
Change the signed int capability fields to u32 to match the
underlying nature of the data. Also update consumers across the IB
core, ULPs, NVMe-oF target, RDS, and NFS/RDMA so the new u32 values
are not forced back through signed int or u8 via min()/min_t() or
narrowing local variables.
The nvmet-rdma consumer of max_srq clamps it against
ib_device.num_comp_vectors, which stays a signed int, so that site
uses min_t() instead of min() to handle the signed/unsigned mismatch.
Suggested-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Erni Sri Satya Vennela <ernis@linux.microsoft.com>
Link: https://patch.msgid.link/20260709055211.2498307-1-ernis@linux.microsoft.com
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Acked-by: Stefan Metzmacher <metze@samba.org> # smbdirect
Signed-off-by: Leon Romanovsky <leon@kernel.org>
The IBTA Software Transport Verbs specification requires the QP,
Memory Window and Memory Region for a Bind Memory Window operation
to belong to the same HCA and protection domain.
rxe only checked the QP and MW protection domain for type 2 MWs.
Move the QP/MW PD check to the common bind path and also reject
binding an MW to an MR from a different PD.
Invalid bind requests continue to fail with IB_WC_MW_BIND_ERR.
Reviewed-by: Zhu Yanjun <yanjun.zhu@linux.dev>
Signed-off-by: Zhiwei Zhang <202275009@qq.com>
Link: https://patch.msgid.link/tencent_FD4FB25AA4FFA845E63F5AC36CF4A46CDC0A@qq.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
siw_create_qp() currently calls siw_qp_add() before the queues, CQ
pointers, state, completion, and device list entry are ready. A QPN
lookup can therefore reach a QP that is still being constructed.
Move siw_qp_add() to the end of siw_create_qp(), after QP
initialization and before adding the QP to the siw device list.
Fixes: f29dd55b02 ("rdma/siw: queue pair methods")
Link: https://patch.msgid.link/r/20260630060040.966461-1-ruoyuw560@gmail.com
Suggested-by: Bernard Metzler <bernard.metzler@linux.dev>
Signed-off-by: Ruoyu Wang <ruoyuw560@gmail.com>
Acked-by: Bernard Metzler <bernard.metzler@linux.dev>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Many AI driven bug fixes, and several big driver API cleanups
- Driver bug fixes and minor cleanups in mlx5, hns, rxe, efa, siw, rtrs,
mana, irdma, mlx4. Commonly error path flows, integer arithmetic
overflows on unsafe data, out of bounds access, and use after free
issues under races.
- Second half of the new udata API for drivers focusing on uAPI response
- bnxt_re supports more options for QP creation that will allow a dv
path in rdma-core
- Untangle the module dependencies so drivers don't link to ib_uverbs.ko
as was originall intended
- Provide a new way to handle umems with a consistent simplified uAPI and
update several drivers to use it. This brings dmabuf support to more
places and more drivers
- Support for mlx5 rate limit and packet pacing for UD and UC
- A batch of fixes for the new shared FRMR pools infrastructure
-----BEGIN PGP SIGNATURE-----
iHUEABYKAB0WIQRRRCHOFoQz/8F5bUaFwuHvBreFYQUCajLevAAKCRCFwuHvBreF
YVFuAP0X6gxaON+6cRGoaHp+KjxNj3zbuRj9VccTOy1OjerUdwD+LsbK/Pa9xrcO
ByA3LVvbw4BoQw9qcQpTu2klFuYV5g8=
=qnpN
-----END PGP SIGNATURE-----
Merge tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma
Pull rdma updates from Jason Gunthorpe:
"Many AI driven bug fixes, and several big driver API cleanups
- Driver bug fixes and minor cleanups in mlx5, hns, rxe, efa, siw,
rtrs, mana, irdma, mlx4. Commonly error path flows, integer
arithmetic overflows on unsafe data, out of bounds access, and use
after free issues under races.
- Second half of the new udata API for drivers focusing on uAPI
response
- bnxt_re supports more options for QP creation that will allow a dv
path in rdma-core
- Untangle the module dependencies so drivers don't link to
ib_uverbs.ko as was originall intended
- Provide a new way to handle umems with a consistent simplified uAPI
and update several drivers to use it. This brings dmabuf support to
more places and more drivers
- Support for mlx5 rate limit and packet pacing for UD and UC
- A batch of fixes for the new shared FRMR pools infrastructure"
* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma: (148 commits)
RDMA/irdma: Replace waitqueue and flag with completion
RDMA/hns: Fix memory leak of bonding resources
RDMA/rtrs-srv: Bound RDMA-Write length to chunk size in rdma_write_sg
docs: infiniband: correct name of option to enable the ib_uverbs module
RDMA/bnxt_re: Reject GET_TOGGLE_MEM when toggle page was not allocated
RDMA/bnxt_re: Fail DBR related page allocation UAPIs if the feature is disabled
RDMA/bnxt_re: Avoid repeated requests to allocate WC pages
RDMA/bnxt_re: Proper rollback if the ioremap fails
RDMA/bnxt_re: Add a max slot check for SQ
RDMA/bnxt_re: Avoid displaying the kernel pointer
RDMA/bnxt_re: Free CQ toggle page after firmware teardown
RDMA/bnxt_re: Free SRQ toggle page after firmware teardown
RDMA/bnxt_re: Initialize dpi variable to zero
ABI: sysfs-class-infiniband: minor cleanup
RDMA/mlx5: Release the HW‑provided UAR index rather than the SW one
RDMA/mlx5: Fix undefined shift of user RQ WQE size
RDMA/mlx5: Remove raw RSS QP restrack tracking
RDMA/mlx5: Remove DCT restrack tracking
RDMA/mlx5: Drop FRMR pool handle on UMR revoke failure
RDMA/core: Add ib_frmr_pool_drop for unrecoverable handles
...
Cross-merge networking fixes after downstream PR (net-7.1-rc8).
Conflicts:
drivers/net/ethernet/wangxun/txgbe/txgbe_aml.c
f67aead16e ("net: txgbe: rework service event handling")
57d39faed4 ("net: txgbe: improve functions of AML 40G devices")
net/rds/info.c
512db8267b ("rds: mark snapshot pages dirty in rds_info_getsockopt()")
6e94eeb2a2 ("rds: convert to getsockopt_iter")
Adjacent changes:
include/net/sock.h
1ee90b77b7 ("net: guard timestamp cmsgs to real error queue skbs")
f0de88303d ("net: make is_skb_wmem() available to modules")
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
If IB_MR_REREG_ACCESS changes from RO to RW then the umem has to be
re-evaluated to ensure it is properly pinned as RW. Since the umem is
hidden inside each driver's mr struct add a ib_umem_check_rereg() function
that each driver has to call before processing IB_MR_REREG_ACCESS.
mlx4 has to retain its duplicate ib_access_writable check because it
implements IB_MR_REREG_ACCESS | IB_MR_REREG_TRANS by changing both items
in place sequentially while the MR is live, so it will continue to not
support this combination.
Cc: stable@vger.kernel.org
Fixes: b40656aa7d ("RDMA/umem: remove FOLL_FORCE usage")
Link: https://patch.msgid.link/r/0-v1-06fb1a2d6cf5+107-rereg_access_jgg@nvidia.com
Reported-by: Philip Tsukerman <philiptsukerman@gmail.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Disassociating a socket from an endpoint via siw_socket_disassoc() may
release the last reference on that endpoint and free it. Therefore, don't
clear the endpoints socket pointer after calling that function, but
within.
This fixes a:
BUG: KASAN: slab-use-after-free in siw_cm_work_handler (drivers/infiniband/sw/siw/siw_cm.c:1053 drivers/infiniband/sw/siw/siw_cm.c:1075)
which occurred after processing a malformed MPA request during connection
establishment, causing the new endpoint to be closed.
Fixes: 6c52fdc244 ("rdma/siw: connection management")
Link: https://patch.msgid.link/r/20260604160808.30948-1-bernard.metzler@linux.dev
Reported-by: Shuangpeng Bai <shuangpeng.kernel@gmail.com>
Signed-off-by: Bernard Metzler <bernard.metzler@linux.dev>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
In drivers/infiniband/sw/siw/siw_qp_rx.c, siw_proc_rresp() places each
inbound Read Response DDP segment at sge->laddr + wqe->processed and then
accumulates wqe->processed, but it never checks the running total against
the sink buffer length on continuation segments. siw_check_sge() resolves
and validates the sink memory only on the first fragment (the if (!*mem)
branch), and siw_rresp_check_ntoh() compares the cumulative length against
wqe->bytes only on the final segment (the !frx->more_ddp_segs guard).
A connected siw peer that answers an outstanding RREAD with Read Response
segments that keep the DDP Last flag clear, carrying more total payload
than the RREAD requested, drives wqe->processed past the validated sink
buffer; the next siw_rx_data() call writes out of bounds at
sge->laddr + wqe->processed. siw runs iWARP over ordinary routable TCP,
so the peer is the remote end of an established RDMA connection and needs
no local privilege.
Bound every segment before placement, exactly as siw_proc_send() and
siw_proc_write() already do for their tagged and untagged paths, and
terminate the connection with a base-or-bounds DDP error when the
Read Response would overrun the sink buffer.
This is the second receive-path length fix for this file. A separate
change rejects an MPA FPDU length that underflows the per-fragment
remainder in the header decode; that guard does not cover this case,
because here each individual segment length is self-consistent and only
the accumulated placement offset overruns the buffer.
Fixes: 8b6a361b8c ("rdma/siw: receive path")
Link: https://patch.msgid.link/r/20260602194700.2273758-1-michael.bommarito@gmail.com
Cc: stable@vger.kernel.org
Assisted-by: Claude:claude-opus-4-8
Signed-off-by: Michael Bommarito <michael.bommarito@gmail.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
This op hasn't followed the normal pattern of passing NULL for udata when
invoked by the kernel. Instead the kernel caller creates a dummy ib_udata
on the stack and passes that in. It does not seem to currently be a bug,
but this flow should be modernized to use the new API flow and in the
process accept NULL as well.
Only mlx4 uses an input request structure, have every other driver call
ib_is_udata_in_empty() to enforce the lack of request structs.
Use ib_respond_empty_udata() in every driver that does not use a response
struct.
Ensure a check for NULL udata before calling ib_respond_udata() in
bnxt_re, efa, and mlx5.
Make mlx4 safe to be called with NULL.
Link: https://patch.msgid.link/r/2-v1-922fa8e828ba+f7-ib_udata_stack_jgg@nvidia.com
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
For non-SRQ QPs, the responder reads WQE fields directly from the
shared queue buffer mapped into userspace. This allows a malicious
user to modify fields like num_sge or sge entries while the kernel
is processing the WQE, leading to out-of-bounds reads in
rxe_resp_check_length() and copy_data().
Introduce get_recv_wqe() that validates num_sge and copies the WQE
to a kernel-local buffer before processing, matching the approach
already used for SRQ WQEs in get_srq_wqe(). The srq_wqe buffer is
reused since SRQ and non-SRQ paths are mutually exclusive per QP.
Fixes: 8700e3e7c4 ("Soft RoCE driver")
Link: https://patch.msgid.link/r/20260518215040.1598586-3-tristan@talencesecurity.com
Signed-off-by: Tristan Madani <tristan@talencesecurity.com>
Reviewed-by: Zhu Yanjun <yanjun.zhu@linux.dev>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
get_srq_wqe() reads wqe->dma.num_sge from the shared receive queue
buffer, which is mapped into userspace. It validates num_sge against
max_sge, but then re-reads the same field to calculate the memcpy
size. A concurrent userspace thread can modify num_sge between
validation and use, causing a heap buffer overflow when copying the
WQE into qp->resp.srq_wqe.
Read num_sge into a local variable and use it for both the bounds
check and the size calculation.
Fixes: 8700e3e7c4 ("Soft RoCE driver")
Link: https://patch.msgid.link/r/20260518215040.1598586-2-tristan@talencesecurity.com
Signed-off-by: Tristan Madani <tristan@talencesecurity.com>
Reviewed-by: Zhu Yanjun <yanjun.zhu@linux.dev>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
The new umem getter family being introduced in follow-up patches
need a fitting name for the central all-source helper that resolves
attributes, legacy fillers and a UHW VA fallback.
Rename the existing VA-pinning helper ib_umem_get() to ib_umem_get_va()
so the name is freed up. The new name is consistent with names of rest
of the helpers that are about to be introduced.
Link: https://patch.msgid.link/r/20260529134312.2836341-2-jiri@resnulli.us
Signed-off-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
-----BEGIN PGP SIGNATURE-----
iQFSBAABCgA8FiEEq68RxlopcLEwq+PEeb4+QwBBGIYFAmoTZAYeHHRvcnZhbGRz
QGxpbnV4LWZvdW5kYXRpb24ub3JnAAoJEHm+PkMAQRiGKZ0H/A8Ti79DXqEl4M6V
S1ZYfteuvyWCdygPfY7+CQsxS1EjfmeRXd0213n95jijUoXvAVe7HJ5FuroTltsF
LYqH3CAedeEWMO1rIZjuDs5J1BzWAh7b0/QoanTlAjI1ZL8c3Btj/o//PoQfztiS
HWlN/aCcsZTi9n0ewV9eQM9jdiuhBlpZujfDX90vYurxAB+3K2c1b0UN+Y4/zMXO
83zaDLMYhIh8cEaU52+23H9UpyfrT0B8n6YgqbU4ZBRIbgA1BC4gloT6B7OHFKfq
PAfjAOBfpe43K8lLUyfbkfNohNblLOboiMmekdf+j2y4wqpCO64o6WigCrZtxtHe
9znkJpM=
=H+I9
-----END PGP SIGNATURE-----
Merge tag 'v7.1-rc5' into rdma.git for-next
For dependencies in the following patches
Resolve conflicts, use the goto labels from the rc tag.
* tag 'v7.1-rc5': (1526 commits)
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Convert the pattern:
ib_copy_to_udata(udata, &resp, sizeof(resp));
Using Coccinelle:
@@
identifier resp;
expression udata;
@@
- ib_copy_to_udata(udata, &resp, sizeof(resp))
+ ib_respond_udata(udata, resp)
Run another pass with AI to propagate the return code correctly and
remove redundant prints.
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Use copy_to_user() directly since the data is not being placed in the
udata response memory.
It is unclear why this is trying to do two copies, but leave it alone.
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Simplify umem allocation by using flexible array member.
Add __counted_by to get extra runtime analysis.
Suggested-by: Rosen Penev <rosenp@gmail.com>
Signed-off-by: Bernard Metzler <bernard.metzler@linux.dev>
Link: https://patch.msgid.link/20260511141149.52362-1-bernard.metzler@linux.dev
Signed-off-by: Leon Romanovsky <leon@kernel.org>
In RXE, hardware counters are already supported, but not in a
standardized manner. For instance, user-space monitoring tools like
atop only read from the *counters* directory. Therefore, it is
necessary to add perf management support to RXE.
Also use rxe_counter_get instead of raw atomic64_read in hw-counters.
Signed-off-by: zhenwei pi <zhenwei.pi@linux.dev>
Link: https://patch.msgid.link/20260414062948.671658-4-zhenwei.pi@linux.dev
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Suggested by Leon, remove the rxe_ib_device_get_netdev() wrapper and
the RXE_PORT definition. These additions do not improve readability,
and RXE has always had only a single port.
Signed-off-by: zhenwei pi <zhenwei.pi@linux.dev>
Link: https://patch.msgid.link/20260414062948.671658-2-zhenwei.pi@linux.dev
Signed-off-by: Leon Romanovsky <leon@kernel.org>
A malicious connected siw peer can send an iWARP FPDU whose MPA length
field (c_hdr->mpa_len, 16 bit big-endian, peer-controlled) is smaller
than the fixed DDP/RDMAP header for the announced opcode. Soft-iWARP
parses the full header in siw_get_hdr() based on iwarp_pktinfo[opcode]
.hdr_len, but never compares mpa_len against that header length.
siw_tcp_rx_data() then derives
srx->fpdu_part_rem = be16_to_cpu(mpa_len) - fpdu_part_rcvd
+ MPA_HDR_SIZE;
where fpdu_part_rcvd equals iwarp_pktinfo[opcode].hdr_len at this
point. For a tagged WRITE (hdr_len 16, MPA_HDR_SIZE 2) the smallest
on-wire mpa_len of 0 yields fpdu_part_rem = -14, and any mpa_len below
hdr_len - MPA_HDR_SIZE underflows to a negative int.
The signed value then flows into siw_proc_write()/siw_proc_rresp() as
bytes = min(srx->fpdu_part_rem, srx->skb_new);
is handed to siw_check_mem() as an int len (whose interval check
addr + len > mem->va + mem->len is satisfied for a valid base when
len is negative), and reaches siw_rx_data() -> siw_rx_kva() /
siw_rx_umem() -> skb_copy_bits() as a signed copy length. The header
copy branch in skb_copy_bits() promotes that to size_t, producing a
multi-gigabyte read.
KASAN under a KUnit harness that drives the real kernel TCP receive
path -- a loopback AF_INET socketpair, the malformed FPDU written via
kernel_sendmsg, sk_data_ready firing in softirq, tcp_read_sock
dispatching to siw_tcp_rx_data -- reports:
BUG: KASAN: use-after-free in skb_copy_bits+0x284/0x480
Read of size 4294967295 at addr ffff888...
Call Trace:
skb_copy_bits
siw_rx_kva
siw_rx_data
siw_check_mem
siw_proc_write
siw_tcp_rx_data
__tcp_read_sock
siw_qp_llp_data_ready
tcp_data_ready
tcp_data_queue
Add the missing invariant at the earliest point where the peer header
is fully assembled. iwarp_pktinfo[*].hdr_len - MPA_HDR_SIZE is exactly
the value the siw transmitter uses as the minimum mpa_len for each
opcode (drivers/infiniband/sw/siw/siw_qp.c:33), so this matches the
protocol contract. Out-of-range FPDUs terminate the connection with
TERM_ERROR_LAYER_LLP / LLP_ETYPE_MPA / LLP_ECODE_FPDU_START -- which
is RFC 5044 Section 8 error code 3 ("Marker and ULPDU Length fields
do not agree on the start of an FPDU"), the correct framing-error
class for this inconsistency.
Fixes: 8b6a361b8c ("rdma/siw: receive path")
Link: https://patch.msgid.link/r/20260513175325.2042630-2-michael.bommarito@gmail.com
Cc: stable@vger.kernel.org
Signed-off-by: Michael Bommarito <michael.bommarito@gmail.com>
Assisted-by: Claude:claude-opus-4-7
Acked-by: Bernard Metzler <bernard.metzler@linux.dev>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
None of the udp_tunnel users need struct socket in their
fast paths; it is only used for tunnel setup / teardown.
Even setup_udp_tunnel_sock() does not need struct socket.
Let's change setup_udp_tunnel_sock() to take struct sock
instead of struct socket.
Signed-off-by: Kuniyuki Iwashima <kuniyu@google.com>
Link: https://patch.msgid.link/20260502031401.3557229-3-kuniyu@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
None of the udp_tunnel users need struct socket in their
fast paths; it is only used for tunnel setup / teardown.
While the UDP tunnel interface accepts struct socket, this
encourages users to store the pointer unnecessarily. This
leads to extra dereferences when accessing struct sock fields
(e.g., sk->sk_user_data instead of sock->sk->sk_user_data).
Furthermore, these dereferences necessitate synchronize_rcu()
in udp_tunnel_sock_release() to protect the fast paths from
sock_orphan() setting sk->sk_socket to NULL.
This overhead can be avoided if users store the struct sock
pointer directly in their private structures.
As a prep, let's change udp_tunnel_sock_release() to take
struct sock instead of struct socket.
Signed-off-by: Kuniyuki Iwashima <kuniyu@google.com>
Link: https://patch.msgid.link/20260502031401.3557229-2-kuniyu@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
atomic_write_reply() at drivers/infiniband/sw/rxe/rxe_resp.c
unconditionally dereferences 8 bytes at payload_addr(pkt):
value = *(u64 *)payload_addr(pkt);
check_rkey() previously accepted an ATOMIC_WRITE request with pktlen ==
resid == 0 because the length validation only compared pktlen against
resid. A remote initiator that sets the RETH length to 0 therefore reaches
atomic_write_reply() with a zero-byte logical payload, and the responder
reads sizeof(u64) bytes from past the logical end of the packet into
skb->head tailroom, then writes those 8 bytes into the attacker's MR via
rxe_mr_do_atomic_write(). That is a remote disclosure of 4 bytes of kernel
tailroom per probe (the other 4 bytes are the packet's own trailing ICRC).
IBA oA19-28 defines ATOMIC_WRITE as exactly 8 bytes. Anything else is
protocol-invalid. Hoist a strict length check into check_rkey() so the
responder never reaches the unchecked dereference, and keep the existing
WRITE-family length logic for the normal RDMA WRITE path.
Reproduced on mainline with an unmodified rxe driver: a sustained
zero-length ATOMIC_WRITE probe repeatedly leaks adjacent skb head-buffer
bytes into the attacker's MR, including recognisable kernel strings and
partial kernel-direct-map pointer words. With this patch applied the
responder rejects the PDU and the MR stays all-zero.
Cc: stable@vger.kernel.org
Fixes: 034e285f8b ("RDMA/rxe: Make responder support atomic write on RC service")
Link: https://patch.msgid.link/r/20260418162141.3610201-1-michael.bommarito@gmail.com
Assisted-by: Claude:claude-opus-4-7
Signed-off-by: Michael Bommarito <michael.bommarito@gmail.com>
Reviewed-by: Zhu Yanjun <yanjun.zhu@linux.dev>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
