Pull BPF fixes from Daniel Borkmann:
- Fix BPF verifier to preserve full pointer state for commuted
scalar += pointer arithmetic (Yiyang Chen, Eduard Zingerman)
- Fix a use-after-free of request sockets in the BPF TCP iterator
batching (Jose Fernandez)
- Fix a use-after-free of sk_redir in the BPF sockmap send verdict
path (Chengfeng Ye)
- Fix a netns reference imbalance in the BPF conntrack kfuncs
(Chengfeng Ye)
- Fix bpf_get_fsverity_digest() dynptr assumptions and silent
digest truncation (Eric Biggers)
- Fix bpf_tcp_{gen,check}_syncookie to check sk_state before
sk_protocol to make sure it is a full socket (Luxiao Xu)
- Fix rqspinlock to reset the tail when preserving the queue
on deadlock (Kumar Kartikeya Dwivedi)
* tag 'bpf-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf:
rqspinlock: Reset tail when preserving queue on deadlock
bpf: Check sk_state before sk_protocol in bpf_tcp_*_syncookie
fsverity: Fix silent truncation in bpf_get_fsverity_digest()
fsverity: Fix bpf_get_fsverity_digest() dynptr assumptions
bpf: tcp: Fix use-after-free in bpf_iter_tcp_established_batch()
bpf: Fix netns reference imbalance in conntrack kfuncs
bpf, sockmap: Fix sk_redir use-after-free in send verdict
selftests/bpf: Cover commuted pointer state propagation
bpf: Propagate untrusted pointer state in commuted arithmetic
bpf: Preserve pointer state for commuted arithmetic
bpf: Simplify sanitize_err() signature
When a signal interrupts a blocking send, tls_tx_records() treats the
resulting -ERESTARTSYS as a transmission failure and marks the socket
errored via tls_err_abort() with the raw error code. Later syscalls
return the kernel-internal errno 512 (ERESTARTSYS) to userspace, as the
signal it stems from is no longer pending during syscall exit and thus
never translated.
An interrupted send is not a connection error: the partially sent record
stays queued and is resent later. Interrupt error codes are therefore
excluded from the abort in the same way as -EAGAIN.
Fixes: b341ca51d2 ("tls: Fix tls_sw_sendmsg error handling")
Signed-off-by: Maximilian Immanuel Brandtner <maxbr@linux.ibm.com>
Link: https://patch.msgid.link/20260805063109.1772314-1-maxbr@linux.ibm.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Technically, it's illegal to take a ref on a netdev just because
we have a pointer on which we already hold a ref, with no other
protection. This is because our simple per-cpu refcount
implementation cannot atomically read the count.
Let's make sure we cancel outstanding work and never queue more
work for a device we know is dead. This way taking a ref on
a dev we know is on the netdev_work_list is always going to be safe.
Jiangshan Yi reports that the issues is caught by ref tracker infra
leading to a warning:
WARNING: lib/ref_tracker.c:322 at ref_tracker_free
WARNING: lib/ref_tracker.c:246 at ref_tracker_dir_exit
Reported-by: Jiangshan Yi <yijiangshan@kylinos.cn>
Link: https://lore.kernel.org/20260731035135.3917308-2-yijiangshan@kylinos.cn
Fixes: 12c765be84 ("net: turn the rx_mode work into a generic netdev_work facility")
Link: https://patch.msgid.link/20260806022821.2079945-1-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
netdev_register_kobject() calls device_add(), which emits KOBJ_ADD and
wakes udev, but register_netdevice() only makes the device findable by
name later, in list_netdevice(). A udev worker that reacts to the uevent
can therefore run against a device that no lookup can find yet.
This used to be harmless because the ethtool ioctl took the rtnl_lock
when looking the device up, and register_netdevice() runs under rtnl, so
the worker simply blocked until registration finished. The commit in the
fixes tag moved the lookup out from under rtnl for ops-locked drivers.
Now there is a short window in register_netdevice() between
netdev_register_kobject() until list_netdevice() when the device is not
findable by name.
This was reproduced with the mlx5 driver on a kernel with KASAN enabled
during devlink reload: systemd-udevd's net_driver builtin gets -ENODEV
from ETHTOOL_GDRVINFO, which was preventing interface renaming.
Suppress the uevent in netdev_register_kobject() and emit it from
register_netdevice() next to rtmsg_ifinfo(). This is the last point in
register_netdevice() where no error can happen, so only fully registered
devices are announced: the registration error paths never reach it, and
the device_del() that unwinds them stays silent as well, leaving
userspace with neither an add nor a remove.
Fixes: f994752b11 ("net: ethtool: optionally skip rtnl_lock on IOCTL path")
Signed-off-by: Dragos Tatulea <dtatulea@nvidia.com>
Reviewed-by: Shahar Shitrit <shshitrit@nvidia.com>
Link: https://patch.msgid.link/20260806080758.2039586-2-dtatulea@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
sctp_make_heartbeat_ack() caches the destination transport in
chunk->transport without taking a reference. When src_out_of_asoc_ok is
enabled, the HEARTBEAT ACK may remain queued on control_chunk_list instead
of being transmitted immediately.
If the peer transport is removed while the chunk is still queued,
sctp_assoc_rm_peer() drops the transport and schedules it for RCU freeing,
but only clears cached transport pointers in out_chunk_list. The queued
control chunk therefore retains a dangling transport pointer.
Once an ASCONF_ACK clears the suppression and the queued control chunk is
transmitted, SCTP dereferences the stale transport pointer, leading to a
use-after-free.
Fix this by also clearing chunk->transport for queued control chunks in
control_chunk_list when removing the transport.
Fixes: 8a07eb0a50 ("sctp: Add ASCONF operation on the single-homed host")
Reported-by: Daniele Linguaglossa <danielelinguaglossa@gmail.com>
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Link: https://patch.msgid.link/7e1168cb722132152a29d47e5eafaeac4a3bf6f3.1785943120.git.lucien.xin@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
vcc_setsockopt() contained an ineffective optlen check:
if (__SO_LEVEL_MATCH(optname, level) && optlen != __SO_SIZE(optname))
return -EINVAL;
If __SO_LEVEL_MATCH(optname, level) evaluated to false (e.g. if the caller
passed a mismatched level), the length check optlen != __SO_SIZE(optname)
was short-circuited and bypassed. Execution then fell through to switch(optname),
calling copy_from_sockptr() assuming optval contained sufficient space.
Furthermore, even if level matched, a cgroup BPF setsockopt filter could shrink
optlen after entry. Because copy_from_sockptr() on kernel pointers uses memcpy(),
this leads to a KASAN slab-out-of-bounds read when optlen is smaller than the
expected structure size.
Fix this by using copy_safe_from_sockptr(), which unconditionally validates
that optlen is at least the expected size before copying. Also change the local
'value' variable type from 'unsigned long' to 'int' so that SO_SETCLP matches
its sizeof(int) ABI encoding on 64-bit systems.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Reported-by: syzbot+53ecc09fb81df10ef4de@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=53ecc09fb81df10ef4de
Signed-off-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20260805131508.3227331-1-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
tpacket_snd() reads dev->hard_header_len independently for skb
allocation and header construction in tpacket_fill_skb(). Concurrent
netdevice reconfiguration can therefore make the reserved headroom
smaller than the amount later pushed, or make copylen - hard_header_len
negative.
Snapshot hard_header_len once before processing ring frames and use it
for the frame limit, headroom allocation, copy length, and skb
construction. Pass the snapshot to tpacket_fill_skb().
The separate SOCK_DGRAM consistency problem between hard_header_len and
header_ops->create is not addressed here.
Fixes: 69e3c75f4d ("net: TX_RING and packet mmap")
Cc: stable@vger.kernel.org
Signed-off-by: Qihang Tang <q.h.hack.winter@gmail.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Link: https://patch.msgid.link/20260805125729.19220-4-q.h.hack.winter@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
packet_snd() reads dev->hard_header_len multiple times while allocating
and constructing an skb. Device reconfiguration can change this value
concurrently, for example through bonding device type changes.
For SOCK_RAW, packet_snd() can save a larger value in reserve and later
allocate headroom using a smaller value. Moving skb->data back by reserve
then places it before skb->head, and the following copy from userspace can
attempt an out-of-bounds write.
packet_sendmsg_spkt() has the same issue because it calculates its
reservation and header offset from separate reads before dropping the RCU
read lock to allocate the skb.
Add LL_RESERVED_SPACE_EX() for callers that already saved a header length.
Read hard_header_len once in packet_snd() and use it for allocation and
construction. In packet_sendmsg_spkt(), preserve the allocation-time value
through the device lookup retry.
The separate SOCK_DGRAM consistency problem between hard_header_len and
header_ops->create is not addressed here.
Fixes: b84bbaf7a6 ("packet: in packet_snd start writing at link layer allocation")
Cc: stable@vger.kernel.org
Signed-off-by: Qihang Tang <q.h.hack.winter@gmail.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Link: https://patch.msgid.link/20260805125729.19220-3-q.h.hack.winter@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
tls_decrypt_sg() advances msg->msg_iter when it maps user pages for
the optimistic TLS 1.3 zero-copy path. If the decrypted record turns
out not to be unpadded application data, tls_decrypt_sw() retries into
a kernel skb, but leaves the iterator advanced.
The subsequent copy from the skb then writes decrypted bytes again at
a later point in the caller iovecs while recvmsg() reports only the
post-retry length. A TLS peer can trigger this after the receiver
enables TLS_RX_EXPECT_NO_PAD.
Revert the iterator by the number of bytes consumed by the optimistic
mapping before retrying without zero-copy.
Add a selftest which sends a TLS 1.3 control record with
TLS_RX_EXPECT_NO_PAD enabled and verifies that recvmsg() does not
overwrite later iovecs beyond the returned length.
Fixes: ce61327ce9 ("tls: rx: support optimistic decrypt to user buffer with TLS 1.3")
Cc: stable@vger.kernel.org
Signed-off-by: Jérémy Jean <Jeremy.Jean@oss.cyber.gouv.fr>
Link: https://patch.msgid.link/20260804125528.2139928-1-Jeremy.Jean@oss.cyber.gouv.fr
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
When the copy path in tls_sw_sendmsg_locked() adds the fragment that fills
the plaintext sk_msg ring, it does not set full_record, so the record is
left full and unpushed. A later splice() then adds to an already full
ring: sk_msg_page_add() has no fullness check of its own, so sg.end wraps
onto sg.start and the ring appears empty. Fragments added after that
overwrite live entries, and sg.size no longer matches what is reachable
between sg.start and sg.end, so pushing the record runs the scatterwalk off
the end of the scatterlist.
An unprivileged user can trigger this on a loopback TCP socket with the
"tls" ULP attached:
BUG: kernel NULL pointer dereference, address: 0000000000000008
RIP: 0010:memcpy_from_scatterwalk+0x32/0xc0
Call Trace:
skcipher_walk_next+0x1d1/0x2c0
gcm_encrypt_aesni_avx+0x1e9/0x220
bpf_exec_tx_verdict+0x3bb/0x860
tls_sw_sendmsg+0xa1a/0xca0
__sys_sendto+0x1da/0x1f0
Set full_record in the copy path when the ring becomes full, and push a
record that is already full on entry to the sendmsg loop.
Suggested-by: Sabrina Dubroca <sd@queasysnail.net>
Fixes: fe1e81d4f7 ("tls/sw: Support MSG_SPLICE_PAGES")
Cc: stable@vger.kernel.org
Signed-off-by: chanyoung <ppoo1220@gmail.com>
Link: https://patch.msgid.link/20260804052837.49015-2-ppoo1220@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
xdpf_clone() clones broadcast copies into a single page and sets
frame_sz to PAGE_SIZE. __xdp_build_skb_from_frame() later treats that
page like a normal XDP frame and expects the usual skb_shared_info
tailroom at the end of the buffer.
The current check only rejects frames whose linear xdp_frame header,
headroom, and packet data exceed PAGE_SIZE. A source frame backed by a
larger allocation can still satisfy that check while extending into the
clone's required shared-info area. When such a clone is converted back
into an skb, build_skb_around() places skb_shared_info over live packet
bytes and later writes can corrupt XDP return metadata.
Reject clones unless their linear area fits inside
SKB_WITH_OVERHEAD(PAGE_SIZE), matching the tailroom requirement already
enforced by the XDP-to-skb conversion path.
Fixes: e624d4ed4a ("xdp: Extend xdp_redirect_map with broadcast support")
Cc: stable@vger.kernel.org
Reported-by: Vega <vega@nebusec.ai>
Signed-off-by: Zhiling Zou <zhilinz@nebusec.ai>
Link: https://patch.msgid.link/6b2afef5d1738763c6965e8e466eb16e43e4f956.1785757386.git.zhilinz@nebusec.ai
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Passive TCP Fast Open accepts a valid-cookie SYN even when it carries
no data. In that case the child socket's receive queue is intentionally
left empty.
mptcp_fastopen_subflow_synack_set_params() set is_mptfo before checking
for queued SYN data. That made data-less TFO SYNs hit a WARN and, if
the warning was non-fatal, left stale MPTFO state behind. The stale
flag could later trigger a state-confusion bug in
check_fully_established().
Only mark the subflow as MPTFO after confirming that an SKB was queued.
Return quietly when the receive queue is empty.
Note that mptcp_subflow_context's is_mptfo field is now not just about
subflows where the TFO was present, but about MPTFO subflow that
consumed SYN data. Only having a valid cookie but not carrying data is
not really "doing TFO".
Fixes: 36b122baf6 ("mptcp: add subflow_v(4,6)_send_synack()")
Cc: stable@vger.kernel.org
Reported-by: Yuan Tan <yuantan098@gmail.com>
Reported-by: Yifan Wu <yifanwucs@gmail.com>
Reported-by: Juefei Pu <tomapufckgml@gmail.com>
Reported-by: Zhengchuan Liang <zcliangcn@gmail.com>
Reported-by: Xin Liu <bird@lzu.edu.cn>
Signed-off-by: Wyatt Feng <bronzed_45_vested@icloud.com>
Signed-off-by: Ren Wei <n05ec@lzu.edu.cn>
Reviewed-by: Matthieu Baerts (NGI0) <matttbe@kernel.org>
Signed-off-by: Matthieu Baerts (NGI0) <matttbe@kernel.org>
Link: https://patch.msgid.link/20260803-net-mptcp-misc-fixes-7-2-rc6-v2-7-b8f496d71664@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
mptcp_pm_destroy() empties msk->pm.anno_list and
msk->pm.userspace_pm_local_addr_list under msk->pm.lock during socket
teardown, dropping the lock between the two.
A concurrent userspace PM genl ANNOUNCE on the same msk holds a sock
reference via mptcp_token_get_sock() and, in
mptcp_pm_nl_announce_doit(), calls
mptcp_userspace_pm_append_new_local_addr() and
mptcp_pm_announced_alloc(). Both take msk->pm.lock briefly to add to
their respective lists. Because the genl handler holds a sock reference,
mptcp_pm_destroy() may run on the same msk via mptcp_disconnect(), which
invokes mptcp_destroy_common() without dropping the sock refcount,
before the handler completes.
If the lock acquisitions interleave such that mptcp_pm_destroy() empties
a list first, the later alloc adds its entry to a list head that nothing
else iterates for this msk, and the entry leaks. kmemleak reports both
mptcp_pm_add_addr objects (from mptcp_pm_announced_alloc()) and
mptcp_pm_addr_entry objects (from
mptcp_userspace_pm_append_new_local_addr()) under sustained concurrent
ANNOUNCE + close load against the userspace PM.
Add an MPTCP_PM_DESTROYING bit in msk->pm.status, set by
mptcp_pm_destroy() under pm.lock before the lists are emptied and
checked under pm.lock by the alloc paths. Either the alloc takes pm.lock
first, in which case its entry is on the list when mptcp_pm_destroy()
frees it; or mptcp_pm_destroy() takes pm.lock first, in which case the
later alloc observes the bit and refuses.
Found by an MPTCP protocol-flow harness extending BRF (arXiv:2305.08782).
Fixes: 9ab4807c84 ("mptcp: netlink: Add MPTCP_PM_CMD_ANNOUNCE")
Cc: stable@vger.kernel.org
Signed-off-by: Shardul Bankar <shardul.b@mpiricsoftware.com>
Reviewed-by: Matthieu Baerts (NGI0) <matttbe@kernel.org>
Signed-off-by: Matthieu Baerts (NGI0) <matttbe@kernel.org>
Link: https://patch.msgid.link/20260803-net-mptcp-misc-fixes-7-2-rc6-v2-6-b8f496d71664@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The timer callback reads entry->retrans_times outside pm.lock to decide
whether to call mptcp_pm_subflow_established(). Since
mptcp_pm_announced_del_timer() can concurrently set retrans_times =
ADD_ADDR_RETRANS_MAX under pm.lock, a race condition exists.
I discovered this issue while studying the code. AI tools helped me to
verify the issue can potentially happen under race conditions.
Use a local 'retransmit' flag set inside pm.lock to capture whether
retransmission is still possible when the lock is taken. This allows to
call mptcp_pm_subflow_established() accordingly, and not depending on
the situation that can be different when checked outside the pm.lock.
Fixes: 348d5c1dec ("mptcp: move to next addr when timeout")
Cc: stable@vger.kernel.org
Signed-off-by: Qing Luo <luoqing@kylinos.cn>
Reviewed-by: Matthieu Baerts (NGI0) <matttbe@kernel.org>
Signed-off-by: Matthieu Baerts (NGI0) <matttbe@kernel.org>
Link: https://patch.msgid.link/20260803-net-mptcp-misc-fixes-7-2-rc6-v2-4-b8f496d71664@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Some MPTCP suboptions are mutually exclusive according to the RFC8684,
but also because in different places, the code doesn't expect some
combinations to be present. That's specially true for suboptions that
would be present twice, but with different attributes.
The new restrictions are the same as the ones applied on the output
side, with mptcp_write_options. The same rules can be reused with a
small fix: an MP_FASTCLOSE can be used with a DSS when the sender picks
this option [1], which is not the case on Linux. Here are the rules:
Which options can be used together?
X: mutually exclusive
O: often used together
C: can be used together in some cases
P: could be used together but we prefer not to (optimisations)
| Opt: | MPC | MPJ | DSS | ADD | RM | PRIO | FAIL | FC |
|------|------|------|------|------|------|------|------|------|
| MPC |------|------|------|------|------|------|------|------|
| MPJ | X |------|------|------|------|------|------|------|
| DSS | X | X |------|------|------|------|------|------|
| ADD | X | X | P |------|------|------|------|------|
| RM | C | C | C | P |------|------|------|------|
| PRIO | X | C | C | C | C |------|------|------|
| FAIL | X | X | C | X | X | X |------|------|
| FC | X | X | P | X | X | X | X |------|
| RST | X | X | X | X | X | X | O | O |
|------|------|------|------|------|------|------|------|------|
The only difference is with the 'P': another stack could send and
ADD_ADDR with other suboptions (DSS, RM_ADDR), and this should be
allowed.
A few points of attention:
- In theory, an MP_CAPABLE could be used with a RM_ADDR, but there is
no reason to add it with a SYN. Note that even with a 4th ACK, it
doesn't seem to be useful, except when IDs are known in advance via
another channel. Better not to break that.
- Now, combining both an MP_CAPABLE and an MP_JOIN will no longer
result to a reject of the two options, but only the second suboption
is ignored. That seems OK to do that for this unexpected error. At
least now all inconsistent combinations are handled the same way.
This could change later in next. This also means the explicit checks
for having both MPC + MPJ in subflow.c will now be unreachable.
That's fine, they will be removed in a follow-up patch.
- In case of conflicting combinations, the extra suboption(s) is/are
ignored: having such combinations either means the remote peer is
buggy, or is evil. The simplest action is then taken in this case:
stop processing the current suboption.
- In mp_opt->suboptions, there is also a bit reserved to the checksum,
which can be used in an MP_CAPABLE and a DSS. Each time a DSS option
can be used in parallel with another option, the checksum can be set,
so the verification is combined into a new OPTIONS_MPTCP_DSS macro.
- An MP_CAPABLE ACK can carry a Data-Level Length, and an optional
Checksum: they are the same as the ones found in a DSS, because a DSS
cannot be used in parallel to an MP_CAPABLE. Similarly, even if there
is room, a DSS cannot be used with an MP_JOIN.
Fixes: eda7acddf8 ("mptcp: Handle MPTCP TCP options")
Cc: stable@vger.kernel.org
Link: https://www.rfc-editor.org/rfc/rfc8684.html#section-3.5-5.1 [1]
Signed-off-by: Matthieu Baerts (NGI0) <matttbe@kernel.org>
Link: https://patch.msgid.link/20260803-net-mptcp-misc-fixes-7-2-rc6-v2-2-b8f496d71664@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Current limit of 64 is not enough for Sony Xperia 10 VII (SM6475).
After merging v6.6.142 into a downstream AOSP device, it's stuck on
boot animation and following log spam can be observed in dmesg:
E qrtr : ctrl_cmd_new_lookup(): QRTR client node exceeds max lookup limit!
E qrtr : qrtr_ns_worker(): failed while handling packet from 1:16600
No idea why it needs more than 64 client lookups, but it appears to
work fine with 128 as it did when there were no limits.
I don't really have a good way to investigate what it needs all
these lookups for as most of the userspace is closed source.
Fixes: 5640227d9a ("net: qrtr: ns: Limit the maximum number of lookups")
Signed-off-by: Łukasz Patron <priv.luk@gmail.com>
Link: https://patch.msgid.link/20260804201919.1148015-1-priv.luk@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
smc_listen_out() reads lsmc->sk.sk_state without the listener lock,
then acquires lock_sock_nested() only after the check passes. This
opens a window where smc_close_active() can transition the listener
to SMC_CLOSED, call smc_close_cleanup_listen() to drain the accept
queue, and release the lock, all between the lockless read and the
delayed lock acquisition:
smc_listen_work (smc_hs_wq) smc_close_active()
------------------------------- -------------------------
release_sock(child)
if (sk_state == SMC_LISTEN) TRUE
lock_sock(listener)
sk_state = SMC_CLOSED
smc_close_cleanup_listen()
release_sock(listener)
flush_work(tcp_listen_work)
lock_sock_nested(listener)
smc_accept_enqueue(listener, child) /* child enqueued on dead listener */
smc_close_active() flushes only tcp_listen_work. Work items already
dispatched onto smc_hs_wq for the CLC handshake continue running
unguarded. smc_accept_enqueue() takes a sock_hold() on the child that
is never released, so the child smc_sock, its clcsock, and the
reference all leak. A remote peer that opens TCP connections while the
server calls close() can exhaust kernel memory.
Move lock_sock_nested() to before the sk_state check so that the test
and the enqueue are atomic under the listener lock.
Fixes: fd57770dd1 ("net/smc: wait for pending work before clcsock release_sock")
Reviewed-by: Mahanta Jambigi <mjambigi@linux.ibm.com>
Signed-off-by: Sidraya Jayagond <sidraya@linux.ibm.com>
Reviewed-by: Breno Leitao <leitao@debian.org>
Reviewed-by: Dust Li <dust.li@linux.alibaba.com>
Link: https://patch.msgid.link/20260803070701.126339-1-sidraya@linux.ibm.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
ip6ip6_err() clones an outer IPv6 ICMP error skb, pulls it to the
quoted inner IPv6 packet, and then passes the clone to icmpv6_send().
The clone still carries the outer packet's inet6_skb_parm in skb->cb.
If the outer packet had a Home Address Option, IP6CB(skb2)->dsthao
remains non-zero after skb_pull(). icmpv6_send() later calls
mip6_addr_swap(), which uses that stale dsthao offset against the quoted
inner packet. A malformed inner destination-options header can then make
the HAO lookup and address swap run past the end of the quoted packet
and corrupt skb_shared_info.
Clear skb2->cb[] before pulling the quoted inner IPv6 packet so the
reply path does not reuse metadata left by the outer IPv6 stack.
Fixes: e490d1d85c ("[IPV6] IP6TUNNEL: Split out generic routine in ip6ip6_err().")
Cc: stable@vger.kernel.org
Reported-by: Vega <vega@nebusec.ai>
Signed-off-by: Zhiling Zou <zhilinz@nebusec.ai>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Link: https://patch.msgid.link/fe1a5e765fbca88d69391887f0ed26a19e3e4d39.1785736562.git.zhilinz@nebusec.ai
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
ncsi_send_cmd_nl() takes the number of bytes to copy from the
attacker-controlled ncsi_pkt_hdr.length field of the in-band packet
header, while the source buffer is the NCSI_ATTR_DATA netlink
attribute whose readable size is nla_len() - sizeof(ncsi_pkt_hdr).
The two length sources are never cross-checked: only
nla_len() >= sizeof(struct ncsi_pkt_hdr) is enforced.
With hdr->length set larger than the attribute payload (up to 65535
against at most 2032 readable bytes), ncsi_cmd_handler_oem() copies
past the end of the netlink attribute buffer with unsafe_memcpy(),
leaking up to ~64KB of kernel heap memory into the transmitted NCSI
command packet. The destination skb is sized by the declared payload,
so the write side does not overflow - this is a pure OOB read /
information leak, reachable with CAP_NET_ADMIN on systems with a
registered NCSI device (e.g. OpenBMC on Aspeed BMC SoCs, where
NET_NCSI=y is standard).
Reject commands whose declared payload extends past the end of the
data attribute.
The issue was found by the autokbug dynamic kernel fuzzer at Tencent
Yunding Lab.
Fixes: 9771b8ccdf ("net/ncsi: Extend NC-SI Netlink interface to allow user space to send NC-SI command")
Reported-by: Henry Martin <bsdhenrymartin@gmail.com>
Signed-off-by: Henry Martin <bsdhenrymartin@gmail.com>
Link: https://patch.msgid.link/20260803043618.3210301-1-bsdhenrymartin@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
mac802154_beacon_worker() reads local->beacon_req under RCU and derives
the sub-interface from the request, but then drops the RCU read lock and
continues to use both sdata and the embedded wpan_dev.
mac802154_stop_beacons_locked() cancels only pending beacon work, clears
local->beacon_req and frees the request. A beacon worker that is already
running can therefore continue after interface teardown and dereference
the freed netdev private area.
The scan worker already pins the netdev before leaving RCU. Apply the
same lifetime rule to the beacon worker: take a netdev reference while
the request is still protected by RCU, and release it on all paths that
continue after the reference is acquired.
Fixes: 3accf47627 ("mac802154: Handle basic beaconing")
Cc: stable@vger.kernel.org
Reported-by: Vega <vega@nebusec.ai>
Signed-off-by: Zihan Xi <zihanx@nebusec.ai>
Reviewed-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://patch.msgid.link/e9a3909c7a6281967961773ca841e860b8ecf40e.1785596603.git.zihanx@nebusec.ai
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Incoming skbs passing through netfilter flowtable offload hooks (or XFRM
offload path) might already carry a ref-counted dst_entry assigned during
earlier RX or routing steps.
Calling skb_dst_set_noref() when skb already holds a ref-counted dst
overwrites skb->_skb_refdst, leaking the previous dst_entry reference
count and triggering a DEBUG_NET_WARN_ON_ONCE assertion in
skb_dst_check_unset():
WARNING: at skb_dst_check_unset include/linux/skbuff.h:1170
WARNING: at skb_dst_set_noref include/linux/skbuff.h:1234
WARNING: at nf_flow_offload_ip_hook+0xf6c/0x2b60 net/netfilter/nf_flow_table_ip.c:864
Drop any existing dst_entry reference with skb_dst_drop(skb) before
setting the non-referenced flowtable destination.
Fixes: 2a79fd3908 ("netfilter: nf_flow_table: attach dst to skbs")
Reported-by: syzbot+76d4e3a055aec3b007ec@syzkaller.appspotmail.com
Closes: https://lore.kernel.org/netdev/6a71b141.9511d2ce.1fc5b9.033b.GAE@google.com/T/#u
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Pablo Neira Ayuso <pablo@netfilter.org>
Link: https://patch.msgid.link/20260804093328.1831847-1-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
A listener's TCP_FASTOPEN max_qlen stops being accurate and lets through
far more pending Fast Open requests than it was configured for.
This only shows up with SO_REUSEPORT listener migration, where closing a
listener hands its still-pending TFO children over to a surviving one.
fastopenq.qlen is charged in tcp_fastopen_create_child() when the child
is created and uncharged in reqsk_fastopen_remove() when the handshake
completes. The uncharge follows rsk_listener of the request the child
points at, and inet_reqsk_clone() has repointed the child at a new
request owned by the new listener, so the ++ and the -- land on two
different sockets. The new listener's qlen drifts negative and its
limit no longer binds.
Charge the new listener during migration, like reqsk_queue_migrated()
already does for queue->young and queue->qlen.
Fixes: 54b92e8419 ("tcp: Migrate TCP_ESTABLISHED/TCP_SYN_RECV sockets in accept queues.")
Signed-off-by: Jiayuan Chen <jiayuan.chen@linux.dev>
Reviewed-by: Kuniyuki Iwashima <kuniyu@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20260803061739.134737-1-jiayuan.chen@linux.dev
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Partial pull of the nf-26-07-31 tag
Pablo says:
====================
The following patchset contains Netfilter/IPVS fixes net, this
includes fixes for ebtables nflog target, ipset hash type,
IPVS kthread estimator
1) Prevent IPVS kthread estimator from draining the est_temp_list
when netns is being dismantled. From Zhiling Zou.
2) Missing module nflog refcount bump from ebtables nflog target from
.checkentry path. Similar dependency exists already in xt_NFLOG and
nft_log. From Chengfeng Ye.
3) Use RCU to fix ipset bookkeeping of cidr values on weakly-ordered
architectures. From Jozsef Kadlecsik.
4) Use atomic64_t for set->ext_size in ipset to fix parallel inserts
and deletes racing on updating it. From Jozsef Kadlecsik.
5) Add small wrappers for hash and bucket size to prepare the update
of ipset hash set types to rhashtable, from Florian Westphal.
6) Add mtype_del_cidr_all() and use it to prepare the migration of
ipset hash types to rhashtable. From Florian Westphal.
7) Replace existing ipset call_rcu() based destruction with rcu_work
api also to ease the transition to rhashtable. Also from Florian.
8) Avoid reading the IPv4 ihl field multiple times to prevent local
attacker to cause out-of-bounds write in ip_vs_nat_icmp(), from
Julian Anastasov.
9) Restore the checksum validations that could be needed by the IPVS
FORWARD hook. Also from Julian.
====================
Link: https://patch.msgid.link/20260731151806.849724-1-pablo@netfilter.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
bpf_tcp_gen_syncookie and bpf_tcp_check_syncookie accept a socket pointer
'sk' with argument type ARG_PTR_TO_BTF_ID_SOCK_COMMON. However, they access
sk->sk_protocol without validating whether 'sk' represents a full socket.
Fix this issue by checking sk->sk_state != TCP_LISTEN before inspecting
sk->sk_protocol in both bpf_tcp_gen_syncookie and bpf_tcp_check_syncookie.
Since mini-sockets are never in the TCP_LISTEN state, the condition
short-circuits and prevents dereferencing fullsock-specific fields.
Fixes: 3990408470 ("bpf: add helper to check for a valid SYN cookie")
Fixes: 70d6624431 ("bpf: add bpf_tcp_gen_syncookie helper")
Reported-by: Vega <vega@nebusec.ai>
Signed-off-by: Luxiao Xu <rakukuip@gmail.com>
Signed-off-by: Ren Wei <enjou1224z@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Kuniyuki Iwashima <kuniyu@google.com>
Link: https://lore.kernel.org/bpf/6218aa3534d0d2d3f448fde70a8dc2769d7a8201.1785823138.git.rakukuip@gmail.com
In sctp_process_asconf(), when sctp_make_asconf_ack() fails to allocate
the ASCONF_ACK chunk due to memory pressure, the code jumps to the
done label where asoc->peer.addip_serial is unconditionally incremented.
This leaves the peer's ASCONF (serial N) unacknowledged while the local
endpoint now expects serial N+1. When the peer retransmits serial N, it
falls into the serial < addip_serial + 1 branch ,
which attempts to look up a cached ACK for serial N. No cached ACK
exists since the allocation failed, so the retransmission is silently
discarded. The peer eventually times out and ABORTs the association.
Move the addip_serial increment inside the if (asconf_ack) block so that
the serial number is only advanced when the ASCONF_ACK is successfully
created and cached. This way, on allocation failure, the serial number
is unchanged and the peer's retransmitted ASCONF will be correctly
re-processed.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Qing Luo <luoqing@kylinos.cn>
Acked-by: Xin Long <lucien.xin@gmail.com>
Link: https://patch.msgid.link/20260804025514.241767-1-l1138897701@163.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
ovs_flow_cmd_new() preallocates the optional reply skb before it takes
ovs_mutex and before it knows which existing flow will be updated.
That is normally fine because the skb is sized from the request flow
identifier. That identifier also becomes the inserted flow's identifier.
For updates, however, a request with a UFID may miss the UFID lookup and
then fall back to the flow key lookup. That lookup can legitimately find
an existing key-identified flow. UFIDs are optional and the flow key is
the primary identifier.
For echoed replies, ovs_flow_cmd_fill_info() writes the matched flow's
identifier, not the request identifier used for the preallocation. A short
request UFID can therefore leave too little room for the key identifier.
The fill can then fail with -EMSGSIZE and hit the BUG_ON(error < 0) in the
update path.
Once the update target has been resolved, reallocate the reply skb if the
matched flow needs a larger reply than the request identifier allowed. Do
this before replacing the actions so the request can still fail cleanly if
the rare extra allocation fails.
Fixes: 74ed7ab926 ("openvswitch: Add support for unique flow IDs.")
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>
Link: https://patch.msgid.link/f7bbd3c30ce81a39156e226b3872d73abed21d2f.1785644623.git.zhilinz@nebusec.ai
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Another challenge with unlocked filters.
There is a short window in tc_new_tfilter where a tcf_proto can be found
and briefly referenced by a totally unrelated, unlocked classifier's request
and cause a race.
Feng created a poc which created this race with two threads, one creating a
u32 filter and other a flower filter in the same chain/prio:
1. Both threads enter tc_new_tfilter, both find the chain empty, both
drop filter_chain_lock
2. u32 finishes tcf_proto_create("u32") first, calls
tcf_chain_tp_insert_unique() -> inserts u32_tp into the chain
3. flower finishes tcf_proto_create("flower") later, calls
tcf_chain_tp_insert_unique() -> tcf_chain_tp_find() now sees u32_tp
already there, takes a reference on it, destroys flower's own tp_new
and returns u32_tp to the caller.
Flower then hits the kind mismatch check (because it requested for kind
"flower" but tp->ops->kind is "u32") and goes through the errout path
which calls tcf_proto_put() on u32_tp. If the u32 thread has already
gone through its own errout (its change() call failed on the PoC's empty
options) and dropped its create and insert refs, flower's put is the
last one and drops u32_tp's refcnt to zero.
At this point tp->ops->destroy() runs in a context that never took
rtnl_lock. When that happens, it might cause a UAF like the following
(illustrated by the PoC):
[ +0.000710] BUG: KASAN: slab-use-after-free in u32_init (net/sched/cls_u32.c:393)
[ +0.000281] Read of size 8 at addr ffff888120022f00 by task poc_feng_xue/524
Call Trace:
u32_init (net/sched/cls_u32.c:393)
tc_new_tfilter (net/sched/cls_api.c:2378)
Allocated by task 526:
u32_init (net/sched/cls_u32.c:378)
tc_new_tfilter (net/sched/cls_api.c:2378)
Freed by task 522:
kfree
u32_destroy (net/sched/cls_u32.c:662)
tcf_proto_destroy (net/sched/cls_api.c:446)
tcf_proto_put (net/sched/cls_api.c:459)
tc_new_tfilter (net/sched/cls_api.c:2459)
Fix this by having tcf_proto_destroy() take rtnl_lock around
tp->ops->destroy() for locked classifiers whenever rtnl is not held.
To explain why I used a temp variable "not_lockless" I'd like to point to a
semi-related note on rtnl_held vs TCF_PROTO_OPS_DOIT_UNLOCKED (adding here
for future cleanup if deemed necessary):
The rtnl_held parameter and the TCF_PROTO_OPS_DOIT_UNLOCKED flag are
redundant sources of truth for whether rtnl_lock is held. Among the nine
classifier destroy(..rtnl_held..) callbacks, only flower consults the
rtnl_held parameter which it propagates to tc_setup_cb_destroy()
and tc_setup_cb_call(). The other eight (u32, flow, bpf, cgroup, route, basic,
fw, mall) ignore it entirely;-> those that call tc_setup_cb_destroy()
(u32, bpf, mall) hardcode true always instead of forwarding the parameter.
A future cleanup should remove the rtnl_held parameter from the destroy callback
signature entirely and have callers rely solely on their knowledge whether
they are running in an unlocked context.
Fixes: 12db03b65c ("net: sched: extend proto ops to support unlocked classifiers")
Reported-by: Feng Xue <feng.xue@outlook.com>
Tested-by: Victor Nogueira <victor@mojatatu.com>
Signed-off-by: Jamal Hadi Salim <jhs@mojatatu.com>
Link: https://patch.msgid.link/20260801125632.360365-1-jhs@mojatatu.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
When a packet arrives on an ARPHRD_NONE device (e.g. TUN),
ovs_flow_key_extract() trusts the user-provided skb->protocol field: if
it is ETH_P_TEB, the packet is classified as MAC_PROTO_ETHERNET and
key_extract() is called without ensuring the skb has ETH_HLEN (14) bytes
of linear data. key_extract() unconditionally pulls 2 * ETH_ALEN bytes
for MAC addresses and parse_ethertype() pulls 2 more, either of which
triggers a kernel BUG in __skb_pull() when the linear area is too small.
kernel BUG at include/linux/skbuff.h:2848!
RIP: 0010:key_extract+0xa7e/0xd90 net/openvswitch/flow.c:933
ovs_flow_key_extract+0x419/0xa70
ovs_vport_receive+0x222/0x390
netdev_frame_hook+0x3e0/0x630
tun_get_user+0x2d0c/0x38e0
Fixed by calling check_header() in key_extract() before accessing the
Ethernet header.
Fixes: 217ac77a3c ("openvswitch: allow L3 netdev ports")
Reported-by: AutonomousCodeSecurity@microsoft.com
Reviewed-by: Eelco Chaudron <echaudro@redhat.com>
Signed-off-by: Cen Zhang (Microsoft) <blbllhy@gmail.com>
Reviewed-by: Ilya Maximets <i.maximets@ovn.org>
Link: https://patch.msgid.link/20260730222006.118652-1-blbllhy@gmail.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
packet_set_ring() updates the RX ring state under sk_receive_queue.lock,
but used to publish the tpacket receive mode through po->prot_hook.func
after releasing that lock. packet_poll() and packet_recvmsg() can then
run the pressure clearing path after the ring has been cleared while
still seeing tpacket_rcv, causing __packet_rcv_has_room() to dereference
stale or NULL ring storage.
Move the existing receive hook assignment into the same
sk_receive_queue.lock section as the ring state update. Keep the
assignment otherwise unchanged, including on TX ring reconfiguration, to
avoid adding behavior changes that are not required for the fix.
Serialize packet_recvmsg() pressure clearing with the same queue lock
only after PACKET_SOCK_PRESSURE has been observed. If the flag is clear
and the socket has moved away from tpacket_rcv, packet_set_ring() has
already detached the socket and waited for synchronize_net(), so no new
packet input can set the flag again.
packet_poll() already holds sk_receive_queue.lock, so it uses the new
unlocked helper directly.
Fixes: 2ccdbaa6d5 ("packet: rollover lock contention avoidance")
Cc: stable@vger.kernel.org
Reported-by: Vega <vega@nebusec.ai>
Assisted-by: Codex:gpt-5.4
Signed-off-by: Zihan Xi <zihanx@nebusec.ai>
Link: https://patch.msgid.link/f90b5688311fa278d1361ea8c6be0bf25967d591.1785247446.git.zihanx@nebusec.ai
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
The sch_cake ACK filter parses packets to find the TCP header and filter
duplicated ACKs if the flow is backlogged. The parsing code contains a
WARN_ON(1) which can be triggered by a malformed IP header in certain
cases. Depending on the system configuration, this leads either to
either spamming dmesg with warnings, or a panic if panic_on_warn is set.
The code already correctly skips the offending packet in the branch that
triggers the warning, so the WARN_ON itself doesn't really serve any
purpose. So just drop it altogether to avoid the inconvenient side
effects.
Fixes: 8b7138814f ("sch_cake: Add optional ACK filter")
Reported-by: Zhiling Zou <zhilinz@nebusec.ai>
Reported-by: Ren Wei <enjou1224z@gmail.com>
Signed-off-by: Toke Høiland-Jørgensen <toke@toke.dk>
Link: https://patch.msgid.link/20260729191417.45665-1-toke@toke.dk
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
__skb_udp_tunnel_segment() gets the UDP header before ensuring the
tunnel header is in the skb head. If the pull reallocates skb->head,
the saved UDP header pointer is no longer valid.
Get the UDP header after the pull to avoid a potential use-after-free.
Fixes: dbef491ebe ("udp: Use uh->len instead of skb->len to compute checksum in segmentation")
Signed-off-by: Xuanqiang Luo <luoxuanqiang@kylinos.cn>
Reviewed-by: Antoine Tenart <atenart@kernel.org>
Link: https://patch.msgid.link/20260730093554.68127-1-xuanqiang.luo@linux.dev
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
fib_nlmsg_size() still estimates nexthop space as if every gateway is
encoded as an IPv4 RTA_GATEWAY attribute. IPv4 routes can also carry an
IPv6 gateway, which fib_nexthop_info() dumps as RTA_VIA.
As a result, route notifications can allocate an skb that is too small.
fib_dump_info() then fails with -EMSGSIZE and rtmsg_fib() hits the
WARN_ON() that marks such failures as a fib_nlmsg_size() bug. With
panic_on_warn set, this becomes a kernel panic.
Mirror the actual nexthop dump layout in fib_nlmsg_size(): account for
IPv6 nexthop gateways dumped as RTA_VIA, for the no-header rtnexthop
layout used inside RTA_MULTIPATH, and for RTA_FLOW only when it is
actually present.
Fixes: d15662682d ("ipv4: Allow ipv6 gateway with ipv4 routes")
Cc: stable@vger.kernel.org
Reported-by: Vega <vega@nebusec.ai>
Signed-off-by: Zihan Xi <zihanx@nebusec.ai>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Link: https://patch.msgid.link/6f53fa797fcaeb26966432ed7ae9bb87c4961f37.1785411220.git.zihanx@nebusec.ai
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
smc_rx_splice() passes pages to splice_to_pipe() before taking the
references that cover the lifetime of each splice entry. In the
VM-backed RMB path, splice_to_pipe() may drop unqueued entries through
smc_rx_spd_release(), while queued entries are released later via the
pipe buffer callback.
The old post-splice accounting also derives the number of queued VM pages
from an offset mutated while building the descriptor, and a multi-page
splice pairs one sock_hold() with multiple sock_put() calls.
Take the page and socket references for every candidate entry before
splice_to_pipe(), and drop the matching private state, page reference,
and socket reference from smc_rx_spd_release() for entries that never
get queued. This fixes a refcount imbalance that can underflow page
refcounts and trigger a use-after-free.
Fixes: 9014db202c ("smc: add support for splice()")
Cc: stable@vger.kernel.org
Reported-by: Vega <vega@nebusec.ai>
Co-developed-by: Xiao Liu <lx24@stu.ynu.edu.cn>
Signed-off-by: Xiao Liu <lx24@stu.ynu.edu.cn>
Signed-off-by: Daming Li <d4n.for.sec@gmail.com>
Signed-off-by: Ren Wei <enjou1224z@gmail.com>
Reviewed-by: Dust Li <dust.li@linux.alibaba.com>
Reviewed-by: Sidraya Jayagond <sidraya@linux.ibm.com>
Link: https://patch.msgid.link/20260730145552.360287-2-enjou1224z@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The zero-copy path validates TX metadata while obtaining the descriptor
context, then reads it again later when preparing the hardware request.
User space can change the metadata between those operations and bypass the
original validation.
Validate the metadata in xsk_tx_metadata_request() and use the resulting
flags snapshot for every feature check. Read request fields once so all
zero-copy drivers process only values observed after successful
validation.
Fixes: ca4419f15a ("xsk: Add launch time hardware offload support to XDP Tx metadata")
Cc: Cen Zhang (Microsoft) <blbllhy@gmail.com>
Signed-off-by: Stanislav Fomichev <sdf@fomichev.me>
Link: https://patch.msgid.link/20260727161959.885642-7-sdf@fomichev.me
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Launch-time metadata extends beyond the first 16 bytes of struct
xsk_tx_metadata. Reject the request when the registered metadata area does
not contain the complete field.
Snapshot the validated flags for the generic transmit path and use that
snapshot for request and completion processing, avoiding inconsistent
decisions if user space changes the flags concurrently.
Note that only xsk_skb_metadata is properly using the flags,
__xsk_buff_get_metadata ignores them. Next commits address that.
Fixes: ca4419f15a ("xsk: Add launch time hardware offload support to XDP Tx metadata")
Cc: Cen Zhang (Microsoft) <blbllhy@gmail.com>
Signed-off-by: Stanislav Fomichev <sdf@fomichev.me>
Link: https://patch.msgid.link/20260727161959.885642-5-sdf@fomichev.me
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
User space can change metadata flags after request processing. Rereading
them during completion can therefore make the kernel write a timestamp
that was not requested when the packet was submitted.
Clear the metadata pointer during request processing unless timestamp
completion is requested. Completion handling can then use the pointer
itself instead of rereading the flags.
On the mlx5 multi-packet WQE path metadata is evaluated per batch:
xsk_tx_metadata_request() runs only for the descriptor that starts a
session, just like the checksum offload that is applied once through the
shared WQE. Only that descriptor's pointer is reset, so completion
handling can record a timestamp for the other descriptors of the session
regardless of their own XDP_TXMD_FLAGS_TIMESTAMP bit. The write stays
inside the metadata area; the single-WQE, other zero-copy, and generic
paths reset the pointer per descriptor and are unaffected.
Fixes: ca4419f15a ("xsk: Add launch time hardware offload support to XDP Tx metadata")
Cc: Cen Zhang (Microsoft) <blbllhy@gmail.com>
Signed-off-by: Stanislav Fomichev <sdf@fomichev.me>
Link: https://patch.msgid.link/20260727161959.885642-4-sdf@fomichev.me
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Commit b917507e5a ("vsock/virtio: stop workers during the .remove()")
made the RX worker jump to its common exit when rx_run is clear. That
exit still refills the RX queue when the buffer count is low, so work
queued across virtio_vsock_vqs_del() can add buffers after the virtqueues
have been deleted.
BUG: KASAN: slab-use-after-free in virtqueue_add_sgs
Read of size 4 by task kworker/0:1
Workqueue: virtio_vsock virtio_transport_rx_work
Call Trace:
virtqueue_add_sgs (drivers/virtio/virtio_ring.c:2796)
virtio_vsock_rx_fill (net/vmw_vsock/virtio_transport.c:332)
virtio_transport_rx_work (net/vmw_vsock/virtio_transport.c:701)
process_one_work (kernel/workqueue.c:3314)
worker_thread (kernel/workqueue.c:3478)
kthread (kernel/kthread.c:436)
ret_from_fork (arch/x86/kernel/process.c:158)
ret_from_fork_asm (arch/x86/entry/entry_64.S:245)
...
Freed by task 141:
kfree (mm/slub.c:6566)
vp_del_vq (drivers/virtio/virtio_pci_common.c:259)
vp_del_vqs (drivers/virtio/virtio_pci_common.c:285)
virtio_vsock_freeze (net/vmw_vsock/virtio_transport.c:912)
virtio_device_freeze (drivers/virtio/virtio.c:658)
virtio_pci_freeze (drivers/virtio/virtio_pci_common.c:601)
pci_pm_freeze (drivers/pci/pci-driver.c:1098)
device_suspend (drivers/base/power/main.c:1968)
Kernel panic - not syncing: KASAN: panic_on_warn set ...
Jump to a no-refill exit when rx_run is clear, leaving the normal exit
to replenish a running queue.
Fixes: b917507e5a ("vsock/virtio: stop workers during the .remove()")
Cc: stable@vger.kernel.org
Reported-by: Xiang Mei <xmei5@asu.edu>
Link: https://lore.kernel.org/r/20260727035804.1860862-1-bestswngs@gmail.com
Suggested-by: Stefano Garzarella <sgarzare@redhat.com>
Signed-off-by: Weiming Shi <bestswngs@gmail.com>
Reviewed-by: Bobby Eshleman <bobbyeshleman@meta.com>
Link: https://patch.msgid.link/f9c8c1d64cad9d262f305d02ffe164c2f900fadf.1785352330.git.bestswngs@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
rt6_route_rcv() validates the Route Information option (RFC 4191) length
against the prefix length, but both checks are off by one.
rinfo->length is the ND option length in units of 8 octets and it
*includes* the 8-byte option header, so an option carrying N bytes of
prefix has length == 1 + N/8. RFC 4191 section 2.3 requires length 3
when Prefix Length is greater than 64, and 2 or 3 when it is greater
than 0. The code accepts length >= 2 and length >= 1 respectively.
ipv6_addr_prefix() then copies prefix_len/8 bytes out of rinfo->prefix,
so a Router Advertisement with (prefix_len=128, length=2) or
(prefix_len=64, length=1) makes the kernel read up to 8 bytes past the
end of the option. Those bytes end up in the prefix of the route that
gets installed, so they are visible to userspace:
# RA with a Route Information option (prefix_len=128, length=2)
# followed by a source link-layer address option, 01 01 de ad be ef ca fe
$ ip -6 route show
2001:db8:dead:beef:101:dead:beef:cafe via fe80::1234 dev veth0 proto ra
^^^^^^^^^^^^^^^^^^ the next option, read out of bounds
When the Route Information option is the last one in the packet, those
eight bytes come from the skb tail room instead.
Reject the option lengths RFC 4191 does not allow.
Fixes: 70ceb4f539 ("[IPV6]: ROUTE: Add experimental support for Route Information Option in RA (RFC4191).")
Cc: stable@vger.kernel.org
Signed-off-by: Yuejie Shi <syjcnss@gmail.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Link: https://patch.msgid.link/20260730035310.74584-1-syjcnss@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
__sctp_outq_flush_rtx() moves a gap-acked chunk onto another transport's
transmitted list without updating chunk->transport:
if (chunk->tsn_gap_acked) {
list_move_tail(&chunk->transmitted_list,
&transport->transmitted);
continue;
}
The chunk then sits on a live transport's list while chunk->transport still
names a different one. If that transport is removed - sctp_assoc_rm_peer()
from an ASCONF Delete-IP - sctp_transport_free() RCU-frees it and the chunk
is left with a dangling pointer. sctp_assoc_rm_peer() scrubs
peer->transmitted and asoc->outqueue.out_chunk_list, but the chunk is on
neither.
The pointer is not followed while tsn_gap_acked is set. A SACK that
reneges on the TSN clears the flag, and the next SACK reaches
tchunk->transport->flight_size -= sctp_data_size(tchunk);
inside the freed transport. KASAN reports a slab-use-after-free read in
sctp_check_transmitted(), freed from sctp_assoc_rm_peer(). Both the
removal and the SACKs come from the association peer.
Set chunk->transport at the move. The ordinary resend path needs nothing:
it reaches its list_move_tail() only after sctp_packet_append_chunk()
returned SCTP_XMIT_OK, and __sctp_packet_append_chunk() has rebound the
chunk by then.
Discovered by XBOW, triaged by Baul Lee <baul.lee@xbow.com>
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Cc: stable@vger.kernel.org
Signed-off-by: Baul Lee <baul.lee@xbow.com>
Acked-by: Xin Long <lucien.xin@gmail.com>
Link: https://patch.msgid.link/20260729160028.54546-1-baul.lee@xbow.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Commit f5da7c4518 ("tcp: adjust rcvq_space after updating scaling
ratio") replaced the direct window_clamp update in tcp_measure_rcv_mss()
with a call to tcp_set_window_clamp(), a helper that implements the
TCP_WINDOW_CLAMP setsockopt. As a side effect, the helper also shrinks
rcv_ssthresh via __tcp_adjust_rcv_ssthresh().
As a result, each scaling_ratio decrease detected by
tcp_measure_rcv_mss() also cuts rcv_ssthresh. Elsewhere in TCP,
rcv_ssthresh is usually cut under memory pressure and grows via
tcp_grow_window().
Flows whose segment sizes vary keep scaling_ratio oscillating, which
leads to an unstable rcv_ssthresh: a dip of rcv_ssthresh only recovers
via tcp_grow_window(), keeping the advertised window at a relatively
low level even after the ratio itself has recovered, and can even stall
the sender.
Observed on a customer's proxy gateway after upgrading from kernel 6.1
to 6.12: in the worst case, rcv_ssthresh was cut in half by a
scaling_ratio dip. P99 latency jumped from <10ms on 6.1 to ~100ms on
6.12, and almost returned to the 6.1 level with this patch applied.
Restore the plain WRITE_ONCE() update of window_clamp, as introduced
in commit a2cbb16039 ("tcp: Update window clamping condition"), and
keep the rcvq_space.space adjustment. Now rcv_ssthresh is decoupled from
scaling_ratio changes in tcp_measure_rcv_mss().
Fixes: f5da7c4518 ("tcp: adjust rcvq_space after updating scaling ratio")
Signed-off-by: Nathan Gao <zcgao@amazon.com>
Link: https://patch.msgid.link/20260725030806.28135-1-zcgao@amazon.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
reqsk_queue_hash_req() publishes a TCP_NEW_SYN_RECV request_sock onto
the ehash chain, drops the bucket lock, and only afterwards sets
rsk_refcnt to 3.
Lockless readers such as __inet_lookup_established() handle this with
refcount_inc_not_zero(), but bpf_iter_tcp_established_batch() uses plain
sock_hold() while holding the bucket lock, on the assumption that the
lock guarantees sk_refcnt > 0. That assumption does not hold for
request_sock:
CPU 0 CPU 1
----- -----
tcp_conn_request()
reqsk_queue_hash_req()
inet_ehash_insert(req)
spin_lock(bucket)
__sk_nulls_add_node_rcu(req) // rsk_refcnt == 0
spin_unlock(bucket)
bpf_iter_tcp_established_batch()
spin_lock(bucket)
sock_hold(req) <-- addition on 0
spin_unlock(bucket)
refcount_set(&req->rsk_refcnt, 3) // clobbers saturated value
which surfaces as:
refcount_t: addition on 0; use-after-free.
WARNING: lib/refcount.c:25 at refcount_warn_saturate+0x48/0x90, CPU#1
Call Trace:
bpf_iter_tcp_established_batch+0x14e/0x170
bpf_iter_tcp_batch+0x53/0x200
bpf_iter_tcp_seq_next+0x27/0x70
bpf_seq_read+0x107/0x410
vfs_read+0xb9/0x380
The iterator's stolen reference is lost when the publishing CPU's
refcount_set() overwrites the count, leaving the socket one reference
short. When the last legitimate owner drops its reference the reqsk is
freed while still reachable, leading to use-after-free.
This reproduces in seconds with tcp_syncookies=0, a handful of threads
doing connect()/close() to a local listener while others read an
iter/tcp link in a tight loop.
Use refcount_inc_not_zero() and skip the socket on failure. A skipped
socket is still part of the bucket, so keep counting it in expected.
The reallocations are sized from expected, and a request sock whose
refcount gets published while the lock is held across the last realloc
must already have room.
A skipped socket is counted in expected but never batched, so end_sk
can be short of expected on a batch that is actually complete. Decide
completeness by whether the walk left any socket behind instead. The
WARN after the locked realloc checks the same, replacing an
end_sk == expected check that could not hold on that path since
commit cdec67a489 ("bpf: tcp: Make sure iter->batch always
contains a full bucket snapshot").
If every matching socket in a bucket is mid-init (refcount 0), end_sk
stays 0. Advance to the next bucket rather than returning a batch entry
that was never filled this round.
Fixes: 04c7820b77 ("bpf: tcp: Bpf iter batching and lock_sock")
Assisted-by: Claude:unspecified
Signed-off-by: Jose Fernandez (Anthropic) <jose.fernandez@linux.dev>
Reviewed-by: Kuniyuki Iwashima <kuniyu@google.com>
Link: https://lore.kernel.org/bpf/20260730-bpf-iter-tcp-refcnt-v3-1-754b9c8a6717@linux.dev
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
The opts argument of the BPF conntrack kfuncs can point to a shared
map value. __bpf_nf_ct_lookup() and __bpf_nf_ct_alloc_entry() read
opts->netns_id separately when acquiring and releasing the network
namespace reference.
The reference imbalance can occur as follows:
CPU 0 CPU 1
read opts->netns_id (-1)
skip get_net_ns_by_id()
write opts->netns_id (id)
read opts->netns_id (id)
put_net(net) /* no matching get */
The reverse transition leaks the reference. Repeating the unmatched put
can destroy a live namespace and crash later users.
The kernel reported:
Oops: general protection fault, probably for non-canonical address
KASAN: null-ptr-deref in range [0x00000000000000e8-0x00000000000000ef]
RIP: 0010:bpf_prog_test_run_xdp+0x52c/0x1700
Call Trace:
__sys_bpf+0x1662/0x50c0
__x64_sys_bpf+0x73/0xb0
do_syscall_64+0xf9/0x540
entry_SYSCALL_64_after_hwframe+0x77/0x7f
Kernel panic - not syncing: Fatal exception
Snapshot every input field of opts with READ_ONCE() before validating or
using it. The netns_id snapshot keeps the namespace get/put pair
balanced, while the other snapshots keep the remaining options from
changing partway through an invocation. The individual reads can still
observe an inconsistent combination during a concurrent update, but each
selected field value remains stable for that invocation.
Fixes: aed8ee7feb ("net: netfilter: Deduplicate code in bpf_{xdp,skb}_ct_lookup")
Fixes: d7e79c97c0 ("net: netfilter: Add kfuncs to allocate and insert CT")
Signed-off-by: Chengfeng Ye <nicoyip.dev@gmail.com>
Reviewed-by: Emil Tsalapatis <emil@etsalapatis.com>
Link: https://lore.kernel.org/bpf/20260731160921.3245840-1-nicoyip.dev@gmail.com
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
sk_psock_msg_verdict() takes a socket reference for psock->sk_redir.
tcp_bpf_send_verdict() copies that pointer while holding the source socket
lock, but does not take a reference for the local copy before dropping the
lock around tcp_bpf_sendmsg_redir().
When apply_bytes keeps the cached verdict active, another sendmsg() on the
same source socket can consume the remaining bytes and release the cached
reference while the first thread still holds only the raw local pointer:
CPU 0 CPU 1
sk_redir = psock->sk_redir
apply_bytes remains nonzero
release_sock(sk)
lock_sock(sk)
apply_bytes reaches zero
psock->sk_redir = NULL
release_sock(sk)
tcp_bpf_sendmsg_redir(sk_redir)
sock_put(sk_redir)
tcp_bpf_sendmsg_redir(sk_redir)
The final sock_put() can free sk_redir before CPU 0 dereferences it.
KASAN reported:
BUG: KASAN: slab-use-after-free in tcp_bpf_sendmsg_redir+0xf39/0x1020
Read of size 8 at addr ffff888108537090 by task poc/87
Call Trace:
tcp_bpf_sendmsg_redir+0xf39/0x1020
tcp_bpf_sendmsg+0x977/0x1a50
__sys_sendto+0x32c/0x3a0
__x64_sys_sendto+0xdb/0x1b0
Allocated by task 85:
sk_prot_alloc+0x56/0x210
sk_clone+0x6f/0x14b0
inet_csk_clone_lock+0x24/0x740
tcp_create_openreq_child+0x25/0x2710
tcp_v4_syn_recv_sock+0x10a/0xe00
Freed by task 0:
__kasan_slab_free+0x43/0x70
slab_free_after_rcu_debug+0xa6/0x1e0
rcu_core+0x50a/0x1850
Last potentially related work creation:
__sk_destruct+0x3da/0x540
sk_psock_destroy+0x81e/0xab0
process_one_work+0x63a/0x1070
Take a temporary socket reference while the source socket lock still
protects psock->sk_redir, and drop it after tcp_bpf_sendmsg_redir()
returns. This keeps each unlocked use independent of cached-verdict
ownership.
Fixes: 604326b41a ("bpf, sockmap: convert to generic sk_msg interface")
Signed-off-by: Chengfeng Ye <nicoyip.dev@gmail.com>
Reviewed-by: John Fastabend <john.fastabend@gmail.com>
Reviewed-by: Emil Tsalapatis <emil@etsalapatis.com>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/bpf/20260719152207.2892156-1-nicoyip.dev@gmail.com
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
