linux-stable.git/net/ipv4, branch v7.2.4 Linux kernel stable tree net: cap advertised IP tunnel headroom 2026-09-07T15:37:21+00:00 Zhiling Zou zhilinz@nebusec.ai 2026-08-12T16:22:35+00:00 9144f2c53a04465a6878172b523f640313c5559e commit 6b222adeb9340306e2ff97127c76117abb9b3df8 upstream. IP tunnel devices derive their advertised needed_headroom from lower output devices. A stack of user-created devices can make the derived value larger than the 16-bit skb header offsets can represent. Once IP output reserves it, skb head expansion can wrap those offsets. The runtime transmit path already caps a growing needed_headroom at 512. Apply the same cap when tunnel configuration publishes needed_headroom derived from a lower output device. Capping the advertised value is safe: IP tunnel transmit still expands the skb when a packet needs more headroom. A nonsensical stacked configuration can therefore incur an extra reallocation, but it cannot publish an unbounded reservation to upper layers. Fixes: 1a37e412a022 ("net: Use 16bits for *_headers fields of struct skbuff") 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/ba04a1fd6bfae2377607fad5d8f80f7eb80fd4c4.1786542637.git.zhilinz@nebusec.ai Signed-off-by: Paolo Abeni <pabeni@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> 
commit 6b222adeb9340306e2ff97127c76117abb9b3df8 upstream.

IP tunnel devices derive their advertised needed_headroom from lower
output devices. A stack of user-created devices can make the derived
value larger than the 16-bit skb header offsets can represent. Once IP
output reserves it, skb head expansion can wrap those offsets.

The runtime transmit path already caps a growing needed_headroom at 512.
Apply the same cap when tunnel configuration publishes needed_headroom
derived from a lower output device.

Capping the advertised value is safe: IP tunnel transmit still expands
the skb when a packet needs more headroom. A nonsensical stacked
configuration can therefore incur an extra reallocation, but it cannot
publish an unbounded reservation to upper layers.

Fixes: 1a37e412a022 ("net: Use 16bits for *_headers fields of struct skbuff")
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/ba04a1fd6bfae2377607fad5d8f80f7eb80fd4c4.1786542637.git.zhilinz@nebusec.ai
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
ip: orphan prefetched skbs before multicast forwarding 2026-09-07T15:36:58+00:00 Zhiling Zou zhilinz@nebusec.ai 2026-08-12T16:36:38+00:00 f1281d4b99089fbaf0d3579bba62a814ebb4de4d commit e36ce6e78fe3fc3c071a26750783b7ba081ce10d upstream. IPv4 and IPv6 input preserve an skb->sk association installed by bpf_sk_assign() so that local delivery can use the selected socket under RCU. Both address families can also prefetch a socket in UDP early demux. In both paths (BPF and UDP early demux) a reference is not guaranteed to be held on the socket. When a multicast packet is not locally deliverable, IPv6 hands the original skb to ip6_mr_input(). IPv4's ip_mr_input() similarly keeps the original skb when local delivery is not needed. Either path can put the skb on an unresolved multicast route queue or forward it after the receive-side RCU section ends. After the prefetched socket is destroyed, a later skb free invokes sock_pfree() and dereferences the stale skb->sk. Orphan the skb before each non-local multicast forwarding path. Local delivery retains the original skb; the existing skb_clone() calls provide multicast forwarding with a socket-free clone. Fixes: cf7fbe660f2d ("bpf: Add socket assign support") Fixes: 08842c43d016 ("udp: no longer touch sk->sk_refcnt in early demux") Cc: stable@vger.kernel.org Reported-by: Vega <vega@nebusec.ai> Signed-off-by: Zhiling Zou <zhilinz@nebusec.ai> 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/0c52eb3d7532aaf8bccf37e0f7c922143c639735.1786552223.git.zhilinz@nebusec.ai Signed-off-by: Paolo Abeni <pabeni@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> 
commit e36ce6e78fe3fc3c071a26750783b7ba081ce10d upstream.

IPv4 and IPv6 input preserve an skb->sk association installed by
bpf_sk_assign() so that local delivery can use the selected socket under
RCU. Both address families can also prefetch a socket in UDP early demux.
In both paths (BPF and UDP early demux) a reference is not guaranteed to
be held on the socket.

When a multicast packet is not locally deliverable, IPv6 hands the
original skb to ip6_mr_input(). IPv4's ip_mr_input() similarly keeps the
original skb when local delivery is not needed. Either path can put the
skb on an unresolved multicast route queue or forward it after the
receive-side RCU section ends.

After the prefetched socket is destroyed, a later skb free invokes
sock_pfree() and dereferences the stale skb->sk. Orphan the skb before
each non-local multicast forwarding path. Local delivery retains the
original skb; the existing skb_clone() calls provide multicast forwarding
with a socket-free clone.

Fixes: cf7fbe660f2d ("bpf: Add socket assign support")
Fixes: 08842c43d016 ("udp: no longer touch sk->sk_refcnt in early demux")
Cc: stable@vger.kernel.org
Reported-by: Vega <vega@nebusec.ai>
Signed-off-by: Zhiling Zou <zhilinz@nebusec.ai>
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/0c52eb3d7532aaf8bccf37e0f7c922143c639735.1786552223.git.zhilinz@nebusec.ai
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
ipip: fix skb leak in collect_md mode when metadata_dst allocation fails 2026-09-07T15:36:58+00:00 Anton Danilov littlesmilingcloud@gmail.com 2026-08-19T10:43:39+00:00 fd01f1a3ed4a189a784654f7c69eb565872f2e12 commit 6776efe4a52f289a3fc18f8adf19b035a7d8e1bb upstream. In collect_md mode ipip_tunnel_rcv() returns 0 without freeing the skb when ip_tun_rx_dst() fails to allocate the metadata_dst. ipip_rcv() and mplsip_rcv() are registered as xfrm_tunnel handlers, so tunnel4_rcv() and tunnelmpls4_rcv() read the zero return as "the packet has been consumed" and do not free it either. The skb is leaked. The other tunnel drivers all dispose of the packet at this point: ip6_tunnel.c jumps to its drop label, ip_gre.c and ip6_gre.c return PACKET_REJECT, which makes gre_rcv() free the skb. Only ipip returns 0. Jump to the existing drop label instead. It frees the skb and still returns 0, so the packet keeps being reported as consumed, which is what we want here: the outer header has already been pulled, and neither the remaining handlers nor an ICMP unreachable have any use for it. Triggering this needs an ipip or mplsip tunnel in collect_md mode and an atomic allocation failure, which is why it has gone unnoticed. Fixes: cfc7381b3002 ("ip_tunnel: add collect_md mode to IPIP tunnel") Cc: stable@vger.kernel.org Signed-off-by: Anton Danilov <littlesmilingcloud@gmail.com> Reviewed-by: Fernando Fernandez Mancera <fmancera@suse.de> Link: https://patch.msgid.link/20260819104338.432631-2-littlesmilingcloud@gmail.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> 
commit 6776efe4a52f289a3fc18f8adf19b035a7d8e1bb upstream.

In collect_md mode ipip_tunnel_rcv() returns 0 without freeing the skb
when ip_tun_rx_dst() fails to allocate the metadata_dst. ipip_rcv() and
mplsip_rcv() are registered as xfrm_tunnel handlers, so tunnel4_rcv()
and tunnelmpls4_rcv() read the zero return as "the packet has been
consumed" and do not free it either. The skb is leaked.

The other tunnel drivers all dispose of the packet at this point:
ip6_tunnel.c jumps to its drop label, ip_gre.c and ip6_gre.c return
PACKET_REJECT, which makes gre_rcv() free the skb. Only ipip returns 0.

Jump to the existing drop label instead. It frees the skb and still
returns 0, so the packet keeps being reported as consumed, which is what
we want here: the outer header has already been pulled, and neither the
remaining handlers nor an ICMP unreachable have any use for it.

Triggering this needs an ipip or mplsip tunnel in collect_md mode and an
atomic allocation failure, which is why it has gone unnoticed.

Fixes: cfc7381b3002 ("ip_tunnel: add collect_md mode to IPIP tunnel")
Cc: stable@vger.kernel.org
Signed-off-by: Anton Danilov <littlesmilingcloud@gmail.com>
Reviewed-by: Fernando Fernandez Mancera <fmancera@suse.de>
Link: https://patch.msgid.link/20260819104338.432631-2-littlesmilingcloud@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
tcp: clamp route advmss to TCP_MIN_MSS 2026-09-02T12:33:19+00:00 Yong Wang edragain@163.com 2026-08-19T15:22:04+00:00 31cf2349361902769dc323e4dbf4b449795ec288 commit 870a9e42ecc6fe1b8c25d87af043cb0d9c178fe1 upstream. tcp_select_initial_window() assumes that callers never pass an MSS smaller than 1, but route-derived advmss values can violate that assumption. A too-small explicit RTAX_ADVMSS is one way to get there, but it is not the only one. The same divide-by-zero can also be reached through the "default advmss" path when RTAX_ADVMSS is left at 0 and the effective advmss is later driven down by route MTU and min_adv_mss. Introduce a tcp_dst_advmss() helper that clamps route advmss to TCP_MIN_MSS before TCP consumes it, and use it in the TCP paths that derive advmss from dst metrics. This keeps the effective MSS from dropping to zero before tcp_select_initial_window() rounds the receive window. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Cc: stable@vger.kernel.org Reported-by: Vega <vega@nebusec.ai> Signed-off-by: Yong Wang <edragain@163.com> Signed-off-by: Ren Wei <weir@nebusec.ai> Link: https://patch.msgid.link/251eaf8277fa7c66364c9815c5da01662d269181.1787074852.git.edragain@163.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> 
commit 870a9e42ecc6fe1b8c25d87af043cb0d9c178fe1 upstream.

tcp_select_initial_window() assumes that callers never pass an MSS
smaller than 1, but route-derived advmss values can violate that
assumption.

A too-small explicit RTAX_ADVMSS is one way to get there, but it is not
the only one. The same divide-by-zero can also be reached through the
"default advmss" path when RTAX_ADVMSS is left at 0 and the effective
advmss is later driven down by route MTU and min_adv_mss.

Introduce a tcp_dst_advmss() helper that clamps route advmss to
TCP_MIN_MSS before TCP consumes it, and use it in the TCP paths that
derive advmss from dst metrics. This keeps the effective MSS from
dropping to zero before tcp_select_initial_window() rounds the receive
window.

Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
Cc: stable@vger.kernel.org
Reported-by: Vega <vega@nebusec.ai>
Signed-off-by: Yong Wang <edragain@163.com>
Signed-off-by: Ren Wei <weir@nebusec.ai>
Link: https://patch.msgid.link/251eaf8277fa7c66364c9815c5da01662d269181.1787074852.git.edragain@163.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
net: advertise TCP MSS from the configured MTU, not the learned PMTU 2026-09-02T12:33:19+00:00 Jiayuan Chen jiayuan.chen@linux.dev 2026-08-15T07:03:36+00:00 713ed6ce111e8f5162454bae12df4e7b27d35708 commit 2640e64195948a601430d230c9864f5426574cde upstream. The MSS a host puts in its SYN tells the peer how big a segment it may send us. Right now we can shrink it with a PMTU we learned on our own send path, which is the wrong direction entirely. On asymmetric paths this bites - think DSR load balancers, where the request side goes through a smaller-MTU overlay. We learn a small PMTU going out, then advertise a small MSS, and the peer stays capped for the whole connection even though its path back to us is wide. MSS only shows up in the SYN and never grows back. On symmetric paths we lose nothing by dropping it either: the peer runs its own PMTU discovery and usually already knows the real path MTU. So work out the advertised MSS from the configured route or device MTU and ignore the learned PMTU. Our send side is unchanged, still clamped by tcp_current_mss(). Add ip_dst_mtu_configured()/ip6_dst_mtu_configured() and use them from the two default_advmss() paths. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Fixes: 164a5e7ad531 ("ipv4: ipv4_default_advmss() should use route mtu") Cc: stable@vger.kernel.org Signed-off-by: Jiayuan Chen <jiayuan.chen@linux.dev> Reviewed-by: Eric Dumazet <edumazet@google.com> Link: https://patch.msgid.link/20260815070413.294559-1-jiayuan.chen@linux.dev Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> 
commit 2640e64195948a601430d230c9864f5426574cde upstream.

The MSS a host puts in its SYN tells the peer how big a segment it may
send us. Right now we can shrink it with a PMTU we learned on our own
send path, which is the wrong direction entirely.

On asymmetric paths this bites - think DSR load balancers, where the
request side goes through a smaller-MTU overlay. We learn a small PMTU
going out, then advertise a small MSS, and the peer stays capped for the
whole connection even though its path back to us is wide. MSS only shows
up in the SYN and never grows back.

On symmetric paths we lose nothing by dropping it either: the peer runs
its own PMTU discovery and usually already knows the real path MTU.

So work out the advertised MSS from the configured route or device MTU
and ignore the learned PMTU. Our send side is unchanged, still clamped by
tcp_current_mss(). Add ip_dst_mtu_configured()/ip6_dst_mtu_configured()
and use them from the two default_advmss() paths.

Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
Fixes: 164a5e7ad531 ("ipv4: ipv4_default_advmss() should use route mtu")
Cc: stable@vger.kernel.org
Signed-off-by: Jiayuan Chen <jiayuan.chen@linux.dev>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20260815070413.294559-1-jiayuan.chen@linux.dev
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
net/tcp-ao: fix use-after-free of current_key on reconnect to another peer 2026-09-02T12:33:19+00:00 Hyunwoo Kim imv4bel@gmail.com 2026-08-16T21:28:42+00:00 2857dcbd03cf3354af0fba1b65c6a260fb43811a commit da4471557f279d0f56605158a625bb6e49ef7d41 upstream. tcp_inbound_ao_hash() is called before bh_lock_sock_nested() is taken, with only rcu_read_lock() held. On the fast path for established sockets, if the rnext_keyid sent by the peer differs from current_key->sndid, the key the peer asked for is looked up and stored in current_key. The lookup is inside the RCU read side, but current_key outlives it. When the socket is disconnected and connect() is called again for another peer, tcp_ao_connect_init() unlinks every key that does not match the new peer and frees it with call_rcu(). If current_key points at such a key, it is cleared to NULL. The fast path reads sk_state only once on entry, so a softirq that got into it while the socket was still established can update current_key after that loop has already run. The update is inside the RCU read side, so it comes before the call_rcu() callback, and once the callback frees the key, current_key is left pointing at freed memory. The next transmission picks that pointer up in tcp_get_current_key(). tcp_ao_transmit_skb() then reads the traffic key from the freed object, which is the use-after-free. Wait for one grace period before unlinking, and only if a key is going to be removed. By the time tcp_connect() runs the socket is already in TCP_SYN_SENT, and TCP_AO_ESTABLISHED does not contain TCPF_SYN_SENT, so a softirq entering after the wait cannot reach the fast path, and the ones already in it have finished. The existing NULL handling in the loop is then enough. Fixes: 0a3a809089eb ("net/tcp: Verify inbound TCP-AO signed segments") Cc: stable@vger.kernel.org Signed-off-by: Hyunwoo Kim <imv4bel@gmail.com> Reviewed-by: Simon Horman <horms@kernel.org> Acked-by: Paolo Abeni <pabeni@redhat.com> Link: https://patch.msgid.link/aoIriv3pHDgII2YR@v4bel Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> 
commit da4471557f279d0f56605158a625bb6e49ef7d41 upstream.

tcp_inbound_ao_hash() is called before bh_lock_sock_nested() is taken,
with only rcu_read_lock() held. On the fast path for established
sockets, if the rnext_keyid sent by the peer differs from
current_key->sndid, the key the peer asked for is looked up and stored
in current_key. The lookup is inside the RCU read side, but current_key
outlives it.

When the socket is disconnected and connect() is called again for
another peer, tcp_ao_connect_init() unlinks every key that does not
match the new peer and frees it with call_rcu(). If current_key points
at such a key, it is cleared to NULL.

The fast path reads sk_state only once on entry, so a softirq that got
into it while the socket was still established can update current_key
after that loop has already run. The update is inside the RCU read side,
so it comes before the call_rcu() callback, and once the callback frees
the key, current_key is left pointing at freed memory.

The next transmission picks that pointer up in tcp_get_current_key().
tcp_ao_transmit_skb() then reads the traffic key from the freed object,
which is the use-after-free.

Wait for one grace period before unlinking, and only if a key is going
to be removed. By the time tcp_connect() runs the socket is already in
TCP_SYN_SENT, and TCP_AO_ESTABLISHED does not contain TCPF_SYN_SENT, so
a softirq entering after the wait cannot reach the fast path, and the
ones already in it have finished. The existing NULL handling in the loop
is then enough.

Fixes: 0a3a809089eb ("net/tcp: Verify inbound TCP-AO signed segments")
Cc: stable@vger.kernel.org
Signed-off-by: Hyunwoo Kim <imv4bel@gmail.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Acked-by: Paolo Abeni <pabeni@redhat.com>
Link: https://patch.msgid.link/aoIriv3pHDgII2YR@v4bel
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
tcp: fix AO info use-after-free in tcp_ao_connect_init() 2026-09-02T12:33:19+00:00 Qing Ming a0yami@mailbox.org 2026-08-25T07:20:33+00:00 d17e88b6b60ff4ef64e0dd1444f935cb13dee1cd commit ea30dc5267e367b8a5e1e06cc074f813bcbf18b2 upstream. tcp_v4_connect() adds a SYN-SENT socket to the ehash before calling tcp_connect(). If TCP-AO is configured, tcp_connect() first verifies that a key matches the peer and the bound device's current L3 master. tcp_ao_connect_init() later resolves the L3 master again and removes keys which do not match it. The socket lock does not stabilize the bound device's VRF membership. Detaching the device from its VRF between the initial validation and the L3-master calculation in tcp_ao_connect_init() can therefore make the validation succeed while initialization observes the default L3 domain and removes the only key. The subsequent AO lookup then fails, so the no-key path clears tp->ao_info and frees it directly. The receive path can find the socket in the ehash and load tp->ao_info under RCU before acquiring the socket lock. A reader which loaded the old pointer can thus continue into tcp_inbound_ao_hash() after the direct free. The issue was found during a static audit of TCP-AO object lifetime. An unprivileged reproducer in self-created user and network namespaces raced connect() with detaching a veth from its VRF while sending TCP-AO segments. It triggered the same KASAN report on two fresh boots: BUG: KASAN: slab-use-after-free in tcp_inbound_ao_hash+0x585/0x19f0 Write of size 8 at addr ffff88800bf88128 by task tcp_ao_vrf_race/232 Call Trace: tcp_inbound_ao_hash+0x585/0x19f0 tcp_inbound_hash+0x677/0xa80 tcp_v4_rcv+0x1c3e/0x3ab0 Allocated by task 235: tcp_ao_alloc_info+0x43/0xf0 tcp_ao_add_cmd+0xdf7/0x13b0 do_tcp_setsockopt+0x168c/0x2640 Freed by task 235: kfree+0x1b8/0x550 tcp_connect+0x252/0x4f00 tcp_v4_connect+0x1114/0x1720 The bad address is 40 bytes inside the freed 128-byte object, matching the tcp_ao_info counters.key_not_found field. The two runs used 1000 attempts each, reached the no-key path 366 and 411 times, and produced one and two KASAN reports respectively. With this change, the same reproducer reached the no-key path 366 times in 1000 attempts without a KASAN report or oops. Use tcp_ao_destroy_sock() for the no-key path. It unpublishes the AO info, updates the socket memory and static-key accounting, and defers the free until after an RCU grace period. Also drop the WARN_ON_ONCE() and its stale comment. The VRF detach race makes the no-key state reachable during normal operation, so it is a handled condition rather than an impossible assertion. On panic_on_warn kernels the WARN would turn this handled race into a kernel panic. Fixes: 248411b8cb89 ("net/tcp: Wire up l3index to TCP-AO") Cc: stable@vger.kernel.org Assisted-by: Codex:gpt-5 Signed-off-by: Qing Ming <a0yami@mailbox.org> Reviewed-by: Eric Dumazet <edumazet@google.com> Link: https://patch.msgid.link/20260825072033.6921-1-a0yami@mailbox.org Signed-off-by: Paolo Abeni <pabeni@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> 
commit ea30dc5267e367b8a5e1e06cc074f813bcbf18b2 upstream.

tcp_v4_connect() adds a SYN-SENT socket to the ehash before calling
tcp_connect().  If TCP-AO is configured, tcp_connect() first verifies that
a key matches the peer and the bound device's current L3 master.
tcp_ao_connect_init() later resolves the L3 master again and removes keys
which do not match it.

The socket lock does not stabilize the bound device's VRF membership.
Detaching the device from its VRF between the initial validation and the
L3-master calculation in tcp_ao_connect_init() can therefore make the
validation succeed while initialization observes the default L3 domain and
removes the only key.  The subsequent AO lookup then fails, so the no-key
path clears tp->ao_info and frees it directly.

The receive path can find the socket in the ehash and load tp->ao_info
under RCU before acquiring the socket lock.  A reader which loaded the old
pointer can thus continue into tcp_inbound_ao_hash() after the direct free.

The issue was found during a static audit of TCP-AO object lifetime.  An
unprivileged reproducer in self-created user and network namespaces raced
connect() with detaching a veth from its VRF while sending TCP-AO segments.
It triggered the same KASAN report on two fresh boots:

  BUG: KASAN: slab-use-after-free in tcp_inbound_ao_hash+0x585/0x19f0
  Write of size 8 at addr ffff88800bf88128 by task tcp_ao_vrf_race/232

  Call Trace:
   tcp_inbound_ao_hash+0x585/0x19f0
   tcp_inbound_hash+0x677/0xa80
   tcp_v4_rcv+0x1c3e/0x3ab0

  Allocated by task 235:
   tcp_ao_alloc_info+0x43/0xf0
   tcp_ao_add_cmd+0xdf7/0x13b0
   do_tcp_setsockopt+0x168c/0x2640

  Freed by task 235:
   kfree+0x1b8/0x550
   tcp_connect+0x252/0x4f00
   tcp_v4_connect+0x1114/0x1720

The bad address is 40 bytes inside the freed 128-byte object, matching the
tcp_ao_info counters.key_not_found field.  The two runs used 1000 attempts
each, reached the no-key path 366 and 411 times, and produced one and two
KASAN reports respectively.  With this change, the same reproducer reached
the no-key path 366 times in 1000 attempts without a KASAN report or oops.

Use tcp_ao_destroy_sock() for the no-key path.  It unpublishes the AO info,
updates the socket memory and static-key accounting, and defers the free
until after an RCU grace period.

Also drop the WARN_ON_ONCE() and its stale comment.  The VRF detach race
makes the no-key state reachable during normal operation, so it is a
handled condition rather than an impossible assertion.  On panic_on_warn
kernels the WARN would turn this handled race into a kernel panic.

Fixes: 248411b8cb89 ("net/tcp: Wire up l3index to TCP-AO")
Cc: stable@vger.kernel.org
Assisted-by: Codex:gpt-5
Signed-off-by: Qing Ming <a0yami@mailbox.org>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20260825072033.6921-1-a0yami@mailbox.org
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
net/tcp: fix TCP-AO key deletion in VRFs 2026-09-02T12:33:19+00:00 Rastislav Szabo rastislav.szabo@isovalent.com 2026-08-22T20:11:18+00:00 a742178889f923281268d3c171fb62447847c520 commit 94ad9e114a1c7b16ea418c1456ac3835e038ab3f upstream. TCP-AO keys with TCP_AO_KEYF_IFINDEX store the VRF L3 interface index in l3index. tcp_ao_del_cmd() validates the supplied ifindex, but does not assign it to its local l3index before matching keys. As a result, deleting a key scoped to a non-default VRF always fails with ENOENT because it is matched against l3index 0. Fixes: 248411b8cb89 ("net/tcp: Wire up l3index to TCP-AO") Cc: stable@vger.kernel.org Signed-off-by: Rastislav Szabo <rastislav.szabo@isovalent.com> Reviewed-by: David Ahern <dsahern@kernel.org> Acked-by: Dmitry Safonov <0x7f454c46@gmail.com> Link: https://patch.msgid.link/20260822201119.272269-1-rastislav.szabo@isovalent.com Signed-off-by: Paolo Abeni <pabeni@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> 
commit 94ad9e114a1c7b16ea418c1456ac3835e038ab3f upstream.

TCP-AO keys with TCP_AO_KEYF_IFINDEX store the VRF L3 interface index in
l3index. tcp_ao_del_cmd() validates the supplied ifindex, but does not
assign it to its local l3index before matching keys.

As a result, deleting a key scoped to a non-default VRF always fails with
ENOENT because it is matched against l3index 0.

Fixes: 248411b8cb89 ("net/tcp: Wire up l3index to TCP-AO")
Cc: stable@vger.kernel.org
Signed-off-by: Rastislav Szabo <rastislav.szabo@isovalent.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Acked-by: Dmitry Safonov <0x7f454c46@gmail.com>
Link: https://patch.msgid.link/20260822201119.272269-1-rastislav.szabo@isovalent.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
inet: frags: strip GSO state from fragments before reassembly 2026-08-28T06:25:22+00:00 Xinyang Ge xinyang@anthropic.com 2026-08-27T14:07:07+00:00 da857e448322a2e871ce3ecc2900027041160d43 commit d5dc1e69fd7258ea605c9952e5d5947539159ae3 upstream. A virtio_net_hdr (tun/tap, or AF_PACKET with PACKET_VNET_HDR) can mark an IPv4 or IPv6 fragment as GSO; nothing relates gso_type to frag_off. inet_frag_reasm_prepare()/inet_frag_reasm_finish() keep the first fragment's skb as the head of the reassembled datagram, including its shinfo->gso_size/gso_type/gso_segs, and chain the remaining fragments on frag_list with whatever linear/paged layout they arrived with. After ip_defrag() (ip_local_deliver(), nf_defrag_ipv4, ...) the reassembled skb therefore still claims to be GSO (SKB_GSO_DODGY), and the next software segmentation point - udp_rcv_segment() on local delivery, validate_xmit_skb(), or the ip_finish_output_gso() slow path - hands it to skb_segment(). skb_segment()'s frag_list walk assumes GRO-shaped input and hits one of its BUG_ON()s. Two writes to a tap by an unprivileged user in its own userns are enough: kernel BUG at net/core/skbuff.c:4899! Oops: invalid opcode: 0000 [#1] SMP KASAN NOPTI CPU: 0 UID: 1000 PID: 82 Comm: poc Not tainted 7.2.0-pentest+ #2 RIP: 0010:skb_segment+0x20ca/0x48b0 Call Trace: <TASK> __udp_gso_segment+0x29a/0x27d0 udp4_ufo_fragment+0x458/0x6c0 inet_gso_segment+0x429/0x1340 skb_mac_gso_segment+0x233/0x4f0 __skb_gso_segment+0x308/0x660 udp_queue_rcv_skb+0x440/0xad0 udp_unicast_rcv_skb+0xc7/0x2c0 udp_rcv+0x16ce/0x2260 ip_protocol_deliver_rcu+0x197/0x2d0 ip_local_deliver+0x430/0x690 ip_rcv+0x16f/0x1f0 __netif_receive_skb_one_core+0x15e/0x1c0 __netif_receive_skb+0x1e/0x110 netif_receive_skb+0xf6/0x5c0 tun_rx_batched.isra.0+0x3ab/0x790 tun_get_user+0x17c3/0x3550 tun_chr_write_iter+0xba/0x1b0 vfs_write+0x646/0x1130 </TASK> Kernel panic - not syncing: Fatal exception in interrupt This runs with BH disabled, so it is a panic rather than an oops. The same is reachable with CAP_NET_RAW in a netns where a defrag point precedes a GSO point, and from a guest whose VMM forwards virtio_net_hdr to a tap. The SKB_GSO_DODGY frag_list checks added by commit 3dcbdb134f32 ("net: gso: Fix skb_segment splat when splitting gso_size mangled skb having linear-headed frag_list") and by commit 9e4b7a99a03a ("net: gso: fix panic on frag_list with mixed head alloc types") do not cover it: page-backed heads skip them, and kmalloc heads skip them when gso_size == skb_headlen(head), which the sender controls. An skb entering a frag queue is an IP fragment by definition and cannot legitimately carry GSO state: GRO does not merge fragments and the stack segments before it fragments, so only untrusted sources are affected. This has been reachable since commit f43798c27684 ("tun: Allow GSO using virtio_net_hdr"), the first path that let userspace attach GSO metadata to an IP fragment. Reset the GSO fields of every fragment as it is queued, in inet_frag_queue_insert(), which IPv4, IPv6, nf_conntrack_reasm and 6lowpan reassembly share; then neither the head nor the frag_list members of the reassembled skb carry them (the members matter too: the ip_do_fragment()/ip6_fragment() fast paths send them out as they are). The head may remain CHECKSUM_PARTIAL; that is already accepted on receive and resolved by skb_checksum_help() in ip_do_fragment()/ip6_fragment() on forward. Tested on top of net.git (dc4b95b8fee9), x86_64: the tap reproducer above, two further IPv4 frag_list geometries that reach BUG_ON(i >= nfrags) and BUG_ON(!list_skb->head_frag), and an IPv6 fragment-header variant (udp6_ufo_fragment()) each panic the unpatched kernel; with this patch all four datagrams are delivered intact and nothing is logged. Fixes: f43798c27684 ("tun: Allow GSO using virtio_net_hdr") Cc: stable@kernel.org Suggested-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Xinyang Ge <xinyang@anthropic.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Link: https://patch.msgid.link/937926e509f2acd8e0e66520dc2b30fd6b4d1687.1787839506.git.pabeni@redhat.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> 
commit d5dc1e69fd7258ea605c9952e5d5947539159ae3 upstream.

A virtio_net_hdr (tun/tap, or AF_PACKET with PACKET_VNET_HDR) can mark
an IPv4 or IPv6 fragment as GSO; nothing relates gso_type to frag_off.
inet_frag_reasm_prepare()/inet_frag_reasm_finish() keep the first
fragment's skb as the head of the reassembled datagram, including its
shinfo->gso_size/gso_type/gso_segs, and chain the remaining fragments
on frag_list with whatever linear/paged layout they arrived with.

After ip_defrag() (ip_local_deliver(), nf_defrag_ipv4, ...) the
reassembled skb therefore still claims to be GSO (SKB_GSO_DODGY), and
the next software segmentation point - udp_rcv_segment() on local
delivery, validate_xmit_skb(), or the ip_finish_output_gso() slow
path - hands it to skb_segment(). skb_segment()'s frag_list walk
assumes GRO-shaped input and hits one of its BUG_ON()s. Two writes to
a tap by an unprivileged user in its own userns are enough:

  kernel BUG at net/core/skbuff.c:4899!
  Oops: invalid opcode: 0000 [#1] SMP KASAN NOPTI
  CPU: 0 UID: 1000 PID: 82 Comm: poc Not tainted 7.2.0-pentest+ #2
  RIP: 0010:skb_segment+0x20ca/0x48b0
  Call Trace:
   <TASK>
   __udp_gso_segment+0x29a/0x27d0
   udp4_ufo_fragment+0x458/0x6c0
   inet_gso_segment+0x429/0x1340
   skb_mac_gso_segment+0x233/0x4f0
   __skb_gso_segment+0x308/0x660
   udp_queue_rcv_skb+0x440/0xad0
   udp_unicast_rcv_skb+0xc7/0x2c0
   udp_rcv+0x16ce/0x2260
   ip_protocol_deliver_rcu+0x197/0x2d0
   ip_local_deliver+0x430/0x690
   ip_rcv+0x16f/0x1f0
   __netif_receive_skb_one_core+0x15e/0x1c0
   __netif_receive_skb+0x1e/0x110
   netif_receive_skb+0xf6/0x5c0
   tun_rx_batched.isra.0+0x3ab/0x790
   tun_get_user+0x17c3/0x3550
   tun_chr_write_iter+0xba/0x1b0
   vfs_write+0x646/0x1130
   </TASK>
  Kernel panic - not syncing: Fatal exception in interrupt

This runs with BH disabled, so it is a panic rather than an oops. The
same is reachable with CAP_NET_RAW in a netns where a defrag point
precedes a GSO point, and from a guest whose VMM forwards
virtio_net_hdr to a tap. The SKB_GSO_DODGY frag_list checks added by
commit 3dcbdb134f32 ("net: gso: Fix skb_segment splat when splitting
gso_size mangled skb having linear-headed frag_list") and by
commit 9e4b7a99a03a ("net: gso: fix panic on frag_list with mixed head
alloc types") do not cover it: page-backed heads skip them, and kmalloc
heads skip them when gso_size == skb_headlen(head), which the sender
controls.

An skb entering a frag queue is an IP fragment by definition and
cannot legitimately carry GSO state: GRO does not merge fragments and
the stack segments before it fragments, so only untrusted sources are
affected. This has been reachable since
commit f43798c27684 ("tun: Allow GSO using virtio_net_hdr"), the first
path that let userspace attach GSO metadata to an IP fragment. Reset
the GSO fields of every fragment as it is queued, in
inet_frag_queue_insert(), which IPv4, IPv6, nf_conntrack_reasm and
6lowpan reassembly share; then neither the head nor the frag_list
members of the reassembled skb carry them (the members matter too:
the ip_do_fragment()/ip6_fragment() fast paths send them out as they
are). The head may remain CHECKSUM_PARTIAL; that is already accepted
on receive and resolved by skb_checksum_help() in
ip_do_fragment()/ip6_fragment() on forward.

Tested on top of net.git (dc4b95b8fee9), x86_64: the tap reproducer
above, two further IPv4 frag_list geometries that reach
BUG_ON(i >= nfrags) and BUG_ON(!list_skb->head_frag), and an IPv6
fragment-header variant (udp6_ufo_fragment()) each panic the unpatched
kernel; with this patch all four datagrams are delivered intact and
nothing is logged.

Fixes: f43798c27684 ("tun: Allow GSO using virtio_net_hdr")
Cc: stable@kernel.org
Suggested-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Xinyang Ge <xinyang@anthropic.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/937926e509f2acd8e0e66520dc2b30fd6b4d1687.1787839506.git.pabeni@redhat.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
ipv4: reject undersized MTUs in ip_do_fragment() 2026-08-27T12:35:24+00:00 Yong Wang edragain@163.com 2026-08-13T17:35:26+00:00 c8a74adccaf028223054633b532593101dfcc581 commit c0726f0caf8c6b3208552949e17d23634a2f3129 upstream. ip_do_fragment() subtracts the IPv4 header length from the effective MTU and passes the resulting payload MTU to ip_frag_next(). If the effective MTU is smaller than hlen + 8, ip_frag_next() rounds the fragment payload length down to zero. The fragmentation state then never makes forward progress: state->left, state->ptr and state->offset stay unchanged while ip_do_fragment() keeps allocating and transmitting header-only fragments until the softlockup detector fires. This is reproducible with a route installed using "mtu lock 20", but it is also reproducible without route MTU lock, for example by forwarding a packet to a device whose MTU is 20. Fix it in ip_do_fragment() by rejecting mtu < hlen + 8 with -EMSGSIZE, matching the existing IPv6 fragmentation check. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Cc: stable@vger.kernel.org Reported-by: Vega <vega@nebusec.ai> Signed-off-by: Yong Wang <edragain@163.com> Signed-off-by: Ren Wei <weir@nebusec.ai> Reviewed-by: Ido Schimmel <idosch@nvidia.com> Link: https://patch.msgid.link/8809ef6314b98913681b0b370a05a85c2b6cd579.1786599079.git.edragain@163.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> 
commit c0726f0caf8c6b3208552949e17d23634a2f3129 upstream.

ip_do_fragment() subtracts the IPv4 header length from the effective
MTU and passes the resulting payload MTU to ip_frag_next().

If the effective MTU is smaller than hlen + 8, ip_frag_next() rounds
the fragment payload length down to zero. The fragmentation state then
never makes forward progress: state->left, state->ptr and state->offset
stay unchanged while ip_do_fragment() keeps allocating and transmitting
header-only fragments until the softlockup detector fires.

This is reproducible with a route installed using "mtu lock 20", but it
is also reproducible without route MTU lock, for example by forwarding a
packet to a device whose MTU is 20.

Fix it in ip_do_fragment() by rejecting mtu < hlen + 8 with -EMSGSIZE,
matching the existing IPv6 fragmentation check.

Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
Cc: stable@vger.kernel.org
Reported-by: Vega <vega@nebusec.ai>
Signed-off-by: Yong Wang <edragain@163.com>
Signed-off-by: Ren Wei <weir@nebusec.ai>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Link: https://patch.msgid.link/8809ef6314b98913681b0b370a05a85c2b6cd579.1786599079.git.edragain@163.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>