[Upstream commit 7fb1291257]
Commit 6693731487 ("vsock/virtio: Allocate nonlinear SKBs for handling
large transmit buffers") converted the virtio vsock transmit path to
utilise nonlinear SKBs when handling large buffers. As part of this
change, virtio_transport_fill_skb() was updated to call
skb_copy_datagram_from_iter() instead of memcpy_from_msg() as the latter
expects a single destination buffer and cannot handle nonlinear SKBs
correctly.
Unfortunately, during this conversion, I overlooked the error case when
the copying function returns -EFAULT due to a fault on the input buffer
in userspace. In this case, memcpy_from_msg() reverts the iterator to
its initial state thanks to copy_from_iter_full() whereas
skb_copy_datagram_from_iter() leaves the iterator partially advanced.
This results in a WARN_ONCE() from the vsock code, which expects the
iterator to stay in sync with the number of bytes transmitted so that
virtio_transport_send_pkt_info() can return -EFAULT when it is called
again:
------------[ cut here ]------------
'send_pkt()' returns 0, but 65536 expected
WARNING: CPU: 0 PID: 5503 at net/vmw_vsock/virtio_transport_common.c:428 virtio_transport_send_pkt_info+0xd11/0xf00 net/vmw_vsock/virtio_transport_common.c:426
Modules linked in:
CPU: 0 UID: 0 PID: 5503 Comm: syz.0.17 Not tainted 6.16.0-syzkaller-12063-g37816488247d #0 PREEMPT(full)
Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.16.3-debian-1.16.3-2~bpo12+1 04/01/2014
Call virtio_transport_fill_skb_full() to restore the previous iterator
behaviour.
Cc: Jason Wang <jasowang@redhat.com>
Cc: Stefano Garzarella <sgarzare@redhat.com>
Fixes: 6693731487 ("vsock/virtio: Allocate nonlinear SKBs for handling large transmit buffers")
Reported-by: syzbot+b4d960daf7a3c7c2b7b1@syzkaller.appspotmail.com
Signed-off-by: Will Deacon <will@kernel.org>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Link: https://patch.msgid.link/20250818180355.29275-3-will@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
[halves: adjust __zerocopy_sg_from_iter() parameters]
Signed-off-by: Heitor Alves de Siqueira <halves@igalia.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[Upstream commit b08a784a5d]
In a similar manner to copy_from_iter()/copy_from_iter_full(), introduce
skb_copy_datagram_from_iter_full() which reverts the iterator to its
initial state when returning an error.
A subsequent fix for a vsock regression will make use of this new
function.
Cc: Christian Brauner <brauner@kernel.org>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Will Deacon <will@kernel.org>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Link: https://patch.msgid.link/20250818180355.29275-2-will@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Heitor Alves de Siqueira <halves@igalia.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[Upstream commit 6693731487]
When transmitting a vsock packet, virtio_transport_send_pkt_info() calls
virtio_transport_alloc_linear_skb() to allocate and fill SKBs with the
transmit data. Unfortunately, these are always linear allocations and
can therefore result in significant pressure on kmalloc() considering
that the maximum packet size (VIRTIO_VSOCK_MAX_PKT_BUF_SIZE +
VIRTIO_VSOCK_SKB_HEADROOM) is a little over 64KiB, resulting in a 128KiB
allocation for each packet.
Rework the vsock SKB allocation so that, for sizes with page order
greater than PAGE_ALLOC_COSTLY_ORDER, a nonlinear SKB is allocated
instead with the packet header in the SKB and the transmit data in the
fragments. Note that this affects both the vhost and virtio transports.
Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>
Signed-off-by: Will Deacon <will@kernel.org>
Message-Id: <20250717090116.11987-10-will@kernel.org>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Heitor Alves de Siqueira <halves@igalia.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[Upstream commit 8ca76151d2]
In preparation for using virtio_vsock_skb_rx_put() when populating SKBs
on the vsock TX path, rename virtio_vsock_skb_rx_put() to
virtio_vsock_skb_put().
No functional change.
Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>
Signed-off-by: Will Deacon <will@kernel.org>
Message-Id: <20250717090116.11987-9-will@kernel.org>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Heitor Alves de Siqueira <halves@igalia.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[Upstream commit 2304c64a28]
In preparation for nonlinear allocations for large SKBs, rename
virtio_vsock_alloc_skb() to virtio_vsock_alloc_linear_skb() to indicate
that it returns linear SKBs unconditionally and switch all callers over
to this new interface for now.
No functional change.
Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>
Signed-off-by: Will Deacon <will@kernel.org>
Message-Id: <20250717090116.11987-6-will@kernel.org>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Heitor Alves de Siqueira <halves@igalia.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[Upstream commit 87dbae5e36]
virtio_vsock_skb_rx_put() only calls skb_put() if the length in the
packet header is not zero even though skb_put() handles this case
gracefully.
Remove the functionally redundant check from virtio_vsock_skb_rx_put()
and, on the assumption that this is a worthwhile optimisation for
handling credit messages, augment the existing length checks in
virtio_transport_rx_work() to elide the call for zero-length payloads.
Since the callers all have the length, extend virtio_vsock_skb_rx_put()
to take it as an additional parameter rather than fish it back out of
the packet header.
Note that the vhost code already has similar logic in
vhost_vsock_alloc_skb().
Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>
Signed-off-by: Will Deacon <will@kernel.org>
Message-Id: <20250717090116.11987-4-will@kernel.org>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Heitor Alves de Siqueira <halves@igalia.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 04a899573f upstream.
Yinhao et al. reported that their fuzzer tool was able to trigger a
skb_warn_bad_offload() from netif_skb_features() -> gso_features_check().
When a BPF program - triggered via BPF test infra - pushes the packet
to the loopback device via bpf_clone_redirect() then mentioned offload
warning can be seen. GSO-related features are then rightfully disabled.
We get into this situation due to convert___skb_to_skb() setting
gso_segs and gso_size but not gso_type. Technically, it makes sense
that this warning triggers since the GSO properties are malformed due
to the gso_type. Potentially, the gso_type could be marked non-trustworthy
through setting it at least to SKB_GSO_DODGY without any other specific
assumptions, but that also feels wrong given we should not go further
into the GSO engine in the first place.
The checks were added in 121d57af30 ("gso: validate gso_type in GSO
handlers") because there were malicious (syzbot) senders that combine
a protocol with a non-matching gso_type. If we would want to drop such
packets, gso_features_check() currently only returns feature flags via
netif_skb_features(), so one location for potentially dropping such skbs
could be validate_xmit_unreadable_skb(), but then otoh it would be
an additional check in the fast-path for a very corner case. Given
bpf_clone_redirect() is the only place where BPF test infra could emit
such packets, lets reject them right there.
Fixes: 850a88cc40 ("bpf: Expose __sk_buff wire_len/gso_segs to BPF_PROG_TEST_RUN")
Fixes: cf62089b0e ("bpf: Add gso_size to __sk_buff")
Reported-by: Yinhao Hu <dddddd@hust.edu.cn>
Reported-by: Kaiyan Mei <M202472210@hust.edu.cn>
Reported-by: Dongliang Mu <dzm91@hust.edu.cn>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
Link: https://patch.msgid.link/20251020075441.127980-1-daniel@iogearbox.net
Signed-off-by: Shung-Hsi Yu <shung-hsi.yu@suse.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 8ee784fdf0 ]
The virtio transports derives its TX credit directly from peer_buf_alloc,
which is set from the remote endpoint's SO_VM_SOCKETS_BUFFER_SIZE value.
On the host side this means that the amount of data we are willing to
queue for a connection is scaled by a guest-chosen buffer size, rather
than the host's own vsock configuration. A malicious guest can advertise
a large buffer and read slowly, causing the host to allocate a
correspondingly large amount of sk_buff memory.
The same thing would happen in the guest with a malicious host, since
virtio transports share the same code base.
Introduce a small helper, virtio_transport_tx_buf_size(), that
returns min(peer_buf_alloc, buf_alloc), and use it wherever we consume
peer_buf_alloc.
This ensures the effective TX window is bounded by both the peer's
advertised buffer and our own buf_alloc (already clamped to
buffer_max_size via SO_VM_SOCKETS_BUFFER_MAX_SIZE), so a remote peer
cannot force the other to queue more data than allowed by its own
vsock settings.
On an unpatched Ubuntu 22.04 host (~64 GiB RAM), running a PoC with
32 guest vsock connections advertising 2 GiB each and reading slowly
drove Slab/SUnreclaim from ~0.5 GiB to ~57 GiB; the system only
recovered after killing the QEMU process. That said, if QEMU memory is
limited with cgroups, the maximum memory used will be limited.
With this patch applied:
Before:
MemFree: ~61.6 GiB
Slab: ~142 MiB
SUnreclaim: ~117 MiB
After 32 high-credit connections:
MemFree: ~61.5 GiB
Slab: ~178 MiB
SUnreclaim: ~152 MiB
Only ~35 MiB increase in Slab/SUnreclaim, no host OOM, and the guest
remains responsive.
Compatibility with non-virtio transports:
- VMCI uses the AF_VSOCK buffer knobs to size its queue pairs per
socket based on the local vsk->buffer_* values; the remote side
cannot enlarge those queues beyond what the local endpoint
configured.
- Hyper-V's vsock transport uses fixed-size VMBus ring buffers and
an MTU bound; there is no peer-controlled credit field comparable
to peer_buf_alloc, and the remote endpoint cannot drive in-flight
kernel memory above those ring sizes.
- The loopback path reuses virtio_transport_common.c, so it
naturally follows the same semantics as the virtio transport.
This change is limited to virtio_transport_common.c and thus affects
virtio-vsock, vhost-vsock, and loopback, bringing them in line with the
"remote window intersected with local policy" behaviour that VMCI and
Hyper-V already effectively have.
Fixes: 06a8fc7836 ("VSOCK: Introduce virtio_vsock_common.ko")
Suggested-by: Stefano Garzarella <sgarzare@redhat.com>
Signed-off-by: Melbin K Mathew <mlbnkm1@gmail.com>
[Stefano: small adjustments after changing the previous patch]
[Stefano: tweak the commit message]
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Reviewed-by: Luigi Leonardi <leonardi@redhat.com>
Link: https://patch.msgid.link/20260121093628.9941-4-sgarzare@redhat.com
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 3ef3d52a1a ]
The credit calculation in virtio_transport_get_credit() uses unsigned
arithmetic:
ret = vvs->peer_buf_alloc - (vvs->tx_cnt - vvs->peer_fwd_cnt);
If the peer shrinks its advertised buffer (peer_buf_alloc) while bytes
are in flight, the subtraction can underflow and produce a large
positive value, potentially allowing more data to be queued than the
peer can handle.
Reuse virtio_transport_has_space() which already handles this case and
add a comment to make it clear why we are doing that.
Fixes: 06a8fc7836 ("VSOCK: Introduce virtio_vsock_common.ko")
Suggested-by: Stefano Garzarella <sgarzare@redhat.com>
Signed-off-by: Melbin K Mathew <mlbnkm1@gmail.com>
[Stefano: use virtio_transport_has_space() instead of duplicating the code]
[Stefano: tweak the commit message]
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Reviewed-by: Luigi Leonardi <leonardi@redhat.com>
Link: https://patch.msgid.link/20260121093628.9941-2-sgarzare@redhat.com
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit cc4816bdb0 ]
In ovs_vport_get_upcall_stats(), some statistics protected by
u64_stats_sync, are read and accumulated in ignorance of possible
u64_stats_fetch_retry() events. These statistics are already accumulated
by u64_stats_inc(). Fix this by reading them into temporary variables
first.
Fixes: 1933ea365a ("net: openvswitch: Add support to count upcall packets")
Signed-off-by: David Yang <mmyangfl@gmail.com>
Acked-by: Ilya Maximets <i.maximets@ovn.org>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Aaron Conole <aconole@redhat.com>
Link: https://patch.msgid.link/20260121072932.2360971-1-mmyangfl@gmail.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit dfca045cd4 ]
Prior to the blamed commit, the bridge_num range was from
0 to ds->max_num_bridges - 1. After the commit, it is from
1 to ds->max_num_bridges.
So this check:
if (bridge_num >= max)
return 0;
must be updated to:
if (bridge_num > max)
return 0;
in order to allow the last bridge_num value (==max) to be used.
This is easiest visible when a driver sets ds->max_num_bridges=1.
The observed behaviour is that even the first created bridge triggers
the netlink extack "Range of offloadable bridges exceeded" warning, and
is handled in software rather than being offloaded.
Fixes: 3f9bb0301d ("net: dsa: make dp->bridge_num one-based")
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Link: https://patch.msgid.link/20260120211039.3228999-1-vladimir.oltean@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 9a063f96d8 ]
syzbot found that ndisc_router_discovery() could read and write
in6_dev->ra_mtu without holding a lock [1]
This looks fine, IFLA_INET6_RA_MTU is best effort.
Add READ_ONCE()/WRITE_ONCE() to document the race.
Note that we might also reject illegal MTU values
(mtu < IPV6_MIN_MTU || mtu > skb->dev->mtu) in a future patch.
[1]
BUG: KCSAN: data-race in ndisc_router_discovery / ndisc_router_discovery
read to 0xffff888119809c20 of 4 bytes by task 25817 on cpu 1:
ndisc_router_discovery+0x151d/0x1c90 net/ipv6/ndisc.c:1558
ndisc_rcv+0x2ad/0x3d0 net/ipv6/ndisc.c:1841
icmpv6_rcv+0xe5a/0x12f0 net/ipv6/icmp.c:989
ip6_protocol_deliver_rcu+0xb2a/0x10d0 net/ipv6/ip6_input.c:438
ip6_input_finish+0xf0/0x1d0 net/ipv6/ip6_input.c:489
NF_HOOK include/linux/netfilter.h:318 [inline]
ip6_input+0x5e/0x140 net/ipv6/ip6_input.c:500
ip6_mc_input+0x27c/0x470 net/ipv6/ip6_input.c:590
dst_input include/net/dst.h:474 [inline]
ip6_rcv_finish+0x336/0x340 net/ipv6/ip6_input.c:79
...
write to 0xffff888119809c20 of 4 bytes by task 25816 on cpu 0:
ndisc_router_discovery+0x155a/0x1c90 net/ipv6/ndisc.c:1559
ndisc_rcv+0x2ad/0x3d0 net/ipv6/ndisc.c:1841
icmpv6_rcv+0xe5a/0x12f0 net/ipv6/icmp.c:989
ip6_protocol_deliver_rcu+0xb2a/0x10d0 net/ipv6/ip6_input.c:438
ip6_input_finish+0xf0/0x1d0 net/ipv6/ip6_input.c:489
NF_HOOK include/linux/netfilter.h:318 [inline]
ip6_input+0x5e/0x140 net/ipv6/ip6_input.c:500
ip6_mc_input+0x27c/0x470 net/ipv6/ip6_input.c:590
dst_input include/net/dst.h:474 [inline]
ip6_rcv_finish+0x336/0x340 net/ipv6/ip6_input.c:79
...
value changed: 0x00000000 -> 0xe5400659
Fixes: 49b99da2c9 ("ipv6: add IFLA_INET6_RA_MTU to expose mtu value")
Reported-by: syzbot <syzkaller@googlegroups.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Rocco Yue <rocco.yue@mediatek.com>
Link: https://patch.msgid.link/20260118152941.2563857-1-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 5dc6975566 ]
S1G beacons don't contain the DA field as per IEEE80211-2024 9.3.4.3,
so the DA broadcast check reads the SA address of the S1G beacon which
will subsequently lead to the beacon being dropped. As a result, passive
scanning is not possible. Fix this by only performing the check on
non-S1G beacons to allow S1G long beacons to be processed during a
passive scan.
Fixes: ddf82e752f ("wifi: mac80211: Allow beacons to update BSS table regardless of scan")
Signed-off-by: Lachlan Hodges <lachlan.hodges@morsemicro.com>
Link: https://patch.msgid.link/20260120031122.309942-1-lachlan.hodges@morsemicro.com
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit d837fbee92 ]
This is more of a preventive patch to make the code more consistent and
to prevent possible exploits that employ child qlen manipulations on qfq.
use cl_is_active instead of relying on the child qdisc's qlen to determine
class activation.
Fixes: 462dbc9101 ("pkt_sched: QFQ Plus: fair-queueing service at DRR cost")
Signed-off-by: Jamal Hadi Salim <jhs@mojatatu.com>
Link: https://patch.msgid.link/20260114160243.913069-3-jhs@mojatatu.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 50da4b9d07 ]
Design intent of teql is that it is only supposed to be used as root qdisc.
We need to check for that constraint.
Although not important, I will describe the scenario that unearthed this
issue for the curious.
GangMin Kim <km.kim1503@gmail.com> managed to concot a scenario as follows:
ROOT qdisc 1:0 (QFQ)
├── class 1:1 (weight=15, lmax=16384) netem with delay 6.4s
└── class 1:2 (weight=1, lmax=1514) teql
GangMin sends a packet which is enqueued to 1:1 (netem).
Any invocation of dequeue by QFQ from this class will not return a packet
until after 6.4s. In the meantime, a second packet is sent and it lands on
1:2. teql's enqueue will return success and this will activate class 1:2.
Main issue is that teql only updates the parent visible qlen (sch->q.qlen)
at dequeue. Since QFQ will only call dequeue if peek succeeds (and teql's
peek always returns NULL), dequeue will never be called and thus the qlen
will remain as 0. With that in mind, when GangMin updates 1:2's lmax value,
the qfq_change_class calls qfq_deact_rm_from_agg. Since the child qdisc's
qlen was not incremented, qfq fails to deactivate the class, but still
frees its pointers from the aggregate. So when the first packet is
rescheduled after 6.4 seconds (netem's delay), a dangling pointer is
accessed causing GangMin's causing a UAF.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Reported-by: GangMin Kim <km.kim1503@gmail.com>
Tested-by: Victor Nogueira <victor@mojatatu.com>
Signed-off-by: Jamal Hadi Salim <jhs@mojatatu.com>
Link: https://patch.msgid.link/20260114160243.913069-2-jhs@mojatatu.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 7a9bc9e3f4 ]
fou_udp_recv() has the same problem mentioned in the previous
patch.
If FOU_ATTR_IPPROTO is set to 0, skb is not freed by
fou_udp_recv() nor "resubmit"-ted in ip_protocol_deliver_rcu().
Let's forbid 0 for FOU_ATTR_IPPROTO.
Fixes: 23461551c0 ("fou: Support for foo-over-udp RX path")
Signed-off-by: Kuniyuki Iwashima <kuniyu@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20260115172533.693652-4-kuniyu@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit a80c9d945a ]
A null-ptr-deref was reported in the SCTP transmit path when SCTP-AUTH key
initialization fails:
==================================================================
KASAN: null-ptr-deref in range [0x0000000000000018-0x000000000000001f]
CPU: 0 PID: 16 Comm: ksoftirqd/0 Tainted: G W 6.6.0 #2
RIP: 0010:sctp_packet_bundle_auth net/sctp/output.c:264 [inline]
RIP: 0010:sctp_packet_append_chunk+0xb36/0x1260 net/sctp/output.c:401
Call Trace:
sctp_packet_transmit_chunk+0x31/0x250 net/sctp/output.c:189
sctp_outq_flush_data+0xa29/0x26d0 net/sctp/outqueue.c:1111
sctp_outq_flush+0xc80/0x1240 net/sctp/outqueue.c:1217
sctp_cmd_interpreter.isra.0+0x19a5/0x62c0 net/sctp/sm_sideeffect.c:1787
sctp_side_effects net/sctp/sm_sideeffect.c:1198 [inline]
sctp_do_sm+0x1a3/0x670 net/sctp/sm_sideeffect.c:1169
sctp_assoc_bh_rcv+0x33e/0x640 net/sctp/associola.c:1052
sctp_inq_push+0x1dd/0x280 net/sctp/inqueue.c:88
sctp_rcv+0x11ae/0x3100 net/sctp/input.c:243
sctp6_rcv+0x3d/0x60 net/sctp/ipv6.c:1127
The issue is triggered when sctp_auth_asoc_init_active_key() fails in
sctp_sf_do_5_1C_ack() while processing an INIT_ACK. In this case, the
command sequence is currently:
- SCTP_CMD_PEER_INIT
- SCTP_CMD_TIMER_STOP (T1_INIT)
- SCTP_CMD_TIMER_START (T1_COOKIE)
- SCTP_CMD_NEW_STATE (COOKIE_ECHOED)
- SCTP_CMD_ASSOC_SHKEY
- SCTP_CMD_GEN_COOKIE_ECHO
If SCTP_CMD_ASSOC_SHKEY fails, asoc->shkey remains NULL, while
asoc->peer.auth_capable and asoc->peer.peer_chunks have already been set by
SCTP_CMD_PEER_INIT. This allows a DATA chunk with auth = 1 and shkey = NULL
to be queued by sctp_datamsg_from_user().
Since command interpretation stops on failure, no COOKIE_ECHO should been
sent via SCTP_CMD_GEN_COOKIE_ECHO. However, the T1_COOKIE timer has already
been started, and it may enqueue a COOKIE_ECHO into the outqueue later. As
a result, the DATA chunk can be transmitted together with the COOKIE_ECHO
in sctp_outq_flush_data(), leading to the observed issue.
Similar to the other places where it calls sctp_auth_asoc_init_active_key()
right after sctp_process_init(), this patch moves the SCTP_CMD_ASSOC_SHKEY
immediately after SCTP_CMD_PEER_INIT, before stopping T1_INIT and starting
T1_COOKIE. This ensures that if shared key generation fails, authenticated
DATA cannot be sent. It also allows the T1_INIT timer to retransmit INIT,
giving the client another chance to process INIT_ACK and retry key setup.
Fixes: 730fc3d05c ("[SCTP]: Implete SCTP-AUTH parameter processing")
Reported-by: Zhen Chen <chenzhen126@huawei.com>
Tested-by: Zhen Chen <chenzhen126@huawei.com>
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Link: https://patch.msgid.link/44881224b375aa8853f5e19b4055a1a56d895813.1768324226.git.lucien.xin@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 0386bd321d ]
vsock/virtio common tries to coalesce buffers in rx queue: if a linear skb
(with a spare tail room) is followed by a small skb (length limited by
GOOD_COPY_LEN = 128), an attempt is made to join them.
Since the introduction of MSG_ZEROCOPY support, assumption that a small skb
will always be linear is incorrect. In the zerocopy case, data is lost and
the linear skb is appended with uninitialized kernel memory.
Of all 3 supported virtio-based transports, only loopback-transport is
affected. G2H virtio-transport rx queue operates on explicitly linear skbs;
see virtio_vsock_alloc_linear_skb() in virtio_vsock_rx_fill(). H2G
vhost-transport may allocate non-linear skbs, but only for sizes that are
not considered for coalescence; see PAGE_ALLOC_COSTLY_ORDER in
virtio_vsock_alloc_skb().
Ensure only linear skbs are coalesced. Note that skb_tailroom(last_skb) > 0
guarantees last_skb is linear.
Fixes: 581512a6dc ("vsock/virtio: MSG_ZEROCOPY flag support")
Signed-off-by: Michal Luczaj <mhal@rbox.co>
Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>
Link: https://patch.msgid.link/20260113-vsock-recv-coalescence-v2-1-552b17837cf4@rbox.co
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit e09299225d upstream.
The following BPF program, simplified from a syzkaller repro, causes a
kernel warning:
r0 = *(u8 *)(r1 + 169);
exit;
With pointer field sk being at offset 168 in __sk_buff. This access is
detected as a narrower read in bpf_skb_is_valid_access because it
doesn't match offsetof(struct __sk_buff, sk). It is therefore allowed
and later proceeds to bpf_convert_ctx_access. Note that for the
"is_narrower_load" case in the convert_ctx_accesses(), the insn->off
is aligned, so the cnt may not be 0 because it matches the
offsetof(struct __sk_buff, sk) in the bpf_convert_ctx_access. However,
the target_size stays 0 and the verifier errors with a kernel warning:
verifier bug: error during ctx access conversion(1)
This patch fixes that to return a proper "invalid bpf_context access
off=X size=Y" error on the load instruction.
The same issue affects multiple other fields in context structures that
allow narrow access. Some other non-affected fields (for sk_msg,
sk_lookup, and sockopt) were also changed to use bpf_ctx_range_ptr for
consistency.
Note this syzkaller crash was reported in the "Closes" link below, which
used to be about a different bug, fixed in
commit fce7bd8e38 ("bpf/verifier: Handle BPF_LOAD_ACQ instructions
in insn_def_regno()"). Because syzbot somehow confused the two bugs,
the new crash and repro didn't get reported to the mailing list.
Fixes: f96da09473 ("bpf: simplify narrower ctx access")
Fixes: 0df1a55afa ("bpf: Warn on internal verifier errors")
Reported-by: syzbot+0ef84a7bdf5301d4cbec@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=0ef84a7bdf5301d4cbec
Signed-off-by: Paul Chaignon <paul.chaignon@gmail.com>
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
Acked-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://patch.msgid.link/3b8dcee67ff4296903351a974ddd9c4dca768b64.1753194596.git.paul.chaignon@gmail.com
Signed-off-by: Shung-Hsi Yu <shung-hsi.yu@suse.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 1809c82aa0 upstream.
Since j1939_session_deactivate_activate_next() in j1939_tp_rxtimer() is
called only when the timer is enabled, we need to call
j1939_session_deactivate_activate_next() if we cancelled the timer.
Otherwise, refcount for j1939_session leaks, which will later appear as
| unregister_netdevice: waiting for vcan0 to become free. Usage count = 2.
problem.
Reported-by: syzbot <syzbot+881d65229ca4f9ae8c84@syzkaller.appspotmail.com>
Closes: https://syzkaller.appspot.com/bug?extid=881d65229ca4f9ae8c84
Signed-off-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Tested-by: Oleksij Rempel <o.rempel@pengutronix.de>
Acked-by: Oleksij Rempel <o.rempel@pengutronix.de>
Fixes: 9d71dd0c70 ("can: add support of SAE J1939 protocol")
Link: https://patch.msgid.link/b1212653-8fa1-44e1-be9d-12f950fb3a07@I-love.SAKURA.ne.jp
Cc: stable@vger.kernel.org
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 7544f3f5b0 upstream.
The bridge maintains a global list of ports behind which a multicast
router resides. The list is consulted during forwarding to ensure
multicast packets are forwarded to these ports even if the ports are not
member in the matching MDB entry.
When per-VLAN multicast snooping is enabled, the per-port multicast
context is disabled on each port and the port is removed from the global
router port list:
# ip link add name br1 up type bridge vlan_filtering 1 mcast_snooping 1
# ip link add name dummy1 up master br1 type dummy
# ip link set dev dummy1 type bridge_slave mcast_router 2
$ bridge -d mdb show | grep router
router ports on br1: dummy1
# ip link set dev br1 type bridge mcast_vlan_snooping 1
$ bridge -d mdb show | grep router
However, the port can be re-added to the global list even when per-VLAN
multicast snooping is enabled:
# ip link set dev dummy1 type bridge_slave mcast_router 0
# ip link set dev dummy1 type bridge_slave mcast_router 2
$ bridge -d mdb show | grep router
router ports on br1: dummy1
Since commit 4b30ae9adb ("net: bridge: mcast: re-implement
br_multicast_{enable, disable}_port functions"), when per-VLAN multicast
snooping is enabled, multicast disablement on a port will disable the
per-{port, VLAN} multicast contexts and not the per-port one. As a
result, a port will remain in the global router port list even after it
is deleted. This will lead to a use-after-free [1] when the list is
traversed (when adding a new port to the list, for example):
# ip link del dev dummy1
# ip link add name dummy2 up master br1 type dummy
# ip link set dev dummy2 type bridge_slave mcast_router 2
Similarly, stale entries can also be found in the per-VLAN router port
list. When per-VLAN multicast snooping is disabled, the per-{port, VLAN}
contexts are disabled on each port and the port is removed from the
per-VLAN router port list:
# ip link add name br1 up type bridge vlan_filtering 1 mcast_snooping 1 mcast_vlan_snooping 1
# ip link add name dummy1 up master br1 type dummy
# bridge vlan add vid 2 dev dummy1
# bridge vlan global set vid 2 dev br1 mcast_snooping 1
# bridge vlan set vid 2 dev dummy1 mcast_router 2
$ bridge vlan global show dev br1 vid 2 | grep router
router ports: dummy1
# ip link set dev br1 type bridge mcast_vlan_snooping 0
$ bridge vlan global show dev br1 vid 2 | grep router
However, the port can be re-added to the per-VLAN list even when
per-VLAN multicast snooping is disabled:
# bridge vlan set vid 2 dev dummy1 mcast_router 0
# bridge vlan set vid 2 dev dummy1 mcast_router 2
$ bridge vlan global show dev br1 vid 2 | grep router
router ports: dummy1
When the VLAN is deleted from the port, the per-{port, VLAN} multicast
context will not be disabled since multicast snooping is not enabled
on the VLAN. As a result, the port will remain in the per-VLAN router
port list even after it is no longer member in the VLAN. This will lead
to a use-after-free [2] when the list is traversed (when adding a new
port to the list, for example):
# ip link add name dummy2 up master br1 type dummy
# bridge vlan add vid 2 dev dummy2
# bridge vlan del vid 2 dev dummy1
# bridge vlan set vid 2 dev dummy2 mcast_router 2
Fix these issues by removing the port from the relevant (global or
per-VLAN) router port list in br_multicast_port_ctx_deinit(). The
function is invoked during port deletion with the per-port multicast
context and during VLAN deletion with the per-{port, VLAN} multicast
context.
Note that deleting the multicast router timer is not enough as it only
takes care of the temporary multicast router states (1 or 3) and not the
permanent one (2).
[1]
BUG: KASAN: slab-out-of-bounds in br_multicast_add_router.part.0+0x3f1/0x560
Write of size 8 at addr ffff888004a67328 by task ip/384
[...]
Call Trace:
<TASK>
dump_stack_lvl+0x6f/0xa0
print_address_description.constprop.0+0x6f/0x350
print_report+0x108/0x205
kasan_report+0xdf/0x110
br_multicast_add_router.part.0+0x3f1/0x560
br_multicast_set_port_router+0x74e/0xac0
br_setport+0xa55/0x1870
br_port_slave_changelink+0x95/0x120
__rtnl_newlink+0x5e8/0xa40
rtnl_newlink+0x627/0xb00
rtnetlink_rcv_msg+0x6fb/0xb70
netlink_rcv_skb+0x11f/0x350
netlink_unicast+0x426/0x710
netlink_sendmsg+0x75a/0xc20
__sock_sendmsg+0xc1/0x150
____sys_sendmsg+0x5aa/0x7b0
___sys_sendmsg+0xfc/0x180
__sys_sendmsg+0x124/0x1c0
do_syscall_64+0xbb/0x360
entry_SYSCALL_64_after_hwframe+0x4b/0x53
[2]
BUG: KASAN: slab-use-after-free in br_multicast_add_router.part.0+0x378/0x560
Read of size 8 at addr ffff888009f00840 by task bridge/391
[...]
Call Trace:
<TASK>
dump_stack_lvl+0x6f/0xa0
print_address_description.constprop.0+0x6f/0x350
print_report+0x108/0x205
kasan_report+0xdf/0x110
br_multicast_add_router.part.0+0x378/0x560
br_multicast_set_port_router+0x6f9/0xac0
br_vlan_process_options+0x8b6/0x1430
br_vlan_rtm_process_one+0x605/0xa30
br_vlan_rtm_process+0x396/0x4c0
rtnetlink_rcv_msg+0x2f7/0xb70
netlink_rcv_skb+0x11f/0x350
netlink_unicast+0x426/0x710
netlink_sendmsg+0x75a/0xc20
__sock_sendmsg+0xc1/0x150
____sys_sendmsg+0x5aa/0x7b0
___sys_sendmsg+0xfc/0x180
__sys_sendmsg+0x124/0x1c0
do_syscall_64+0xbb/0x360
entry_SYSCALL_64_after_hwframe+0x4b/0x53
Fixes: 2796d846d7 ("net: bridge: vlan: convert mcast router global option to per-vlan entry")
Fixes: 4b30ae9adb ("net: bridge: mcast: re-implement br_multicast_{enable, disable}_port functions")
Reported-by: syzbot+7bfa4b72c6a5da128d32@syzkaller.appspotmail.com
Closes: https://lore.kernel.org/all/684c18bd.a00a0220.279073.000b.GAE@google.com/T/
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Link: https://patch.msgid.link/20250619182228.1656906-1-idosch@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Lee Jones <lee@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit e67c577d89 ]
Analog to commit db5b4e39c4 ("ip6_gre: make ip6gre_header() robust")
Over the years, syzbot found many ways to crash the kernel
in ipgre_header() [1].
This involves team or bonding drivers ability to dynamically
change their dev->needed_headroom and/or dev->hard_header_len
In this particular crash mld_newpack() allocated an skb
with a too small reserve/headroom, and by the time mld_sendpack()
was called, syzbot managed to attach an ipgre device.
[1]
skbuff: skb_under_panic: text:ffffffff89ea3cb7 len:2030915468 put:2030915372 head:ffff888058b43000 data:ffff887fdfa6e194 tail:0x120 end:0x6c0 dev:team0
kernel BUG at net/core/skbuff.c:213 !
Oops: invalid opcode: 0000 [#1] SMP KASAN PTI
CPU: 1 UID: 0 PID: 1322 Comm: kworker/1:9 Not tainted syzkaller #0 PREEMPT(full)
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 10/25/2025
Workqueue: mld mld_ifc_work
RIP: 0010:skb_panic+0x157/0x160 net/core/skbuff.c:213
Call Trace:
<TASK>
skb_under_panic net/core/skbuff.c:223 [inline]
skb_push+0xc3/0xe0 net/core/skbuff.c:2641
ipgre_header+0x67/0x290 net/ipv4/ip_gre.c:897
dev_hard_header include/linux/netdevice.h:3436 [inline]
neigh_connected_output+0x286/0x460 net/core/neighbour.c:1618
NF_HOOK_COND include/linux/netfilter.h:307 [inline]
ip6_output+0x340/0x550 net/ipv6/ip6_output.c:247
NF_HOOK+0x9e/0x380 include/linux/netfilter.h:318
mld_sendpack+0x8d4/0xe60 net/ipv6/mcast.c:1855
mld_send_cr net/ipv6/mcast.c:2154 [inline]
mld_ifc_work+0x83e/0xd60 net/ipv6/mcast.c:2693
process_one_work kernel/workqueue.c:3257 [inline]
process_scheduled_works+0xad1/0x1770 kernel/workqueue.c:3340
worker_thread+0x8a0/0xda0 kernel/workqueue.c:3421
kthread+0x711/0x8a0 kernel/kthread.c:463
ret_from_fork+0x510/0xa50 arch/x86/kernel/process.c:158
ret_from_fork_asm+0x1a/0x30 arch/x86/entry/entry_64.S:246
Fixes: c544193214 ("GRE: Refactor GRE tunneling code.")
Reported-by: syzbot+7c134e1c3aa3283790b9@syzkaller.appspotmail.com
Closes: https://www.spinics.net/lists/netdev/msg1147302.html
Signed-off-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20260108190214.1667040-1-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit eb74c19fe1 ]
Add missing entries in netdev_lock_type[] and netdev_lock_name[] :
CAN, MCTP, RAWIP, CAIF, IP6GRE, 6LOWPAN, NETLINK, VSOCKMON,
IEEE802154_MONITOR.
Also add a WARN_ONCE() in netdev_lock_pos() to help future bug hunting
next time a protocol is added without updating these arrays.
Fixes: 1a33e10e4a ("net: partially revert dynamic lockdep key changes")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20260108093244.830280-1-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit c196def07b ]
The XFRM_STATE_NOPMTUDISC flag is only meaningful for output SAs, but
it was being applied regardless of the SA direction when the sysctl
ip_no_pmtu_disc is enabled. This can unintentionally affect input SAs.
Limit setting XFRM_STATE_NOPMTUDISC to output SAs when the SA direction
is configured.
Closes: https://github.com/strongswan/strongswan/issues/2946
Fixes: a4a87fa4e9 ("xfrm: Add Direction to the SA in or out")
Signed-off-by: Antony Antony <antony.antony@secunet.com>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 3d5221af9c ]
Commit 61fafbee6c ("xfrm: Determine inner GSO type from packet inner
protocol") attempted to fix GSO segmentation by reading the inner
protocol from XFRM_MODE_SKB_CB(skb)->protocol. This was incorrect
because the field holds the inner L4 protocol (TCP/UDP) instead of the
required tunnel protocol. Also, the memory location (shared by
XFRM_SKB_CB(skb) which could be overwritten by xfrm_replay_overflow())
is prone to corruption. This combination caused the kernel to select
the wrong inner mode and get the wrong address family.
The correct value is in xfrm_offload(skb)->proto, which is set from
the outer tunnel header's protocol field by esp[4|6]_gso_encap(). It
is initialized by xfrm[4|6]_tunnel_encap_add() to either IPPROTO_IPIP
or IPPROTO_IPV6, using xfrm_af2proto() and correctly reflects the
inner packet's address family.
Fixes: 61fafbee6c ("xfrm: Determine inner GSO type from packet inner protocol")
Signed-off-by: Jianbo Liu <jianbol@nvidia.com>
Reviewed-by: Sabrina Dubroca <sd@queasysnail.net>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit 7f9ee5fc97 upstream.
Fix a memory leak in bpf_prog_test_run_xdp() where the context buffer
allocated by bpf_ctx_init() is not freed when the function returns early
due to a data size check.
On the failing path:
ctx = bpf_ctx_init(...);
if (kattr->test.data_size_in - meta_sz < ETH_HLEN)
return -EINVAL;
The early return bypasses the cleanup label that kfree()s ctx, leading to a
leak detectable by kmemleak under fuzzing. Change the return to jump to the
existing free_ctx label.
Fixes: fe9544ed1a ("bpf: Support specifying linear xdp packet data size for BPF_PROG_TEST_RUN")
Reported-by: BPF Runtime Fuzzer (BRF)
Signed-off-by: Shardul Bankar <shardulsb08@gmail.com>
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Link: https://patch.msgid.link/20251014120037.1981316-1-shardulsb08@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 5d5602236f ]
syzbot is still reporting
unregister_netdevice: waiting for vcan0 to become free. Usage count = 2
even after commit 93a27b5891 ("can: j1939: add missing calls in
NETDEV_UNREGISTER notification handler") was added. A debug printk() patch
found that j1939_session_activate() can succeed even after
j1939_cancel_active_session() from j1939_netdev_notify(NETDEV_UNREGISTER)
has completed.
Since j1939_cancel_active_session() is processed with the session list lock
held, checking ndev->reg_state in j1939_session_activate() with the session
list lock held can reliably close the race window.
Reported-by: syzbot <syzbot+881d65229ca4f9ae8c84@syzkaller.appspotmail.com>
Closes: https://syzkaller.appspot.com/bug?extid=881d65229ca4f9ae8c84
Signed-off-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Acked-by: Oleksij Rempel <o.rempel@pengutronix.de>
Link: https://patch.msgid.link/b9653191-d479-4c8b-8536-1326d028db5c@I-love.SAKURA.ne.jp
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 8e1a1bc4f5 ]
Hamza Mahfooz reports cpu soft lock-ups in
nft_chain_validate():
watchdog: BUG: soft lockup - CPU#1 stuck for 27s! [iptables-nft-re:37547]
[..]
RIP: 0010:nft_chain_validate+0xcb/0x110 [nf_tables]
[..]
nft_immediate_validate+0x36/0x50 [nf_tables]
nft_chain_validate+0xc9/0x110 [nf_tables]
nft_immediate_validate+0x36/0x50 [nf_tables]
nft_chain_validate+0xc9/0x110 [nf_tables]
nft_immediate_validate+0x36/0x50 [nf_tables]
nft_chain_validate+0xc9/0x110 [nf_tables]
nft_immediate_validate+0x36/0x50 [nf_tables]
nft_chain_validate+0xc9/0x110 [nf_tables]
nft_immediate_validate+0x36/0x50 [nf_tables]
nft_chain_validate+0xc9/0x110 [nf_tables]
nft_immediate_validate+0x36/0x50 [nf_tables]
nft_chain_validate+0xc9/0x110 [nf_tables]
nft_table_validate+0x6b/0xb0 [nf_tables]
nf_tables_validate+0x8b/0xa0 [nf_tables]
nf_tables_commit+0x1df/0x1eb0 [nf_tables]
[..]
Currently nf_tables will traverse the entire table (chain graph), starting
from the entry points (base chains), exploring all possible paths
(chain jumps). But there are cases where we could avoid revalidation.
Consider:
1 input -> j2 -> j3
2 input -> j2 -> j3
3 input -> j1 -> j2 -> j3
Then the second rule does not need to revalidate j2, and, by extension j3,
because this was already checked during validation of the first rule.
We need to validate it only for rule 3.
This is needed because chain loop detection also ensures we do not exceed
the jump stack: Just because we know that j2 is cycle free, its last jump
might now exceed the allowed stack size. We also need to update all
reachable chains with the new largest observed call depth.
Care has to be taken to revalidate even if the chain depth won't be an
issue: chain validation also ensures that expressions are not called from
invalid base chains. For example, the masquerade expression can only be
called from NAT postrouting base chains.
Therefore we also need to keep record of the base chain context (type,
hooknum) and revalidate if the chain becomes reachable from a different
hook location.
Reported-by: Hamza Mahfooz <hamzamahfooz@linux.microsoft.com>
Closes: https://lore.kernel.org/netfilter-devel/20251118221735.GA5477@linuxonhyperv3.guj3yctzbm1etfxqx2vob5hsef.xx.internal.cloudapp.net/
Tested-by: Hamza Mahfooz <hamzamahfooz@linux.microsoft.com>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit ec69daabe4 ]
syzbot is reporting
unregister_netdevice: waiting for sit0 to become free. Usage count = 2
problem. A debug printk() patch found that a refcount is obtained at
xdp_convert_md_to_buff() from bpf_prog_test_run_xdp().
According to commit ec94670fcb ("bpf: Support specifying ingress via
xdp_md context in BPF_PROG_TEST_RUN"), the refcount obtained by
xdp_convert_md_to_buff() will be released by xdp_convert_buff_to_md().
Therefore, we can consider that the error handling path introduced by
commit 1c19499825 ("bpf: introduce frags support to
bpf_prog_test_run_xdp()") forgot to call xdp_convert_buff_to_md().
Reported-by: syzbot+881d65229ca4f9ae8c84@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=881d65229ca4f9ae8c84
Fixes: 1c19499825 ("bpf: introduce frags support to bpf_prog_test_run_xdp()")
Signed-off-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Reviewed-by: Toke Høiland-Jørgensen <toke@redhat.com>
Link: https://lore.kernel.org/r/af090e53-9d9b-4412-8acb-957733b3975c@I-love.SAKURA.ne.jp
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit e558cca217 ]
The xdp_frame structure takes up part of the XDP frame headroom,
limiting the size of the metadata. However, in bpf_test_run, we don't
take this into account, which makes it possible for userspace to supply
a metadata size that is too large (taking up the entire headroom).
If userspace supplies such a large metadata size in live packet mode,
the xdp_update_frame_from_buff() call in xdp_test_run_init_page() call
will fail, after which packet transmission proceeds with an
uninitialised frame structure, leading to the usual Bad Stuff.
The commit in the Fixes tag fixed a related bug where the second check
in xdp_update_frame_from_buff() could fail, but did not add any
additional constraints on the metadata size. Complete the fix by adding
an additional check on the metadata size. Reorder the checks slightly to
make the logic clearer and add a comment.
Link: https://lore.kernel.org/r/fa2be179-bad7-4ee3-8668-4903d1853461@hust.edu.cn
Fixes: b6f1f780b3 ("bpf, test_run: Fix packet size check for live packet mode")
Reported-by: Yinhao Hu <dddddd@hust.edu.cn>
Reported-by: Kaiyan Mei <M202472210@hust.edu.cn>
Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com>
Reviewed-by: Amery Hung <ameryhung@gmail.com>
Link: https://lore.kernel.org/r/20260105114747.1358750-1-toke@redhat.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit fe9544ed1a ]
To test bpf_xdp_pull_data(), an xdp packet containing fragments as well
as free linear data area after xdp->data_end needs to be created.
However, bpf_prog_test_run_xdp() always fills the linear area with
data_in before creating fragments, leaving no space to pull data. This
patch will allow users to specify the linear data size through
ctx->data_end.
Currently, ctx_in->data_end must match data_size_in and will not be the
final ctx->data_end seen by xdp programs. This is because ctx->data_end
is populated according to the xdp_buff passed to test_run. The linear
data area available in an xdp_buff, max_linear_sz, is alawys filled up
before copying data_in into fragments.
This patch will allow users to specify the size of data that goes into
the linear area. When ctx_in->data_end is different from data_size_in,
only ctx_in->data_end bytes of data will be put into the linear area when
creating the xdp_buff.
While ctx_in->data_end will be allowed to be different from data_size_in,
it cannot be larger than the data_size_in as there will be no data to
copy from user space. If it is larger than the maximum linear data area
size, the layout suggested by the user will not be honored. Data beyond
max_linear_sz bytes will still be copied into fragments.
Finally, since it is possible for a NIC to produce a xdp_buff with empty
linear data area, allow it when calling bpf_test_init() from
bpf_prog_test_run_xdp() so that we can test XDP kfuncs with such
xdp_buff. This is done by moving lower-bound check to callers as most of
them already do except bpf_prog_test_run_skb(). The change also fixes a
bug that allows passing an xdp_buff with data < ETH_HLEN. This can
happen when ctx is used and metadata is at least ETH_HLEN.
Signed-off-by: Amery Hung <ameryhung@gmail.com>
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
Link: https://patch.msgid.link/20250922233356.3356453-7-ameryhung@gmail.com
Stable-dep-of: e558cca217 ("bpf, test_run: Subtract size of xdp_frame from allowed metadata size")
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 7eb83bff02 ]
Change the variable naming in bpf_prog_test_run_xdp() to make the
overall logic less confusing. As different modes were added to the
function over the time, some variables got overloaded, making
it hard to understand and changing the code becomes error-prone.
Replace "size" with "linear_sz" where it refers to the size of metadata
and data. If "size" refers to input data size, use test.data_size_in
directly.
Replace "max_data_sz" with "max_linear_sz" to better reflect the fact
that it is the maximum size of metadata and data (i.e., linear_sz). Also,
xdp_rxq.frags_size is always PAGE_SIZE, so just set it directly instead
of subtracting headroom and tailroom and adding them back.
Signed-off-by: Amery Hung <ameryhung@gmail.com>
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
Link: https://patch.msgid.link/20250922233356.3356453-6-ameryhung@gmail.com
Stable-dep-of: e558cca217 ("bpf, test_run: Subtract size of xdp_frame from allowed metadata size")
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 4fc012daf9 ]
The bpf selftest xdp_adjust_tail/xdp_adjust_frags_tail_grow failed on
arm64 with 64KB page:
xdp_adjust_tail/xdp_adjust_frags_tail_grow:FAIL
In bpf_prog_test_run_xdp(), the xdp->frame_sz is set to 4K, but later on
when constructing frags, with 64K page size, the frag data_len could
be more than 4K. This will cause problems in bpf_xdp_frags_increase_tail().
To fix the failure, the xdp->frame_sz is set to be PAGE_SIZE so kernel
can test different page size properly. With the kernel change, the user
space and bpf prog needs adjustment. Currently, the MAX_SKB_FRAGS default
value is 17, so for 4K page, the maximum packet size will be less than 68K.
To test 64K page, a bigger maximum packet size than 68K is desired. So two
different functions are implemented for subtest xdp_adjust_frags_tail_grow.
Depending on different page size, different data input/output sizes are used
to adapt with different page size.
Signed-off-by: Yonghong Song <yonghong.song@linux.dev>
Link: https://lore.kernel.org/r/20250612035032.2207498-1-yonghong.song@linux.dev
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Stable-dep-of: e558cca217 ("bpf, test_run: Subtract size of xdp_frame from allowed metadata size")
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit c92510f5e3 ]
arp_create() is the only dev_hard_header() caller
making assumption about skb->head being unchanged.
A recent commit broke this assumption.
Initialize @arp pointer after dev_hard_header() call.
Fixes: db5b4e39c4 ("ip6_gre: make ip6gre_header() robust")
Reported-by: syzbot+58b44a770a1585795351@syzkaller.appspotmail.com
Signed-off-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20260107212250.384552-1-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 238e03d046 ]
When skb_segment_list() is called during packet forwarding, it handles
packets that were aggregated by the GRO engine.
Historically, the segmentation logic in skb_segment_list assumes that
individual segments are split from a parent SKB and may need to carry
their own socket memory accounting. Accordingly, the code transfers
truesize from the parent to the newly created segments.
Prior to commit ed4cccef64 ("gro: fix ownership transfer"), this
truesize subtraction in skb_segment_list() was valid because fragments
still carry a reference to the original socket.
However, commit ed4cccef64 ("gro: fix ownership transfer") changed
this behavior by ensuring that fraglist entries are explicitly
orphaned (skb->sk = NULL) to prevent illegal orphaning later in the
stack. This change meant that the entire socket memory charge remained
with the head SKB, but the corresponding accounting logic in
skb_segment_list() was never updated.
As a result, the current code unconditionally adds each fragment's
truesize to delta_truesize and subtracts it from the parent SKB. Since
the fragments are no longer charged to the socket, this subtraction
results in an effective under-count of memory when the head is freed.
This causes sk_wmem_alloc to remain non-zero, preventing socket
destruction and leading to a persistent memory leak.
The leak can be observed via KMEMLEAK when tearing down the networking
environment:
unreferenced object 0xffff8881e6eb9100 (size 2048):
comm "ping", pid 6720, jiffies 4295492526
backtrace:
kmem_cache_alloc_noprof+0x5c6/0x800
sk_prot_alloc+0x5b/0x220
sk_alloc+0x35/0xa00
inet6_create.part.0+0x303/0x10d0
__sock_create+0x248/0x640
__sys_socket+0x11b/0x1d0
Since skb_segment_list() is exclusively used for SKB_GSO_FRAGLIST
packets constructed by GRO, the truesize adjustment is removed.
The call to skb_release_head_state() must be preserved. As documented in
commit cf673ed0e0 ("net: fix fraglist segmentation reference count
leak"), it is still required to correctly drop references to SKB
extensions that may be overwritten during __copy_skb_header().
Fixes: ed4cccef64 ("gro: fix ownership transfer")
Signed-off-by: Mohammad Heib <mheib@redhat.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Link: https://patch.msgid.link/20260104213101.352887-1-mheib@redhat.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit ce5e612dd4 ]
SO_ZEROCOPY handling in vsock_connectible_setsockopt() does not get called
on accept()ed sockets due to a missing flag. Flip it.
Fixes: e0718bd82e ("vsock: enable setting SO_ZEROCOPY")
Signed-off-by: Michal Luczaj <mhal@rbox.co>
Link: https://patch.msgid.link/20251229-vsock-child-sock-custom-sockopt-v2-1-64778d6c4f88@rbox.co
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 4c0856c225 ]
When the ping program uses an IPPROTO_ICMP socket to send ICMP_ECHO
messages, ICMP_MIB_OUTMSGS is counted twice.
ping_v4_sendmsg
ping_v4_push_pending_frames
ip_push_pending_frames
ip_finish_skb
__ip_make_skb
icmp_out_count(net, icmp_type); // first count
icmp_out_count(sock_net(sk), user_icmph.type); // second count
However, when the ping program uses an IPPROTO_RAW socket,
ICMP_MIB_OUTMSGS is counted correctly only once.
Therefore, the first count should be removed.
Fixes: c319b4d76b ("net: ipv4: add IPPROTO_ICMP socket kind")
Signed-off-by: yuan.gao <yuan.gao@ucloud.cn>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Tested-by: Ido Schimmel <idosch@nvidia.com>
Link: https://patch.msgid.link/20251224063145.3615282-1-yuan.gao@ucloud.cn
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 3128df6be1 ]
When using an 802.1ad bridge with vlan_tunnel, the C-VLAN tag is
incorrectly stripped from frames during egress processing.
br_handle_egress_vlan_tunnel() uses skb_vlan_pop() to remove the S-VLAN
from hwaccel before VXLAN encapsulation. However, skb_vlan_pop() also
moves any "next" VLAN from the payload into hwaccel:
/* move next vlan tag to hw accel tag */
__skb_vlan_pop(skb, &vlan_tci);
__vlan_hwaccel_put_tag(skb, vlan_proto, vlan_tci);
For QinQ frames where the C-VLAN sits in the payload, this moves it to
hwaccel where it gets lost during VXLAN encapsulation.
Fix by calling __vlan_hwaccel_clear_tag() directly, which clears only
the hwaccel S-VLAN and leaves the payload untouched.
This path is only taken when vlan_tunnel is enabled and tunnel_info
is configured, so 802.1Q bridges are unaffected.
Tested with 802.1ad bridge + VXLAN vlan_tunnel, verified C-VLAN
preserved in VXLAN payload via tcpdump.
Fixes: 11538d039a ("bridge: vlan dst_metadata hooks in ingress and egress paths")
Signed-off-by: Alexandre Knecht <knecht.alexandre@gmail.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Acked-by: Nikolay Aleksandrov <razor@blackwall.org>
Link: https://patch.msgid.link/20251228020057.2788865-1-knecht.alexandre@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 7811ba4524 ]
Currently last_gc is being updated everytime a new connection is
tracked, that means that it is updated even if a GC wasn't performed.
With a sufficiently high packet rate, it is possible to always bypass
the GC, causing the list to grow infinitely.
Update the last_gc value only when a GC has been actually performed.
Fixes: d265929930 ("netfilter: nf_conncount: reduce unnecessary GC")
Signed-off-by: Fernando Fernandez Mancera <fmancera@suse.de>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Will Deacon
Daniel Borkmann
Jeongjun Park
Eric Dumazet
Melbin K Mathew
David Yang
Vladimir Oltean
Lachlan Hodges
Jamal Hadi Salim
Kuniyuki Iwashima
Xin Long
Michal Luczaj
Paul Chaignon
Tetsuo Handa
Ido Schimmel
Antony Antony
Jianbo Liu
Shardul Bankar
Florian Westphal
Toke Høiland-Jørgensen
Amery Hung
Yonghong Song
Xiang Mei
Mohammad Heib
Weiming Shi
yuan.gao
Alexandre Knecht
Fernando Fernandez Mancera