[SRU][J/N][PATCH 2/2] vsock: Orphan socket after transport release

[Tim Whisonant]

From: Michal Luczaj <mhal at rbox.co>

During socket release, sock_orphan() is called without considering that it
sets sk->sk_wq to NULL. Later, if SO_LINGER is enabled, this leads to a
null pointer dereferenced in virtio_transport_wait_close().

Orphan the socket only after transport release.

Partially reverts the 'Fixes:' commit.

KASAN: null-ptr-deref in range [0x0000000000000018-0x000000000000001f]
 lock_acquire+0x19e/0x500
 _raw_spin_lock_irqsave+0x47/0x70
 add_wait_queue+0x46/0x230
 virtio_transport_release+0x4e7/0x7f0
 __vsock_release+0xfd/0x490
 vsock_release+0x90/0x120
 __sock_release+0xa3/0x250
 sock_close+0x14/0x20
 __fput+0x35e/0xa90
 __x64_sys_close+0x78/0xd0
 do_syscall_64+0x93/0x1b0
 entry_SYSCALL_64_after_hwframe+0x76/0x7e

Reported-by: syzbot+9d55b199192a4be7d02c at syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=9d55b199192a4be7d02c
Fixes: fcdd2242c023 ("vsock: Keep the binding until socket destruction")
Tested-by: Luigi Leonardi <leonardi at redhat.com>
Reviewed-by: Luigi Leonardi <leonardi at redhat.com>
Signed-off-by: Michal Luczaj <mhal at rbox.co>
Link: https://patch.msgid.link/20250210-vsock-linger-nullderef-v3-1-ef6244d02b54@rbox.co
Signed-off-by: Jakub Kicinski <kuba at kernel.org>
(backported from commit 78dafe1cf3afa02ed71084b350713b07e72a18fb)
[tswhison: Adjusted context due to missing commit
135ffc7becc8 ("bpf, vsock: Invoke proto::close on close()")]
CVE-2025-21756
Signed-off-by: Tim Whisonant <tim.whisonant at canonical.com>
---
 net/vmw_vsock/af_vsock.c | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

diff --git a/net/vmw_vsock/af_vsock.c b/net/vmw_vsock/af_vsock.c
index 2a7e364edb97..2052f348f92b 100644
--- a/net/vmw_vsock/af_vsock.c
+++ b/net/vmw_vsock/af_vsock.c
@@ -817,13 +817,19 @@ static void __vsock_release(struct sock *sk, int level)
 		 */
 		lock_sock_nested(sk, level);
 
-		sock_orphan(sk);
+		/* Indicate to vsock_remove_sock() that the socket is being released and
+		 * can be removed from the bound_table. Unlike transport reassignment
+		 * case, where the socket must remain bound despite vsock_remove_sock()
+		 * being called from the transport release() callback.
+		 */
+		sock_set_flag(sk, SOCK_DEAD);
 
 		if (vsk->transport)
 			vsk->transport->release(vsk);
 		else if (sock_type_connectible(sk->sk_type))
 			vsock_remove_sock(vsk);
 
+		sock_orphan(sk);
 		sk->sk_shutdown = SHUTDOWN_MASK;
 
 		skb_queue_purge(&sk->sk_receive_queue);
-- 
2.43.0