Re: [PATCH] vsock/virtio: fix potential unbounded skb queue

["Michael S. Tsirkin" <mst@redhat.com>]

On Fri, May 15, 2026 at 04:22:12PM +0200, Luigi Leonardi wrote:
> From: Eric Dumazet <edumazet@google.com>
> 
> Upstream commit 059b7dbd20a6f0c539a45ddff1573cb8946685b5
> 
> virtio_transport_inc_rx_pkt() checks vvs->rx_bytes + len > vvs->buf_alloc.
> 
> virtio_transport_recv_enqueue() skips coalescing for packets
> with VIRTIO_VSOCK_SEQ_EOM.
> 
> If fed with packets with len == 0 and VIRTIO_VSOCK_SEQ_EOM,
> a very large number of packets can be queued
> because vvs->rx_bytes stays at 0.
> 
> Fix this by estimating the skb metadata size:
> 
> 	(Number of skbs in the queue) * SKB_TRUESIZE(0)
> 
> Fixes: 077706165717 ("virtio/vsock: don't use skbuff state to account credit")
> Signed-off-by: Eric Dumazet <edumazet@google.com>
> Cc: Arseniy Krasnov <AVKrasnov@sberdevices.ru>
> Cc: Stefan Hajnoczi <stefanha@redhat.com>
> Cc: Stefano Garzarella <sgarzare@redhat.com>
> Cc: Michael S. Tsirkin <mst@redhat.com>
> Cc: Jason Wang <jasowang@redhat.com>
> Cc: Xuan Zhuo <xuanzhuo@linux.alibaba.com>
> Cc: Eugenio Pérez <eperezma@redhat.com>
> Cc: virtualization@lists.linux.dev
> Link: https://patch.msgid.link/20260430122653.554058-1-edumazet@google.com
> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
> [LL: Fixed conflict since this tree does not use buf_used added by commit
>  45ca7e9f0730 ("vsock/virtio: fix `rx_bytes` accounting for stream sockets")]
> Signed-off-by: Luigi Leonardi <leonardi@redhat.com>
> ---
>  net/vmw_vsock/virtio_transport_common.c | 4 +++-
>  1 file changed, 3 insertions(+), 1 deletion(-)
> 
> diff --git a/net/vmw_vsock/virtio_transport_common.c b/net/vmw_vsock/virtio_transport_common.c
> index 4c374c36c29d..86e3051d000e 100644
> --- a/net/vmw_vsock/virtio_transport_common.c
> +++ b/net/vmw_vsock/virtio_transport_common.c
> @@ -283,7 +283,9 @@ static int virtio_transport_send_pkt_info(struct vsock_sock *vsk,
>  static bool virtio_transport_inc_rx_pkt(struct virtio_vsock_sock *vvs,
>  					u32 len)
>  {
> -	if (vvs->rx_bytes + len > vvs->buf_alloc)
> +	u64 skb_overhead = (skb_queue_len(&vvs->rx_queue) + 1) * SKB_TRUESIZE(0);
> +
> +	if (skb_overhead + vvs->rx_bytes + len > vvs->buf_alloc)
>  		return false;
> 
>  	vvs->rx_bytes += len;
> 
> ---
> base-commit: 3b9f64db049687c0d38b4b3ef2f297f0642179af
> change-id: 20260515-dumazet-07c0c855a9e2
> 
> Best regards,
> --
> Luigi Leonardi <leonardi@redhat.com>


I am not sure we should queue this for stable yet. It causes regressions
that we are now trying to fix.

-- 
MST