On Thu, Dec 11, 2025 at 09:30:06AM +0100, Stefano Garzarella wrote:

On Thu, Dec 11, 2025 at 10:45:47AM +0300, Dan Carpenter wrote:

quoted

Return a negative error code if the transport doesn't match.  Don't
return success.

Fixes: c0cfa2d8a788 ("vsock: add multi-transports support")
Cc: stable@vger.kernel.org
Signed-off-by: Dan Carpenter <redacted>
---
From static analysis.  Not tested.

net/vmw_vsock/virtio_transport_common.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/net/vmw_vsock/virtio_transport_common.c b/net/vmw_vsock/virtio_transport_common.c
index dcc8a1d5851e..77fbc6c541bf 100644
--- a/net/vmw_vsock/virtio_transport_common.c
+++ b/net/vmw_vsock/virtio_transport_common.c

@@ -1550,7 +1550,7 @@ virtio_transport_recv_listen(struct sock *sk, struct sk_buff *skb,

		release_sock(child);
		virtio_transport_reset_no_sock(t, skb);
		sock_put(child);
-		return ret;
+		return ret ?: -EINVAL;

Thanks for this fix. I think we have a similar issue also in
net/vmw_vsock/vmci_transport.c introduced by the same commit.
In net/vmw_vsock/hyperv_transport.c we have a similar pattern, but the
calling function return void, so no issue there.

Do you mind to fix also that one?

Sure.  I will resend a v2.

The check doesn't catch that one because the != comparison is
hidden inside the vmci_check_transport() call.  So I would have missed
it.  Thanks for catching it.

regards,
dan carpenter