Linux Virtualization - Mar 2018 - [PATCH net] vhost_net: add missing lock nesting notation

We try to hold TX virtqueue mutex in vhost_net_rx_peek_head_len()
after RX virtqueue mutex is held in handle_rx(). This requires an
appropriate lock nesting notation to calm down deadlock detector.

Fixes: 0308813724606 ("vhost_net: basic polling support")
Reported-by: syzbot+7f073540b1384a614e09 at syzkaller.appspotmail.com
Signed-off-by: Jason Wang <jasowang at redhat.com>
---
 drivers/vhost/net.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/vhost/net.c b/drivers/vhost/net.c
index 8139bc7..12bcfba 100644
--- a/drivers/vhost/net.c
+++ b/drivers/vhost/net.c
@@ -630,7 +630,7 @@ static int vhost_net_rx_peek_head_len(struct vhost_net *net,
struct sock *sk)
 
 	if (!len && vq->busyloop_timeout) {
 		/* Both tx vq and rx socket were polled here */
-		mutex_lock(&vq->mutex);
+		mutex_lock_nested(&vq->mutex, 1);
 		vhost_disable_notify(&net->dev, vq);
 
 		preempt_disable();
@@ -763,7 +763,7 @@ static void handle_rx(struct vhost_net *net)
 	struct iov_iter fixup;
 	__virtio16 num_buffers;
 
-	mutex_lock(&vq->mutex);
+	mutex_lock_nested(&vq->mutex, 0);
 	sock = vq->private_data;
 	if (!sock)
 		goto out;
-- 
2.7.4

On Mon, Mar 26, 2018 at 04:10:23PM +0800, Jason Wang
wrote:
> We try to hold TX virtqueue mutex in vhost_net_rx_peek_head_len()
> after RX virtqueue mutex is held in handle_rx(). This requires an
> appropriate lock nesting notation to calm down deadlock detector.
> 
> Fixes: 0308813724606 ("vhost_net: basic polling support")
> Reported-by: syzbot+7f073540b1384a614e09 at syzkaller.appspotmail.com
> Signed-off-by: Jason Wang <jasowang at redhat.com>
Acked-by: Michael S. Tsirkin <mst at redhat.com>
> ---
>  drivers/vhost/net.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/vhost/net.c b/drivers/vhost/net.c
> index 8139bc7..12bcfba 100644
> --- a/drivers/vhost/net.c
> +++ b/drivers/vhost/net.c
> @@ -630,7 +630,7 @@ static int vhost_net_rx_peek_head_len(struct vhost_net
*net, struct sock *sk)
>  
>  	if (!len && vq->busyloop_timeout) {
>  		/* Both tx vq and rx socket were polled here */
> -		mutex_lock(&vq->mutex);
> +		mutex_lock_nested(&vq->mutex, 1);
>  		vhost_disable_notify(&net->dev, vq);
>  
>  		preempt_disable();
> @@ -763,7 +763,7 @@ static void handle_rx(struct vhost_net *net)
>  	struct iov_iter fixup;
>  	__virtio16 num_buffers;
>  
> -	mutex_lock(&vq->mutex);
> +	mutex_lock_nested(&vq->mutex, 0);
>  	sock = vq->private_data;
>  	if (!sock)
>  		goto out;
> -- 
> 2.7.4

From: Jason Wang <jasowang at redhat.com>
Date: Mon, 26 Mar 2018 16:10:23 +0800
> We try to hold TX virtqueue mutex in vhost_net_rx_peek_head_len()
> after RX virtqueue mutex is held in handle_rx(). This requires an
> appropriate lock nesting notation to calm down deadlock detector.
> 
> Fixes: 0308813724606 ("vhost_net: basic polling support")
> Reported-by: syzbot+7f073540b1384a614e09 at syzkaller.appspotmail.com
> Signed-off-by: Jason Wang <jasowang at redhat.com>
Applied and queued up for -stable, thanks Jason.