Re: [PATCH 1/2] vhost - Fix error path in vhost_net_set_backend

[Michael S. Tsirkin]

On Thu, Mar 04, 2010 at 04:10:14PM -0500, Jeff Dike wrote:
> commit 7bee327f8020f81a52ecb362f1fec7a4494dc98b
> Author: Jeff Dike <jdike@xxxxxxxxxxxxxxxxxxxxx>
> Date:   Thu Mar 4 14:48:38 2010 -0500
> 
>     Fix vhost_net_set_backend error path
>     
>     An error could cause vhost_net_set_backend to exit without unlocking
>     vq->mutex.
>     
>     Signed-off-by: Jeff Dike <jdike@xxxxxxxxxxxxxxx>

Thanks, applied.

> diff --git a/drivers/vhost/net.c b/drivers/vhost/net.c
> index ad37da2..4ada4b4 100644
> --- a/drivers/vhost/net.c
> +++ b/drivers/vhost/net.c
> @@ -508,12 +508,12 @@ static long vhost_net_set_backend(struct vhost_net *n, unsigned index, int fd)
>  	/* Verify that ring has been setup correctly. */
>  	if (!vhost_vq_access_ok(vq)) {
>  		r = -EFAULT;
> -		goto err;
> +		goto err_unlock_vq;
>  	}
>  	sock = get_socket(fd);
>  	if (IS_ERR(sock)) {
>  		r = PTR_ERR(sock);
> -		goto err;
> +		goto err_unlock_vq;
>  	}
>  
>  	/* start polling new socket */
> @@ -524,12 +524,14 @@ static long vhost_net_set_backend(struct vhost_net *n, unsigned index, int fd)
>  	vhost_net_disable_vq(n, vq);
>  	rcu_assign_pointer(vq->private_data, sock);
>  	vhost_net_enable_vq(n, vq);
> -	mutex_unlock(&vq->mutex);
>  done:
>  	if (oldsock) {
>  		vhost_net_flush_vq(n, index);
>  		fput(oldsock->file);
>  	}
> +
> +err_unlock_vq:
> +	mutex_unlock(&vq->mutex);
>  err:
>  	mutex_unlock(&n->dev.mutex);
>  	return r;
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/