Linux Virtualization - Apr 2018 - [PATCH v2 net 2/3] virtio_net: fix adding vids on big-endian

Programming vids (adding or removing them) still passes
guest-endian values in the DMA buffer. That's wrong
if guest is big-endian and when virtio 1 is enabled.

Note: this is on top of a previous patch:
	virtio_net: split out ctrl buffer

Fixes: 9465a7a6f ("virtio_net: enable v1.0 support")
Signed-off-by: Michael S. Tsirkin <mst at redhat.com>
---
 drivers/net/virtio_net.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c
index 3d0eff53..f84fe04 100644
--- a/drivers/net/virtio_net.c
+++ b/drivers/net/virtio_net.c
@@ -154,7 +154,7 @@ struct control_buf {
 	struct virtio_net_ctrl_mq mq;
 	u8 promisc;
 	u8 allmulti;
-	u16 vid;
+	__virtio16 vid;
 	u64 offloads;
 };
 
@@ -1718,7 +1718,7 @@ static int virtnet_vlan_rx_add_vid(struct net_device *dev,
 	struct virtnet_info *vi = netdev_priv(dev);
 	struct scatterlist sg;
 
-	vi->ctrl->vid = vid;
+	vi->ctrl->vid = cpu_to_virtio16(vi->vdev, vid);
 	sg_init_one(&sg, &vi->ctrl->vid, sizeof(vi->ctrl->vid));
 
 	if (!virtnet_send_command(vi, VIRTIO_NET_CTRL_VLAN,
@@ -1733,7 +1733,7 @@ static int virtnet_vlan_rx_kill_vid(struct net_device
*dev,
 	struct virtnet_info *vi = netdev_priv(dev);
 	struct scatterlist sg;
 
-	vi->ctrl->vid = vid;
+	vi->ctrl->vid = cpu_to_virtio16(vi->vdev, vid);
 	sg_init_one(&sg, &vi->ctrl->vid, sizeof(vi->ctrl->vid));
 
 	if (!virtnet_send_command(vi, VIRTIO_NET_CTRL_VLAN,
-- 
MST

On 2018?04?19? 13:30, Michael S. Tsirkin wrote:
> Programming vids (adding or removing them) still passes
> guest-endian values in the DMA buffer. That's wrong
> if guest is big-endian and when virtio 1 is enabled.
>
> Note: this is on top of a previous patch:
> 	virtio_net: split out ctrl buffer
>
> Fixes: 9465a7a6f ("virtio_net: enable v1.0 support")
> Signed-off-by: Michael S. Tsirkin <mst at redhat.com>
> ---
>   drivers/net/virtio_net.c | 6 +++---
>   1 file changed, 3 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c
> index 3d0eff53..f84fe04 100644
> --- a/drivers/net/virtio_net.c
> +++ b/drivers/net/virtio_net.c
> @@ -154,7 +154,7 @@ struct control_buf {
>   	struct virtio_net_ctrl_mq mq;
>   	u8 promisc;
>   	u8 allmulti;
> -	u16 vid;
> +	__virtio16 vid;
>   	u64 offloads;
>   };
>   
> @@ -1718,7 +1718,7 @@ static int virtnet_vlan_rx_add_vid(struct net_device
*dev,
>   	struct virtnet_info *vi = netdev_priv(dev);
>   	struct scatterlist sg;
>   
> -	vi->ctrl->vid = vid;
> +	vi->ctrl->vid = cpu_to_virtio16(vi->vdev, vid);
>   	sg_init_one(&sg, &vi->ctrl->vid,
sizeof(vi->ctrl->vid));
>   
>   	if (!virtnet_send_command(vi, VIRTIO_NET_CTRL_VLAN,
> @@ -1733,7 +1733,7 @@ static int virtnet_vlan_rx_kill_vid(struct net_device
*dev,
>   	struct virtnet_info *vi = netdev_priv(dev);
>   	struct scatterlist sg;
>   
> -	vi->ctrl->vid = vid;
> +	vi->ctrl->vid = cpu_to_virtio16(vi->vdev, vid);
>   	sg_init_one(&sg, &vi->ctrl->vid,
sizeof(vi->ctrl->vid));
>   
>   	if (!virtnet_send_command(vi, VIRTIO_NET_CTRL_VLAN,
Acked-by: Jason Wang <jasowang at redhat.com>

On Thu, 19 Apr 2018 08:30:49 +0300
"Michael S. Tsirkin" <mst at redhat.com> wrote:
> Programming vids (adding or removing them) still passes
> guest-endian values in the DMA buffer. That's wrong
> if guest is big-endian and when virtio 1 is enabled.
> 
> Note: this is on top of a previous patch:
> 	virtio_net: split out ctrl buffer
> 
> Fixes: 9465a7a6f ("virtio_net: enable v1.0 support")
> Signed-off-by: Michael S. Tsirkin <mst at redhat.com>
> ---
>  drivers/net/virtio_net.c | 6 +++---
>  1 file changed, 3 insertions(+), 3 deletions(-)
Ouch. Have you seen any bug reports for that?

On Thu, Apr 19, 2018 at 03:26:41PM +0200, Cornelia Huck
wrote:
> On Thu, 19 Apr 2018 08:30:49 +0300
> "Michael S. Tsirkin" <mst at redhat.com> wrote:
> 
> > Programming vids (adding or removing them) still passes
> > guest-endian values in the DMA buffer. That's wrong
> > if guest is big-endian and when virtio 1 is enabled.
> > 
> > Note: this is on top of a previous patch:
> > 	virtio_net: split out ctrl buffer
> > 
> > Fixes: 9465a7a6f ("virtio_net: enable v1.0 support")
> > Signed-off-by: Michael S. Tsirkin <mst at redhat.com>
> > ---
> >  drivers/net/virtio_net.c | 6 +++---
> >  1 file changed, 3 insertions(+), 3 deletions(-)
> 
> Ouch. Have you seen any bug reports for that?
No, but then vlans within VMs aren't used too often
(as opposed to attaching vlans by the HV).

-- 
MST