Virtualization - Mar 2017 - [PATCH] virtio_net: enable big packets for large MTU values

If one enables e.g. jumbo frames without mergeable
buffers, packets won't fit in 1500 byte buffers
we use. Switch to big packet mode instead.
TODO: make sizing more exact, possibly extend small
packet mode to use larger pages.

Signed-off-by: Michael S. Tsirkin <mst at redhat.com>
---
 drivers/net/virtio_net.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c
index e0fb3707..9dc31dc 100644
--- a/drivers/net/virtio_net.c
+++ b/drivers/net/virtio_net.c
@@ -2428,6 +2428,10 @@ static int virtnet_probe(struct virtio_device *vdev)
 			dev->mtu = mtu;
 			dev->max_mtu = mtu;
 		}
+
+		/* TODO: size buffers correctly in this case. */
+		if (dev->mtu > ETH_DATA_LEN)
+			vi->big_packets = true;
 	}
 
 	if (vi->any_header_sg)
-- 
MST

On Wed, Mar 29, 2017 at 03:38:09PM +0300, Michael S. Tsirkin
wrote:
> If one enables e.g. jumbo frames without mergeable
> buffers, packets won't fit in 1500 byte buffers
> we use. Switch to big packet mode instead.
> TODO: make sizing more exact, possibly extend small
> packet mode to use larger pages.
> 
> Signed-off-by: Michael S. Tsirkin <mst at redhat.com>
Fixes: 14de9d114a82 ("virtio-net: Add initial MTU advice feature")
> ---
>  drivers/net/virtio_net.c | 4 ++++
>  1 file changed, 4 insertions(+)
> 
> diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c
> index e0fb3707..9dc31dc 100644
> --- a/drivers/net/virtio_net.c
> +++ b/drivers/net/virtio_net.c
> @@ -2428,6 +2428,10 @@ static int virtnet_probe(struct virtio_device *vdev)
>  			dev->mtu = mtu;
>  			dev->max_mtu = mtu;
>  		}
> +
> +		/* TODO: size buffers correctly in this case. */
> +		if (dev->mtu > ETH_DATA_LEN)
> +			vi->big_packets = true;
>  	}
>  
>  	if (vi->any_header_sg)
> -- 
> MST

On 2017?03?29? 20:38, Michael S. Tsirkin wrote:
> If one enables e.g. jumbo frames without mergeable
> buffers, packets won't fit in 1500 byte buffers
> we use. Switch to big packet mode instead.
> TODO: make sizing more exact, possibly extend small
> packet mode to use larger pages.
>
> Signed-off-by: Michael S. Tsirkin <mst at redhat.com>
> ---
>   drivers/net/virtio_net.c | 4 ++++
>   1 file changed, 4 insertions(+)
>
> diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c
> index e0fb3707..9dc31dc 100644
> --- a/drivers/net/virtio_net.c
> +++ b/drivers/net/virtio_net.c
> @@ -2428,6 +2428,10 @@ static int virtnet_probe(struct virtio_device *vdev)
>   			dev->mtu = mtu;
>   			dev->max_mtu = mtu;
>   		}
> +
> +		/* TODO: size buffers correctly in this case. */
> +		if (dev->mtu > ETH_DATA_LEN)
> +			vi->big_packets = true;
>   	}
>   
>   	if (vi->any_header_sg)
Ok, I think maybe we should fail the probe of small buffer in this case 
and decrease the max_mtu to ETH_DATA_LEN if small buffer is used (since 
user are allowed to increase the mtu).

Thanks