Re: [virtio-dev] [PATCH v9] virtio-net: add Max MTU configuration field

[Cornelia Huck <cornelia.huck@de.ibm.com>]

On Wed, 7 Sep 2016 17:15:13 +0300
"Michael S. Tsirkin" <mst@redhat.com> wrote:

> On Wed, Sep 07, 2016 at 11:18:22AM +0200, Cornelia Huck wrote:
> > On Tue,  6 Sep 2016 10:31:01 -0400
> > Aaron Conole <aconole@redhat.com> wrote:

> > > @@ -3153,6 +3166,18 @@ struct virtio_net_config {
> > >  The device MUST set \field{max_virtqueue_pairs} to between 1 and 0x8000 inclusive,
> > >  if it offers VIRTIO_NET_F_MQ.
> > > 
> > > +The device MUST set \field{mtu} to between 68 and 65535 inclusive,
> > > +if it offers VIRTIO_NET_F_MTU.
> > > +
> > > +The device MUST NOT modify \field{mtu} once it has been set.
> > 
> > Should this rather be relaxed to "after VIRTIO_NET_F_MTU has been
> > successfully negotiated"? I don't think the driver should assume
> > anything until after feature negotiation is complete.
> 
> I don't think this matters much, but generally it's best
> if the config space can be read at any time.
> For example, this can allow validating MTU and not negotiating
> it if it's e.g. too small.
> Do you see value in tweaking the MTU after negotiation?

Not really. Let's keep it that way.