Re: [PATCH 2/2] virtio_net: remove send completion interrupts and avoid TX queue overrun through packet drop

[Rusty Russell <rusty@rustcorp.com.au>]

On Sun, 27 Mar 2011 09:52:54 +0200, "Michael S. Tsirkin" <mst@redhat.com> wrote:
> > Though IIRC, qemu's virtio barfs if the first descriptor isn't just the
> > hdr (barf...).
> 
> Maybe we can try fixing this before adding more flags,
> then e.g. publish used flag can be resued to also
> tell us layout is flexible. Or just add a feature flag for that.

We should probably do this at some stage, yes.

> > > 2. I didn't have time to work on virtio2 ideas presented
> > >    at the kvm forum yet, any takers?
> > 
> > I didn't even attend.
> 
> Hmm, right. But what was presented there was discussed on list as well:
> a single R/W descriptor ring with valid bit instead of 2 rings
> + a descriptor array.

I'll be happy when we reach the point that the extra cacheline is
hurting us :)

Then we should do direct descriptors w/ a cookie as the value to hand
back when finished.  That seems to be close to optimal.

> I agree absolutely that not all lessons has been learned,
> playing with different ring layouts would make at least
> an interesting paper IMO.

Yes, I'd like to see the results...

Thanks,
Rusty.