Re: [rfc net-next v6 2/3] virtio_net: multiqueue support

From: Ben Hutchings <bhutchings@solarflare.com>
To: "Michael S. Tsirkin" <mst@redhat.com>
Cc: krkumar2@in.ibm.com, kvm@vger.kernel.org, netdev@vger.kernel.org,
	linux-kernel@vger.kernel.org,
	virtualization@lists.linux-foundation.org, edumazet@google.com,
	davem@davemloft.net
Subject: Re: [rfc net-next v6 2/3] virtio_net: multiqueue support
Date: Mon, 19 Nov 2012 18:44:21 +0000	[thread overview]
Message-ID: <1353350661.4156.0.camel__18811.354696709$1353350692$gmane$org@bwh-desktop.uk.solarflarecom.com> (raw)
In-Reply-To: <20121118091354.GA31775@redhat.com>

On Sun, 2012-11-18 at 11:13 +0200, Michael S. Tsirkin wrote:
> On Sat, Nov 17, 2012 at 12:35:29AM +0000, Ben Hutchings wrote:
> > On Tue, 2012-11-13 at 08:40 +0200, Michael S. Tsirkin wrote:
> > > On Mon, Nov 05, 2012 at 11:38:39AM +1030, Rusty Russell wrote:
> > > > > @@ -924,11 +1032,10 @@ static void virtnet_get_ringparam(struct net_device *dev,
> > > > >  {
> > > > >  	struct virtnet_info *vi = netdev_priv(dev);
> > > > >  
> > > > > -	ring->rx_max_pending = virtqueue_get_vring_size(vi->rvq);
> > > > > -	ring->tx_max_pending = virtqueue_get_vring_size(vi->svq);
> > > > > +	ring->rx_max_pending = virtqueue_get_vring_size(vi->rq[0].vq);
> > > > > +	ring->tx_max_pending = virtqueue_get_vring_size(vi->sq[0].vq);
> > > > >  	ring->rx_pending = ring->rx_max_pending;
> > > > >  	ring->tx_pending = ring->tx_max_pending;
> > > > > -
> > > > >  }
> > > > 
> > > > This assumes all vqs are the same size.  I think this should probably
> > > > check: for mq mode, use the first vq, otherewise use the 0th.
> > > 
> > > For rx_pending/tx_pending I think what is required here is the
> > > actual number of outstanding buffers.
> > > Dave, Eric - right?
> > > 
> > > So this should be the total over all rings and to be useful,
> > > rx_max_pending/tx_max_pending should be the total too.
> > 
> > So far as I know, all current implementations use the number of
> > descriptors per ring here. virtio_net should be consistent with this.
> > 
> > Ben.
> 
> Problem is, it could in theory be different between rings. I guess we
> could use the maximum.
> 
> What's the right thing to do for rx_pending - I am guessing
> we want the current outstanding packets right?

The 'max_pending' fields are for the maximum ring sizes supported; the
'pending' fields are for the current ring sizes.  The current number of
descriptors pending is not included (would be a bit pointless).

Ben.

-- 
Ben Hutchings, Staff Engineer, Solarflare
Not speaking for my employer; that's the marketing department's job.
They asked us to note that Solarflare product names are trademarked.