On Thu, 30 Mar 2017 17:31:37 +0300
"Michael S. Tsirkin" <m...@redhat.com> wrote:
> On Thu, Mar 30, 2017 at 09:26:51AM +0200, Cornelia Huck wrote:
> > On Wed, 29 Mar 2017 23:48:54 +0300
> > "Michael S. Tsirkin" <m...@redhat.com> wrote:
> >
> > > With mergeable buffers we never use s/g for rx,
> > > so allow specifying context in that case.
> > >
> > > Signed-off-by: Michael S. Tsirkin <m...@redhat.com>
> > > ---
> > > drivers/net/virtio_net.c | 15 ++++++++++++++-
> > > 1 file changed, 14 insertions(+), 1 deletion(-)
> > >
> > > diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c
> > > index 6802169..340f737 100644
> > > --- a/drivers/net/virtio_net.c
> > > +++ b/drivers/net/virtio_net.c
> > > @@ -2044,6 +2044,7 @@ static int virtnet_find_vqs(struct virtnet_info *vi)
> > > int ret = -ENOMEM;
> > > int i, total_vqs;
> > > const char **names;
> > > + const bool *ctx;
> > >
> > > /* We expect 1 RX virtqueue followed by 1 TX virtqueue, followed by
> > > * possible N-1 RX/TX queue pairs used in multiqueue mode, followed by
> > > @@ -2062,6 +2063,13 @@ static int virtnet_find_vqs(struct virtnet_info
> > > *vi)
> > > names = kmalloc(total_vqs * sizeof(*names), GFP_KERNEL);
> > > if (!names)
> > > goto err_names;
> > > + if (vi->mergeable_rx_bufs) {
> > > + ctx = kzalloc(total_vqs * sizeof(*ctx), GFP_KERNEL);
> > > + if (!ctx)
> > > + goto err_ctx;
> > > + } else {
> > > + ctx = NULL;
> > > + }
> > >
> > > /* Parameters for control virtqueue, if any */
> > > if (vi->has_cvq) {
> > > @@ -2077,9 +2085,12 @@ static int virtnet_find_vqs(struct virtnet_info
> > > *vi)
> > > sprintf(vi->sq[i].name, "output.%d", i);
> > > names[rxq2vq(i)] = vi->rq[i].name;
> > > names[txq2vq(i)] = vi->sq[i].name;
> > > + if (ctx)
> > > + ctx[rxq2vq(i)] = true;
> > > }
> > >
> > > - ret = virtio_find_vqs(vi->vdev, total_vqs, vqs, callbacks, names, NULL);
> > > + ret = vi->vdev->config->find_vqs(vi->vdev, total_vqs, vqs, callbacks,
> > > + names, ctx, NULL);
> >
> > virtio_find_vqs_ctx()? (Needs to be exported, obviously.)
>
> I guess I can do that but there's a single user ATM.
> Do you think it's worth doing right now, or wait until
> it gets more users?
Given that you introduce it in patch 2, I'd just make it an exported
function from the start. (It also looks nicer IMO.)
