Re: [V4 PATCH] virtio-net: send gratuitous packet when needed

["Michael S. Tsirkin" <mst@redhat.com>]

On Mon, Mar 19, 2012 at 12:46:29PM +1030, Rusty Russell wrote:
> On Tue, 13 Mar 2012 16:33:31 +0200, "Michael S. Tsirkin" <mst@redhat.com> wrote:
> > > diff --git a/include/linux/virtio_net.h b/include/linux/virtio_net.h
> > > index 970d5a2..44a38d6 100644
> > > --- a/include/linux/virtio_net.h
> > > +++ b/include/linux/virtio_net.h
> > > @@ -49,8 +49,10 @@
> > >  #define VIRTIO_NET_F_CTRL_RX	18	/* Control channel RX mode support */
> > >  #define VIRTIO_NET_F_CTRL_VLAN	19	/* Control channel VLAN filtering */
> > >  #define VIRTIO_NET_F_CTRL_RX_EXTRA 20	/* Extra RX mode control support */
> > > +#define VIRTIO_NET_F_GUEST_ANNOUNCE 21  /* Guest can send gratituous packet */
> > >  
> > >  #define VIRTIO_NET_S_LINK_UP	1	/* Link is up */
> > > +#define VIRTIO_NET_S_ANNOUNCE   2       /* Announcement is needed */
> > 
> > I would put this in bit 8 (0x100), this way low status byte
> > is RO, high byte is RW.
> 
> The whole idea of acking by clearing the bit is unreliable, moving to a
> separate byte just controls the damage.
> 
> How about you use bits 8-15 as a counter?  It's still theoretically
> unreliable if 256 notifications pass before the guest notices, but it's
> probably better and clearer than this.
> 
> I leave the final call to MST though.
> 
> Thanks,
> Rusty.

I guess the point was that we want a single packet
so we don't care if multiple notifications are coalesced
into a single one.

> -- 
>   How could I marry someone with more hair than me?  http://baldalex.org