Re: [virtio-dev] RFC: Doorbell suppression, packed-ring mode and hardware offload

[David Riddoch <driddoch@solarflare.com>]

> > > Maybe we should use a flag in event suppression structure.
> > > This way device can switch between being notification driven
> > > and being driven by index reads.
> > On first thought it seems hard to avoid races: On receiving a doorbell the
> > device wishes to transition from doorbells to polling, so driver->avail_idx
> > currently has an old value; device writes device->flags; a little later
> > device reads driver->avail_idx. It might have a new value written by the
> > driver, or an old value...the device can't tell. Needs more thought.
> I would say device writes flags and then reads the descriptor
> ring. This is exactly the same sequence that needs to happen
> when enabling driver notifications.

Yes fair enough. I was trying to allow a device that never has to read 
descriptors speculatively, but such a device can always have the 
avail_idx update flag enabled permanently.


--
David Riddoch  <driddoch@solarflare.com> -- Chief Architect, Solarflare