Re: [virtio] [OASIS Issue Tracker] Created: (VIRTIO-35) race condition with multi-dword config accesses

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

On Tue, Oct 08, 2013 at 02:26:35PM +0100, Pawel Moll wrote:
> On Tue, 2013-10-08 at 14:15 +0100, Michael S. Tsirkin wrote:
> > > > Transitional devices should support both old and
> > > > new guests.
> > > 
> > > We've discussed it before and I have a different opinion.
> > 
> > It's merely a definition (we voted this into spec btw :) ).
> > If you don't support legacy drivers
> > you have a non transitional device.
> > That's fine I guess.
> 
> This is what I meant - I don't mind the "transitional" idea in the spec.
> I simply don't want to bother MMIO devices authors with maintaining
> backward compatibility :-) Simply see no value of that in MMIO case.
> 
> Paweł
> 

Well basically what I am saying is that spec should say that drivers
MUST support both revision 1 and revision 2 and MUST always look at the
VIRTIO_1 feature bit to check whether device supports the new interface
(as opposed to checking the revision).

This removes info duplication between revision and feature bit
and will allow transitional devices if we ever want them
after all, without adding complexity.

Is this acceptable to you?


-- 
MST