On Mon, Aug 06, 2012 at 11:15:31AM -0700, Ben Pfaff wrote: > On Wed, Aug 01, 2012 at 04:01:46PM +0900, Simon Horman wrote: > > This may be used by both ofputil_decode_flow_mod() and > > ofputil_decode_packet_in() > > > > Signed-off-by: Simon Horman <ho...@verge.net.au> > > Do you think see a benefit in prohibiting use of OXM in OF1.1? I have > been planning to allow any supported match format in any version of > the protocol, just because I don't know of a drawback.
I think that generally I side towards disallowing things, which explains the approach that I have taken in this patch. I am comfortable with allowing OXM for OF1.1, even though I think it is is out of spec. And I think that doing so would slightly simplify the code. Should I update the patch? _______________________________________________ dev mailing list dev@openvswitch.org http://openvswitch.org/mailman/listinfo/dev
