On Wed, Nov 20, 2013 at 09:28:47AM -0800, Ben Pfaff wrote: > On Mon, Nov 18, 2013 at 03:48:29PM +0900, Simon Horman wrote: > > On Fri, Nov 15, 2013 at 02:33:14PM -0800, Ben Pfaff wrote: > > > Until now ofpacts_check() has been told either to enforce consistency or > > > not, but that means that the caller has to know exactly what protocol is > > > going to be in use (because some protocols require consistency to be > > > enforced and others don't). This commit changes ofpacts_check() to just > > > rule out protocols that require enforcement when it detects > > > inconsistencies. > > > > > > Signed-off-by: Ben Pfaff <b...@nicira.com> > > > > this approach seems reasonable to me. > > > > The problem I see with verification of VLAN actions that follow push_mpls > > when OpenFlow1.3 is not that different protocols require consistency to > > enforced while others don't. Rather the problem I as is that what is > > consistent is different for pre-OpenFlow1.3 and OpenFlow1.3+ tag ordering. > > Hmm, good point. > > > I believe that logic to handle that difference can be built on top of your > > patch and what follows is a prototype of such logic. I'm not sure that this > > is the direction you want us to take so I would appreciate feedback on this > > approach. > > I think the approach, that is, to check both semantics, is reasonable > enough. I didn't really review the code, since that isn't what you seem > to be looking for, yet. > > Thanks for taking a look, and for the very detailed feedback! I'll > apply this soon.
Likewise, thanks for spending time on this problem. I have updated the prototype I posted here earlier to fix up various details. I will include it in a fresh post of the MPLS patchset (v2.51). _______________________________________________ dev mailing list dev@openvswitch.org http://openvswitch.org/mailman/listinfo/dev
