On Wed, Aug 24, 2011 at 09:29:29AM +0800, Jesse Gross wrote: > On Wed, Aug 24, 2011 at 2:22 AM, Ben Pfaff <b...@nicira.com> wrote: > > On Tue, Aug 23, 2011 at 11:17:22AM -0700, Pravin Shelar wrote: > >> On Tue, Aug 23, 2011 at 10:55 AM, Ben Pfaff <b...@nicira.com> wrote: > >> > On Tue, Aug 23, 2011 at 10:48:42AM -0700, Pravin Shelar wrote: > >> >> Currently the kernel vlan actions mirror those used by OpenFlow 1.0. > >> >> i.e. > >> >> MODIFY and STRIP. More flexible approach is to have an action to push a > >> >> tag and pop a tag off, so that it can handle multiple levels of vlan > >> >> tags. > >> >> Plus it aligns with newer version of OpenFlow. > >> >> ? ? ? As this patch replaces MODIFY with PUSH semantic, > >> >> action ?mapping done in userpace is fixed accordingly. > >> >> ? ? ? GSO handling for multiple levels of vlan tags is also added as > >> >> Jesse suggested before. > >> >> > >> >> Signed-off-by: Pravin B Shelar <pshe...@nicira.com> > >> > > >> > I'm surprised that the new "push" action doesn't accept a tpid as part > >> > of its argument, so that it could be transparently extended later to > >> > allow pushing a nonstandard (pre-standardization) tpid such as 0x9100. > >> > >> We could have 32 bit argument for push action. I guess this is all can > >> be at this moment. > > > > The ovs_key_8021q structure might be an appropriate type for the > > argument. ??(It's still 32 bits.) > > The plan is to do that in a second step. I thought about it and I > think your idea of using some generic actions with the ovs_key > structures make sense. The ability (at least as a placeholder) to use > non-standard TPIDs will fall out of that naturally, so this is just > about the push/pop semantics.
That makes sense. Thank you for the explanation. _______________________________________________ dev mailing list dev@openvswitch.org http://openvswitch.org/mailman/listinfo/dev
