On Tue, Feb 3, 2015 at 1:18 PM, Thomas Graf <tg...@noironetworks.com> wrote: > On 02/03/15 at 10:08am, Pravin Shelar wrote: >> On Fri, Jan 30, 2015 at 6:36 AM, Thomas Graf <tg...@noironetworks.com> wrote: >> > + flags = TUNNEL_KEY; >> > + vxlan_port = vxlan_vport(vport); >> > + if (vxlan_port->exts & VXLAN_F_GBP) >> > + flags |= TUNNEL_VXLAN_OPT; >> > >> This does not allow GBP and non-GBP tunnel share vxlan port, since OVS >> we have single vport shared between all vxlan end-points for given UDP >> port. > > Correct, a non-GBP tunnel will ignore the GBP bits. Though a GBP tunnel > will continue to receive non-GBP frames. This reflects what has been > merged in the upstream kernel. The reason to require enabling > individual extensions is to be able to support overlapping extensions, > e.g. although using a different identifier bit, VXLAN-GPE is not > compatible with VXLAN-GBP and the user can only enable one of the two > extensions on a particular tunnel.
I agree port will continue to accept non GBP frames but sets TUNNEL_VXLAN_OPT which is odd. Since it is checking extension for the vport. Why not set TUNNEL_VXLAN_OPT according to received packet header/metadata? _______________________________________________ dev mailing list dev@openvswitch.org http://openvswitch.org/mailman/listinfo/dev
