On Sat, Feb 27, 2016 at 11:44 AM, Jiri Benc <jb...@redhat.com> wrote: > On Sat, 27 Feb 2016 20:21:59 +0100, Jiri Benc wrote: >> You mean returning ETH_P_TEB in skb->protocol? That's not much useful, >> unfortunately. You won't get such packet processed by the kernel IP >> stack, rendering the VXLAN-GPE device unusable outside of ovs. It would >> effectively became a packet sink when used standalone, as it cannot be >> added to bridge and received packets are not processed by anything - >> there's no protocol handler for ETH_P_TEB. > > Actually, I can do that in the L3 mode (or whatever we'll call it). It > won't hurt anything and may be useful for openvswitch. Ovs will have to > special case VXLAN-GPE vport (or perhaps any ARPHRD_NONE port) and set > skb->protocol to ETH_P_TEB on xmit and dissect correctly the ETH_P_TEB > packet on rcv. > > The L2 mode (or whatever we'll call it) will need to stay, though, for > non-ovs use cases.
Sorry for the delay in responding but this plan sounds good to me.
