[ovs-dev] [PATCH 2/2] ofproto-dpif: Store patch port peer in struct ofport_dpif.

Ben Pfaff blp at nicira.com
Tue Jun 18 20:12:24 UTC 2013


On Tue, Jun 18, 2013 at 01:11:22PM -0700, Ethan Jackson wrote:
> > In compose_output_action__(), the reason for the new test against
> > OVSP_NONE isn't clear to me.  Is this where we end up if a patch port
> > doesn't have a peer?  (How does it get OVSP_NONE?)
> 
> Right, all patch ports are initialized with OVSP_NONE.  If the patch
> port has a peer, it will be caught by the patch port implementation
> code.  Otherwise it will fall through and try to output to OVSP_NONE.

That's fine then, thanks.

Acked-by: Ben Pfaff <blp at nicira.com>



More information about the dev mailing list