On Thu, May 01, 2014 at 08:44:29AM -0700, Ben Pfaff wrote: > On Tue, Apr 22, 2014 at 05:54:52PM +0900, Simon Horman wrote: > > Rather then prohibiting hash actions dpif_execute_helper_cb() > > execute them by passing them on to aux->dpif->dpif_class->execute() > > and save the updated packet metadata afterwards. It is assumed > > that if a hash is calculated it will be sored in the dp_hash field > > of the metadata. > > Not sure where this is going. Userspace can't necessarily reproduce the > hash values computed by the datapath; the particular hash function is > unspecified.
Indeed. This was going towards executing actions in user-space if the in_port is not present in the datapath and a recirculation action is added. That discussion has moved on a bit and it now seems that the approach I took is not the favoured one. I'll remove the patches in this patchset that implement that approach, rebase the remaining patches and repost. _______________________________________________ dev mailing list dev@openvswitch.org http://openvswitch.org/mailman/listinfo/dev
