It looks to me like there may have been some rebasing issues in this patch. The general approach seems fine though.
> - connmgr_send_port_status(p->connmgr, &ofport->opp, OFPPR_DELETE); > - > - netdev_monitor_remove(p->netdev_monitor, ofport->netdev); > - hmap_remove(&p->ports, &ofport->hmap_node); > - shash_delete(&p->port_by_name, > - shash_find(&p->port_by_name, > - netdev_get_name(ofport->netdev))); > - if (p->sflow) { > - ofproto_sflow_del_port(p->sflow, ofport->odp_port); > - } > - > + connmgr_send_port_status(ofport->ofproto->connmgr, &ofport->opp, > + OFPPR_DELETE); > ofport_free(ofport); > } Does this code deletion belong in this patch or a separate one? It's not clear to me what it has to do with adding a pointer up to the parent ofproto. > @@ -1341,7 +1334,7 @@ ofport_remove_with_name(struct ofproto *ofproto, const > char *name) > * Does not handle a name or port number change. The caller must implement > * such a change as a delete followed by an add. */ > static void > -ofport_modified(struct ofproto *ofproto, struct ofport *port, > +ofport_modified(struct ofport *port, > struct netdev *netdev, struct ofp_phy_port *opp) > { The code doesn't compile because ofport_modified still needs a reference to the parent ofproto. Should be a simple "struct ofproto *ofproto = port->ofproto;" at the beginning of the funciton. Ethan _______________________________________________ dev mailing list dev@openvswitch.org http://openvswitch.org/mailman/listinfo/dev