bgpd: Continue processing rpki nodes

We are prematurely stopping processing of bgp_path_info's
when bgp_update returns a -1.  The only place that this
happens when we cross some prefix limit.  But we still
need to continiue processing other nodes in the list too.

Signed-off-by: Donald Sharp <sharpd@nvidia.com>

diff --git a/bgpd/bgp_rpki.c b/bgpd/bgp_rpki.c
index 451264ef7dcdc5cb7576ed52914abb0bbe3a818d..04ff8ac7c91c428c73e2ee3cc93f900e11571c2f 100644 (file)
--- a/bgpd/bgp_rpki.c
+++ b/bgpd/bgp_rpki.c
@@ -421,7 +421,6 @@ static void revalidate_bgp_node(struct bgp_dest *bgp_dest, afi_t afi,
        struct bgp_adj_in *ain;
 
        for (ain = bgp_dest->adj_in; ain; ain = ain->next) {
-               int ret;
                struct bgp_path_info *path =
                        bgp_dest_get_bgp_path_info(bgp_dest);
                mpls_label_t *label = NULL;
@@ -431,13 +430,10 @@ static void revalidate_bgp_node(struct bgp_dest *bgp_dest, afi_t afi,
                        label = path->extra->label;
                        num_labels = path->extra->num_labels;
                }
-               ret = bgp_update(ain->peer, bgp_dest_get_prefix(bgp_dest),
+               (void)bgp_update(ain->peer, bgp_dest_get_prefix(bgp_dest),
                                 ain->addpath_rx_id, ain->attr, afi, safi,
                                 ZEBRA_ROUTE_BGP, BGP_ROUTE_NORMAL, NULL, label,
                                 num_labels, 1, NULL);
-
-               if (ret < 0)
-                       return;
        }
 }