Adding upstream version 8.4.4.upstream/8.4.4upstream

Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>

1 files changed, 683 insertions, 0 deletions

diff --git a/ospfd/ospf_ia.c b/ospfd/ospf_ia.c
new file mode 100644
index 0000000..f9869a6
--- /dev/null
+++ b/ospfd/ospf_ia.c
@@ -0,0 +1,683 @@
+/*
+ * OSPF inter-area routing.
+ * Copyright (C) 1999, 2000 Alex Zinin, Toshiaki Takada
+ *
+ * This file is part of GNU Zebra.
+ *
+ * GNU Zebra is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the
+ * Free Software Foundation; either version 2, or (at your option) any
+ * later version.
+ *
+ * GNU Zebra is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License along
+ * with this program; see the file COPYING; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
+ */
+
+
+#include <zebra.h>
+
+#include "thread.h"
+#include "memory.h"
+#include "hash.h"
+#include "linklist.h"
+#include "prefix.h"
+#include "table.h"
+#include "log.h"
+
+#include "ospfd/ospfd.h"
+#include "ospfd/ospf_interface.h"
+#include "ospfd/ospf_ism.h"
+#include "ospfd/ospf_asbr.h"
+#include "ospfd/ospf_lsa.h"
+#include "ospfd/ospf_lsdb.h"
+#include "ospfd/ospf_neighbor.h"
+#include "ospfd/ospf_nsm.h"
+#include "ospfd/ospf_spf.h"
+#include "ospfd/ospf_route.h"
+#include "ospfd/ospf_ase.h"
+#include "ospfd/ospf_abr.h"
+#include "ospfd/ospf_ia.h"
+#include "ospfd/ospf_dump.h"
+
+static struct ospf_route *ospf_find_abr_route(struct route_table *rtrs,
+					      struct prefix_ipv4 *abr,
+					      struct ospf_area *area)
+{
+	struct route_node *rn;
+	struct ospf_route * or ;
+	struct listnode *node;
+
+	if ((rn = route_node_lookup(rtrs, (struct prefix *)abr)) == NULL)
+		return NULL;
+
+	route_unlock_node(rn);
+
+	for (ALL_LIST_ELEMENTS_RO((struct list *)rn->info, node, or))
+		if (IPV4_ADDR_SAME(& or->u.std.area_id, &area->area_id)
+		    && (or->u.std.flags & ROUTER_LSA_BORDER))
+			return or ;
+
+	return NULL;
+}
+
+static void ospf_ia_network_route(struct ospf *ospf, struct route_table *rt,
+				  struct prefix_ipv4 *p,
+				  struct ospf_route *new_or,
+				  struct ospf_route *abr_or)
+{
+	struct route_node *rn1;
+	struct ospf_route * or ;
+
+	if (IS_DEBUG_OSPF_EVENT)
+		zlog_debug("%s: processing summary route to %pFX", __func__, p);
+
+	/* Find a route to the same dest */
+	if ((rn1 = route_node_lookup(rt, (struct prefix *)p))) {
+		int res;
+
+		route_unlock_node(rn1);
+
+		if ((or = rn1->info)) {
+			if (IS_DEBUG_OSPF_EVENT)
+				zlog_debug(
+					"%s: Found a route to the same network",
+					__func__);
+			/* Check the existing route. */
+			if ((res = ospf_route_cmp(ospf, new_or, or)) < 0) {
+				/* New route is better, so replace old one. */
+				ospf_route_subst(rn1, new_or, abr_or);
+			} else if (res == 0) {
+				/* New and old route are equal, so next hops can
+				 * be added. */
+				route_lock_node(rn1);
+				ospf_route_copy_nexthops(or, abr_or->paths);
+				route_unlock_node(rn1);
+
+				/* new route can be deleted, because existing
+				 * route has been updated. */
+				ospf_route_free(new_or);
+			} else {
+				/* New route is worse, so free it. */
+				ospf_route_free(new_or);
+				return;
+			}
+		} /* if (or)*/
+	}	 /*if (rn1)*/
+	else {    /* no route */
+		if (IS_DEBUG_OSPF_EVENT)
+			zlog_debug("%s: add new route to %pFX", __func__, p);
+		ospf_route_add(rt, p, new_or, abr_or);
+	}
+}
+
+static void ospf_ia_router_route(struct ospf *ospf, struct route_table *rtrs,
+				 struct prefix_ipv4 *p,
+				 struct ospf_route *new_or,
+				 struct ospf_route *abr_or)
+{
+	struct ospf_route * or = NULL;
+	struct route_node *rn;
+	int ret;
+
+	if (IS_DEBUG_OSPF_EVENT)
+		zlog_debug("%s: considering %pFX", __func__, p);
+
+	/* Find a route to the same dest */
+	rn = route_node_get(rtrs, (struct prefix *)p);
+
+	if (rn->info == NULL)
+		/* This is a new route */
+		rn->info = list_new();
+	else {
+		struct ospf_area *or_area;
+		or_area = ospf_area_lookup_by_area_id(ospf,
+						      new_or->u.std.area_id);
+		assert(or_area);
+		/* This is an additional route */
+		route_unlock_node(rn);
+		or = ospf_find_asbr_route_through_area(rtrs, p, or_area);
+	}
+
+	if (or) {
+		if (IS_DEBUG_OSPF_EVENT)
+			zlog_debug(
+				"%s: a route to the same ABR through the same area exists",
+				__func__);
+		/* New route is better */
+		if ((ret = ospf_route_cmp(ospf, new_or, or)) < 0) {
+			listnode_delete(rn->info, or);
+			ospf_route_free(or);
+			/* proceed down */
+		}
+		/* Routes are the same */
+		else if (ret == 0) {
+			if (IS_DEBUG_OSPF_EVENT)
+				zlog_debug("%s: merging the new route",
+					   __func__);
+
+			ospf_route_copy_nexthops(or, abr_or->paths);
+			ospf_route_free(new_or);
+			return;
+		}
+		/* New route is worse */
+		else {
+			if (IS_DEBUG_OSPF_EVENT)
+				zlog_debug("%s: skipping the new route",
+					   __func__);
+			ospf_route_free(new_or);
+			return;
+		}
+	}
+
+	ospf_route_copy_nexthops(new_or, abr_or->paths);
+
+	if (IS_DEBUG_OSPF_EVENT)
+		zlog_debug("%s: adding the new route", __func__);
+
+	listnode_add(rn->info, new_or);
+}
+
+
+static int process_summary_lsa(struct ospf_area *area, struct route_table *rt,
+			       struct route_table *rtrs, struct ospf_lsa *lsa)
+{
+	struct ospf *ospf = area->ospf;
+	struct ospf_area_range *range;
+	struct ospf_route *abr_or, *new_or;
+	struct summary_lsa *sl;
+	struct prefix_ipv4 p, abr;
+	uint32_t metric;
+
+	if (lsa == NULL)
+		return 0;
+
+	sl = (struct summary_lsa *)lsa->data;
+
+	if (IS_DEBUG_OSPF_EVENT)
+		zlog_debug("%s: LS ID: %pI4", __func__, &sl->header.id);
+
+	metric = GET_METRIC(sl->metric);
+
+	if (metric == OSPF_LS_INFINITY)
+		return 0;
+
+	if (IS_LSA_MAXAGE(lsa))
+		return 0;
+
+	if (ospf_lsa_is_self_originated(area->ospf, lsa))
+		return 0;
+
+	p.family = AF_INET;
+	p.prefix = sl->header.id;
+
+	if (sl->header.type == OSPF_SUMMARY_LSA)
+		p.prefixlen = ip_masklen(sl->mask);
+	else
+		p.prefixlen = IPV4_MAX_BITLEN;
+
+	apply_mask_ipv4(&p);
+
+	if (sl->header.type == OSPF_SUMMARY_LSA
+	    && (range = ospf_area_range_match_any(ospf, &p))
+	    && ospf_area_range_active(range))
+		return 0;
+
+	/* XXX: This check seems dubious to me. If an ABR has already decided
+	 * to consider summaries received in this area, then why would one wish
+	 * to exclude default?
+	 */
+	if (IS_OSPF_ABR(ospf) && ospf->abr_type != OSPF_ABR_STAND
+	    && area->external_routing != OSPF_AREA_DEFAULT
+	    && p.prefix.s_addr == OSPF_DEFAULT_DESTINATION && p.prefixlen == 0)
+		return 0; /* Ignore summary default from a stub area */
+
+	abr.family = AF_INET;
+	abr.prefix = sl->header.adv_router;
+	abr.prefixlen = IPV4_MAX_BITLEN;
+	apply_mask_ipv4(&abr);
+
+	abr_or = ospf_find_abr_route(rtrs, &abr, area);
+
+	if (abr_or == NULL)
+		return 0;
+
+	new_or = ospf_route_new();
+	new_or->type = OSPF_DESTINATION_NETWORK;
+	new_or->id = sl->header.id;
+	new_or->mask = sl->mask;
+	new_or->u.std.options = sl->header.options;
+	new_or->u.std.origin = (struct lsa_header *)sl;
+	new_or->cost = abr_or->cost + metric;
+	new_or->u.std.area_id = area->area_id;
+	new_or->u.std.external_routing = area->external_routing;
+	new_or->path_type = OSPF_PATH_INTER_AREA;
+
+	if (sl->header.type == OSPF_SUMMARY_LSA)
+		ospf_ia_network_route(ospf, rt, &p, new_or, abr_or);
+	else {
+		new_or->type = OSPF_DESTINATION_ROUTER;
+		new_or->u.std.flags = ROUTER_LSA_EXTERNAL;
+		ospf_ia_router_route(ospf, rtrs, &p, new_or, abr_or);
+	}
+
+	return 0;
+}
+
+static void ospf_examine_summaries(struct ospf_area *area,
+				   struct route_table *lsdb_rt,
+				   struct route_table *rt,
+				   struct route_table *rtrs)
+{
+	struct ospf_lsa *lsa;
+	struct route_node *rn;
+
+	LSDB_LOOP (lsdb_rt, rn, lsa)
+		process_summary_lsa(area, rt, rtrs, lsa);
+}
+
+int ospf_area_is_transit(struct ospf_area *area)
+{
+	return (area->transit == OSPF_TRANSIT_TRUE)
+	       || ospf_full_virtual_nbrs(
+			  area); /* Cisco forgets to set the V-bit :( */
+}
+
+static void ospf_update_network_route(struct ospf *ospf, struct route_table *rt,
+				      struct route_table *rtrs,
+				      struct summary_lsa *lsa,
+				      struct prefix_ipv4 *p,
+				      struct ospf_area *area)
+{
+	struct route_node *rn;
+	struct ospf_route * or, *abr_or, *new_or;
+	struct prefix_ipv4 abr;
+	uint32_t cost;
+
+	abr.family = AF_INET;
+	abr.prefix = lsa->header.adv_router;
+	abr.prefixlen = IPV4_MAX_BITLEN;
+	apply_mask_ipv4(&abr);
+
+	abr_or = ospf_find_abr_route(rtrs, &abr, area);
+
+	if (abr_or == NULL) {
+		if (IS_DEBUG_OSPF_EVENT)
+			zlog_debug("%s: can't find a route to the ABR",
+				   __func__);
+		return;
+	}
+
+	cost = abr_or->cost + GET_METRIC(lsa->metric);
+
+	rn = route_node_lookup(rt, (struct prefix *)p);
+
+	if (!rn) {
+		if (ospf->abr_type != OSPF_ABR_SHORTCUT)
+			return; /* Standard ABR can update only already
+				   installed
+				   backbone paths */
+		if (IS_DEBUG_OSPF_EVENT)
+			zlog_debug("%s: Allowing Shortcut ABR to add new route",
+				   __func__);
+		new_or = ospf_route_new();
+		new_or->type = OSPF_DESTINATION_NETWORK;
+		new_or->id = lsa->header.id;
+		new_or->mask = lsa->mask;
+		new_or->u.std.options = lsa->header.options;
+		new_or->u.std.origin = (struct lsa_header *)lsa;
+		new_or->cost = cost;
+		new_or->u.std.area_id = area->area_id;
+		new_or->u.std.external_routing = area->external_routing;
+		new_or->path_type = OSPF_PATH_INTER_AREA;
+		ospf_route_add(rt, p, new_or, abr_or);
+
+		return;
+	} else {
+		route_unlock_node(rn);
+		if (rn->info == NULL)
+			return;
+	}
+
+	or = rn->info;
+
+	if (or->path_type != OSPF_PATH_INTRA_AREA &&
+	    or->path_type != OSPF_PATH_INTER_AREA) {
+		if (IS_DEBUG_OSPF_EVENT)
+			zlog_debug("%s: ERR: path type is wrong", __func__);
+		return;
+	}
+
+	if (ospf->abr_type == OSPF_ABR_SHORTCUT) {
+		if (
+			or->path_type == OSPF_PATH_INTRA_AREA
+				  && !OSPF_IS_AREA_ID_BACKBONE(
+					     or->u.std.area_id)) {
+			if (IS_DEBUG_OSPF_EVENT)
+				zlog_debug(
+					"%s: Shortcut: this intra-area path is not backbone",
+					__func__);
+			return;
+		}
+	} else /* Not Shortcut ABR */
+	{
+		if (!OSPF_IS_AREA_ID_BACKBONE(or->u.std.area_id)) {
+			if (IS_DEBUG_OSPF_EVENT)
+				zlog_debug("%s: route is not BB-associated",
+					   __func__);
+			return; /* We can update only BB routes */
+		}
+	}
+
+	if (or->cost < cost) {
+		if (IS_DEBUG_OSPF_EVENT)
+			zlog_debug("%s: new route is worse", __func__);
+		return;
+	}
+
+	if (or->cost == cost) {
+		if (IS_DEBUG_OSPF_EVENT)
+			zlog_debug(
+				"%s: new route is same distance, adding nexthops",
+				__func__);
+		ospf_route_copy_nexthops(or, abr_or->paths);
+	}
+
+	if (or->cost > cost) {
+		if (IS_DEBUG_OSPF_EVENT)
+			zlog_debug(
+				"%s: new route is better, overriding nexthops",
+				__func__);
+		ospf_route_subst_nexthops(or, abr_or->paths);
+		or->cost = cost;
+
+		if ((ospf->abr_type == OSPF_ABR_SHORTCUT)
+		    && !OSPF_IS_AREA_ID_BACKBONE(or->u.std.area_id)) {
+			or->path_type = OSPF_PATH_INTER_AREA;
+			or->u.std.area_id = area->area_id;
+			or->u.std.external_routing = area->external_routing;
+			/* Note that we can do this only in Shortcut ABR mode,
+			   because standard ABR must leave the route type and
+			   area
+			   unchanged
+			*/
+		}
+	}
+}
+
+static void ospf_update_router_route(struct ospf *ospf,
+				     struct route_table *rtrs,
+				     struct summary_lsa *lsa,
+				     struct prefix_ipv4 *p,
+				     struct ospf_area *area)
+{
+	struct ospf_route * or, *abr_or, *new_or;
+	struct prefix_ipv4 abr;
+	uint32_t cost;
+
+	abr.family = AF_INET;
+	abr.prefix = lsa->header.adv_router;
+	abr.prefixlen = IPV4_MAX_BITLEN;
+	apply_mask_ipv4(&abr);
+
+	abr_or = ospf_find_abr_route(rtrs, &abr, area);
+
+	if (abr_or == NULL) {
+		if (IS_DEBUG_OSPF_EVENT)
+			zlog_debug("%s: can't find a route to the ABR",
+				   __func__);
+		return;
+	}
+
+	cost = abr_or->cost + GET_METRIC(lsa->metric);
+
+	/* First try to find a backbone path,
+	   because standard ABR can update only BB-associated paths */
+
+	if ((ospf->backbone == NULL) && (ospf->abr_type != OSPF_ABR_SHORTCUT))
+		return; /* no BB area, not Shortcut ABR, exiting */
+
+	/* find the backbone route, if possible */
+	if ((ospf->backbone == NULL)
+	    || !(or = ospf_find_asbr_route_through_area(rtrs, p,
+							ospf->backbone))) {
+		if (ospf->abr_type != OSPF_ABR_SHORTCUT)
+
+			/* route to ASBR through the BB not found
+			   the router is not Shortcut ABR, exiting */
+
+			return;
+		else
+		/* We're a Shortcut ABR*/
+		{
+			/* Let it either add a new router or update the route
+			   through the same (non-BB) area. */
+
+			new_or = ospf_route_new();
+			new_or->type = OSPF_DESTINATION_ROUTER;
+			new_or->id = lsa->header.id;
+			new_or->mask = lsa->mask;
+			new_or->u.std.options = lsa->header.options;
+			new_or->u.std.origin = (struct lsa_header *)lsa;
+			new_or->cost = cost;
+			new_or->u.std.area_id = area->area_id;
+			new_or->u.std.external_routing = area->external_routing;
+			new_or->path_type = OSPF_PATH_INTER_AREA;
+			new_or->u.std.flags = ROUTER_LSA_EXTERNAL;
+			ospf_ia_router_route(ospf, rtrs, p, new_or, abr_or);
+
+			return;
+		}
+	}
+
+	/* At this point the "or" is always bb-associated */
+
+	if (!(or->u.std.flags & ROUTER_LSA_EXTERNAL)) {
+		if (IS_DEBUG_OSPF_EVENT)
+			zlog_debug("%s: the remote router is not an ASBR",
+				   __func__);
+		return;
+	}
+
+	if (or->path_type != OSPF_PATH_INTRA_AREA &&
+	    or->path_type != OSPF_PATH_INTER_AREA)
+		return;
+
+	if (or->cost < cost)
+		return;
+
+	else if (or->cost == cost)
+		ospf_route_copy_nexthops(or, abr_or->paths);
+
+	else if (or->cost > cost) {
+		ospf_route_subst_nexthops(or, abr_or->paths);
+		or->cost = cost;
+
+		/* Even if the ABR runs in Shortcut mode, we can't change
+		   the path type and area, because the "or" is always
+		   bb-associated
+		   at this point and even Shortcut ABR can't change these
+		   attributes */
+	}
+}
+
+static int process_transit_summary_lsa(struct ospf_area *area,
+				       struct route_table *rt,
+				       struct route_table *rtrs,
+				       struct ospf_lsa *lsa)
+{
+	struct ospf *ospf = area->ospf;
+	struct summary_lsa *sl;
+	struct prefix_ipv4 p;
+	uint32_t metric;
+
+	if (lsa == NULL)
+		return 0;
+
+	sl = (struct summary_lsa *)lsa->data;
+
+	if (IS_DEBUG_OSPF_EVENT)
+		zlog_debug("%s: LS ID: %pI4", __func__, &lsa->data->id);
+	metric = GET_METRIC(sl->metric);
+
+	if (metric == OSPF_LS_INFINITY) {
+		if (IS_DEBUG_OSPF_EVENT)
+			zlog_debug("%s: metric is infinity, skip", __func__);
+		return 0;
+	}
+
+	if (IS_LSA_MAXAGE(lsa)) {
+		if (IS_DEBUG_OSPF_EVENT)
+			zlog_debug("%s: This LSA is too old", __func__);
+		return 0;
+	}
+
+	if (ospf_lsa_is_self_originated(area->ospf, lsa)) {
+		if (IS_DEBUG_OSPF_EVENT)
+			zlog_debug("%s: This LSA is mine, skip", __func__);
+		return 0;
+	}
+
+	p.family = AF_INET;
+	p.prefix = sl->header.id;
+
+	if (sl->header.type == OSPF_SUMMARY_LSA)
+		p.prefixlen = ip_masklen(sl->mask);
+	else
+		p.prefixlen = IPV4_MAX_BITLEN;
+
+	apply_mask_ipv4(&p);
+
+	if (sl->header.type == OSPF_SUMMARY_LSA)
+		ospf_update_network_route(ospf, rt, rtrs, sl, &p, area);
+	else
+		ospf_update_router_route(ospf, rtrs, sl, &p, area);
+
+	return 0;
+}
+
+static void ospf_examine_transit_summaries(struct ospf_area *area,
+					   struct route_table *lsdb_rt,
+					   struct route_table *rt,
+					   struct route_table *rtrs)
+{
+	struct ospf_lsa *lsa;
+	struct route_node *rn;
+
+	LSDB_LOOP (lsdb_rt, rn, lsa)
+		process_transit_summary_lsa(area, rt, rtrs, lsa);
+}
+
+void ospf_ia_routing(struct ospf *ospf, struct route_table *rt,
+		     struct route_table *rtrs)
+{
+	struct listnode *node;
+	struct ospf_area *area;
+
+	if (IS_DEBUG_OSPF_EVENT)
+		zlog_debug("%s:start", __func__);
+
+	if (IS_OSPF_ABR(ospf)) {
+		switch (ospf->abr_type) {
+		case OSPF_ABR_STAND:
+			if (IS_DEBUG_OSPF_EVENT)
+				zlog_debug("%s:Standard ABR", __func__);
+
+			if ((area = ospf->backbone)) {
+				if (IS_DEBUG_OSPF_EVENT) {
+					zlog_debug(
+						"%s:backbone area found, examining summaries",
+						__func__);
+				}
+
+				OSPF_EXAMINE_SUMMARIES_ALL(area, rt, rtrs);
+
+				for (ALL_LIST_ELEMENTS_RO(ospf->areas, node,
+							  area))
+					if (area != ospf->backbone)
+						if (ospf_area_is_transit(area))
+							OSPF_EXAMINE_TRANSIT_SUMMARIES_ALL(
+								area, rt, rtrs);
+			} else if (IS_DEBUG_OSPF_EVENT)
+				zlog_debug("%s:backbone area NOT found",
+					   __func__);
+			break;
+		case OSPF_ABR_IBM:
+		case OSPF_ABR_CISCO:
+			if (IS_DEBUG_OSPF_EVENT)
+				zlog_debug("%s:Alternative Cisco/IBM ABR",
+					   __func__);
+			area = ospf->backbone; /* Find the BB */
+
+			/* If we have an active BB connection */
+			if (area && ospf_act_bb_connection(ospf)) {
+				if (IS_DEBUG_OSPF_EVENT) {
+					zlog_debug(
+						"%s: backbone area found, examining BB summaries",
+						__func__);
+				}
+
+				OSPF_EXAMINE_SUMMARIES_ALL(area, rt, rtrs);
+
+				for (ALL_LIST_ELEMENTS_RO(ospf->areas, node,
+							  area))
+					if (area != ospf->backbone)
+						if (ospf_area_is_transit(area))
+							OSPF_EXAMINE_TRANSIT_SUMMARIES_ALL(
+								area, rt, rtrs);
+			} else { /* No active BB connection--consider all areas
+				    */
+				if (IS_DEBUG_OSPF_EVENT)
+					zlog_debug(
+						"%s: Active BB connection not found",
+						__func__);
+				for (ALL_LIST_ELEMENTS_RO(ospf->areas, node,
+							  area))
+					OSPF_EXAMINE_SUMMARIES_ALL(area, rt,
+								   rtrs);
+			}
+			break;
+		case OSPF_ABR_SHORTCUT:
+			if (IS_DEBUG_OSPF_EVENT)
+				zlog_debug("%s:Alternative Shortcut", __func__);
+			area = ospf->backbone; /* Find the BB */
+
+			/* If we have an active BB connection */
+			if (area && ospf_act_bb_connection(ospf)) {
+				if (IS_DEBUG_OSPF_EVENT) {
+					zlog_debug(
+						"%s: backbone area found, examining BB summaries",
+						__func__);
+				}
+				OSPF_EXAMINE_SUMMARIES_ALL(area, rt, rtrs);
+			}
+
+			for (ALL_LIST_ELEMENTS_RO(ospf->areas, node, area))
+				if (area != ospf->backbone)
+					if (ospf_area_is_transit(area)
+					    || ((area->shortcut_configured
+						 != OSPF_SHORTCUT_DISABLE)
+						&& ((ospf->backbone == NULL)
+						    || ((area->shortcut_configured
+							 == OSPF_SHORTCUT_ENABLE)
+							&& area->shortcut_capability))))
+						OSPF_EXAMINE_TRANSIT_SUMMARIES_ALL(
+							area, rt, rtrs);
+			break;
+		default:
+			break;
+		}
+	} else {
+		if (IS_DEBUG_OSPF_EVENT)
+			zlog_debug("%s:not ABR, considering all areas",
+				   __func__);
+
+		for (ALL_LIST_ELEMENTS_RO(ospf->areas, node, area))
+			OSPF_EXAMINE_SUMMARIES_ALL(area, rt, rtrs);
+	}
+}