1 files changed, 94 insertions, 21 deletions

diff --git a/net/ipv6/ip6_fib.c b/net/ipv6/ip6_fib.c
index 7bae6a91b487..58fbde244381 100644
--- a/net/ipv6/ip6_fib.c
+++ b/net/ipv6/ip6_fib.c
@@ -370,6 +370,21 @@ static int call_fib6_entry_notifier(struct notifier_block *nb,
 	return call_fib6_notifier(nb, event_type, &info.info);
 }
 
+static int call_fib6_multipath_entry_notifier(struct notifier_block *nb,
+					      enum fib_event_type event_type,
+					      struct fib6_info *rt,
+					      unsigned int nsiblings,
+					      struct netlink_ext_ack *extack)
+{
+	struct fib6_entry_notifier_info info = {
+		.info.extack = extack,
+		.rt = rt,
+		.nsiblings = nsiblings,
+	};
+
+	return call_fib6_notifier(nb, event_type, &info.info);
+}
+
 int call_fib6_entry_notifiers(struct net *net,
 			      enum fib_event_type event_type,
 			      struct fib6_info *rt,
@@ -400,6 +415,17 @@ int call_fib6_multipath_entry_notifiers(struct net *net,
 	return call_fib6_notifiers(net, event_type, &info.info);
 }
 
+int call_fib6_entry_notifiers_replace(struct net *net, struct fib6_info *rt)
+{
+	struct fib6_entry_notifier_info info = {
+		.rt = rt,
+		.nsiblings = rt->fib6_nsiblings,
+	};
+
+	rt->fib6_table->fib_seq++;
+	return call_fib6_notifiers(net, FIB_EVENT_ENTRY_REPLACE, &info.info);
+}
+
 struct fib6_dump_arg {
 	struct net *net;
 	struct notifier_block *nb;
@@ -408,22 +434,29 @@ struct fib6_dump_arg {
 
 static int fib6_rt_dump(struct fib6_info *rt, struct fib6_dump_arg *arg)
 {
-	if (rt == arg->net->ipv6.fib6_null_entry)
+	enum fib_event_type fib_event = FIB_EVENT_ENTRY_REPLACE;
+	int err;
+
+	if (!rt || rt == arg->net->ipv6.fib6_null_entry)
 		return 0;
-	return call_fib6_entry_notifier(arg->nb, FIB_EVENT_ENTRY_ADD,
-					rt, arg->extack);
+
+	if (rt->fib6_nsiblings)
+		err = call_fib6_multipath_entry_notifier(arg->nb, fib_event,
+							 rt,
+							 rt->fib6_nsiblings,
+							 arg->extack);
+	else
+		err = call_fib6_entry_notifier(arg->nb, fib_event, rt,
+					       arg->extack);
+
+	return err;
 }
 
 static int fib6_node_dump(struct fib6_walker *w)
 {
-	struct fib6_info *rt;
-	int err = 0;
+	int err;
 
-	for_each_fib6_walker_rt(w) {
-		err = fib6_rt_dump(rt, w->args);
-		if (err)
-			break;
-	}
+	err = fib6_rt_dump(w->leaf, w->args);
 	w->leaf = NULL;
 	return err;
 }
@@ -1039,6 +1072,7 @@ static int fib6_add_rt2node(struct fib6_node *fn, struct fib6_info *rt,
 		   (info->nlh->nlmsg_flags & NLM_F_CREATE));
 	int found = 0;
 	bool rt_can_ecmp = rt6_qualify_for_ecmp(rt);
+	bool notify_sibling_rt = false;
 	u16 nlflags = NLM_F_EXCL;
 	int err;
 
@@ -1130,6 +1164,7 @@ next_iter:
 
 		/* Find the first route that have the same metric */
 		sibling = leaf;
+		notify_sibling_rt = true;
 		while (sibling) {
 			if (sibling->fib6_metric == rt->fib6_metric &&
 			    rt6_qualify_for_ecmp(sibling)) {
@@ -1139,6 +1174,7 @@ next_iter:
 			}
 			sibling = rcu_dereference_protected(sibling->fib6_next,
 				    lockdep_is_held(&rt->fib6_table->tb6_lock));
+			notify_sibling_rt = false;
 		}
 		/* For each sibling in the list, increment the counter of
 		 * siblings. BUG() if counters does not match, list of siblings
@@ -1165,10 +1201,21 @@ next_iter:
 add:
 		nlflags |= NLM_F_CREATE;
 
-		if (!info->skip_notify_kernel) {
+		/* The route should only be notified if it is the first
+		 * route in the node or if it is added as a sibling
+		 * route to the first route in the node.
+		 */
+		if (!info->skip_notify_kernel &&
+		    (notify_sibling_rt || ins == &fn->leaf)) {
+			enum fib_event_type fib_event;
+
+			if (notify_sibling_rt)
+				fib_event = FIB_EVENT_ENTRY_APPEND;
+			else
+				fib_event = FIB_EVENT_ENTRY_REPLACE;
 			err = call_fib6_entry_notifiers(info->nl_net,
-							FIB_EVENT_ENTRY_ADD,
-							rt, extack);
+							fib_event, rt,
+							extack);
 			if (err) {
 				struct fib6_info *sibling, *next_sibling;
 
@@ -1212,7 +1259,7 @@ add:
 			return -ENOENT;
 		}
 
-		if (!info->skip_notify_kernel) {
+		if (!info->skip_notify_kernel && ins == &fn->leaf) {
 			err = call_fib6_entry_notifiers(info->nl_net,
 							FIB_EVENT_ENTRY_REPLACE,
 							rt, extack);
@@ -1845,13 +1892,29 @@ static struct fib6_node *fib6_repair_tree(struct net *net,
 static void fib6_del_route(struct fib6_table *table, struct fib6_node *fn,
 			   struct fib6_info __rcu **rtp, struct nl_info *info)
 {
+	struct fib6_info *leaf, *replace_rt = NULL;
 	struct fib6_walker *w;
 	struct fib6_info *rt = rcu_dereference_protected(*rtp,
 				    lockdep_is_held(&table->tb6_lock));
 	struct net *net = info->nl_net;
+	bool notify_del = false;
 
 	RT6_TRACE("fib6_del_route\n");
 
+	/* If the deleted route is the first in the node and it is not part of
+	 * a multipath route, then we need to replace it with the next route
+	 * in the node, if exists.
+	 */
+	leaf = rcu_dereference_protected(fn->leaf,
+					 lockdep_is_held(&table->tb6_lock));
+	if (leaf == rt && !rt->fib6_nsiblings) {
+		if (rcu_access_pointer(rt->fib6_next))
+			replace_rt = rcu_dereference_protected(rt->fib6_next,
+					    lockdep_is_held(&table->tb6_lock));
+		else
+			notify_del = true;
+	}
+
 	/* Unlink it */
 	*rtp = rt->fib6_next;
 	rt->fib6_node = NULL;
@@ -1869,6 +1932,14 @@ static void fib6_del_route(struct fib6_table *table, struct fib6_node *fn,
 	if (rt->fib6_nsiblings) {
 		struct fib6_info *sibling, *next_sibling;
 
+		/* The route is deleted from a multipath route. If this
+		 * multipath route is the first route in the node, then we need
+		 * to emit a delete notification. Otherwise, we need to skip
+		 * the notification.
+		 */
+		if (rt->fib6_metric == leaf->fib6_metric &&
+		    rt6_qualify_for_ecmp(leaf))
+			notify_del = true;
 		list_for_each_entry_safe(sibling, next_sibling,
 					 &rt->fib6_siblings, fib6_siblings)
 			sibling->fib6_nsiblings--;
@@ -1904,8 +1975,13 @@ static void fib6_del_route(struct fib6_table *table, struct fib6_node *fn,
 
 	fib6_purge_rt(rt, fn, net);
 
-	if (!info->skip_notify_kernel)
-		call_fib6_entry_notifiers(net, FIB_EVENT_ENTRY_DEL, rt, NULL);
+	if (!info->skip_notify_kernel) {
+		if (notify_del)
+			call_fib6_entry_notifiers(net, FIB_EVENT_ENTRY_DEL,
+						  rt, NULL);
+		else if (replace_rt)
+			call_fib6_entry_notifiers_replace(net, replace_rt);
+	}
 	if (!info->skip_notify)
 		inet6_rt_notify(RTM_DELROUTE, rt, info, 0);
 
@@ -2495,14 +2571,13 @@ static void *ipv6_route_seq_next(struct seq_file *seq, void *v, loff_t *pos)
 	struct net *net = seq_file_net(seq);
 	struct ipv6_route_iter *iter = seq->private;
 
+	++(*pos);
 	if (!v)
 		goto iter_table;
 
 	n = rcu_dereference_bh(((struct fib6_info *)v)->fib6_next);
-	if (n) {
-		++*pos;
+	if (n)
 		return n;
-	}
 
 iter_table:
 	ipv6_route_check_sernum(iter);
@@ -2510,8 +2585,6 @@ iter_table:
 	r = fib6_walk_continue(&iter->w);
 	spin_unlock_bh(&iter->tbl->tb6_lock);
 	if (r > 0) {
-		if (v)
-			++*pos;
 		return iter->w.leaf;
 	} else if (r < 0) {
 		fib6_walker_unlink(net, &iter->w);