[TIPC]: Fixed link switchover bugs

Incorporates several related fixes: - switchover now occurs when switching from an active link to a standby link - failure of a standby link no longer initiates switchover - links now display correct # of received packtes following reactivation Signed-off-by: Allan Stephens <allan.stephens@windriver.com> Signed-off-by: Per Liden <per.liden@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>

[TIPC]: Fixed link switchover bugs
Incorporates several related fixes: - switchover now occurs when switching from an active link to a standby link - failure of a standby link no longer initiates switchover - links now display correct # of received packtes following reactivation Signed-off-by: Allan Stephens <allan.stephens@windriver.com> Signed-off-by: Per Liden <per.liden@ericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Allan Stephens · David S. Miller
1 parent a10bd924a4
Showing 3 changed files with 31 additions and 8 deletions Side-by-side Diff
net/tipc/link.c
net/tipc/node.c
net/tipc/node.h
@@ -690,6 +690,7 @@
 	struct sk_buff *buf;
 	u32 prev_state = l_ptr->state;
 	u32 checkpoint = l_ptr->next_in_no;
+	int was_active_link = tipc_link_is_active(l_ptr);
  
 	msg_set_session(l_ptr->pmsg, msg_session(l_ptr->pmsg) + 1);
  
@@ -711,7 +712,7 @@
 	tipc_printf(TIPC_CONS, "\nReset link <%s>\n", l_ptr->name);
 	dbg_link_dump();
 #endif
-	if (tipc_node_has_active_links(l_ptr->owner) &&
+	if (was_active_link && tipc_node_has_active_links(l_ptr->owner) &&
 	    l_ptr->owner->permit_changeover) {
 		l_ptr->reset_checkpoint = checkpoint;
 		l_ptr->exp_msg_count = START_CHANGEOVER;
@@ -754,7 +755,7 @@
  
 static void link_activate(struct link *l_ptr)
 {
-	l_ptr->next_in_no = 1;
+	l_ptr->next_in_no = l_ptr->stats.recv_info = 1;
 	tipc_node_link_up(l_ptr->owner, l_ptr);
 	tipc_bearer_add_dest(l_ptr->b_ptr, l_ptr->addr);
 	link_send_event(tipc_cfg_link_event, l_ptr, 1);
  
  
  
@@ -2303,12 +2304,18 @@
 	u32 length = msg_size(msg);
  
 	tunnel = l_ptr->owner->active_links[selector & 1];
-	if (!tipc_link_is_up(tunnel))
+	if (!tipc_link_is_up(tunnel)) {
+		warn("Link changeover error, "
+		     "tunnel link no longer available\n");
 		return;
+	}
 	msg_set_size(tunnel_hdr, length + INT_H_SIZE);
 	buf = buf_acquire(length + INT_H_SIZE);
-	if (!buf)
+	if (!buf) {
+		warn("Link changeover error, "
+		     "unable to send tunnel msg\n");
 		return;
+	}
 	memcpy(buf->data, (unchar *)tunnel_hdr, INT_H_SIZE);
 	memcpy(buf->data + INT_H_SIZE, (unchar *)msg, length);
 	dbg("%c->%c:", l_ptr->b_ptr->net_plane, tunnel->b_ptr->net_plane);
  
  
  
  
@@ -2328,19 +2335,23 @@
 	u32 msgcount = l_ptr->out_queue_size;
 	struct sk_buff *crs = l_ptr->first_out;
 	struct link *tunnel = l_ptr->owner->active_links[0];
-	int split_bundles = tipc_node_has_redundant_links(l_ptr->owner);
 	struct tipc_msg tunnel_hdr;
+	int split_bundles;
  
 	if (!tunnel)
 		return;
  
-	if (!l_ptr->owner->permit_changeover)
+	if (!l_ptr->owner->permit_changeover) {
+		warn("Link changeover error, "
+		     "peer did not permit changeover\n");
 		return;
+	}
  
 	msg_init(&tunnel_hdr, CHANGEOVER_PROTOCOL,
 		 ORIGINAL_MSG, TIPC_OK, INT_H_SIZE, l_ptr->addr);
 	msg_set_bearer_id(&tunnel_hdr, l_ptr->peer_bearer_id);
 	msg_set_msgcnt(&tunnel_hdr, msgcount);
+	dbg("Link changeover requires %u tunnel messages\n", msgcount);
  
 	if (!l_ptr->first_out) {
 		struct sk_buff *buf;
@@ -2360,6 +2371,9 @@
 		return;
 	}
  
+	split_bundles = (l_ptr->owner->active_links[0] != 
+			 l_ptr->owner->active_links[1]);
+
 	while (crs) {
 		struct tipc_msg *msg = buf_msg(crs);
  
  
@@ -2497,11 +2511,13 @@
 		     dest_link->name);
 		tipc_link_reset(dest_link);
 		dest_link->exp_msg_count = msg_count;
+		dbg("Expecting %u tunnelled messages\n", msg_count);
 		if (!msg_count)
 			goto exit;
 	} else if (dest_link->exp_msg_count == START_CHANGEOVER) {
 		msg_dbg(tunnel_msg, "BLK/FIRST/<REC<");
 		dest_link->exp_msg_count = msg_count;
+		dbg("Expecting %u tunnelled messages\n", msg_count);
 		if (!msg_count)
 			goto exit;
 	}
@@ -2509,6 +2525,8 @@
 	/* Receive original message */
  
 	if (dest_link->exp_msg_count == 0) {
+		warn("Link switchover error, "
+		     "got too many tunnelled messages\n");
 		msg_dbg(tunnel_msg, "OVERDUE/DROP/<REC<");
 		dbg_print_link(dest_link, "LINK:");
 		goto exit;
@@ -125,6 +125,8 @@
 {
 	struct link **active = &n_ptr->active_links[0];
  
+	n_ptr->working_links++;
+
 	info("Established link <%s> on network plane %c\n",
 	     l_ptr->name, l_ptr->b_ptr->net_plane);
  
@@ -185,6 +187,8 @@
 {
 	struct link **active;
  
+	n_ptr->working_links--;
+
 	if (!tipc_link_is_active(l_ptr)) {
 		info("Lost standby link <%s> on network plane %c\n",
 		     l_ptr->name, l_ptr->b_ptr->net_plane);
@@ -214,8 +218,7 @@
  
 int tipc_node_has_redundant_links(struct node *n_ptr)
 {
-	return (tipc_node_has_active_links(n_ptr) &&
-		(n_ptr->active_links[0] != n_ptr->active_links[1]));
+	return (n_ptr->working_links > 1);
 }
  
 static int tipc_node_has_active_routes(struct node *n_ptr)
@@ -51,6 +51,7 @@
  * @nsub: list of "node down" subscriptions monitoring node
  * @active_links: pointers to active links to node
  * @links: pointers to all links to node
+ * @working_links: number of working links to node (both active and standby)
  * @link_cnt: number of links to node
  * @permit_changeover: non-zero if node has redundant links to this system
  * @routers: bitmap (used for multicluster communication)
@@ -76,6 +77,7 @@
 	struct link *active_links[2];
 	struct link *links[MAX_BEARERS];
 	int link_cnt;
+	int working_links;
 	int permit_changeover;
 	u32 routers[512/32];
 	int last_router;
...	...	@@ -690,6 +690,7 @@
690	690	struct sk_buff *buf;
691	691	u32 prev_state = l_ptr->state;
692	692	u32 checkpoint = l_ptr->next_in_no;
	693	+ int was_active_link = tipc_link_is_active(l_ptr);
693	694
694	695	msg_set_session(l_ptr->pmsg, msg_session(l_ptr->pmsg) + 1);
695	696
...	...	@@ -711,7 +712,7 @@
711	712	tipc_printf(TIPC_CONS, "\nReset link <%s>\n", l_ptr->name);
712	713	dbg_link_dump();
713	714	#endif
714		- if (tipc_node_has_active_links(l_ptr->owner) &&
	715	+ if (was_active_link && tipc_node_has_active_links(l_ptr->owner) &&
715	716	l_ptr->owner->permit_changeover) {
716	717	l_ptr->reset_checkpoint = checkpoint;
717	718	l_ptr->exp_msg_count = START_CHANGEOVER;
...	...	@@ -754,7 +755,7 @@
754	755
755	756	static void link_activate(struct link *l_ptr)
756	757	{
757		- l_ptr->next_in_no = 1;
	758	+ l_ptr->next_in_no = l_ptr->stats.recv_info = 1;
758	759	tipc_node_link_up(l_ptr->owner, l_ptr);
759	760	tipc_bearer_add_dest(l_ptr->b_ptr, l_ptr->addr);
760	761	link_send_event(tipc_cfg_link_event, l_ptr, 1);
761	762
762	763
763	764
...	...	@@ -2303,12 +2304,18 @@
2303	2304	u32 length = msg_size(msg);
2304	2305
2305	2306	tunnel = l_ptr->owner->active_links[selector & 1];
2306		- if (!tipc_link_is_up(tunnel))
	2307	+ if (!tipc_link_is_up(tunnel)) {
	2308	+ warn("Link changeover error, "
	2309	+ "tunnel link no longer available\n");
2307	2310	return;
	2311	+ }
2308	2312	msg_set_size(tunnel_hdr, length + INT_H_SIZE);
2309	2313	buf = buf_acquire(length + INT_H_SIZE);
2310		- if (!buf)
	2314	+ if (!buf) {
	2315	+ warn("Link changeover error, "
	2316	+ "unable to send tunnel msg\n");
2311	2317	return;
	2318	+ }
2312	2319	memcpy(buf->data, (unchar *)tunnel_hdr, INT_H_SIZE);
2313	2320	memcpy(buf->data + INT_H_SIZE, (unchar *)msg, length);
2314	2321	dbg("%c->%c:", l_ptr->b_ptr->net_plane, tunnel->b_ptr->net_plane);
2315	2322
2316	2323
2317	2324
2318	2325
...	...	@@ -2328,19 +2335,23 @@
2328	2335	u32 msgcount = l_ptr->out_queue_size;
2329	2336	struct sk_buff *crs = l_ptr->first_out;
2330	2337	struct link *tunnel = l_ptr->owner->active_links[0];
2331		- int split_bundles = tipc_node_has_redundant_links(l_ptr->owner);
2332	2338	struct tipc_msg tunnel_hdr;
	2339	+ int split_bundles;
2333	2340
2334	2341	if (!tunnel)
2335	2342	return;
2336	2343
2337		- if (!l_ptr->owner->permit_changeover)
	2344	+ if (!l_ptr->owner->permit_changeover) {
	2345	+ warn("Link changeover error, "
	2346	+ "peer did not permit changeover\n");
2338	2347	return;
	2348	+ }
2339	2349
2340	2350	msg_init(&tunnel_hdr, CHANGEOVER_PROTOCOL,
2341	2351	ORIGINAL_MSG, TIPC_OK, INT_H_SIZE, l_ptr->addr);
2342	2352	msg_set_bearer_id(&tunnel_hdr, l_ptr->peer_bearer_id);
2343	2353	msg_set_msgcnt(&tunnel_hdr, msgcount);
	2354	+ dbg("Link changeover requires %u tunnel messages\n", msgcount);
2344	2355
2345	2356	if (!l_ptr->first_out) {
2346	2357	struct sk_buff *buf;
...	...	@@ -2360,6 +2371,9 @@
2360	2371	return;
2361	2372	}
2362	2373
	2374	+ split_bundles = (l_ptr->owner->active_links[0] !=
	2375	+ l_ptr->owner->active_links[1]);
	2376	+
2363	2377	while (crs) {
2364	2378	struct tipc_msg *msg = buf_msg(crs);
2365	2379
2366	2380
...	...	@@ -2497,11 +2511,13 @@
2497	2511	dest_link->name);
2498	2512	tipc_link_reset(dest_link);
2499	2513	dest_link->exp_msg_count = msg_count;
	2514	+ dbg("Expecting %u tunnelled messages\n", msg_count);
2500	2515	if (!msg_count)
2501	2516	goto exit;
2502	2517	} else if (dest_link->exp_msg_count == START_CHANGEOVER) {
2503	2518	msg_dbg(tunnel_msg, "BLK/FIRST/<REC<");
2504	2519	dest_link->exp_msg_count = msg_count;
	2520	+ dbg("Expecting %u tunnelled messages\n", msg_count);
2505	2521	if (!msg_count)
2506	2522	goto exit;
2507	2523	}
...	...	@@ -2509,6 +2525,8 @@
2509	2525	/* Receive original message */
2510	2526
2511	2527	if (dest_link->exp_msg_count == 0) {
	2528	+ warn("Link switchover error, "
	2529	+ "got too many tunnelled messages\n");
2512	2530	msg_dbg(tunnel_msg, "OVERDUE/DROP/<REC<");
2513	2531	dbg_print_link(dest_link, "LINK:");
2514	2532	goto exit;
...	...	@@ -125,6 +125,8 @@
125	125	{
126	126	struct link **active = &n_ptr->active_links[0];
127	127
	128	+ n_ptr->working_links++;
	129	+
128	130	info("Established link <%s> on network plane %c\n",
129	131	l_ptr->name, l_ptr->b_ptr->net_plane);
130	132
...	...	@@ -185,6 +187,8 @@
185	187	{
186	188	struct link **active;
187	189
	190	+ n_ptr->working_links--;
	191	+
188	192	if (!tipc_link_is_active(l_ptr)) {
189	193	info("Lost standby link <%s> on network plane %c\n",
190	194	l_ptr->name, l_ptr->b_ptr->net_plane);
...	...	@@ -214,8 +218,7 @@
214	218
215	219	int tipc_node_has_redundant_links(struct node *n_ptr)
216	220	{
217		- return (tipc_node_has_active_links(n_ptr) &&
218		- (n_ptr->active_links[0] != n_ptr->active_links[1]));
	221	+ return (n_ptr->working_links > 1);
219	222	}
220	223
221	224	static int tipc_node_has_active_routes(struct node *n_ptr)
...	...	@@ -51,6 +51,7 @@
51	51	* @nsub: list of "node down" subscriptions monitoring node
52	52	* @active_links: pointers to active links to node
53	53	* @links: pointers to all links to node
	54	+ * @working_links: number of working links to node (both active and standby)
54	55	* @link_cnt: number of links to node
55	56	* @permit_changeover: non-zero if node has redundant links to this system
56	57	* @routers: bitmap (used for multicluster communication)
...	...	@@ -76,6 +77,7 @@
76	77	struct link *active_links[2];
77	78	struct link *links[MAX_BEARERS];
78	79	int link_cnt;
	80	+ int working_links;
79	81	int permit_changeover;
80	82	u32 routers[512/32];
81	83	int last_router;