[Cluster-devel] cluster/cmirror-kernel/src dm-cmirror-client.c ...

[jbrassow at sourceware.org]

CVSROOT:	/cvs/cluster
Module name:	cluster
Branch: 	RHEL4
Changes by:	jbrassow at sourceware.org	2007-02-26 17:38:06

Modified files:
	cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-server.c 

Log message:
	add locking around the log list.  There was a small window of opportunity
	for the log server to look up a log in the list while another entry was
	being deleted (bad for the server).
	
	Bug 229715 Processed: cmirror panic in dm_cmirror:cluster_log_serverd

Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.38&r2=1.1.2.39
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.24&r2=1.1.2.25

--- cluster/cmirror-kernel/src/Attic/dm-cmirror-client.c	2007/02/20 19:35:10	1.1.2.38
+++ cluster/cmirror-kernel/src/Attic/dm-cmirror-client.c	2007/02/26 17:38:06	1.1.2.39
@@ -28,6 +28,7 @@
 #include "dm-cmirror-server.h"
 #include "dm-cmirror-cman.h"
 
+spinlock_t log_list_lock;
 LIST_HEAD(log_list_head);
 
 struct region_state {
@@ -635,6 +636,7 @@
 	atomic_set(&lc->in_sync, -1);
 	lc->uuid_ref = 1;
 
+	spin_lock(&log_list_lock);
 	list_for_each_entry(tmp_lc, &log_list_head, log_list){
 		if(!strncmp(tmp_lc->uuid, lc->uuid, MAX_NAME_LEN)){
 			lc->uuid_ref = (lc->uuid_ref > tmp_lc->uuid_ref) ?
@@ -647,6 +649,7 @@
 	       lc->uuid_ref);
 
 	list_add(&lc->log_list, &log_list_head);
+	spin_unlock(&log_list_lock);
 	INIT_LIST_HEAD(&lc->region_users);
 
 	lc->server_id = 0xDEAD;
@@ -757,7 +760,9 @@
 	if (!list_empty(&clear_region_list))
 		DMINFO("Leaving while clear region requests remain.");
 
+	spin_lock(&log_list_lock);
 	list_del_init(&lc->log_list);
+	spin_unlock(&log_list_lock);
 
 	if ((lc->server_id == my_id) && !atomic_read(&lc->suspended))
 		consult_server(lc, 0, LRT_MASTER_LEAVING, NULL);
@@ -1204,9 +1209,11 @@
 
 	atomic_set(&suspend_client, 1);
 
+	spin_lock(&log_list_lock);
 	list_for_each_entry(lc, &log_list_head, log_list) {
 		atomic_set(&lc->in_sync, 0);
 	}
+	spin_unlock(&log_list_lock);
 	
 	if (likely(!shutting_down))
 		suspend_server();
@@ -1238,6 +1245,7 @@
 	switch(type){
 	case SERVICE_NODE_LEAVE:
 	case SERVICE_NODE_FAILED:
+		spin_lock(&log_list_lock);
 		list_for_each_entry(lc, &log_list_head, log_list){
 			for(i=0, server = 0xDEAD; i < count; i++){
 				if(lc->server_id == nodeids[i]){
@@ -1247,6 +1255,8 @@
 			/* ATTENTION -- need locking around this ? */
 			lc->server_id = server;
 		}
+		spin_unlock(&log_list_lock);
+
 		break;
 	case SERVICE_NODE_JOIN:
 		break;
@@ -1387,6 +1397,7 @@
 	INIT_LIST_HEAD(&marked_region_list);
 
 	spin_lock_init(&region_state_lock);
+	spin_lock_init(&log_list_lock);
 	region_state_pool = mempool_create(20, region_state_alloc,
 					   region_state_free, NULL);
 	if(!region_state_pool){
--- cluster/cmirror-kernel/src/Attic/dm-cmirror-server.c	2007/02/21 17:14:44	1.1.2.24
+++ cluster/cmirror-kernel/src/Attic/dm-cmirror-server.c	2007/02/26 17:38:06	1.1.2.25
@@ -47,6 +47,7 @@
 static atomic_t _do_requests;
 
 static int debug_disk_write = 0;
+extern spinlock_t log_list_lock;
 extern struct list_head log_list_head;
 
 static void *region_user_alloc(int gfp_mask, void *pool_data){
@@ -649,6 +650,7 @@
 static struct log_c *get_log_context(char *uuid, int uuid_ref){
 	struct log_c *lc, *r = NULL;
 
+	spin_lock(&log_list_lock);
 	list_for_each_entry(lc, &log_list_head, log_list){
 		if (!strncmp(lc->uuid, uuid, MAX_NAME_LEN) &&
 		    (uuid_ref == lc->uuid_ref)) {
@@ -658,6 +660,7 @@
 				r = lc;
 		}
 	}
+	spin_unlock(&log_list_lock);
 
 	return r;
 }
@@ -1079,6 +1082,7 @@
 			if (atomic_read(&restart_event_type) == SERVICE_NODE_FAILED)
 				DMINFO("A cluster mirror log member has failed.");
 			
+			spin_lock(&log_list_lock);
 			list_for_each_entry(lc, &log_list_head, log_list){
 				if(lc->server_id == my_id){
 					if (atomic_read(&lc->suspended)) {
@@ -1088,6 +1092,8 @@
 					}
 				}
 			}
+			spin_unlock(&log_list_lock);
+
 			break;
 		default:
 			/* Someone has joined, or there is no event */