ocfs2/dlm: do not insert a new mle when another process is already migrating

When two processes are migrating the same lockres,
dlm_add_migration_mle() return -EEXIST, but insert a new mle in hash
list.  dlm_migrate_lockres() will detach the old mle and free the new
one which is already in hash list, that will destroy the list.

Signed-off-by: Jiufei Xue <xuejiufei@huawei.com>
Reviewed-by: Joseph Qi <joseph.qi@huawei.com>
Cc: Mark Fasheh <mfasheh@suse.de>
Cc: Joel Becker <jlbec@evilplan.org>
Reviewed-by: Junxiao Bi <junxiao.bi@oracle.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

fs/ocfs2/dlm/dlmmaster.c

@@ -2549,7 +2549,7 @@ static int dlm_migrate_lockres(struct dlm_ctxt *dlm,
 	}
 
 fail:
-	if (oldmle) {
+	if (ret != -EEXIST && oldmle) {
 		/* master is known, detach if not already detached */
 		dlm_mle_detach_hb_events(dlm, oldmle);
 		dlm_put_mle(oldmle);
@@ -3136,7 +3136,8 @@ static int dlm_add_migration_mle(struct dlm_ctxt *dlm,
 				mlog(0, "tried to migrate %.*s, but some "
 				     "process beat me to it\n",
 				     namelen, name);
-				ret = -EEXIST;
+				spin_unlock(&tmp->spinlock);
+				return -EEXIST;
 			} else {
 				/* bad.  2 NODES are trying to migrate! */
 				mlog(ML_ERROR, "migration error  mle: "