From: Massimo Dal Zotto <dz@cs.unitn.it>

I have this small patch which fixes a typo in my previous lock patches.
Fortunately the bug affects only user locks.

-

diff --git a/src/backend/storage/lmgr/lock.c b/src/backend/storage/lmgr/lock.c
index b35008a3bb8afc84d393eddbd4d8e3a07c7ade8e..52ddd635832e6a2ebc044044225b9234545ddd34 100644 (file)
--- a/src/backend/storage/lmgr/lock.c
+++ b/src/backend/storage/lmgr/lock.c
@@ -7,7 +7,7 @@
  *
  *
  * IDENTIFICATION
- *       $Header: /cvsroot/pgsql/src/backend/storage/lmgr/lock.c,v 1.33 1998/08/25 21:20:26 scrappy Exp $
+ *       $Header: /cvsroot/pgsql/src/backend/storage/lmgr/lock.c,v 1.34 1998/08/28 12:08:03 scrappy Exp $
  *
  * NOTES
  *       Outside modules can create a lock table and acquire/release
@@ -668,7 +668,7 @@ LockAcquire(LOCKMETHOD lockmethod, LOCKTAG *locktag, LOCKMODE lockmode)
                        {
                                SHMQueueDelete(&result->queue);
                                result = (XIDLookupEnt *) hash_search(xidTable,
-                                                                                                         (Pointer) &result,
+                                                                                                         (Pointer) result,
                                                                                                          HASH_REMOVE, &found);
                                if (!result || !found) {
                                        elog(NOTICE, "LockAcquire: remove xid, table corrupted");