Re: DISCARD ALL failing to acquire locks on pg_listen

Matteo Beccati <php(at)beccati(dot)com> writes:
>> Seems like we could/should fix UNLISTEN * to not do anything if it is
>> known that the current backend never did any LISTENs.

> Here's my proposed patch, both for HEAD and 8.3:

This seems a bit overcomplicated. I had in mind something like this...

Index: src/backend/commands/async.c
===================================================================
RCS file: /cvsroot/pgsql/src/backend/commands/async.c,v
retrieving revision 1.145
diff -c -r1.145 async.c
*** src/backend/commands/async.c 1 Jan 2009 17:23:37 -0000 1.145
--- src/backend/commands/async.c 12 Feb 2009 18:28:43 -0000
***************
*** 277,282 ****
--- 277,286 ----
if (Trace_notify)
elog(DEBUG1, "Async_Unlisten(%s,%d)", relname, MyProcPid);

+ /* If we couldn't possibly be listening, no need to queue anything */
+ if (pendingActions == NIL && !unlistenExitRegistered)
+ return;
+
queue_listen(LISTEN_UNLISTEN, relname);
}

***************
*** 291,296 ****
--- 295,304 ----
if (Trace_notify)
elog(DEBUG1, "Async_UnlistenAll(%d)", MyProcPid);

+ /* If we couldn't possibly be listening, no need to queue anything */
+ if (pendingActions == NIL && !unlistenExitRegistered)
+ return;
+
queue_listen(LISTEN_UNLISTEN_ALL, "");
}

regards, tom lane