On Tue, 2007-04-03 at 10:15 +0300, Marko Kreen wrote:
> New commands:
>
> CLOSE ALL -- close all cursors
> DEALLOCATE ALL -- close all prepared stmts
> RESET PLANS -- drop all plans
> RESET TEMP | TEMPORARY -- drop all temp tables
>
> RESET SESSION -- drop/close/free everything
+ void
+ ResetTempTableNamespace(void)
+ {
+ char namespaceName[NAMEDATALEN];
+ Oid namespaceId;
+
+ /* If not allowed to create, no point proceeding */
+ if (pg_database_aclcheck(MyDatabaseId, GetUserId(),
+ ACL_CREATE_TEMP) != ACLCHECK_OK)
+ return;
ISTM this is buggy: if the user's TEMPORARY privilege is revoked between
the time that they create a temporary table and when they execute RESET
SESSION, the temporary table won't be cleaned up.
> * RESET SESSION does not ABORT anymore, instead fails if in transaction.
I think it's quite bizarre to have RESET SESSION fail if used in a
transaction, but to allow an equivalent sequence of commands to be
executed by hand inside a transaction.
guc.c is missing some #includes.
> * DEALLOCATE PREPARE ALL gives bison conflicts. Is that even needed?
Seems best to have it, for the sake of consistency. The shift/reduce
conflict is easy to workaround, provided you're content to duplicate the
body of the DEALLOCATE ALL rule -- e.g. see the attached incremental
diff.
> * Are the CommandComplete changes needed?
Seems warranted to me. BTW, why is CLOSE's command tag "CLOSE CURSOR",
not "CLOSE"? That seems needlessly verbose, and inconsistent with other
commands (e.g. DEALLOCATE).
> * ResetPlanCache() is implemented as PlanCacheCallback((Datum)0,
> InvalidOid); That seems to leave plans for utility commands untouched.
> Is it problem?
Yes, I'd think you'd also want to cleanup plans for utility commands.
-Neil
diff -u src/backend/parser/gram.y src/backend/parser/gram.y
--- src/backend/parser/gram.y 3 Apr 2007 07:09:31 -0000
+++ src/backend/parser/gram.y 7 Apr 2007 20:14:48 -0000
@@ -5596,6 +5596,12 @@
n->name = NULL;
$$ = (Node *) n;
}
+ | DEALLOCATE PREPARE ALL
+ {
+ DeallocateStmt *n = makeNode(DeallocateStmt);
+ n->name = NULL;
+ $$ = (Node *) n;
+ }
;
/*****************************************************************************
---------------------------(end of broadcast)---------------------------
TIP 2: Don't 'kill -9' the postmaster
