Re: [HACKERS] Unexpected subselect result. - Mailing list pgsql-hackers

From Vadim B. Mikheev
Subject Re: [HACKERS] Unexpected subselect result.
Date
Msg-id 34EB9F38.2D4F1004@sable.krasnoyarsk.su
Whole thread Raw
In response to Unexpected subselect result.  (Keith Parks <emkxp01@mtcc.demon.co.uk>)
List pgsql-hackers
Sorry, I decided to copy subselect' tuple unconditionally...

Vadim
*** nodeSubplan.c.orig    Thu Feb 19 08:54:19 1998
--- nodeSubplan.c    Thu Feb 19 09:51:53 1998
***************
*** 193,198 ****
--- 193,207 ----
              break;
          }

+         /*
+          * If this is uncorrelated subquery then its plan will be closed
+          * (see below) and this tuple will be free-ed - bad for not byval
+          * types... But is free-ing possible in the next ExecProcNode in
+          * this loop ? Who knows... Someday we'll keep track of saved
+          * tuples...
+          */
+         tup = heap_copytuple (tup);
+
          foreach (lst, node->setParam)
          {
              ParamExecData   *prm = &(plan->state->es_param_exec_vals[lfirsti(lst)]);

pgsql-hackers by date:

Previous
From: Peter Stockwell
Date:
Subject: Status of Alpha/Unix
Next
From: Brook Milligan
Date:
Subject: a few patches for the upcoming 6.3 release