Fix within-function memory leaks in the various PLs' interfaces to

SPI_prepare: they all save the prepared plan into topCxt, and so the
procCxt copy that's actually returned by SPI_prepare ought to be freed.
Diagnosis and plpython fix by Nigel Andrews, followup for other PLs
by Tom Lane.

Author: tglsfdc

src/pl/plpgsql/src/pl_exec.c

@@ -3,7 +3,7 @@
  *			  procedural language
  *
  * IDENTIFICATION
- *	  $Header: /cvsroot/pgsql/src/pl/plpgsql/src/pl_exec.c,v 1.64 2002/09/05 00:43:07 tgl Exp $
+ *	  $Header: /cvsroot/pgsql/src/pl/plpgsql/src/pl_exec.c,v 1.65 2002/10/19 22:10:58 tgl Exp $
  *
  *	  This software is copyrighted by Jan Wieck - Hamburg.
  *
@@ -2018,6 +2018,7 @@ exec_prepare_plan(PLpgSQL_execstate * estate,
 	expr->plan_simple_expr = NULL;
 	exec_simple_check_plan(expr);
 
+	SPI_freeplan(plan);
 	pfree(argtypes);
 }
 
@@ -2544,10 +2545,14 @@ exec_stmt_open(PLpgSQL_execstate * estate, PLpgSQL_stmt_open * stmt)
 		 * ----------
 		 */
 		curplan = SPI_prepare(querystr, 0, NULL);
+		if (curplan == NULL)
+			elog(ERROR, "SPI_prepare() failed for dynamic query \"%s\"",
+				 querystr);
 		portal = SPI_cursor_open(curname, curplan, NULL, NULL);
 		if (portal == NULL)
 			elog(ERROR, "Failed to open cursor");
 		pfree(querystr);
+		SPI_freeplan(curplan);
 
 		/* ----------
 		 * Store the eventually assigned cursor name in the cursor variable

src/pl/plpython/plpython.c

@@ -29,7 +29,7 @@
  * MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
  *
  * IDENTIFICATION
- *	$Header: /cvsroot/pgsql/src/pl/plpython/plpython.c,v 1.25 2002/10/14 04:20:52 momjian Exp $
+ *	$Header: /cvsroot/pgsql/src/pl/plpython/plpython.c,v 1.26 2002/10/19 22:10:58 tgl Exp $
  *
  *********************************************************************
  */
@@ -1837,21 +1837,7 @@ PLy_plan_dealloc(PyObject * arg)
 	enter();
 
 	if (ob->plan)
-	{
-		/*
-		 * free the plan... pfree(ob->plan);
-		 *
-		 * FIXME -- leaks saved plan on object destruction.  can this be
-		 * avoided?
-		 * I think so. A function prepares and then execp's a statement.
-		 * When we come to deallocate the 'statement' object we obviously
-		 * no long need the plan. Even if we did, without the object
-		 * we're never going to be able to use it again.
-		 * In the against arguments: SPI_saveplan has stuck this under
-		 * the top context so there must be a reason for doing that.
-		 */
-		pfree(ob->plan);
-	}
+		SPI_freeplan(ob->plan);
 	if (ob->types)
 		PLy_free(ob->types);
 	if (ob->args)
@@ -2023,6 +2009,7 @@ PLy_spi_prepare(PyObject * self, PyObject * args)
 	PyObject   *list = NULL;
 	PyObject   *volatile optr = NULL;
 	char	   *query;
+	void	   *tmpplan;
 
 	enter();
 
@@ -2062,7 +2049,6 @@ PLy_spi_prepare(PyObject * self, PyObject * args)
 		int			nargs,
 					i;
 
-
 		nargs = PySequence_Length(list);
 		if (nargs > 0)
 		{
@@ -2125,7 +2111,10 @@ PLy_spi_prepare(PyObject * self, PyObject * args)
 		RAISE_EXC(1);
 	}
 
-	plan->plan = SPI_saveplan(plan->plan);
+	/* transfer plan from procCxt to topCxt */
+	tmpplan = plan->plan;
+	plan->plan = SPI_saveplan(tmpplan);
+	SPI_freeplan(tmpplan);
 	if (plan->plan == NULL)
 	{
 		PLy_exception_set(PLy_exc_spi_error,

src/pl/tcl/pltcl.c

@@ -31,7 +31,7 @@
  *	  ENHANCEMENTS, OR MODIFICATIONS.
  *
  * IDENTIFICATION
- *	  $Header: /cvsroot/pgsql/src/pl/tcl/pltcl.c,v 1.65 2002/10/14 04:20:52 momjian Exp $
+ *	  $Header: /cvsroot/pgsql/src/pl/tcl/pltcl.c,v 1.66 2002/10/19 22:10:58 tgl Exp $
  *
  **********************************************************************/
 
@@ -1285,7 +1285,7 @@ pltcl_elog(ClientData cdata, Tcl_Interp *interp,
 	else if (strcmp(argv[1], "NOTICE") == 0)
 		level = NOTICE;
 	else if (strcmp(argv[1], "WARNING") == 0)
-		level = ERROR;
+		level = WARNING;
 	else if (strcmp(argv[1], "ERROR") == 0)
 		level = ERROR;
 	else if (strcmp(argv[1], "FATAL") == 0)
@@ -1837,7 +1837,8 @@ pltcl_SPI_prepare(ClientData cdata, Tcl_Interp *interp,
 	}
 
 	/************************************************************
-	 * Save the plan
+	 * Save the plan into permanent memory (right now it's in the
+	 * SPI procCxt, which will go away at function end).
 	 ************************************************************/
 	qdesc->plan = SPI_saveplan(plan);
 	if (qdesc->plan == NULL)
@@ -1866,6 +1867,8 @@ pltcl_SPI_prepare(ClientData cdata, Tcl_Interp *interp,
 
 		elog(ERROR, "pltcl: SPI_saveplan() failed - %s", reason);
 	}
+	/* Release the procCxt copy to avoid within-function memory leak */
+	SPI_freeplan(plan);
 
 	/************************************************************
 	 * Insert a hashtable entry for the plan and return