Code review for HeapTupleHeader changes. Add version number to page headers

(overlaying low byte of page size) and add HEAP_HASOID bit to t_infomask,
per earlier discussion.  Simplify scheme for overlaying fields in tuple
header (no need for cmax to live in more than one place).  Don't try to
clear infomask status bits in tqual.c --- not safe to do it there.  Don't
try to force output table of a SELECT INTO to have OIDs, either.  Get rid
of unnecessarily complex three-state scheme for TupleDesc.tdhasoids, which
has already caused one recent failure.  Improve documentation.

src/backend/executor/execMain.c

@@ -27,7 +27,7 @@
  *
  *
  * IDENTIFICATION
- *	  $Header: /cvsroot/pgsql/src/backend/executor/execMain.c,v 1.176 2002/08/29 00:17:03 tgl Exp $
+ *	  $Header: /cvsroot/pgsql/src/backend/executor/execMain.c,v 1.177 2002/09/02 01:05:05 tgl Exp $
  *
  *-------------------------------------------------------------------------
  */
@@ -712,22 +712,25 @@ InitPlan(CmdType operation, Query *parseTree, Plan *plan, EState *estate)
 					aclcheck_error(aclresult,
 								   get_namespace_name(namespaceId));
 
-				/*
-				 * new "INTO" table is created WITH OIDS
-				 */
-				tupType->tdhasoid = WITHOID;
 				/*
 				 * have to copy tupType to get rid of constraints
 				 */
 				tupdesc = CreateTupleDescCopy(tupType);
 
+				/*
+				 * Formerly we forced the output table to have OIDs, but
+				 * as of 7.3 it will not have OIDs, because it's too late
+				 * here to change the tupdescs of the already-initialized
+				 * plan tree.  (Perhaps we could recurse and change them
+				 * all, but it's not really worth the trouble IMHO...)
+				 */
+
 				intoRelationId =
 					heap_create_with_catalog(intoName,
 											 namespaceId,
 											 tupdesc,
 											 RELKIND_RELATION,
 											 false,
-											 true,
 											 allowSystemTableMods);
 
 				FreeTupleDesc(tupdesc);