copy_relation_data was mistakenly assuming that the source relation

would always be already open at the smgr level.  Per bug report from
Fabien Coelho.

src/backend/commands/tablecmds.c

@@ -8,7 +8,7 @@
  *
  *
  * IDENTIFICATION
- *	  $PostgreSQL: pgsql/src/backend/commands/tablecmds.c,v 1.129 2004/08/29 05:06:41 momjian Exp $
+ *	  $PostgreSQL: pgsql/src/backend/commands/tablecmds.c,v 1.130 2004/08/31 15:56:39 tgl Exp $
  *
  *-------------------------------------------------------------------------
  */
@@ -5434,7 +5434,7 @@ ATExecSetTableSpace(Oid tableOid, Oid newTableSpace)
 static void
 copy_relation_data(Relation rel, SMgrRelation dst)
 {
-	SMgrRelation src = rel->rd_smgr;
+	SMgrRelation src;
 	bool		use_wal;
 	BlockNumber nblocks;
 	BlockNumber blkno;
@@ -5457,6 +5457,9 @@ copy_relation_data(Relation rel, SMgrRelation dst)
 	use_wal = XLogArchivingActive() && !rel->rd_istemp;
 
 	nblocks = RelationGetNumberOfBlocks(rel);
+	/* RelationGetNumberOfBlocks will certainly have opened rd_smgr */
+	src = rel->rd_smgr;
+
 	for (blkno = 0; blkno < nblocks; blkno++)
 	{
 		smgrread(src, blkno, buf);