Derive latestRemovedXid for btree deletes by reading heap pages. The

WAL record for btree delete contains a list of tids, even when backup
blocks are present. We follow the tids to their heap tuples, taking
care to follow LP_REDIRECT tuples. We ignore LP_DEAD tuples on the
understanding that they will always have xmin/xmax earlier than any
LP_NORMAL tuples referred to by killed index tuples. Iff all tuples
are LP_DEAD we return InvalidTransactionId. The heap relfilenode is
added to the WAL record, requiring API changes to pass down the heap
Relation. XLOG_PAGE_MAGIC updated.

Author: simonat2ndQuadrant

src/backend/access/nbtree/nbtinsert.c

@@ -8,7 +8,7 @@
  *
  *
  * IDENTIFICATION
- *	  $PostgreSQL: pgsql/src/backend/access/nbtree/nbtinsert.c,v 1.177 2010/02/26 02:00:34 momjian Exp $
+ *	  $PostgreSQL: pgsql/src/backend/access/nbtree/nbtinsert.c,v 1.178 2010/03/28 09:27:01 sriggs Exp $
  *
  *-------------------------------------------------------------------------
  */
@@ -57,7 +57,8 @@ static void _bt_findinsertloc(Relation rel,
 				  OffsetNumber *offsetptr,
 				  int keysz,
 				  ScanKey scankey,
-				  IndexTuple newtup);
+				  IndexTuple newtup,
+				  Relation heapRel);
 static void _bt_insertonpg(Relation rel, Buffer buf,
 			   BTStack stack,
 			   IndexTuple itup,
@@ -78,7 +79,7 @@ static void _bt_pgaddtup(Relation rel, Page page,
 			 OffsetNumber itup_off, const char *where);
 static bool _bt_isequal(TupleDesc itupdesc, Page page, OffsetNumber offnum,
 			int keysz, ScanKey scankey);
-static void _bt_vacuum_one_page(Relation rel, Buffer buffer);
+static void _bt_vacuum_one_page(Relation rel, Buffer buffer, Relation heapRel);
 
 
 /*
@@ -175,7 +176,7 @@ _bt_doinsert(Relation rel, IndexTuple itup,
 	if (checkUnique != UNIQUE_CHECK_EXISTING)
 	{
 		/* do the insertion */
-		_bt_findinsertloc(rel, &buf, &offset, natts, itup_scankey, itup);
+		_bt_findinsertloc(rel, &buf, &offset, natts, itup_scankey, itup, heapRel);
 		_bt_insertonpg(rel, buf, stack, itup, offset, false);
 	}
 	else
@@ -491,7 +492,8 @@ _bt_findinsertloc(Relation rel,
 				  OffsetNumber *offsetptr,
 				  int keysz,
 				  ScanKey scankey,
-				  IndexTuple newtup)
+				  IndexTuple newtup,
+				  Relation heapRel)
 {
 	Buffer		buf = *bufptr;
 	Page		page = BufferGetPage(buf);
@@ -556,7 +558,7 @@ _bt_findinsertloc(Relation rel,
 		 */
 		if (P_ISLEAF(lpageop) && P_HAS_GARBAGE(lpageop))
 		{
-			_bt_vacuum_one_page(rel, buf);
+			_bt_vacuum_one_page(rel, buf, heapRel);
 
 			/*
 			 * remember that we vacuumed this page, because that makes the
@@ -1998,7 +2000,7 @@ _bt_isequal(TupleDesc itupdesc, Page page, OffsetNumber offnum,
  * super-exclusive "cleanup" lock (see nbtree/README).
  */
 static void
-_bt_vacuum_one_page(Relation rel, Buffer buffer)
+_bt_vacuum_one_page(Relation rel, Buffer buffer, Relation heapRel)
 {
 	OffsetNumber deletable[MaxOffsetNumber];
 	int			ndeletable = 0;
@@ -2025,7 +2027,7 @@ _bt_vacuum_one_page(Relation rel, Buffer buffer)
 	}
 
 	if (ndeletable > 0)
-		_bt_delitems(rel, buffer, deletable, ndeletable, false, 0);
+		_bt_delitems_delete(rel, buffer, deletable, ndeletable, heapRel);
 
 	/*
 	 * Note: if we didn't find any LP_DEAD items, then the page's

src/backend/access/nbtree/nbtpage.c

@@ -9,7 +9,7 @@
  *
  *
  * IDENTIFICATION
- *	  $PostgreSQL: pgsql/src/backend/access/nbtree/nbtpage.c,v 1.121 2010/03/19 10:41:21 sriggs Exp $
+ *	  $PostgreSQL: pgsql/src/backend/access/nbtree/nbtpage.c,v 1.122 2010/03/28 09:27:01 sriggs Exp $
  *
  *	NOTES
  *	   Postgres btree pages look like ordinary relation pages.	The opaque
@@ -719,15 +719,12 @@ _bt_page_recyclable(Page page)
  * ensure correct locking.
  */
 void
-_bt_delitems(Relation rel, Buffer buf,
-			 OffsetNumber *itemnos, int nitems, bool isVacuum,
-			 BlockNumber lastBlockVacuumed)
+_bt_delitems_vacuum(Relation rel, Buffer buf,
+			 OffsetNumber *itemnos, int nitems, BlockNumber lastBlockVacuumed)
 {
 	Page		page = BufferGetPage(buf);
 	BTPageOpaque opaque;
 
-	Assert(isVacuum || lastBlockVacuumed == 0);
-
 	/* No ereport(ERROR) until changes are logged */
 	START_CRIT_SECTION();
 
@@ -759,35 +756,14 @@ _bt_delitems(Relation rel, Buffer buf,
 		XLogRecPtr	recptr;
 		XLogRecData rdata[2];
 
-		if (isVacuum)
-		{
-			xl_btree_vacuum xlrec_vacuum;
-
-			xlrec_vacuum.node = rel->rd_node;
-			xlrec_vacuum.block = BufferGetBlockNumber(buf);
-
-			xlrec_vacuum.lastBlockVacuumed = lastBlockVacuumed;
-			rdata[0].data = (char *) &xlrec_vacuum;
-			rdata[0].len = SizeOfBtreeVacuum;
-		}
-		else
-		{
-			xl_btree_delete xlrec_delete;
-
-			xlrec_delete.node = rel->rd_node;
-			xlrec_delete.block = BufferGetBlockNumber(buf);
+		xl_btree_vacuum xlrec_vacuum;
 
-			/*
-			 * XXX: We would like to set an accurate latestRemovedXid, but
-			 * there is no easy way of obtaining a useful value. So we punt
-			 * and store InvalidTransactionId, which forces the standby to
-			 * wait for/cancel all currently running transactions.
-			 */
-			xlrec_delete.latestRemovedXid = InvalidTransactionId;
-			rdata[0].data = (char *) &xlrec_delete;
-			rdata[0].len = SizeOfBtreeDelete;
-		}
+		xlrec_vacuum.node = rel->rd_node;
+		xlrec_vacuum.block = BufferGetBlockNumber(buf);
 
+		xlrec_vacuum.lastBlockVacuumed = lastBlockVacuumed;
+		rdata[0].data = (char *) &xlrec_vacuum;
+		rdata[0].len = SizeOfBtreeVacuum;
 		rdata[0].buffer = InvalidBuffer;
 		rdata[0].next = &(rdata[1]);
 
@@ -810,10 +786,82 @@ _bt_delitems(Relation rel, Buffer buf,
 		rdata[1].buffer_std = true;
 		rdata[1].next = NULL;
 
-		if (isVacuum)
-			recptr = XLogInsert(RM_BTREE_ID, XLOG_BTREE_VACUUM, rdata);
-		else
-			recptr = XLogInsert(RM_BTREE_ID, XLOG_BTREE_DELETE, rdata);
+		recptr = XLogInsert(RM_BTREE_ID, XLOG_BTREE_VACUUM, rdata);
+
+		PageSetLSN(page, recptr);
+		PageSetTLI(page, ThisTimeLineID);
+	}
+
+	END_CRIT_SECTION();
+}
+
+void
+_bt_delitems_delete(Relation rel, Buffer buf,
+			 OffsetNumber *itemnos, int nitems, Relation heapRel)
+{
+	Page		page = BufferGetPage(buf);
+	BTPageOpaque opaque;
+
+	Assert(nitems > 0);
+
+	/* No ereport(ERROR) until changes are logged */
+	START_CRIT_SECTION();
+
+	/* Fix the page */
+	PageIndexMultiDelete(page, itemnos, nitems);
+
+	/*
+	 * We can clear the vacuum cycle ID since this page has certainly been
+	 * processed by the current vacuum scan.
+	 */
+	opaque = (BTPageOpaque) PageGetSpecialPointer(page);
+	opaque->btpo_cycleid = 0;
+
+	/*
+	 * Mark the page as not containing any LP_DEAD items.  This is not
+	 * certainly true (there might be some that have recently been marked, but
+	 * weren't included in our target-item list), but it will almost always be
+	 * true and it doesn't seem worth an additional page scan to check it.
+	 * Remember that BTP_HAS_GARBAGE is only a hint anyway.
+	 */
+	opaque->btpo_flags &= ~BTP_HAS_GARBAGE;
+
+	MarkBufferDirty(buf);
+
+	/* XLOG stuff */
+	if (!rel->rd_istemp)
+	{
+		XLogRecPtr	recptr;
+		XLogRecData rdata[3];
+
+		xl_btree_delete xlrec_delete;
+
+		xlrec_delete.node = rel->rd_node;
+		xlrec_delete.hnode = heapRel->rd_node;
+		xlrec_delete.block = BufferGetBlockNumber(buf);
+		xlrec_delete.nitems = nitems;
+
+		rdata[0].data = (char *) &xlrec_delete;
+		rdata[0].len = SizeOfBtreeDelete;
+		rdata[0].buffer = InvalidBuffer;
+		rdata[0].next = &(rdata[1]);
+
+		/*
+		 * We need the target-offsets array whether or not we store the
+		 * to allow us to find the latestRemovedXid on a standby server.
+		 */
+		rdata[1].data = (char *) itemnos;
+		rdata[1].len = nitems * sizeof(OffsetNumber);
+		rdata[1].buffer = InvalidBuffer;
+		rdata[1].next = &(rdata[2]);
+
+		rdata[2].data = NULL;
+		rdata[2].len = 0;
+		rdata[2].buffer = buf;
+		rdata[2].buffer_std = true;
+		rdata[2].next = NULL;
+
+		recptr = XLogInsert(RM_BTREE_ID, XLOG_BTREE_DELETE, rdata);
 
 		PageSetLSN(page, recptr);
 		PageSetTLI(page, ThisTimeLineID);

src/backend/access/nbtree/nbtree.c

@@ -12,7 +12,7 @@
  * Portions Copyright (c) 1994, Regents of the University of California
  *
  * IDENTIFICATION
- *	  $PostgreSQL: pgsql/src/backend/access/nbtree/nbtree.c,v 1.176 2010/02/26 02:00:34 momjian Exp $
+ *	  $PostgreSQL: pgsql/src/backend/access/nbtree/nbtree.c,v 1.177 2010/03/28 09:27:01 sriggs Exp $
  *
  *-------------------------------------------------------------------------
  */
@@ -708,7 +708,7 @@ btvacuumscan(IndexVacuumInfo *info, IndexBulkDeleteResult *stats,
 		buf = ReadBufferExtended(rel, MAIN_FORKNUM, num_pages - 1, RBM_NORMAL,
 								 info->strategy);
 		LockBufferForCleanup(buf);
-		_bt_delitems(rel, buf, NULL, 0, true, vstate.lastBlockVacuumed);
+		_bt_delitems_vacuum(rel, buf, NULL, 0, vstate.lastBlockVacuumed);
 		_bt_relbuf(rel, buf);
 	}
 
@@ -889,7 +889,7 @@ btvacuumpage(BTVacState *vstate, BlockNumber blkno, BlockNumber orig_blkno)
 		{
 			BlockNumber lastBlockVacuumed = BufferGetBlockNumber(buf);
 
-			_bt_delitems(rel, buf, deletable, ndeletable, true, vstate->lastBlockVacuumed);
+			_bt_delitems_vacuum(rel, buf, deletable, ndeletable, vstate->lastBlockVacuumed);
 
 			/*
 			 * Keep track of the block number of the lastBlockVacuumed, so we