Fix multiple problems in WAL replay.

Most of the replay functions for WAL record types that modify more than
one page failed to ensure that those pages were locked correctly to ensure
that concurrent queries could not see inconsistent page states.  This is
a hangover from coding decisions made long before Hot Standby was added,
when it was hardly necessary to acquire buffer locks during WAL replay
at all, let alone hold them for carefully-chosen periods.

The key problem was that RestoreBkpBlocks was written to hold lock on each
page restored from a full-page image for only as long as it took to update
that page.  This was guaranteed to break any WAL replay function in which
there was any update-ordering constraint between pages, because even if the
nominal order of the pages is the right one, any mixture of full-page and
non-full-page updates in the same record would result in out-of-order
updates.  Moreover, it wouldn't work for situations where there's a
requirement to maintain lock on one page while updating another.  Failure
to honor an update ordering constraint in this way is thought to be the
cause of bug #7648 from Daniel Farina: what seems to have happened there
is that a btree page being split was rewritten from a full-page image
before the new right sibling page was written, and because lock on the
original page was not maintained it was possible for hot standby queries to
try to traverse the page's right-link to the not-yet-existing sibling page.

To fix, get rid of RestoreBkpBlocks as such, and instead create a new
function RestoreBackupBlock that restores just one full-page image at a
time.  This function can be invoked by WAL replay functions at the points
where they would otherwise perform non-full-page updates; in this way, the
physical order of page updates remains the same no matter which pages are
replaced by full-page images.  We can then further adjust the logic in
individual replay functions if it is necessary to hold buffer locks
for overlapping periods.  A side benefit is that we can simplify the
handling of concurrency conflict resolution by moving that code into the
record-type-specfic functions; there's no more need to contort the code
layout to keep conflict resolution in front of the RestoreBkpBlocks call.

In connection with that, standardize on zero-based numbering rather than
one-based numbering for referencing the full-page images.  In HEAD, I
removed the macros XLR_BKP_BLOCK_1 through XLR_BKP_BLOCK_4.  They are
still there in the header files in previous branches, but are no longer
used by the code.

In addition, fix some other bugs identified in the course of making these
changes:

spgRedoAddNode could fail to update the parent downlink at all, if the
parent tuple is in the same page as either the old or new split tuple and
we're not doing a full-page image: it would get fooled by the LSN having
been advanced already.  This would result in permanent index corruption,
not just transient failure of concurrent queries.

Also, ginHeapTupleFastInsert's "merge lists" case failed to mark the old
tail page as a candidate for a full-page image; in the worst case this
could result in torn-page corruption.

heap_xlog_freeze() was inconsistent about using a cleanup lock or plain
exclusive lock: it did the former in the normal path but the latter for a
full-page image.  A plain exclusive lock seems sufficient, so change to
that.

Also, remove gistRedoPageDeleteRecord(), which has been dead code since
VACUUM FULL was rewritten.

Back-patch to 9.0, where hot standby was introduced.  Note however that 9.0
had a significantly different WAL-logging scheme for GIST index updates,
and it doesn't appear possible to make that scheme safe for concurrent hot
standby queries, because it can leave inconsistent states in the index even
between WAL records.  Given the lack of complaints from the field, we won't
work too hard on fixing that branch.

Author: tglsfdc

src/backend/access/gin/ginfast.c

@@ -289,7 +289,7 @@ ginHeapTupleFastInsert(GinState *ginstate, GinTupleCollector *collector)
 		if (metadata->head == InvalidBlockNumber)
 		{
 			/*
-			 * Main list is empty, so just copy sublist into main list
+			 * Main list is empty, so just insert sublist as main list
 			 */
 			START_CRIT_SECTION();
 
@@ -312,6 +312,14 @@ ginHeapTupleFastInsert(GinState *ginstate, GinTupleCollector *collector)
 			LockBuffer(buffer, GIN_EXCLUSIVE);
 			page = BufferGetPage(buffer);
 
+			rdata[0].next = rdata + 1;
+
+			rdata[1].buffer = buffer;
+			rdata[1].buffer_std = true;
+			rdata[1].data = NULL;
+			rdata[1].len = 0;
+			rdata[1].next = NULL;
+
 			Assert(GinPageGetOpaque(page)->rightlink == InvalidBlockNumber);
 
 			START_CRIT_SECTION();

src/backend/access/gin/ginxlog.c

@@ -79,6 +79,9 @@ ginRedoCreateIndex(XLogRecPtr lsn, XLogRecord *record)
 				MetaBuffer;
 	Page		page;
 
+	/* Backup blocks are not used in create_index records */
+	Assert(!(record->xl_info & XLR_BKP_BLOCK_MASK));
+
 	MetaBuffer = XLogReadBuffer(*node, GIN_METAPAGE_BLKNO, true);
 	Assert(BufferIsValid(MetaBuffer));
 	page = (Page) BufferGetPage(MetaBuffer);
@@ -109,6 +112,9 @@ ginRedoCreatePTree(XLogRecPtr lsn, XLogRecord *record)
 	Buffer		buffer;
 	Page		page;
 
+	/* Backup blocks are not used in create_ptree records */
+	Assert(!(record->xl_info & XLR_BKP_BLOCK_MASK));
+
 	buffer = XLogReadBuffer(data->node, data->blkno, true);
 	Assert(BufferIsValid(buffer));
 	page = (Page) BufferGetPage(buffer);
@@ -158,9 +164,12 @@ ginRedoInsert(XLogRecPtr lsn, XLogRecord *record)
 		}
 	}
 
-	/* nothing else to do if page was backed up */
+	/* If we have a full-page image, restore it and we're done */
 	if (IsBkpBlockApplied(record, 0))
+	{
+		(void) RestoreBackupBlock(lsn, record, 0, false, false);
 		return;
+	}
 
 	buffer = XLogReadBuffer(data->node, data->blkno, false);
 	if (!BufferIsValid(buffer))
@@ -254,6 +263,9 @@ ginRedoSplit(XLogRecPtr lsn, XLogRecord *record)
 	if (data->isData)
 		flags |= GIN_DATA;
 
+	/* Backup blocks are not used in split records */
+	Assert(!(record->xl_info & XLR_BKP_BLOCK_MASK));
+
 	lbuffer = XLogReadBuffer(data->node, data->lblkno, true);
 	Assert(BufferIsValid(lbuffer));
 	lpage = (Page) BufferGetPage(lbuffer);
@@ -364,9 +376,11 @@ ginRedoVacuumPage(XLogRecPtr lsn, XLogRecord *record)
 	Buffer		buffer;
 	Page		page;
 
-	/* nothing to do if page was backed up (and no info to do it with) */
 	if (IsBkpBlockApplied(record, 0))
+	{
+		(void) RestoreBackupBlock(lsn, record, 0, false, false);
 		return;
+	}
 
 	buffer = XLogReadBuffer(data->node, data->blkno, false);
 	if (!BufferIsValid(buffer))
@@ -414,60 +428,71 @@ static void
 ginRedoDeletePage(XLogRecPtr lsn, XLogRecord *record)
 {
 	ginxlogDeletePage *data = (ginxlogDeletePage *) XLogRecGetData(record);
-	Buffer		buffer;
+	Buffer		dbuffer;
+	Buffer		pbuffer;
+	Buffer		lbuffer;
 	Page		page;
 
-	if (!(IsBkpBlockApplied(record, 0)))
+	if (IsBkpBlockApplied(record, 0))
+		dbuffer = RestoreBackupBlock(lsn, record, 0, false, true);
+	else
 	{
-		buffer = XLogReadBuffer(data->node, data->blkno, false);
-		if (BufferIsValid(buffer))
+		dbuffer = XLogReadBuffer(data->node, data->blkno, false);
+		if (BufferIsValid(dbuffer))
 		{
-			page = BufferGetPage(buffer);
+			page = BufferGetPage(dbuffer);
 			if (!XLByteLE(lsn, PageGetLSN(page)))
 			{
 				Assert(GinPageIsData(page));
 				GinPageGetOpaque(page)->flags = GIN_DELETED;
 				PageSetLSN(page, lsn);
-				MarkBufferDirty(buffer);
+				MarkBufferDirty(dbuffer);
 			}
-			UnlockReleaseBuffer(buffer);
 		}
 	}
 
-	if (!(IsBkpBlockApplied(record, 1)))
+	if (IsBkpBlockApplied(record, 1))
+		pbuffer = RestoreBackupBlock(lsn, record, 1, false, true);
+	else
 	{
-		buffer = XLogReadBuffer(data->node, data->parentBlkno, false);
-		if (BufferIsValid(buffer))
+		pbuffer = XLogReadBuffer(data->node, data->parentBlkno, false);
+		if (BufferIsValid(pbuffer))
 		{
-			page = BufferGetPage(buffer);
+			page = BufferGetPage(pbuffer);
 			if (!XLByteLE(lsn, PageGetLSN(page)))
 			{
 				Assert(GinPageIsData(page));
 				Assert(!GinPageIsLeaf(page));
 				GinPageDeletePostingItem(page, data->parentOffset);
 				PageSetLSN(page, lsn);
-				MarkBufferDirty(buffer);
+				MarkBufferDirty(pbuffer);
 			}
-			UnlockReleaseBuffer(buffer);
 		}
 	}
 
-	if (!(IsBkpBlockApplied(record, 2)) && data->leftBlkno != InvalidBlockNumber)
+	if (IsBkpBlockApplied(record, 2))
+		(void) RestoreBackupBlock(lsn, record, 2, false, false);
+	else if (data->leftBlkno != InvalidBlockNumber)
 	{
-		buffer = XLogReadBuffer(data->node, data->leftBlkno, false);
-		if (BufferIsValid(buffer))
+		lbuffer = XLogReadBuffer(data->node, data->leftBlkno, false);
+		if (BufferIsValid(lbuffer))
 		{
-			page = BufferGetPage(buffer);
+			page = BufferGetPage(lbuffer);
 			if (!XLByteLE(lsn, PageGetLSN(page)))
 			{
 				Assert(GinPageIsData(page));
 				GinPageGetOpaque(page)->rightlink = data->rightLink;
 				PageSetLSN(page, lsn);
-				MarkBufferDirty(buffer);
+				MarkBufferDirty(lbuffer);
 			}
-			UnlockReleaseBuffer(buffer);
+			UnlockReleaseBuffer(lbuffer);
 		}
 	}
+
+	if (BufferIsValid(pbuffer))
+		UnlockReleaseBuffer(pbuffer);
+	if (BufferIsValid(dbuffer))
+		UnlockReleaseBuffer(dbuffer);
 }
 
 static void
@@ -495,7 +520,9 @@ ginRedoUpdateMetapage(XLogRecPtr lsn, XLogRecord *record)
 		/*
 		 * insert into tail page
 		 */
-		if (!(record->xl_info & XLR_BKP_BLOCK_1))
+		if (record->xl_info & XLR_BKP_BLOCK(0))
+			(void) RestoreBackupBlock(lsn, record, 0, false, false);
+		else
 		{
 			buffer = XLogReadBuffer(data->node, data->metadata.tail, false);
 			if (BufferIsValid(buffer))
@@ -542,19 +569,24 @@ ginRedoUpdateMetapage(XLogRecPtr lsn, XLogRecord *record)
 		/*
 		 * New tail
 		 */
-		buffer = XLogReadBuffer(data->node, data->prevTail, false);
-		if (BufferIsValid(buffer))
+		if (record->xl_info & XLR_BKP_BLOCK(0))
+			(void) RestoreBackupBlock(lsn, record, 0, false, false);
+		else
 		{
-			Page		page = BufferGetPage(buffer);
-
-			if (!XLByteLE(lsn, PageGetLSN(page)))
+			buffer = XLogReadBuffer(data->node, data->prevTail, false);
+			if (BufferIsValid(buffer))
 			{
-				GinPageGetOpaque(page)->rightlink = data->newRightlink;
+				Page		page = BufferGetPage(buffer);
 
-				PageSetLSN(page, lsn);
-				MarkBufferDirty(buffer);
+				if (!XLByteLE(lsn, PageGetLSN(page)))
+				{
+					GinPageGetOpaque(page)->rightlink = data->newRightlink;
+
+					PageSetLSN(page, lsn);
+					MarkBufferDirty(buffer);
+				}
+				UnlockReleaseBuffer(buffer);
 			}
-			UnlockReleaseBuffer(buffer);
 		}
 	}
 
@@ -573,8 +605,12 @@ ginRedoInsertListPage(XLogRecPtr lsn, XLogRecord *record)
 				tupsize;
 	IndexTuple	tuples = (IndexTuple) (XLogRecGetData(record) + sizeof(ginxlogInsertListPage));
 
-	if (record->xl_info & XLR_BKP_BLOCK_1)
+	/* If we have a full-page image, restore it and we're done */
+	if (record->xl_info & XLR_BKP_BLOCK(0))
+	{
+		(void) RestoreBackupBlock(lsn, record, 0, false, false);
 		return;
+	}
 
 	buffer = XLogReadBuffer(data->node, data->blkno, true);
 	Assert(BufferIsValid(buffer));
@@ -619,6 +655,9 @@ ginRedoDeleteListPages(XLogRecPtr lsn, XLogRecord *record)
 	Page		metapage;
 	int			i;
 
+	/* Backup blocks are not used in delete_listpage records */
+	Assert(!(record->xl_info & XLR_BKP_BLOCK_MASK));
+
 	metabuffer = XLogReadBuffer(data->node, GIN_METAPAGE_BLKNO, false);
 	if (!BufferIsValid(metabuffer))
 		return;					/* assume index was deleted, nothing to do */
@@ -631,6 +670,16 @@ ginRedoDeleteListPages(XLogRecPtr lsn, XLogRecord *record)
 		MarkBufferDirty(metabuffer);
 	}
 
+	/*
+	 * In normal operation, shiftList() takes exclusive lock on all the
+	 * pages-to-be-deleted simultaneously.	During replay, however, it should
+	 * be all right to lock them one at a time.  This is dependent on the fact
+	 * that we are deleting pages from the head of the list, and that readers
+	 * share-lock the next page before releasing the one they are on. So we
+	 * cannot get past a reader that is on, or due to visit, any page we are
+	 * going to delete.  New incoming readers will block behind our metapage
+	 * lock and then see a fully updated page list.
+	 */
 	for (i = 0; i < data->ndeleted; i++)
 	{
 		Buffer		buffer = XLogReadBuffer(data->node, data->toDelete[i], false);
@@ -663,7 +712,6 @@ gin_redo(XLogRecPtr beginLoc, XLogRecPtr lsn, XLogRecord *record)
 	 * implement a similar optimization as we have in b-tree, and remove
 	 * killed tuples outside VACUUM, we'll need to handle that here.
 	 */
-	RestoreBkpBlocks(lsn, record, false);
 
 	topCtx = MemoryContextSwitchTo(opCtx);
 	switch (info)