From 09c2e7cd2ff0b884625c37ce8249832820c58710 Mon Sep 17 00:00:00 2001
From: Robert Haas <rhaas@postgresql.org>
Date: Mon, 17 Jul 2017 12:03:35 -0400
Subject: [PATCH] hash: Fix write-ahead logging bugs related to init forks.

One, logging for CREATE INDEX was oblivious to the fact that when
an unlogged table is created, *only* operations on the init fork
should be logged.

Two, init fork buffers need to be flushed after they are written;
otherwise, a filesystem-level copy following recovery may do the
wrong thing.  (There may be a better fix for this issue than the
one used here, but this is transposed from the similar logic already
present in XLogReadBufferForRedoExtended, and a broader refactoring
after beta2 seems inadvisable.)

Amit Kapila, reviewed by Ashutosh Sharma, Kyotaro Horiguchi,
and Michael Paquier

Discussion: http://postgr.es/m/CAA4eK1JpcMsEtOL_J7WODumeEfyrPi7FPYHeVdS7fyyrCrgp4w@mail.gmail.com
---
 src/backend/access/hash/hash_xlog.c | 27 +++++++++++++++++++++++++++
 src/backend/access/hash/hashpage.c  | 23 ++++++++++++++++-------
 2 files changed, 43 insertions(+), 7 deletions(-)

diff --git a/src/backend/access/hash/hash_xlog.c b/src/backend/access/hash/hash_xlog.c
index 429af11f4d..67a856c142 100644
--- a/src/backend/access/hash/hash_xlog.c
+++ b/src/backend/access/hash/hash_xlog.c
@@ -33,6 +33,7 @@ hash_xlog_init_meta_page(XLogReaderState *record)
 	XLogRecPtr	lsn = record->EndRecPtr;
 	Page		page;
 	Buffer		metabuf;
+	ForkNumber	forknum;
 
 	xl_hash_init_meta_page *xlrec = (xl_hash_init_meta_page *) XLogRecGetData(record);
 
@@ -44,6 +45,17 @@ hash_xlog_init_meta_page(XLogReaderState *record)
 	page = (Page) BufferGetPage(metabuf);
 	PageSetLSN(page, lsn);
 	MarkBufferDirty(metabuf);
+
+	/*
+	 * Force the on-disk state of init forks to always be in sync with the
+	 * state in shared buffers.  See XLogReadBufferForRedoExtended.  We need
+	 * special handling for init forks as create index operations don't log a
+	 * full page image of the metapage.
+	 */
+	XLogRecGetBlockTag(record, 0, NULL, &forknum, NULL);
+	if (forknum == INIT_FORKNUM)
+		FlushOneBuffer(metabuf);
+
 	/* all done */
 	UnlockReleaseBuffer(metabuf);
 }
@@ -60,6 +72,7 @@ hash_xlog_init_bitmap_page(XLogReaderState *record)
 	Page		page;
 	HashMetaPage metap;
 	uint32		num_buckets;
+	ForkNumber	forknum;
 
 	xl_hash_init_bitmap_page *xlrec = (xl_hash_init_bitmap_page *) XLogRecGetData(record);
 
@@ -70,6 +83,16 @@ hash_xlog_init_bitmap_page(XLogReaderState *record)
 	_hash_initbitmapbuffer(bitmapbuf, xlrec->bmsize, true);
 	PageSetLSN(BufferGetPage(bitmapbuf), lsn);
 	MarkBufferDirty(bitmapbuf);
+
+	/*
+	 * Force the on-disk state of init forks to always be in sync with the
+	 * state in shared buffers.  See XLogReadBufferForRedoExtended.  We need
+	 * special handling for init forks as create index operations don't log a
+	 * full page image of the metapage.
+	 */
+	XLogRecGetBlockTag(record, 0, NULL, &forknum, NULL);
+	if (forknum == INIT_FORKNUM)
+		FlushOneBuffer(bitmapbuf);
 	UnlockReleaseBuffer(bitmapbuf);
 
 	/* add the new bitmap page to the metapage's list of bitmaps */
@@ -90,6 +113,10 @@ hash_xlog_init_bitmap_page(XLogReaderState *record)
 
 		PageSetLSN(page, lsn);
 		MarkBufferDirty(metabuf);
+
+		XLogRecGetBlockTag(record, 1, NULL, &forknum, NULL);
+		if (forknum == INIT_FORKNUM)
+			FlushOneBuffer(metabuf);
 	}
 	if (BufferIsValid(metabuf))
 		UnlockReleaseBuffer(metabuf);
diff --git a/src/backend/access/hash/hashpage.c b/src/backend/access/hash/hashpage.c
index 1cb18a7513..d5b6502775 100644
--- a/src/backend/access/hash/hashpage.c
+++ b/src/backend/access/hash/hashpage.c
@@ -345,12 +345,20 @@ _hash_init(Relation rel, double num_tuples, ForkNumber forkNum)
 	int32		ffactor;
 	uint32		num_buckets;
 	uint32		i;
+	bool		use_wal;
 
 	/* safety check */
 	if (RelationGetNumberOfBlocksInFork(rel, forkNum) != 0)
 		elog(ERROR, "cannot initialize non-empty hash index \"%s\"",
 			 RelationGetRelationName(rel));
 
+	/*
+	 * WAL log creation of pages if the relation is persistent, or this is the
+	 * init fork.  Init forks for unlogged relations always need to be WAL
+	 * logged.
+	 */
+	use_wal = RelationNeedsWAL(rel) || forkNum == INIT_FORKNUM;
+
 	/*
 	 * Determine the target fill factor (in tuples per bucket) for this index.
 	 * The idea is to make the fill factor correspond to pages about as full
@@ -384,7 +392,7 @@ _hash_init(Relation rel, double num_tuples, ForkNumber forkNum)
 	metap = HashPageGetMeta(pg);
 
 	/* XLOG stuff */
-	if (RelationNeedsWAL(rel))
+	if (use_wal)
 	{
 		xl_hash_init_meta_page xlrec;
 		XLogRecPtr	recptr;
@@ -427,11 +435,12 @@ _hash_init(Relation rel, double num_tuples, ForkNumber forkNum)
 		_hash_initbuf(buf, metap->hashm_maxbucket, i, LH_BUCKET_PAGE, false);
 		MarkBufferDirty(buf);
 
-		log_newpage(&rel->rd_node,
-					forkNum,
-					blkno,
-					BufferGetPage(buf),
-					true);
+		if (use_wal)
+			log_newpage(&rel->rd_node,
+						forkNum,
+						blkno,
+						BufferGetPage(buf),
+						true);
 		_hash_relbuf(rel, buf);
 	}
 
@@ -459,7 +468,7 @@ _hash_init(Relation rel, double num_tuples, ForkNumber forkNum)
 	MarkBufferDirty(metabuf);
 
 	/* XLOG stuff */
-	if (RelationNeedsWAL(rel))
+	if (use_wal)
 	{
 		xl_hash_init_bitmap_page xlrec;
 		XLogRecPtr	recptr;