Modify RelationGetBufferForTuple() so that we only do lseek and lock

when we need to move to a new page; as long as we can insert the new
tuple on the same page as before, we only need LockBuffer and not the
expensive stuff.  Also, twiddle bufmgr interfaces to avoid redundant
lseeks in RelationGetBufferForTuple and BufferAlloc.  Successive inserts
now require one lseek per page added, rather than one per tuple with
several additional ones at each page boundary as happened before.
Lock contention when multiple backends are inserting in same table
is also greatly reduced.

Author: tglsfdc

src/backend/access/heap/heapam.c

@@ -8,7 +8,7 @@
  *
  *
  * IDENTIFICATION
- *	  $Header: /cvsroot/pgsql/src/backend/access/heap/heapam.c,v 1.113 2001/03/25 23:23:58 tgl Exp $
+ *	  $Header: /cvsroot/pgsql/src/backend/access/heap/heapam.c,v 1.114 2001/05/12 19:58:27 tgl Exp $
  *
  *
  * INTERFACE ROUTINES
@@ -487,7 +487,7 @@ heapgettup(Relation relation,
 			return;
 		}
 
-		*buffer = ReleaseAndReadBuffer(*buffer, relation, page);
+		*buffer = ReleaseAndReadBuffer(*buffer, relation, page, false);
 
 		if (!BufferIsValid(*buffer))
 			elog(ERROR, "heapgettup: failed ReadBuffer");

src/backend/access/heap/hio.c

@@ -8,7 +8,7 @@
  *
  *
  * IDENTIFICATION
- *	  $Id: hio.c,v 1.37 2001/03/22 06:16:07 momjian Exp $
+ *	  $Id: hio.c,v 1.38 2001/05/12 19:58:27 tgl Exp $
  *
  *-------------------------------------------------------------------------
  */
@@ -66,7 +66,7 @@ RelationPutHeapTuple(Relation relation,
 /*
  * RelationGetBufferForTuple
  *
- * Returns (locked) buffer with free space >= given len.
+ * Returns exclusive-locked buffer with free space >= given len.
  *
  * Note that we use LockPage to lock relation for extension. We can
  * do this as long as in all other places we use page-level locking
@@ -75,14 +75,14 @@ RelationPutHeapTuple(Relation relation,
  *
  * ELOG(ERROR) is allowed here, so this routine *must* be called
  * before any (unlogged) changes are made in buffer pool.
- *
  */
 Buffer
 RelationGetBufferForTuple(Relation relation, Size len)
 {
-	Buffer		buffer;
+	Buffer		buffer = InvalidBuffer;
 	Page		pageHeader;
-	BlockNumber lastblock;
+	BlockNumber lastblock,
+				oldnblocks;
 
 	len = MAXALIGN(len);		/* be conservative */
 
@@ -93,59 +93,102 @@ RelationGetBufferForTuple(Relation relation, Size len)
 		elog(ERROR, "Tuple is too big: size %lu, max size %ld",
 			 (unsigned long) len, MaxTupleSize);
 
-	if (!relation->rd_myxactonly)
-		LockPage(relation, 0, ExclusiveLock);
-
 	/*
-	 * XXX This does an lseek - VERY expensive - but at the moment it is
-	 * the only way to accurately determine how many blocks are in a
-	 * relation.  A good optimization would be to get this to actually
-	 * work properly.
+	 * First, use relcache's record of table length to guess where the
+	 * last page is, and try to put the tuple there.  This cached value
+	 * may be out of date, in which case we'll be inserting into a non-last
+	 * page, but that should be OK.  Note that in a newly created relcache
+	 * entry, rd_nblocks may be zero; if so, we'll set it correctly below.
 	 */
-	lastblock = RelationGetNumberOfBlocks(relation);
-
-	/*
-	 * Get the last existing page --- may need to create the first one if
-	 * this is a virgin relation.
-	 */
-	if (lastblock == 0)
-	{
-		buffer = ReadBuffer(relation, P_NEW);
-		LockBuffer(buffer, BUFFER_LOCK_EXCLUSIVE);
-		pageHeader = (Page) BufferGetPage(buffer);
-		Assert(PageIsNew((PageHeader) pageHeader));
-		PageInit(pageHeader, BufferGetPageSize(buffer), 0);
-	}
-	else
+	if (relation->rd_nblocks > 0)
 	{
-		buffer = ReadBuffer(relation, lastblock - 1);
+		lastblock = relation->rd_nblocks - 1;
+		buffer = ReadBuffer(relation, lastblock);
 		LockBuffer(buffer, BUFFER_LOCK_EXCLUSIVE);
 		pageHeader = (Page) BufferGetPage(buffer);
+		if (len <= PageGetFreeSpace(pageHeader))
+			return buffer;
+		/*
+		 * Doesn't fit, so we'll have to try someplace else.
+		 */
+		LockBuffer(buffer, BUFFER_LOCK_UNLOCK);
+		/* buffer release will happen below... */
 	}
 
 	/*
-	 * Is there room on the last existing page?
+	 * Before extending relation, make sure no one else has done
+	 * so more recently than our last rd_nblocks update.  (If we
+	 * blindly extend the relation here, then probably most of the
+	 * page the other guy added will end up going to waste.)
+	 *
+	 * We have to use a lock to ensure no one else is extending the
+	 * rel at the same time, else we will both try to initialize the
+	 * same new page.
 	 */
-	if (len > PageGetFreeSpace(pageHeader))
+	if (!relation->rd_myxactonly)
+		LockPage(relation, 0, ExclusiveLock);
+
+	oldnblocks = relation->rd_nblocks;
+	/*
+	 * XXX This does an lseek - rather expensive - but at the moment it is
+	 * the only way to accurately determine how many blocks are in a
+	 * relation.  Is it worth keeping an accurate file length in shared
+	 * memory someplace, rather than relying on the kernel to do it for us?
+	 */
+	relation->rd_nblocks = RelationGetNumberOfBlocks(relation);
+
+	if (relation->rd_nblocks > oldnblocks)
 	{
-		LockBuffer(buffer, BUFFER_LOCK_UNLOCK);
-		buffer = ReleaseAndReadBuffer(buffer, relation, P_NEW);
+		/*
+		 * Someone else has indeed extended the relation recently.
+		 * Try to fit our tuple into the new last page.
+		 */
+		lastblock = relation->rd_nblocks - 1;
+		buffer = ReleaseAndReadBuffer(buffer, relation, lastblock, false);
 		LockBuffer(buffer, BUFFER_LOCK_EXCLUSIVE);
 		pageHeader = (Page) BufferGetPage(buffer);
-		Assert(PageIsNew((PageHeader) pageHeader));
-		PageInit(pageHeader, BufferGetPageSize(buffer), 0);
-
-		if (len > PageGetFreeSpace(pageHeader))
+		if (len <= PageGetFreeSpace(pageHeader))
 		{
-			/* We should not get here given the test at the top */
-			elog(STOP, "Tuple is too big: size %lu",
-				 (unsigned long) len);
+			/* OK, we don't need to extend again. */
+			if (!relation->rd_myxactonly)
+				UnlockPage(relation, 0, ExclusiveLock);
+			return buffer;
 		}
+		/*
+		 * Doesn't fit, so we'll have to extend the relation (again).
+		 */
+		LockBuffer(buffer, BUFFER_LOCK_UNLOCK);
+		/* buffer release will happen below... */
 	}
 
+	/*
+	 * Extend the relation by one page and update rd_nblocks for next time.
+	 */
+	lastblock = relation->rd_nblocks;
+	buffer = ReleaseAndReadBuffer(buffer, relation, lastblock, true);
+	relation->rd_nblocks = lastblock + 1;
+
+	/*
+	 * We need to initialize the empty new page.
+	 */
+	LockBuffer(buffer, BUFFER_LOCK_EXCLUSIVE);
+	pageHeader = (Page) BufferGetPage(buffer);
+	Assert(PageIsNew((PageHeader) pageHeader));
+	PageInit(pageHeader, BufferGetPageSize(buffer), 0);
+
+	/*
+	 * Release the file-extension lock; it's now OK for someone else
+	 * to extend the relation some more.
+	 */
 	if (!relation->rd_myxactonly)
 		UnlockPage(relation, 0, ExclusiveLock);
 
-	return (buffer);
+	if (len > PageGetFreeSpace(pageHeader))
+	{
+		/* We should not get here given the test at the top */
+		elog(STOP, "Tuple is too big: size %lu",
+			 (unsigned long) len);
+	}
 
+	return buffer;
 }