/*
 * Post vacuum, iterate over all entries in index, check if the h_tid
 * of each entry exists and is not dead.  For specific system tables,
 * also ensure that the key in index entry matches the corresponding
 * attribute in the heap tuple.
 */
void
_bt_validate_vacuum(Relation irel, Relation hrel, TransactionId oldest_xmin)
{
	MIRROREDLOCK_BUFMGR_DECLARE;

	BlockNumber blkno;
	BlockNumber num_pages;
	Buffer ibuf = InvalidBuffer;
	Buffer hbuf = InvalidBuffer;
	Page ipage;
	BTPageOpaque opaque;
	IndexTuple itup;
	HeapTupleData htup;
	OffsetNumber maxoff,
			minoff,
			offnum;
	Oid ioid,
		hoid;
	bool isnull;

	blkno = BTREE_METAPAGE + 1;
	num_pages = RelationGetNumberOfBlocks(irel);

	elog(LOG, "btvalidatevacuum: index %s, heap %s",
		 RelationGetRelationName(irel), RelationGetRelationName(hrel));

	MIRROREDLOCK_BUFMGR_LOCK;
	for (; blkno < num_pages; blkno++)
	{
		ibuf = ReadBuffer(irel, blkno);
		ipage = BufferGetPage(ibuf);
		opaque = (BTPageOpaque) PageGetSpecialPointer(ipage);
		if (!PageIsNew(ipage))
			_bt_checkpage(irel, ibuf);
		if (P_ISLEAF(opaque))
		{
			minoff = P_FIRSTDATAKEY(opaque);
			maxoff = PageGetMaxOffsetNumber(ipage);
			for (offnum = minoff;
				 offnum <= maxoff;
				 offnum = OffsetNumberNext(offnum))
			{
				itup = (IndexTuple) PageGetItem(ipage,
												PageGetItemId(ipage, offnum));
				ItemPointerCopy(&itup->t_tid, &htup.t_self);
				/*
				 * TODO: construct a tid bitmap based on index tids
				 * and fetch heap tids in order afterwards.  That will
				 * also allow validating if a heap tid appears twice
				 * in a unique index.
				 */
				if (!heap_release_fetch(hrel, SnapshotAny, &htup,
										&hbuf, true, NULL))
				{
					elog(ERROR, "btvalidatevacuum: tid (%d,%d) from index %s "
						 "not found in heap %s",
						 ItemPointerGetBlockNumber(&itup->t_tid),
						 ItemPointerGetOffsetNumber(&itup->t_tid),
						 RelationGetRelationName(irel),
						 RelationGetRelationName(hrel));
				}
				switch (HeapTupleSatisfiesVacuum(hrel, htup.t_data, oldest_xmin, hbuf))
				{
					case HEAPTUPLE_RECENTLY_DEAD:
					case HEAPTUPLE_LIVE:
					case HEAPTUPLE_INSERT_IN_PROGRESS:
					case HEAPTUPLE_DELETE_IN_PROGRESS:
						/* these tuples are considered alive by vacuum */
						break;
					case HEAPTUPLE_DEAD:
						elog(ERROR, "btvalidatevacuum: vacuum did not remove "
							 "dead tuple (%d,%d) from heap %s and index %s",
							 ItemPointerGetBlockNumber(&itup->t_tid),
							 ItemPointerGetOffsetNumber(&itup->t_tid),
							 RelationGetRelationName(hrel),
							 RelationGetRelationName(irel));
						break;
					default:
						elog(ERROR, "btvalidatevacuum: invalid visibility");
						break;
				}
				switch(RelationGetRelid(irel))
				{
					case DatabaseOidIndexId:
					case TypeOidIndexId:
					case ClassOidIndexId:
					case ConstraintOidIndexId:
						hoid = HeapTupleGetOid(&htup);
						ioid = index_getattr(itup, 1, RelationGetDescr(irel), &isnull);
						if (hoid != ioid)
						{
							elog(ERROR,
								 "btvalidatevacuum: index oid(%d) != heap oid(%d)"
								 " tuple (%d,%d) index %s", ioid, hoid,
//.........这里部分代码省略.........

/*
 * Rescan end pages to verify that they are (still) empty of tuples.
 *
 * Returns number of nondeletable pages (last nonempty page + 1).
 */
static BlockNumber
count_nondeletable_pages(Relation onerel, LVRelStats *vacrelstats)
{
	BlockNumber blkno;

	/* Strange coding of loop control is needed because blkno is unsigned */
	blkno = vacrelstats->rel_pages;
	while (blkno > vacrelstats->nonempty_pages)
	{
		Buffer		buf;
		Page		page;
		OffsetNumber offnum,
					maxoff;
		bool		hastup;

		/*
		 * We don't insert a vacuum delay point here, because we have an
		 * exclusive lock on the table which we want to hold for as short a
		 * time as possible.  We still need to check for interrupts however.
		 */
		CHECK_FOR_INTERRUPTS();

		blkno--;

		buf = ReadBufferExtended(onerel, MAIN_FORKNUM, blkno,
								 RBM_NORMAL, vac_strategy);

		/* In this phase we only need shared access to the buffer */
		LockBuffer(buf, BUFFER_LOCK_SHARE);

		page = BufferGetPage(buf);

		if (PageIsNew(page) || PageIsEmpty(page))
		{
			/* PageIsNew probably shouldn't happen... */
			UnlockReleaseBuffer(buf);
			continue;
		}

		hastup = false;
		maxoff = PageGetMaxOffsetNumber(page);
		for (offnum = FirstOffsetNumber;
			 offnum <= maxoff;
			 offnum = OffsetNumberNext(offnum))
		{
			ItemId		itemid;

			itemid = PageGetItemId(page, offnum);

			/*
			 * Note: any non-unused item should be taken as a reason to keep
			 * this page.  We formerly thought that DEAD tuples could be
			 * thrown away, but that's not so, because we'd not have cleaned
			 * out their index entries.
			 */
			if (ItemIdIsUsed(itemid))
			{
				hastup = true;
				break;			/* can stop scanning */
			}
		}						/* scan along page */

		UnlockReleaseBuffer(buf);

		/* Done scanning if we found a tuple here */
		if (hastup)
			return blkno + 1;
	}

	/*
	 * If we fall out of the loop, all the previously-thought-to-be-empty
	 * pages still are; we need not bother to look at the last known-nonempty
	 * page.
	 */
	return vacrelstats->nonempty_pages;
}

/*
 * VACUUM cleanup: update FSM
 */
IndexBulkDeleteResult *
gistvacuumcleanup(IndexVacuumInfo *info, IndexBulkDeleteResult *stats)
{
	Relation	rel = info->index;
	BlockNumber npages,
				blkno;
	BlockNumber totFreePages;
	bool		needLock;

	/* No-op in ANALYZE ONLY mode */
	if (info->analyze_only)
		return stats;

	/* Set up all-zero stats if gistbulkdelete wasn't called */
	if (stats == NULL)
	{
		stats = (IndexBulkDeleteResult *) palloc0(sizeof(IndexBulkDeleteResult));
		/* use heap's tuple count */
		stats->num_index_tuples = info->num_heap_tuples;
		stats->estimated_count = info->estimated_count;

		/*
		 * XXX the above is wrong if index is partial.  Would it be OK to just
		 * return NULL, or is there work we must do below?
		 */
	}

	/*
	 * Need lock unless it's local to this backend.
	 */
	needLock = !RELATION_IS_LOCAL(rel);

	/* try to find deleted pages */
	if (needLock)
		LockRelationForExtension(rel, ExclusiveLock);
	npages = RelationGetNumberOfBlocks(rel);
	if (needLock)
		UnlockRelationForExtension(rel, ExclusiveLock);

	totFreePages = 0;
	for (blkno = GIST_ROOT_BLKNO + 1; blkno < npages; blkno++)
	{
		Buffer		buffer;
		Page		page;

		vacuum_delay_point();

		buffer = ReadBufferExtended(rel, MAIN_FORKNUM, blkno, RBM_NORMAL,
									info->strategy);
		LockBuffer(buffer, GIST_SHARE);
		page = (Page) BufferGetPage(buffer);

		if (PageIsNew(page) || GistPageIsDeleted(page))
		{
			totFreePages++;
			RecordFreeIndexPage(rel, blkno);
		}
		UnlockReleaseBuffer(buffer);
	}

	/* Finally, vacuum the FSM */
	IndexFreeSpaceMapVacuum(info->index);

	/* return statistics */
	stats->pages_free = totFreePages;
	if (needLock)
		LockRelationForExtension(rel, ExclusiveLock);
	stats->num_pages = RelationGetNumberOfBlocks(rel);
	if (needLock)
		UnlockRelationForExtension(rel, ExclusiveLock);

	return stats;
}

Datum
gistvacuumcleanup(PG_FUNCTION_ARGS)
{
	IndexVacuumInfo *info = (IndexVacuumInfo *) PG_GETARG_POINTER(0);
	GistBulkDeleteResult *stats = (GistBulkDeleteResult *) PG_GETARG_POINTER(1);
	Relation	rel = info->index;
	BlockNumber npages,
				blkno;
	BlockNumber totFreePages,
				nFreePages,
			   *freePages,
				maxFreePages;
	BlockNumber lastBlock = GIST_ROOT_BLKNO,
				lastFilledBlock = GIST_ROOT_BLKNO;
	bool		needLock;

	/* Set up all-zero stats if gistbulkdelete wasn't called */
	if (stats == NULL)
	{
		stats = (GistBulkDeleteResult *) palloc0(sizeof(GistBulkDeleteResult));
		/* use heap's tuple count */
		Assert(info->num_heap_tuples >= 0);
		stats->std.num_index_tuples = info->num_heap_tuples;

		/*
		 * XXX the above is wrong if index is partial.	Would it be OK to just
		 * return NULL, or is there work we must do below?
		 */
	}

	/* gistVacuumUpdate may cause hard work */
	if (info->vacuum_full)
	{
		GistVacuum	gv;
		ArrayTuple	res;

		/* note: vacuum.c already acquired AccessExclusiveLock on index */

		gv.index = rel;
		initGISTstate(&(gv.giststate), rel);
		gv.opCtx = createTempGistContext();
		gv.result = stats;
		gv.strategy = info->strategy;

		/* walk through the entire index for update tuples */
		res = gistVacuumUpdate(&gv, GIST_ROOT_BLKNO, false);
		/* cleanup */
		if (res.itup)
		{
			int			i;

			for (i = 0; i < res.ituplen; i++)
				pfree(res.itup[i]);
			pfree(res.itup);
		}
		freeGISTstate(&(gv.giststate));
		MemoryContextDelete(gv.opCtx);
	}
	else if (stats->needFullVacuum)
		ereport(NOTICE,
				(errmsg("index \"%s\" needs VACUUM FULL or REINDEX to finish crash recovery",
						RelationGetRelationName(rel))));

	/*
	 * If vacuum full, we already have exclusive lock on the index. Otherwise,
	 * need lock unless it's local to this backend.
	 */
	if (info->vacuum_full)
		needLock = false;
	else
		needLock = !RELATION_IS_LOCAL(rel);

	/* try to find deleted pages */
	if (needLock)
		LockRelationForExtension(rel, ExclusiveLock);
	npages = RelationGetNumberOfBlocks(rel);
	if (needLock)
		UnlockRelationForExtension(rel, ExclusiveLock);

	maxFreePages = npages;
	if (maxFreePages > MaxFSMPages)
		maxFreePages = MaxFSMPages;

	totFreePages = nFreePages = 0;
	freePages = (BlockNumber *) palloc(sizeof(BlockNumber) * maxFreePages);

	for (blkno = GIST_ROOT_BLKNO + 1; blkno < npages; blkno++)
	{
		Buffer		buffer;
		Page		page;

		vacuum_delay_point();

		buffer = ReadBufferWithStrategy(rel, blkno, info->strategy);
		LockBuffer(buffer, GIST_SHARE);
		page = (Page) BufferGetPage(buffer);

		if (PageIsNew(page) || GistPageIsDeleted(page))
		{
			if (nFreePages < maxFreePages)
//.........这里部分代码省略.........

/*
 * Process the pending-TID list between pages of the main scan
 */
static void
spgprocesspending(spgBulkDeleteState *bds)
{
	Relation	index = bds->info->index;
	spgVacPendingItem *pitem;
	spgVacPendingItem *nitem;
	BlockNumber blkno;
	Buffer		buffer;
	Page		page;

	for (pitem = bds->pendingList; pitem != NULL; pitem = pitem->next)
	{
		if (pitem->done)
			continue;			/* ignore already-done items */

		/* call vacuum_delay_point while not holding any buffer lock */
		vacuum_delay_point();

		/* examine the referenced page */
		blkno = ItemPointerGetBlockNumber(&pitem->tid);
		buffer = ReadBufferExtended(index, MAIN_FORKNUM, blkno,
									RBM_NORMAL, bds->info->strategy);
		LockBuffer(buffer, BUFFER_LOCK_EXCLUSIVE);
		page = (Page) BufferGetPage(buffer);

		if (PageIsNew(page) || SpGistPageIsDeleted(page))
		{
			/* Probably shouldn't happen, but ignore it */
		}
		else if (SpGistPageIsLeaf(page))
		{
			if (SpGistBlockIsRoot(blkno))
			{
				/* this should definitely not happen */
				elog(ERROR, "redirection leads to root page of index \"%s\"",
					 RelationGetRelationName(index));
			}

			/* deal with any deletable tuples */
			vacuumLeafPage(bds, index, buffer, true);
			/* might as well do this while we are here */
			vacuumRedirectAndPlaceholder(index, buffer);

			SpGistSetLastUsedPage(index, buffer);

			/*
			 * We can mark as done not only this item, but any later ones
			 * pointing at the same page, since we vacuumed the whole page.
			 */
			pitem->done = true;
			for (nitem = pitem->next; nitem != NULL; nitem = nitem->next)
			{
				if (ItemPointerGetBlockNumber(&nitem->tid) == blkno)
					nitem->done = true;
			}
		}
		else
		{
			/*
			 * On an inner page, visit the referenced inner tuple and add all
			 * its downlinks to the pending list.  We might have pending items
			 * for more than one inner tuple on the same page (in fact this is
			 * pretty likely given the way space allocation works), so get
			 * them all while we are here.
			 */
			for (nitem = pitem; nitem != NULL; nitem = nitem->next)
			{
				if (nitem->done)
					continue;
				if (ItemPointerGetBlockNumber(&nitem->tid) == blkno)
				{
					OffsetNumber offset;
					SpGistInnerTuple innerTuple;

					offset = ItemPointerGetOffsetNumber(&nitem->tid);
					innerTuple = (SpGistInnerTuple) PageGetItem(page,
												PageGetItemId(page, offset));
					if (innerTuple->tupstate == SPGIST_LIVE)
					{
						SpGistNodeTuple node;
						int			i;

						SGITITERATE(innerTuple, i, node)
						{
							if (ItemPointerIsValid(&node->t_tid))
								spgAddPendingTID(bds, &node->t_tid);
						}
					}
					else if (innerTuple->tupstate == SPGIST_REDIRECT)
					{
						/* transfer attention to redirect point */
						spgAddPendingTID(bds,
								   &((SpGistDeadTuple) innerTuple)->pointer);
					}
					else
						elog(ERROR, "unexpected SPGiST tuple state: %d",
							 innerTuple->tupstate);
//.........这里部分代码省略.........

IndexBulkDeleteResult *
ginvacuumcleanup(IndexVacuumInfo *info, IndexBulkDeleteResult *stats)
{
	Relation	index = info->index;
	bool		needLock;
	BlockNumber npages,
				blkno;
	BlockNumber totFreePages;
	GinState	ginstate;
	GinStatsData idxStat;

	/*
	 * In an autovacuum analyze, we want to clean up pending insertions.
	 * Otherwise, an ANALYZE-only call is a no-op.
	 */
	if (info->analyze_only)
	{
		if (IsAutoVacuumWorkerProcess())
		{
			initGinState(&ginstate, index);
			ginInsertCleanup(&ginstate, true, stats);
		}
		return stats;
	}

	/*
	 * Set up all-zero stats and cleanup pending inserts if ginbulkdelete
	 * wasn't called
	 */
	if (stats == NULL)
	{
		stats = (IndexBulkDeleteResult *) palloc0(sizeof(IndexBulkDeleteResult));
		initGinState(&ginstate, index);
		ginInsertCleanup(&ginstate, false, stats);
	}

	memset(&idxStat, 0, sizeof(idxStat));

	/*
	 * XXX we always report the heap tuple count as the number of index
	 * entries.  This is bogus if the index is partial, but it's real hard to
	 * tell how many distinct heap entries are referenced by a GIN index.
	 */
	stats->num_index_tuples = info->num_heap_tuples;
	stats->estimated_count = info->estimated_count;

	/*
	 * Need lock unless it's local to this backend.
	 */
	needLock = !RELATION_IS_LOCAL(index);

	if (needLock)
		LockRelationForExtension(index, ExclusiveLock);
	npages = RelationGetNumberOfBlocks(index);
	if (needLock)
		UnlockRelationForExtension(index, ExclusiveLock);

	totFreePages = 0;

	for (blkno = GIN_ROOT_BLKNO; blkno < npages; blkno++)
	{
		Buffer		buffer;
		Page		page;

		vacuum_delay_point();

		buffer = ReadBufferExtended(index, MAIN_FORKNUM, blkno,
									RBM_NORMAL, info->strategy);
		LockBuffer(buffer, GIN_SHARE);
		page = (Page) BufferGetPage(buffer);

		if (PageIsNew(page) || GinPageIsDeleted(page))
		{
			Assert(blkno != GIN_ROOT_BLKNO);
			RecordFreeIndexPage(index, blkno);
			totFreePages++;
		}
		else if (GinPageIsData(page))
		{
			idxStat.nDataPages++;
		}
		else if (!GinPageIsList(page))
		{
			idxStat.nEntryPages++;

			if (GinPageIsLeaf(page))
				idxStat.nEntries += PageGetMaxOffsetNumber(page);
		}

		UnlockReleaseBuffer(buffer);
	}

	/* Update the metapage with accurate page and entry counts */
	idxStat.nTotalPages = npages;
	ginUpdateStats(info->index, &idxStat);

	/* Finally, vacuum the FSM */
	IndexFreeSpaceMapVacuum(info->index);

	stats->pages_free = totFreePages;
//.........这里部分代码省略.........

//.........这里部分代码省略.........
		 * code above.
		 */
		LockBuffer(buffer, BUFFER_LOCK_UNLOCK);
		if (otherBuffer == InvalidBuffer)
			ReleaseBuffer(buffer);
		else if (otherBlock != targetBlock)
		{
			LockBuffer(otherBuffer, BUFFER_LOCK_UNLOCK);
			ReleaseBuffer(buffer);
		}

		/* Without FSM, always fall out of the loop and extend */
		if (!use_fsm)
			break;

		/*
		 * Update FSM as to condition of this page, and ask for another page
		 * to try.
		 */
		targetBlock = RecordAndGetPageWithFreeSpace(&relation->rd_node,
													targetBlock,
													pageFreeSpace,
													len + saveFreeSpace);
	}

	/*
	 * Have to extend the relation.
	 *
	 * 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.  We
	 * can skip locking for new or temp relations, however, since no one else
	 * could be accessing them.
	 */
	needLock = !RELATION_IS_LOCAL(relation);

	if (needLock)
		LockRelationForExtension(relation, ExclusiveLock);

	/*
	 * 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?
	 */
	buffer = ReadBuffer(relation, P_NEW);

	/*
	 * We can be certain that locking the otherBuffer first is OK, since it
	 * must have a lower page number.
	 */
	if (otherBuffer != InvalidBuffer)
		LockBuffer(otherBuffer, BUFFER_LOCK_EXCLUSIVE);

	/*
	 * Now acquire lock on the new page.
	 */
	LockBuffer(buffer, BUFFER_LOCK_EXCLUSIVE);

	/*
	 * Release the file-extension lock; it's now OK for someone else to extend
	 * the relation some more.	Note that we cannot release this lock before
	 * we have buffer lock on the new page, or we risk a race condition
	 * against vacuumlazy.c --- see comments therein.
	 */
	if (needLock)
		UnlockRelationForExtension(relation, ExclusiveLock);

	/*
	 * We need to initialize the empty new page.  Double-check that it really
	 * is empty (this should never happen, but if it does we don't want to
	 * risk wiping out valid data).
	 */
	pageHeader = (Page) BufferGetPage(buffer);

	if (!PageIsNew((PageHeader) pageHeader))
		elog(ERROR, "page %u of relation \"%s\" should be empty but is not",
			 BufferGetBlockNumber(buffer),
			 RelationGetRelationName(relation));

	PageInit(pageHeader, BufferGetPageSize(buffer), 0);

	if (len > PageGetFreeSpace(pageHeader))
	{
		/* We should not get here given the test at the top */
		elog(PANIC, "tuple is too big: size %lu", (unsigned long) len);
	}

	/*
	 * Remember the new page as our target for future insertions.
	 *
	 * XXX should we enter the new page into the free space map immediately,
	 * or just keep it for this backend's exclusive use in the short run
	 * (until VACUUM sees it)?	Seems to depend on whether you expect the
	 * current backend to make more insertions or not, which is probably a
	 * good bet most of the time.  So for now, don't add it to FSM yet.
	 */
	relation->rd_targblock = BufferGetBlockNumber(buffer);

	return buffer;
}

/*
 * Rescan end pages to verify that they are (still) empty of tuples.
 *
 * Returns number of nondeletable pages (last nonempty page + 1).
 */
static BlockNumber
count_nondeletable_pages(Relation onerel, LVRelStats *vacrelstats)
{
	BlockNumber blkno;
	instr_time	starttime;
	instr_time	currenttime;
	instr_time	elapsed;

	/* Initialize the starttime if we check for conflicting lock requests */
	INSTR_TIME_SET_CURRENT(starttime);

	/* Strange coding of loop control is needed because blkno is unsigned */
	blkno = vacrelstats->rel_pages;
	while (blkno > vacrelstats->nonempty_pages)
	{
		Buffer		buf;
		Page		page;
		OffsetNumber offnum,
					maxoff;
		bool		hastup;

		/*
		 * Check if another process requests a lock on our relation. We are
		 * holding an AccessExclusiveLock here, so they will be waiting. We
		 * only do this in autovacuum_truncate_lock_check millisecond
		 * intervals, and we only check if that interval has elapsed once
		 * every 32 blocks to keep the number of system calls and actual
		 * shared lock table lookups to a minimum.
		 */
		if ((blkno % 32) == 0)
		{
			INSTR_TIME_SET_CURRENT(currenttime);
			elapsed = currenttime;
			INSTR_TIME_SUBTRACT(elapsed, starttime);
			if ((INSTR_TIME_GET_MICROSEC(elapsed) / 1000)
				>= AUTOVACUUM_TRUNCATE_LOCK_CHECK_INTERVAL)
			{
				if (LockHasWaitersRelation(onerel, AccessExclusiveLock))
				{
					ereport(elevel,
							(errmsg("\"%s\": suspending truncate "
									"due to conflicting lock request",
									RelationGetRelationName(onerel))));

					vacrelstats->lock_waiter_detected = true;
					return blkno;
				}
				starttime = currenttime;
			}
		}

		/*
		 * We don't insert a vacuum delay point here, because we have an
		 * exclusive lock on the table which we want to hold for as short a
		 * time as possible.  We still need to check for interrupts however.
		 */
		CHECK_FOR_INTERRUPTS();

		blkno--;

		buf = ReadBufferExtended(onerel, MAIN_FORKNUM, blkno,
								 RBM_NORMAL, vac_strategy);

		/* In this phase we only need shared access to the buffer */
		LockBuffer(buf, BUFFER_LOCK_SHARE);

		page = BufferGetPage(buf);

		if (PageIsNew(page) || PageIsEmpty(page))
		{
			/* PageIsNew probably shouldn't happen... */
			UnlockReleaseBuffer(buf);
			continue;
		}

		hastup = false;
		maxoff = PageGetMaxOffsetNumber(page);
		for (offnum = FirstOffsetNumber;
			 offnum <= maxoff;
			 offnum = OffsetNumberNext(offnum))
		{
			ItemId		itemid;

			itemid = PageGetItemId(page, offnum);

			/*
			 * Note: any non-unused item should be taken as a reason to keep
			 * this page.  We formerly thought that DEAD tuples could be
			 * thrown away, but that's not so, because we'd not have cleaned
			 * out their index entries.
			 */
			if (ItemIdIsUsed(itemid))
			{
				hastup = true;
				break;			/* can stop scanning */
//.........这里部分代码省略.........

/*
 * Extend a relation by multiple blocks to avoid future contention on the
 * relation extension lock.  Our goal is to pre-extend the relation by an
 * amount which ramps up as the degree of contention ramps up, but limiting
 * the result to some sane overall value.
 */
static void
RelationAddExtraBlocks(Relation relation, BulkInsertState bistate)
{
	BlockNumber blockNum,
				firstBlock = InvalidBlockNumber;
	int			extraBlocks;
	int			lockWaiters;

	/* Use the length of the lock wait queue to judge how much to extend. */
	lockWaiters = RelationExtensionLockWaiterCount(relation);
	if (lockWaiters <= 0)
		return;

	/*
	 * It might seem like multiplying the number of lock waiters by as much as
	 * 20 is too aggressive, but benchmarking revealed that smaller numbers
	 * were insufficient.  512 is just an arbitrary cap to prevent
	 * pathological results.
	 */
	extraBlocks = Min(512, lockWaiters * 20);

	do
	{
		Buffer		buffer;
		Page		page;
		Size		freespace;

		/*
		 * Extend by one page.  This should generally match the main-line
		 * extension code in RelationGetBufferForTuple, except that we hold
		 * the relation extension lock throughout.
		 */
		buffer = ReadBufferBI(relation, P_NEW, bistate);

		LockBuffer(buffer, BUFFER_LOCK_EXCLUSIVE);
		page = BufferGetPage(buffer);

		if (!PageIsNew(page))
			elog(ERROR, "page %u of relation \"%s\" should be empty but is not",
				 BufferGetBlockNumber(buffer),
				 RelationGetRelationName(relation));

		PageInit(page, BufferGetPageSize(buffer), 0);

		/*
		 * We mark all the new buffers dirty, but do nothing to write them
		 * out; they'll probably get used soon, and even if they are not, a
		 * crash will leave an okay all-zeroes page on disk.
		 */
		MarkBufferDirty(buffer);

		/* we'll need this info below */
		blockNum = BufferGetBlockNumber(buffer);
		freespace = PageGetHeapFreeSpace(page);

		UnlockReleaseBuffer(buffer);

		/* Remember first block number thus added. */
		if (firstBlock == InvalidBlockNumber)
			firstBlock = blockNum;

		/*
		 * Immediately update the bottom level of the FSM.  This has a good
		 * chance of making this page visible to other concurrently inserting
		 * backends, and we want that to happen without delay.
		 */
		RecordPageWithFreeSpace(relation, blockNum, freespace);
	}
	while (--extraBlocks > 0);

	/*
	 * Updating the upper levels of the free space map is too expensive to do
	 * for every block, but it's worth doing once at the end to make sure that
	 * subsequent insertion activity sees all of those nifty free pages we
	 * just inserted.
	 */
	FreeSpaceMapVacuumRange(relation, firstBlock, blockNum + 1);
}

/*
 * _bitmap_init() -- initialize the bitmap index.
 *
 * Create the meta page, a new heap which stores the distinct values for
 * the attributes to be indexed, a btree index on this new heap for searching
 * those distinct values, and the first LOV page.
 */
void
_bitmap_init(Relation rel, Oid comptypeOid,
			 Oid heapOid, Oid indexOid,
			 Oid heapRelfilenode, Oid indexRelfilenode,
			 bool use_wal)
{
	MIRROREDLOCK_BUFMGR_DECLARE;

	BMMetaPage		metapage;
	Buffer			metabuf;
	Page			page;
	Buffer			buf;
	BMLOVItem 		lovItem;
	OffsetNumber	newOffset;
	Page			currLovPage;
	OffsetNumber	o;
  
	/* sanity check */
	if (RelationGetNumberOfBlocks(rel) != 0)
		ereport(ERROR,
				(errcode(ERRCODE_INDEX_CORRUPTED),
				errmsg("cannot initialize non-empty bitmap index \"%s\"",
				RelationGetRelationName(rel)),
				errSendAlert(true)));
	
	// -------- MirroredLock ----------
	MIRROREDLOCK_BUFMGR_LOCK;
	
	/* create the metapage */
	metabuf = _bitmap_getbuf(rel, P_NEW, BM_WRITE);
	page = BufferGetPage(metabuf);
	Assert(PageIsNew(page));

	/* initialize the LOV metadata */
	_bitmap_create_lov_heapandindex(rel, comptypeOid,
									&(heapOid),
									&(indexOid),
									heapRelfilenode,
									indexRelfilenode);

	START_CRIT_SECTION();

	MarkBufferDirty(metabuf);

	/* initialize the metapage */
	PageInit(page, BufferGetPageSize(metabuf), 0);
	metapage = (BMMetaPage) PageGetContents(page);
	
	metapage->bm_magic = BITMAP_MAGIC;
	metapage->bm_version = BITMAP_VERSION;
	metapage->bm_lov_heapId = heapOid;
	metapage->bm_lov_indexId = indexOid;

	if (use_wal)
		_bitmap_log_metapage(rel, page);

	/* allocate the first LOV page. */
	buf = _bitmap_getbuf(rel, P_NEW, BM_WRITE);
	_bitmap_init_lovpage(rel, buf);

	MarkBufferDirty(buf);

	currLovPage = BufferGetPage(buf);

	/* set the first item to support NULL value */
	lovItem = _bitmap_formitem(0);
	newOffset = OffsetNumberNext(PageGetMaxOffsetNumber(currLovPage));

	/*
	 * XXX: perhaps this could be a special page, with more efficient storage
	 * after all, we have fixed size data
	 */
	o = PageAddItem(currLovPage, (Item)lovItem, sizeof(BMLOVItemData),
                    newOffset, LP_USED);

	if (o == InvalidOffsetNumber)
		ereport(ERROR,
				(errcode(ERRCODE_INTERNAL_ERROR),
				 errmsg("failed to add LOV item to \"%s\"",
				 RelationGetRelationName(rel))));

	metapage->bm_lov_lastpage = BufferGetBlockNumber(buf);

	if(use_wal)
		_bitmap_log_lovitem(rel, buf, newOffset, lovItem, metabuf, true);

	END_CRIT_SECTION();

	_bitmap_wrtbuf(buf);
	_bitmap_wrtbuf(metabuf);
	
	MIRROREDLOCK_BUFMGR_UNLOCK;
	// -------- MirroredLock ----------
//.........这里部分代码省略.........

/*
 * Read a FSM page.
 *
 * If the page doesn't exist, InvalidBuffer is returned, or if 'extend' is
 * true, the FSM file is extended.
 */
static Buffer
fsm_readbuf(Relation rel, FSMAddress addr, bool extend)
{
	BlockNumber blkno = fsm_logical_to_physical(addr);
	Buffer		buf;

	RelationOpenSmgr(rel);

	/*
	 * If we haven't cached the size of the FSM yet, check it first.  Also
	 * recheck if the requested block seems to be past end, since our cached
	 * value might be stale.  (We send smgr inval messages on truncation, but
	 * not on extension.)
	 */
	if (rel->rd_smgr->smgr_fsm_nblocks == InvalidBlockNumber ||
		blkno >= rel->rd_smgr->smgr_fsm_nblocks)
	{
		if (smgrexists(rel->rd_smgr, FSM_FORKNUM))
			rel->rd_smgr->smgr_fsm_nblocks = smgrnblocks(rel->rd_smgr,
														 FSM_FORKNUM);
		else
			rel->rd_smgr->smgr_fsm_nblocks = 0;
	}

	/* Handle requests beyond EOF */
	if (blkno >= rel->rd_smgr->smgr_fsm_nblocks)
	{
		if (extend)
			fsm_extend(rel, blkno + 1);
		else
			return InvalidBuffer;
	}

	/*
	 * Use ZERO_ON_ERROR mode, and initialize the page if necessary. The FSM
	 * information is not accurate anyway, so it's better to clear corrupt
	 * pages than error out. Since the FSM changes are not WAL-logged, the
	 * so-called torn page problem on crash can lead to pages with corrupt
	 * headers, for example.
	 *
	 * The initialize-the-page part is trickier than it looks, because of the
	 * possibility of multiple backends doing this concurrently, and our
	 * desire to not uselessly take the buffer lock in the normal path where
	 * the page is OK.  We must take the lock to initialize the page, so
	 * recheck page newness after we have the lock, in case someone else
	 * already did it.  Also, because we initially check PageIsNew with no
	 * lock, it's possible to fall through and return the buffer while someone
	 * else is still initializing the page (i.e., we might see pd_upper as set
	 * but other page header fields are still zeroes).  This is harmless for
	 * callers that will take a buffer lock themselves, but some callers
	 * inspect the page without any lock at all.  The latter is OK only so
	 * long as it doesn't depend on the page header having correct contents.
	 * Current usage is safe because PageGetContents() does not require that.
	 */
	buf = ReadBufferExtended(rel, FSM_FORKNUM, blkno, RBM_ZERO_ON_ERROR, NULL);
	if (PageIsNew(BufferGetPage(buf)))
	{
		LockBuffer(buf, BUFFER_LOCK_EXCLUSIVE);
		if (PageIsNew(BufferGetPage(buf)))
			PageInit(BufferGetPage(buf), BLCKSZ, 0);
		LockBuffer(buf, BUFFER_LOCK_UNLOCK);
	}
	return buf;
}

/*
 * Insert all matching tuples into a bitmap.
 */
int64
blgetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
{
	int64		ntids = 0;
	BlockNumber blkno = BLOOM_HEAD_BLKNO,
				npages;
	int			i;
	BufferAccessStrategy bas;
	BloomScanOpaque so = (BloomScanOpaque) scan->opaque;

	if (so->sign == NULL)
	{
		/* New search: have to calculate search signature */
		ScanKey		skey = scan->keyData;

		so->sign = palloc0(sizeof(BloomSignatureWord) * so->state.opts.bloomLength);

		for (i = 0; i < scan->numberOfKeys; i++)
		{
			/*
			 * Assume bloom-indexable operators to be strict, so nothing could
			 * be found for NULL key.
			 */
			if (skey->sk_flags & SK_ISNULL)
			{
				pfree(so->sign);
				so->sign = NULL;
				return 0;
			}

			/* Add next value to the signature */
			signValue(&so->state, so->sign, skey->sk_argument,
					  skey->sk_attno - 1);

			skey++;
		}
	}

	/*
	 * We're going to read the whole index. This is why we use appropriate
	 * buffer access strategy.
	 */
	bas = GetAccessStrategy(BAS_BULKREAD);
	npages = RelationGetNumberOfBlocks(scan->indexRelation);

	for (blkno = BLOOM_HEAD_BLKNO; blkno < npages; blkno++)
	{
		Buffer		buffer;
		Page		page;

		buffer = ReadBufferExtended(scan->indexRelation, MAIN_FORKNUM,
									blkno, RBM_NORMAL, bas);

		LockBuffer(buffer, BUFFER_LOCK_SHARE);
		page = BufferGetPage(buffer);
		TestForOldSnapshot(scan->xs_snapshot, scan->indexRelation, page);

		if (!PageIsNew(page) && !BloomPageIsDeleted(page))
		{
			OffsetNumber offset,
						maxOffset = BloomPageGetMaxOffset(page);

			for (offset = 1; offset <= maxOffset; offset++)
			{
				BloomTuple *itup = BloomPageGetTuple(&so->state, page, offset);
				bool		res = true;

				/* Check index signature with scan signature */
				for (i = 0; i < so->state.opts.bloomLength; i++)
				{
					if ((itup->sign[i] & so->sign[i]) != so->sign[i])
					{
						res = false;
						break;
					}
				}

				/* Add matching tuples to bitmap */
				if (res)
				{
					tbm_add_tuples(tbm, &itup->heapPtr, 1, true);
					ntids++;
				}
			}
		}

		UnlockReleaseBuffer(buffer);
		CHECK_FOR_INTERRUPTS();
	}
	FreeAccessStrategy(bas);

	return ntids;
}

/*
 * For a newly inserted heap tid, check if an entry with this tid
 * already exists in a unique index.  If it does, abort the inserting
 * transaction.
 */
static void
_bt_validate_tid(Relation irel, ItemPointer h_tid)
{
	MIRROREDLOCK_BUFMGR_DECLARE;

	BlockNumber blkno;
	BlockNumber num_pages;
	Buffer buf;
	Page page;
	BTPageOpaque opaque;
	IndexTuple itup;
	OffsetNumber maxoff,
			minoff,
			offnum;

	elog(DEBUG1, "validating tid (%d,%d) for index (%s)",
		 ItemPointerGetBlockNumber(h_tid), ItemPointerGetOffsetNumber(h_tid),
		 RelationGetRelationName(irel));

	blkno = BTREE_METAPAGE + 1;
	num_pages = RelationGetNumberOfBlocks(irel);

	MIRROREDLOCK_BUFMGR_LOCK;
	for (; blkno < num_pages; blkno++)
	{
		buf = ReadBuffer(irel, blkno);
		page = BufferGetPage(buf);
		opaque = (BTPageOpaque) PageGetSpecialPointer(page);
		if (!PageIsNew(page))
			_bt_checkpage(irel, buf);
		if (P_ISLEAF(opaque))
		{
			minoff = P_FIRSTDATAKEY(opaque);
			maxoff = PageGetMaxOffsetNumber(page);
			for (offnum = minoff;
				 offnum <= maxoff;
				 offnum = OffsetNumberNext(offnum))
			{
				itup = (IndexTuple) PageGetItem(page,
												PageGetItemId(page, offnum));
				if (ItemPointerEquals(&itup->t_tid, h_tid))
				{
					Form_pg_attribute key_att = RelationGetDescr(irel)->attrs[0];
					Oid key = InvalidOid;
					bool isnull;
					if (key_att->atttypid == OIDOID)
					{
						key = DatumGetInt32(
								index_getattr(itup, 1, RelationGetDescr(irel), &isnull));
						elog(ERROR, "found tid (%d,%d), %s (%d) already in index (%s)",
							 ItemPointerGetBlockNumber(h_tid), ItemPointerGetOffsetNumber(h_tid),
							 NameStr(key_att->attname), key, RelationGetRelationName(irel));
					}
					else
					{
						elog(ERROR, "found tid (%d,%d) already in index (%s)",
							 ItemPointerGetBlockNumber(h_tid), ItemPointerGetOffsetNumber(h_tid),
							 RelationGetRelationName(irel));
					}
				}
			}
		}
		ReleaseBuffer(buf);
	}
	MIRROREDLOCK_BUFMGR_UNLOCK;
}

/*
 * btvacuumpage --- VACUUM one page
 *
 * This processes a single page for btvacuumscan().  In some cases we
 * must go back and re-examine previously-scanned pages; this routine
 * recurses when necessary to handle that case.
 *
 * blkno is the page to process.  orig_blkno is the highest block number
 * reached by the outer btvacuumscan loop (the same as blkno, unless we
 * are recursing to re-examine a previous page).
 */
static void
btvacuumpage(BTVacState *vstate, BlockNumber blkno, BlockNumber orig_blkno)
{
	IndexVacuumInfo *info = vstate->info;
	IndexBulkDeleteResult *stats = vstate->stats;
	IndexBulkDeleteCallback callback = vstate->callback;
	void	   *callback_state = vstate->callback_state;
	Relation	rel = info->index;
	bool		delete_now;
	BlockNumber recurse_to;
	Buffer		buf;
	Page		page;
	BTPageOpaque opaque;

restart:
	delete_now = false;
	recurse_to = P_NONE;

	/* call vacuum_delay_point while not holding any buffer lock */
	vacuum_delay_point();

	/*
	 * We can't use _bt_getbuf() here because it always applies
	 * _bt_checkpage(), which will barf on an all-zero page. We want to
	 * recycle all-zero pages, not fail.  Also, we want to use a nondefault
	 * buffer access strategy.
	 */
	buf = ReadBufferExtended(rel, MAIN_FORKNUM, blkno, RBM_NORMAL,
							 info->strategy);
	LockBuffer(buf, BT_READ);
	page = BufferGetPage(buf);
	opaque = (BTPageOpaque) PageGetSpecialPointer(page);
	if (!PageIsNew(page))
		_bt_checkpage(rel, buf);

	/*
	 * If we are recursing, the only case we want to do anything with is a
	 * live leaf page having the current vacuum cycle ID.  Any other state
	 * implies we already saw the page (eg, deleted it as being empty).
	 */
	if (blkno != orig_blkno)
	{
		if (_bt_page_recyclable(page) ||
			P_IGNORE(opaque) ||
			!P_ISLEAF(opaque) ||
			opaque->btpo_cycleid != vstate->cycleid)
		{
			_bt_relbuf(rel, buf);
			return;
		}
	}

	/* If the page is in use, update lastUsedPage */
	if (!_bt_page_recyclable(page) && vstate->lastUsedPage < blkno)
		vstate->lastUsedPage = blkno;

	/* Page is valid, see what to do with it */
	if (_bt_page_recyclable(page))
	{
		/* Okay to recycle this page */
		RecordFreeIndexPage(rel, blkno);
		vstate->totFreePages++;
		stats->pages_deleted++;
	}
	else if (P_ISDELETED(opaque))
	{
		/* Already deleted, but can't recycle yet */
		stats->pages_deleted++;
	}
	else if (P_ISHALFDEAD(opaque))
	{
		/* Half-dead, try to delete */
		delete_now = true;
	}
	else if (P_ISLEAF(opaque))
	{
		OffsetNumber deletable[MaxOffsetNumber];
		int			ndeletable;
		OffsetNumber offnum,
					minoff,
					maxoff;

		/*
		 * Trade in the initial read lock for a super-exclusive write lock on
		 * this page.  We must get such a lock on every leaf page over the
		 * course of the vacuum scan, whether or not it actually contains any
		 * deletable tuples --- see nbtree/README.
		 */
		LockBuffer(buf, BUFFER_LOCK_UNLOCK);
//.........这里部分代码省略.........

/*
 * Insert new tuple to the bloom index.
 */
bool
blinsert(Relation index, Datum *values, bool *isnull,
		 ItemPointer ht_ctid, Relation heapRel,
		 IndexUniqueCheck checkUnique,
		 IndexInfo *indexInfo)
{
	BloomState	blstate;
	BloomTuple *itup;
	MemoryContext oldCtx;
	MemoryContext insertCtx;
	BloomMetaPageData *metaData;
	Buffer		buffer,
				metaBuffer;
	Page		page,
				metaPage;
	BlockNumber blkno = InvalidBlockNumber;
	OffsetNumber nStart;
	GenericXLogState *state;

	insertCtx = AllocSetContextCreate(CurrentMemoryContext,
									  "Bloom insert temporary context",
									  ALLOCSET_DEFAULT_SIZES);

	oldCtx = MemoryContextSwitchTo(insertCtx);

	initBloomState(&blstate, index);
	itup = BloomFormTuple(&blstate, ht_ctid, values, isnull);

	/*
	 * At first, try to insert new tuple to the first page in notFullPage
	 * array.  If successful, we don't need to modify the meta page.
	 */
	metaBuffer = ReadBuffer(index, BLOOM_METAPAGE_BLKNO);
	LockBuffer(metaBuffer, BUFFER_LOCK_SHARE);
	metaData = BloomPageGetMeta(BufferGetPage(metaBuffer));

	if (metaData->nEnd > metaData->nStart)
	{
		Page		page;

		blkno = metaData->notFullPage[metaData->nStart];
		Assert(blkno != InvalidBlockNumber);

		/* Don't hold metabuffer lock while doing insert */
		LockBuffer(metaBuffer, BUFFER_LOCK_UNLOCK);

		buffer = ReadBuffer(index, blkno);
		LockBuffer(buffer, BUFFER_LOCK_EXCLUSIVE);

		state = GenericXLogStart(index);
		page = GenericXLogRegisterBuffer(state, buffer, 0);

		/*
		 * We might have found a page that was recently deleted by VACUUM.  If
		 * so, we can reuse it, but we must reinitialize it.
		 */
		if (PageIsNew(page) || BloomPageIsDeleted(page))
			BloomInitPage(page, 0);

		if (BloomPageAddItem(&blstate, page, itup))
		{
			/* Success!  Apply the change, clean up, and exit */
			GenericXLogFinish(state);
			UnlockReleaseBuffer(buffer);
			ReleaseBuffer(metaBuffer);
			MemoryContextSwitchTo(oldCtx);
			MemoryContextDelete(insertCtx);
			return false;
		}

		/* Didn't fit, must try other pages */
		GenericXLogAbort(state);
		UnlockReleaseBuffer(buffer);
	}
	else
	{
		/* No entries in notFullPage */
		LockBuffer(metaBuffer, BUFFER_LOCK_UNLOCK);
	}

	/*
	 * Try other pages in notFullPage array.  We will have to change nStart in
	 * metapage.  Thus, grab exclusive lock on metapage.
	 */
	LockBuffer(metaBuffer, BUFFER_LOCK_EXCLUSIVE);

	/* nStart might have changed while we didn't have lock */
	nStart = metaData->nStart;

	/* Skip first page if we already tried it above */
	if (nStart < metaData->nEnd &&
		blkno == metaData->notFullPage[nStart])
		nStart++;

	/*
	 * This loop iterates for each page we try from the notFullPage array, and
	 * will also initialize a GenericXLogState for the fallback case of having
//.........这里部分代码省略.........

/*
 *	lazy_scan_heap() -- scan an open heap relation
 *
 *		This routine sets commit status bits, builds lists of dead tuples
 *		and pages with free space, and calculates statistics on the number
 *		of live tuples in the heap.  When done, or when we run low on space
 *		for dead-tuple TIDs, invoke vacuuming of indexes and heap.
 *
 *		If there are no indexes then we just vacuum each dirty page as we
 *		process it, since there's no point in gathering many tuples.
 */
static void
lazy_scan_heap(Relation onerel, LVRelStats *vacrelstats,
			   Relation *Irel, int nindexes, List *updated_stats)
{
	MIRROREDLOCK_BUFMGR_DECLARE;

	BlockNumber nblocks,
				blkno;
	HeapTupleData tuple;
	char	   *relname;
	BlockNumber empty_pages,
				vacuumed_pages;
	double		num_tuples,
				tups_vacuumed,
				nkeep,
				nunused;
	IndexBulkDeleteResult **indstats;
	int			i;
	int reindex_count = 1;
	PGRUsage	ru0;

	/* Fetch gp_persistent_relation_node information that will be added to XLOG record. */
	RelationFetchGpRelationNodeForXLog(onerel);

	pg_rusage_init(&ru0);

	relname = RelationGetRelationName(onerel);
	ereport(elevel,
			(errmsg("vacuuming \"%s.%s\"",
					get_namespace_name(RelationGetNamespace(onerel)),
					relname)));

	empty_pages = vacuumed_pages = 0;
	num_tuples = tups_vacuumed = nkeep = nunused = 0;

	indstats = (IndexBulkDeleteResult **)
		palloc0(nindexes * sizeof(IndexBulkDeleteResult *));

	nblocks = RelationGetNumberOfBlocks(onerel);
	vacrelstats->rel_pages = nblocks;
	vacrelstats->nonempty_pages = 0;

	lazy_space_alloc(vacrelstats, nblocks);

	for (blkno = 0; blkno < nblocks; blkno++)
	{
		Buffer		buf;
		Page		page;
		OffsetNumber offnum,
					maxoff;
		bool		tupgone,
					hastup;
		int			prev_dead_count;
		OffsetNumber frozen[MaxOffsetNumber];
		int			nfrozen;

		vacuum_delay_point();

		/*
		 * If we are close to overrunning the available space for dead-tuple
		 * TIDs, pause and do a cycle of vacuuming before we tackle this page.
		 */
		if ((vacrelstats->max_dead_tuples - vacrelstats->num_dead_tuples) < MaxHeapTuplesPerPage &&
			vacrelstats->num_dead_tuples > 0)
		{
			/* Remove index entries */
			for (i = 0; i < nindexes; i++)
				lazy_vacuum_index(Irel[i], &indstats[i], vacrelstats);

			reindex_count++;

			/* Remo