/*
 * Delete item(s) from a btree page.
 *
 * This must only be used for deleting leaf items.	Deleting an item on a
 * non-leaf page has to be done as part of an atomic action that includes
 * deleting the page it points to.
 *
 * This routine assumes that the caller has pinned and locked the buffer.
 * Also, the given itemnos *must* appear in increasing order in the array.
 */
void
_bt_delitems(Relation rel, Buffer buf,
			 OffsetNumber *itemnos, int nitems,
			 bool inVacuum)
{
	Page		page;
	BTPageOpaque opaque;

	MIRROREDLOCK_BUFMGR_MUST_ALREADY_BE_HELD;

	page = BufferGetPage(buf);

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

	/* No ereport(ERROR) until changes are logged */
	START_CRIT_SECTION();

	/* Fix the page */
	PageIndexMultiDelete(page, itemnos, nitems);

	/*
	 * If this is within VACUUM, we can clear the vacuum cycleID since this
	 * page has certainly been processed by the current vacuum scan.
	 */
	opaque = (BTPageOpaque) PageGetSpecialPointer(page);
	if (inVacuum)
		opaque->btpo_cycleid = 0;

	/*
	 * Mark the page as not containing any LP_DELETE 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)
	{
		xl_btree_delete xlrec;
		XLogRecPtr	recptr;
		XLogRecData rdata[2];

		xl_btreenode_set(&(xlrec.btreenode), rel);
		xlrec.block = BufferGetBlockNumber(buf);

		rdata[0].data = (char *) &xlrec;
		rdata[0].len = SizeOfBtreeDelete;
		rdata[0].buffer = InvalidBuffer;
		rdata[0].next = &(rdata[1]);

		/*
		 * The target-offsets array is not in the buffer, but pretend that it
		 * is.	When XLogInsert stores the whole buffer, the offsets array
		 * need not be stored too.
		 */
		if (nitems > 0)
		{
			rdata[1].data = (char *) itemnos;
			rdata[1].len = nitems * sizeof(OffsetNumber);
		}
		else
		{
			rdata[1].data = NULL;
			rdata[1].len = 0;
		}
		rdata[1].buffer = buf;
		rdata[1].buffer_std = true;
		rdata[1].next = NULL;

		recptr = XLogInsert(RM_BTREE_ID, XLOG_BTREE_DELETE, rdata);

		PageSetLSN(page, recptr);
		PageSetTLI(page, ThisTimeLineID);
	}

	END_CRIT_SECTION();
}

//.........这里部分代码省略.........

	LWLockAcquire(slot->io_in_progress_lock, LW_EXCLUSIVE);

	/* silence valgrind :( */
	memset(&cp, 0, sizeof(ReplicationSlotOnDisk));

	sprintf(tmppath, "%s/state.tmp", dir);
	sprintf(path, "%s/state", dir);

	fd = OpenTransientFile(tmppath,
						   O_CREAT | O_EXCL | O_WRONLY | PG_BINARY,
						   S_IRUSR | S_IWUSR);
	if (fd < 0)
	{
		ereport(elevel,
				(errcode_for_file_access(),
				 errmsg("could not create file \"%s\": %m",
						tmppath)));
		return;
	}

	cp.magic = SLOT_MAGIC;
	INIT_CRC32C(cp.checksum);
	cp.version = SLOT_VERSION;
	cp.length = ReplicationSlotOnDiskV2Size;

	SpinLockAcquire(&slot->mutex);

	memcpy(&cp.slotdata, &slot->data, sizeof(ReplicationSlotPersistentData));

	SpinLockRelease(&slot->mutex);

	COMP_CRC32C(cp.checksum,
				(char *) (&cp) + SnapBuildOnDiskNotChecksummedSize,
				SnapBuildOnDiskChecksummedSize);
	FIN_CRC32C(cp.checksum);

	if ((write(fd, &cp, sizeof(cp))) != sizeof(cp))
	{
		int			save_errno = errno;

		CloseTransientFile(fd);
		errno = save_errno;
		ereport(elevel,
				(errcode_for_file_access(),
				 errmsg("could not write to file \"%s\": %m",
						tmppath)));
		return;
	}

	/* fsync the temporary file */
	if (pg_fsync(fd) != 0)
	{
		int			save_errno = errno;

		CloseTransientFile(fd);
		errno = save_errno;
		ereport(elevel,
				(errcode_for_file_access(),
				 errmsg("could not fsync file \"%s\": %m",
						tmppath)));
		return;
	}

	CloseTransientFile(fd);

	/* rename to permanent file, fsync file and directory */
	if (rename(tmppath, path) != 0)
	{
		ereport(elevel,
				(errcode_for_file_access(),
				 errmsg("could not rename file \"%s\" to \"%s\": %m",
						tmppath, path)));
		return;
	}

	/* Check CreateSlot() for the reasoning of using a crit. section. */
	START_CRIT_SECTION();

	fsync_fname(path, false);
	fsync_fname((char *) dir, true);
	fsync_fname("pg_replslot", true);

	END_CRIT_SECTION();

	/*
	 * Successfully wrote, unset dirty bit, unless somebody dirtied again
	 * already.
	 */
	{
		volatile ReplicationSlot *vslot = slot;

		SpinLockAcquire(&vslot->mutex);
		if (!vslot->just_dirtied)
			vslot->dirty = false;
		SpinLockRelease(&vslot->mutex);
	}

	LWLockRelease(slot->io_in_progress_lock);
}

/*
 * Physical write of a page from a buffer slot
 *
 * On failure, we cannot just ereport(ERROR) since caller has put state in
 * shared memory that must be undone.  So, we return FALSE and save enough
 * info in static variables to let SlruReportIOError make the report.
 *
 * For now, assume it's not worth keeping a file pointer open across
 * independent read/write operations.  We do batch operations during
 * SimpleLruFlush, though.
 *
 * fdata is NULL for a standalone write, pointer to open-file info during
 * SimpleLruFlush.
 */
static bool
SlruPhysicalWritePage(SlruCtl ctl, int pageno, int slotno, SlruFlush fdata)
{
	SlruShared	shared = ctl->shared;
	int			segno = pageno / SLRU_PAGES_PER_SEGMENT;
	int			rpageno = pageno % SLRU_PAGES_PER_SEGMENT;
	int			offset = rpageno * BLCKSZ;
	char		path[MAXPGPATH];
	int			fd = -1;

	/*
	 * Honor the write-WAL-before-data rule, if appropriate, so that we do not
	 * write out data before associated WAL records.  This is the same action
	 * performed during FlushBuffer() in the main buffer manager.
	 */
	if (shared->group_lsn != NULL)
	{
		/*
		 * We must determine the largest async-commit LSN for the page. This
		 * is a bit tedious, but since this entire function is a slow path
		 * anyway, it seems better to do this here than to maintain a per-page
		 * LSN variable (which'd need an extra comparison in the
		 * transaction-commit path).
		 */
		XLogRecPtr	max_lsn;
		int			lsnindex,
					lsnoff;

		lsnindex = slotno * shared->lsn_groups_per_page;
		max_lsn = shared->group_lsn[lsnindex++];
		for (lsnoff = 1; lsnoff < shared->lsn_groups_per_page; lsnoff++)
		{
			XLogRecPtr	this_lsn = shared->group_lsn[lsnindex++];

			if (max_lsn < this_lsn)
				max_lsn = this_lsn;
		}

		if (!XLogRecPtrIsInvalid(max_lsn))
		{
			/*
			 * As noted above, elog(ERROR) is not acceptable here, so if
			 * XLogFlush were to fail, we must PANIC.  This isn't much of a
			 * restriction because XLogFlush is just about all critical
			 * section anyway, but let's make sure.
			 */
			START_CRIT_SECTION();
			XLogFlush(max_lsn);
			END_CRIT_SECTION();
		}
	}

	/*
	 * During a Flush, we may already have the desired file open.
	 */
	if (fdata)
	{
		int			i;

		for (i = 0; i < fdata->num_files; i++)
		{
			if (fdata->segno[i] == segno)
			{
				fd = fdata->fd[i];
				break;
			}
		}
	}

	if (fd < 0)
	{
		/*
		 * If the file doesn't already exist, we should create it.  It is
		 * possible for this to need to happen when writing a page that's not
		 * first in its segment; we assume the OS can cope with that. (Note:
		 * it might seem that it'd be okay to create files only when
		 * SimpleLruZeroPage is called for the first page of a segment.
		 * However, if after a crash and restart the REDO logic elects to
		 * replay the log from a checkpoint before the latest one, then it's
		 * possible that we will get commands to set transaction status of
		 * transactions that have already been truncated from the commit log.
		 * Easiest way to deal with that is to accept references to
		 * nonexistent files here and in SlruPhysicalReadPage.)
		 *
		 * Note: it is possible for more than one backend to be executing this
		 * code simultaneously for different pages of the same file. Hence,
//.........这里部分代码省略.........

/*
 * Build a pending-list page from the given array of tuples, and write it out.
 *
 * Returns amount of free space left on the page.
 */
static int32
writeListPage(Relation index, Buffer buffer,
			  IndexTuple *tuples, int32 ntuples, BlockNumber rightlink)
{
	Page		page = BufferGetPage(buffer);
	int32		i,
				freesize,
				size = 0;
	OffsetNumber l,
				off;
	char	   *workspace;
	char	   *ptr;

	/* workspace could be a local array; we use palloc for alignment */
	workspace = palloc(BLCKSZ);

	START_CRIT_SECTION();

	GinInitBuffer(buffer, GIN_LIST);

	off = FirstOffsetNumber;
	ptr = workspace;

	for (i = 0; i < ntuples; i++)
	{
		int			this_size = IndexTupleSize(tuples[i]);

		memcpy(ptr, tuples[i], this_size);
		ptr += this_size;
		size += this_size;

		l = PageAddItem(page, (Item) tuples[i], this_size, off, false, false);

		if (l == InvalidOffsetNumber)
			elog(ERROR, "failed to add item to index page in \"%s\"",
				 RelationGetRelationName(index));

		off++;
	}

	Assert(size <= BLCKSZ);		/* else we overran workspace */

	GinPageGetOpaque(page)->rightlink = rightlink;

	/*
	 * tail page may contain only whole row(s) or final part of row placed on
	 * previous pages (a "row" here meaning all the index tuples generated for
	 * one heap tuple)
	 */
	if (rightlink == InvalidBlockNumber)
	{
		GinPageSetFullRow(page);
		GinPageGetOpaque(page)->maxoff = 1;
	}
	else
	{
		GinPageGetOpaque(page)->maxoff = 0;
	}

	MarkBufferDirty(buffer);

	if (RelationNeedsWAL(index))
	{
		ginxlogInsertListPage data;
		XLogRecPtr	recptr;

		data.rightlink = rightlink;
		data.ntuples = ntuples;

		XLogBeginInsert();
		XLogRegisterData((char *) &data, sizeof(ginxlogInsertListPage));

		XLogRegisterBuffer(0, buffer, REGBUF_WILL_INIT);
		XLogRegisterBufData(0, workspace, size);

		recptr = XLogInsert(RM_GIN_ID, XLOG_GIN_INSERT_LISTPAGE);
		PageSetLSN(page, recptr);
	}

	/* get free space before releasing buffer */
	freesize = PageGetExactFreeSpace(page);

	UnlockReleaseBuffer(buffer);

	END_CRIT_SECTION();

	pfree(workspace);

	return freesize;
}

/*
 * Load a single slot from disk into memory.
 */
static void
RestoreSlotFromDisk(const char *name)
{
	ReplicationSlotOnDisk cp;
	int			i;
	char		path[MAXPGPATH];
	int			fd;
	bool		restored = false;
	int			readBytes;
	pg_crc32c	checksum;

	/* no need to lock here, no concurrent access allowed yet */

	/* delete temp file if it exists */
	sprintf(path, "pg_replslot/%s/state.tmp", name);
	if (unlink(path) < 0 && errno != ENOENT)
		ereport(PANIC,
				(errcode_for_file_access(),
				 errmsg("could not remove file \"%s\": %m", path)));

	sprintf(path, "pg_replslot/%s/state", name);

	elog(DEBUG1, "restoring replication slot from \"%s\"", path);

	fd = OpenTransientFile(path, O_RDWR | PG_BINARY, 0);

	/*
	 * We do not need to handle this as we are rename()ing the directory into
	 * place only after we fsync()ed the state file.
	 */
	if (fd < 0)
		ereport(PANIC,
				(errcode_for_file_access(),
				 errmsg("could not open file \"%s\": %m", path)));

	/*
	 * Sync state file before we're reading from it. We might have crashed
	 * while it wasn't synced yet and we shouldn't continue on that basis.
	 */
	if (pg_fsync(fd) != 0)
	{
		CloseTransientFile(fd);
		ereport(PANIC,
				(errcode_for_file_access(),
				 errmsg("could not fsync file \"%s\": %m",
						path)));
	}

	/* Also sync the parent directory */
	START_CRIT_SECTION();
	fsync_fname(path, true);
	END_CRIT_SECTION();

	/* read part of statefile that's guaranteed to be version independent */
	readBytes = read(fd, &cp, ReplicationSlotOnDiskConstantSize);
	if (readBytes != ReplicationSlotOnDiskConstantSize)
	{
		int			saved_errno = errno;

		CloseTransientFile(fd);
		errno = saved_errno;
		ereport(PANIC,
				(errcode_for_file_access(),
				 errmsg("could not read file \"%s\", read %d of %u: %m",
						path, readBytes,
						(uint32) ReplicationSlotOnDiskConstantSize)));
	}

	/* verify magic */
	if (cp.magic != SLOT_MAGIC)
		ereport(PANIC,
				(errcode_for_file_access(),
				 errmsg("replication slot file \"%s\" has wrong magic %u instead of %u",
						path, cp.magic, SLOT_MAGIC)));

	/* verify version */
	if (cp.version != SLOT_VERSION)
		ereport(PANIC,
				(errcode_for_file_access(),
			errmsg("replication slot file \"%s\" has unsupported version %u",
				   path, cp.version)));

	/* boundary check on length */
	if (cp.length != ReplicationSlotOnDiskV2Size)
		ereport(PANIC,
				(errcode_for_file_access(),
			   errmsg("replication slot file \"%s\" has corrupted length %u",
					  path, cp.length)));

	/* Now that we know the size, read the entire file */
	readBytes = read(fd,
					 (char *) &cp + ReplicationSlotOnDiskConstantSize,
					 cp.length);
	if (readBytes != cp.length)
	{
		int			saved_errno = errno;

//.........这里部分代码省略.........

IndexBuildResult *
ginbuild(Relation heap, Relation index, IndexInfo *indexInfo)
{
	IndexBuildResult *result;
	double		reltuples;
	GinBuildState buildstate;
	Buffer		RootBuffer,
				MetaBuffer;
	ItemPointerData *list;
	Datum		key;
	GinNullCategory category;
	uint32		nlist;
	MemoryContext oldCtx;
	OffsetNumber attnum;

	if (RelationGetNumberOfBlocks(index) != 0)
		elog(ERROR, "index \"%s\" already contains data",
			 RelationGetRelationName(index));

	initGinState(&buildstate.ginstate, index);
	buildstate.indtuples = 0;
	memset(&buildstate.buildStats, 0, sizeof(GinStatsData));

	/* initialize the meta page */
	MetaBuffer = GinNewBuffer(index);

	/* initialize the root page */
	RootBuffer = GinNewBuffer(index);

	START_CRIT_SECTION();
	GinInitMetabuffer(MetaBuffer);
	MarkBufferDirty(MetaBuffer);
	GinInitBuffer(RootBuffer, GIN_LEAF);
	MarkBufferDirty(RootBuffer);

	if (RelationNeedsWAL(index))
	{
		XLogRecPtr	recptr;
		Page		page;

		XLogBeginInsert();
		XLogRegisterBuffer(0, MetaBuffer, REGBUF_WILL_INIT | REGBUF_STANDARD);
		XLogRegisterBuffer(1, RootBuffer, REGBUF_WILL_INIT);

		recptr = XLogInsert(RM_GIN_ID, XLOG_GIN_CREATE_INDEX);

		page = BufferGetPage(RootBuffer);
		PageSetLSN(page, recptr);

		page = BufferGetPage(MetaBuffer);
		PageSetLSN(page, recptr);
	}

	UnlockReleaseBuffer(MetaBuffer);
	UnlockReleaseBuffer(RootBuffer);
	END_CRIT_SECTION();

	/* count the root as first entry page */
	buildstate.buildStats.nEntryPages++;

	/*
	 * create a temporary memory context that is used to hold data not yet
	 * dumped out to the index
	 */
	buildstate.tmpCtx = AllocSetContextCreate(CurrentMemoryContext,
											  "Gin build temporary context",
											  ALLOCSET_DEFAULT_SIZES);

	/*
	 * create a temporary memory context that is used for calling
	 * ginExtractEntries(), and can be reset after each tuple
	 */
	buildstate.funcCtx = AllocSetContextCreate(CurrentMemoryContext,
											   "Gin build temporary context for user-defined function",
											   ALLOCSET_DEFAULT_SIZES);

	buildstate.accum.ginstate = &buildstate.ginstate;
	ginInitBA(&buildstate.accum);

	/*
	 * Do the heap scan.  We disallow sync scan here because dataPlaceToPage
	 * prefers to receive tuples in TID order.
	 */
	reltuples = IndexBuildHeapScan(heap, index, indexInfo, false,
								   ginBuildCallback, (void *) &buildstate);

	/* dump remaining entries to the index */
	oldCtx = MemoryContextSwitchTo(buildstate.tmpCtx);
	ginBeginBAScan(&buildstate.accum);
	while ((list = ginGetBAEntry(&buildstate.accum,
								 &attnum, &key, &category, &nlist)) != NULL)
	{
		/* there could be many entries, so be willing to abort here */
		CHECK_FOR_INTERRUPTS();
		ginEntryInsert(&buildstate.ginstate, attnum, key, category,
					   list, nlist, &buildstate.buildStats);
	}
	MemoryContextSwitchTo(oldCtx);

	MemoryContextDelete(buildstate.funcCtx);
//.........这里部分代码省略.........

//.........这里部分代码省略.........
		char	   *ptr;
		char	   *collectordata;

		buffer = ReadBuffer(index, metadata->tail);
		LockBuffer(buffer, GIN_EXCLUSIVE);
		page = BufferGetPage(buffer);

		off = (PageIsEmpty(page)) ? FirstOffsetNumber :
			OffsetNumberNext(PageGetMaxOffsetNumber(page));

		collectordata = ptr = (char *) palloc(collector->sumsize);

		data.ntuples = collector->ntuples;

		if (needWal)
			XLogBeginInsert();

		START_CRIT_SECTION();

		/*
		 * Increase counter of heap tuples
		 */
		Assert(GinPageGetOpaque(page)->maxoff <= metadata->nPendingHeapTuples);
		GinPageGetOpaque(page)->maxoff++;
		metadata->nPendingHeapTuples++;

		for (i = 0; i < collector->ntuples; i++)
		{
			tupsize = IndexTupleSize(collector->tuples[i]);
			l = PageAddItem(page, (Item) collector->tuples[i], tupsize, off, false, false);

			if (l == InvalidOffsetNumber)
				elog(ERROR, "failed to add item to index page in \"%s\"",
					 RelationGetRelationName(index));

			memcpy(ptr, collector->tuples[i], tupsize);
			ptr += tupsize;

			off++;
		}

		Assert((ptr - collectordata) <= collector->sumsize);
		if (needWal)
		{
			XLogRegisterBuffer(1, buffer, REGBUF_STANDARD);
			XLogRegisterBufData(1, collectordata, collector->sumsize);
		}

		metadata->tailFreeSize = PageGetExactFreeSpace(page);

		MarkBufferDirty(buffer);
	}

	/*
	 * Write metabuffer, make xlog entry
	 */
	MarkBufferDirty(metabuffer);

	if (needWal)
	{
		XLogRecPtr	recptr;

		memcpy(&data.metadata, metadata, sizeof(GinMetaPageData));

		XLogRegisterBuffer(0, metabuffer, REGBUF_WILL_INIT);
		XLogRegisterData((char *) &data, sizeof(ginxlogUpdateMeta));

		recptr = XLogInsert(RM_GIN_ID, XLOG_GIN_UPDATE_META_PAGE);
		PageSetLSN(metapage, recptr);

		if (buffer != InvalidBuffer)
		{
			PageSetLSN(page, recptr);
		}
	}

	if (buffer != InvalidBuffer)
		UnlockReleaseBuffer(buffer);

	/*
	 * Force pending list cleanup when it becomes too long. And,
	 * ginInsertCleanup could take significant amount of time, so we prefer to
	 * call it when it can do all the work in a single collection cycle. In
	 * non-vacuum mode, it shouldn't require maintenance_work_mem, so fire it
	 * while pending list is still small enough to fit into
	 * gin_pending_list_limit.
	 *
	 * ginInsertCleanup() should not be called inside our CRIT_SECTION.
	 */
	cleanupSize = GinGetPendingListCleanupSize(index);
	if (metadata->nPendingPages * GIN_PAGE_FREESIZE > cleanupSize * 1024L)
		needCleanup = true;

	UnlockReleaseBuffer(metabuffer);

	END_CRIT_SECTION();

	if (needCleanup)
		ginInsertCleanup(ginstate, true, NULL);
}

//.........这里部分代码省略.........

		nextrdata->data = NULL;
		nextrdata->len = 0;
		nextrdata->next = nextrdata + 1;
		nextrdata->buffer = pbuf;
		nextrdata->buffer_std = true;
		nextrdata++;

		nextrdata->data = NULL;
		nextrdata->len = 0;
		nextrdata->buffer = rbuf;
		nextrdata->buffer_std = true;
		nextrdata->next = NULL;

		if (BufferIsValid(lbuf))
		{
			nextrdata->next = nextrdata + 1;
			nextrdata++;
			nextrdata->data = NULL;
			nextrdata->len = 0;
			nextrdata->buffer = lbuf;
			nextrdata->buffer_std = true;
			nextrdata->next = NULL;
		}

		recptr = XLogInsert(RM_BTREE_ID, xlinfo, rdata);

		if (BufferIsValid(metabuf))
		{
			PageSetLSN(metapg, recptr);
			PageSetTLI(metapg, ThisTimeLineID);
		}
		page = BufferGetPage(pbuf);
		PageSetLSN(page, recptr);
		PageSetTLI(page, ThisTimeLineID);
		page = BufferGetPage(rbuf);
		PageSetLSN(page, recptr);
		PageSetTLI(page, ThisTimeLineID);
		page = BufferGetPage(buf);
		PageSetLSN(page, recptr);
		PageSetTLI(page, ThisTimeLineID);
		if (BufferIsValid(lbuf))
		{
			page = BufferGetPage(lbuf);
			PageSetLSN(page, recptr);
			PageSetTLI(page, ThisTimeLineID);
		}
	}

	END_CRIT_SECTION();

	/* release metapage; send out relcache inval if metapage changed */
	if (BufferIsValid(metabuf))
	{
		CacheInvalidateRelcache(rel);
		_bt_relbuf(rel, metabuf);
	}
	/* can always release leftsib immediately */
	if (BufferIsValid(lbuf))
		_bt_relbuf(rel, lbuf);

	/*
	 * If parent became half dead, recurse to delete it. Otherwise, if right
	 * sibling is empty and is now the last child of the parent, recurse to
	 * try to delete it.  (These cases cannot apply at the same time, though
	 * the second case might itself recurse to the first.)
	 *
	 * When recursing to parent, we hold the lock on the target page until
	 * done.  This delays any insertions into the keyspace that was just
	 * effectively reassigned to the parent's right sibling.  If we allowed
	 * that, and there were enough such insertions before we finish deleting
	 * the parent, page splits within that keyspace could lead to inserting
	 * out-of-order keys into the grandparent level.  It is thought that that
	 * wouldn't have any serious consequences, but it still seems like a
	 * pretty bad idea.
	 */
	if (parent_half_dead)
	{
		/* recursive call will release pbuf */
		_bt_relbuf(rel, rbuf);
		result = _bt_pagedel(rel, pbuf, stack->bts_parent) + 1;
		_bt_relbuf(rel, buf);
	}
	else if (parent_one_child && rightsib_empty)
	{
		_bt_relbuf(rel, pbuf);
		_bt_relbuf(rel, buf);
		/* recursive call will release rbuf */
		result = _bt_pagedel(rel, rbuf, stack) + 1;
	}
	else
	{
		_bt_relbuf(rel, pbuf);
		_bt_relbuf(rel, buf);
		_bt_relbuf(rel, rbuf);
		result = 1;
	}

	return result;
}

/*
 * Creates new posting tree containing the given TIDs. Returns the page
 * number of the root of the new posting tree.
 *
 * items[] must be in sorted order with no duplicates.
 */
BlockNumber
createPostingTree(Relation index, ItemPointerData *items, uint32 nitems,
				  GinStatsData *buildStats)
{
	BlockNumber blkno;
	Buffer		buffer;
	Page		page;
	int			nrootitems;

	/* Calculate how many TIDs will fit on first page. */
	nrootitems = Min(nitems, GinMaxLeafDataItems);

	/*
	 * Create the root page.
	 */
	buffer = GinNewBuffer(index);
	page = BufferGetPage(buffer);
	blkno = BufferGetBlockNumber(buffer);

	START_CRIT_SECTION();

	GinInitBuffer(buffer, GIN_DATA | GIN_LEAF);
	memcpy(GinDataPageGetData(page), items, sizeof(ItemPointerData) * nrootitems);
	GinPageGetOpaque(page)->maxoff = nrootitems;

	MarkBufferDirty(buffer);

	if (RelationNeedsWAL(index))
	{
		XLogRecPtr	recptr;
		XLogRecData rdata[2];
		ginxlogCreatePostingTree data;

		data.node = index->rd_node;
		data.blkno = blkno;
		data.nitem = nrootitems;

		rdata[0].buffer = InvalidBuffer;
		rdata[0].data = (char *) &data;
		rdata[0].len = sizeof(ginxlogCreatePostingTree);
		rdata[0].next = &rdata[1];

		rdata[1].buffer = InvalidBuffer;
		rdata[1].data = (char *) items;
		rdata[1].len = sizeof(ItemPointerData) * nrootitems;
		rdata[1].next = NULL;

		recptr = XLogInsert(RM_GIN_ID, XLOG_GIN_CREATE_PTREE, rdata);
		PageSetLSN(page, recptr);
	}

	UnlockReleaseBuffer(buffer);

	END_CRIT_SECTION();

	/* During index build, count the newly-added data page */
	if (buildStats)
		buildStats->nDataPages++;

	/*
	 * Add any remaining TIDs to the newly-created posting tree.
	 */
	if (nitems > nrootitems)
	{
		ginInsertItemPointers(index, blkno,
							  items + nrootitems,
							  nitems - nrootitems,
							  buildStats);
	}

	return blkno;
}

//.........这里部分代码省略.........

		metad->btm_root = rootblkno;
		metad->btm_level = 0;
		metad->btm_fastroot = rootblkno;
		metad->btm_fastlevel = 0;

		MarkBufferDirty(rootbuf);
		MarkBufferDirty(metabuf);

		/* XLOG stuff */
		if (!rel->rd_istemp)
		{
			xl_btree_newroot xlrec;
			XLogRecPtr	recptr;
			XLogRecData rdata;

			xlrec.node = rel->rd_node;
			xlrec.rootblk = rootblkno;
			xlrec.level = 0;

			rdata.data = (char *) &xlrec;
			rdata.len = SizeOfBtreeNewroot;
			rdata.buffer = InvalidBuffer;
			rdata.next = NULL;

			recptr = XLogInsert(RM_BTREE_ID, XLOG_BTREE_NEWROOT, &rdata);

			PageSetLSN(rootpage, recptr);
			PageSetTLI(rootpage, ThisTimeLineID);
			PageSetLSN(metapg, recptr);
			PageSetTLI(metapg, ThisTimeLineID);
		}

		END_CRIT_SECTION();

		/*
		 * Send out relcache inval for metapage change (probably unnecessary
		 * here, but let's be safe).
		 */
		CacheInvalidateRelcache(rel);

		/*
		 * swap root write lock for read lock.	There is no danger of anyone
		 * else accessing the new root page while it's unlocked, since no one
		 * else knows where it is yet.
		 */
		LockBuffer(rootbuf, BUFFER_LOCK_UNLOCK);
		LockBuffer(rootbuf, BT_READ);

		/* okay, metadata is correct, release lock on it */
		_bt_relbuf(rel, metabuf);
	}
	else
	{
		rootblkno = metad->btm_fastroot;
		Assert(rootblkno != P_NONE);
		rootlevel = metad->btm_fastlevel;

		/*
		 * Cache the metapage data for next time
		 */
		rel->rd_amcache = MemoryContextAlloc(rel->rd_indexcxt,
											 sizeof(BTMetaPageData));
		memcpy(rel->rd_amcache, metad, sizeof(BTMetaPageData));

		/*

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);
	}

	END_CRIT_SECTION();
}

/*
 * Delete item(s) from a btree page.
 *
 * This must only be used for deleting leaf items.	Deleting an item on a
 * non-leaf page has to be done as part of an atomic action that includes
 * deleting the page it points to.
 *
 * This routine assumes that the caller has pinned and locked the buffer.
 * Also, the given itemnos *must* appear in increasing order in the array.
 *
 * We record VACUUMs and b-tree deletes differently in WAL. InHotStandby
 * we need to be able to pin all of the blocks in the btree in physical
 * order when replaying the effects of a VACUUM, just as we do for the
 * original VACUUM itself. lastBlockVacuumed allows us to tell whether an
 * intermediate range of blocks has had no changes at all by VACUUM,
 * and so must be scanned anyway during replay. We always write a WAL record
 * for the last block in the index, whether or not it contained any items
 * to be removed. This allows us to scan right up to end of index to
 * ensure correct locking.
 */
void
_bt_delitems_vacuum(Relation rel, Buffer buf,
			OffsetNumber *itemnos, int nitems, BlockNumber lastBlockVacuumed)
{
	Page		page = BufferGetPage(buf);
	BTPageOpaque opaque;

	/* No ereport(ERROR) until changes are logged */
	START_CRIT_SECTION();

	/* Fix the page */
	if (nitems > 0)
		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[2];

		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;
		rdata[0].buffer = InvalidBuffer;
		rdata[0].next = &(rdata[1]);

		/*
		 * The target-offsets array is not in the buffer, but pretend that it
		 * is.	When XLogInsert stores the whole buffer, the offsets array
		 * need not be stored too.
		 */
		if (nitems > 0)
		{
			rdata[1].data = (char *) itemnos;
			rdata[1].len = nitems * sizeof(OffsetNumber);
		}
		else
		{
			rdata[1].data = NULL;
			rdata[1].len = 0;
		}
		rdata[1].buffer = buf;
		rdata[1].buffer_std = true;
		rdata[1].next = NULL;

		recptr = XLogInsert(RM_BTREE_ID, XLOG_BTREE_VACUUM, rdata);

		PageSetLSN(page, recptr);
		PageSetTLI(page, ThisTimeLineID);
	}

	END_CRIT_SECTION();
}

/*
 * Delete a posting tree page.
 */
static void
ginDeletePage(GinVacuumState *gvs, BlockNumber deleteBlkno, BlockNumber leftBlkno,
			  BlockNumber parentBlkno, OffsetNumber myoff, bool isParentRoot)
{
	Buffer		dBuffer;
	Buffer		lBuffer;
	Buffer		pBuffer;
	Page		page,
				parentPage;
	BlockNumber rightlink;

	/*
	 * Lock the pages in the same order as an insertion would, to avoid
	 * deadlocks: left, then right, then parent.
	 */
	lBuffer = ReadBufferExtended(gvs->index, MAIN_FORKNUM, leftBlkno,
								 RBM_NORMAL, gvs->strategy);
	dBuffer = ReadBufferExtended(gvs->index, MAIN_FORKNUM, deleteBlkno,
								 RBM_NORMAL, gvs->strategy);
	pBuffer = ReadBufferExtended(gvs->index, MAIN_FORKNUM, parentBlkno,
								 RBM_NORMAL, gvs->strategy);

	LockBuffer(lBuffer, GIN_EXCLUSIVE);
	LockBuffer(dBuffer, GIN_EXCLUSIVE);
	if (!isParentRoot)			/* parent is already locked by
								 * LockBufferForCleanup() */
		LockBuffer(pBuffer, GIN_EXCLUSIVE);

	START_CRIT_SECTION();

	/* Unlink the page by changing left sibling's rightlink */
	page = BufferGetPage(dBuffer);
	rightlink = GinPageGetOpaque(page)->rightlink;

	page = BufferGetPage(lBuffer);
	GinPageGetOpaque(page)->rightlink = rightlink;

	/* Delete downlink from parent */
	parentPage = BufferGetPage(pBuffer);
#ifdef USE_ASSERT_CHECKING
	do
	{
		PostingItem *tod = GinDataPageGetPostingItem(parentPage, myoff);

		Assert(PostingItemGetBlockNumber(tod) == deleteBlkno);
	} while (0);
#endif
	GinPageDeletePostingItem(parentPage, myoff);

	page = BufferGetPage(dBuffer);

	/*
	 * we shouldn't change rightlink field to save workability of running
	 * search scan
	 */
	GinPageGetOpaque(page)->flags = GIN_DELETED;

	MarkBufferDirty(pBuffer);
	MarkBufferDirty(lBuffer);
	MarkBufferDirty(dBuffer);

	if (RelationNeedsWAL(gvs->index))
	{
		XLogRecPtr	recptr;
		ginxlogDeletePage data;

		/*
		 * We can't pass REGBUF_STANDARD for the deleted page, because we
		 * didn't set pd_lower on pre-9.4 versions. The page might've been
		 * binary-upgraded from an older version, and hence not have pd_lower
		 * set correctly. Ditto for the left page, but removing the item from
		 * the parent updated its pd_lower, so we know that's OK at this
		 * point.
		 */
		XLogBeginInsert();
		XLogRegisterBuffer(0, dBuffer, 0);
		XLogRegisterBuffer(1, pBuffer, REGBUF_STANDARD);
		XLogRegisterBuffer(2, lBuffer, 0);

		data.parentOffset = myoff;
		data.rightLink = GinPageGetOpaque(page)->rightlink;

		XLogRegisterData((char *) &data, sizeof(ginxlogDeletePage));

		recptr = XLogInsert(RM_GIN_ID, XLOG_GIN_DELETE_PAGE);
		PageSetLSN(page, recptr);
		PageSetLSN(parentPage, recptr);
		PageSetLSN(BufferGetPage(lBuffer), recptr);
	}

	if (!isParentRoot)
		LockBuffer(pBuffer, GIN_UNLOCK);
	ReleaseBuffer(pBuffer);
	UnlockReleaseBuffer(lBuffer);
	UnlockReleaseBuffer(dBuffer);

	END_CRIT_SECTION();
//.........这里部分代码省略.........

IndexBulkDeleteResult *
ginbulkdelete(IndexVacuumInfo *info, IndexBulkDeleteResult *stats,
			  IndexBulkDeleteCallback callback, void *callback_state)
{
	Relation	index = info->index;
	BlockNumber blkno = GIN_ROOT_BLKNO;
	GinVacuumState gvs;
	Buffer		buffer;
	BlockNumber rootOfPostingTree[BLCKSZ / (sizeof(IndexTupleData) + sizeof(ItemId))];
	uint32		nRoot;

	gvs.tmpCxt = AllocSetContextCreate(CurrentMemoryContext,
									   "Gin vacuum temporary context",
									   ALLOCSET_DEFAULT_MINSIZE,
									   ALLOCSET_DEFAULT_INITSIZE,
									   ALLOCSET_DEFAULT_MAXSIZE);
	gvs.index = index;
	gvs.callback = callback;
	gvs.callback_state = callback_state;
	gvs.strategy = info->strategy;
	initGinState(&gvs.ginstate, index);

	/* first time through? */
	if (stats == NULL)
	{
		/* Yes, so initialize stats to zeroes */
		stats = (IndexBulkDeleteResult *) palloc0(sizeof(IndexBulkDeleteResult));
		/*
		 * and cleanup any pending inserts  */
		ginInsertCleanup(&gvs.ginstate, !IsAutoVacuumWorkerProcess(),
						 false, stats);
	}

	/* we'll re-count the tuples each time */
	stats->num_index_tuples = 0;
	gvs.result = stats;

	buffer = ReadBufferExtended(index, MAIN_FORKNUM, blkno,
								RBM_NORMAL, info->strategy);

	/* find leaf page */
	for (;;)
	{
		Page		page = BufferGetPage(buffer);
		IndexTuple	itup;

		LockBuffer(buffer, GIN_SHARE);

		Assert(!GinPageIsData(page));

		if (GinPageIsLeaf(page))
		{
			LockBuffer(buffer, GIN_UNLOCK);
			LockBuffer(buffer, GIN_EXCLUSIVE);

			if (blkno == GIN_ROOT_BLKNO && !GinPageIsLeaf(page))
			{
				LockBuffer(buffer, GIN_UNLOCK);
				continue;		/* check it one more */
			}
			break;
		}

		Assert(PageGetMaxOffsetNumber(page) >= FirstOffsetNumber);

		itup = (IndexTuple) PageGetItem(page, PageGetItemId(page, FirstOffsetNumber));
		blkno = GinGetDownlink(itup);
		Assert(blkno != InvalidBlockNumber);

		UnlockReleaseBuffer(buffer);
		buffer = ReadBufferExtended(index, MAIN_FORKNUM, blkno,
									RBM_NORMAL, info->strategy);
	}

	/* right now we found leftmost page in entry's BTree */

	for (;;)
	{
		Page		page = BufferGetPage(buffer);
		Page		resPage;
		uint32		i;

		Assert(!GinPageIsData(page));

		resPage = ginVacuumEntryPage(&gvs, buffer, rootOfPostingTree, &nRoot);

		blkno = GinPageGetOpaque(page)->rightlink;

		if (resPage)
		{
			START_CRIT_SECTION();
			PageRestoreTempPage(resPage, page);
			MarkBufferDirty(buffer);
			xlogVacuumPage(gvs.index, buffer);
			UnlockReleaseBuffer(buffer);
			END_CRIT_SECTION();
		}
		else
		{
			UnlockReleaseBuffer(buffer);
//.........这里部分代码省略.........

//.........这里部分代码省略.........
	Page		metapage;
	GinMetaPageData *metadata;
	BlockNumber blknoToDelete;

	metapage = BufferGetPage(metabuffer);
	metadata = GinPageGetMeta(metapage);
	blknoToDelete = metadata->head;

	do
	{
		Page		page;
		int			i;
		int64		nDeletedHeapTuples = 0;
		ginxlogDeleteListPages data;
		XLogRecData rdata[1];
		Buffer		buffers[GIN_NDELETE_AT_ONCE];

		data.node = index->rd_node;

		rdata[0].buffer = InvalidBuffer;
		rdata[0].data = (char *) &data;
		rdata[0].len = sizeof(ginxlogDeleteListPages);
		rdata[0].next = NULL;

		data.ndeleted = 0;
		while (data.ndeleted < GIN_NDELETE_AT_ONCE && blknoToDelete != newHead)
		{
			data.toDelete[data.ndeleted] = blknoToDelete;
			buffers[data.ndeleted] = ReadBuffer(index, blknoToDelete);
			LockBuffer(buffers[data.ndeleted], GIN_EXCLUSIVE);
			page = BufferGetPage(buffers[data.ndeleted]);

			data.ndeleted++;

			if (GinPageIsDeleted(page))
			{
				/* concurrent cleanup process is detected */
				for (i = 0; i < data.ndeleted; i++)
					UnlockReleaseBuffer(buffers[i]);

				return true;
			}

			nDeletedHeapTuples += GinPageGetOpaque(page)->maxoff;
			blknoToDelete = GinPageGetOpaque(page)->rightlink;
		}

		if (stats)
			stats->pages_deleted += data.ndeleted;

		START_CRIT_SECTION();

		metadata->head = blknoToDelete;

		Assert(metadata->nPendingPages >= data.ndeleted);
		metadata->nPendingPages -= data.ndeleted;
		Assert(metadata->nPendingHeapTuples >= nDeletedHeapTuples);
		metadata->nPendingHeapTuples -= nDeletedHeapTuples;

		if (blknoToDelete == InvalidBlockNumber)
		{
			metadata->tail = InvalidBlockNumber;
			metadata->tailFreeSize = 0;
			metadata->nPendingPages = 0;
			metadata->nPendingHeapTuples = 0;
		}

		MarkBufferDirty(metabuffer);

		for (i = 0; i < data.ndeleted; i++)
		{
			page = BufferGetPage(buffers[i]);
			GinPageGetOpaque(page)->flags = GIN_DELETED;
			MarkBufferDirty(buffers[i]);
		}

		if (RelationNeedsWAL(index))
		{
			XLogRecPtr	recptr;

			memcpy(&data.metadata, metadata, sizeof(GinMetaPageData));

			recptr = XLogInsert(RM_GIN_ID, XLOG_GIN_DELETE_LISTPAGE, rdata);
			PageSetLSN(metapage, recptr);

			for (i = 0; i < data.ndeleted; i++)
			{
				page = BufferGetPage(buffers[i]);
				PageSetLSN(page, recptr);
			}
		}

		for (i = 0; i < data.ndeleted; i++)
			UnlockReleaseBuffer(buffers[i]);

		END_CRIT_SECTION();
	} while (blknoToDelete != newHead);

	return false;
}

/*
 *	visibilitymap_set - set a bit on a previously pinned page
 *
 * recptr is the LSN of the XLOG record we're replaying, if we're in recovery,
 * or InvalidXLogRecPtr in normal running.	The page LSN is advanced to the
 * one provided; in normal running, we generate a new XLOG record and set the
 * page LSN to that value.	cutoff_xid is the largest xmin on the page being
 * marked all-visible; it is needed for Hot Standby, and can be
 * InvalidTransactionId if the page contains no tuples.
 *
 * Caller is expected to set the heap page's PD_ALL_VISIBLE bit before calling
 * this function. Except in recovery, caller should also pass the heap
 * buffer. When checksums are enabled and we're not in recovery, we must add
 * the heap buffer to the WAL chain to protect it from being torn.
 *
 * You must pass a buffer containing the correct map page to this function.
 * Call visibilitymap_pin first to pin the right one. This function doesn't do
 * any I/O.
 */
void
visibilitymap_set(Relation rel, BlockNumber heapBlk, Buffer heapBuf,
				  XLogRecPtr recptr, Buffer vmBuf, TransactionId cutoff_xid)
{
	BlockNumber mapBlock = HEAPBLK_TO_MAPBLOCK(heapBlk);
	uint32		mapByte = HEAPBLK_TO_MAPBYTE(heapBlk);
	uint8		mapBit = HEAPBLK_TO_MAPBIT(heapBlk);
	Page		page;
	char	   *map;

#ifdef TRACE_VISIBILITYMAP
	elog(DEBUG1, "vm_set %s %d", RelationGetRelationName(rel), heapBlk);
#endif

	Assert(InRecovery || XLogRecPtrIsInvalid(recptr));
	Assert(InRecovery || BufferIsValid(heapBuf));

	/* Check that we have the right heap page pinned, if present */
	if (BufferIsValid(heapBuf) && BufferGetBlockNumber(heapBuf) != heapBlk)
		elog(ERROR, "wrong heap buffer passed to visibilitymap_set");

	/* Check that we have the right VM page pinned */
	if (!BufferIsValid(vmBuf) || BufferGetBlockNumber(vmBuf) != mapBlock)
		elog(ERROR, "wrong VM buffer passed to visibilitymap_set");

	page = BufferGetPage(vmBuf);
	map = PageGetContents(page);
	LockBuffer(vmBuf, BUFFER_LOCK_EXCLUSIVE);

	if (!(map[mapByte] & (1 << mapBit)))
	{
		START_CRIT_SECTION();

		map[mapByte] |= (1 << mapBit);
		MarkBufferDirty(vmBuf);

		if (RelationNeedsWAL(rel))
		{
			if (XLogRecPtrIsInvalid(recptr))
			{
				Assert(!InRecovery);
				recptr = log_heap_visible(rel->rd_node, heapBuf, vmBuf,
										  cutoff_xid);

				/*
				 * If data checksums are enabled, we need to protect the heap
				 * page from being torn.
				 */
				if (DataChecksumsEnabled())
				{
					Page		heapPage = BufferGetPage(heapBuf);

					/* caller is expected to set PD_ALL_VISIBLE first */
					Assert(PageIsAllVisible(heapPage));
					PageSetLSN(heapPage, recptr);
				}
			}
			PageSetLSN(page, recptr);
		}

		END_CRIT_SECTION();
	}

	LockBuffer(vmBuf, BUFFER_LOCK_UNLOCK);
}