C++ PageAddItem函数代码示例

/*
 * Place tuple and split page, original buffer(lbuf) leaves untouched,
 * returns shadow pages filled with new data.
 * Tuples are distributed between pages by equal size on its, not
 * an equal number!
 */
static void
entrySplitPage(GinBtree btree, Buffer origbuf,
			   GinBtreeStack *stack,
			   void *insertPayload,
			   BlockNumber updateblkno, XLogRecData **prdata,
			   Page *newlpage, Page *newrpage)
{
	GinBtreeEntryInsertData *insertData = insertPayload;
	OffsetNumber off = stack->off;
	OffsetNumber i,
				maxoff,
				separator = InvalidOffsetNumber;
	Size		totalsize = 0;
	Size		tupstoresize;
	Size		lsize = 0,
				size;
	char	   *ptr;
	IndexTuple	itup;
	Page		page;
	Page		lpage = PageGetTempPageCopy(BufferGetPage(origbuf));
	Page		rpage = PageGetTempPageCopy(BufferGetPage(origbuf));
	Size		pageSize = PageGetPageSize(lpage);

	/* these must be static so they can be returned to caller */
	static XLogRecData rdata[2];
	static ginxlogSplitEntry data;
	static char tupstore[2 * BLCKSZ];

	*prdata = rdata;
	entryPreparePage(btree, lpage, off, insertData, updateblkno);

	/*
	 * First, append all the existing tuples and the new tuple we're inserting
	 * one after another in a temporary workspace.
	 */
	maxoff = PageGetMaxOffsetNumber(lpage);
	ptr = tupstore;
	for (i = FirstOffsetNumber; i <= maxoff; i++)
	{
		if (i == off)
		{
			size = MAXALIGN(IndexTupleSize(insertData->entry));
			memcpy(ptr, insertData->entry, size);
			ptr += size;
			totalsize += size + sizeof(ItemIdData);
		}

		itup = (IndexTuple) PageGetItem(lpage, PageGetItemId(lpage, i));
		size = MAXALIGN(IndexTupleSize(itup));
		memcpy(ptr, itup, size);
		ptr += size;
		totalsize += size + sizeof(ItemIdData);
	}

	if (off == maxoff + 1)
	{
		size = MAXALIGN(IndexTupleSize(insertData->entry));
		memcpy(ptr, insertData->entry, size);
		ptr += size;
		totalsize += size + sizeof(ItemIdData);
	}
	tupstoresize = ptr - tupstore;

	/*
	 * Initialize the left and right pages, and copy all the tuples back to
	 * them.
	 */
	GinInitPage(rpage, GinPageGetOpaque(lpage)->flags, pageSize);
	GinInitPage(lpage, GinPageGetOpaque(rpage)->flags, pageSize);

	ptr = tupstore;
	maxoff++;
	lsize = 0;

	page = lpage;
	for (i = FirstOffsetNumber; i <= maxoff; i++)
	{
		itup = (IndexTuple) ptr;

		if (lsize > totalsize / 2)
		{
			if (separator == InvalidOffsetNumber)
				separator = i - 1;
			page = rpage;
		}
		else
		{
			lsize += MAXALIGN(IndexTupleSize(itup)) + sizeof(ItemIdData);
		}

		if (PageAddItem(page, (Item) itup, IndexTupleSize(itup), InvalidOffsetNumber, false, false) == InvalidOffsetNumber)
			elog(ERROR, "failed to add item to index page in \"%s\"",
				 RelationGetRelationName(btree->index));
		ptr += MAXALIGN(IndexTupleSize(itup));
//.........这里部分代码省略.........

/*
 * Insert a tuple to the new relation.	This has to track heap_insert
 * and its subsidiary functions!
 *
 * t_self of the tuple is set to the new TID of the tuple. If t_ctid of the
 * tuple is invalid on entry, it's replaced with the new TID as well (in
 * the inserted data only, not in the caller's copy).
 */
static void
raw_heap_insert(RewriteState state, HeapTuple tup)
{
	Page		page = state->rs_buffer;
	Size		pageFreeSpace,
				saveFreeSpace;
	Size		len;
	OffsetNumber newoff;
	HeapTuple	heaptup;

	/*
	 * If the new tuple is too big for storage or contains already toasted
	 * out-of-line attributes from some other relation, invoke the toaster.
	 *
	 * Note: below this point, heaptup is the data we actually intend to store
	 * into the relation; tup is the caller's original untoasted data.
	 */
	if (state->rs_new_rel->rd_rel->relkind == RELKIND_TOASTVALUE)
	{
		/* toast table entries should never be recursively toasted */
		Assert(!HeapTupleHasExternal(tup));
		heaptup = tup;
	}
	else if (HeapTupleHasExternal(tup) || tup->t_len > TOAST_TUPLE_THRESHOLD)
		heaptup = toast_insert_or_update(state->rs_new_rel, tup, NULL,
										 HEAP_INSERT_SKIP_FSM |
										 (state->rs_use_wal ?
										  0 : HEAP_INSERT_SKIP_WAL));
	else
		heaptup = tup;

	len = MAXALIGN(heaptup->t_len);		/* be conservative */

	/*
	 * If we're gonna fail for oversize tuple, do it right away
	 */
	if (len > MaxHeapTupleSize)
		ereport(ERROR,
				(errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED),
				 errmsg("row is too big: size %zu, maximum size %zu",
						len, MaxHeapTupleSize)));

	/* Compute desired extra freespace due to fillfactor option */
	saveFreeSpace = RelationGetTargetPageFreeSpace(state->rs_new_rel,
												   HEAP_DEFAULT_FILLFACTOR);

	/* Now we can check to see if there's enough free space already. */
	if (state->rs_buffer_valid)
	{
		pageFreeSpace = PageGetHeapFreeSpace(page);

		if (len + saveFreeSpace > pageFreeSpace)
		{
			/* Doesn't fit, so write out the existing page */

			/* XLOG stuff */
			if (state->rs_use_wal)
				log_newpage(&state->rs_new_rel->rd_node,
							MAIN_FORKNUM,
							state->rs_blockno,
							page,
							true);

			/*
			 * Now write the page. We say isTemp = true even if it's not a
			 * temp table, because there's no need for smgr to schedule an
			 * fsync for this write; we'll do it ourselves in
			 * end_heap_rewrite.
			 */
			RelationOpenSmgr(state->rs_new_rel);

			PageSetChecksumInplace(page, state->rs_blockno);

			smgrextend(state->rs_new_rel->rd_smgr, MAIN_FORKNUM,
					   state->rs_blockno, (char *) page, true);

			state->rs_blockno++;
			state->rs_buffer_valid = false;
		}
	}

	if (!state->rs_buffer_valid)
	{
		/* Initialize a new empty page */
		PageInit(page, BLCKSZ, 0);
		state->rs_buffer_valid = true;
	}

	/* And now we can insert the tuple into the page */
	newoff = PageAddItem(page, (Item) heaptup->t_data, heaptup->t_len,
						 InvalidOffsetNumber, false, true);
	if (newoff == InvalidOffsetNumber)
//.........这里部分代码省略.........

//.........这里部分代码省略.........
	tmppage = origpage;

	*nroot = 0;

	for (i = FirstOffsetNumber; i <= maxoff; i++)
	{
		IndexTuple	itup = (IndexTuple) PageGetItem(tmppage, PageGetItemId(tmppage, i));

		if (GinIsPostingTree(itup))
		{
			/*
			 * store posting tree's roots for further processing, we can't
			 * vacuum it just now due to risk of deadlocks with scans/inserts
			 */
			roots[*nroot] = GinGetDownlink(itup);
			(*nroot)++;
		}
		else if (GinGetNPosting(itup) > 0)
		{
			int			nitems;
			ItemPointer items_orig;
			bool		free_items_orig;
			ItemPointer items;

			/* Get list of item pointers from the tuple. */
			if (GinItupIsCompressed(itup))
			{
				items_orig = ginPostingListDecode((GinPostingList *) GinGetPosting(itup), &nitems);
				free_items_orig = true;
			}
			else
			{
				items_orig = (ItemPointer) GinGetPosting(itup);
				nitems = GinGetNPosting(itup);
				free_items_orig = false;
			}

			/* Remove any items from the list that need to be vacuumed. */
			items = ginVacuumItemPointers(gvs, items_orig, nitems, &nitems);

			if (free_items_orig)
				pfree(items_orig);

			/* If any item pointers were removed, recreate the tuple. */
			if (items)
			{
				OffsetNumber attnum;
				Datum		key;
				GinNullCategory category;
				GinPostingList *plist;
				int			plistsize;

				if (nitems > 0)
				{
					plist = ginCompressPostingList(items, nitems, GinMaxItemSize, NULL);
					plistsize = SizeOfGinPostingList(plist);
				}
				else
				{
					plist = NULL;
					plistsize = 0;
				}

				/*
				 * if we already created a temporary page, make changes in
				 * place
				 */
				if (tmppage == origpage)
				{
					/*
					 * On first difference, create a temporary copy of the
					 * page and copy the tuple's posting list to it.
					 */
					tmppage = PageGetTempPageCopy(origpage);

					/* set itup pointer to new page */
					itup = (IndexTuple) PageGetItem(tmppage, PageGetItemId(tmppage, i));
				}

				attnum = gintuple_get_attrnum(&gvs->ginstate, itup);
				key = gintuple_get_key(&gvs->ginstate, itup, &category);
				itup = GinFormTuple(&gvs->ginstate, attnum, key, category,
									(char *) plist, plistsize,
									nitems, true);
				if (plist)
					pfree(plist);
				PageIndexTupleDelete(tmppage, i);

				if (PageAddItem(tmppage, (Item) itup, IndexTupleSize(itup), i, false, false) != i)
					elog(ERROR, "failed to add item to index page in \"%s\"",
						 RelationGetRelationName(gvs->index));

				pfree(itup);
				pfree(items);
			}
		}
	}

	return (tmppage == origpage) ? NULL : tmppage;
}

//.........这里部分代码省略.........
		 * Insert into tail page.  Metapage is already locked
		 */
		OffsetNumber l,
					off;
		int			i,
					tupsize;
		char	   *ptr;

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

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

		rdata[0].next = rdata + 1;

		rdata[1].buffer = buffer;
		rdata[1].buffer_std = true;
		ptr = rdata[1].data = (char *) palloc(collector->sumsize);
		rdata[1].len = collector->sumsize;
		rdata[1].next = NULL;

		data.ntuples = collector->ntuples;

		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 - rdata[1].data) <= collector->sumsize);

		metadata->tailFreeSize = PageGetExactFreeSpace(page);

		MarkBufferDirty(buffer);
	}

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

	if (RelationNeedsWAL(index))
	{
		XLogRecPtr	recptr;

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

		recptr = XLogInsert(RM_GIN_ID, XLOG_GIN_UPDATE_META_PAGE, rdata);
		PageSetLSN(metapage, recptr);
		PageSetTLI(metapage, ThisTimeLineID);

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

	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 work_mem.
	 *
	 * ginInsertCleanup() should not be called inside our CRIT_SECTION.
	 */
	if (metadata->nPendingPages * GIN_PAGE_FREESIZE > work_mem * 1024L)
		needCleanup = true;

	UnlockReleaseBuffer(metabuffer);

	END_CRIT_SECTION();

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

static void
ginRedoUpdateMetapage(XLogReaderState *record)
{
	XLogRecPtr	lsn = record->EndRecPtr;
	ginxlogUpdateMeta *data = (ginxlogUpdateMeta *) XLogRecGetData(record);
	Buffer		metabuffer;
	Page		metapage;
	Buffer		buffer;

	/*
	 * Restore the metapage. This is essentially the same as a full-page
	 * image, so restore the metapage unconditionally without looking at the
	 * LSN, to avoid torn page hazards.
	 */
	metabuffer = XLogInitBufferForRedo(record, 0);
	Assert(BufferGetBlockNumber(metabuffer) == GIN_METAPAGE_BLKNO);
	metapage = BufferGetPage(metabuffer);

	memcpy(GinPageGetMeta(metapage), &data->metadata, sizeof(GinMetaPageData));
	PageSetLSN(metapage, lsn);
	MarkBufferDirty(metabuffer);

	if (data->ntuples > 0)
	{
		/*
		 * insert into tail page
		 */
		if (XLogReadBufferForRedo(record, 1, &buffer) == BLK_NEEDS_REDO)
		{
			Page		page = BufferGetPage(buffer);
			OffsetNumber off;
			int			i;
			Size		tupsize;
			char	   *payload;
			IndexTuple	tuples;
			Size		totaltupsize;

			payload = XLogRecGetBlockData(record, 1, &totaltupsize);
			tuples = (IndexTuple) payload;

			if (PageIsEmpty(page))
				off = FirstOffsetNumber;
			else
				off = OffsetNumberNext(PageGetMaxOffsetNumber(page));

			for (i = 0; i < data->ntuples; i++)
			{
				tupsize = IndexTupleSize(tuples);

				if (PageAddItem(page, (Item) tuples, tupsize, off,
								false, false) == InvalidOffsetNumber)
					elog(ERROR, "failed to add item to index page");

				tuples = (IndexTuple) (((char *) tuples) + tupsize);

				off++;
			}
			Assert(payload + totaltupsize == (char *) tuples);

			/*
			 * Increase counter of heap tuples
			 */
			GinPageGetOpaque(page)->maxoff++;

			PageSetLSN(page, lsn);
			MarkBufferDirty(buffer);
		}
		if (BufferIsValid(buffer))
			UnlockReleaseBuffer(buffer);
	}
	else if (data->prevTail != InvalidBlockNumber)
	{
		/*
		 * New tail
		 */
		if (XLogReadBufferForRedo(record, 1, &buffer) == BLK_NEEDS_REDO)
		{
			Page		page = BufferGetPage(buffer);

			GinPageGetOpaque(page)->rightlink = data->newRightlink;

			PageSetLSN(page, lsn);
			MarkBufferDirty(buffer);
		}
		if (BufferIsValid(buffer))
			UnlockReleaseBuffer(buffer);
	}

	UnlockReleaseBuffer(metabuffer);
}

static void
btree_xlog_mark_page_halfdead(uint8 info, XLogReaderState *record)
{
	XLogRecPtr	lsn = record->EndRecPtr;
	xl_btree_mark_page_halfdead *xlrec = (xl_btree_mark_page_halfdead *) XLogRecGetData(record);
	Buffer		buffer;
	Page		page;
	BTPageOpaque pageop;
	IndexTupleData trunctuple;

	/*
	 * In normal operation, we would lock all the pages this WAL record
	 * touches before changing any of them.  In WAL replay, it should be okay
	 * to lock just one page at a time, since no concurrent index updates can
	 * be happening, and readers should not care whether they arrive at the
	 * target page or not (since it's surely empty).
	 */

	/* parent page */
	if (XLogReadBufferForRedo(record, 1, &buffer) == BLK_NEEDS_REDO)
	{
		OffsetNumber poffset;
		ItemId		itemid;
		IndexTuple	itup;
		OffsetNumber nextoffset;
		BlockNumber rightsib;

		page = (Page) BufferGetPage(buffer);
		pageop = (BTPageOpaque) PageGetSpecialPointer(page);

		poffset = xlrec->poffset;

		nextoffset = OffsetNumberNext(poffset);
		itemid = PageGetItemId(page, nextoffset);
		itup = (IndexTuple) PageGetItem(page, itemid);
		rightsib = ItemPointerGetBlockNumber(&itup->t_tid);

		itemid = PageGetItemId(page, poffset);
		itup = (IndexTuple) PageGetItem(page, itemid);
		ItemPointerSet(&(itup->t_tid), rightsib, P_HIKEY);
		nextoffset = OffsetNumberNext(poffset);
		PageIndexTupleDelete(page, nextoffset);

		PageSetLSN(page, lsn);
		MarkBufferDirty(buffer);
	}
	if (BufferIsValid(buffer))
		UnlockReleaseBuffer(buffer);

	/* Rewrite the leaf page as a halfdead page */
	buffer = XLogInitBufferForRedo(record, 0);
	page = (Page) BufferGetPage(buffer);

	_bt_pageinit(page, BufferGetPageSize(buffer));
	pageop = (BTPageOpaque) PageGetSpecialPointer(page);

	pageop->btpo_prev = xlrec->leftblk;
	pageop->btpo_next = xlrec->rightblk;
	pageop->btpo.level = 0;
	pageop->btpo_flags = BTP_HALF_DEAD | BTP_LEAF;
	pageop->btpo_cycleid = 0;

	/*
	 * Construct a dummy hikey item that points to the next parent to be
	 * deleted (if any).
	 */
	MemSet(&trunctuple, 0, sizeof(IndexTupleData));
	trunctuple.t_info = sizeof(IndexTupleData);
	if (xlrec->topparent != InvalidBlockNumber)
		ItemPointerSet(&trunctuple.t_tid, xlrec->topparent, P_HIKEY);
	else
		ItemPointerSetInvalid(&trunctuple.t_tid);
	if (PageAddItem(page, (Item) &trunctuple, sizeof(IndexTupleData), P_HIKEY,
					false, false) == InvalidOffsetNumber)
		elog(ERROR, "could not add dummy high key to half-dead page");

	PageSetLSN(page, lsn);
	MarkBufferDirty(buffer);
	UnlockReleaseBuffer(buffer);
}

static ArrayTuple
vacuumSplitPage(GistVacuum *gv, Page tempPage, Buffer buffer, IndexTuple *addon, int curlenaddon)
{
	ArrayTuple	res = {NULL, 0, false};
	IndexTuple *vec;
	SplitedPageLayout *dist = NULL,
			   *ptr;
	int			i,
				veclen = 0;
	BlockNumber blkno = BufferGetBlockNumber(buffer);
	MemoryContext oldCtx = MemoryContextSwitchTo(gv->opCtx);

	vec = gistextractpage(tempPage, &veclen);
	vec = gistjoinvector(vec, &veclen, addon, curlenaddon);
	dist = gistSplit(gv->index, tempPage, vec, veclen, &(gv->giststate));

	MemoryContextSwitchTo(oldCtx);

	if (blkno != GIST_ROOT_BLKNO)
	{
		/* if non-root split then we should not allocate new buffer */
		dist->buffer = buffer;
		dist->page = tempPage;
		/* during vacuum we never split leaf page */
		GistPageGetOpaque(dist->page)->flags = 0;
	}
	else
		pfree(tempPage);

	res.itup = (IndexTuple *) palloc(sizeof(IndexTuple) * veclen);
	res.ituplen = 0;

	/* make new pages and fills them */
	for (ptr = dist; ptr; ptr = ptr->next)
	{
		char	   *data;

		if (ptr->buffer == InvalidBuffer)
		{
			ptr->buffer = gistNewBuffer(gv->index);
			GISTInitBuffer(ptr->buffer, 0);
			ptr->page = BufferGetPage(ptr->buffer);
		}
		ptr->block.blkno = BufferGetBlockNumber(ptr->buffer);

		data = (char *) (ptr->list);
		for (i = 0; i < ptr->block.num; i++)
		{
			if (PageAddItem(ptr->page, (Item) data, IndexTupleSize((IndexTuple) data), i + FirstOffsetNumber, false, false) == InvalidOffsetNumber)
				elog(ERROR, "failed to add item to index page in \"%s\"", RelationGetRelationName(gv->index));
			data += IndexTupleSize((IndexTuple) data);
		}

		ItemPointerSetBlockNumber(&(ptr->itup->t_tid), ptr->block.blkno);
		res.itup[res.ituplen] = (IndexTuple) palloc(IndexTupleSize(ptr->itup));
		memcpy(res.itup[res.ituplen], ptr->itup, IndexTupleSize(ptr->itup));
		res.ituplen++;
	}

	START_CRIT_SECTION();

	for (ptr = dist; ptr; ptr = ptr->next)
	{
		MarkBufferDirty(ptr->buffer);
		GistPageGetOpaque(ptr->page)->rightlink = InvalidBlockNumber;
	}

	/* restore splitted non-root page */
	if (blkno != GIST_ROOT_BLKNO)
	{
		PageRestoreTempPage(dist->page, BufferGetPage(dist->buffer));
		dist->page = BufferGetPage(dist->buffer);
	}

	if (!gv->index->rd_istemp)
	{
		XLogRecPtr	recptr;
		XLogRecData *rdata;
		ItemPointerData key;	/* set key for incomplete insert */
		char	   *xlinfo;

		ItemPointerSet(&key, blkno, TUPLE_IS_VALID);

		rdata = formSplitRdata(gv->index->rd_node, blkno,
							   false, &key, dist);
		xlinfo = rdata->data;

		recptr = XLogInsert(RM_GIST_ID, XLOG_GIST_PAGE_SPLIT, rdata);
		for (ptr = dist; ptr; ptr = ptr->next)
		{
			PageSetLSN(BufferGetPage(ptr->buffer), recptr);
			PageSetTLI(BufferGetPage(ptr->buffer), ThisTimeLineID);
		}

		pfree(xlinfo);
		pfree(rdata);
	}
	else
	{
		for (ptr = dist; ptr; ptr = ptr->next)
//.........这里部分代码省略.........

/*
 * redo any page update (except page split)
 */
static void
gistRedoPageUpdateRecord(XLogRecPtr lsn, XLogRecord *record)
{
	char	   *begin = XLogRecGetData(record);
	gistxlogPageUpdate *xldata = (gistxlogPageUpdate *) begin;
	Buffer		buffer;
	Page		page;
	char	   *data;

	if (BlockNumberIsValid(xldata->leftchild))
		gistRedoClearFollowRight(xldata->node, lsn, xldata->leftchild);

	/* nothing more to do if page was backed up (and no info to do it with) */
	if (record->xl_info & XLR_BKP_BLOCK_1)
		return;

	buffer = XLogReadBuffer(xldata->node, xldata->blkno, false);
	if (!BufferIsValid(buffer))
		return;
	page = (Page) BufferGetPage(buffer);

	if (XLByteLE(lsn, PageGetLSN(page)))
	{
		UnlockReleaseBuffer(buffer);
		return;
	}

	data = begin + sizeof(gistxlogPageUpdate);

	/* Delete old tuples */
	if (xldata->ntodelete > 0)
	{
		int			i;
		OffsetNumber *todelete = (OffsetNumber *) data;

		data += sizeof(OffsetNumber) * xldata->ntodelete;

		for (i = 0; i < xldata->ntodelete; i++)
			PageIndexTupleDelete(page, todelete[i]);
		if (GistPageIsLeaf(page))
			GistMarkTuplesDeleted(page);
	}

	/* add tuples */
	if (data - begin < record->xl_len)
	{
		OffsetNumber off = (PageIsEmpty(page)) ? FirstOffsetNumber :
		OffsetNumberNext(PageGetMaxOffsetNumber(page));

		while (data - begin < record->xl_len)
		{
			IndexTuple	itup = (IndexTuple) data;
			Size		sz = IndexTupleSize(itup);
			OffsetNumber l;

			data += sz;

			l = PageAddItem(page, (Item) itup, sz, off, false, false);
			if (l == InvalidOffsetNumber)
				elog(ERROR, "failed to add item to GiST index page, size %d bytes",
					 (int) sz);
			off++;
		}
	}
	else
	{
		/*
		 * special case: leafpage, nothing to insert, nothing to delete, then
		 * vacuum marks page
		 */
		if (GistPageIsLeaf(page) && xldata->ntodelete == 0)
			GistClearTuplesDeleted(page);
	}

	if (!GistPageIsLeaf(page) && PageGetMaxOffsetNumber(page) == InvalidOffsetNumber && xldata->blkno == GIST_ROOT_BLKNO)

		/*
		 * all links on non-leaf root page was deleted by vacuum full, so root
		 * page becomes a leaf
		 */
		GistPageSetLeaf(page);

	GistPageGetOpaque(page)->rightlink = InvalidBlockNumber;
	PageSetLSN(page, lsn);
	PageSetTLI(page, ThisTimeLineID);
	MarkBufferDirty(buffer);
	UnlockReleaseBuffer(buffer);
}

/*
 * Insert an index tuple into the index relation.  The revmap is updated to
 * mark the range containing the given page as pointing to the inserted entry.
 * A WAL record is written.
 *
 * The buffer, if valid, is first checked for free space to insert the new
 * entry; if there isn't enough, a new buffer is obtained and pinned.  No
 * buffer lock must be held on entry, no buffer lock is held on exit.
 *
 * Return value is the offset number where the tuple was inserted.
 */
OffsetNumber
brin_doinsert(Relation idxrel, BlockNumber pagesPerRange,
			  BrinRevmap *revmap, Buffer *buffer, BlockNumber heapBlk,
			  BrinTuple *tup, Size itemsz)
{
	Page		page;
	BlockNumber blk;
	OffsetNumber off;
	Buffer		revmapbuf;
	ItemPointerData tid;
	bool		extended;

	Assert(itemsz == MAXALIGN(itemsz));

	/* If the item is oversized, don't even bother. */
	if (itemsz > BrinMaxItemSize)
	{
		ereport(ERROR,
				(errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED),
			errmsg("index row size %lu exceeds maximum %lu for index \"%s\"",
				   (unsigned long) itemsz,
				   (unsigned long) BrinMaxItemSize,
				   RelationGetRelationName(idxrel))));
		return InvalidOffsetNumber;		/* keep compiler quiet */
	}

	/* Make sure the revmap is long enough to contain the entry we need */
	brinRevmapExtend(revmap, heapBlk);

	/*
	 * Acquire lock on buffer supplied by caller, if any.  If it doesn't have
	 * enough space, unpin it to obtain a new one below.
	 */
	if (BufferIsValid(*buffer))
	{
		/*
		 * It's possible that another backend (or ourselves!) extended the
		 * revmap over the page we held a pin on, so we cannot assume that
		 * it's still a regular page.
		 */
		LockBuffer(*buffer, BUFFER_LOCK_EXCLUSIVE);
		if (br_page_get_freespace(BufferGetPage(*buffer)) < itemsz)
		{
			UnlockReleaseBuffer(*buffer);
			*buffer = InvalidBuffer;
		}
	}

	/*
	 * If we still don't have a usable buffer, have brin_getinsertbuffer
	 * obtain one for us.
	 */
	if (!BufferIsValid(*buffer))
	{
		do
			*buffer = brin_getinsertbuffer(idxrel, InvalidBuffer, itemsz, &extended);
		while (!BufferIsValid(*buffer));
	}
	else
		extended = false;

	/* Now obtain lock on revmap buffer */
	revmapbuf = brinLockRevmapPageForUpdate(revmap, heapBlk);

	page = BufferGetPage(*buffer);
	blk = BufferGetBlockNumber(*buffer);

	/* Execute the actual insertion */
	START_CRIT_SECTION();
	if (extended)
		brin_page_init(BufferGetPage(*buffer), BRIN_PAGETYPE_REGULAR);
	off = PageAddItem(page, (Item) tup, itemsz, InvalidOffsetNumber,
					  false, false);
	if (off == InvalidOffsetNumber)
		elog(ERROR, "could not insert new index tuple to page");
	MarkBufferDirty(*buffer);

	BRIN_elog((DEBUG2, "inserted tuple (%u,%u) for range starting at %u",
			   blk, off, heapBlk));

	ItemPointerSet(&tid, blk, off);
	brinSetHeapBlockItemptr(revmapbuf, pagesPerRange, heapBlk, tid);
	MarkBufferDirty(revmapbuf);

	/* XLOG stuff */
	if (RelationNeedsWAL(idxrel))
	{
		xl_brin_insert xlrec;
		XLogRecPtr	recptr;
//.........这里部分代码省略.........

static void
spgRedoAddLeaf(XLogRecPtr lsn, XLogRecord *record)
{
	char	   *ptr = XLogRecGetData(record);
	spgxlogAddLeaf *xldata = (spgxlogAddLeaf *) ptr;
	SpGistLeafTuple leafTuple;
	Buffer		buffer;
	Page		page;

	/* we assume this is adequately aligned */
	ptr += sizeof(spgxlogAddLeaf);
	leafTuple = (SpGistLeafTuple) ptr;

	if (!(record->xl_info & XLR_BKP_BLOCK_1))
	{
		buffer = XLogReadBuffer(xldata->node, xldata->blknoLeaf,
								xldata->newPage);
		if (BufferIsValid(buffer))
		{
			page = BufferGetPage(buffer);

			if (xldata->newPage)
				SpGistInitBuffer(buffer, SPGIST_LEAF);

			if (!XLByteLE(lsn, PageGetLSN(page)))
			{
				/* insert new tuple */
				if (xldata->offnumLeaf != xldata->offnumHeadLeaf)
				{
					/* normal cases, tuple was added by SpGistPageAddNewItem */
					addOrReplaceTuple(page, (Item) leafTuple, leafTuple->size,
									  xldata->offnumLeaf);

					/* update head tuple's chain link if needed */
					if (xldata->offnumHeadLeaf != InvalidOffsetNumber)
					{
						SpGistLeafTuple head;

						head = (SpGistLeafTuple) PageGetItem(page,
														 PageGetItemId(page, xldata->offnumHeadLeaf));
						Assert(head->nextOffset == leafTuple->nextOffset);
						head->nextOffset = xldata->offnumLeaf;
					}
				}
				else
				{
					/* replacing a DEAD tuple */
					PageIndexTupleDelete(page, xldata->offnumLeaf);
					if (PageAddItem(page,
									(Item) leafTuple, leafTuple->size,
									xldata->offnumLeaf, false, false) != xldata->offnumLeaf)
						elog(ERROR, "failed to add item of size %u to SPGiST index page",
							 leafTuple->size);
				}

				PageSetLSN(page, lsn);
				PageSetTLI(page, ThisTimeLineID);
				MarkBufferDirty(buffer);
			}
			UnlockReleaseBuffer(buffer);
		}
	}

	/* update parent downlink if necessary */
	if (xldata->blknoParent != InvalidBlockNumber &&
		!(record->xl_info & XLR_BKP_BLOCK_2))
	{
		buffer = XLogReadBuffer(xldata->node, xldata->blknoParent, false);
		if (BufferIsValid(buffer))
		{
			page = BufferGetPage(buffer);
			if (!XLByteLE(lsn, PageGetLSN(page)))
			{
				SpGistInnerTuple tuple;

				tuple = (SpGistInnerTuple) PageGetItem(page,
													   PageGetItemId(page, xldata->offnumParent));

				spgUpdateNodeLink(tuple, xldata->nodeI,
								  xldata->blknoLeaf, xldata->offnumLeaf);

				PageSetLSN(page, lsn);
				PageSetTLI(page, ThisTimeLineID);
				MarkBufferDirty(buffer);
			}
			UnlockReleaseBuffer(buffer);
		}
	}
}

/*
 * redo any page update (except page split)
 */
static void
gistRedoPageUpdateRecord(XLogReaderState *record)
{
	XLogRecPtr	lsn = record->EndRecPtr;
	gistxlogPageUpdate *xldata = (gistxlogPageUpdate *) XLogRecGetData(record);
	Buffer		buffer;
	Page		page;

	if (XLogReadBufferForRedo(record, 0, &buffer) == BLK_NEEDS_REDO)
	{
		char	   *begin;
		char	   *data;
		Size		datalen;
		int			ninserted = 0;

		data = begin = XLogRecGetBlockData(record, 0, &datalen);

		page = (Page) BufferGetPage(buffer);

		/* Delete old tuples */
		if (xldata->ntodelete > 0)
		{
			int			i;
			OffsetNumber *todelete = (OffsetNumber *) data;

			data += sizeof(OffsetNumber) * xldata->ntodelete;

			for (i = 0; i < xldata->ntodelete; i++)
				PageIndexTupleDelete(page, todelete[i]);
			if (GistPageIsLeaf(page))
				GistMarkTuplesDeleted(page);
		}

		/* add tuples */
		if (data - begin < datalen)
		{
			OffsetNumber off = (PageIsEmpty(page)) ? FirstOffsetNumber :
			OffsetNumberNext(PageGetMaxOffsetNumber(page));

			while (data - begin < datalen)
			{
				IndexTuple	itup = (IndexTuple) data;
				Size		sz = IndexTupleSize(itup);
				OffsetNumber l;

				data += sz;

				l = PageAddItem(page, (Item) itup, sz, off, false, false);
				if (l == InvalidOffsetNumber)
					elog(ERROR, "failed to add item to GiST index page, size %d bytes",
						 (int) sz);
				off++;
				ninserted++;
			}
		}

		Assert(ninserted == xldata->ntoinsert);

		PageSetLSN(page, lsn);
		MarkBufferDirty(buffer);
	}

	/*
	 * Fix follow-right data on left child page
	 *
	 * This must be done while still holding the lock on the target page. Note
	 * that even if the target page no longer exists, we still attempt to
	 * replay the change on the child page.
	 */
	if (XLogRecHasBlockRef(record, 1))
		gistRedoClearFollowRight(record, 1);

	if (BufferIsValid(buffer))
		UnlockReleaseBuffer(buffer);
}

static void
spgRedoSplitTuple(XLogRecPtr lsn, XLogRecord *record)
{
	char	   *ptr = XLogRecGetData(record);
	spgxlogSplitTuple *xldata = (spgxlogSplitTuple *) ptr;
	SpGistInnerTuple prefixTuple;
	SpGistInnerTuple postfixTuple;
	Buffer		buffer;
	Page		page;

	/* we assume this is adequately aligned */
	ptr += sizeof(spgxlogSplitTuple);
	prefixTuple = (SpGistInnerTuple) ptr;
	ptr += prefixTuple->size;
	postfixTuple = (SpGistInnerTuple) ptr;

	/* insert postfix tuple first to avoid dangling link */
	if (xldata->blknoPostfix != xldata->blknoPrefix &&
		!(record->xl_info & XLR_BKP_BLOCK_2))
	{
		buffer = XLogReadBuffer(xldata->node, xldata->blknoPostfix,
								xldata->newPage);
		if (BufferIsValid(buffer))
		{
			page = BufferGetPage(buffer);

			if (xldata->newPage)
				SpGistInitBuffer(buffer, 0);

			if (!XLByteLE(lsn, PageGetLSN(page)))
			{
				addOrReplaceTuple(page, (Item) postfixTuple,
								  postfixTuple->size, xldata->offnumPostfix);

				PageSetLSN(page, lsn);
				PageSetTLI(page, ThisTimeLineID);
				MarkBufferDirty(buffer);
			}
			UnlockReleaseBuffer(buffer);
		}
	}

	/* now handle the original page */
	if (!(record->xl_info & XLR_BKP_BLOCK_1))
	{
		buffer = XLogReadBuffer(xldata->node, xldata->blknoPrefix, false);
		if (BufferIsValid(buffer))
		{
			page = BufferGetPage(buffer);
			if (!XLByteLE(lsn, PageGetLSN(page)))
			{
				PageIndexTupleDelete(page, xldata->offnumPrefix);
				if (PageAddItem(page, (Item) prefixTuple, prefixTuple->size,
								xldata->offnumPrefix, false, false) != xldata->offnumPrefix)
					elog(ERROR, "failed to add item of size %u to SPGiST index page",
						 prefixTuple->size);

				if (xldata->blknoPostfix == xldata->blknoPrefix)
					addOrReplaceTuple(page, (Item) postfixTuple,
									  postfixTuple->size,
									  xldata->offnumPostfix);

				PageSetLSN(page, lsn);
				PageSetTLI(page, ThisTimeLineID);
				MarkBufferDirty(buffer);
			}
			UnlockReleaseBuffer(buffer);
		}
	}
}

static void
spgRedoAddNode(XLogRecPtr lsn, XLogRecord *record)
{
	char	   *ptr = XLogRecGetData(record);
	spgxlogAddNode *xldata = (spgxlogAddNode *) ptr;
	SpGistInnerTuple innerTuple;
	SpGistState state;
	Buffer		buffer;
	Page		page;
	int			bbi;

	/* we assume this is adequately aligned */
	ptr += sizeof(spgxlogAddNode);
	innerTuple = (SpGistInnerTuple) ptr;

	fillFakeState(&state, xldata->stateSrc);

	if (xldata->blknoNew == InvalidBlockNumber)
	{
		/* update in place */
		Assert(xldata->blknoParent == InvalidBlockNumber);
		if (!(record->xl_info & XLR_BKP_BLOCK_1))
		{
			buffer = XLogReadBuffer(xldata->node, xldata->blkno, false);
			if (BufferIsValid(buffer))
			{
				page = BufferGetPage(buffer);
				if (!XLByteLE(lsn, PageGetLSN(page)))
				{
					PageIndexTupleDelete(page, xldata->offnum);
					if (PageAddItem(page, (Item) innerTuple, innerTuple->size,
									xldata->offnum,
									false, false) != xldata->offnum)
						elog(ERROR, "failed to add item of size %u to SPGiST index page",
							 innerTuple->size);

					PageSetLSN(page, lsn);
					PageSetTLI(page, ThisTimeLineID);
					MarkBufferDirty(buffer);
				}
				UnlockReleaseBuffer(buffer);
			}
		}
	}
	else
	{
		/* Install new tuple first so redirect is valid */
		if (!(record->xl_info & XLR_BKP_BLOCK_2))
		{
			buffer = XLogReadBuffer(xldata->node, xldata->blknoNew,
									xldata->newPage);
			if (BufferIsValid(buffer))
			{
				page = BufferGetPage(buffer);

				if (xldata->newPage)
					SpGistInitBuffer(buffer, 0);

				if (!XLByteLE(lsn, PageGetLSN(page)))
				{
					addOrReplaceTuple(page, (Item) innerTuple,
									  innerTuple->size, xldata->offnumNew);

					PageSetLSN(page, lsn);
					PageSetTLI(page, ThisTimeLineID);
					MarkBufferDirty(buffer);
				}
				UnlockReleaseBuffer(buffer);
			}
		}

		/* Delete old tuple, replacing it with redirect or placeholder tuple */
		if (!(record->xl_info & XLR_BKP_BLOCK_1))
		{
			buffer = XLogReadBuffer(xldata->node, xldata->blkno, false);
			if (BufferIsValid(buffer))
			{
				page = BufferGetPage(buffer);
				if (!XLByteLE(lsn, PageGetLSN(page)))
				{
					SpGistDeadTuple dt;

					if (state.isBuild)
						dt = spgFormDeadTuple(&state, SPGIST_PLACEHOLDER,
											  InvalidBlockNumber,
											  InvalidOffsetNumber);
					else
						dt = spgFormDeadTuple(&state, SPGIST_REDIRECT,
											  xldata->blknoNew,
											  xldata->offnumNew);

					PageIndexTupleDelete(page, xldata->offnum);
					if (PageAddItem(page, (Item) dt, dt->size,
									xldata->offnum,
									false, false) != xldata->offnum)
						elog(ERROR, "failed to add item of size %u to SPGiST index page",
							 dt->size);

					if (state.isBuild)
						SpGistPageGetOpaque(page)->nPlaceholder++;
//.........这里部分代码省略.........

/*
 * redo any page update (except page split)
 */
static void
gistRedoPageUpdateRecord(XLogRecPtr lsn, XLogRecord *record)
{
	char	   *begin = XLogRecGetData(record);
	gistxlogPageUpdate *xldata = (gistxlogPageUpdate *) begin;
	Buffer		buffer;
	Page		page;
	char	   *data;

	/*
	 * We need to acquire and hold lock on target page while updating the left
	 * child page.	If we have a full-page image of target page, getting the
	 * lock is a side-effect of restoring that image.  Note that even if the
	 * target page no longer exists, we'll still attempt to replay the change
	 * on the child page.
	 */
	if (record->xl_info & XLR_BKP_BLOCK(0))
		buffer = RestoreBackupBlock(lsn, record, 0, false, true);
	else
		buffer = XLogReadBuffer(xldata->node, xldata->blkno, false);

	/* Fix follow-right data on left child page */
	if (BlockNumberIsValid(xldata->leftchild))
		gistRedoClearFollowRight(lsn, record, 1,
								 xldata->node, xldata->leftchild);

	/* Done if target page no longer exists */
	if (!BufferIsValid(buffer))
		return;

	/* nothing more to do if page was backed up (and no info to do it with) */
	if (record->xl_info & XLR_BKP_BLOCK(0))
	{
		UnlockReleaseBuffer(buffer);
		return;
	}

	page = (Page) BufferGetPage(buffer);

	/* nothing more to do if change already applied */
	if (lsn <= PageGetLSN(page))
	{
		UnlockReleaseBuffer(buffer);
		return;
	}

	data = begin + sizeof(gistxlogPageUpdate);

	/* Delete old tuples */
	if (xldata->ntodelete > 0)
	{
		int			i;
		OffsetNumber *todelete = (OffsetNumber *) data;

		data += sizeof(OffsetNumber) * xldata->ntodelete;

		for (i = 0; i < xldata->ntodelete; i++)
			PageIndexTupleDelete(page, todelete[i]);
		if (GistPageIsLeaf(page))
			GistMarkTuplesDeleted(page);
	}

	/* add tuples */
	if (data - begin < record->xl_len)
	{
		OffsetNumber off = (PageIsEmpty(page)) ? FirstOffsetNumber :
		OffsetNumberNext(PageGetMaxOffsetNumber(page));

		while (data - begin < record->xl_len)
		{
			IndexTuple	itup = (IndexTuple) data;
			Size		sz = IndexTupleSize(itup);
			OffsetNumber l;

			data += sz;

			l = PageAddItem(page, (Item) itup, sz, off, false, false);
			if (l == InvalidOffsetNumber)
				elog(ERROR, "failed to add item to GiST index page, size %d bytes",
					 (int) sz);
			off++;
		}
	}
	else
	{
		/*
		 * special case: leafpage, nothing to insert, nothing to delete, then
		 * vacuum marks page
		 */
		if (GistPageIsLeaf(page) && xldata->ntodelete == 0)
			GistClearTuplesDeleted(page);
	}

	if (!GistPageIsLeaf(page) &&
		PageGetMaxOffsetNumber(page) == InvalidOffsetNumber &&
		xldata->blkno == GIST_ROOT_BLKNO)
	{
//.........这里部分代码省略.........

//.........这里部分代码省略.........
		if (ooffnum > omaxoffnum)
		{
			/* at end of page, but check for an(other) overflow page */
			oblkno = oopaque->hasho_nextblkno;
			if (!BlockNumberIsValid(oblkno))
				break;
			/*
			 * we ran out of tuples on this particular page, but we
			 * have more overflow pages; advance to next page.
			 */
			_hash_wrtbuf(rel, obuf);

			obuf = _hash_getbuf(rel, oblkno, HASH_WRITE);
			opage = BufferGetPage(obuf);
			_hash_checkpage(rel, opage, LH_OVERFLOW_PAGE);
			oopaque = (HashPageOpaque) PageGetSpecialPointer(opage);
			ooffnum = FirstOffsetNumber;
			omaxoffnum = PageGetMaxOffsetNumber(opage);
			continue;
		}

		/*
		 * Re-hash the tuple to determine which bucket it now belongs in.
		 *
		 * It is annoying to call the hash function while holding locks,
		 * but releasing and relocking the page for each tuple is unappealing
		 * too.
		 */
		hitem = (HashItem) PageGetItem(opage, PageGetItemId(opage, ooffnum));
		itup = &(hitem->hash_itup);
		datum = index_getattr(itup, 1, itupdesc, &null);
		Assert(!null);

		bucket = _hash_hashkey2bucket(_hash_datum2hashkey(rel, datum),
									  maxbucket, highmask, lowmask);

		if (bucket == nbucket)
		{
			/*
			 * insert the tuple into the new bucket.  if it doesn't fit on
			 * the current page in the new bucket, we must allocate a new
			 * overflow page and place the tuple on that page instead.
			 */
			itemsz = IndexTupleDSize(hitem->hash_itup)
				+ (sizeof(HashItemData) - sizeof(IndexTupleData));

			itemsz = MAXALIGN(itemsz);

			if (PageGetFreeSpace(npage) < itemsz)
			{
				/* write out nbuf and drop lock, but keep pin */
				_hash_chgbufaccess(rel, nbuf, HASH_WRITE, HASH_NOLOCK);
				/* chain to a new overflow page */
				nbuf = _hash_addovflpage(rel, metabuf, nbuf);
				npage = BufferGetPage(nbuf);
				_hash_checkpage(rel, npage, LH_OVERFLOW_PAGE);
				/* we don't need nopaque within the loop */
			}

			noffnum = OffsetNumberNext(PageGetMaxOffsetNumber(npage));
			if (PageAddItem(npage, (Item) hitem, itemsz, noffnum, LP_USED)
				== InvalidOffsetNumber)
				elog(ERROR, "failed to add index item to \"%s\"",
					 RelationGetRelationName(rel));

			/*
			 * now delete the tuple from the old bucket.  after this
			 * section of code, 'ooffnum' will actually point to the
			 * ItemId to which we would point if we had advanced it before
			 * the deletion (PageIndexTupleDelete repacks the ItemId
			 * array).	this also means that 'omaxoffnum' is exactly one
			 * less than it used to be, so we really can just decrement it
			 * instead of calling PageGetMaxOffsetNumber.
			 */
			PageIndexTupleDelete(opage, ooffnum);
			omaxoffnum = OffsetNumberPrev(omaxoffnum);
		}
		else
		{
			/*
			 * the tuple stays on this page.  we didn't move anything, so
			 * we didn't delete anything and therefore we don't have to
			 * change 'omaxoffnum'.
			 */
			Assert(bucket == obucket);
			ooffnum = OffsetNumberNext(ooffnum);
		}
	}

	/*
	 * We're at the end of the old bucket chain, so we're done partitioning
	 * the tuples.  Before quitting, call _hash_squeezebucket to ensure the
	 * tuples remaining in the old bucket (including the overflow pages) are
	 * packed as tightly as possible.  The new bucket is already tight.
	 */
	_hash_wrtbuf(rel, obuf);
	_hash_wrtbuf(rel, nbuf);

	_hash_squeezebucket(rel, obucket, start_oblkno);
}