C++ BlockNumberIsValid函數代碼示例

/*
 *	_hash_addovflpage
 *
 *	Add an overflow page to the bucket whose last page is pointed to by 'buf'.
 *
 *	On entry, the caller must hold a pin but no lock on 'buf'.	The pin is
 *	dropped before exiting (we assume the caller is not interested in 'buf'
 *	anymore).  The returned overflow page will be pinned and write-locked;
 *	it is guaranteed to be empty.
 *
 *	The caller must hold a pin, but no lock, on the metapage buffer.
 *	That buffer is returned in the same state.
 *
 *	The caller must hold at least share lock on the bucket, to ensure that
 *	no one else tries to compact the bucket meanwhile.	This guarantees that
 *	'buf' won't stop being part of the bucket while it's unlocked.
 *
 * NB: since this could be executed concurrently by multiple processes,
 * one should not assume that the returned overflow page will be the
 * immediate successor of the originally passed 'buf'.	Additional overflow
 * pages might have been added to the bucket chain in between.
 */
Buffer
_hash_addovflpage(Relation rel, Buffer metabuf, Buffer buf)
{
    Buffer		ovflbuf;
    Page		page;
    Page		ovflpage;
    HashPageOpaque pageopaque;
    HashPageOpaque ovflopaque;

    /* allocate and lock an empty overflow page */
    ovflbuf = _hash_getovflpage(rel, metabuf);

    /*
     * Write-lock the tail page.  It is okay to hold two buffer locks here
     * since there cannot be anyone else contending for access to ovflbuf.
     */
    _hash_chgbufaccess(rel, buf, HASH_NOLOCK, HASH_WRITE);

    /* probably redundant... */
    _hash_checkpage(rel, buf, LH_BUCKET_PAGE | LH_OVERFLOW_PAGE);

    /* loop to find current tail page, in case someone else inserted too */
    for (;;)
    {
        BlockNumber nextblkno;

        page = BufferGetPage(buf);
        pageopaque = (HashPageOpaque) PageGetSpecialPointer(page);
        nextblkno = pageopaque->hasho_nextblkno;

        if (!BlockNumberIsValid(nextblkno))
            break;

        /* we assume we do not need to write the unmodified page */
        _hash_relbuf(rel, buf);

        buf = _hash_getbuf(rel, nextblkno, HASH_WRITE, LH_OVERFLOW_PAGE);
    }

    /* now that we have correct backlink, initialize new overflow page */
    ovflpage = BufferGetPage(ovflbuf);
    ovflopaque = (HashPageOpaque) PageGetSpecialPointer(ovflpage);
    ovflopaque->hasho_prevblkno = BufferGetBlockNumber(buf);
    ovflopaque->hasho_nextblkno = InvalidBlockNumber;
    ovflopaque->hasho_bucket = pageopaque->hasho_bucket;
    ovflopaque->hasho_flag = LH_OVERFLOW_PAGE;
    ovflopaque->hasho_page_id = HASHO_PAGE_ID;

    MarkBufferDirty(ovflbuf);

    /* logically chain overflow page to previous page */
    pageopaque->hasho_nextblkno = BufferGetBlockNumber(ovflbuf);
    _hash_wrtbuf(rel, buf);

    return ovflbuf;
}

/*
 * Advance to next page in a bucket, if any.
 */
static void
_hash_readnext(Relation rel,
			   Buffer *bufp, Page *pagep, HashPageOpaque *opaquep)
{
	BlockNumber blkno;

	blkno = (*opaquep)->hasho_nextblkno;
	_hash_relbuf(rel, *bufp);
	*bufp = InvalidBuffer;
	if (BlockNumberIsValid(blkno))
	{
		*bufp = _hash_getbuf(rel, blkno, HASH_READ, LH_OVERFLOW_PAGE);
		*pagep = BufferGetPage(*bufp);
		*opaquep = (HashPageOpaque) PageGetSpecialPointer(*pagep);
	}
}

/*
 * Advance to next page in a bucket, if any.
 */
static void
_hash_readnext(Relation rel,
			   Buffer *bufp, Page *pagep, HashPageOpaque *opaquep)
{
	BlockNumber blkno;

	blkno = (*opaquep)->hasho_nextblkno;
	_hash_relbuf(rel, *bufp);
	*bufp = InvalidBuffer;
	/* check for interrupts while we're not holding any buffer lock */
	CHECK_FOR_INTERRUPTS();
	if (BlockNumberIsValid(blkno))
	{
		*bufp = _hash_getbuf(rel, blkno, HASH_READ, LH_OVERFLOW_PAGE);
		*pagep = BufferGetPage(*bufp);
		*opaquep = (HashPageOpaque) PageGetSpecialPointer(*pagep);
	}
}

/*
 * Advance to previous page in a bucket, if any.
 */
static void
_hash_readprev(Relation rel,
			   Buffer *bufp, Page *pagep, HashPageOpaque *opaquep)
{
	BlockNumber blkno;

	MIRROREDLOCK_BUFMGR_MUST_ALREADY_BE_HELD;

	blkno = (*opaquep)->hasho_prevblkno;
	_hash_relbuf(rel, *bufp);
	*bufp = InvalidBuffer;
	/* check for interrupts while we're not holding any buffer lock */
	CHECK_FOR_INTERRUPTS();
	if (BlockNumberIsValid(blkno))
	{
		*bufp = _hash_getbuf(rel, blkno, HASH_READ);
		_hash_checkpage(rel, *bufp, LH_BUCKET_PAGE | LH_OVERFLOW_PAGE);
		*pagep = BufferGetPage(*bufp);
		*opaquep = (HashPageOpaque) PageGetSpecialPointer(*pagep);
	}
}

/*
 *	_hash_freeovflpage() -
 *
 *	Remove this overflow page from its bucket's chain, and mark the page as
 *	free.  On entry, ovflbuf is write-locked; it is released before exiting.
 *
 *	Since this function is invoked in VACUUM, we provide an access strategy
 *	parameter that controls fetches of the bucket pages.
 *
 *	Returns the block number of the page that followed the given page
 *	in the bucket, or InvalidBlockNumber if no following page.
 *
 *	NB: caller must not hold lock on metapage, nor on either page that's
 *	adjacent in the bucket chain.  The caller had better hold exclusive lock
 *	on the bucket, too.
 */
BlockNumber
_hash_freeovflpage(Relation rel, Buffer ovflbuf,
				   BufferAccessStrategy bstrategy)
{
	HashMetaPage metap;
	Buffer		metabuf;
	Buffer		mapbuf;
	BlockNumber ovflblkno;
	BlockNumber prevblkno;
	BlockNumber blkno;
	BlockNumber nextblkno;
	HashPageOpaque ovflopaque;
	Page		ovflpage;
	Page		mappage;
	uint32	   *freep;
	uint32		ovflbitno;
	int32		bitmappage,
				bitmapbit;
	/*CS3223*/
	int 			index;
	int 			bitIndexInElement;
	uint32 			ovflElement;
	uint32 			temp, temp2;
	int 			i;
	BlockNumber 	nextblkno_temp;
	HashPageOpaque	pageopaque;
	Page			page;
	uint32 			*tempPointer;
	
	Bucket bucket PG_USED_FOR_ASSERTS_ONLY;

	/* Get information from the doomed page */
	_hash_checkpage(rel, ovflbuf, LH_OVERFLOW_PAGE);
	ovflblkno = BufferGetBlockNumber(ovflbuf);
	ovflpage = BufferGetPage(ovflbuf);
	ovflopaque = (HashPageOpaque) PageGetSpecialPointer(ovflpage);
	nextblkno = ovflopaque->hasho_nextblkno;
	prevblkno = ovflopaque->hasho_prevblkno;
	bucket = ovflopaque->hasho_bucket;
	
	/*CS3223*/
	/* find the length of the bucket chain*/
	
	while (i>=0)
	{
		//nextblkno_temp;

		page = BufferGetPage(ovflbuf);
		pageopaque = (HashPageOpaque) PageGetSpecialPointer(page);
		nextblkno_temp = pageopaque->hasho_nextblkno;

		if (!BlockNumberIsValid(nextblkno_temp))
			break;

		/* we assume we do not need to write the unmodified page */
		_hash_relbuf(rel, ovflbuf);

		ovflbuf = _hash_getbuf(rel, nextblkno_temp, HASH_WRITE, LH_OVERFLOW_PAGE);
		/*CS3223*/
		i++;
	}

	/*
	 * Zero the page for debugging's sake; then write and release it. (Note:
	 * if we failed to zero the page here, we'd have problems with the Assert
	 * in _hash_pageinit() when the page is reused.)
	 */
	MemSet(ovflpage, 0, BufferGetPageSize(ovflbuf));
	_hash_wrtbuf(rel, ovflbuf);

	/*
	 * Fix up the bucket chain.  this is a doubly-linked list, so we must fix
	 * up the bucket chain members behind and ahead of the overflow page being
	 * deleted.  No concurrency issues since we hold exclusive lock on the
	 * entire bucket.
	 */
	if (BlockNumberIsValid(prevblkno))
	{
		Buffer		prevbuf = _hash_getbuf_with_strategy(rel,
														 prevblkno,
														 HASH_WRITE,
										   LH_BUCKET_PAGE | LH_OVERFLOW_PAGE,
														 bstrategy);
		Page		prevpage = BufferGetPage(prevbuf);
//.........這裏部分代碼省略.........

/*
 *	_hash_freeovflpage() -
 *
 *	Remove this overflow page from its bucket's chain, and mark the page as
 *	free.  On entry, ovflbuf is write-locked; it is released before exiting.
 *
 *	Since this function is invoked in VACUUM, we provide an access strategy
 *	parameter that controls fetches of the bucket pages.
 *
 *	Returns the block number of the page that followed the given page
 *	in the bucket, or InvalidBlockNumber if no following page.
 *
 *	NB: caller must not hold lock on metapage, nor on either page that's
 *	adjacent in the bucket chain.  The caller had better hold exclusive lock
 *	on the bucket, too.
 */
BlockNumber
_hash_freeovflpage(Relation rel, Buffer ovflbuf,
                   BufferAccessStrategy bstrategy)
{
    HashMetaPage metap;
    Buffer		metabuf;
    Buffer		mapbuf;
    BlockNumber ovflblkno;
    BlockNumber prevblkno;
    BlockNumber blkno;
    BlockNumber nextblkno;
    HashPageOpaque ovflopaque;
    Page		ovflpage;
    Page		mappage;
    uint32	   *freep;
    uint32		ovflbitno;
    int32		bitmappage,
                bitmapbit;
    Bucket bucket PG_USED_FOR_ASSERTS_ONLY;

    /* Get information from the doomed page */
    _hash_checkpage(rel, ovflbuf, LH_OVERFLOW_PAGE);
    ovflblkno = BufferGetBlockNumber(ovflbuf);
    ovflpage = BufferGetPage(ovflbuf);
    ovflopaque = (HashPageOpaque) PageGetSpecialPointer(ovflpage);
    nextblkno = ovflopaque->hasho_nextblkno;
    prevblkno = ovflopaque->hasho_prevblkno;
    bucket = ovflopaque->hasho_bucket;

    /*
     * Zero the page for debugging's sake; then write and release it. (Note:
     * if we failed to zero the page here, we'd have problems with the Assert
     * in _hash_pageinit() when the page is reused.)
     */
    MemSet(ovflpage, 0, BufferGetPageSize(ovflbuf));
    _hash_wrtbuf(rel, ovflbuf);

    /*
     * Fix up the bucket chain.  this is a doubly-linked list, so we must fix
     * up the bucket chain members behind and ahead of the overflow page being
     * deleted.  No concurrency issues since we hold exclusive lock on the
     * entire bucket.
     */
    if (BlockNumberIsValid(prevblkno))
    {
        Buffer		prevbuf = _hash_getbuf_with_strategy(rel,
                              prevblkno,
                              HASH_WRITE,
                              LH_BUCKET_PAGE | LH_OVERFLOW_PAGE,
                              bstrategy);
        Page		prevpage = BufferGetPage(prevbuf);
        HashPageOpaque prevopaque = (HashPageOpaque) PageGetSpecialPointer(prevpage);

        Assert(prevopaque->hasho_bucket == bucket);
        prevopaque->hasho_nextblkno = nextblkno;
        _hash_wrtbuf(rel, prevbuf);
    }
    if (BlockNumberIsValid(nextblkno))
    {
        Buffer		nextbuf = _hash_getbuf_with_strategy(rel,
                              nextblkno,
                              HASH_WRITE,
                              LH_OVERFLOW_PAGE,
                              bstrategy);
        Page		nextpage = BufferGetPage(nextbuf);
        HashPageOpaque nextopaque = (HashPageOpaque) PageGetSpecialPointer(nextpage);

        Assert(nextopaque->hasho_bucket == bucket);
        nextopaque->hasho_prevblkno = prevblkno;
        _hash_wrtbuf(rel, nextbuf);
    }

    /* Note: bstrategy is intentionally not used for metapage and bitmap */

    /* Read the metapage so we can determine which bitmap page to use */
    metabuf = _hash_getbuf(rel, HASH_METAPAGE, HASH_READ, LH_META_PAGE);
    metap = HashPageGetMeta(BufferGetPage(metabuf));

    /* Identify which bit to set */
    ovflbitno = blkno_to_bitno(metap, ovflblkno);

    bitmappage = ovflbitno >> BMPG_SHIFT(metap);
    bitmapbit = ovflbitno & BMPG_MASK(metap);

//.........這裏部分代碼省略.........

//.........這裏部分代碼省略.........
	HashPageOpaque opaque;
	HashMetaPage metap;
	IndexTuple	itup;
	ItemPointer current;
	OffsetNumber offnum;

	MIRROREDLOCK_BUFMGR_MUST_ALREADY_BE_HELD;

	pgstat_count_index_scan(rel);

	current = &(scan->currentItemData);
	ItemPointerSetInvalid(current);

	/*
	 * We do not support hash scans with no index qualification, because we
	 * would have to read the whole index rather than just one bucket. That
	 * creates a whole raft of problems, since we haven't got a practical way
	 * to lock all the buckets against splits or compactions.
	 */
	if (scan->numberOfKeys < 1)
		ereport(ERROR,
				(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
				 errmsg("hash indexes do not support whole-index scans")));

	/*
	 * If the constant in the index qual is NULL, assume it cannot match any
	 * items in the index.
	 */
	if (scan->keyData[0].sk_flags & SK_ISNULL)
		return false;

	/*
	 * Okay to compute the hash key.  We want to do this before acquiring any
	 * locks, in case a user-defined hash function happens to be slow.
	 */
	hashkey = _hash_datum2hashkey(rel, scan->keyData[0].sk_argument);

	/*
	 * Acquire shared split lock so we can compute the target bucket safely
	 * (see README).
	 */
	_hash_getlock(rel, 0, HASH_SHARE);

	/* Read the metapage */
	metabuf = _hash_getbuf(rel, HASH_METAPAGE, HASH_READ);
	_hash_checkpage(rel, metabuf, LH_META_PAGE);
	metap = (HashMetaPage) BufferGetPage(metabuf);

	/*
	 * Compute the target bucket number, and convert to block number.
	 */
	bucket = _hash_hashkey2bucket(hashkey,
								  metap->hashm_maxbucket,
								  metap->hashm_highmask,
								  metap->hashm_lowmask);

	blkno = BUCKET_TO_BLKNO(metap, bucket);

	/* done with the metapage */
	_hash_relbuf(rel, metabuf);

	/*
	 * Acquire share lock on target bucket; then we can release split lock.
	 */
	_hash_getlock(rel, blkno, HASH_SHARE);

	_hash_droplock(rel, 0, HASH_SHARE);

	/* Update scan opaque state to show we have lock on the bucket */
	so->hashso_bucket = bucket;
	so->hashso_bucket_valid = true;
	so->hashso_bucket_blkno = blkno;

	/* Fetch the primary bucket page for the bucket */
	buf = _hash_getbuf(rel, blkno, HASH_READ);
	_hash_checkpage(rel, buf, LH_BUCKET_PAGE);
	page = BufferGetPage(buf);
	opaque = (HashPageOpaque) PageGetSpecialPointer(page);
	Assert(opaque->hasho_bucket == bucket);

	/* If a backwards scan is requested, move to the end of the chain */
	if (ScanDirectionIsBackward(dir))
	{
		while (BlockNumberIsValid(opaque->hasho_nextblkno))
			_hash_readnext(rel, &buf, &page, &opaque);
	}

	/* Now find the first tuple satisfying the qualification */
	if (!_hash_step(scan, &buf, dir))
		return false;

	/* if we're here, _hash_step found a valid tuple */
	offnum = ItemPointerGetOffsetNumber(current);
	_hash_checkpage(rel, buf, LH_BUCKET_PAGE | LH_OVERFLOW_PAGE);
	page = BufferGetPage(buf);
	itup = (IndexTuple) PageGetItem(page, PageGetItemId(page, offnum));
	scan->xs_ctup.t_self = itup->t_tid;

	return true;
}

/*
 * _hash_splitbucket -- split 'obucket' into 'obucket' and 'nbucket'
 *
 * We are splitting a bucket that consists of a base bucket page and zero
 * or more overflow (bucket chain) pages.  We must relocate tuples that
 * belong in the new bucket, and compress out any free space in the old
 * bucket.
 *
 * The caller must hold exclusive locks on both buckets to ensure that
 * no one else is trying to access them (see README).
 *
 * The caller must hold a pin, but no lock, on the metapage buffer.
 * The buffer is returned in the same state.  (The metapage is only
 * touched if it becomes necessary to add or remove overflow pages.)
 */
static void
_hash_splitbucket(Relation rel,
				  Buffer metabuf,
				  Bucket obucket,
				  Bucket nbucket,
				  BlockNumber start_oblkno,
				  BlockNumber start_nblkno,
				  uint32 maxbucket,
				  uint32 highmask,
				  uint32 lowmask)
{
	Bucket		bucket;
	Buffer		obuf;
	Buffer		nbuf;
	BlockNumber oblkno;
	BlockNumber nblkno;
	bool		null;
	Datum		datum;
	HashItem	hitem;
	HashPageOpaque oopaque;
	HashPageOpaque nopaque;
	IndexTuple	itup;
	Size		itemsz;
	OffsetNumber ooffnum;
	OffsetNumber noffnum;
	OffsetNumber omaxoffnum;
	Page		opage;
	Page		npage;
	TupleDesc	itupdesc = RelationGetDescr(rel);

	/*
	 * It should be okay to simultaneously write-lock pages from each
	 * bucket, since no one else can be trying to acquire buffer lock
	 * on pages of either bucket.
	 */
	oblkno = start_oblkno;
	nblkno = start_nblkno;
	obuf = _hash_getbuf(rel, oblkno, HASH_WRITE);
	nbuf = _hash_getbuf(rel, nblkno, HASH_WRITE);
	opage = BufferGetPage(obuf);
	npage = BufferGetPage(nbuf);

	_hash_checkpage(rel, opage, LH_BUCKET_PAGE);
	oopaque = (HashPageOpaque) PageGetSpecialPointer(opage);

	/* initialize the new bucket's primary page */
	_hash_pageinit(npage, BufferGetPageSize(nbuf));
	nopaque = (HashPageOpaque) PageGetSpecialPointer(npage);
	nopaque->hasho_prevblkno = InvalidBlockNumber;
	nopaque->hasho_nextblkno = InvalidBlockNumber;
	nopaque->hasho_bucket = nbucket;
	nopaque->hasho_flag = LH_BUCKET_PAGE;
	nopaque->hasho_filler = HASHO_FILL;

	/*
	 * Partition the tuples in the old bucket between the old bucket and the
	 * new bucket, advancing along the old bucket's overflow bucket chain
	 * and adding overflow pages to the new bucket as needed.
	 */
	ooffnum = FirstOffsetNumber;
	omaxoffnum = PageGetMaxOffsetNumber(opage);
	for (;;)
	{
		/*
		 * at each iteration through this loop, each of these variables
		 * should be up-to-date: obuf opage oopaque ooffnum omaxoffnum
		 */

		/* check if we're at the end of the page */
		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);
//.........這裏部分代碼省略.........

//.........這裏部分代碼省略.........
            }

            if (kill_tuple)
            {
                /* mark the item for deletion */
                deletable[ndeletable++] = offno;
            }
            else
            {
                /* we're keeping it, so count it */
                if (num_index_tuples)
                    *num_index_tuples += 1;
            }
        }

        /* retain the pin on primary bucket page till end of bucket scan */
        if (blkno == bucket_blkno)
            retain_pin = true;
        else
            retain_pin = false;

        blkno = opaque->hasho_nextblkno;

        /*
         * Apply deletions, advance to next page and write page if needed.
         */
        if (ndeletable > 0)
        {
            PageIndexMultiDelete(page, deletable, ndeletable);
            bucket_dirty = true;
            curr_page_dirty = true;
        }

        /* bail out if there are no more pages to scan. */
        if (!BlockNumberIsValid(blkno))
            break;

        next_buf = _hash_getbuf_with_strategy(rel, blkno, HASH_WRITE,
                                              LH_OVERFLOW_PAGE,
                                              bstrategy);

        /*
         * release the lock on previous page after acquiring the lock on next
         * page
         */
        if (curr_page_dirty)
        {
            if (retain_pin)
                _hash_chgbufaccess(rel, buf, HASH_WRITE, HASH_NOLOCK);
            else
                _hash_wrtbuf(rel, buf);
            curr_page_dirty = false;
        }
        else if (retain_pin)
            _hash_chgbufaccess(rel, buf, HASH_READ, HASH_NOLOCK);
        else
            _hash_relbuf(rel, buf);

        buf = next_buf;
    }

    /*
     * lock the bucket page to clear the garbage flag and squeeze the bucket.
     * if the current buffer is same as bucket buffer, then we already have
     * lock on bucket page.
     */
    if (buf != bucket_buf)
    {
        _hash_relbuf(rel, buf);
        _hash_chgbufaccess(rel, bucket_buf, HASH_NOLOCK, HASH_WRITE);
    }

    /*
     * Clear the garbage flag from bucket after deleting the tuples that are
     * moved by split.  We purposefully clear the flag before squeeze bucket,
     * so that after restart, vacuum shouldn't again try to delete the moved
     * by split tuples.
     */
    if (split_cleanup)
    {
        HashPageOpaque bucket_opaque;
        Page		page;

        page = BufferGetPage(bucket_buf);
        bucket_opaque = (HashPageOpaque) PageGetSpecialPointer(page);

        bucket_opaque->hasho_flag &= ~LH_BUCKET_NEEDS_SPLIT_CLEANUP;
    }

    /*
     * If we have deleted anything, try to compact free space.  For squeezing
     * the bucket, we must have a cleanup lock, else it can impact the
     * ordering of tuples for a scan that has started before it.
     */
    if (bucket_dirty && IsBufferCleanupOK(bucket_buf))
        _hash_squeezebucket(rel, cur_bucket, bucket_blkno, bucket_buf,
                            bstrategy);
    else
        _hash_chgbufaccess(rel, bucket_buf, HASH_WRITE, HASH_NOLOCK);
}

/*
 *	_hash_addovflpage
 *
 *	Add an overflow page to the bucket whose last page is pointed to by 'buf'.
 *
 *	On entry, the caller must hold a pin but no lock on 'buf'.  The pin is
 *	dropped before exiting (we assume the caller is not interested in 'buf'
 *	anymore) if not asked to retain.  The pin will be retained only for the
 *	primary bucket.  The returned overflow page will be pinned and
 *	write-locked; it is guaranteed to be empty.
 *
 *	The caller must hold a pin, but no lock, on the metapage buffer.
 *	That buffer is returned in the same state.
 *
 * NB: since this could be executed concurrently by multiple processes,
 * one should not assume that the returned overflow page will be the
 * immediate successor of the originally passed 'buf'.  Additional overflow
 * pages might have been added to the bucket chain in between.
 */
Buffer
_hash_addovflpage(Relation rel, Buffer metabuf, Buffer buf, bool retain_pin)
{
	Buffer		ovflbuf;
	Page		page;
	Page		ovflpage;
	HashPageOpaque pageopaque;
	HashPageOpaque ovflopaque;
	HashMetaPage metap;
	Buffer		mapbuf = InvalidBuffer;
	Buffer		newmapbuf = InvalidBuffer;
	BlockNumber blkno;
	uint32		orig_firstfree;
	uint32		splitnum;
	uint32	   *freep = NULL;
	uint32		max_ovflpg;
	uint32		bit;
	uint32		bitmap_page_bit;
	uint32		first_page;
	uint32		last_bit;
	uint32		last_page;
	uint32		i,
				j;
	bool		page_found = false;

	/*
	 * Write-lock the tail page.  Here, we need to maintain locking order such
	 * that, first acquire the lock on tail page of bucket, then on meta page
	 * to find and lock the bitmap page and if it is found, then lock on meta
	 * page is released, then finally acquire the lock on new overflow buffer.
	 * We need this locking order to avoid deadlock with backends that are
	 * doing inserts.
	 *
	 * Note: We could have avoided locking many buffers here if we made two
	 * WAL records for acquiring an overflow page (one to allocate an overflow
	 * page and another to add it to overflow bucket chain).  However, doing
	 * so can leak an overflow page, if the system crashes after allocation.
	 * Needless to say, it is better to have a single record from a
	 * performance point of view as well.
	 */
	LockBuffer(buf, BUFFER_LOCK_EXCLUSIVE);

	/* probably redundant... */
	_hash_checkpage(rel, buf, LH_BUCKET_PAGE | LH_OVERFLOW_PAGE);

	/* loop to find current tail page, in case someone else inserted too */
	for (;;)
	{
		BlockNumber nextblkno;

		page = BufferGetPage(buf);
		pageopaque = (HashPageOpaque) PageGetSpecialPointer(page);
		nextblkno = pageopaque->hasho_nextblkno;

		if (!BlockNumberIsValid(nextblkno))
			break;

		/* we assume we do not need to write the unmodified page */
		if (retain_pin)
		{
			/* pin will be retained only for the primary bucket page */
			Assert((pageopaque->hasho_flag & LH_PAGE_TYPE) == LH_BUCKET_PAGE);
			LockBuffer(buf, BUFFER_LOCK_UNLOCK);
		}
		else
			_hash_relbuf(rel, buf);

		retain_pin = false;

		buf = _hash_getbuf(rel, nextblkno, HASH_WRITE, LH_OVERFLOW_PAGE);
	}

	/* Get exclusive lock on the meta page */
	LockBuffer(metabuf, BUFFER_LOCK_EXCLUSIVE);

	_hash_checkpage(rel, metabuf, LH_META_PAGE);
	metap = HashPageGetMeta(BufferGetPage(metabuf));

	/* start search at hashm_firstfree */
	orig_firstfree = metap->hashm_firstfree;
	first_page = orig_firstfree >> BMPG_SHIFT(metap);
//.........這裏部分代碼省略.........

static void
btree_xlog_vacuum(XLogReaderState *record)
{
	XLogRecPtr	lsn = record->EndRecPtr;
	Buffer		buffer;
	Page		page;
	BTPageOpaque opaque;
#ifdef UNUSED
	xl_btree_vacuum *xlrec = (xl_btree_vacuum *) XLogRecGetData(record);

	/*
	 * This section of code is thought to be no longer needed, after analysis
	 * of the calling paths. It is retained to allow the code to be reinstated
	 * if a flaw is revealed in that thinking.
	 *
	 * If we are running non-MVCC scans using this index we need to do some
	 * additional work to ensure correctness, which is known as a "pin scan"
	 * described in more detail in next paragraphs. We used to do the extra
	 * work in all cases, whereas we now avoid that work in most cases. If
	 * lastBlockVacuumed is set to InvalidBlockNumber then we skip the
	 * additional work required for the pin scan.
	 *
	 * Avoiding this extra work is important since it requires us to touch
	 * every page in the index, so is an O(N) operation. Worse, it is an
	 * operation performed in the foreground during redo, so it delays
	 * replication directly.
	 *
	 * If queries might be active then we need to ensure every leaf page is
	 * unpinned between the lastBlockVacuumed and the current block, if there
	 * are any.  This prevents replay of the VACUUM from reaching the stage of
	 * removing heap tuples while there could still be indexscans "in flight"
	 * to those particular tuples for those scans which could be confused by
	 * finding new tuples at the old TID locations (see nbtree/README).
	 *
	 * It might be worth checking if there are actually any backends running;
	 * if not, we could just skip this.
	 *
	 * Since VACUUM can visit leaf pages out-of-order, it might issue records
	 * with lastBlockVacuumed >= block; that's not an error, it just means
	 * nothing to do now.
	 *
	 * Note: since we touch all pages in the range, we will lock non-leaf
	 * pages, and also any empty (all-zero) pages that may be in the index. It
	 * doesn't seem worth the complexity to avoid that.  But it's important
	 * that HotStandbyActiveInReplay() will not return true if the database
	 * isn't yet consistent; so we need not fear reading still-corrupt blocks
	 * here during crash recovery.
	 */
	if (HotStandbyActiveInReplay() && BlockNumberIsValid(xlrec->lastBlockVacuumed))
	{
		RelFileNode thisrnode;
		BlockNumber thisblkno;
		BlockNumber blkno;

		XLogRecGetBlockTag(record, 0, &thisrnode, NULL, &thisblkno);

		for (blkno = xlrec->lastBlockVacuumed + 1; blkno < thisblkno; blkno++)
		{
			/*
			 * We use RBM_NORMAL_NO_LOG mode because it's not an error
			 * condition to see all-zero pages.  The original btvacuumpage
			 * scan would have skipped over all-zero pages, noting them in FSM
			 * but not bothering to initialize them just yet; so we mustn't
			 * throw an error here.  (We could skip acquiring the cleanup lock
			 * if PageIsNew, but it's probably not worth the cycles to test.)
			 *
			 * XXX we don't actually need to read the block, we just need to
			 * confirm it is unpinned. If we had a special call into the
			 * buffer manager we could optimise this so that if the block is
			 * not in shared_buffers we confirm it as unpinned. Optimizing
			 * this is now moot, since in most cases we avoid the scan.
			 */
			buffer = XLogReadBufferExtended(thisrnode, MAIN_FORKNUM, blkno,
											RBM_NORMAL_NO_LOG);
			if (BufferIsValid(buffer))
			{
				LockBufferForCleanup(buffer);
				UnlockReleaseBuffer(buffer);
			}
		}
	}
#endif

	/*
	 * Like in btvacuumpage(), we need to take a cleanup lock on every leaf
	 * page. See nbtree/README for details.
	 */
	if (XLogReadBufferForRedoExtended(record, 0, RBM_NORMAL, true, &buffer)
		== BLK_NEEDS_REDO)
	{
		char	   *ptr;
		Size		len;

		ptr = XLogRecGetBlockData(record, 0, &len);

		page = (Page) BufferGetPage(buffer);

		if (len > 0)
		{
			OffsetNumber *unused;
//.........這裏部分代碼省略.........

/*
 *	_hash_squeezebucket(rel, bucket)
 *
 *	Try to squeeze the tuples onto pages occurring earlier in the
 *	bucket chain in an attempt to free overflow pages. When we start
 *	the "squeezing", the page from which we start taking tuples (the
 *	"read" page) is the last bucket in the bucket chain and the page
 *	onto which we start squeezing tuples (the "write" page) is the
 *	first page in the bucket chain.  The read page works backward and
 *	the write page works forward; the procedure terminates when the
 *	read page and write page are the same page.
 *
 *	At completion of this procedure, it is guaranteed that all pages in
 *	the bucket are nonempty, unless the bucket is totally empty (in
 *	which case all overflow pages will be freed).  The original implementation
 *	required that to be true on entry as well, but it's a lot easier for
 *	callers to leave empty overflow pages and let this guy clean it up.
 *
 *	Caller must hold exclusive lock on the target bucket.  This allows
 *	us to safely lock multiple pages in the bucket.
 */
void
_hash_squeezebucket(Relation rel,
					Bucket bucket,
					BlockNumber bucket_blkno)
{
	Buffer		wbuf;
	Buffer		rbuf = 0;
	BlockNumber wblkno;
	BlockNumber rblkno;
	Page		wpage;
	Page		rpage;
	HashPageOpaque wopaque;
	HashPageOpaque ropaque;
	OffsetNumber woffnum;
	OffsetNumber roffnum;
	IndexTuple	itup;
	Size		itemsz;

	/*
	 * start squeezing into the base bucket page.
	 */
	wblkno = bucket_blkno;
	wbuf = _hash_getbuf(rel, wblkno, HASH_WRITE);
	_hash_checkpage(rel, wbuf, LH_BUCKET_PAGE);
	wpage = BufferGetPage(wbuf);
	wopaque = (HashPageOpaque) PageGetSpecialPointer(wpage);

	/*
	 * if there aren't any overflow pages, there's nothing to squeeze.
	 */
	if (!BlockNumberIsValid(wopaque->hasho_nextblkno))
	{
		_hash_relbuf(rel, wbuf);
		return;
	}

	/*
	 * find the last page in the bucket chain by starting at the base bucket
	 * page and working forward.
	 */
	ropaque = wopaque;
	do
	{
		rblkno = ropaque->hasho_nextblkno;
		if (ropaque != wopaque)
			_hash_relbuf(rel, rbuf);
		rbuf = _hash_getbuf(rel, rblkno, HASH_WRITE);
		_hash_checkpage(rel, rbuf, LH_OVERFLOW_PAGE);
		rpage = BufferGetPage(rbuf);
		ropaque = (HashPageOpaque) PageGetSpecialPointer(rpage);
		Assert(ropaque->hasho_bucket == bucket);
	} while (BlockNumberIsValid(ropaque->hasho_nextblkno));

	/*
	 * squeeze the tuples.
	 */
	roffnum = FirstOffsetNumber;
	for (;;)
	{
		/* this test is needed in case page is empty on entry */
		if (roffnum <= PageGetMaxOffsetNumber(rpage))
		{
			itup = (IndexTuple) PageGetItem(rpage,
											PageGetItemId(rpage, roffnum));
			itemsz = IndexTupleDSize(*itup);
			itemsz = MAXALIGN(itemsz);

			/*
			 * Walk up the bucket chain, looking for a page big enough for
			 * this item.  Exit if we reach the read page.
			 */
			while (PageGetFreeSpace(wpage) < itemsz)
			{
				Assert(!PageIsEmpty(wpage));

				wblkno = wopaque->hasho_nextblkno;
				Assert(BlockNumberIsValid(wblkno));

				_hash_wrtbuf(rel, wbuf);
//.........這裏部分代碼省略.........

/*
 *	_hash_doinsert() -- Handle insertion of a single index tuple.
 *
 *		This routine is called by the public interface routines, hashbuild
 *		and hashinsert.  By here, itup is completely filled in.
 */
void
_hash_doinsert(Relation rel, IndexTuple itup)
{
	Buffer		buf;
	Buffer		metabuf;
	HashMetaPage metap;
	BlockNumber blkno;
	Page		page;
	HashPageOpaque pageopaque;
	Size		itemsz;
	bool		do_expand;
	uint32		hashkey;
	Bucket		bucket;

	/*
	 * Get the hash key for the item (it's stored in the index tuple itself).
	 */
	hashkey = _hash_get_indextuple_hashkey(itup);

	/* compute item size too */
	itemsz = IndexTupleDSize(*itup);
	itemsz = MAXALIGN(itemsz);	/* be safe, PageAddItem will do this but we
								 * need to be consistent */

	/*
	 * Acquire shared split lock so we can compute the target bucket safely
	 * (see README).
	 */
	_hash_getlock(rel, 0, HASH_SHARE);

	/* Read the metapage */
	metabuf = _hash_getbuf(rel, HASH_METAPAGE, HASH_READ, LH_META_PAGE);
	metap = HashPageGetMeta(BufferGetPage(metabuf));

	/*
	 * Check whether the item can fit on a hash page at all. (Eventually, we
	 * ought to try to apply TOAST methods if not.)  Note that at this point,
	 * itemsz doesn't include the ItemId.
	 *
	 * XXX this is useless code if we are only storing hash keys.
	 */
	if (itemsz > HashMaxItemSize((Page) metap))
		ereport(ERROR,
				(errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED),
				 errmsg("index row size %lu exceeds hash maximum %lu",
						(unsigned long) itemsz,
						(unsigned long) HashMaxItemSize((Page) metap)),
			errhint("Values larger than a buffer page cannot be indexed.")));

	/*
	 * Compute the target bucket number, and convert to block number.
	 */
	bucket = _hash_hashkey2bucket(hashkey,
								  metap->hashm_maxbucket,
								  metap->hashm_highmask,
								  metap->hashm_lowmask);

	blkno = BUCKET_TO_BLKNO(metap, bucket);

	/* release lock on metapage, but keep pin since we'll need it again */
	_hash_chgbufaccess(rel, metabuf, HASH_READ, HASH_NOLOCK);

	/*
	 * Acquire share lock on target bucket; then we can release split lock.
	 */
	_hash_getlock(rel, blkno, HASH_SHARE);

	_hash_droplock(rel, 0, HASH_SHARE);

	/* Fetch the primary bucket page for the bucket */
	buf = _hash_getbuf(rel, blkno, HASH_WRITE, LH_BUCKET_PAGE);
	page = BufferGetPage(buf);
	pageopaque = (HashPageOpaque) PageGetSpecialPointer(page);
	Assert(pageopaque->hasho_bucket == bucket);

	/* Do the insertion */
	while (PageGetFreeSpace(page) < itemsz)
	{
		/*
		 * no space on this page; check for an overflow page
		 */
		BlockNumber nextblkno = pageopaque->hasho_nextblkno;

		if (BlockNumberIsValid(nextblkno))
		{
			/*
			 * ovfl page exists; go get it.  if it doesn't have room, we'll
			 * find out next pass through the loop test above.
			 */
			_hash_relbuf(rel, buf);
			buf = _hash_getbuf(rel, nextblkno, HASH_WRITE, LH_OVERFLOW_PAGE);
			page = BufferGetPage(buf);
		}
		else
//.........這裏部分代碼省略.........

/*
 *	_hash_finish_split() -- Finish the previously interrupted split operation
 *
 * To complete the split operation, we form the hash table of TIDs in new
 * bucket which is then used by split operation to skip tuples that are
 * already moved before the split operation was previously interrupted.
 *
 * The caller must hold a pin, but no lock, on the metapage and old bucket's
 * primary page buffer.  The buffers are returned in the same state.  (The
 * metapage is only touched if it becomes necessary to add or remove overflow
 * pages.)
 */
void
_hash_finish_split(Relation rel, Buffer metabuf, Buffer obuf, Bucket obucket,
				   uint32 maxbucket, uint32 highmask, uint32 lowmask)
{
	HASHCTL		hash_ctl;
	HTAB	   *tidhtab;
	Buffer		bucket_nbuf = InvalidBuffer;
	Buffer		nbuf;
	Page		npage;
	BlockNumber nblkno;
	BlockNumber bucket_nblkno;
	HashPageOpaque npageopaque;
	Bucket		nbucket;
	bool		found;

	/* Initialize hash tables used to track TIDs */
	memset(&hash_ctl, 0, sizeof(hash_ctl));
	hash_ctl.keysize = sizeof(ItemPointerData);
	hash_ctl.entrysize = sizeof(ItemPointerData);
	hash_ctl.hcxt = CurrentMemoryContext;

	tidhtab =
		hash_create("bucket ctids",
					256,		/* arbitrary initial size */
					&hash_ctl,
					HASH_ELEM | HASH_BLOBS | HASH_CONTEXT);

	bucket_nblkno = nblkno = _hash_get_newblock_from_oldbucket(rel, obucket);

	/*
	 * Scan the new bucket and build hash table of TIDs
	 */
	for (;;)
	{
		OffsetNumber noffnum;
		OffsetNumber nmaxoffnum;

		nbuf = _hash_getbuf(rel, nblkno, HASH_READ,
							LH_BUCKET_PAGE | LH_OVERFLOW_PAGE);

		/* remember the primary bucket buffer to acquire cleanup lock on it. */
		if (nblkno == bucket_nblkno)
			bucket_nbuf = nbuf;

		npage = BufferGetPage(nbuf);
		npageopaque = (HashPageOpaque) PageGetSpecialPointer(npage);

		/* Scan each tuple in new page */
		nmaxoffnum = PageGetMaxOffsetNumber(npage);
		for (noffnum = FirstOffsetNumber;
			 noffnum <= nmaxoffnum;
			 noffnum = OffsetNumberNext(noffnum))
		{
			IndexTuple	itup;

			/* Fetch the item's TID and insert it in hash table. */
			itup = (IndexTuple) PageGetItem(npage,
											PageGetItemId(npage, noffnum));

			(void) hash_search(tidhtab, &itup->t_tid, HASH_ENTER, &found);

			Assert(!found);
		}

		nblkno = npageopaque->hasho_nextblkno;

		/*
		 * release our write lock without modifying buffer and ensure to
		 * retain the pin on primary bucket.
		 */
		if (nbuf == bucket_nbuf)
			LockBuffer(nbuf, BUFFER_LOCK_UNLOCK);
		else
			_hash_relbuf(rel, nbuf);

		/* Exit loop if no more overflow pages in new bucket */
		if (!BlockNumberIsValid(nblkno))
			break;
	}

	/*
	 * Conditionally get the cleanup lock on old and new buckets to perform
	 * the split operation.  If we don't get the cleanup locks, silently give
	 * up and next insertion on old bucket will try again to complete the
	 * split.
	 */
	if (!ConditionalLockBufferForCleanup(obuf))
	{
//.........這裏部分代碼省略.........

//.........這裏部分代碼省略.........
					/* be tidy */
					for (i = 0; i < nitups; i++)
						pfree(itups[i]);
					nitups = 0;
					all_tups_size = 0;

					/* chain to a new overflow page */
					nbuf = _hash_addovflpage(rel, metabuf, nbuf, (nbuf == bucket_nbuf) ? true : false);
					npage = BufferGetPage(nbuf);
					nopaque = (HashPageOpaque) PageGetSpecialPointer(npage);
				}

				itups[nitups++] = new_itup;
				all_tups_size += itemsz;
			}
			else
			{
				/*
				 * the tuple stays on this page, so nothing to do.
				 */
				Assert(bucket == obucket);
			}
		}

		oblkno = oopaque->hasho_nextblkno;

		/* retain the pin on the old primary bucket */
		if (obuf == bucket_obuf)
			LockBuffer(obuf, BUFFER_LOCK_UNLOCK);
		else
			_hash_relbuf(rel, obuf);

		/* Exit loop if no more overflow pages in old bucket */
		if (!BlockNumberIsValid(oblkno))
		{
			/*
			 * Change the shared buffer state in critical section, otherwise
			 * any error could make it unrecoverable.
			 */
			START_CRIT_SECTION();

			_hash_pgaddmultitup(rel, nbuf, itups, itup_offsets, nitups);
			MarkBufferDirty(nbuf);
			/* log the split operation before releasing the lock */
			log_split_page(rel, nbuf);

			END_CRIT_SECTION();

			if (nbuf == bucket_nbuf)
				LockBuffer(nbuf, BUFFER_LOCK_UNLOCK);
			else
				_hash_relbuf(rel, nbuf);

			/* be tidy */
			for (i = 0; i < nitups; i++)
				pfree(itups[i]);
			break;
		}

		/* Else, advance to next old page */
		obuf = _hash_getbuf(rel, oblkno, HASH_READ, LH_OVERFLOW_PAGE);
		opage = BufferGetPage(obuf);
		oopaque = (HashPageOpaque) PageGetSpecialPointer(opage);
	}

	/*