C++ RelationOpenSmgr函数代码示例

/*
 * SetMatViewToPopulated
 *		Indicate that the materialized view has been populated by its query.
 *
 * NOTE: The heap starts out in a state that doesn't look scannable, and can
 * only transition from there to scannable at the time a new heap is created.
 *
 * NOTE: caller must be holding an appropriate lock on the relation.
 */
void
SetMatViewToPopulated(Relation relation)
{
	Page        page;

	Assert(relation->rd_rel->relkind == RELKIND_MATVIEW);
	Assert(relation->rd_ispopulated == false);

	page = (Page) palloc(BLCKSZ);
	PageInit(page, BLCKSZ, 0);

	if (RelationNeedsWAL(relation))
		log_newpage(&(relation->rd_node), MAIN_FORKNUM, 0, page);

	RelationOpenSmgr(relation);

	PageSetChecksumInplace(page, 0);
	smgrextend(relation->rd_smgr, MAIN_FORKNUM, 0, (char *) page, true);

	pfree(page);

	smgrimmedsync(relation->rd_smgr, MAIN_FORKNUM);

	RelationCacheInvalidateEntry(relation->rd_id);
}

/*
 * Remove the visibility map fork for a relation.  If there turn out to be
 * any bugs in the visibility map code that require rebuilding the VM, this
 * provides users with a way to do it that is cleaner than shutting down the
 * server and removing files by hand.
 *
 * This is a cut-down version of RelationTruncate.
 */
Datum
pg_truncate_visibility_map(PG_FUNCTION_ARGS)
{
    Oid			relid = PG_GETARG_OID(0);
    Relation	rel;

    rel = relation_open(relid, AccessExclusiveLock);

    if (rel->rd_rel->relkind != RELKIND_RELATION &&
            rel->rd_rel->relkind != RELKIND_MATVIEW &&
            rel->rd_rel->relkind != RELKIND_TOASTVALUE)
        ereport(ERROR,
                (errcode(ERRCODE_WRONG_OBJECT_TYPE),
                 errmsg("\"%s\" is not a table, materialized view, or TOAST table",
                        RelationGetRelationName(rel))));

    RelationOpenSmgr(rel);
    rel->rd_smgr->smgr_vm_nblocks = InvalidBlockNumber;

    visibilitymap_truncate(rel, 0);

    if (RelationNeedsWAL(rel))
    {
        xl_smgr_truncate xlrec;

        xlrec.blkno = 0;
        xlrec.rnode = rel->rd_node;
        xlrec.flags = SMGR_TRUNCATE_VM;

        XLogBeginInsert();
        XLogRegisterData((char *) &xlrec, sizeof(xlrec));

        XLogInsert(RM_SMGR_ID, XLOG_SMGR_TRUNCATE | XLR_SPECIAL_REL_UPDATE);
    }

    /*
     * Release the lock right away, not at commit time.
     *
     * It would be a problem to release the lock prior to commit if this
     * truncate operation sends any transactional invalidation messages. Other
     * backends would potentially be able to lock the relation without
     * processing them in the window of time between when we release the lock
     * here and when we sent the messages at our eventual commit.  However,
     * we're currently only sending a non-transactional smgr invalidation,
     * which will have been posted to shared memory immediately from within
     * visibilitymap_truncate.  Therefore, there should be no race here.
     *
     * The reason why it's desirable to release the lock early here is because
     * of the possibility that someone will need to use this to blow away many
     * visibility map forks at once.  If we can't release the lock until
     * commit time, the transaction doing this will accumulate
     * AccessExclusiveLocks on all of those relations at the same time, which
     * is undesirable. However, if this turns out to be unsafe we may have no
     * choice...
     */
    relation_close(rel, AccessExclusiveLock);

    /* Nothing to return. */
    PG_RETURN_VOID();
}

/*
 * Ensure that the visibility map fork is at least vm_nblocks long, extending
 * it if necessary with zeroed pages.
 */
static void
vm_extend(Relation rel, BlockNumber vm_nblocks)
{
	BlockNumber vm_nblocks_now;
	Page		pg;

	pg = (Page) palloc(BLCKSZ);
	PageInit(pg, BLCKSZ, 0);

	/*
	 * We use the relation extension lock to lock out other backends trying to
	 * extend the visibility map at the same time. It also locks out extension
	 * of the main fork, unnecessarily, but extending the visibility map
	 * happens seldom enough that it doesn't seem worthwhile to have a
	 * separate lock tag type for it.
	 *
	 * Note that another backend might have extended or created the relation
	 * by the time we get the lock.
	 */
	LockRelationForExtension(rel, ExclusiveLock);

	/* Might have to re-open if a cache flush happened */
	RelationOpenSmgr(rel);

	/*
	 * Create the file first if it doesn't exist.  If smgr_vm_nblocks is
	 * positive then it must exist, no need for an smgrexists call.
	 */
	if ((rel->rd_smgr->smgr_vm_nblocks == 0 ||
		 rel->rd_smgr->smgr_vm_nblocks == InvalidBlockNumber) &&
		!smgrexists(rel->rd_smgr, VISIBILITYMAP_FORKNUM))
		smgrcreate(rel->rd_smgr, VISIBILITYMAP_FORKNUM, false);

	vm_nblocks_now = smgrnblocks(rel->rd_smgr, VISIBILITYMAP_FORKNUM);

	/* Now extend the file */
	while (vm_nblocks_now < vm_nblocks)
	{
		smgrextend(rel->rd_smgr, VISIBILITYMAP_FORKNUM, vm_nblocks_now,
				   (char *) pg, false);
		vm_nblocks_now++;
	}

	/*
	 * Send a shared-inval message to force other backends to close any smgr
	 * references they may have for this rel, which we are about to change.
	 * This is a useful optimization because it means that backends don't have
	 * to keep checking for creation or extension of the file, which happens
	 * infrequently.
	 */
	CacheInvalidateSmgr(rel->rd_smgr->smgr_rnode);

	/* Update local cache with the up-to-date size */
	rel->rd_smgr->smgr_vm_nblocks = vm_nblocks_now;

	UnlockRelationForExtension(rel, ExclusiveLock);

	pfree(pg);
}

/*
 * FreeSpaceMapTruncateRel - adjust for truncation of a relation.
 *
 * The caller must hold AccessExclusiveLock on the relation, to ensure that
 * other backends receive the smgr invalidation event that this function sends
 * before they access the FSM again.
 *
 * nblocks is the new___ size of the heap.
 */
void
FreeSpaceMapTruncateRel(Relation rel, BlockNumber nblocks)
{
	BlockNumber new_nfsmblocks;
	FSMAddress	first_removed_address;
	uint16		first_removed_slot;
	Buffer		buf;

	RelationOpenSmgr(rel);

	/*
	 * If no FSM has been created yet for this relation, there's nothing to
	 * truncate.
	 */
	if (!smgrexists(rel->rd_smgr, FSM_FORKNUM))
		return;

	/* Get the location in the FSM of the first removed heap block */
	first_removed_address = fsm_get_location(nblocks, &first_removed_slot);

	/*
	 * Zero out the tail of the last remaining FSM page. If the slot
	 * representing the first removed heap block is at a page boundary, as the
	 * first slot on the FSM page that first_removed_address points to, we can
	 * just truncate that page altogether.
	 */
	if (first_removed_slot > 0)
	{
		buf = fsm_readbuf(rel, first_removed_address, false);
		if (!BufferIsValid(buf))
			return;				/* nothing to do; the FSM was already smaller */
		LockBuffer(buf, BUFFER_LOCK_EXCLUSIVE);
		fsm_truncate_avail(BufferGetPage(buf), first_removed_slot);
		MarkBufferDirtyHint(buf, false);
		UnlockReleaseBuffer(buf);

		new_nfsmblocks = fsm_logical_to_physical(first_removed_address) + 1;
	}
	else
	{
		new_nfsmblocks = fsm_logical_to_physical(first_removed_address);
		if (smgrnblocks(rel->rd_smgr, FSM_FORKNUM) <= new_nfsmblocks)
			return;				/* nothing to do; the FSM was already smaller */
	}

	/* Truncate the unused FSM pages, and send smgr inval message */
	smgrtruncate(rel->rd_smgr, FSM_FORKNUM, new_nfsmblocks);

	/*
	 * We might as well update the local smgr_fsm_nblocks setting.
	 * smgrtruncate sent an smgr cache inval message, which will cause other
	 * backends to invalidate their copy of smgr_fsm_nblocks, and this one too
	 * at the next command boundary.  But this ensures it isn't outright wrong
	 * until then.
	 */
	if (rel->rd_smgr)
		rel->rd_smgr->smgr_fsm_nblocks = new_nfsmblocks;
}

/*
 * emit a completed btree page, and release the working storage.
 */
static void
_bt_blwritepage(BTWriteState *wstate, Page page, BlockNumber blkno)
{
	/* Ensure rd_smgr is open (could have been closed by relcache flush!) */
	RelationOpenSmgr(wstate->index);

	/* XLOG stuff */
	if (wstate->btws_use_wal)
	{
		/* We use the heap NEWPAGE record type for this */
		log_newpage(&wstate->index->rd_node, MAIN_FORKNUM, blkno, page);
	}
	else
	{
		/* Leave the page LSN zero if not WAL-logged, but set TLI anyway */
		PageSetTLI(page, ThisTimeLineID);
	}

	/*
	 * If we have to write pages nonsequentially, fill in the space with
	 * zeroes until we come back and overwrite.  This is not logically
	 * necessary on standard Unix filesystems (unwritten space will read as
	 * zeroes anyway), but it should help to avoid fragmentation. The dummy
	 * pages aren't WAL-logged though.
	 */
	while (blkno > wstate->btws_pages_written)
	{
		if (!wstate->btws_zeropage)
			wstate->btws_zeropage = (Page) palloc0(BLCKSZ);
		smgrextend(wstate->index->rd_smgr, MAIN_FORKNUM,
				   wstate->btws_pages_written++,
				   (char *) wstate->btws_zeropage,
				   true);
	}

	/*
	 * Now write the page.	We say isTemp = true even if it's not a temp
	 * index, because there's no need for smgr to schedule an fsync for this
	 * write; we'll do it ourselves before ending the build.
	 */
	if (blkno == wstate->btws_pages_written)
	{
		/* extending the file... */
		smgrextend(wstate->index->rd_smgr, MAIN_FORKNUM, blkno,
				   (char *) page, true);
		wstate->btws_pages_written++;
	}
	else
	{
		/* overwriting a block we zero-filled before */
		smgrwrite(wstate->index->rd_smgr, MAIN_FORKNUM, blkno,
				  (char *) page, true);
	}

	pfree(page);
}

/*
 * End a rewrite.
 *
 * state and any other resources are freed.
 */
void
end_heap_rewrite(RewriteState state)
{
	HASH_SEQ_STATUS seq_status;
	UnresolvedTup unresolved;

	/*
	 * Write any remaining tuples in the UnresolvedTups table. If we have any
	 * left, they should in fact be dead, but let's err on the safe side.
	 */
	hash_seq_init(&seq_status, state->rs_unresolved_tups);

	while ((unresolved = hash_seq_search(&seq_status)) != NULL)
	{
		ItemPointerSetInvalid(&unresolved->tuple->t_data->t_ctid);
		raw_heap_insert(state, unresolved->tuple);
	}

	/* Write the last page, if any */
	if (state->rs_buffer_valid)
	{
		if (state->rs_use_wal)
			log_newpage(&state->rs_new_rel->rd_node,
						MAIN_FORKNUM,
						state->rs_blockno,
						state->rs_buffer,
						true);
		RelationOpenSmgr(state->rs_new_rel);

		PageSetChecksumInplace(state->rs_buffer, state->rs_blockno);

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

	/*
	 * If the rel is WAL-logged, must fsync before commit.	We use heap_sync
	 * to ensure that the toast table gets fsync'd too.
	 *
	 * It's obvious that we must do this when not WAL-logging. It's less
	 * obvious that we have to do it even if we did WAL-log the pages. The
	 * reason is the same as in tablecmds.c's copy_relation_data(): we're
	 * writing data that's not in shared buffers, and so a CHECKPOINT
	 * occurring during the rewriteheap operation won't have fsync'd data we
	 * wrote before the checkpoint.
	 */
	if (RelationNeedsWAL(state->rs_new_rel))
		heap_sync(state->rs_new_rel);

	/* Deleting the context frees everything */
	MemoryContextDelete(state->rs_cxt);
}

/*
 * Ensure that the FSM fork is at least fsm_nblocks long, extending
 * it if necessary with empty pages. And by empty, I mean pages filled
 * with zeros, meaning there's no free space.
 */
static void
fsm_extend(Relation rel, BlockNumber fsm_nblocks)
{
	BlockNumber fsm_nblocks_now;
	Page		pg;

	pg = (Page) palloc(BLCKSZ);
	PageInit(pg, BLCKSZ, 0);

	/*
	 * We use the relation extension lock to lock out other backends trying to
	 * extend the FSM at the same time. It also locks out extension of the
	 * main fork, unnecessarily, but extending the FSM happens seldom enough
	 * that it doesn't seem worthwhile to have a separate lock tag type for
	 * it.
	 *
	 * Note that another backend might have extended or created the relation
	 * by the time we get the lock.
	 */
	LockRelationForExtension(rel, ExclusiveLock);

	/* Might have to re-open if a cache flush happened */
	RelationOpenSmgr(rel);

	/*
	 * Create the FSM file first if it doesn't exist.  If smgr_fsm_nblocks is
	 * positive then it must exist, no need for an smgrexists call.
	 */
	if ((rel->rd_smgr->smgr_fsm_nblocks == 0 ||
		 rel->rd_smgr->smgr_fsm_nblocks == InvalidBlockNumber) &&
		!smgrexists(rel->rd_smgr, FSM_FORKNUM))
		smgrcreate(rel->rd_smgr, FSM_FORKNUM, false);

	fsm_nblocks_now = smgrnblocks(rel->rd_smgr, FSM_FORKNUM);

	while (fsm_nblocks_now < fsm_nblocks)
	{
		PageSetChecksumInplace(pg, fsm_nblocks_now);

		smgrextend(rel->rd_smgr, FSM_FORKNUM, fsm_nblocks_now,
				   (char *) pg, false);
		fsm_nblocks_now++;
	}

	/* Update local cache with the up-to-date size */
	rel->rd_smgr->smgr_fsm_nblocks = fsm_nblocks_now;

	UnlockRelationForExtension(rel, ExclusiveLock);

	pfree(pg);
}

/*
 * For heaps, we prevent creation of the FSM unless the number of pages
 * exceeds HEAP_FSM_CREATION_THRESHOLD.  For tables that don't already have
 * a FSM, this will save an inode and a few kB of space.
 *
 * XXX The API is a little awkward -- if the caller passes a valid nblocks
 * value, it can avoid invoking a system call.  If the caller passes
 * InvalidBlockNumber and receives a false return value, it can get an
 * up-to-date relation size from get_nblocks.  This saves a few cycles in
 * the caller, which would otherwise need to get the relation size by itself.
 */
static bool
fsm_allow_writes(Relation rel, BlockNumber heapblk,
				 BlockNumber nblocks, BlockNumber *get_nblocks)
{
	bool		skip_get_nblocks;

	if (heapblk >= HEAP_FSM_CREATION_THRESHOLD)
		return true;

	/* Non-heap rels can always create a FSM. */
	if (rel->rd_rel->relkind != RELKIND_RELATION &&
		rel->rd_rel->relkind != RELKIND_TOASTVALUE)
		return true;

	/*
	 * If the caller knows nblocks, we can avoid a system call later. If it
	 * doesn't, maybe we have relpages from a previous VACUUM. Since the table
	 * may have extended since then, we still have to count the pages later if
	 * we can't return now.
	 */
	if (nblocks != InvalidBlockNumber)
	{
		if (nblocks > HEAP_FSM_CREATION_THRESHOLD)
			return true;
		else
			skip_get_nblocks = true;
	}
	else
	{
		if (rel->rd_rel->relpages != InvalidBlockNumber &&
			rel->rd_rel->relpages > HEAP_FSM_CREATION_THRESHOLD)
			return true;
		else
			skip_get_nblocks = false;
	}

	RelationOpenSmgr(rel);
	if (smgrexists(rel->rd_smgr, FSM_FORKNUM))
		return true;

	if (skip_get_nblocks)
		return false;

	/* last resort */
	*get_nblocks = RelationGetNumberOfBlocks(rel);
	if (*get_nblocks > HEAP_FSM_CREATION_THRESHOLD)
		return true;
	else
		return false;
}

/*
 * Read a visibility map page.
 *
 * If the page doesn't exist, InvalidBuffer is returned, or if 'extend' is
 * true, the visibility map file is extended.
 */
static Buffer
vm_readbuf(Relation rel, BlockNumber blkno, bool extend)
{
	Buffer		buf;

	/*
	 * We might not have opened the relation at the smgr level yet, or we
	 * might have been forced to close it by a sinval message.  The code below
	 * won't necessarily notice relation extension immediately when extend =
	 * false, so we rely on sinval messages to ensure that our ideas about the
	 * size of the map aren't too far out of date.
	 */
	RelationOpenSmgr(rel);

	/*
	 * If we haven't cached the size of the visibility map fork yet, check it
	 * first.
	 */
	if (rel->rd_smgr->smgr_vm_nblocks == InvalidBlockNumber)
	{
		if (smgrexists(rel->rd_smgr, VISIBILITYMAP_FORKNUM))
			rel->rd_smgr->smgr_vm_nblocks = smgrnblocks(rel->rd_smgr,
													  VISIBILITYMAP_FORKNUM);
		else
			rel->rd_smgr->smgr_vm_nblocks = 0;
	}

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

	/*
	 * Use ZERO_ON_ERROR mode, and initialize the page if necessary. It's
	 * always safe to clear bits, so it's better to clear corrupt pages than
	 * error out.
	 */
	buf = ReadBufferExtended(rel, VISIBILITYMAP_FORKNUM, blkno,
							 RBM_ZERO_ON_ERROR, NULL);
	if (PageIsNew(BufferGetPage(buf)))
		PageInit(BufferGetPage(buf), BLCKSZ, 0);
	return buf;
}

/*
 * _hash_alloc_buckets -- allocate a new splitpoint's worth of bucket pages
 *
 * This does not need to initialize the new bucket pages; we'll do that as
 * each one is used by _hash_expandtable().  But we have to extend the logical
 * EOF to the end of the splitpoint; this keeps smgr's idea of the EOF in
 * sync with ours, so that we don't get complaints from smgr.
 *
 * We do this by writing a page of zeroes at the end of the splitpoint range.
 * We expect that the filesystem will ensure that the intervening pages read
 * as zeroes too.  On many filesystems this "hole" will not be allocated
 * immediately, which means that the index file may end up more fragmented
 * than if we forced it all to be allocated now; but since we don't scan
 * hash indexes sequentially anyway, that probably doesn't matter.
 *
 * XXX It's annoying that this code is executed with the metapage lock held.
 * We need to interlock against _hash_addovflpage() adding a new overflow page
 * concurrently, but it'd likely be better to use LockRelationForExtension
 * for the purpose.  OTOH, adding a splitpoint is a very infrequent operation,
 * so it may not be worth worrying about.
 *
 * Returns TRUE if successful, or FALSE if allocation failed due to
 * BlockNumber overflow.
 */
static bool
_hash_alloc_buckets(Relation rel, BlockNumber firstblock, uint32 nblocks)
{
	BlockNumber lastblock;
	char		zerobuf[BLCKSZ];
	Page		page;
	HashPageOpaque ovflopaque;

	lastblock = firstblock + nblocks - 1;

	/*
	 * Check for overflow in block number calculation; if so, we cannot extend
	 * the index anymore.
	 */
	if (lastblock < firstblock || lastblock == InvalidBlockNumber)
		return false;

	page = (Page) zerobuf;

	/*
	 * Initialize the page.  Just zeroing the page won't work; see
	 * _hash_freeovflpage for similar usage.  We take care to make the special
	 * space valid for the benefit of tools such as pageinspect.
	 */
	_hash_pageinit(page, BLCKSZ);

	ovflopaque = (HashPageOpaque) PageGetSpecialPointer(page);

	ovflopaque->hasho_prevblkno = InvalidBlockNumber;
	ovflopaque->hasho_nextblkno = InvalidBlockNumber;
	ovflopaque->hasho_bucket = -1;
	ovflopaque->hasho_flag = LH_UNUSED_PAGE;
	ovflopaque->hasho_page_id = HASHO_PAGE_ID;

	if (RelationNeedsWAL(rel))
		log_newpage(&rel->rd_node,
					MAIN_FORKNUM,
					lastblock,
					zerobuf,
					true);

	RelationOpenSmgr(rel);
	smgrextend(rel->rd_smgr, MAIN_FORKNUM, lastblock, zerobuf, false);

	return true;
}

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

	RelationOpenSmgr(rel);

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

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

	/*
	 * Use ZERO_ON_ERROR mode, and initialize the page if necessary. The FSM
	 * information is not accurate anyway, so it's better to clear corrupt
	 * pages than error out. Since the FSM changes are not WAL-logged, the
	 * so-called torn page problem on crash can lead to pages with corrupt
	 * headers, for example.
	 */
	buf = ReadBufferExtended(rel, FSM_FORKNUM, blkno, RBM_ZERO_ON_ERROR, NULL);
	if (PageIsNew(BufferGetPage(buf)))
		PageInit(BufferGetPage(buf), BLCKSZ, 0);
	return buf;
}

/*
 * Read a visibility map page.
 *
 * If the page doesn't exist, InvalidBuffer is returned, or if 'extend' is
 * true, the visibility map file is extended.
 */
static Buffer
vm_readbuf(Relation rel, BlockNumber blkno, bool extend)
{
	Buffer		buf;

	RelationOpenSmgr(rel);

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

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

	/*
	 * Use ZERO_ON_ERROR mode, and initialize the page if necessary. It's
	 * always safe to clear bits, so it's better to clear corrupt pages than
	 * error out.
	 */
	buf = ReadBufferExtended(rel, VISIBILITYMAP_FORKNUM, blkno,
							 RBM_ZERO_ON_ERROR, NULL);
	if (PageIsNew(BufferGetPage(buf)))
		PageInit(BufferGetPage(buf), BLCKSZ, 0);
	return buf;
}

/*
 * Read a visibility map page.
 *
 * If the page doesn't exist, InvalidBuffer is returned, or if 'extend' is
 * true, the visibility map file is extended.
 */
static Buffer
vm_readbuf(Relation rel, BlockNumber blkno, bool extend)
{
	Buffer		buf;

	RelationOpenSmgr(rel);

	/*
	 * The current size of the visibility map fork is kept in relcache, to
	 * avoid reading beyond EOF. If we haven't cached the size of the map yet,
	 * do that first.
	 */
	if (rel->rd_vm_nblocks == InvalidBlockNumber)
	{
		if (smgrexists(rel->rd_smgr, VISIBILITYMAP_FORKNUM))
			rel->rd_vm_nblocks = smgrnblocks(rel->rd_smgr,
											 VISIBILITYMAP_FORKNUM);
		else
			rel->rd_vm_nblocks = 0;
	}

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

	/*
	 * Use ZERO_ON_ERROR mode, and initialize the page if necessary. It's
	 * always safe to clear bits, so it's better to clear corrupt pages than
	 * error out.
	 */
	buf = ReadBufferExtended(rel, VISIBILITYMAP_FORKNUM, blkno,
							 RBM_ZERO_ON_ERROR, NULL);
	if (PageIsNew(BufferGetPage(buf)))
		PageInit(BufferGetPage(buf), BLCKSZ, 0);
	return buf;
}

/*
 * _hash_alloc_buckets -- allocate a new splitpoint's worth of bucket pages
 *
 * This does not need to initialize the new bucket pages; we'll do that as
 * each one is used by _hash_expandtable().  But we have to extend the logical
 * EOF to the end of the splitpoint; this keeps smgr's idea of the EOF in
 * sync with ours, so that we don't get complaints from smgr.
 *
 * We do this by writing a page of zeroes at the end of the splitpoint range.
 * We expect that the filesystem will ensure that the intervening pages read
 * as zeroes too.  On many filesystems this "hole" will not be allocated
 * immediately, which means that the index file may end up more fragmented
 * than if we forced it all to be allocated now; but since we don't scan
 * hash indexes sequentially anyway, that probably doesn't matter.
 *
 * XXX It's annoying that this code is executed with the metapage lock held.
 * We need to interlock against _hash_getovflpage() adding a new overflow page
 * concurrently, but it'd likely be better to use LockRelationForExtension
 * for the purpose.  OTOH, adding a splitpoint is a very infrequent operation,
 * so it may not be worth worrying about.
 *
 * Returns TRUE if successful, or FALSE if allocation failed due to
 * BlockNumber overflow.
 */
static bool
_hash_alloc_buckets(Relation rel, BlockNumber firstblock, uint32 nblocks)
{
    BlockNumber lastblock;
    char		zerobuf[BLCKSZ];

    lastblock = firstblock + nblocks - 1;

    /*
     * Check for overflow in block number calculation; if so, we cannot extend
     * the index anymore.
     */
    if (lastblock < firstblock || lastblock == InvalidBlockNumber)
        return false;

    MemSet(zerobuf, 0, sizeof(zerobuf));

    RelationOpenSmgr(rel);
    smgrextend(rel->rd_smgr, MAIN_FORKNUM, lastblock, zerobuf, false);

    return true;
}

/*
 * _hash_alloc_buckets -- allocate a new splitpoint's worth of bucket pages
 *
 * This does not need to initialize the new bucket pages; we'll do that as
 * each one is used by _hash_expandtable().  But we have to extend the logical
 * EOF to the end of the splitpoint; otherwise the first overflow page
 * allocated beyond the splitpoint will represent a noncontiguous access,
 * which can confuse md.c (and will probably be forbidden by future changes
 * to md.c).
 *
 * We do this by writing a page of zeroes at the end of the splitpoint range.
 * We expect that the filesystem will ensure that the intervening pages read
 * as zeroes too.  On many filesystems this "hole" will not be allocated
 * immediately, which means that the index file may end up more fragmented
 * than if we forced it all to be allocated now; but since we don't scan
 * hash indexes sequentially anyway, that probably doesn't matter.
 *
 * XXX It's annoying that this code is executed with the metapage lock held.
 * We need to interlock against _hash_getovflpage() adding a new overflow page
 * concurrently, but it'd likely be better to use LockRelationForExtension
 * for the purpose.  OTOH, adding a splitpoint is a very infrequent operation,
 * so it may not be worth worrying about.
 *
 * Returns TRUE if successful, or FALSE if allocation failed due to
 * BlockNumber overflow.
 */
static bool
_hash_alloc_buckets(Relation rel, BlockNumber firstblock, uint32 nblocks)
{
	BlockNumber	lastblock;
	BlockNumber	endblock;
	char		zerobuf[BLCKSZ];

	lastblock = firstblock + nblocks - 1;

	/*
	 * Check for overflow in block number calculation; if so, we cannot
	 * extend the index anymore.
	 */
	if (lastblock < firstblock || lastblock == InvalidBlockNumber)
		return false;

	MemSet(zerobuf, 0, sizeof(zerobuf));

	RelationOpenSmgr(rel);

	/*
	 * XXX If the extension results in creation of new segment files,
	 * we have to make sure that each non-last file is correctly filled out to
	 * RELSEG_SIZE blocks.  This ought to be done inside mdextend, but
	 * changing the smgr API seems best left for development cycle not late
	 * beta.  Temporary fix for bug #2737.
	 */
#ifndef LET_OS_MANAGE_FILESIZE
	for (endblock = firstblock | (RELSEG_SIZE - 1);
		 endblock < lastblock;
		 endblock += RELSEG_SIZE)
		smgrextend(rel->rd_smgr, endblock, zerobuf, rel->rd_istemp);
#endif

	smgrextend(rel->rd_smgr, lastblock, zerobuf, rel->rd_istemp);

	return true;
}