/*
* Update index metapage's lastUsedPages info from local cache, if possible
*
* Updating meta page isn't critical for index working, so
* 1 use ConditionalLockBuffer to improve concurrency
* 2 don't WAL-log metabuffer changes to decrease WAL traffic
*/
void
SpGistUpdateMetaPage(Relation index)
{
SpGistCache *cache = (SpGistCache *) index->rd_amcache;
if (cache != NULL)
{
Buffer metabuffer;
SpGistMetaPageData *metadata;
metabuffer = ReadBuffer(index, SPGIST_METAPAGE_BLKNO);
if (ConditionalLockBuffer(metabuffer))
{
metadata = SpGistPageGetMeta(BufferGetPage(metabuffer));
metadata->lastUsedPages = cache->lastUsedPages;
MarkBufferDirty(metabuffer);
UnlockReleaseBuffer(metabuffer);
}
else
{
ReleaseBuffer(metabuffer);
}
}
}
/*
* Allocate a new page (either by recycling, or by extending the index file).
*
* The returned buffer is already pinned and exclusive-locked.
* Caller is responsible for initializing the page by calling SpGistInitBuffer.
*/
Buffer
SpGistNewBuffer(Relation index)
{
Buffer buffer;
bool needLock;
/* First, try to get a page from FSM */
for (;;)
{
BlockNumber blkno = GetFreeIndexPage(index);
if (blkno == InvalidBlockNumber)
break; /* nothing known to FSM */
/*
* The fixed pages shouldn't ever be listed in FSM, but just in case
* one is, ignore it.
*/
if (SpGistBlockIsFixed(blkno))
continue;
buffer = ReadBuffer(index, blkno);
/*
* We have to guard against the possibility that someone else already
* recycled this page; the buffer may be locked if so.
*/
if (ConditionalLockBuffer(buffer))
{
Page page = BufferGetPage(buffer);
if (PageIsNew(page))
return buffer; /* OK to use, if never initialized */
if (SpGistPageIsDeleted(page) || PageIsEmpty(page))
return buffer; /* OK to use */
LockBuffer(buffer, BUFFER_LOCK_UNLOCK);
}
/* Can't use it, so release buffer and try again */
ReleaseBuffer(buffer);
}
/* Must extend the file */
needLock = !RELATION_IS_LOCAL(index);
if (needLock)
LockRelationForExtension(index, ExclusiveLock);
buffer = ReadBuffer(index, P_NEW);
LockBuffer(buffer, BUFFER_LOCK_EXCLUSIVE);
if (needLock)
UnlockRelationForExtension(index, ExclusiveLock);
return buffer;
}
/*
* Allocate a new page (either by recycling, or by extending the index file)
*
* The returned buffer is already pinned and exclusive-locked
*
* Caller is responsible for initializing the page by calling GISTInitBuffer
*/
Buffer
gistNewBuffer(Relation r)
{
Buffer buffer;
bool needLock;
MIRROREDLOCK_BUFMGR_MUST_ALREADY_BE_HELD;
/* First, try to get a page from FSM */
for (;;)
{
BlockNumber blkno = GetFreeIndexPage(&r->rd_node);
if (blkno == InvalidBlockNumber)
break; /* nothing left in FSM */
buffer = ReadBuffer(r, blkno);
/*
* We have to guard against the possibility that someone else already
* recycled this page; the buffer may be locked if so.
*/
if (ConditionalLockBuffer(buffer))
{
Page page = BufferGetPage(buffer);
if (PageIsNew(page))
return buffer; /* OK to use, if never initialized */
gistcheckpage(r, buffer);
if (GistPageIsDeleted(page))
return buffer; /* OK to use */
LockBuffer(buffer, GIST_UNLOCK);
}
/* Can't use it, so release buffer and try again */
ReleaseBuffer(buffer);
}
/* Must extend the file */
needLock = !RELATION_IS_LOCAL(r);
if (needLock)
LockRelationForExtension(r, ExclusiveLock);
buffer = ReadBuffer(r, P_NEW);
LockBuffer(buffer, GIST_EXCLUSIVE);
if (needLock)
UnlockRelationForExtension(r, ExclusiveLock);
return buffer;
}
Buffer
GinNewBuffer(Relation index)
{
Buffer buffer;
bool needLock;
/* First, try to get a page from FSM */
for (;;)
{
BlockNumber blkno = GetFreeIndexPage(index);
if (blkno == InvalidBlockNumber)
break;
buffer = ReadBuffer(index, blkno);
/*
* We have to guard against the possibility that someone else already
* recycled this page; the buffer may be locked if so.
*/
if (ConditionalLockBuffer(buffer))
{
Page page = BufferGetPage(buffer);
if (PageIsNew(page))
return buffer; /* OK to use, if never initialized */
if (GinPageIsDeleted(page))
return buffer; /* OK to use */
LockBuffer(buffer, GIN_UNLOCK);
}
/* Can't use it, so release buffer and try again */
ReleaseBuffer(buffer);
}
/* Must extend the file */
needLock = !RELATION_IS_LOCAL(index);
if (needLock)
LockRelationForExtension(index, ExclusiveLock);
buffer = ReadBuffer(index, P_NEW);
LockBuffer(buffer, GIN_EXCLUSIVE);
if (needLock)
UnlockRelationForExtension(index, ExclusiveLock);
return buffer;
}
/*
* _bt_getbuf() -- Get a buffer by block number for read or write.
*
* blkno == P_NEW means to get an unallocated index page. The page
* will be initialized before returning it.
*
* When this routine returns, the appropriate lock is set on the
* requested buffer and its reference count has been incremented
* (ie, the buffer is "locked and pinned"). Also, we apply
* _bt_checkpage to sanity-check the page (except in P_NEW case).
*/
Buffer
_bt_getbuf(Relation rel, BlockNumber blkno, int access)
{
Buffer buf;
MIRROREDLOCK_BUFMGR_MUST_ALREADY_BE_HELD;
if (blkno != P_NEW)
{
/* Read an existing block of the relation */
buf = ReadBuffer(rel, blkno);
LockBuffer(buf, access);
_bt_checkpage(rel, buf);
}
else
{
bool needLock;
Page page;
Assert(access == BT_WRITE);
/*
* First see if the FSM knows of any free pages.
*
* We can't trust the FSM's report unreservedly; we have to check that
* the page is still free. (For example, an already-free page could
* have been re-used between the time the last VACUUM scanned it and
* the time the VACUUM made its FSM updates.)
*
* In fact, it's worse than that: we can't even assume that it's safe
* to take a lock on the reported page. If somebody else has a lock
* on it, or even worse our own caller does, we could deadlock. (The
* own-caller scenario is actually not improbable. Consider an index
* on a serial or timestamp column. Nearly all splits will be at the
* rightmost page, so it's entirely likely that _bt_split will call us
* while holding a lock on the page most recently acquired from FSM. A
* VACUUM running concurrently with the previous split could well have
* placed that page back in FSM.)
*
* To get around that, we ask for only a conditional lock on the
* reported page. If we fail, then someone else is using the page,
* and we may reasonably assume it's not free. (If we happen to be
* wrong, the worst consequence is the page will be lost to use till
* the next VACUUM, which is no big problem.)
*/
for (;;)
{
blkno = GetFreeIndexPage(&rel->rd_node);
if (blkno == InvalidBlockNumber)
break;
buf = ReadBuffer(rel, blkno);
if (ConditionalLockBuffer(buf))
{
page = BufferGetPage(buf);
if (_bt_page_recyclable(page))
{
/* Okay to use page. Re-initialize and return it */
_bt_pageinit(page, BufferGetPageSize(buf));
return buf;
}
elog(DEBUG2, "FSM returned nonrecyclable page");
_bt_relbuf(rel, buf);
}
else
{
elog(DEBUG2, "FSM returned nonlockable page");
/* couldn't get lock, so just drop pin */
ReleaseBuffer(buf);
}
}
/*
* Extend the relation by one page.
*
* We have to use a lock to ensure no one else is extending the rel at
* the same time, else we will both try to initialize the same new
* page. We can skip locking for new or temp relations, however,
* since no one else could be accessing them.
*/
needLock = !RELATION_IS_LOCAL(rel);
if (needLock)
LockRelationForExtension(rel, ExclusiveLock);
buf = ReadBuffer(rel, P_NEW);
/* Acquire buffer lock on new page */
LockBuffer(buf, BT_WRITE);
/*
* Release the file-extension lock; it's now OK for someone else to
* extend the relation some more. Note that we cannot release this
* lock before we have buffer lock on the new page, or we risk a race
* condition against btvacuumscan --- see comments therein.
*/
if (needLock)
UnlockRelationForExtension(rel, ExclusiveLock);
/* Initialize the new page before returning it */
page = BufferGetPage(buf);
Assert(PageIsNew((PageHeader) page));
_bt_pageinit(page, BufferGetPageSize(buf));
}
/* ref count and lock type are correct */
return buf;
}
/*
* Get a buffer of the type and parity specified by flags, having at least
* as much free space as indicated by needSpace. We use the lastUsedPages
* cache to assign the same buffer previously requested when possible.
* The returned buffer is already pinned and exclusive-locked.
*
* *isNew is set true if the page was initialized here, false if it was
* already valid.
*/
Buffer
SpGistGetBuffer(Relation index, int flags, int needSpace, bool *isNew)
{
SpGistCache *cache = spgGetCache(index);
SpGistLastUsedPage *lup;
/* Bail out if even an empty page wouldn't meet the demand */
if (needSpace > SPGIST_PAGE_CAPACITY)
elog(ERROR, "desired SPGiST tuple size is too big");
/*
* If possible, increase the space request to include relation's
* fillfactor. This ensures that when we add unrelated tuples to a page,
* we try to keep 100-fillfactor% available for adding tuples that are
* related to the ones already on it. But fillfactor mustn't cause an
* error for requests that would otherwise be legal.
*/
needSpace += RelationGetTargetPageFreeSpace(index,
SPGIST_DEFAULT_FILLFACTOR);
needSpace = Min(needSpace, SPGIST_PAGE_CAPACITY);
/* Get the cache entry for this flags setting */
lup = GET_LUP(cache, flags);
/* If we have nothing cached, just turn it over to allocNewBuffer */
if (lup->blkno == InvalidBlockNumber)
{
*isNew = true;
return allocNewBuffer(index, flags);
}
/* fixed pages should never be in cache */
Assert(!SpGistBlockIsFixed(lup->blkno));
/* If cached freeSpace isn't enough, don't bother looking at the page */
if (lup->freeSpace >= needSpace)
{
Buffer buffer;
Page page;
buffer = ReadBuffer(index, lup->blkno);
if (!ConditionalLockBuffer(buffer))
{
/*
* buffer is locked by another process, so return a new buffer
*/
ReleaseBuffer(buffer);
*isNew = true;
return allocNewBuffer(index, flags);
}
page = BufferGetPage(buffer);
if (PageIsNew(page) || SpGistPageIsDeleted(page) || PageIsEmpty(page))
{
/* OK to initialize the page */
uint16 pageflags = 0;
if (GBUF_REQ_LEAF(flags))
pageflags |= SPGIST_LEAF;
if (GBUF_REQ_NULLS(flags))
pageflags |= SPGIST_NULLS;
SpGistInitBuffer(buffer, pageflags);
lup->freeSpace = PageGetExactFreeSpace(page) - needSpace;
*isNew = true;
return buffer;
}
/*
* Check that page is of right type and has enough space. We must
* recheck this since our cache isn't necessarily up to date.
*/
if ((GBUF_REQ_LEAF(flags) ? SpGistPageIsLeaf(page) : !SpGistPageIsLeaf(page)) &&
(GBUF_REQ_NULLS(flags) ? SpGistPageStoresNulls(page) : !SpGistPageStoresNulls(page)))
{
int freeSpace = PageGetExactFreeSpace(page);
if (freeSpace >= needSpace)
{
/* Success, update freespace info and return the buffer */
lup->freeSpace = freeSpace - needSpace;
*isNew = false;
return buffer;
}
}
/*
* fallback to allocation of new buffer
*/
UnlockReleaseBuffer(buffer);
}
/* No success with cache, so return a new buffer */
*isNew = true;
return allocNewBuffer(index, flags);
}