/*
* Delete tuple on leaf page if tuples was existed and we
* should update it, update old child blkno to new right page
* if child split is occured
*/
static BlockNumber
entryPreparePage(GinBtree btree, Page page, OffsetNumber off)
{
BlockNumber ret = InvalidBlockNumber;
Assert(btree->entry);
Assert(!GinPageIsData(page));
if (btree->isDelete)
{
Assert(GinPageIsLeaf(page));
PageIndexTupleDelete(page, off);
}
if (!GinPageIsLeaf(page) && btree->rightblkno != InvalidBlockNumber)
{
IndexTuple itup = (IndexTuple) PageGetItem(page, PageGetItemId(page, off));
ItemPointerSet(&itup->t_tid, btree->rightblkno, InvalidOffsetNumber);
ret = btree->rightblkno;
}
btree->rightblkno = InvalidBlockNumber;
return ret;
}
/*
* Locks buffer by needed method for search.
*/
static int
ginTraverseLock(Buffer buffer, bool searchMode)
{
Page page;
int access = GIN_SHARE;
LockBuffer(buffer, GIN_SHARE);
page = BufferGetPage(buffer);
if (GinPageIsLeaf(page))
{
if (searchMode == FALSE)
{
/* we should relock our page */
LockBuffer(buffer, GIN_UNLOCK);
LockBuffer(buffer, GIN_EXCLUSIVE);
/* But root can become non-leaf during relock */
if (!GinPageIsLeaf(page))
{
/* restore old lock type (very rare) */
LockBuffer(buffer, GIN_UNLOCK);
LockBuffer(buffer, GIN_SHARE);
}
else
access = GIN_EXCLUSIVE;
}
}
return access;
}
/*
* Step right from current page.
*
* The next page is locked first, before releasing the current page. This is
* crucial to protect from concurrent page deletion (see comment in
* ginDeletePage).
*/
Buffer
ginStepRight(Buffer buffer, Relation index, int lockmode)
{
Buffer nextbuffer;
Page page = BufferGetPage(buffer);
bool isLeaf = GinPageIsLeaf(page);
bool isData = GinPageIsData(page);
BlockNumber blkno = GinPageGetOpaque(page)->rightlink;
nextbuffer = ReadBuffer(index, blkno);
LockBuffer(nextbuffer, lockmode);
UnlockReleaseBuffer(buffer);
/* Sanity check that the page we stepped to is of similar kind. */
page = BufferGetPage(nextbuffer);
if (isLeaf != GinPageIsLeaf(page) || isData != GinPageIsData(page))
elog(ERROR, "right sibling of GIN page is of different type");
/*
* Given the proper lock sequence above, we should never land on a deleted
* page.
*/
if (GinPageIsDeleted(page))
elog(ERROR, "right sibling of GIN page was deleted");
return nextbuffer;
}
static void
ginRedoInsertData(Buffer buffer, bool isLeaf, BlockNumber rightblkno, void *rdata)
{
Page page = BufferGetPage(buffer);
if (isLeaf)
{
ginxlogRecompressDataLeaf *data = (ginxlogRecompressDataLeaf *) rdata;
Assert(GinPageIsLeaf(page));
ginRedoRecompress(page, data);
}
else
{
ginxlogInsertDataInternal *data = (ginxlogInsertDataInternal *) rdata;
PostingItem *oldpitem;
Assert(!GinPageIsLeaf(page));
/* update link to right page after split */
oldpitem = GinDataPageGetPostingItem(page, data->offset);
PostingItemSetBlockNumber(oldpitem, rightblkno);
GinDataPageAddPostingItem(page, &data->newitem, data->offset);
}
}
/*
* add ItemPointer to a leaf page.
*/
void
GinDataPageAddItemPointer(Page page, ItemPointer data, OffsetNumber offset)
{
OffsetNumber maxoff = GinPageGetOpaque(page)->maxoff;
char *ptr;
Assert(ItemPointerIsValid(data));
Assert(GinPageIsLeaf(page));
if (offset == InvalidOffsetNumber)
{
ptr = (char *) GinDataPageGetItemPointer(page, maxoff + 1);
}
else
{
ptr = (char *) GinDataPageGetItemPointer(page, offset);
if (maxoff + 1 - offset != 0)
memmove(ptr + sizeof(ItemPointerData),
ptr,
(maxoff - offset + 1) * sizeof(ItemPointerData));
}
memcpy(ptr, data, sizeof(ItemPointerData));
GinPageGetOpaque(page)->maxoff++;
}
/*
* add PostingItem to a non-leaf page.
*/
void
GinDataPageAddPostingItem(Page page, PostingItem *data, OffsetNumber offset)
{
OffsetNumber maxoff = GinPageGetOpaque(page)->maxoff;
char *ptr;
Assert(PostingItemGetBlockNumber(data) != InvalidBlockNumber);
Assert(!GinPageIsLeaf(page));
if (offset == InvalidOffsetNumber)
{
ptr = (char *) GinDataPageGetPostingItem(page, maxoff + 1);
}
else
{
ptr = (char *) GinDataPageGetPostingItem(page, offset);
if (maxoff + 1 - offset != 0)
memmove(ptr + sizeof(PostingItem),
ptr,
(maxoff - offset + 1) * sizeof(PostingItem));
}
memcpy(ptr, data, sizeof(PostingItem));
GinPageGetOpaque(page)->maxoff++;
}
/*
* Form a non-leaf entry tuple by copying the key data from the given tuple,
* which can be either a leaf or non-leaf entry tuple.
*
* Any posting list in the source tuple is not copied. The specified child
* block number is inserted into t_tid.
*/
static IndexTuple
GinFormInteriorTuple(IndexTuple itup, Page page, BlockNumber childblk)
{
IndexTuple nitup;
if (GinPageIsLeaf(page) && !GinIsPostingTree(itup))
{
/* Tuple contains a posting list, just copy stuff before that */
uint32 origsize = GinGetPostingOffset(itup);
origsize = MAXALIGN(origsize);
nitup = (IndexTuple) palloc(origsize);
memcpy(nitup, itup, origsize);
/* ... be sure to fix the size header field ... */
nitup->t_info &= ~INDEX_SIZE_MASK;
nitup->t_info |= origsize;
}
else
{
/* Copy the tuple as-is */
nitup = (IndexTuple) palloc(IndexTupleSize(itup));
memcpy(nitup, itup, IndexTupleSize(itup));
}
/* Now insert the correct downlink */
GinSetDownlink(nitup, childblk);
return nitup;
}
static void
ginRedoVacuumDataLeafPage(XLogRecPtr lsn, XLogRecord *record)
{
ginxlogVacuumDataLeafPage *xlrec = (ginxlogVacuumDataLeafPage *) XLogRecGetData(record);
Buffer buffer;
Page page;
/* If we have a full-page image, restore it and we're done */
if (record->xl_info & XLR_BKP_BLOCK(0))
{
(void) RestoreBackupBlock(lsn, record, 0, false, false);
return;
}
buffer = XLogReadBuffer(xlrec->node, xlrec->blkno, false);
if (!BufferIsValid(buffer))
return;
page = (Page) BufferGetPage(buffer);
Assert(GinPageIsLeaf(page));
Assert(GinPageIsData(page));
if (lsn > PageGetLSN(page))
{
ginRedoRecompress(page, &xlrec->data);
PageSetLSN(page, lsn);
MarkBufferDirty(buffer);
}
UnlockReleaseBuffer(buffer);
}
/*
* Searches correct position for value on leaf page.
* Page should be correctly chosen.
* Returns true if value found on page.
*/
static bool
entryLocateLeafEntry(GinBtree btree, GinBtreeStack *stack)
{
Page page = BufferGetPage(stack->buffer, NULL, NULL,
BGP_NO_SNAPSHOT_TEST);
OffsetNumber low,
high;
Assert(GinPageIsLeaf(page));
Assert(!GinPageIsData(page));
if (btree->fullScan)
{
stack->off = FirstOffsetNumber;
return TRUE;
}
low = FirstOffsetNumber;
high = PageGetMaxOffsetNumber(page);
if (high < low)
{
stack->off = FirstOffsetNumber;
return false;
}
high++;
while (high > low)
{
OffsetNumber mid = low + ((high - low) / 2);
IndexTuple itup;
OffsetNumber attnum;
Datum key;
GinNullCategory category;
int result;
itup = (IndexTuple) PageGetItem(page, PageGetItemId(page, mid));
attnum = gintuple_get_attrnum(btree->ginstate, itup);
key = gintuple_get_key(btree->ginstate, itup, &category);
result = ginCompareAttEntries(btree->ginstate,
btree->entryAttnum,
btree->entryKey,
btree->entryCategory,
attnum, key, category);
if (result == 0)
{
stack->off = mid;
return true;
}
else if (result > 0)
low = mid + 1;
else
high = mid;
}
stack->off = high;
return false;
}
/*
* Place tuple on page and fills WAL record
*/
static void
entryPlaceToPage(GinBtree btree, Buffer buf, OffsetNumber off, XLogRecData **prdata)
{
Page page = BufferGetPage(buf);
static XLogRecData rdata[3];
OffsetNumber placed;
static ginxlogInsert data;
int cnt = 0;
*prdata = rdata;
data.updateBlkno = entryPreparePage(btree, page, off);
placed = PageAddItem(page, (Item) btree->entry, IndexTupleSize(btree->entry), off, false, false);
if (placed != off)
elog(ERROR, "failed to add item to index page in \"%s\"",
RelationGetRelationName(btree->index));
data.node = btree->index->rd_node;
data.blkno = BufferGetBlockNumber(buf);
data.offset = off;
data.nitem = 1;
data.isDelete = btree->isDelete;
data.isData = false;
data.isLeaf = GinPageIsLeaf(page) ? TRUE : FALSE;
/*
* Prevent full page write if child's split occurs. That is needed to
* remove incomplete splits while replaying WAL
*
* data.updateBlkno contains new block number (of newly created right
* page) for recently splited page.
*/
if (data.updateBlkno == InvalidBlockNumber)
{
rdata[0].buffer = buf;
rdata[0].buffer_std = TRUE;
rdata[0].data = NULL;
rdata[0].len = 0;
rdata[0].next = &rdata[1];
cnt++;
}
rdata[cnt].buffer = InvalidBuffer;
rdata[cnt].data = (char *) &data;
rdata[cnt].len = sizeof(ginxlogInsert);
rdata[cnt].next = &rdata[cnt + 1];
cnt++;
rdata[cnt].buffer = InvalidBuffer;
rdata[cnt].data = (char *) btree->entry;
rdata[cnt].len = IndexTupleSize(btree->entry);
rdata[cnt].next = NULL;
btree->entry = NULL;
}
/*
* Delete tuple on leaf page if tuples existed and we
* should update it, update old child blkno to new right page
* if child split occurred
*/
static void
entryPreparePage(GinBtree btree, Page page, OffsetNumber off,
GinBtreeEntryInsertData *insertData, BlockNumber updateblkno)
{
Assert(insertData->entry);
Assert(!GinPageIsData(page));
if (insertData->isDelete)
{
Assert(GinPageIsLeaf(page));
PageIndexTupleDelete(page, off);
}
if (!GinPageIsLeaf(page) && updateblkno != InvalidBlockNumber)
{
IndexTuple itup = (IndexTuple) PageGetItem(page, PageGetItemId(page, off));
GinSetDownlink(itup, updateblkno);
}
}
/*
* returns blkno of leftmost child
*/
static BlockNumber
dataGetLeftMostPage(GinBtree btree, Page page)
{
PostingItem *pitem;
Assert(!GinPageIsLeaf(page));
Assert(GinPageIsData(page));
Assert(GinPageGetOpaque(page)->maxoff >= FirstOffsetNumber);
pitem = (PostingItem *) GinDataPageGetItem(page, FirstOffsetNumber);
return PostingItemGetBlockNumber(pitem);
}
/*
* Searches correct position for value on leaf page.
* Page should be corrrectly choosen.
* Returns true if value found on page.
*/
static bool
entryLocateLeafEntry(GinBtree btree, GinBtreeStack *stack)
{
Page page = BufferGetPage(stack->buffer);
OffsetNumber low,
high;
IndexTuple itup;
Assert(GinPageIsLeaf(page));
Assert(!GinPageIsData(page));
if (btree->fullScan)
{
stack->off = FirstOffsetNumber;
return TRUE;
}
low = FirstOffsetNumber;
high = PageGetMaxOffsetNumber(page);
if (high < low)
{
stack->off = FirstOffsetNumber;
return false;
}
high++;
while (high > low)
{
OffsetNumber mid = low + ((high - low) / 2);
int result;
itup = (IndexTuple) PageGetItem(page, PageGetItemId(page, mid));
result = ginCompareAttEntries(btree->ginstate,
btree->entryAttnum,
btree->entryValue,
gintuple_get_attrnum(btree->ginstate, itup),
gin_index_getattr(btree->ginstate, itup));
if (result == 0)
{
stack->off = mid;
return true;
}
else if (result > 0)
low = mid + 1;
else
high = mid;
}
stack->off = high;
return false;
}
static BlockNumber
entryGetLeftMostPage(GinBtree btree, Page page)
{
IndexTuple itup;
Assert(!GinPageIsLeaf(page));
Assert(!GinPageIsData(page));
Assert(PageGetMaxOffsetNumber(page) >= FirstOffsetNumber);
itup = (IndexTuple) PageGetItem(page, PageGetItemId(page, FirstOffsetNumber));
return GinGetDownlink(itup);
}
static void
ginRedoDeletePage(XLogReaderState *record)
{
XLogRecPtr lsn = record->EndRecPtr;
ginxlogDeletePage *data = (ginxlogDeletePage *) XLogRecGetData(record);
Buffer dbuffer;
Buffer pbuffer;
Buffer lbuffer;
Page page;
/*
* Lock left page first in order to prevent possible deadlock with
* ginStepRight().
*/
if (XLogReadBufferForRedo(record, 2, &lbuffer) == BLK_NEEDS_REDO)
{
page = BufferGetPage(lbuffer);
Assert(GinPageIsData(page));
GinPageGetOpaque(page)->rightlink = data->rightLink;
PageSetLSN(page, lsn);
MarkBufferDirty(lbuffer);
}
if (XLogReadBufferForRedo(record, 0, &dbuffer) == BLK_NEEDS_REDO)
{
page = BufferGetPage(dbuffer);
Assert(GinPageIsData(page));
GinPageGetOpaque(page)->flags = GIN_DELETED;
GinPageSetDeleteXid(page, data->deleteXid);
PageSetLSN(page, lsn);
MarkBufferDirty(dbuffer);
}
if (XLogReadBufferForRedo(record, 1, &pbuffer) == BLK_NEEDS_REDO)
{
page = BufferGetPage(pbuffer);
Assert(GinPageIsData(page));
Assert(!GinPageIsLeaf(page));
GinPageDeletePostingItem(page, data->parentOffset);
PageSetLSN(page, lsn);
MarkBufferDirty(pbuffer);
}
if (BufferIsValid(lbuffer))
UnlockReleaseBuffer(lbuffer);
if (BufferIsValid(pbuffer))
UnlockReleaseBuffer(pbuffer);
if (BufferIsValid(dbuffer))
UnlockReleaseBuffer(dbuffer);
}
/*
* Deletes posting item from non-leaf page
*/
void
GinPageDeletePostingItem(Page page, OffsetNumber offset)
{
OffsetNumber maxoff = GinPageGetOpaque(page)->maxoff;
Assert(!GinPageIsLeaf(page));
Assert(offset >= FirstOffsetNumber && offset <= maxoff);
if (offset != maxoff)
memmove(GinDataPageGetItem(page, offset), GinDataPageGetItem(page, offset + 1),
sizeof(PostingItem) * (maxoff - offset));
GinPageGetOpaque(page)->maxoff--;
}
static void
ginRedoInsertEntry(Buffer buffer, bool isLeaf, BlockNumber rightblkno, void *rdata)
{
Page page = BufferGetPage(buffer);
ginxlogInsertEntry *data = (ginxlogInsertEntry *) rdata;
OffsetNumber offset = data->offset;
IndexTuple itup;
if (rightblkno != InvalidBlockNumber)
{
/* update link to right page after split */
Assert(!GinPageIsLeaf(page));
Assert(offset >= FirstOffsetNumber && offset <= PageGetMaxOffsetNumber(page));
itup = (IndexTuple) PageGetItem(page, PageGetItemId(page, offset));
GinSetDownlink(itup, rightblkno);
}
if (data->isDelete)
{
Assert(GinPageIsLeaf(page));
Assert(offset >= FirstOffsetNumber && offset <= PageGetMaxOffsetNumber(page));
PageIndexTupleDelete(page, offset);
}
itup = &data->tuple;
if (PageAddItem(page, (Item) itup, IndexTupleSize(itup), offset, false, false) == InvalidOffsetNumber)
{
RelFileNode node;
ForkNumber forknum;
BlockNumber blknum;
BufferGetTag(buffer, &node, &forknum, &blknum);
elog(ERROR, "failed to add item to index page in %u/%u/%u",
node.spcNode, node.dbNode, node.relNode);
}
}
/*
* Searches correct position for value on leaf page.
* Page should be correctly chosen.
* Returns true if value found on page.
*/
static bool
dataLocateLeafItem(GinBtree btree, GinBtreeStack *stack)
{
Page page = BufferGetPage(stack->buffer);
OffsetNumber low,
high;
int result;
Assert(GinPageIsLeaf(page));
Assert(GinPageIsData(page));
if (btree->fullScan)
{
stack->off = FirstOffsetNumber;
return TRUE;
}
low = FirstOffsetNumber;
high = GinPageGetOpaque(page)->maxoff;
if (high < low)
{
stack->off = FirstOffsetNumber;
return false;
}
high++;
while (high > low)
{
OffsetNumber mid = low + ((high - low) / 2);
result = ginCompareItemPointers(&btree->itemptr,
GinDataPageGetItemPointer(page, mid));
if (result == 0)
{
stack->off = mid;
return true;
}
else if (result > 0)
low = mid + 1;
else
high = mid;
}
stack->off = high;
return false;
}
/*
* In case of previous split update old child blkno to
* new right page
* item pointer never deletes!
*/
static BlockNumber
dataPrepareData(GinBtree btree, Page page, OffsetNumber off)
{
BlockNumber ret = InvalidBlockNumber;
Assert(GinPageIsData(page));
if (!GinPageIsLeaf(page) && btree->rightblkno != InvalidBlockNumber)
{
PostingItem *pitem = (PostingItem *) GinDataPageGetItem(page, off);
PostingItemSetBlockNumber(pitem, btree->rightblkno);
ret = btree->rightblkno;
}
btree->rightblkno = InvalidBlockNumber;
return ret;
}
/*
* Returns new tuple with copied value from source tuple.
* New tuple will not store posting list
*/
static IndexTuple
copyIndexTuple(IndexTuple itup, Page page)
{
IndexTuple nitup;
if (GinPageIsLeaf(page) && !GinIsPostingTree(itup))
{
nitup = (IndexTuple) palloc(MAXALIGN(GinGetOrigSizePosting(itup)));
memcpy(nitup, itup, GinGetOrigSizePosting(itup));
nitup->t_info &= ~INDEX_SIZE_MASK;
nitup->t_info |= GinGetOrigSizePosting(itup);
}
else
{
nitup = (IndexTuple) palloc(MAXALIGN(IndexTupleSize(itup)));
memcpy(nitup, itup, IndexTupleSize(itup));
}
return nitup;
}
/*
* Finds links to blkno on non-leaf page, returns
* offset of PostingItem
*/
static OffsetNumber
dataFindChildPtr(GinBtree btree, Page page, BlockNumber blkno, OffsetNumber storedOff)
{
OffsetNumber i,
maxoff = GinPageGetOpaque(page)->maxoff;
PostingItem *pitem;
Assert(!GinPageIsLeaf(page));
Assert(GinPageIsData(page));
/* if page isn't changed, we return storedOff */
if (storedOff >= FirstOffsetNumber && storedOff <= maxoff)
{
pitem = (PostingItem *) GinDataPageGetItem(page, storedOff);
if (PostingItemGetBlockNumber(pitem) == blkno)
return storedOff;
/*
* we hope, that needed pointer goes to right. It's true if there
* wasn't a deletion
*/
for (i = storedOff + 1; i <= maxoff; i++)
{
pitem = (PostingItem *) GinDataPageGetItem(page, i);
if (PostingItemGetBlockNumber(pitem) == blkno)
return i;
}
maxoff = storedOff - 1;
}
/* last chance */
for (i = FirstOffsetNumber; i <= maxoff; i++)
{
pitem = (PostingItem *) GinDataPageGetItem(page, i);
if (PostingItemGetBlockNumber(pitem) == blkno)
return i;
}
return InvalidOffsetNumber;
}
static OffsetNumber
entryFindChildPtr(GinBtree btree, Page page, BlockNumber blkno, OffsetNumber storedOff)
{
OffsetNumber i,
maxoff = PageGetMaxOffsetNumber(page);
IndexTuple itup;
Assert(!GinPageIsLeaf(page));
Assert(!GinPageIsData(page));
/* if page isn't changed, we returns storedOff */
if (storedOff >= FirstOffsetNumber && storedOff <= maxoff)
{
itup = (IndexTuple) PageGetItem(page, PageGetItemId(page, storedOff));
if (GinGetDownlink(itup) == blkno)
return storedOff;
/*
* we hope, that needed pointer goes to right. It's true if there
* wasn't a deletion
*/
for (i = storedOff + 1; i <= maxoff; i++)
{
itup = (IndexTuple) PageGetItem(page, PageGetItemId(page, i));
if (GinGetDownlink(itup) == blkno)
return i;
}
maxoff = storedOff - 1;
}
/* last chance */
for (i = FirstOffsetNumber; i <= maxoff; i++)
{
itup = (IndexTuple) PageGetItem(page, PageGetItemId(page, i));
if (GinGetDownlink(itup) == blkno)
return i;
}
return InvalidOffsetNumber;
}
static void
ginRedoSplitData(Page lpage, Page rpage, void *rdata)
{
bool isleaf = GinPageIsLeaf(lpage);
if (isleaf)
{
ginxlogSplitDataLeaf *data = (ginxlogSplitDataLeaf *) rdata;
Pointer lptr = (Pointer) rdata + sizeof(ginxlogSplitDataLeaf);
Pointer rptr = lptr + data->lsize;
Assert(data->lsize > 0 && data->lsize <= GinDataPageMaxDataSize);
Assert(data->rsize > 0 && data->rsize <= GinDataPageMaxDataSize);
memcpy(GinDataLeafPageGetPostingList(lpage), lptr, data->lsize);
memcpy(GinDataLeafPageGetPostingList(rpage), rptr, data->rsize);
GinDataPageSetDataSize(lpage, data->lsize);
GinDataPageSetDataSize(rpage, data->rsize);
*GinDataPageGetRightBound(lpage) = data->lrightbound;
*GinDataPageGetRightBound(rpage) = data->rrightbound;
}
else
{
ginxlogSplitDataInternal *data = (ginxlogSplitDataInternal *) rdata;
PostingItem *items = (PostingItem *) ((char *) rdata + sizeof(ginxlogSplitDataInternal));
OffsetNumber i;
OffsetNumber maxoff;
for (i = 0; i < data->separator; i++)
GinDataPageAddPostingItem(lpage, &items[i], InvalidOffsetNumber);
for (i = data->separator; i < data->nitem; i++)
GinDataPageAddPostingItem(rpage, &items[i], InvalidOffsetNumber);
/* set up right key */
maxoff = GinPageGetOpaque(lpage)->maxoff;
*GinDataPageGetRightBound(lpage) = GinDataPageGetPostingItem(lpage, maxoff)->key;
*GinDataPageGetRightBound(rpage) = data->rightbound;
}
}
/*
* Create a WAL record for vacuuming entry tree leaf page.
*/
static void
xlogVacuumPage(Relation index, Buffer buffer)
{
Page page = BufferGetPage(buffer);
XLogRecPtr recptr;
/* This is only used for entry tree leaf pages. */
Assert(!GinPageIsData(page));
Assert(GinPageIsLeaf(page));
if (!RelationNeedsWAL(index))
return;
/*
* Always create a full image, we don't track the changes on the page at
* any more fine-grained level. This could obviously be improved...
*/
XLogBeginInsert();
XLogRegisterBuffer(0, buffer, REGBUF_FORCE_IMAGE | REGBUF_STANDARD);
recptr = XLogInsert(RM_GIN_ID, XLOG_GIN_VACUUM_PAGE);
PageSetLSN(page, recptr);
}
/*
* checks space to install new value,
* item pointer never deletes!
*/
static bool
dataIsEnoughSpace(GinBtree btree, Buffer buf, OffsetNumber off)
{
Page page = BufferGetPage(buf);
Assert(GinPageIsData(page));
Assert(!btree->isDelete);
if (GinPageIsLeaf(page))
{
if (GinPageRightMost(page) && off > GinPageGetOpaque(page)->maxoff)
{
if ((btree->nitem - btree->curitem) * sizeof(ItemPointerData) <= GinDataPageGetFreeSpace(page))
return true;
}
else if (sizeof(ItemPointerData) <= GinDataPageGetFreeSpace(page))
return true;
}
else if (sizeof(PostingItem) <= GinDataPageGetFreeSpace(page))
return true;
return false;
}
static void
ginRedoVacuumDataLeafPage(XLogReaderState *record)
{
XLogRecPtr lsn = record->EndRecPtr;
Buffer buffer;
if (XLogReadBufferForRedo(record, 0, &buffer) == BLK_NEEDS_REDO)
{
Page page = BufferGetPage(buffer);
Size len;
ginxlogVacuumDataLeafPage *xlrec;
xlrec = (ginxlogVacuumDataLeafPage *) XLogRecGetBlockData(record, 0, &len);
Assert(GinPageIsLeaf(page));
Assert(GinPageIsData(page));
ginRedoRecompress(page, &xlrec->data);
PageSetLSN(page, lsn);
MarkBufferDirty(buffer);
}
if (BufferIsValid(buffer))
UnlockReleaseBuffer(buffer);
}
/*
* fills WAL record for vacuum leaf page
*/
static void
xlogVacuumPage(Relation index, Buffer buffer)
{
Page page = BufferGetPage(buffer);
XLogRecPtr recptr;
XLogRecData rdata[3];
ginxlogVacuumPage data;
char *backup;
char itups[BLCKSZ];
uint32 len = 0;
Assert(GinPageIsLeaf(page));
if (!RelationNeedsWAL(index))
return;
data.node = index->rd_node;
data.blkno = BufferGetBlockNumber(buffer);
if (GinPageIsData(page))
{
backup = GinDataPageGetData(page);
data.nitem = GinPageGetOpaque(page)->maxoff;
if (data.nitem)
len = MAXALIGN(sizeof(ItemPointerData) * data.nitem);
}
else
{
char *ptr;
OffsetNumber i;
ptr = backup = itups;
for (i = FirstOffsetNumber; i <= PageGetMaxOffsetNumber(page); i++)
{
IndexTuple itup = (IndexTuple) PageGetItem(page, PageGetItemId(page, i));
memcpy(ptr, itup, IndexTupleSize(itup));
ptr += MAXALIGN(IndexTupleSize(itup));
}
data.nitem = PageGetMaxOffsetNumber(page);
len = ptr - backup;
}
rdata[0].buffer = buffer;
rdata[0].buffer_std = (GinPageIsData(page)) ? FALSE : TRUE;
rdata[0].len = 0;
rdata[0].data = NULL;
rdata[0].next = rdata + 1;
rdata[1].buffer = InvalidBuffer;
rdata[1].len = sizeof(ginxlogVacuumPage);
rdata[1].data = (char *) &data;
if (len == 0)
{
rdata[1].next = NULL;
}
else
{
rdata[1].next = rdata + 2;
rdata[2].buffer = InvalidBuffer;
rdata[2].len = len;
rdata[2].data = backup;
rdata[2].next = NULL;
}
recptr = XLogInsert(RM_GIN_ID, XLOG_GIN_VACUUM_PAGE, rdata);
PageSetLSN(page, recptr);
PageSetTLI(page, ThisTimeLineID);
}
Datum
ginvacuumcleanup(PG_FUNCTION_ARGS)
{
IndexVacuumInfo *info = (IndexVacuumInfo *) PG_GETARG_POINTER(0);
IndexBulkDeleteResult *stats = (IndexBulkDeleteResult *) PG_GETARG_POINTER(1);
Relation index = info->index;
bool needLock;
BlockNumber npages,
blkno;
BlockNumber totFreePages;
GinState ginstate;
GinStatsData idxStat;
/*
* In an autovacuum analyze, we want to clean up pending insertions.
* Otherwise, an ANALYZE-only call is a no-op.
*/
if (info->analyze_only)
{
if (IsAutoVacuumWorkerProcess())
{
initGinState(&ginstate, index);
ginInsertCleanup(&ginstate, true, stats);
}
PG_RETURN_POINTER(stats);
}
/*
* Set up all-zero stats and cleanup pending inserts if ginbulkdelete
* wasn't called
*/
if (stats == NULL)
{
stats = (IndexBulkDeleteResult *) palloc0(sizeof(IndexBulkDeleteResult));
initGinState(&ginstate, index);
ginInsertCleanup(&ginstate, true, stats);
}
memset(&idxStat, 0, sizeof(idxStat));
/*
* XXX we always report the heap tuple count as the number of index
* entries. This is bogus if the index is partial, but it's real hard to
* tell how many distinct heap entries are referenced by a GIN index.
*/
stats->num_index_tuples = info->num_heap_tuples;
stats->estimated_count = info->estimated_count;
/*
* Need lock unless it's local to this backend.
*/
needLock = !RELATION_IS_LOCAL(index);
if (needLock)
LockRelationForExtension(index, ExclusiveLock);
npages = RelationGetNumberOfBlocks(index);
if (needLock)
UnlockRelationForExtension(index, ExclusiveLock);
totFreePages = 0;
for (blkno = GIN_ROOT_BLKNO; blkno < npages; blkno++)
{
Buffer buffer;
Page page;
vacuum_delay_point();
buffer = ReadBufferExtended(index, MAIN_FORKNUM, blkno,
RBM_NORMAL, info->strategy);
LockBuffer(buffer, GIN_SHARE);
page = (Page) BufferGetPage(buffer);
if (GinPageIsDeleted(page))
{
Assert(blkno != GIN_ROOT_BLKNO);
RecordFreeIndexPage(index, blkno);
totFreePages++;
}
else if (GinPageIsData(page))
{
idxStat.nDataPages++;
}
else if (!GinPageIsList(page))
{
idxStat.nEntryPages++;
if (GinPageIsLeaf(page))
idxStat.nEntries += PageGetMaxOffsetNumber(page);
}
UnlockReleaseBuffer(buffer);
}
/* Update the metapage with accurate page and entry counts */
idxStat.nTotalPages = npages;
ginUpdateStats(info->index, &idxStat);
/* Finally, vacuum the FSM */
IndexFreeSpaceMapVacuum(info->index);
stats->pages_free = totFreePages;
if (needLock)
LockRelationForExtension(index, ExclusiveLock);
stats->num_pages = RelationGetNumberOfBlocks(index);
if (needLock)
UnlockRelationForExtension(index, ExclusiveLock);
PG_RETURN_POINTER(stats);
}
Datum
ginbulkdelete(PG_FUNCTION_ARGS)
{
IndexVacuumInfo *info = (IndexVacuumInfo *) PG_GETARG_POINTER(0);
IndexBulkDeleteResult *stats = (IndexBulkDeleteResult *) PG_GETARG_POINTER(1);
IndexBulkDeleteCallback callback = (IndexBulkDeleteCallback) PG_GETARG_POINTER(2);
void *callback_state = (void *) PG_GETARG_POINTER(3);
Relation index = info->index;
BlockNumber blkno = GIN_ROOT_BLKNO;
GinVacuumState gvs;
Buffer buffer;
BlockNumber rootOfPostingTree[BLCKSZ / (sizeof(IndexTupleData) + sizeof(ItemId))];
uint32 nRoot;
gvs.index = index;
gvs.callback = callback;
gvs.callback_state = callback_state;
gvs.strategy = info->strategy;
initGinState(&gvs.ginstate, index);
/* first time through? */
if (stats == NULL)
{
/* Yes, so initialize stats to zeroes */
stats = (IndexBulkDeleteResult *) palloc0(sizeof(IndexBulkDeleteResult));
/* and cleanup any pending inserts */
ginInsertCleanup(&gvs.ginstate, true, stats);
}
/* we'll re-count the tuples each time */
stats->num_index_tuples = 0;
gvs.result = stats;
buffer = ReadBufferExtended(index, MAIN_FORKNUM, blkno,
RBM_NORMAL, info->strategy);
/* find leaf page */
for (;;)
{
Page page = BufferGetPage(buffer);
IndexTuple itup;
LockBuffer(buffer, GIN_SHARE);
Assert(!GinPageIsData(page));
if (GinPageIsLeaf(page))
{
LockBuffer(buffer, GIN_UNLOCK);
LockBuffer(buffer, GIN_EXCLUSIVE);
if (blkno == GIN_ROOT_BLKNO && !GinPageIsLeaf(page))
{
LockBuffer(buffer, GIN_UNLOCK);
continue; /* check it one more */
}
break;
}
Assert(PageGetMaxOffsetNumber(page) >= FirstOffsetNumber);
itup = (IndexTuple) PageGetItem(page, PageGetItemId(page, FirstOffsetNumber));
blkno = GinGetDownlink(itup);
Assert(blkno != InvalidBlockNumber);
UnlockReleaseBuffer(buffer);
buffer = ReadBufferExtended(index, MAIN_FORKNUM, blkno,
RBM_NORMAL, info->strategy);
}
/* right now we found leftmost page in entry's BTree */
for (;;)
{
Page page = BufferGetPage(buffer);
Page resPage;
uint32 i;
Assert(!GinPageIsData(page));
resPage = ginVacuumEntryPage(&gvs, buffer, rootOfPostingTree, &nRoot);
blkno = GinPageGetOpaque(page)->rightlink;
if (resPage)
{
START_CRIT_SECTION();
PageRestoreTempPage(resPage, page);
MarkBufferDirty(buffer);
xlogVacuumPage(gvs.index, buffer);
UnlockReleaseBuffer(buffer);
END_CRIT_SECTION();
}
else
{
UnlockReleaseBuffer(buffer);
}
vacuum_delay_point();
for (i = 0; i < nRoot; i++)
{
ginVacuumPostingTree(&gvs, rootOfPostingTree[i]);
vacuum_delay_point();
}
if (blkno == InvalidBlockNumber) /* rightmost page */
break;
buffer = ReadBufferExtended(index, MAIN_FORKNUM, blkno,
RBM_NORMAL, info->strategy);
LockBuffer(buffer, GIN_EXCLUSIVE);
}
PG_RETURN_POINTER(gvs.result);
}
/*
* scans posting tree and deletes empty pages
*/
static bool
ginScanToDelete(GinVacuumState *gvs, BlockNumber blkno, bool isRoot, DataPageDeleteStack *parent, OffsetNumber myoff)
{
DataPageDeleteStack *me;
Buffer buffer;
Page page;
bool meDelete = FALSE;
if (isRoot)
{
me = parent;
}
else
{
if (!parent->child)
{
me = (DataPageDeleteStack *) palloc0(sizeof(DataPageDeleteStack));
me->parent = parent;
parent->child = me;
me->leftBlkno = InvalidBlockNumber;
}
else
me = parent->child;
}
buffer = ReadBufferExtended(gvs->index, MAIN_FORKNUM, blkno,
RBM_NORMAL, gvs->strategy);
page = BufferGetPage(buffer);
Assert(GinPageIsData(page));
if (!GinPageIsLeaf(page))
{
OffsetNumber i;
me->blkno = blkno;
for (i = FirstOffsetNumber; i <= GinPageGetOpaque(page)->maxoff; i++)
{
PostingItem *pitem = (PostingItem *) GinDataPageGetItem(page, i);
if (ginScanToDelete(gvs, PostingItemGetBlockNumber(pitem), FALSE, me, i))
i--;
}
}
if (GinPageGetOpaque(page)->maxoff < FirstOffsetNumber)
{
if (!(me->leftBlkno == InvalidBlockNumber && GinPageRightMost(page)))
{
/* we never delete right most branch */
Assert(!isRoot);
if (GinPageGetOpaque(page)->maxoff < FirstOffsetNumber)
{
ginDeletePage(gvs, blkno, me->leftBlkno, me->parent->blkno, myoff, me->parent->isRoot);
meDelete = TRUE;
}
}
}
ReleaseBuffer(buffer);
if (!meDelete)
me->leftBlkno = blkno;
return meDelete;
}