/*
* _bt_pagedel() -- Delete a page from the b-tree, if legal to do so.
*
* This action unlinks the page from the b-tree structure, removing all
* pointers leading to it --- but not touching its own left and right links.
* The page cannot be physically reclaimed right away, since other processes
* may currently be trying to follow links leading to the page; they have to
* be allowed to use its right-link to recover. See nbtree/README.
*
* On entry, the target buffer must be pinned and locked (either read or write
* lock is OK). This lock and pin will be dropped before exiting.
*
* The "stack" argument can be a search stack leading (approximately) to the
* target page, or NULL --- outside callers typically pass NULL since they
* have not done such a search, but internal recursion cases pass the stack
* to avoid duplicated search effort.
*
* Returns the number of pages successfully deleted (zero if page cannot
* be deleted now; could be more than one if parent pages were deleted too).
*
* NOTE: this leaks memory. Rather than trying to clean up everything
* carefully, it's better to run it in a temp context that can be reset
* frequently.
*/
int
_bt_pagedel(Relation rel, Buffer buf, BTStack stack)
{
int result;
BlockNumber target,
leftsib,
rightsib,
parent;
OffsetNumber poffset,
maxoff;
uint32 targetlevel,
ilevel;
ItemId itemid;
IndexTuple targetkey,
itup;
ScanKey itup_scankey;
Buffer lbuf,
rbuf,
pbuf;
bool parent_half_dead;
bool parent_one_child;
bool rightsib_empty;
Buffer metabuf = InvalidBuffer;
Page metapg = NULL;
BTMetaPageData *metad = NULL;
Page page;
BTPageOpaque opaque;
/*
* We can never delete rightmost pages nor root pages. While at it, check
* that page is not already deleted and is empty.
*/
page = BufferGetPage(buf);
opaque = (BTPageOpaque) PageGetSpecialPointer(page);
if (P_RIGHTMOST(opaque) || P_ISROOT(opaque) || P_ISDELETED(opaque) ||
P_FIRSTDATAKEY(opaque) <= PageGetMaxOffsetNumber(page))
{
/* Should never fail to delete a half-dead page */
Assert(!P_ISHALFDEAD(opaque));
_bt_relbuf(rel, buf);
return 0;
}
/*
* Save info about page, including a copy of its high key (it must have
* one, being non-rightmost).
*/
target = BufferGetBlockNumber(buf);
targetlevel = opaque->btpo.level;
leftsib = opaque->btpo_prev;
itemid = PageGetItemId(page, P_HIKEY);
targetkey = CopyIndexTuple((IndexTuple) PageGetItem(page, itemid));
/*
* To avoid deadlocks, we'd better drop the target page lock before going
* further.
*/
_bt_relbuf(rel, buf);
/*
* We need an approximate pointer to the page's parent page. We use the
* standard search mechanism to search for the page's high key; this will
* give us a link to either the current parent or someplace to its left
* (if there are multiple equal high keys). In recursion cases, the
* caller already generated a search stack and we can just re-use that
* work.
*/
if (stack == NULL)
{
if (!InRecovery)
{
/* we need an insertion scan key to do our search, so build one */
itup_scankey = _bt_mkscankey(rel, targetkey);
/* find the leftmost leaf page containing this key */
stack = _bt_search(rel, rel->rd_rel->relnatts, itup_scankey, false,
&lbuf, BT_READ);
/* don't need a pin on that either */
_bt_relbuf(rel, lbuf);
/*
* If we are trying to delete an interior page, _bt_search did
* more than we needed. Locate the stack item pointing to our
* parent level.
*/
ilevel = 0;
for (;;)
{
if (stack == NULL)
elog(ERROR, "not enough stack items");
if (ilevel == targetlevel)
break;
stack = stack->bts_parent;
ilevel++;
}
}
else
{
/*
* During WAL recovery, we can't use _bt_search (for one reason,
* it might invoke user-defined comparison functions that expect
* facilities not available in recovery mode). Instead, just set
* up a dummy stack pointing to the left end of the parent tree
* level, from which _bt_getstackbuf will walk right to the parent
* page. Painful, but we don't care too much about performance in
* this scenario.
*/
pbuf = _bt_get_endpoint(rel, targetlevel + 1, false);
stack = (BTStack) palloc(sizeof(BTStackData));
stack->bts_blkno = BufferGetBlockNumber(pbuf);
stack->bts_offset = InvalidOffsetNumber;
/* bts_btentry will be initialized below */
stack->bts_parent = NULL;
_bt_relbuf(rel, pbuf);
}
}
/*
* We cannot delete a page that is the rightmost child of its immediate
* parent, unless it is the only child --- in which case the parent has to
* be deleted too, and the same condition applies recursively to it. We
* have to check this condition all the way up before trying to delete. We
* don't need to re-test when deleting a non-leaf page, though.
*/
if (targetlevel == 0 &&
!_bt_parent_deletion_safe(rel, target, stack))
return 0;
/*
* We have to lock the pages we need to modify in the standard order:
* moving right, then up. Else we will deadlock against other writers.
*
* So, we need to find and write-lock the current left sibling of the
* target page. The sibling that was current a moment ago could have
* split, so we may have to move right. This search could fail if either
* the sibling or the target page was deleted by someone else meanwhile;
* if so, give up. (Right now, that should never happen, since page
* deletion is only done in VACUUM and there shouldn't be multiple VACUUMs
* concurrently on the same table.)
*/
if (leftsib != P_NONE)
{
lbuf = _bt_getbuf(rel, leftsib, BT_WRITE);
page = BufferGetPage(lbuf);
opaque = (BTPageOpaque) PageGetSpecialPointer(page);
while (P_ISDELETED(opaque) || opaque->btpo_next != target)
{
/* step right one page */
leftsib = opaque->btpo_next;
_bt_relbuf(rel, lbuf);
if (leftsib == P_NONE)
{
elog(LOG, "no left sibling (concurrent deletion?) in \"%s\"",
RelationGetRelationName(rel));
return 0;
}
lbuf = _bt_getbuf(rel, leftsib, BT_WRITE);
page = BufferGetPage(lbuf);
opaque = (BTPageOpaque) PageGetSpecialPointer(page);
}
}
else
lbuf = InvalidBuffer;
/*
* Next write-lock the target page itself. It should be okay to take just
* a write lock not a superexclusive lock, since no scans would stop on an
* empty page.
*/
buf = _bt_getbuf(rel, target, BT_WRITE);
page = BufferGetPage(buf);
opaque = (BTPageOpaque) PageGetSpecialPointer(page);
/*
* Check page is still empty etc, else abandon deletion. The empty check
* is necessary since someone else might have inserted into it while we
* didn't have it locked; the others are just for paranoia's sake.
*/
if (P_RIGHTMOST(opaque) || P_ISROOT(opaque) || P_ISDELETED(opaque) ||
P_FIRSTDATAKEY(opaque) <= PageGetMaxOffsetNumber(page))
{
_bt_relbuf(rel, buf);
if (BufferIsValid(lbuf))
_bt_relbuf(rel, lbuf);
return 0;
}
if (opaque->btpo_prev != leftsib)
elog(ERROR, "left link changed unexpectedly in block %u of index \"%s\"",
target, RelationGetRelationName(rel));
/*
* And next write-lock the (current) right sibling.
*/
rightsib = opaque->btpo_next;
rbuf = _bt_getbuf(rel, rightsib, BT_WRITE);
page = BufferGetPage(rbuf);
opaque = (BTPageOpaque) PageGetSpecialPointer(page);
if (opaque->btpo_prev != target)
elog(ERROR, "right sibling's left-link doesn't match: "
"block %u links to %u instead of expected %u in index \"%s\"",
rightsib, opaque->btpo_prev, target,
RelationGetRelationName(rel));
/*
* Any insert which would have gone on the target block will now go to the
* right sibling block.
*/
PredicateLockPageCombine(rel, target, rightsib);
/*
* Next find and write-lock the current parent of the target page. This is
* essentially the same as the corresponding step of splitting.
*/
ItemPointerSet(&(stack->bts_btentry.t_tid), target, P_HIKEY);
pbuf = _bt_getstackbuf(rel, stack, BT_WRITE);
if (pbuf == InvalidBuffer)
elog(ERROR, "failed to re-find parent key in index \"%s\" for deletion target page %u",
RelationGetRelationName(rel), target);
parent = stack->bts_blkno;
poffset = stack->bts_offset;
/*
* If the target is the rightmost child of its parent, then we can't
* delete, unless it's also the only child --- in which case the parent
* changes to half-dead status. The "can't delete" case should have been
* detected by _bt_parent_deletion_safe, so complain if we see it now.
*/
page = BufferGetPage(pbuf);
opaque = (BTPageOpaque) PageGetSpecialPointer(page);
maxoff = PageGetMaxOffsetNumber(page);
parent_half_dead = false;
parent_one_child = false;
if (poffset >= maxoff)
{
if (poffset == P_FIRSTDATAKEY(opaque))
parent_half_dead = true;
else
elog(ERROR, "failed to delete rightmost child %u of block %u in index \"%s\"",
target, parent, RelationGetRelationName(rel));
}
else
{
/* Will there be exactly one child left in this parent? */
if (OffsetNumberNext(P_FIRSTDATAKEY(opaque)) == maxoff)
parent_one_child = true;
}
/*
* If we are deleting the next-to-last page on the target's level, then
* the rightsib is a candidate to become the new fast root. (In theory, it
* might be possible to push the fast root even further down, but the odds
* of doing so are slim, and the locking considerations daunting.)
*
* We don't support handling this in the case where the parent is becoming
* half-dead, even though it theoretically could occur.
*
* We can safely acquire a lock on the metapage here --- see comments for
* _bt_newroot().
*/
if (leftsib == P_NONE && !parent_half_dead)
{
page = BufferGetPage(rbuf);
opaque = (BTPageOpaque) PageGetSpecialPointer(page);
Assert(opaque->btpo.level == targetlevel);
if (P_RIGHTMOST(opaque))
{
/* rightsib will be the only one left on the level */
metabuf = _bt_getbuf(rel, BTREE_METAPAGE, BT_WRITE);
metapg = BufferGetPage(metabuf);
metad = BTPageGetMeta(metapg);
/*
* The expected case here is btm_fastlevel == targetlevel+1; if
* the fastlevel is <= targetlevel, something is wrong, and we
* choose to overwrite it to fix it.
*/
if (metad->btm_fastlevel > targetlevel + 1)
{
/* no update wanted */
_bt_relbuf(rel, metabuf);
metabuf = InvalidBuffer;
}
}
}
/*
* Check that the parent-page index items we're about to delete/overwrite
* contain what we expect. This can fail if the index has become
* corrupt for some reason. We want to throw any error before entering
* the critical section --- otherwise it'd be a PANIC.
*
* The test on the target item is just an Assert because _bt_getstackbuf
* should have guaranteed it has the expected contents. The test on the
* next-child downlink is known to sometimes fail in the field, though.
*/
page = BufferGetPage(pbuf);
opaque = (BTPageOpaque) PageGetSpecialPointer(page);
#ifdef USE_ASSERT_CHECKING
itemid = PageGetItemId(page, poffset);
itup = (IndexTuple) PageGetItem(page, itemid);
Assert(ItemPointerGetBlockNumber(&(itup->t_tid)) == target);
#endif
if (!parent_half_dead)
{
OffsetNumber nextoffset;
nextoffset = OffsetNumberNext(poffset);
itemid = PageGetItemId(page, nextoffset);
itup = (IndexTuple) PageGetItem(page, itemid);
if (ItemPointerGetBlockNumber(&(itup->t_tid)) != rightsib)
elog(ERROR, "right sibling %u of block %u is not next child %u of block %u in index \"%s\"",
rightsib, target, ItemPointerGetBlockNumber(&(itup->t_tid)),
parent, RelationGetRelationName(rel));
}
/*
* Here we begin doing the deletion.
*/
/* No ereport(ERROR) until changes are logged */
START_CRIT_SECTION();
/*
* Update parent. The normal case is a tad tricky because we want to
* delete the target's downlink and the *following* key. Easiest way is
* to copy the right sibling's downlink over the target downlink, and then
* delete the following item.
*/
if (parent_half_dead)
{
PageIndexTupleDelete(page, poffset);
opaque->btpo_flags |= BTP_HALF_DEAD;
}
else
{
OffsetNumber nextoffset;
itemid = PageGetItemId(page, poffset);
itup = (IndexTuple) PageGetItem(page, itemid);
ItemPointerSet(&(itup->t_tid), rightsib, P_HIKEY);
nextoffset = OffsetNumberNext(poffset);
PageIndexTupleDelete(page, nextoffset);
}
/*
* Update siblings' side-links. Note the target page's side-links will
* continue to point to the siblings. Asserts here are just rechecking
* things we already verified above.
*/
if (BufferIsValid(lbuf))
{
page = BufferGetPage(lbuf);
opaque = (BTPageOpaque) PageGetSpecialPointer(page);
Assert(opaque->btpo_next == target);
opaque->btpo_next = rightsib;
}
page = BufferGetPage(rbuf);
opaque = (BTPageOpaque) PageGetSpecialPointer(page);
Assert(opaque->btpo_prev == target);
opaque->btpo_prev = leftsib;
rightsib_empty = (P_FIRSTDATAKEY(opaque) > PageGetMaxOffsetNumber(page));
/*
* Mark the page itself deleted. It can be recycled when all current
* transactions are gone.
*/
page = BufferGetPage(buf);
opaque = (BTPageOpaque) PageGetSpecialPointer(page);
opaque->btpo_flags &= ~BTP_HALF_DEAD;
opaque->btpo_flags |= BTP_DELETED;
opaque->btpo.xact = ReadNewTransactionId();
/* And update the metapage, if needed */
if (BufferIsValid(metabuf))
{
metad->btm_fastroot = rightsib;
metad->btm_fastlevel = targetlevel;
MarkBufferDirty(metabuf);
}
/* Must mark buffers dirty before XLogInsert */
MarkBufferDirty(pbuf);
MarkBufferDirty(rbuf);
MarkBufferDirty(buf);
if (BufferIsValid(lbuf))
MarkBufferDirty(lbuf);
/* XLOG stuff */
if (RelationNeedsWAL(rel))
{
xl_btree_delete_page xlrec;
xl_btree_metadata xlmeta;
uint8 xlinfo;
XLogRecPtr recptr;
XLogRecData rdata[5];
XLogRecData *nextrdata;
xlrec.target.node = rel->rd_node;
ItemPointerSet(&(xlrec.target.tid), parent, poffset);
xlrec.deadblk = target;
xlrec.leftblk = leftsib;
xlrec.rightblk = rightsib;
xlrec.btpo_xact = opaque->btpo.xact;
rdata[0].data = (char *) &xlrec;
rdata[0].len = SizeOfBtreeDeletePage;
rdata[0].buffer = InvalidBuffer;
rdata[0].next = nextrdata = &(rdata[1]);
if (BufferIsValid(metabuf))
{
xlmeta.root = metad->btm_root;
xlmeta.level = metad->btm_level;
xlmeta.fastroot = metad->btm_fastroot;
xlmeta.fastlevel = metad->btm_fastlevel;
nextrdata->data = (char *) &xlmeta;
nextrdata->len = sizeof(xl_btree_metadata);
nextrdata->buffer = InvalidBuffer;
nextrdata->next = nextrdata + 1;
nextrdata++;
xlinfo = XLOG_BTREE_DELETE_PAGE_META;
}
else if (parent_half_dead)
xlinfo = XLOG_BTREE_DELETE_PAGE_HALF;
else
xlinfo = XLOG_BTREE_DELETE_PAGE;
nextrdata->data = NULL;
nextrdata->len = 0;
nextrdata->next = nextrdata + 1;
nextrdata->buffer = pbuf;
nextrdata->buffer_std = true;
nextrdata++;
nextrdata->data = NULL;
nextrdata->len = 0;
nextrdata->buffer = rbuf;
nextrdata->buffer_std = true;
nextrdata->next = NULL;
if (BufferIsValid(lbuf))
{
nextrdata->next = nextrdata + 1;
nextrdata++;
nextrdata->data = NULL;
nextrdata->len = 0;
nextrdata->buffer = lbuf;
nextrdata->buffer_std = true;
nextrdata->next = NULL;
}
recptr = XLogInsert(RM_BTREE_ID, xlinfo, rdata);
if (BufferIsValid(metabuf))
{
PageSetLSN(metapg, recptr);
PageSetTLI(metapg, ThisTimeLineID);
}
page = BufferGetPage(pbuf);
PageSetLSN(page, recptr);
PageSetTLI(page, ThisTimeLineID);
page = BufferGetPage(rbuf);
PageSetLSN(page, recptr);
PageSetTLI(page, ThisTimeLineID);
page = BufferGetPage(buf);
PageSetLSN(page, recptr);
PageSetTLI(page, ThisTimeLineID);
if (BufferIsValid(lbuf))
{
page = BufferGetPage(lbuf);
PageSetLSN(page, recptr);
PageSetTLI(page, ThisTimeLineID);
}
}
END_CRIT_SECTION();
/* release metapage; send out relcache inval if metapage changed */
if (BufferIsValid(metabuf))
{
CacheInvalidateRelcache(rel);
_bt_relbuf(rel, metabuf);
}
/* can always release leftsib immediately */
if (BufferIsValid(lbuf))
_bt_relbuf(rel, lbuf);
/*
* If parent became half dead, recurse to delete it. Otherwise, if right
* sibling is empty and is now the last child of the parent, recurse to
* try to delete it. (These cases cannot apply at the same time, though
* the second case might itself recurse to the first.)
*
* When recursing to parent, we hold the lock on the target page until
* done. This delays any insertions into the keyspace that was just
* effectively reassigned to the parent's right sibling. If we allowed
* that, and there were enough such insertions before we finish deleting
* the parent, page splits within that keyspace could lead to inserting
* out-of-order keys into the grandparent level. It is thought that that
* wouldn't have any serious consequences, but it still seems like a
* pretty bad idea.
*/
if (parent_half_dead)
{
/* recursive call will release pbuf */
_bt_relbuf(rel, rbuf);
result = _bt_pagedel(rel, pbuf, stack->bts_parent) + 1;
_bt_relbuf(rel, buf);
}
else if (parent_one_child && rightsib_empty)
{
_bt_relbuf(rel, pbuf);
_bt_relbuf(rel, buf);
/* recursive call will release rbuf */
result = _bt_pagedel(rel, rbuf, stack) + 1;
}
else
{
_bt_relbuf(rel, pbuf);
_bt_relbuf(rel, buf);
_bt_relbuf(rel, rbuf);
result = 1;
}
return result;
}
/*
* visibilitymap_set - set a bit on a previously pinned page
*
* recptr is the LSN of the XLOG record we're replaying, if we're in recovery,
* or InvalidXLogRecPtr in normal running. The page LSN is advanced to the
* one provided; in normal running, we generate a new XLOG record and set the
* page LSN to that value. cutoff_xid is the largest xmin on the page being
* marked all-visible; it is needed for Hot Standby, and can be
* InvalidTransactionId if the page contains no tuples.
*
* Caller is expected to set the heap page's PD_ALL_VISIBLE bit before calling
* this function. Except in recovery, caller should also pass the heap
* buffer. When checksums are enabled and we're not in recovery, we must add
* the heap buffer to the WAL chain to protect it from being torn.
*
* You must pass a buffer containing the correct map page to this function.
* Call visibilitymap_pin first to pin the right one. This function doesn't do
* any I/O.
*/
void
visibilitymap_set(Relation rel, BlockNumber heapBlk, Buffer heapBuf,
XLogRecPtr recptr, Buffer vmBuf, TransactionId cutoff_xid)
{
BlockNumber mapBlock = HEAPBLK_TO_MAPBLOCK(heapBlk);
uint32 mapByte = HEAPBLK_TO_MAPBYTE(heapBlk);
uint8 mapBit = HEAPBLK_TO_MAPBIT(heapBlk);
Page page;
char *map;
#ifdef TRACE_VISIBILITYMAP
elog(DEBUG1, "vm_set %s %d", RelationGetRelationName(rel), heapBlk);
#endif
Assert(InRecovery || XLogRecPtrIsInvalid(recptr));
Assert(InRecovery || BufferIsValid(heapBuf));
/* Check that we have the right heap page pinned, if present */
if (BufferIsValid(heapBuf) && BufferGetBlockNumber(heapBuf) != heapBlk)
elog(ERROR, "wrong heap buffer passed to visibilitymap_set");
/* Check that we have the right VM page pinned */
if (!BufferIsValid(vmBuf) || BufferGetBlockNumber(vmBuf) != mapBlock)
elog(ERROR, "wrong VM buffer passed to visibilitymap_set");
page = BufferGetPage(vmBuf);
map = PageGetContents(page);
LockBuffer(vmBuf, BUFFER_LOCK_EXCLUSIVE);
if (!(map[mapByte] & (1 << mapBit)))
{
START_CRIT_SECTION();
map[mapByte] |= (1 << mapBit);
MarkBufferDirty(vmBuf);
if (RelationNeedsWAL(rel))
{
if (XLogRecPtrIsInvalid(recptr))
{
Assert(!InRecovery);
recptr = log_heap_visible(rel->rd_node, heapBuf, vmBuf,
cutoff_xid);
/*
* If data checksums are enabled (or wal_log_hints=on), we
* need to protect the heap page from being torn.
*/
if (XLogHintBitIsNeeded())
{
Page heapPage = BufferGetPage(heapBuf);
/* caller is expected to set PD_ALL_VISIBLE first */
Assert(PageIsAllVisible(heapPage));
PageSetLSN(heapPage, recptr);
}
}
PageSetLSN(page, recptr);
}
END_CRIT_SECTION();
}
LockBuffer(vmBuf, BUFFER_LOCK_UNLOCK);
}
/*
* Main entry point to GiST index build. Initially calls insert over and over,
* but switches to more efficient buffering build algorithm after a certain
* number of tuples (unless buffering mode is disabled).
*/
IndexBuildResult *
gistbuild(Relation heap, Relation index, IndexInfo *indexInfo)
{
IndexBuildResult *result;
double reltuples;
GISTBuildState buildstate;
Buffer buffer;
Page page;
MemoryContext oldcxt = CurrentMemoryContext;
int fillfactor;
buildstate.indexrel = index;
if (index->rd_options)
{
/* Get buffering mode from the options string */
GiSTOptions *options = (GiSTOptions *) index->rd_options;
char *bufferingMode = (char *) options + options->bufferingModeOffset;
if (strcmp(bufferingMode, "on") == 0)
buildstate.bufferingMode = GIST_BUFFERING_STATS;
else if (strcmp(bufferingMode, "off") == 0)
buildstate.bufferingMode = GIST_BUFFERING_DISABLED;
else
buildstate.bufferingMode = GIST_BUFFERING_AUTO;
fillfactor = options->fillfactor;
}
else
{
/*
* By default, switch to buffering mode when the index grows too large
* to fit in cache.
*/
buildstate.bufferingMode = GIST_BUFFERING_AUTO;
fillfactor = GIST_DEFAULT_FILLFACTOR;
}
/* Calculate target amount of free space to leave on pages */
buildstate.freespace = BLCKSZ * (100 - fillfactor) / 100;
/*
* We expect to be called exactly once for any index relation. If that's
* not the case, big trouble's what we have.
*/
if (RelationGetNumberOfBlocks(index) != 0)
elog(ERROR, "index \"%s\" already contains data",
RelationGetRelationName(index));
/* no locking is needed */
buildstate.giststate = initGISTstate(index);
/*
* Create a temporary memory context that is reset once for each tuple
* processed. (Note: we don't bother to make this a child of the
* giststate's scanCxt, so we have to delete it separately at the end.)
*/
buildstate.giststate->tempCxt = createTempGistContext();
/* initialize the root page */
buffer = gistNewBuffer(index);
Assert(BufferGetBlockNumber(buffer) == GIST_ROOT_BLKNO);
page = BufferGetPage(buffer, NULL, NULL, BGP_NO_SNAPSHOT_TEST);
START_CRIT_SECTION();
GISTInitBuffer(buffer, F_LEAF);
MarkBufferDirty(buffer);
if (RelationNeedsWAL(index))
{
XLogRecPtr recptr;
XLogBeginInsert();
XLogRegisterBuffer(0, buffer, REGBUF_WILL_INIT);
recptr = XLogInsert(RM_GIST_ID, XLOG_GIST_CREATE_INDEX);
PageSetLSN(page, recptr);
}
else
PageSetLSN(page, gistGetFakeLSN(heap));
UnlockReleaseBuffer(buffer);
END_CRIT_SECTION();
/* build the index */
buildstate.indtuples = 0;
buildstate.indtuplesSize = 0;
/*
* Do the heap scan.
*/
reltuples = IndexBuildHeapScan(heap, index, indexInfo, true,
gistBuildCallback, (void *) &buildstate);
/*
* If buffering was used, flush out all the tuples that are still in the
* buffers.
*/
if (buildstate.bufferingMode == GIST_BUFFERING_ACTIVE)
{
elog(DEBUG1, "all tuples processed, emptying buffers");
gistEmptyAllBuffers(&buildstate);
gistFreeBuildBuffers(buildstate.gfbb);
}
/* okay, all heap tuples are indexed */
MemoryContextSwitchTo(oldcxt);
MemoryContextDelete(buildstate.giststate->tempCxt);
freeGISTstate(buildstate.giststate);
/*
* Return statistics
*/
result = (IndexBuildResult *) palloc(sizeof(IndexBuildResult));
result->heap_tuples = reltuples;
result->index_tuples = (double) buildstate.indtuples;
return result;
}
/*
* Build a pending-list page from the given array of tuples, and write it out.
*
* Returns amount of free space left on the page.
*/
static int32
writeListPage(Relation index, Buffer buffer,
IndexTuple *tuples, int32 ntuples, BlockNumber rightlink)
{
Page page = BufferGetPage(buffer);
int32 i,
freesize,
size = 0;
OffsetNumber l,
off;
char *workspace;
char *ptr;
/* workspace could be a local array; we use palloc for alignment */
workspace = palloc(BLCKSZ);
START_CRIT_SECTION();
GinInitBuffer(buffer, GIN_LIST);
off = FirstOffsetNumber;
ptr = workspace;
for (i = 0; i < ntuples; i++)
{
int this_size = IndexTupleSize(tuples[i]);
memcpy(ptr, tuples[i], this_size);
ptr += this_size;
size += this_size;
l = PageAddItem(page, (Item) tuples[i], this_size, off, false, false);
if (l == InvalidOffsetNumber)
elog(ERROR, "failed to add item to index page in \"%s\"",
RelationGetRelationName(index));
off++;
}
Assert(size <= BLCKSZ); /* else we overran workspace */
GinPageGetOpaque(page)->rightlink = rightlink;
/*
* tail page may contain only whole row(s) or final part of row placed on
* previous pages (a "row" here meaning all the index tuples generated for
* one heap tuple)
*/
if (rightlink == InvalidBlockNumber)
{
GinPageSetFullRow(page);
GinPageGetOpaque(page)->maxoff = 1;
}
else
{
GinPageGetOpaque(page)->maxoff = 0;
}
MarkBufferDirty(buffer);
if (RelationNeedsWAL(index))
{
ginxlogInsertListPage data;
XLogRecPtr recptr;
data.rightlink = rightlink;
data.ntuples = ntuples;
XLogBeginInsert();
XLogRegisterData((char *) &data, sizeof(ginxlogInsertListPage));
XLogRegisterBuffer(0, buffer, REGBUF_WILL_INIT);
XLogRegisterBufData(0, workspace, size);
recptr = XLogInsert(RM_GIN_ID, XLOG_GIN_INSERT_LISTPAGE);
PageSetLSN(page, recptr);
}
/* get free space before releasing buffer */
freesize = PageGetExactFreeSpace(page);
UnlockReleaseBuffer(buffer);
END_CRIT_SECTION();
pfree(workspace);
return freesize;
}
/*
* Place tuples from 'itup' to 'buffer'. If 'oldoffnum' is valid, the tuple
* at that offset is atomically removed along with inserting the new tuples.
* This is used to replace a tuple with a new one.
*
* If 'leftchildbuf' is valid, we're inserting the downlink for the page
* to the right of 'leftchildbuf', or updating the downlink for 'leftchildbuf'.
* F_FOLLOW_RIGHT flag on 'leftchildbuf' is cleared and NSN is set.
*
* If 'markfollowright' is true and the page is split, the left child is
* marked with F_FOLLOW_RIGHT flag. That is the normal case. During buffered
* index build, however, there is no concurrent access and the page splitting
* is done in a slightly simpler fashion, and false is passed.
*
* If there is not enough room on the page, it is split. All the split
* pages are kept pinned and locked and returned in *splitinfo, the caller
* is responsible for inserting the downlinks for them. However, if
* 'buffer' is the root page and it needs to be split, gistplacetopage()
* performs the split as one atomic operation, and *splitinfo is set to NIL.
* In that case, we continue to hold the root page locked, and the child
* pages are released; note that new tuple(s) are *not* on the root page
* but in one of the new child pages.
*
* If 'newblkno' is not NULL, returns the block number of page the first
* new/updated tuple was inserted to. Usually it's the given page, but could
* be its right sibling if the page was split.
*
* Returns 'true' if the page was split, 'false' otherwise.
*/
bool
gistplacetopage(Relation rel, Size freespace, GISTSTATE *giststate,
Buffer buffer,
IndexTuple *itup, int ntup, OffsetNumber oldoffnum,
BlockNumber *newblkno,
Buffer leftchildbuf,
List **splitinfo,
bool markfollowright,
Relation heapRel,
bool is_build)
{
BlockNumber blkno = BufferGetBlockNumber(buffer);
Page page = BufferGetPage(buffer);
bool is_leaf = (GistPageIsLeaf(page)) ? true : false;
XLogRecPtr recptr;
int i;
bool is_split;
/*
* Refuse to modify a page that's incompletely split. This should not
* happen because we finish any incomplete splits while we walk down the
* tree. However, it's remotely possible that another concurrent inserter
* splits a parent page, and errors out before completing the split. We
* will just throw an error in that case, and leave any split we had in
* progress unfinished too. The next insert that comes along will clean up
* the mess.
*/
if (GistFollowRight(page))
elog(ERROR, "concurrent GiST page split was incomplete");
*splitinfo = NIL;
/*
* if isupdate, remove old key: This node's key has been modified, either
* because a child split occurred or because we needed to adjust our key
* for an insert in a child node. Therefore, remove the old version of
* this node's key.
*
* for WAL replay, in the non-split case we handle this by setting up a
* one-element todelete array; in the split case, it's handled implicitly
* because the tuple vector passed to gistSplit won't include this tuple.
*/
is_split = gistnospace(page, itup, ntup, oldoffnum, freespace);
/*
* If leaf page is full, try at first to delete dead tuples. And then
* check again.
*/
if (is_split && GistPageIsLeaf(page) && GistPageHasGarbage(page))
{
gistprunepage(rel, page, buffer, heapRel);
is_split = gistnospace(page, itup, ntup, oldoffnum, freespace);
}
if (is_split)
{
/* no space for insertion */
IndexTuple *itvec;
int tlen;
SplitedPageLayout *dist = NULL,
*ptr;
BlockNumber oldrlink = InvalidBlockNumber;
GistNSN oldnsn = 0;
SplitedPageLayout rootpg;
bool is_rootsplit;
int npage;
is_rootsplit = (blkno == GIST_ROOT_BLKNO);
/*
* Form index tuples vector to split. If we're replacing an old tuple,
* remove the old version from the vector.
*/
itvec = gistextractpage(page, &tlen);
if (OffsetNumberIsValid(oldoffnum))
{
/* on inner page we should remove old tuple */
int pos = oldoffnum - FirstOffsetNumber;
tlen--;
if (pos != tlen)
memmove(itvec + pos, itvec + pos + 1, sizeof(IndexTuple) * (tlen - pos));
}
itvec = gistjoinvector(itvec, &tlen, itup, ntup);
dist = gistSplit(rel, page, itvec, tlen, giststate);
/*
* Check that split didn't produce too many pages.
*/
npage = 0;
for (ptr = dist; ptr; ptr = ptr->next)
npage++;
/* in a root split, we'll add one more page to the list below */
if (is_rootsplit)
npage++;
if (npage > GIST_MAX_SPLIT_PAGES)
elog(ERROR, "GiST page split into too many halves (%d, maximum %d)",
npage, GIST_MAX_SPLIT_PAGES);
/*
* Set up pages to work with. Allocate new buffers for all but the
* leftmost page. The original page becomes the new leftmost page, and
* is just replaced with the new contents.
*
* For a root-split, allocate new buffers for all child pages, the
* original page is overwritten with new root page containing
* downlinks to the new child pages.
*/
ptr = dist;
if (!is_rootsplit)
{
/* save old rightlink and NSN */
oldrlink = GistPageGetOpaque(page)->rightlink;
oldnsn = GistPageGetNSN(page);
dist->buffer = buffer;
dist->block.blkno = BufferGetBlockNumber(buffer);
dist->page = PageGetTempPageCopySpecial(BufferGetPage(buffer));
/* clean all flags except F_LEAF */
GistPageGetOpaque(dist->page)->flags = (is_leaf) ? F_LEAF : 0;
ptr = ptr->next;
}
for (; ptr; ptr = ptr->next)
{
/* Allocate new page */
ptr->buffer = gistNewBuffer(rel);
GISTInitBuffer(ptr->buffer, (is_leaf) ? F_LEAF : 0);
ptr->page = BufferGetPage(ptr->buffer);
ptr->block.blkno = BufferGetBlockNumber(ptr->buffer);
PredicateLockPageSplit(rel,
BufferGetBlockNumber(buffer),
BufferGetBlockNumber(ptr->buffer));
}
/*
* Now that we know which blocks the new pages go to, set up downlink
* tuples to point to them.
*/
for (ptr = dist; ptr; ptr = ptr->next)
{
ItemPointerSetBlockNumber(&(ptr->itup->t_tid), ptr->block.blkno);
GistTupleSetValid(ptr->itup);
}
/*
* If this is a root split, we construct the new root page with the
* downlinks here directly, instead of requiring the caller to insert
* them. Add the new root page to the list along with the child pages.
*/
if (is_rootsplit)
{
IndexTuple *downlinks;
int ndownlinks = 0;
int i;
rootpg.buffer = buffer;
rootpg.page = PageGetTempPageCopySpecial(BufferGetPage(rootpg.buffer));
GistPageGetOpaque(rootpg.page)->flags = 0;
/* Prepare a vector of all the downlinks */
for (ptr = dist; ptr; ptr = ptr->next)
ndownlinks++;
downlinks = palloc(sizeof(IndexTuple) * ndownlinks);
for (i = 0, ptr = dist; ptr; ptr = ptr->next)
downlinks[i++] = ptr->itup;
rootpg.block.blkno = GIST_ROOT_BLKNO;
rootpg.block.num = ndownlinks;
rootpg.list = gistfillitupvec(downlinks, ndownlinks,
&(rootpg.lenlist));
rootpg.itup = NULL;
rootpg.next = dist;
dist = &rootpg;
}
else
{
/* Prepare split-info to be returned to caller */
for (ptr = dist; ptr; ptr = ptr->next)
{
GISTPageSplitInfo *si = palloc(sizeof(GISTPageSplitInfo));
si->buf = ptr->buffer;
si->downlink = ptr->itup;
*splitinfo = lappend(*splitinfo, si);
}
}
/*
* Fill all pages. All the pages are new, ie. freshly allocated empty
* pages, or a temporary copy of the old page.
*/
for (ptr = dist; ptr; ptr = ptr->next)
{
char *data = (char *) (ptr->list);
for (i = 0; i < ptr->block.num; i++)
{
IndexTuple thistup = (IndexTuple) data;
if (PageAddItem(ptr->page, (Item) data, IndexTupleSize(thistup), i + FirstOffsetNumber, false, false) == InvalidOffsetNumber)
elog(ERROR, "failed to add item to index page in \"%s\"", RelationGetRelationName(rel));
/*
* If this is the first inserted/updated tuple, let the caller
* know which page it landed on.
*/
if (newblkno && ItemPointerEquals(&thistup->t_tid, &(*itup)->t_tid))
*newblkno = ptr->block.blkno;
data += IndexTupleSize(thistup);
}
/* Set up rightlinks */
if (ptr->next && ptr->block.blkno != GIST_ROOT_BLKNO)
GistPageGetOpaque(ptr->page)->rightlink =
ptr->next->block.blkno;
else
GistPageGetOpaque(ptr->page)->rightlink = oldrlink;
/*
* Mark the all but the right-most page with the follow-right
* flag. It will be cleared as soon as the downlink is inserted
* into the parent, but this ensures that if we error out before
* that, the index is still consistent. (in buffering build mode,
* any error will abort the index build anyway, so this is not
* needed.)
*/
if (ptr->next && !is_rootsplit && markfollowright)
GistMarkFollowRight(ptr->page);
else
GistClearFollowRight(ptr->page);
/*
* Copy the NSN of the original page to all pages. The
* F_FOLLOW_RIGHT flags ensure that scans will follow the
* rightlinks until the downlinks are inserted.
*/
GistPageSetNSN(ptr->page, oldnsn);
}
/*
* gistXLogSplit() needs to WAL log a lot of pages, prepare WAL
* insertion for that. NB: The number of pages and data segments
* specified here must match the calculations in gistXLogSplit()!
*/
if (!is_build && RelationNeedsWAL(rel))
XLogEnsureRecordSpace(npage, 1 + npage * 2);
START_CRIT_SECTION();
/*
* Must mark buffers dirty before XLogInsert, even though we'll still
* be changing their opaque fields below.
*/
for (ptr = dist; ptr; ptr = ptr->next)
MarkBufferDirty(ptr->buffer);
if (BufferIsValid(leftchildbuf))
MarkBufferDirty(leftchildbuf);
/*
* The first page in the chain was a temporary working copy meant to
* replace the old page. Copy it over the old page.
*/
PageRestoreTempPage(dist->page, BufferGetPage(dist->buffer));
dist->page = BufferGetPage(dist->buffer);
/*
* Write the WAL record.
*
* If we're building a new index, however, we don't WAL-log changes
* yet. The LSN-NSN interlock between parent and child requires that
* LSNs never move backwards, so set the LSNs to a value that's
* smaller than any real or fake unlogged LSN that might be generated
* later. (There can't be any concurrent scans during index build, so
* we don't need to be able to detect concurrent splits yet.)
*/
if (is_build)
recptr = GistBuildLSN;
else
{
if (RelationNeedsWAL(rel))
recptr = gistXLogSplit(is_leaf,
dist, oldrlink, oldnsn, leftchildbuf,
markfollowright);
else
recptr = gistGetFakeLSN(rel);
}
for (ptr = dist; ptr; ptr = ptr->next)
PageSetLSN(ptr->page, recptr);
/*
* Return the new child buffers to the caller.
*
* If this was a root split, we've already inserted the downlink
* pointers, in the form of a new root page. Therefore we can release
* all the new buffers, and keep just the root page locked.
*/
if (is_rootsplit)
{
for (ptr = dist->next; ptr; ptr = ptr->next)
UnlockReleaseBuffer(ptr->buffer);
}
}
else
{
/*
* Enough space. We always get here if ntup==0.
*/
START_CRIT_SECTION();
/*
* Delete old tuple if any, then insert new tuple(s) if any. If
* possible, use the fast path of PageIndexTupleOverwrite.
*/
if (OffsetNumberIsValid(oldoffnum))
{
if (ntup == 1)
{
/* One-for-one replacement, so use PageIndexTupleOverwrite */
if (!PageIndexTupleOverwrite(page, oldoffnum, (Item) *itup,
IndexTupleSize(*itup)))
elog(ERROR, "failed to add item to index page in \"%s\"",
RelationGetRelationName(rel));
}
else
{
/* Delete old, then append new tuple(s) to page */
PageIndexTupleDelete(page, oldoffnum);
gistfillbuffer(page, itup, ntup, InvalidOffsetNumber);
}
}
else
{
/* Just append new tuples at the end of the page */
gistfillbuffer(page, itup, ntup, InvalidOffsetNumber);
}
MarkBufferDirty(buffer);
if (BufferIsValid(leftchildbuf))
MarkBufferDirty(leftchildbuf);
if (is_build)
recptr = GistBuildLSN;
else
{
if (RelationNeedsWAL(rel))
{
OffsetNumber ndeloffs = 0,
deloffs[1];
if (OffsetNumberIsValid(oldoffnum))
{
deloffs[0] = oldoffnum;
ndeloffs = 1;
}
recptr = gistXLogUpdate(buffer,
deloffs, ndeloffs, itup, ntup,
leftchildbuf);
}
else
recptr = gistGetFakeLSN(rel);
}
PageSetLSN(page, recptr);
if (newblkno)
*newblkno = blkno;
}
/*
* If we inserted the downlink for a child page, set NSN and clear
* F_FOLLOW_RIGHT flag on the left child, so that concurrent scans know to
* follow the rightlink if and only if they looked at the parent page
* before we inserted the downlink.
*
* Note that we do this *after* writing the WAL record. That means that
* the possible full page image in the WAL record does not include these
* changes, and they must be replayed even if the page is restored from
* the full page image. There's a chicken-and-egg problem: if we updated
* the child pages first, we wouldn't know the recptr of the WAL record
* we're about to write.
*/
if (BufferIsValid(leftchildbuf))
{
Page leftpg = BufferGetPage(leftchildbuf);
GistPageSetNSN(leftpg, recptr);
GistClearFollowRight(leftpg);
PageSetLSN(leftpg, recptr);
}
END_CRIT_SECTION();
return is_split;
}
/*
* Prune and repair fragmentation in the specified page.
*
* Caller must have pin and buffer cleanup lock on the page.
*
* OldestXmin is the cutoff XID used to distinguish whether tuples are DEAD
* or RECENTLY_DEAD (see HeapTupleSatisfiesVacuum).
*
* If report_stats is true then we send the number of reclaimed heap-only
* tuples to pgstats. (This must be FALSE during vacuum, since vacuum will
* send its own new total to pgstats, and we don't want this delta applied
* on top of that.)
*
* Returns the number of tuples deleted from the page and sets
* latestRemovedXid.
*/
int
heap_page_prune(Relation relation, Buffer buffer, TransactionId OldestXmin,
bool report_stats, TransactionId *latestRemovedXid)
{
int ndeleted = 0;
Page page = BufferGetPage(buffer);
OffsetNumber offnum,
maxoff;
PruneState prstate;
/*
* Our strategy is to scan the page and make lists of items to change,
* then apply the changes within a critical section. This keeps as much
* logic as possible out of the critical section, and also ensures that
* WAL replay will work the same as the normal case.
*
* First, initialize the new pd_prune_xid value to zero (indicating no
* prunable tuples). If we find any tuples which may soon become
* prunable, we will save the lowest relevant XID in new_prune_xid. Also
* initialize the rest of our working state.
*/
prstate.new_prune_xid = InvalidTransactionId;
prstate.latestRemovedXid = *latestRemovedXid;
prstate.nredirected = prstate.ndead = prstate.nunused = 0;
memset(prstate.marked, 0, sizeof(prstate.marked));
/* Scan the page */
maxoff = PageGetMaxOffsetNumber(page);
for (offnum = FirstOffsetNumber;
offnum <= maxoff;
offnum = OffsetNumberNext(offnum))
{
ItemId itemid;
/* Ignore items already processed as part of an earlier chain */
if (prstate.marked[offnum])
continue;
/* Nothing to do if slot is empty or already dead */
itemid = PageGetItemId(page, offnum);
if (!ItemIdIsUsed(itemid) || ItemIdIsDead(itemid))
continue;
/* Process this item or chain of items */
ndeleted += heap_prune_chain(relation, buffer, offnum,
OldestXmin,
&prstate);
}
/* Any error while applying the changes is critical */
START_CRIT_SECTION();
/* Have we found any prunable items? */
if (prstate.nredirected > 0 || prstate.ndead > 0 || prstate.nunused > 0)
{
/*
* Apply the planned item changes, then repair page fragmentation, and
* update the page's hint bit about whether it has free line pointers.
*/
heap_page_prune_execute(buffer,
prstate.redirected, prstate.nredirected,
prstate.nowdead, prstate.ndead,
prstate.nowunused, prstate.nunused);
/*
* Update the page's pd_prune_xid field to either zero, or the lowest
* XID of any soon-prunable tuple.
*/
((PageHeader) page)->pd_prune_xid = prstate.new_prune_xid;
/*
* Also clear the "page is full" flag, since there's no point in
* repeating the prune/defrag process until something else happens to
* the page.
*/
PageClearFull(page);
MarkBufferDirty(buffer);
/*
* Emit a WAL HEAP_CLEAN record showing what we did
*/
if (RelationNeedsWAL(relation))
{
XLogRecPtr recptr;
recptr = log_heap_clean(relation, buffer,
prstate.redirected, prstate.nredirected,
prstate.nowdead, prstate.ndead,
prstate.nowunused, prstate.nunused,
prstate.latestRemovedXid);
PageSetLSN(BufferGetPage(buffer), recptr);
}
}
else
{
/*
* If we didn't prune anything, but have found a new value for the
* pd_prune_xid field, update it and mark the buffer dirty. This is
* treated as a non-WAL-logged hint.
*
* Also clear the "page is full" flag if it is set, since there's no
* point in repeating the prune/defrag process until something else
* happens to the page.
*/
if (((PageHeader) page)->pd_prune_xid != prstate.new_prune_xid ||
PageIsFull(page))
{
((PageHeader) page)->pd_prune_xid = prstate.new_prune_xid;
PageClearFull(page);
MarkBufferDirtyHint(buffer, true);
}
}
END_CRIT_SECTION();
/*
* If requested, report the number of tuples reclaimed to pgstats. This is
* ndeleted minus ndead, because we don't want to count a now-DEAD root
* item as a deletion for this purpose.
*/
if (report_stats && ndeleted > prstate.ndead)
pgstat_update_heap_dead_tuples(relation, ndeleted - prstate.ndead);
*latestRemovedXid = prstate.latestRemovedXid;
/*
* XXX Should we update the FSM information of this page ?
*
* There are two schools of thought here. We may not want to update FSM
* information so that the page is not used for unrelated UPDATEs/INSERTs
* and any free space in this page will remain available for further
* UPDATEs in *this* page, thus improving chances for doing HOT updates.
*
* But for a large table and where a page does not receive further UPDATEs
* for a long time, we might waste this space by not updating the FSM
* information. The relation may get extended and fragmented further.
*
* One possibility is to leave "fillfactor" worth of space in this page
* and update FSM with the remaining space.
*/
return ndeleted;
}
/*
* Write the index tuples contained in *collector into the index's
* pending list.
*
* Function guarantees that all these tuples will be inserted consecutively,
* preserving order
*/
void
ginHeapTupleFastInsert(GinState *ginstate, GinTupleCollector *collector)
{
Relation index = ginstate->index;
Buffer metabuffer;
Page metapage;
GinMetaPageData *metadata = NULL;
Buffer buffer = InvalidBuffer;
Page page = NULL;
ginxlogUpdateMeta data;
bool separateList = false;
bool needCleanup = false;
int cleanupSize;
bool needWal;
if (collector->ntuples == 0)
return;
needWal = RelationNeedsWAL(index);
data.node = index->rd_node;
data.ntuples = 0;
data.newRightlink = data.prevTail = InvalidBlockNumber;
metabuffer = ReadBuffer(index, GIN_METAPAGE_BLKNO);
metapage = BufferGetPage(metabuffer);
if (collector->sumsize + collector->ntuples * sizeof(ItemIdData) > GinListPageSize)
{
/*
* Total size is greater than one page => make sublist
*/
separateList = true;
}
else
{
LockBuffer(metabuffer, GIN_EXCLUSIVE);
metadata = GinPageGetMeta(metapage);
if (metadata->head == InvalidBlockNumber ||
collector->sumsize + collector->ntuples * sizeof(ItemIdData) > metadata->tailFreeSize)
{
/*
* Pending list is empty or total size is greater than freespace
* on tail page => make sublist
*
* We unlock metabuffer to keep high concurrency
*/
separateList = true;
LockBuffer(metabuffer, GIN_UNLOCK);
}
}
if (separateList)
{
/*
* We should make sublist separately and append it to the tail
*/
GinMetaPageData sublist;
memset(&sublist, 0, sizeof(GinMetaPageData));
makeSublist(index, collector->tuples, collector->ntuples, &sublist);
if (needWal)
XLogBeginInsert();
/*
* metapage was unlocked, see above
*/
LockBuffer(metabuffer, GIN_EXCLUSIVE);
metadata = GinPageGetMeta(metapage);
if (metadata->head == InvalidBlockNumber)
{
/*
* Main list is empty, so just insert sublist as main list
*/
START_CRIT_SECTION();
metadata->head = sublist.head;
metadata->tail = sublist.tail;
metadata->tailFreeSize = sublist.tailFreeSize;
metadata->nPendingPages = sublist.nPendingPages;
metadata->nPendingHeapTuples = sublist.nPendingHeapTuples;
}
else
{
/*
* Merge lists
*/
data.prevTail = metadata->tail;
data.newRightlink = sublist.head;
buffer = ReadBuffer(index, metadata->tail);
LockBuffer(buffer, GIN_EXCLUSIVE);
page = BufferGetPage(buffer);
Assert(GinPageGetOpaque(page)->rightlink == InvalidBlockNumber);
START_CRIT_SECTION();
GinPageGetOpaque(page)->rightlink = sublist.head;
MarkBufferDirty(buffer);
metadata->tail = sublist.tail;
metadata->tailFreeSize = sublist.tailFreeSize;
metadata->nPendingPages += sublist.nPendingPages;
metadata->nPendingHeapTuples += sublist.nPendingHeapTuples;
if (needWal)
XLogRegisterBuffer(1, buffer, REGBUF_STANDARD);
}
}
else
{
/*
* Insert into tail page. Metapage is already locked
*/
OffsetNumber l,
off;
int i,
tupsize;
char *ptr;
char *collectordata;
buffer = ReadBuffer(index, metadata->tail);
LockBuffer(buffer, GIN_EXCLUSIVE);
page = BufferGetPage(buffer);
off = (PageIsEmpty(page)) ? FirstOffsetNumber :
OffsetNumberNext(PageGetMaxOffsetNumber(page));
collectordata = ptr = (char *) palloc(collector->sumsize);
data.ntuples = collector->ntuples;
if (needWal)
XLogBeginInsert();
START_CRIT_SECTION();
/*
* Increase counter of heap tuples
*/
Assert(GinPageGetOpaque(page)->maxoff <= metadata->nPendingHeapTuples);
GinPageGetOpaque(page)->maxoff++;
metadata->nPendingHeapTuples++;
for (i = 0; i < collector->ntuples; i++)
{
tupsize = IndexTupleSize(collector->tuples[i]);
l = PageAddItem(page, (Item) collector->tuples[i], tupsize, off, false, false);
if (l == InvalidOffsetNumber)
elog(ERROR, "failed to add item to index page in \"%s\"",
RelationGetRelationName(index));
memcpy(ptr, collector->tuples[i], tupsize);
ptr += tupsize;
off++;
}
Assert((ptr - collectordata) <= collector->sumsize);
if (needWal)
{
XLogRegisterBuffer(1, buffer, REGBUF_STANDARD);
XLogRegisterBufData(1, collectordata, collector->sumsize);
}
metadata->tailFreeSize = PageGetExactFreeSpace(page);
MarkBufferDirty(buffer);
}
/*
* Set pd_lower just past the end of the metadata. This is essential,
* because without doing so, metadata will be lost if xlog.c compresses
* the page. (We must do this here because pre-v11 versions of PG did not
* set the metapage's pd_lower correctly, so a pg_upgraded index might
* contain the wrong value.)
*/
((PageHeader) metapage)->pd_lower =
((char *) metadata + sizeof(GinMetaPageData)) - (char *) metapage;
/*
* Write metabuffer, make xlog entry
*/
MarkBufferDirty(metabuffer);
if (needWal)
{
XLogRecPtr recptr;
memcpy(&data.metadata, metadata, sizeof(GinMetaPageData));
XLogRegisterBuffer(0, metabuffer, REGBUF_WILL_INIT | REGBUF_STANDARD);
XLogRegisterData((char *) &data, sizeof(ginxlogUpdateMeta));
recptr = XLogInsert(RM_GIN_ID, XLOG_GIN_UPDATE_META_PAGE);
PageSetLSN(metapage, recptr);
if (buffer != InvalidBuffer)
{
PageSetLSN(page, recptr);
}
}
if (buffer != InvalidBuffer)
UnlockReleaseBuffer(buffer);
/*
* Force pending list cleanup when it becomes too long. And,
* ginInsertCleanup could take significant amount of time, so we prefer to
* call it when it can do all the work in a single collection cycle. In
* non-vacuum mode, it shouldn't require maintenance_work_mem, so fire it
* while pending list is still small enough to fit into
* gin_pending_list_limit.
*
* ginInsertCleanup() should not be called inside our CRIT_SECTION.
*/
cleanupSize = GinGetPendingListCleanupSize(index);
if (metadata->nPendingPages * GIN_PAGE_FREESIZE > cleanupSize * 1024L)
needCleanup = true;
UnlockReleaseBuffer(metabuffer);
END_CRIT_SECTION();
/*
* Since it could contend with concurrent cleanup process we cleanup
* pending list not forcibly.
*/
if (needCleanup)
ginInsertCleanup(ginstate, false, true, false, NULL);
}
/*
* brinbuild() -- build a new BRIN index.
*/
IndexBuildResult *
brinbuild(Relation heap, Relation index, IndexInfo *indexInfo)
{
IndexBuildResult *result;
double reltuples;
double idxtuples;
BrinRevmap *revmap;
BrinBuildState *state;
Buffer meta;
BlockNumber pagesPerRange;
/*
* We expect to be called exactly once for any index relation.
*/
if (RelationGetNumberOfBlocks(index) != 0)
elog(ERROR, "index \"%s\" already contains data",
RelationGetRelationName(index));
/*
* Critical section not required, because on error the creation of the
* whole relation will be rolled back.
*/
meta = ReadBuffer(index, P_NEW);
Assert(BufferGetBlockNumber(meta) == BRIN_METAPAGE_BLKNO);
LockBuffer(meta, BUFFER_LOCK_EXCLUSIVE);
brin_metapage_init(BufferGetPage(meta), BrinGetPagesPerRange(index),
BRIN_CURRENT_VERSION);
MarkBufferDirty(meta);
if (RelationNeedsWAL(index))
{
xl_brin_createidx xlrec;
XLogRecPtr recptr;
Page page;
xlrec.version = BRIN_CURRENT_VERSION;
xlrec.pagesPerRange = BrinGetPagesPerRange(index);
XLogBeginInsert();
XLogRegisterData((char *) &xlrec, SizeOfBrinCreateIdx);
XLogRegisterBuffer(0, meta, REGBUF_WILL_INIT | REGBUF_STANDARD);
recptr = XLogInsert(RM_BRIN_ID, XLOG_BRIN_CREATE_INDEX);
page = BufferGetPage(meta);
PageSetLSN(page, recptr);
}
UnlockReleaseBuffer(meta);
/*
* Initialize our state, including the deformed tuple state.
*/
revmap = brinRevmapInitialize(index, &pagesPerRange, NULL);
state = initialize_brin_buildstate(index, revmap, pagesPerRange);
/*
* Now scan the relation. No syncscan allowed here because we want the
* heap blocks in physical order.
*/
reltuples = IndexBuildHeapScan(heap, index, indexInfo, false,
brinbuildCallback, (void *) state, NULL);
/* process the final batch */
form_and_insert_tuple(state);
/* release resources */
idxtuples = state->bs_numtuples;
brinRevmapTerminate(state->bs_rmAccess);
terminate_brin_buildstate(state);
/*
* Return statistics
*/
result = (IndexBuildResult *) palloc(sizeof(IndexBuildResult));
result->heap_tuples = reltuples;
result->index_tuples = idxtuples;
return result;
}
/*
* Delete a posting tree page.
*/
static void
ginDeletePage(GinVacuumState *gvs, BlockNumber deleteBlkno, BlockNumber leftBlkno,
BlockNumber parentBlkno, OffsetNumber myoff, bool isParentRoot)
{
Buffer dBuffer;
Buffer lBuffer;
Buffer pBuffer;
Page page,
parentPage;
BlockNumber rightlink;
/*
* Lock the pages in the same order as an insertion would, to avoid
* deadlocks: left, then right, then parent.
*/
lBuffer = ReadBufferExtended(gvs->index, MAIN_FORKNUM, leftBlkno,
RBM_NORMAL, gvs->strategy);
dBuffer = ReadBufferExtended(gvs->index, MAIN_FORKNUM, deleteBlkno,
RBM_NORMAL, gvs->strategy);
pBuffer = ReadBufferExtended(gvs->index, MAIN_FORKNUM, parentBlkno,
RBM_NORMAL, gvs->strategy);
LockBuffer(lBuffer, GIN_EXCLUSIVE);
LockBuffer(dBuffer, GIN_EXCLUSIVE);
if (!isParentRoot) /* parent is already locked by
* LockBufferForCleanup() */
LockBuffer(pBuffer, GIN_EXCLUSIVE);
START_CRIT_SECTION();
/* Unlink the page by changing left sibling's rightlink */
page = BufferGetPage(dBuffer);
rightlink = GinPageGetOpaque(page)->rightlink;
page = BufferGetPage(lBuffer);
GinPageGetOpaque(page)->rightlink = rightlink;
/* Delete downlink from parent */
parentPage = BufferGetPage(pBuffer);
#ifdef USE_ASSERT_CHECKING
do
{
PostingItem *tod = GinDataPageGetPostingItem(parentPage, myoff);
Assert(PostingItemGetBlockNumber(tod) == deleteBlkno);
} while (0);
#endif
GinPageDeletePostingItem(parentPage, myoff);
page = BufferGetPage(dBuffer);
/*
* we shouldn't change rightlink field to save workability of running
* search scan
*/
GinPageGetOpaque(page)->flags = GIN_DELETED;
MarkBufferDirty(pBuffer);
if (leftBlkno != InvalidBlockNumber)
MarkBufferDirty(lBuffer);
MarkBufferDirty(dBuffer);
if (RelationNeedsWAL(gvs->index))
{
XLogRecPtr recptr;
XLogRecData rdata[4];
ginxlogDeletePage data;
int n;
data.node = gvs->index->rd_node;
data.blkno = deleteBlkno;
data.parentBlkno = parentBlkno;
data.parentOffset = myoff;
data.leftBlkno = leftBlkno;
data.rightLink = GinPageGetOpaque(page)->rightlink;
rdata[0].buffer = dBuffer;
rdata[0].buffer_std = FALSE;
rdata[0].data = NULL;
rdata[0].len = 0;
rdata[0].next = rdata + 1;
rdata[1].buffer = pBuffer;
rdata[1].buffer_std = FALSE;
rdata[1].data = NULL;
rdata[1].len = 0;
rdata[1].next = rdata + 2;
if (leftBlkno != InvalidBlockNumber)
{
rdata[2].buffer = lBuffer;
rdata[2].buffer_std = FALSE;
rdata[2].data = NULL;
rdata[2].len = 0;
rdata[2].next = rdata + 3;
n = 3;
}
else
n = 2;
rdata[n].buffer = InvalidBuffer;
rdata[n].buffer_std = FALSE;
rdata[n].len = sizeof(ginxlogDeletePage);
rdata[n].data = (char *) &data;
rdata[n].next = NULL;
recptr = XLogInsert(RM_GIN_ID, XLOG_GIN_DELETE_PAGE, rdata);
PageSetLSN(page, recptr);
PageSetLSN(parentPage, recptr);
if (leftBlkno != InvalidBlockNumber)
{
page = BufferGetPage(lBuffer);
PageSetLSN(page, recptr);
}
}
if (!isParentRoot)
LockBuffer(pBuffer, GIN_UNLOCK);
ReleaseBuffer(pBuffer);
UnlockReleaseBuffer(lBuffer);
UnlockReleaseBuffer(dBuffer);
END_CRIT_SECTION();
gvs->result->pages_deleted++;
}
/*
* RelationTruncate
* Physically truncate a relation to the specified number of blocks.
*
* This includes getting rid of any buffers for the blocks that are to be
* dropped.
*/
void
RelationTruncate(Relation rel, BlockNumber nblocks)
{
bool fsm;
bool vm;
/* Open it at the smgr level if not already done */
RelationOpenSmgr(rel);
/*
* Make sure smgr_targblock etc aren't pointing somewhere past new end
*/
rel->rd_smgr->smgr_targblock = InvalidBlockNumber;
rel->rd_smgr->smgr_fsm_nblocks = InvalidBlockNumber;
rel->rd_smgr->smgr_vm_nblocks = InvalidBlockNumber;
/* Truncate the FSM first if it exists */
fsm = smgrexists(rel->rd_smgr, FSM_FORKNUM);
if (fsm)
FreeSpaceMapTruncateRel(rel, nblocks);
/* Truncate the visibility map too if it exists. */
vm = smgrexists(rel->rd_smgr, VISIBILITYMAP_FORKNUM);
if (vm)
visibilitymap_truncate(rel, nblocks);
/*
* We WAL-log the truncation before actually truncating, which means
* trouble if the truncation fails. If we then crash, the WAL replay
* likely isn't going to succeed in the truncation either, and cause a
* PANIC. It's tempting to put a critical section here, but that cure
* would be worse than the disease. It would turn a usually harmless
* failure to truncate, that might spell trouble at WAL replay, into a
* certain PANIC.
*/
if (RelationNeedsWAL(rel))
{
/*
* Make an XLOG entry reporting the file truncation.
*/
XLogRecPtr lsn;
xl_smgr_truncate xlrec;
xlrec.blkno = nblocks;
xlrec.rnode = rel->rd_node;
xlrec.flags = SMGR_TRUNCATE_ALL;
XLogBeginInsert();
XLogRegisterData((char *) &xlrec, sizeof(xlrec));
lsn = XLogInsert(RM_SMGR_ID,
XLOG_SMGR_TRUNCATE | XLR_SPECIAL_REL_UPDATE);
/*
* Flush, because otherwise the truncation of the main relation might
* hit the disk before the WAL record, and the truncation of the FSM
* or visibility map. If we crashed during that window, we'd be left
* with a truncated heap, but the FSM or visibility map would still
* contain entries for the non-existent heap pages.
*/
if (fsm || vm)
XLogFlush(lsn);
}
/* Do the real work */
smgrtruncate(rel->rd_smgr, MAIN_FORKNUM, nblocks);
}
/*
* Insert a new item to a page.
*
* Returns true if the insertion was finished. On false, the page was split and
* the parent needs to be updated. (A root split returns true as it doesn't
* need any further action by the caller to complete.)
*
* When inserting a downlink to an internal page, 'childbuf' contains the
* child page that was split. Its GIN_INCOMPLETE_SPLIT flag will be cleared
* atomically with the insert. Also, the existing item at offset stack->off
* in the target page is updated to point to updateblkno.
*
* stack->buffer is locked on entry, and is kept locked.
* Likewise for childbuf, if given.
*/
static bool
ginPlaceToPage(GinBtree btree, GinBtreeStack *stack,
void *insertdata, BlockNumber updateblkno,
Buffer childbuf, GinStatsData *buildStats)
{
Page page = BufferGetPage(stack->buffer);
bool result;
GinPlaceToPageRC rc;
uint16 xlflags = 0;
Page childpage = NULL;
Page newlpage = NULL,
newrpage = NULL;
void *ptp_workspace = NULL;
MemoryContext tmpCxt;
MemoryContext oldCxt;
/*
* We do all the work of this function and its subfunctions in a temporary
* memory context. This avoids leakages and simplifies APIs, since some
* subfunctions allocate storage that has to survive until we've finished
* the WAL insertion.
*/
tmpCxt = AllocSetContextCreate(CurrentMemoryContext,
"ginPlaceToPage temporary context",
ALLOCSET_DEFAULT_MINSIZE,
ALLOCSET_DEFAULT_INITSIZE,
ALLOCSET_DEFAULT_MAXSIZE);
oldCxt = MemoryContextSwitchTo(tmpCxt);
if (GinPageIsData(page))
xlflags |= GIN_INSERT_ISDATA;
if (GinPageIsLeaf(page))
{
xlflags |= GIN_INSERT_ISLEAF;
Assert(!BufferIsValid(childbuf));
Assert(updateblkno == InvalidBlockNumber);
}
else
{
Assert(BufferIsValid(childbuf));
Assert(updateblkno != InvalidBlockNumber);
childpage = BufferGetPage(childbuf);
}
/*
* See if the incoming tuple will fit on the page. beginPlaceToPage will
* decide if the page needs to be split, and will compute the split
* contents if so. See comments for beginPlaceToPage and execPlaceToPage
* functions for more details of the API here.
*/
rc = btree->beginPlaceToPage(btree, stack->buffer, stack,
insertdata, updateblkno,
&ptp_workspace,
&newlpage, &newrpage);
if (rc == GPTP_NO_WORK)
{
/* Nothing to do */
result = true;
}
else if (rc == GPTP_INSERT)
{
/* It will fit, perform the insertion */
START_CRIT_SECTION();
if (RelationNeedsWAL(btree->index))
{
XLogBeginInsert();
XLogRegisterBuffer(0, stack->buffer, REGBUF_STANDARD);
if (BufferIsValid(childbuf))
XLogRegisterBuffer(1, childbuf, REGBUF_STANDARD);
}
/* Perform the page update, and register any extra WAL data */
btree->execPlaceToPage(btree, stack->buffer, stack,
insertdata, updateblkno, ptp_workspace);
MarkBufferDirty(stack->buffer);
/* An insert to an internal page finishes the split of the child. */
if (BufferIsValid(childbuf))
{
GinPageGetOpaque(childpage)->flags &= ~GIN_INCOMPLETE_SPLIT;
MarkBufferDirty(childbuf);
}
if (RelationNeedsWAL(btree->index))
{
XLogRecPtr recptr;
ginxlogInsert xlrec;
BlockIdData childblknos[2];
xlrec.flags = xlflags;
XLogRegisterData((char *) &xlrec, sizeof(ginxlogInsert));
/*
* Log information about child if this was an insertion of a
* downlink.
*/
if (BufferIsValid(childbuf))
{
BlockIdSet(&childblknos[0], BufferGetBlockNumber(childbuf));
BlockIdSet(&childblknos[1], GinPageGetOpaque(childpage)->rightlink);
XLogRegisterData((char *) childblknos,
sizeof(BlockIdData) * 2);
}
recptr = XLogInsert(RM_GIN_ID, XLOG_GIN_INSERT);
PageSetLSN(page, recptr);
if (BufferIsValid(childbuf))
PageSetLSN(childpage, recptr);
}
END_CRIT_SECTION();
/* Insertion is complete. */
result = true;
}
else if (rc == GPTP_SPLIT)
{
/*
* Didn't fit, need to split. The split has been computed in newlpage
* and newrpage, which are pointers to palloc'd pages, not associated
* with buffers. stack->buffer is not touched yet.
*/
Buffer rbuffer;
BlockNumber savedRightLink;
ginxlogSplit data;
Buffer lbuffer = InvalidBuffer;
Page newrootpg = NULL;
/* Get a new index page to become the right page */
rbuffer = GinNewBuffer(btree->index);
/* During index build, count the new page */
if (buildStats)
{
if (btree->isData)
buildStats->nDataPages++;
else
buildStats->nEntryPages++;
}
savedRightLink = GinPageGetOpaque(page)->rightlink;
/* Begin setting up WAL record */
data.node = btree->index->rd_node;
data.flags = xlflags;
if (BufferIsValid(childbuf))
{
data.leftChildBlkno = BufferGetBlockNumber(childbuf);
data.rightChildBlkno = GinPageGetOpaque(childpage)->rightlink;
}
else
data.leftChildBlkno = data.rightChildBlkno = InvalidBlockNumber;
if (stack->parent == NULL)
{
/*
* splitting the root, so we need to allocate new left page and
* place pointers to left and right page on root page.
*/
lbuffer = GinNewBuffer(btree->index);
/* During index build, count the new left page */
if (buildStats)
{
if (btree->isData)
buildStats->nDataPages++;
else
buildStats->nEntryPages++;
}
data.rrlink = InvalidBlockNumber;
data.flags |= GIN_SPLIT_ROOT;
GinPageGetOpaque(newrpage)->rightlink = InvalidBlockNumber;
GinPageGetOpaque(newlpage)->rightlink = BufferGetBlockNumber(rbuffer);
/*
* Construct a new root page containing downlinks to the new left
* and right pages. (Do this in a temporary copy rather than
* overwriting the original page directly, since we're not in the
* critical section yet.)
*/
newrootpg = PageGetTempPage(newrpage);
GinInitPage(newrootpg, GinPageGetOpaque(newlpage)->flags & ~(GIN_LEAF | GIN_COMPRESSED), BLCKSZ);
btree->fillRoot(btree, newrootpg,
BufferGetBlockNumber(lbuffer), newlpage,
BufferGetBlockNumber(rbuffer), newrpage);
}
else
{
/* splitting a non-root page */
data.rrlink = savedRightLink;
GinPageGetOpaque(newrpage)->rightlink = savedRightLink;
GinPageGetOpaque(newlpage)->flags |= GIN_INCOMPLETE_SPLIT;
GinPageGetOpaque(newlpage)->rightlink = BufferGetBlockNumber(rbuffer);
}
/*
* OK, we have the new contents of the left page in a temporary copy
* now (newlpage), and likewise for the new contents of the
* newly-allocated right block. The original page is still unchanged.
*
* If this is a root split, we also have a temporary page containing
* the new contents of the root.
*/
START_CRIT_SECTION();
MarkBufferDirty(rbuffer);
MarkBufferDirty(stack->buffer);
/*
* Restore the temporary copies over the real buffers.
*/
if (stack->parent == NULL)
{
/* Splitting the root, three pages to update */
MarkBufferDirty(lbuffer);
memcpy(page, newrootpg, BLCKSZ);
memcpy(BufferGetPage(lbuffer), newlpage, BLCKSZ);
memcpy(BufferGetPage(rbuffer), newrpage, BLCKSZ);
}
else
{
/* Normal split, only two pages to update */
memcpy(page, newlpage, BLCKSZ);
memcpy(BufferGetPage(rbuffer), newrpage, BLCKSZ);
}
/* We also clear childbuf's INCOMPLETE_SPLIT flag, if passed */
if (BufferIsValid(childbuf))
{
GinPageGetOpaque(childpage)->flags &= ~GIN_INCOMPLETE_SPLIT;
MarkBufferDirty(childbuf);
}
/* write WAL record */
if (RelationNeedsWAL(btree->index))
{
XLogRecPtr recptr;
XLogBeginInsert();
/*
* We just take full page images of all the split pages. Splits
* are uncommon enough that it's not worth complicating the code
* to be more efficient.
*/
if (stack->parent == NULL)
{
XLogRegisterBuffer(0, lbuffer, REGBUF_FORCE_IMAGE | REGBUF_STANDARD);
XLogRegisterBuffer(1, rbuffer, REGBUF_FORCE_IMAGE | REGBUF_STANDARD);
XLogRegisterBuffer(2, stack->buffer, REGBUF_FORCE_IMAGE | REGBUF_STANDARD);
}
else
{
XLogRegisterBuffer(0, stack->buffer, REGBUF_FORCE_IMAGE | REGBUF_STANDARD);
XLogRegisterBuffer(1, rbuffer, REGBUF_FORCE_IMAGE | REGBUF_STANDARD);
}
if (BufferIsValid(childbuf))
XLogRegisterBuffer(3, childbuf, REGBUF_STANDARD);
XLogRegisterData((char *) &data, sizeof(ginxlogSplit));
recptr = XLogInsert(RM_GIN_ID, XLOG_GIN_SPLIT);
PageSetLSN(page, recptr);
PageSetLSN(BufferGetPage(rbuffer), recptr);
if (stack->parent == NULL)
PageSetLSN(BufferGetPage(lbuffer), recptr);
if (BufferIsValid(childbuf))
PageSetLSN(childpage, recptr);
}
END_CRIT_SECTION();
/*
* We can release the locks/pins on the new pages now, but keep
* stack->buffer locked. childbuf doesn't get unlocked either.
*/
UnlockReleaseBuffer(rbuffer);
if (stack->parent == NULL)
UnlockReleaseBuffer(lbuffer);
/*
* If we split the root, we're done. Otherwise the split is not
* complete until the downlink for the new page has been inserted to
* the parent.
*/
result = (stack->parent == NULL);
}
else
{
elog(ERROR, "invalid return code from GIN placeToPage method: %d", rc);
result = false; /* keep compiler quiet */
}
/* Clean up temp context */
MemoryContextSwitchTo(oldCxt);
MemoryContextDelete(tmpCxt);
return result;
}
/*
* Attempt to expand the hash table by creating one new bucket.
*
* This will silently do nothing if we don't get cleanup lock on old or
* new bucket.
*
* Complete the pending splits and remove the tuples from old bucket,
* if there are any left over from the previous split.
*
* The caller must hold a pin, but no lock, on the metapage buffer.
* The buffer is returned in the same state.
*/
void
_hash_expandtable(Relation rel, Buffer metabuf)
{
HashMetaPage metap;
Bucket old_bucket;
Bucket new_bucket;
uint32 spare_ndx;
BlockNumber start_oblkno;
BlockNumber start_nblkno;
Buffer buf_nblkno;
Buffer buf_oblkno;
Page opage;
Page npage;
HashPageOpaque oopaque;
HashPageOpaque nopaque;
uint32 maxbucket;
uint32 highmask;
uint32 lowmask;
bool metap_update_masks = false;
bool metap_update_splitpoint = false;
restart_expand:
/*
* Write-lock the meta page. It used to be necessary to acquire a
* heavyweight lock to begin a split, but that is no longer required.
*/
LockBuffer(metabuf, BUFFER_LOCK_EXCLUSIVE);
_hash_checkpage(rel, metabuf, LH_META_PAGE);
metap = HashPageGetMeta(BufferGetPage(metabuf));
/*
* Check to see if split is still needed; someone else might have already
* done one while we waited for the lock.
*
* Make sure this stays in sync with _hash_doinsert()
*/
if (metap->hashm_ntuples <=
(double) metap->hashm_ffactor * (metap->hashm_maxbucket + 1))
goto fail;
/*
* Can't split anymore if maxbucket has reached its maximum possible
* value.
*
* Ideally we'd allow bucket numbers up to UINT_MAX-1 (no higher because
* the calculation maxbucket+1 mustn't overflow). Currently we restrict
* to half that because of overflow looping in _hash_log2() and
* insufficient space in hashm_spares[]. It's moot anyway because an
* index with 2^32 buckets would certainly overflow BlockNumber and hence
* _hash_alloc_buckets() would fail, but if we supported buckets smaller
* than a disk block then this would be an independent constraint.
*
* If you change this, see also the maximum initial number of buckets in
* _hash_init().
*/
if (metap->hashm_maxbucket >= (uint32) 0x7FFFFFFE)
goto fail;
/*
* Determine which bucket is to be split, and attempt to take cleanup lock
* on the old bucket. If we can't get the lock, give up.
*
* The cleanup lock protects us not only against other backends, but
* against our own backend as well.
*
* The cleanup lock is mainly to protect the split from concurrent
* inserts. See src/backend/access/hash/README, Lock Definitions for
* further details. Due to this locking restriction, if there is any
* pending scan, the split will give up which is not good, but harmless.
*/
new_bucket = metap->hashm_maxbucket + 1;
old_bucket = (new_bucket & metap->hashm_lowmask);
start_oblkno = BUCKET_TO_BLKNO(metap, old_bucket);
buf_oblkno = _hash_getbuf_with_condlock_cleanup(rel, start_oblkno, LH_BUCKET_PAGE);
if (!buf_oblkno)
goto fail;
opage = BufferGetPage(buf_oblkno);
oopaque = (HashPageOpaque) PageGetSpecialPointer(opage);
/*
* We want to finish the split from a bucket as there is no apparent
* benefit by not doing so and it will make the code complicated to finish
* the split that involves multiple buckets considering the case where new
* split also fails. We don't need to consider the new bucket for
* completing the split here as it is not possible that a re-split of new
* bucket starts when there is still a pending split from old bucket.
*/
if (H_BUCKET_BEING_SPLIT(oopaque))
{
/*
* Copy bucket mapping info now; refer the comment in code below where
* we copy this information before calling _hash_splitbucket to see
* why this is okay.
*/
maxbucket = metap->hashm_maxbucket;
highmask = metap->hashm_highmask;
lowmask = metap->hashm_lowmask;
/*
* Release the lock on metapage and old_bucket, before completing the
* split.
*/
LockBuffer(metabuf, BUFFER_LOCK_UNLOCK);
LockBuffer(buf_oblkno, BUFFER_LOCK_UNLOCK);
_hash_finish_split(rel, metabuf, buf_oblkno, old_bucket, maxbucket,
highmask, lowmask);
/* release the pin on old buffer and retry for expand. */
_hash_dropbuf(rel, buf_oblkno);
goto restart_expand;
}
/*
* Clean the tuples remained from the previous split. This operation
* requires cleanup lock and we already have one on the old bucket, so
* let's do it. We also don't want to allow further splits from the bucket
* till the garbage of previous split is cleaned. This has two
* advantages; first, it helps in avoiding the bloat due to garbage and
* second is, during cleanup of bucket, we are always sure that the
* garbage tuples belong to most recently split bucket. On the contrary,
* if we allow cleanup of bucket after meta page is updated to indicate
* the new split and before the actual split, the cleanup operation won't
* be able to decide whether the tuple has been moved to the newly created
* bucket and ended up deleting such tuples.
*/
if (H_NEEDS_SPLIT_CLEANUP(oopaque))
{
/*
* Copy bucket mapping info now; refer to the comment in code below
* where we copy this information before calling _hash_splitbucket to
* see why this is okay.
*/
maxbucket = metap->hashm_maxbucket;
highmask = metap->hashm_highmask;
lowmask = metap->hashm_lowmask;
/* Release the metapage lock. */
LockBuffer(metabuf, BUFFER_LOCK_UNLOCK);
hashbucketcleanup(rel, old_bucket, buf_oblkno, start_oblkno, NULL,
maxbucket, highmask, lowmask, NULL, NULL, true,
NULL, NULL);
_hash_dropbuf(rel, buf_oblkno);
goto restart_expand;
}
/*
* There shouldn't be any active scan on new bucket.
*
* Note: it is safe to compute the new bucket's blkno here, even though we
* may still need to update the BUCKET_TO_BLKNO mapping. This is because
* the current value of hashm_spares[hashm_ovflpoint] correctly shows
* where we are going to put a new splitpoint's worth of buckets.
*/
start_nblkno = BUCKET_TO_BLKNO(metap, new_bucket);
/*
* If the split point is increasing we need to allocate a new batch of
* bucket pages.
*/
spare_ndx = _hash_spareindex(new_bucket + 1);
if (spare_ndx > metap->hashm_ovflpoint)
{
uint32 buckets_to_add;
Assert(spare_ndx == metap->hashm_ovflpoint + 1);
/*
* We treat allocation of buckets as a separate WAL-logged action.
* Even if we fail after this operation, won't leak bucket pages;
* rather, the next split will consume this space. In any case, even
* without failure we don't use all the space in one split operation.
*/
buckets_to_add = _hash_get_totalbuckets(spare_ndx) - new_bucket;
if (!_hash_alloc_buckets(rel, start_nblkno, buckets_to_add))
{
/* can't split due to BlockNumber overflow */
_hash_relbuf(rel, buf_oblkno);
goto fail;
}
}
/*
* Physically allocate the new bucket's primary page. We want to do this
* before changing the metapage's mapping info, in case we can't get the
* disk space. Ideally, we don't need to check for cleanup lock on new
* bucket as no other backend could find this bucket unless meta page is
* updated. However, it is good to be consistent with old bucket locking.
*/
buf_nblkno = _hash_getnewbuf(rel, start_nblkno, MAIN_FORKNUM);
if (!IsBufferCleanupOK(buf_nblkno))
{
_hash_relbuf(rel, buf_oblkno);
_hash_relbuf(rel, buf_nblkno);
goto fail;
}
/*
* Since we are scribbling on the pages in the shared buffers, establish a
* critical section. Any failure in this next code leaves us with a big
* problem: the metapage is effectively corrupt but could get written back
* to disk.
*/
START_CRIT_SECTION();
/*
* Okay to proceed with split. Update the metapage bucket mapping info.
*/
metap->hashm_maxbucket = new_bucket;
if (new_bucket > metap->hashm_highmask)
{
/* Starting a new doubling */
metap->hashm_lowmask = metap->hashm_highmask;
metap->hashm_highmask = new_bucket | metap->hashm_lowmask;
metap_update_masks = true;
}
/*
* If the split point is increasing we need to adjust the hashm_spares[]
* array and hashm_ovflpoint so that future overflow pages will be created
* beyond this new batch of bucket pages.
*/
if (spare_ndx > metap->hashm_ovflpoint)
{
metap->hashm_spares[spare_ndx] = metap->hashm_spares[metap->hashm_ovflpoint];
metap->hashm_ovflpoint = spare_ndx;
metap_update_splitpoint = true;
}
MarkBufferDirty(metabuf);
/*
* Copy bucket mapping info now; this saves re-accessing the meta page
* inside _hash_splitbucket's inner loop. Note that once we drop the
* split lock, other splits could begin, so these values might be out of
* date before _hash_splitbucket finishes. That's okay, since all it
* needs is to tell which of these two buckets to map hashkeys into.
*/
maxbucket = metap->hashm_maxbucket;
highmask = metap->hashm_highmask;
lowmask = metap->hashm_lowmask;
opage = BufferGetPage(buf_oblkno);
oopaque = (HashPageOpaque) PageGetSpecialPointer(opage);
/*
* Mark the old bucket to indicate that split is in progress. (At
* operation end, we will clear the split-in-progress flag.) Also, for a
* primary bucket page, hasho_prevblkno stores the number of buckets that
* existed as of the last split, so we must update that value here.
*/
oopaque->hasho_flag |= LH_BUCKET_BEING_SPLIT;
oopaque->hasho_prevblkno = maxbucket;
MarkBufferDirty(buf_oblkno);
npage = BufferGetPage(buf_nblkno);
/*
* initialize the new bucket's primary page and mark it to indicate that
* split is in progress.
*/
nopaque = (HashPageOpaque) PageGetSpecialPointer(npage);
nopaque->hasho_prevblkno = maxbucket;
nopaque->hasho_nextblkno = InvalidBlockNumber;
nopaque->hasho_bucket = new_bucket;
nopaque->hasho_flag = LH_BUCKET_PAGE | LH_BUCKET_BEING_POPULATED;
nopaque->hasho_page_id = HASHO_PAGE_ID;
MarkBufferDirty(buf_nblkno);
/* XLOG stuff */
if (RelationNeedsWAL(rel))
{
xl_hash_split_allocate_page xlrec;
XLogRecPtr recptr;
xlrec.new_bucket = maxbucket;
xlrec.old_bucket_flag = oopaque->hasho_flag;
xlrec.new_bucket_flag = nopaque->hasho_flag;
xlrec.flags = 0;
XLogBeginInsert();
XLogRegisterBuffer(0, buf_oblkno, REGBUF_STANDARD);
XLogRegisterBuffer(1, buf_nblkno, REGBUF_WILL_INIT);
XLogRegisterBuffer(2, metabuf, REGBUF_STANDARD);
if (metap_update_masks)
{
xlrec.flags |= XLH_SPLIT_META_UPDATE_MASKS;
XLogRegisterBufData(2, (char *) &metap->hashm_lowmask, sizeof(uint32));
XLogRegisterBufData(2, (char *) &metap->hashm_highmask, sizeof(uint32));
}
if (metap_update_splitpoint)
{
xlrec.flags |= XLH_SPLIT_META_UPDATE_SPLITPOINT;
XLogRegisterBufData(2, (char *) &metap->hashm_ovflpoint,
sizeof(uint32));
XLogRegisterBufData(2,
(char *) &metap->hashm_spares[metap->hashm_ovflpoint],
sizeof(uint32));
}
XLogRegisterData((char *) &xlrec, SizeOfHashSplitAllocPage);
recptr = XLogInsert(RM_HASH_ID, XLOG_HASH_SPLIT_ALLOCATE_PAGE);
PageSetLSN(BufferGetPage(buf_oblkno), recptr);
PageSetLSN(BufferGetPage(buf_nblkno), recptr);
PageSetLSN(BufferGetPage(metabuf), recptr);
}
END_CRIT_SECTION();
/* drop lock, but keep pin */
LockBuffer(metabuf, BUFFER_LOCK_UNLOCK);
/* Relocate records to the new bucket */
_hash_splitbucket(rel, metabuf,
old_bucket, new_bucket,
buf_oblkno, buf_nblkno, NULL,
maxbucket, highmask, lowmask);
/* all done, now release the pins on primary buckets. */
_hash_dropbuf(rel, buf_oblkno);
_hash_dropbuf(rel, buf_nblkno);
return;
/* Here if decide not to split or fail to acquire old bucket lock */
fail:
/* We didn't write the metapage, so just drop lock */
LockBuffer(metabuf, BUFFER_LOCK_UNLOCK);
}
/*
* _hash_init() -- Initialize the metadata page of a hash index,
* the initial buckets, and the initial bitmap page.
*
* The initial number of buckets is dependent on num_tuples, an estimate
* of the number of tuples to be loaded into the index initially. The
* chosen number of buckets is returned.
*
* We are fairly cavalier about locking here, since we know that no one else
* could be accessing this index. In particular the rule about not holding
* multiple buffer locks is ignored.
*/
uint32
_hash_init(Relation rel, double num_tuples, ForkNumber forkNum)
{
Buffer metabuf;
Buffer buf;
Buffer bitmapbuf;
Page pg;
HashMetaPage metap;
RegProcedure procid;
int32 data_width;
int32 item_width;
int32 ffactor;
uint32 num_buckets;
uint32 i;
bool use_wal;
/* safety check */
if (RelationGetNumberOfBlocksInFork(rel, forkNum) != 0)
elog(ERROR, "cannot initialize non-empty hash index \"%s\"",
RelationGetRelationName(rel));
/*
* WAL log creation of pages if the relation is persistent, or this is the
* init fork. Init forks for unlogged relations always need to be WAL
* logged.
*/
use_wal = RelationNeedsWAL(rel) || forkNum == INIT_FORKNUM;
/*
* Determine the target fill factor (in tuples per bucket) for this index.
* The idea is to make the fill factor correspond to pages about as full
* as the user-settable fillfactor parameter says. We can compute it
* exactly since the index datatype (i.e. uint32 hash key) is fixed-width.
*/
data_width = sizeof(uint32);
item_width = MAXALIGN(sizeof(IndexTupleData)) + MAXALIGN(data_width) +
sizeof(ItemIdData); /* include the line pointer */
ffactor = RelationGetTargetPageUsage(rel, HASH_DEFAULT_FILLFACTOR) / item_width;
/* keep to a sane range */
if (ffactor < 10)
ffactor = 10;
procid = index_getprocid(rel, 1, HASHSTANDARD_PROC);
/*
* We initialize the metapage, the first N bucket pages, and the first
* bitmap page in sequence, using _hash_getnewbuf to cause smgrextend()
* calls to occur. This ensures that the smgr level has the right idea of
* the physical index length.
*
* Critical section not required, because on error the creation of the
* whole relation will be rolled back.
*/
metabuf = _hash_getnewbuf(rel, HASH_METAPAGE, forkNum);
_hash_init_metabuffer(metabuf, num_tuples, procid, ffactor, false);
MarkBufferDirty(metabuf);
pg = BufferGetPage(metabuf);
metap = HashPageGetMeta(pg);
/* XLOG stuff */
if (use_wal)
{
xl_hash_init_meta_page xlrec;
XLogRecPtr recptr;
xlrec.num_tuples = num_tuples;
xlrec.procid = metap->hashm_procid;
xlrec.ffactor = metap->hashm_ffactor;
XLogBeginInsert();
XLogRegisterData((char *) &xlrec, SizeOfHashInitMetaPage);
XLogRegisterBuffer(0, metabuf, REGBUF_WILL_INIT);
recptr = XLogInsert(RM_HASH_ID, XLOG_HASH_INIT_META_PAGE);
PageSetLSN(BufferGetPage(metabuf), recptr);
}
num_buckets = metap->hashm_maxbucket + 1;
/*
* Release buffer lock on the metapage while we initialize buckets.
* Otherwise, we'll be in interrupt holdoff and the CHECK_FOR_INTERRUPTS
* won't accomplish anything. It's a bad idea to hold buffer locks for
* long intervals in any case, since that can block the bgwriter.
*/
LockBuffer(metabuf, BUFFER_LOCK_UNLOCK);
/*
* Initialize and WAL Log the first N buckets
*/
for (i = 0; i < num_buckets; i++)
{
BlockNumber blkno;
/* Allow interrupts, in case N is huge */
CHECK_FOR_INTERRUPTS();
blkno = BUCKET_TO_BLKNO(metap, i);
buf = _hash_getnewbuf(rel, blkno, forkNum);
_hash_initbuf(buf, metap->hashm_maxbucket, i, LH_BUCKET_PAGE, false);
MarkBufferDirty(buf);
if (use_wal)
log_newpage(&rel->rd_node,
forkNum,
blkno,
BufferGetPage(buf),
true);
_hash_relbuf(rel, buf);
}
/* Now reacquire buffer lock on metapage */
LockBuffer(metabuf, BUFFER_LOCK_EXCLUSIVE);
/*
* Initialize bitmap page
*/
bitmapbuf = _hash_getnewbuf(rel, num_buckets + 1, forkNum);
_hash_initbitmapbuffer(bitmapbuf, metap->hashm_bmsize, false);
MarkBufferDirty(bitmapbuf);
/* add the new bitmap page to the metapage's list of bitmaps */
/* metapage already has a write lock */
if (metap->hashm_nmaps >= HASH_MAX_BITMAPS)
ereport(ERROR,
(errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED),
errmsg("out of overflow pages in hash index \"%s\"",
RelationGetRelationName(rel))));
metap->hashm_mapp[metap->hashm_nmaps] = num_buckets + 1;
metap->hashm_nmaps++;
MarkBufferDirty(metabuf);
/* XLOG stuff */
if (use_wal)
{
xl_hash_init_bitmap_page xlrec;
XLogRecPtr recptr;
xlrec.bmsize = metap->hashm_bmsize;
XLogBeginInsert();
XLogRegisterData((char *) &xlrec, SizeOfHashInitBitmapPage);
XLogRegisterBuffer(0, bitmapbuf, REGBUF_WILL_INIT);
/*
* This is safe only because nobody else can be modifying the index at
* this stage; it's only visible to the transaction that is creating
* it.
*/
XLogRegisterBuffer(1, metabuf, REGBUF_STANDARD);
recptr = XLogInsert(RM_HASH_ID, XLOG_HASH_INIT_BITMAP_PAGE);
PageSetLSN(BufferGetPage(bitmapbuf), recptr);
PageSetLSN(BufferGetPage(metabuf), recptr);
}
/* all done */
_hash_relbuf(rel, bitmapbuf);
_hash_relbuf(rel, metabuf);
return num_buckets;
}
/*
* _hash_splitbucket -- split 'obucket' into 'obucket' and 'nbucket'
*
* This routine is used to partition the tuples between old and new bucket and
* is used to finish the incomplete split operations. To finish the previously
* interrupted split operation, the caller needs to fill htab. If htab is set,
* then we skip the movement of tuples that exists in htab, otherwise NULL
* value of htab indicates movement of all the tuples that belong to the new
* bucket.
*
* 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.
*
* The caller must hold cleanup 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.)
*
* Split needs to retain pin on primary bucket pages of both old and new
* buckets till end of operation. This is to prevent vacuum from starting
* while a split is in progress.
*
* In addition, the caller must have created the new bucket's base page,
* which is passed in buffer nbuf, pinned and write-locked. The lock will be
* released here and pin must be released by the caller. (The API is set up
* this way because we must do _hash_getnewbuf() before releasing the metapage
* write lock. So instead of passing the new bucket's start block number, we
* pass an actual buffer.)
*/
static void
_hash_splitbucket(Relation rel,
Buffer metabuf,
Bucket obucket,
Bucket nbucket,
Buffer obuf,
Buffer nbuf,
HTAB *htab,
uint32 maxbucket,
uint32 highmask,
uint32 lowmask)
{
Buffer bucket_obuf;
Buffer bucket_nbuf;
Page opage;
Page npage;
HashPageOpaque oopaque;
HashPageOpaque nopaque;
OffsetNumber itup_offsets[MaxIndexTuplesPerPage];
IndexTuple itups[MaxIndexTuplesPerPage];
Size all_tups_size = 0;
int i;
uint16 nitups = 0;
bucket_obuf = obuf;
opage = BufferGetPage(obuf);
oopaque = (HashPageOpaque) PageGetSpecialPointer(opage);
bucket_nbuf = nbuf;
npage = BufferGetPage(nbuf);
nopaque = (HashPageOpaque) PageGetSpecialPointer(npage);
/*
* 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. Outer loop iterates
* once per page in old bucket.
*/
for (;;)
{
BlockNumber oblkno;
OffsetNumber ooffnum;
OffsetNumber omaxoffnum;
/* Scan each tuple in old page */
omaxoffnum = PageGetMaxOffsetNumber(opage);
for (ooffnum = FirstOffsetNumber;
ooffnum <= omaxoffnum;
ooffnum = OffsetNumberNext(ooffnum))
{
IndexTuple itup;
Size itemsz;
Bucket bucket;
bool found = false;
/* skip dead tuples */
if (ItemIdIsDead(PageGetItemId(opage, ooffnum)))
continue;
/*
* Before inserting a tuple, probe the hash table containing TIDs
* of tuples belonging to new bucket, if we find a match, then
* skip that tuple, else fetch the item's hash key (conveniently
* stored in the item) and determine which bucket it now belongs
* in.
*/
itup = (IndexTuple) PageGetItem(opage,
PageGetItemId(opage, ooffnum));
if (htab)
(void) hash_search(htab, &itup->t_tid, HASH_FIND, &found);
if (found)
continue;
bucket = _hash_hashkey2bucket(_hash_get_indextuple_hashkey(itup),
maxbucket, highmask, lowmask);
if (bucket == nbucket)
{
IndexTuple new_itup;
/*
* make a copy of index tuple as we have to scribble on it.
*/
new_itup = CopyIndexTuple(itup);
/*
* mark the index tuple as moved by split, such tuples are
* skipped by scan if there is split in progress for a bucket.
*/
new_itup->t_info |= INDEX_MOVED_BY_SPLIT_MASK;
/*
* insert the tuple into the new bucket. if it doesn't fit on
* the current page in the new bucket, we must allocate a new
* overflow page and place the tuple on that page instead.
*/
itemsz = IndexTupleDSize(*new_itup);
itemsz = MAXALIGN(itemsz);
if (PageGetFreeSpaceForMultipleTuples(npage, nitups + 1) < (all_tups_size + itemsz))
{
/*
* 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();
/* drop lock, but keep pin */
LockBuffer(nbuf, BUFFER_LOCK_UNLOCK);
/* 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);
}
/*
* We're at the end of the old bucket chain, so we're done partitioning
* the tuples. Mark the old and new buckets to indicate split is
* finished.
*
* To avoid deadlocks due to locking order of buckets, first lock the old
* bucket and then the new bucket.
*/
LockBuffer(bucket_obuf, BUFFER_LOCK_EXCLUSIVE);
opage = BufferGetPage(bucket_obuf);
oopaque = (HashPageOpaque) PageGetSpecialPointer(opage);
LockBuffer(bucket_nbuf, BUFFER_LOCK_EXCLUSIVE);
npage = BufferGetPage(bucket_nbuf);
nopaque = (HashPageOpaque) PageGetSpecialPointer(npage);
START_CRIT_SECTION();
oopaque->hasho_flag &= ~LH_BUCKET_BEING_SPLIT;
nopaque->hasho_flag &= ~LH_BUCKET_BEING_POPULATED;
/*
* After the split is finished, mark the old bucket to indicate that it
* contains deletable tuples. We will clear split-cleanup flag after
* deleting such tuples either at the end of split or at the next split
* from old bucket or at the time of vacuum.
*/
oopaque->hasho_flag |= LH_BUCKET_NEEDS_SPLIT_CLEANUP;
/*
* now write the buffers, here we don't release the locks as caller is
* responsible to release locks.
*/
MarkBufferDirty(bucket_obuf);
MarkBufferDirty(bucket_nbuf);
if (RelationNeedsWAL(rel))
{
XLogRecPtr recptr;
xl_hash_split_complete xlrec;
xlrec.old_bucket_flag = oopaque->hasho_flag;
xlrec.new_bucket_flag = nopaque->hasho_flag;
XLogBeginInsert();
XLogRegisterData((char *) &xlrec, SizeOfHashSplitComplete);
XLogRegisterBuffer(0, bucket_obuf, REGBUF_STANDARD);
XLogRegisterBuffer(1, bucket_nbuf, REGBUF_STANDARD);
recptr = XLogInsert(RM_HASH_ID, XLOG_HASH_SPLIT_COMPLETE);
PageSetLSN(BufferGetPage(bucket_obuf), recptr);
PageSetLSN(BufferGetPage(bucket_nbuf), recptr);
}
END_CRIT_SECTION();
/*
* If possible, clean up the old bucket. We might not be able to do this
* if someone else has a pin on it, but if not then we can go ahead. This
* isn't absolutely necessary, but it reduces bloat; if we don't do it
* now, VACUUM will do it eventually, but maybe not until new overflow
* pages have been allocated. Note that there's no need to clean up the
* new bucket.
*/
if (IsBufferCleanupOK(bucket_obuf))
{
LockBuffer(bucket_nbuf, BUFFER_LOCK_UNLOCK);
hashbucketcleanup(rel, obucket, bucket_obuf,
BufferGetBlockNumber(bucket_obuf), NULL,
maxbucket, highmask, lowmask, NULL, NULL, true,
NULL, NULL);
}
else
{
LockBuffer(bucket_nbuf, BUFFER_LOCK_UNLOCK);
LockBuffer(bucket_obuf, BUFFER_LOCK_UNLOCK);
}
}
/*
* Update tuple origtup (size origsz), located in offset oldoff of buffer
* oldbuf, to newtup (size newsz) as summary tuple for the page range starting
* at heapBlk. oldbuf must not be locked on entry, and is not locked at exit.
*
* If samepage is true, attempt to put the new tuple in the same page, but if
* there's no room, use some other one.
*
* If the update is successful, return true; the revmap is updated to point to
* the new tuple. If the update is not done for whatever reason, return false.
* Caller may retry the update if this happens.
*/
bool
brin_doupdate(Relation idxrel, BlockNumber pagesPerRange,
BrinRevmap *revmap, BlockNumber heapBlk,
Buffer oldbuf, OffsetNumber oldoff,
const BrinTuple *origtup, Size origsz,
const BrinTuple *newtup, Size newsz,
bool samepage)
{
Page oldpage;
ItemId oldlp;
BrinTuple *oldtup;
Size oldsz;
Buffer newbuf;
BrinSpecialSpace *special;
bool extended = false;
newsz = MAXALIGN(newsz);
/* make sure the revmap is long enough to contain the entry we need */
brinRevmapExtend(revmap, heapBlk);
if (!samepage)
{
/* need a page on which to put the item */
newbuf = brin_getinsertbuffer(idxrel, oldbuf, newsz, &extended);
/* XXX delay vacuuming FSM until locks are released? */
if (extended)
FreeSpaceMapVacuum(idxrel);
if (!BufferIsValid(newbuf))
return false;
/*
* Note: it's possible (though unlikely) that the returned newbuf is
* the same as oldbuf, if brin_getinsertbuffer determined that the old
* buffer does in fact have enough space.
*/
if (newbuf == oldbuf)
newbuf = InvalidBuffer;
}
else
{
LockBuffer(oldbuf, BUFFER_LOCK_EXCLUSIVE);
newbuf = InvalidBuffer;
}
oldpage = BufferGetPage(oldbuf);
oldlp = PageGetItemId(oldpage, oldoff);
/*
* Check that the old tuple wasn't updated concurrently: it might have
* moved someplace else entirely ...
*/
if (!ItemIdIsNormal(oldlp))
{
LockBuffer(oldbuf, BUFFER_LOCK_UNLOCK);
if (BufferIsValid(newbuf))
UnlockReleaseBuffer(newbuf);
return false;
}
oldsz = ItemIdGetLength(oldlp);
oldtup = (BrinTuple *) PageGetItem(oldpage, oldlp);
/*
* ... or it might have been updated in place to different contents.
*/
if (!brin_tuples_equal(oldtup, oldsz, origtup, origsz))
{
LockBuffer(oldbuf, BUFFER_LOCK_UNLOCK);
if (BufferIsValid(newbuf))
UnlockReleaseBuffer(newbuf);
return false;
}
special = (BrinSpecialSpace *) PageGetSpecialPointer(oldpage);
/*
* Great, the old tuple is intact. We can proceed with the update.
*
* If there's enough room in the old page for the new tuple, replace it.
*
* Note that there might now be enough space on the page even though the
* caller told us there isn't, if a concurrent update moved another tuple
* elsewhere or replaced a tuple with a smaller one.
*/
if (((special->flags & BRIN_EVACUATE_PAGE) == 0) &&
brin_can_do_samepage_update(oldbuf, origsz, newsz))
{
if (BufferIsValid(newbuf))
UnlockReleaseBuffer(newbuf);
START_CRIT_SECTION();
PageIndexDeleteNoCompact(oldpage, &oldoff, 1);
if (PageAddItem(oldpage, (Item) newtup, newsz, oldoff, true,
false) == InvalidOffsetNumber)
elog(ERROR, "failed to add BRIN tuple");
MarkBufferDirty(oldbuf);
/* XLOG stuff */
if (RelationNeedsWAL(idxrel))
{
xl_brin_samepage_update xlrec;
XLogRecPtr recptr;
uint8 info = XLOG_BRIN_SAMEPAGE_UPDATE;
xlrec.offnum = oldoff;
XLogBeginInsert();
XLogRegisterData((char *) &xlrec, SizeOfBrinSamepageUpdate);
XLogRegisterBuffer(0, oldbuf, REGBUF_STANDARD);
XLogRegisterBufData(0, (char *) newtup, newsz);
recptr = XLogInsert(RM_BRIN_ID, info);
PageSetLSN(oldpage, recptr);
}
END_CRIT_SECTION();
LockBuffer(oldbuf, BUFFER_LOCK_UNLOCK);
return true;
}
else if (newbuf == InvalidBuffer)
{
/*
* Not enough space, but caller said that there was. Tell them to
* start over.
*/
LockBuffer(oldbuf, BUFFER_LOCK_UNLOCK);
return false;
}
else
{
/*
* Not enough free space on the oldpage. Put the new tuple on the new
* page, and update the revmap.
*/
Page newpage = BufferGetPage(newbuf);
Buffer revmapbuf;
ItemPointerData newtid;
OffsetNumber newoff;
revmapbuf = brinLockRevmapPageForUpdate(revmap, heapBlk);
START_CRIT_SECTION();
PageIndexDeleteNoCompact(oldpage, &oldoff, 1);
newoff = PageAddItem(newpage, (Item) newtup, newsz,
InvalidOffsetNumber, false, false);
if (newoff == InvalidOffsetNumber)
elog(ERROR, "failed to add BRIN tuple to new page");
MarkBufferDirty(oldbuf);
MarkBufferDirty(newbuf);
ItemPointerSet(&newtid, BufferGetBlockNumber(newbuf), newoff);
brinSetHeapBlockItemptr(revmapbuf, pagesPerRange, heapBlk, newtid);
MarkBufferDirty(revmapbuf);
/* XLOG stuff */
if (RelationNeedsWAL(idxrel))
{
xl_brin_update xlrec;
XLogRecPtr recptr;
uint8 info;
info = XLOG_BRIN_UPDATE | (extended ? XLOG_BRIN_INIT_PAGE : 0);
xlrec.insert.offnum = newoff;
xlrec.insert.heapBlk = heapBlk;
xlrec.insert.pagesPerRange = pagesPerRange;
xlrec.oldOffnum = oldoff;
XLogBeginInsert();
/* new page */
XLogRegisterData((char *) &xlrec, SizeOfBrinUpdate);
XLogRegisterBuffer(0, newbuf, REGBUF_STANDARD | (extended ? REGBUF_WILL_INIT : 0));
XLogRegisterBufData(0, (char *) newtup, newsz);
/* revmap page */
XLogRegisterBuffer(1, revmapbuf, REGBUF_STANDARD);
/* old page */
XLogRegisterBuffer(2, oldbuf, REGBUF_STANDARD);
recptr = XLogInsert(RM_BRIN_ID, info);
PageSetLSN(oldpage, recptr);
PageSetLSN(newpage, recptr);
PageSetLSN(BufferGetPage(revmapbuf), recptr);
}
END_CRIT_SECTION();
LockBuffer(revmapbuf, BUFFER_LOCK_UNLOCK);
LockBuffer(oldbuf, BUFFER_LOCK_UNLOCK);
UnlockReleaseBuffer(newbuf);
return true;
}
}
Datum
ginbuild(PG_FUNCTION_ARGS)
{
Relation heap = (Relation) PG_GETARG_POINTER(0);
Relation index = (Relation) PG_GETARG_POINTER(1);
IndexInfo *indexInfo = (IndexInfo *) PG_GETARG_POINTER(2);
IndexBuildResult *result;
double reltuples;
GinBuildState buildstate;
Buffer RootBuffer,
MetaBuffer;
ItemPointerData *list;
Datum entry;
uint32 nlist;
MemoryContext oldCtx;
OffsetNumber attnum;
if (RelationGetNumberOfBlocks(index) != 0)
elog(ERROR, "index \"%s\" already contains data",
RelationGetRelationName(index));
initGinState(&buildstate.ginstate, index);
buildstate.indtuples = 0;
memset(&buildstate.buildStats, 0, sizeof(GinStatsData));
/* initialize the meta page */
MetaBuffer = GinNewBuffer(index);
/* initialize the root page */
RootBuffer = GinNewBuffer(index);
START_CRIT_SECTION();
GinInitMetabuffer(MetaBuffer);
MarkBufferDirty(MetaBuffer);
GinInitBuffer(RootBuffer, GIN_LEAF);
MarkBufferDirty(RootBuffer);
if (RelationNeedsWAL(index))
{
XLogRecPtr recptr;
XLogRecData rdata;
Page page;
rdata.buffer = InvalidBuffer;
rdata.data = (char *) &(index->rd_node);
rdata.len = sizeof(RelFileNode);
rdata.next = NULL;
recptr = XLogInsert(RM_GIN_ID, XLOG_GIN_CREATE_INDEX, &rdata);
page = BufferGetPage(RootBuffer);
PageSetLSN(page, recptr);
PageSetTLI(page, ThisTimeLineID);
page = BufferGetPage(MetaBuffer);
PageSetLSN(page, recptr);
PageSetTLI(page, ThisTimeLineID);
}
UnlockReleaseBuffer(MetaBuffer);
UnlockReleaseBuffer(RootBuffer);
END_CRIT_SECTION();
/* count the root as first entry page */
buildstate.buildStats.nEntryPages++;
/*
* create a temporary memory context that is reset once for each tuple
* inserted into the index
*/
buildstate.tmpCtx = AllocSetContextCreate(CurrentMemoryContext,
"Gin build temporary context",
ALLOCSET_DEFAULT_MINSIZE,
ALLOCSET_DEFAULT_INITSIZE,
ALLOCSET_DEFAULT_MAXSIZE);
buildstate.funcCtx = AllocSetContextCreate(buildstate.tmpCtx,
"Gin build temporary context for user-defined function",
ALLOCSET_DEFAULT_MINSIZE,
ALLOCSET_DEFAULT_INITSIZE,
ALLOCSET_DEFAULT_MAXSIZE);
buildstate.accum.ginstate = &buildstate.ginstate;
ginInitBA(&buildstate.accum);
/*
* Do the heap scan. We disallow sync scan here because dataPlaceToPage
* prefers to receive tuples in TID order.
*/
reltuples = IndexBuildHeapScan(heap, index, indexInfo, false,
ginBuildCallback, (void *) &buildstate);
/* dump remaining entries to the index */
oldCtx = MemoryContextSwitchTo(buildstate.tmpCtx);
ginBeginBAScan(&buildstate.accum);
while ((list = ginGetEntry(&buildstate.accum, &attnum, &entry, &nlist)) != NULL)
{
/* there could be many entries, so be willing to abort here */
CHECK_FOR_INTERRUPTS();
ginEntryInsert(index, &buildstate.ginstate, attnum, entry,
list, nlist, &buildstate.buildStats);
}
MemoryContextSwitchTo(oldCtx);
MemoryContextDelete(buildstate.tmpCtx);
/*
* Update metapage stats
*/
buildstate.buildStats.nTotalPages = RelationGetNumberOfBlocks(index);
ginUpdateStats(index, &buildstate.buildStats);
/*
* Return statistics
*/
result = (IndexBuildResult *) palloc(sizeof(IndexBuildResult));
result->heap_tuples = reltuples;
result->index_tuples = buildstate.indtuples;
PG_RETURN_POINTER(result);
}
IndexBuildResult *
ginbuild(Relation heap, Relation index, IndexInfo *indexInfo)
{
IndexBuildResult *result;
double reltuples;
GinBuildState buildstate;
Buffer RootBuffer,
MetaBuffer;
ItemPointerData *list;
Datum key;
GinNullCategory category;
uint32 nlist;
MemoryContext oldCtx;
OffsetNumber attnum;
if (RelationGetNumberOfBlocks(index) != 0)
elog(ERROR, "index \"%s\" already contains data",
RelationGetRelationName(index));
initGinState(&buildstate.ginstate, index);
buildstate.indtuples = 0;
memset(&buildstate.buildStats, 0, sizeof(GinStatsData));
/* initialize the meta page */
MetaBuffer = GinNewBuffer(index);
/* initialize the root page */
RootBuffer = GinNewBuffer(index);
START_CRIT_SECTION();
GinInitMetabuffer(MetaBuffer);
MarkBufferDirty(MetaBuffer);
GinInitBuffer(RootBuffer, GIN_LEAF);
MarkBufferDirty(RootBuffer);
if (RelationNeedsWAL(index))
{
XLogRecPtr recptr;
Page page;
XLogBeginInsert();
XLogRegisterBuffer(0, MetaBuffer, REGBUF_WILL_INIT);
XLogRegisterBuffer(1, RootBuffer, REGBUF_WILL_INIT);
recptr = XLogInsert(RM_GIN_ID, XLOG_GIN_CREATE_INDEX);
page = BufferGetPage(RootBuffer);
PageSetLSN(page, recptr);
page = BufferGetPage(MetaBuffer);
PageSetLSN(page, recptr);
}
UnlockReleaseBuffer(MetaBuffer);
UnlockReleaseBuffer(RootBuffer);
END_CRIT_SECTION();
/* count the root as first entry page */
buildstate.buildStats.nEntryPages++;
/*
* create a temporary memory context that is used to hold data not yet
* dumped out to the index
*/
buildstate.tmpCtx = AllocSetContextCreate(CurrentMemoryContext,
"Gin build temporary context",
ALLOCSET_DEFAULT_MINSIZE,
ALLOCSET_DEFAULT_INITSIZE,
ALLOCSET_DEFAULT_MAXSIZE);
/*
* create a temporary memory context that is used for calling
* ginExtractEntries(), and can be reset after each tuple
*/
buildstate.funcCtx = AllocSetContextCreate(CurrentMemoryContext,
"Gin build temporary context for user-defined function",
ALLOCSET_DEFAULT_MINSIZE,
ALLOCSET_DEFAULT_INITSIZE,
ALLOCSET_DEFAULT_MAXSIZE);
buildstate.accum.ginstate = &buildstate.ginstate;
ginInitBA(&buildstate.accum);
/*
* Do the heap scan. We disallow sync scan here because dataPlaceToPage
* prefers to receive tuples in TID order.
*/
reltuples = IndexBuildHeapScan(heap, index, indexInfo, false,
ginBuildCallback, (void *) &buildstate);
/* dump remaining entries to the index */
oldCtx = MemoryContextSwitchTo(buildstate.tmpCtx);
ginBeginBAScan(&buildstate.accum);
while ((list = ginGetBAEntry(&buildstate.accum,
&attnum, &key, &category, &nlist)) != NULL)
{
/* there could be many entries, so be willing to abort here */
CHECK_FOR_INTERRUPTS();
ginEntryInsert(&buildstate.ginstate, attnum, key, category,
list, nlist, &buildstate.buildStats);
}
MemoryContextSwitchTo(oldCtx);
MemoryContextDelete(buildstate.funcCtx);
MemoryContextDelete(buildstate.tmpCtx);
/*
* Update metapage stats
*/
buildstate.buildStats.nTotalPages = RelationGetNumberOfBlocks(index);
ginUpdateStats(index, &buildstate.buildStats);
/*
* Return statistics
*/
result = (IndexBuildResult *) palloc(sizeof(IndexBuildResult));
result->heap_tuples = reltuples;
result->index_tuples = buildstate.indtuples;
return result;
}
/*
* Delete a posting tree page.
*/
static void
ginDeletePage(GinVacuumState *gvs, BlockNumber deleteBlkno, BlockNumber leftBlkno,
BlockNumber parentBlkno, OffsetNumber myoff, bool isParentRoot)
{
Buffer dBuffer;
Buffer lBuffer;
Buffer pBuffer;
Page page,
parentPage;
BlockNumber rightlink;
/*
* Lock the pages in the same order as an insertion would, to avoid
* deadlocks: left, then right, then parent.
*/
lBuffer = ReadBufferExtended(gvs->index, MAIN_FORKNUM, leftBlkno,
RBM_NORMAL, gvs->strategy);
dBuffer = ReadBufferExtended(gvs->index, MAIN_FORKNUM, deleteBlkno,
RBM_NORMAL, gvs->strategy);
pBuffer = ReadBufferExtended(gvs->index, MAIN_FORKNUM, parentBlkno,
RBM_NORMAL, gvs->strategy);
LockBuffer(lBuffer, GIN_EXCLUSIVE);
LockBuffer(dBuffer, GIN_EXCLUSIVE);
if (!isParentRoot) /* parent is already locked by
* LockBufferForCleanup() */
LockBuffer(pBuffer, GIN_EXCLUSIVE);
START_CRIT_SECTION();
/* Unlink the page by changing left sibling's rightlink */
page = BufferGetPage(dBuffer);
rightlink = GinPageGetOpaque(page)->rightlink;
page = BufferGetPage(lBuffer);
GinPageGetOpaque(page)->rightlink = rightlink;
/* Delete downlink from parent */
parentPage = BufferGetPage(pBuffer);
#ifdef USE_ASSERT_CHECKING
do
{
PostingItem *tod = GinDataPageGetPostingItem(parentPage, myoff);
Assert(PostingItemGetBlockNumber(tod) == deleteBlkno);
} while (0);
#endif
GinPageDeletePostingItem(parentPage, myoff);
page = BufferGetPage(dBuffer);
/*
* we shouldn't change rightlink field to save workability of running
* search scan
*/
GinPageGetOpaque(page)->flags = GIN_DELETED;
MarkBufferDirty(pBuffer);
MarkBufferDirty(lBuffer);
MarkBufferDirty(dBuffer);
if (RelationNeedsWAL(gvs->index))
{
XLogRecPtr recptr;
XLogRecData rdata[4];
ginxlogDeletePage data;
data.node = gvs->index->rd_node;
data.blkno = deleteBlkno;
data.parentBlkno = parentBlkno;
data.parentOffset = myoff;
data.leftBlkno = leftBlkno;
data.rightLink = GinPageGetOpaque(page)->rightlink;
/*
* We can't pass buffer_std = TRUE, because we didn't set pd_lower on
* pre-9.4 versions. The page might've been binary-upgraded from an
* older version, and hence not have pd_lower set correctly. Ditto for
* the left page, but removing the item from the parent updated its
* pd_lower, so we know that's OK at this point.
*/
rdata[0].buffer = dBuffer;
rdata[0].buffer_std = FALSE;
rdata[0].data = NULL;
rdata[0].len = 0;
rdata[0].next = rdata + 1;
rdata[1].buffer = pBuffer;
rdata[1].buffer_std = TRUE;
rdata[1].data = NULL;
rdata[1].len = 0;
rdata[1].next = rdata + 2;
rdata[2].buffer = lBuffer;
rdata[2].buffer_std = FALSE;
rdata[2].data = NULL;
rdata[2].len = 0;
rdata[2].next = rdata + 3;
rdata[3].buffer = InvalidBuffer;
rdata[3].buffer_std = FALSE;
rdata[3].len = sizeof(ginxlogDeletePage);
rdata[3].data = (char *) &data;
rdata[3].next = NULL;
recptr = XLogInsert(RM_GIN_ID, XLOG_GIN_DELETE_PAGE, rdata);
PageSetLSN(page, recptr);
PageSetLSN(parentPage, recptr);
PageSetLSN(BufferGetPage(lBuffer), recptr);
}
if (!isParentRoot)
LockBuffer(pBuffer, GIN_UNLOCK);
ReleaseBuffer(pBuffer);
UnlockReleaseBuffer(lBuffer);
UnlockReleaseBuffer(dBuffer);
END_CRIT_SECTION();
gvs->result->pages_deleted++;
}
/*
* lazy_scan_heap() -- scan an open heap relation
*
* This routine sets commit status bits, builds lists of dead tuples
* and pages with free space, and calculates statistics on the number
* of live tuples in the heap. When done, or when we run low on space
* for dead-tuple TIDs, invoke vacuuming of indexes and heap.
*
* If there are no indexes then we just vacuum each dirty page as we
* process it, since there's no point in gathering many tuples.
*/
static void
lazy_scan_heap(Relation onerel, LVRelStats *vacrelstats,
Relation *Irel, int nindexes, bool scan_all)
{
BlockNumber nblocks,
blkno;
HeapTupleData tuple;
char *relname;
BlockNumber empty_pages,
vacuumed_pages;
double num_tuples,
tups_vacuumed,
nkeep,
nunused;
IndexBulkDeleteResult **indstats;
int i;
PGRUsage ru0;
Buffer vmbuffer = InvalidBuffer;
BlockNumber next_not_all_visible_block;
bool skipping_all_visible_blocks;
pg_rusage_init(&ru0);
relname = RelationGetRelationName(onerel);
ereport(elevel,
(errmsg("vacuuming \"%s.%s\"",
get_namespace_name(RelationGetNamespace(onerel)),
relname)));
empty_pages = vacuumed_pages = 0;
num_tuples = tups_vacuumed = nkeep = nunused = 0;
indstats = (IndexBulkDeleteResult **)
palloc0(nindexes * sizeof(IndexBulkDeleteResult *));
nblocks = RelationGetNumberOfBlocks(onerel);
vacrelstats->rel_pages = nblocks;
vacrelstats->scanned_pages = 0;
vacrelstats->nonempty_pages = 0;
vacrelstats->latestRemovedXid = InvalidTransactionId;
lazy_space_alloc(vacrelstats, nblocks);
/*
* We want to skip pages that don't require vacuuming according to the
* visibility map, but only when we can skip at least SKIP_PAGES_THRESHOLD
* consecutive pages. Since we're reading sequentially, the OS should be
* doing readahead for us, so there's no gain in skipping a page now and
* then; that's likely to disable readahead and so be counterproductive.
* Also, skipping even a single page means that we can't update
* relfrozenxid, so we only want to do it if we can skip a goodly number
* of pages.
*
* Before entering the main loop, establish the invariant that
* next_not_all_visible_block is the next block number >= blkno that's not
* all-visible according to the visibility map, or nblocks if there's no
* such block. Also, we set up the skipping_all_visible_blocks flag,
* which is needed because we need hysteresis in the decision: once we've
* started skipping blocks, we may as well skip everything up to the next
* not-all-visible block.
*
* Note: if scan_all is true, we won't actually skip any pages; but we
* maintain next_not_all_visible_block anyway, so as to set up the
* all_visible_according_to_vm flag correctly for each page.
*/
for (next_not_all_visible_block = 0;
next_not_all_visible_block < nblocks;
next_not_all_visible_block++)
{
if (!visibilitymap_test(onerel, next_not_all_visible_block, &vmbuffer))
break;
vacuum_delay_point();
}
if (next_not_all_visible_block >= SKIP_PAGES_THRESHOLD)
skipping_all_visible_blocks = true;
else
skipping_all_visible_blocks = false;
for (blkno = 0; blkno < nblocks; blkno++)
{
Buffer buf;
Page page;
OffsetNumber offnum,
maxoff;
bool tupgone,
hastup;
int prev_dead_count;
OffsetNumber frozen[MaxOffsetNumber];
int nfrozen;
Size freespace;
bool all_visible_according_to_vm;
bool all_visible;
bool has_dead_tuples;
if (blkno == next_not_all_visible_block)
{
/* Time to advance next_not_all_visible_block */
for (next_not_all_visible_block++;
next_not_all_visible_block < nblocks;
next_not_all_visible_block++)
{
if (!visibilitymap_test(onerel, next_not_all_visible_block,
&vmbuffer))
break;
vacuum_delay_point();
}
/*
* We know we can't skip the current block. But set up
* skipping_all_visible_blocks to do the right thing at the
* following blocks.
*/
if (next_not_all_visible_block - blkno > SKIP_PAGES_THRESHOLD)
skipping_all_visible_blocks = true;
else
skipping_all_visible_blocks = false;
all_visible_according_to_vm = false;
}
else
{
/* Current block is all-visible */
if (skipping_all_visible_blocks && !scan_all)
continue;
all_visible_according_to_vm = true;
}
vacuum_delay_point();
vacrelstats->scanned_pages++;
/*
* If we are close to overrunning the available space for dead-tuple
* TIDs, pause and do a cycle of vacuuming before we tackle this page.
*/
if ((vacrelstats->max_dead_tuples - vacrelstats->num_dead_tuples) < MaxHeapTuplesPerPage &&
vacrelstats->num_dead_tuples > 0)
{
/* Log cleanup info before we touch indexes */
vacuum_log_cleanup_info(onerel, vacrelstats);
/* Remove index entries */
for (i = 0; i < nindexes; i++)
lazy_vacuum_index(Irel[i],
&indstats[i],
vacrelstats);
/* Remove tuples from heap */
lazy_vacuum_heap(onerel, vacrelstats);
/*
* Forget the now-vacuumed tuples, and press on, but be careful
* not to reset latestRemovedXid since we want that value to be
* valid.
*/
vacrelstats->num_dead_tuples = 0;
vacrelstats->num_index_scans++;
}
buf = ReadBufferExtended(onerel, MAIN_FORKNUM, blkno,
RBM_NORMAL, vac_strategy);
/* We need buffer cleanup lock so that we can prune HOT chains. */
LockBufferForCleanup(buf);
page = BufferGetPage(buf);
if (PageIsNew(page))
{
/*
* An all-zeroes page could be left over if a backend extends the
* relation but crashes before initializing the page. Reclaim such
* pages for use.
*
* We have to be careful here because we could be looking at a
* page that someone has just added to the relation and not yet
* been able to initialize (see RelationGetBufferForTuple). To
* protect against that, release the buffer lock, grab the
* relation extension lock momentarily, and re-lock the buffer. If
* the page is still uninitialized by then, it must be left over
* from a crashed backend, and we can initialize it.
*
* We don't really need the relation lock when this is a new or
* temp relation, but it's probably not worth the code space to
* check that, since this surely isn't a critical path.
*
* Note: the comparable code in vacuum.c need not worry because
* it's got exclusive lock on the whole relation.
*/
LockBuffer(buf, BUFFER_LOCK_UNLOCK);
LockRelationForExtension(onerel, ExclusiveLock);
UnlockRelationForExtension(onerel, ExclusiveLock);
LockBufferForCleanup(buf);
if (PageIsNew(page))
{
ereport(WARNING,
(errmsg("relation \"%s\" page %u is uninitialized --- fixing",
relname, blkno)));
PageInit(page, BufferGetPageSize(buf), 0);
empty_pages++;
}
freespace = PageGetHeapFreeSpace(page);
MarkBufferDirty(buf);
UnlockReleaseBuffer(buf);
RecordPageWithFreeSpace(onerel, blkno, freespace);
continue;
}
if (PageIsEmpty(page))
{
empty_pages++;
freespace = PageGetHeapFreeSpace(page);
if (!PageIsAllVisible(page))
{
PageSetAllVisible(page);
SetBufferCommitInfoNeedsSave(buf);
}
LockBuffer(buf, BUFFER_LOCK_UNLOCK);
/* Update the visibility map */
if (!all_visible_according_to_vm)
{
visibilitymap_pin(onerel, blkno, &vmbuffer);
LockBuffer(buf, BUFFER_LOCK_SHARE);
if (PageIsAllVisible(page))
visibilitymap_set(onerel, blkno, PageGetLSN(page), &vmbuffer);
LockBuffer(buf, BUFFER_LOCK_UNLOCK);
}
ReleaseBuffer(buf);
RecordPageWithFreeSpace(onerel, blkno, freespace);
continue;
}
/*
* Prune all HOT-update chains in this page.
*
* We count tuples removed by the pruning step as removed by VACUUM.
*/
tups_vacuumed += heap_page_prune(onerel, buf, OldestXmin, false,
&vacrelstats->latestRemovedXid);
/*
* Now scan the page to collect vacuumable items and check for tuples
* requiring freezing.
*/
all_visible = true;
has_dead_tuples = false;
nfrozen = 0;
hastup = false;
prev_dead_count = vacrelstats->num_dead_tuples;
maxoff = PageGetMaxOffsetNumber(page);
for (offnum = FirstOffsetNumber;
offnum <= maxoff;
offnum = OffsetNumberNext(offnum))
{
ItemId itemid;
itemid = PageGetItemId(page, offnum);
/* Unused items require no processing, but we count 'em */
if (!ItemIdIsUsed(itemid))
{
nunused += 1;
continue;
}
/* Redirect items mustn't be touched */
if (ItemIdIsRedirected(itemid))
{
hastup = true; /* this page won't be truncatable */
continue;
}
ItemPointerSet(&(tuple.t_self), blkno, offnum);
/*
* DEAD item pointers are to be vacuumed normally; but we don't
* count them in tups_vacuumed, else we'd be double-counting (at
* least in the common case where heap_page_prune() just freed up
* a non-HOT tuple).
*/
if (ItemIdIsDead(itemid))
{
lazy_record_dead_tuple(vacrelstats, &(tuple.t_self));
all_visible = false;
continue;
}
Assert(ItemIdIsNormal(itemid));
tuple.t_data = (HeapTupleHeader) PageGetItem(page, itemid);
tuple.t_len = ItemIdGetLength(itemid);
tupgone = false;
switch (HeapTupleSatisfiesVacuum(tuple.t_data, OldestXmin, buf))
{
case HEAPTUPLE_DEAD:
/*
* Ordinarily, DEAD tuples would have been removed by
* heap_page_prune(), but it's possible that the tuple
* state changed since heap_page_prune() looked. In
* particular an INSERT_IN_PROGRESS tuple could have
* changed to DEAD if the inserter aborted. So this
* cannot be considered an error condition.
*
* If the tuple is HOT-updated then it must only be
* removed by a prune operation; so we keep it just as if
* it were RECENTLY_DEAD. Also, if it's a heap-only
* tuple, we choose to keep it, because it'll be a lot
* cheaper to get rid of it in the next pruning pass than
* to treat it like an indexed tuple.
*/
if (HeapTupleIsHotUpdated(&tuple) ||
HeapTupleIsHeapOnly(&tuple))
nkeep += 1;
else
tupgone = true; /* we can delete the tuple */
all_visible = false;
break;
case HEAPTUPLE_LIVE:
/* Tuple is good --- but let's do some validity checks */
if (onerel->rd_rel->relhasoids &&
!OidIsValid(HeapTupleGetOid(&tuple)))
elog(WARNING, "relation \"%s\" TID %u/%u: OID is invalid",
relname, blkno, offnum);
/*
* Is the tuple definitely visible to all transactions?
*
* NB: Like with per-tuple hint bits, we can't set the
* PD_ALL_VISIBLE flag if the inserter committed
* asynchronously. See SetHintBits for more info. Check
* that the HEAP_XMIN_COMMITTED hint bit is set because of
* that.
*/
if (all_visible)
{
TransactionId xmin;
if (!(tuple.t_data->t_infomask & HEAP_XMIN_COMMITTED))
{
all_visible = false;
break;
}
/*
* The inserter definitely committed. But is it old
* enough that everyone sees it as committed?
*/
xmin = HeapTupleHeaderGetXmin(tuple.t_data);
if (!TransactionIdPrecedes(xmin, OldestXmin))
{
all_visible = false;
break;
}
}
break;
case HEAPTUPLE_RECENTLY_DEAD:
/*
* If tuple is recently deleted then we must not remove it
* from relation.
*/
nkeep += 1;
all_visible = false;
break;
case HEAPTUPLE_INSERT_IN_PROGRESS:
/* This is an expected case during concurrent vacuum */
all_visible = false;
break;
case HEAPTUPLE_DELETE_IN_PROGRESS:
/* This is an expected case during concurrent vacuum */
all_visible = false;
break;
default:
elog(ERROR, "unexpected HeapTupleSatisfiesVacuum result");
break;
}
if (tupgone)
{
lazy_record_dead_tuple(vacrelstats, &(tuple.t_self));
HeapTupleHeaderAdvanceLatestRemovedXid(tuple.t_data,
&vacrelstats->latestRemovedXid);
tups_vacuumed += 1;
has_dead_tuples = true;
}
else
{
num_tuples += 1;
hastup = true;
/*
* Each non-removable tuple must be checked to see if it needs
* freezing. Note we already have exclusive buffer lock.
*/
if (heap_freeze_tuple(tuple.t_data, FreezeLimit,
InvalidBuffer))
frozen[nfrozen++] = offnum;
}
} /* scan along page */
/*
* If we froze any tuples, mark the buffer dirty, and write a WAL
* record recording the changes. We must log the changes to be
* crash-safe against future truncation of CLOG.
*/
if (nfrozen > 0)
{
MarkBufferDirty(buf);
if (RelationNeedsWAL(onerel))
{
XLogRecPtr recptr;
recptr = log_heap_freeze(onerel, buf, FreezeLimit,
frozen, nfrozen);
PageSetLSN(page, recptr);
PageSetTLI(page, ThisTimeLineID);
}
}
/*
* If there are no indexes then we can vacuum the page right now
* instead of doing a second scan.
*/
if (nindexes == 0 &&
vacrelstats->num_dead_tuples > 0)
{
/* Remove tuples from heap */
lazy_vacuum_page(onerel, blkno, buf, 0, vacrelstats);
has_dead_tuples = false;
/*
* Forget the now-vacuumed tuples, and press on, but be careful
* not to reset latestRemovedXid since we want that value to be
* valid.
*/
vacrelstats->num_dead_tuples = 0;
vacuumed_pages++;
}
freespace = PageGetHeapFreeSpace(page);
/* Update the all-visible flag on the page */
if (!PageIsAllVisible(page) && all_visible)
{
PageSetAllVisible(page);
SetBufferCommitInfoNeedsSave(buf);
}
/*
* It's possible for the value returned by GetOldestXmin() to move
* backwards, so it's not wrong for us to see tuples that appear to
* not be visible to everyone yet, while PD_ALL_VISIBLE is already
* set. The real safe xmin value never moves backwards, but
* GetOldestXmin() is conservative and sometimes returns a value
* that's unnecessarily small, so if we see that contradiction it just
* means that the tuples that we think are not visible to everyone yet
* actually are, and the PD_ALL_VISIBLE flag is correct.
*
* There should never be dead tuples on a page with PD_ALL_VISIBLE
* set, however.
*/
else if (PageIsAllVisible(page) && has_dead_tuples)
{
elog(WARNING, "page containing dead tuples is marked as all-visible in relation \"%s\" page %u",
relname, blkno);
PageClearAllVisible(page);
SetBufferCommitInfoNeedsSave(buf);
/*
* Normally, we would drop the lock on the heap page before
* updating the visibility map, but since this case shouldn't
* happen anyway, don't worry about that.
*/
visibilitymap_clear(onerel, blkno);
}
LockBuffer(buf, BUFFER_LOCK_UNLOCK);
/* Update the visibility map */
if (!all_visible_according_to_vm && all_visible)
{
visibilitymap_pin(onerel, blkno, &vmbuffer);
LockBuffer(buf, BUFFER_LOCK_SHARE);
if (PageIsAllVisible(page))
visibilitymap_set(onerel, blkno, PageGetLSN(page), &vmbuffer);
LockBuffer(buf, BUFFER_LOCK_UNLOCK);
}
ReleaseBuffer(buf);
/* Remember the location of the last page with nonremovable tuples */
if (hastup)
vacrelstats->nonempty_pages = blkno + 1;
/*
* If we remembered any tuples for deletion, then the page will be
* visited again by lazy_vacuum_heap, which will compute and record
* its post-compaction free space. If not, then we're done with this
* page, so remember its free space as-is. (This path will always be
* taken if there are no indexes.)
*/
if (vacrelstats->num_dead_tuples == prev_dead_count)
RecordPageWithFreeSpace(onerel, blkno, freespace);
}
/* save stats for use later */
vacrelstats->scanned_tuples = num_tuples;
vacrelstats->tuples_deleted = tups_vacuumed;
/* now we can compute the new value for pg_class.reltuples */
vacrelstats->new_rel_tuples = vac_estimate_reltuples(onerel, false,
nblocks,
vacrelstats->scanned_pages,
num_tuples);
/*
* Release any remaining pin on visibility map page.
*/
if (BufferIsValid(vmbuffer))
{
ReleaseBuffer(vmbuffer);
vmbuffer = InvalidBuffer;
}
/* If any tuples need to be deleted, perform final vacuum cycle */
/* XXX put a threshold on min number of tuples here? */
if (vacrelstats->num_dead_tuples > 0)
{
/* Log cleanup info before we touch indexes */
vacuum_log_cleanup_info(onerel, vacrelstats);
/* Remove index entries */
for (i = 0; i < nindexes; i++)
lazy_vacuum_index(Irel[i],
&indstats[i],
vacrelstats);
/* Remove tuples from heap */
lazy_vacuum_heap(onerel, vacrelstats);
vacrelstats->num_index_scans++;
}
/* Do post-vacuum cleanup and statistics update for each index */
for (i = 0; i < nindexes; i++)
lazy_cleanup_index(Irel[i], indstats[i], vacrelstats);
/* If no indexes, make log report that lazy_vacuum_heap would've made */
if (vacuumed_pages)
ereport(elevel,
(errmsg("\"%s\": removed %.0f row versions in %u pages",
RelationGetRelationName(onerel),
tups_vacuumed, vacuumed_pages)));
ereport(elevel,
(errmsg("\"%s\": found %.0f removable, %.0f nonremovable row versions in %u out of %u pages",
RelationGetRelationName(onerel),
tups_vacuumed, num_tuples,
vacrelstats->scanned_pages, nblocks),
errdetail("%.0f dead row versions cannot be removed yet.\n"
"There were %.0f unused item pointers.\n"
"%u pages are entirely empty.\n"
"%s.",
nkeep,
nunused,
empty_pages,
pg_rusage_show(&ru0))));
}
/*
* Create a WAL record for vacuuming entry tree leaf page.
*/
static void
xlogVacuumPage(Relation index, Buffer buffer)
{
Page page = BufferGetPage(buffer);
XLogRecPtr recptr;
XLogRecData rdata[3];
ginxlogVacuumPage xlrec;
uint16 lower;
uint16 upper;
/* This is only used for entry tree leaf pages. */
Assert(!GinPageIsData(page));
Assert(GinPageIsLeaf(page));
if (!RelationNeedsWAL(index))
return;
xlrec.node = index->rd_node;
xlrec.blkno = BufferGetBlockNumber(buffer);
/* Assume we can omit data between pd_lower and pd_upper */
lower = ((PageHeader) page)->pd_lower;
upper = ((PageHeader) page)->pd_upper;
Assert(lower < BLCKSZ);
Assert(upper < BLCKSZ);
if (lower >= SizeOfPageHeaderData &&
upper > lower &&
upper <= BLCKSZ)
{
xlrec.hole_offset = lower;
xlrec.hole_length = upper - lower;
}
else
{
/* No "hole" to compress out */
xlrec.hole_offset = 0;
xlrec.hole_length = 0;
}
rdata[0].data = (char *) &xlrec;
rdata[0].len = sizeof(ginxlogVacuumPage);
rdata[0].buffer = InvalidBuffer;
rdata[0].next = &rdata[1];
if (xlrec.hole_length == 0)
{
rdata[1].data = (char *) page;
rdata[1].len = BLCKSZ;
rdata[1].buffer = InvalidBuffer;
rdata[1].next = NULL;
}
else
{
/* must skip the hole */
rdata[1].data = (char *) page;
rdata[1].len = xlrec.hole_offset;
rdata[1].buffer = InvalidBuffer;
rdata[1].next = &rdata[2];
rdata[2].data = (char *) page + (xlrec.hole_offset + xlrec.hole_length);
rdata[2].len = BLCKSZ - (xlrec.hole_offset + xlrec.hole_length);
rdata[2].buffer = InvalidBuffer;
rdata[2].next = NULL;
}
recptr = XLogInsert(RM_GIN_ID, XLOG_GIN_VACUUM_PAGE, rdata);
PageSetLSN(page, recptr);
}
/*
* Deletes pending list pages up to (not including) newHead page.
* If newHead == InvalidBlockNumber then function drops the whole list.
*
* metapage is pinned and exclusive-locked throughout this function.
*/
static void
shiftList(Relation index, Buffer metabuffer, BlockNumber newHead,
bool fill_fsm, IndexBulkDeleteResult *stats)
{
Page metapage;
GinMetaPageData *metadata;
BlockNumber blknoToDelete;
metapage = BufferGetPage(metabuffer);
metadata = GinPageGetMeta(metapage);
blknoToDelete = metadata->head;
do
{
Page page;
int i;
int64 nDeletedHeapTuples = 0;
ginxlogDeleteListPages data;
Buffer buffers[GIN_NDELETE_AT_ONCE];
BlockNumber freespace[GIN_NDELETE_AT_ONCE];
data.ndeleted = 0;
while (data.ndeleted < GIN_NDELETE_AT_ONCE && blknoToDelete != newHead)
{
freespace[data.ndeleted] = blknoToDelete;
buffers[data.ndeleted] = ReadBuffer(index, blknoToDelete);
LockBuffer(buffers[data.ndeleted], GIN_EXCLUSIVE);
page = BufferGetPage(buffers[data.ndeleted]);
data.ndeleted++;
Assert(!GinPageIsDeleted(page));
nDeletedHeapTuples += GinPageGetOpaque(page)->maxoff;
blknoToDelete = GinPageGetOpaque(page)->rightlink;
}
if (stats)
stats->pages_deleted += data.ndeleted;
/*
* This operation touches an unusually large number of pages, so
* prepare the XLogInsert machinery for that before entering the
* critical section.
*/
if (RelationNeedsWAL(index))
XLogEnsureRecordSpace(data.ndeleted, 0);
START_CRIT_SECTION();
metadata->head = blknoToDelete;
Assert(metadata->nPendingPages >= data.ndeleted);
metadata->nPendingPages -= data.ndeleted;
Assert(metadata->nPendingHeapTuples >= nDeletedHeapTuples);
metadata->nPendingHeapTuples -= nDeletedHeapTuples;
if (blknoToDelete == InvalidBlockNumber)
{
metadata->tail = InvalidBlockNumber;
metadata->tailFreeSize = 0;
metadata->nPendingPages = 0;
metadata->nPendingHeapTuples = 0;
}
/*
* Set pd_lower just past the end of the metadata. This is essential,
* because without doing so, metadata will be lost if xlog.c
* compresses the page. (We must do this here because pre-v11
* versions of PG did not set the metapage's pd_lower correctly, so a
* pg_upgraded index might contain the wrong value.)
*/
((PageHeader) metapage)->pd_lower =
((char *) metadata + sizeof(GinMetaPageData)) - (char *) metapage;
MarkBufferDirty(metabuffer);
for (i = 0; i < data.ndeleted; i++)
{
page = BufferGetPage(buffers[i]);
GinPageGetOpaque(page)->flags = GIN_DELETED;
MarkBufferDirty(buffers[i]);
}
if (RelationNeedsWAL(index))
{
XLogRecPtr recptr;
XLogBeginInsert();
XLogRegisterBuffer(0, metabuffer,
REGBUF_WILL_INIT | REGBUF_STANDARD);
for (i = 0; i < data.ndeleted; i++)
XLogRegisterBuffer(i + 1, buffers[i], REGBUF_WILL_INIT);
memcpy(&data.metadata, metadata, sizeof(GinMetaPageData));
XLogRegisterData((char *) &data,
sizeof(ginxlogDeleteListPages));
recptr = XLogInsert(RM_GIN_ID, XLOG_GIN_DELETE_LISTPAGE);
PageSetLSN(metapage, recptr);
for (i = 0; i < data.ndeleted; i++)
{
page = BufferGetPage(buffers[i]);
PageSetLSN(page, recptr);
}
}
for (i = 0; i < data.ndeleted; i++)
UnlockReleaseBuffer(buffers[i]);
END_CRIT_SECTION();
for (i = 0; fill_fsm && i < data.ndeleted; i++)
RecordFreeIndexPage(index, freespace[i]);
} while (blknoToDelete != newHead);
}
/*
* 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);
/* NO EREPORT(ERROR) from here till changes are logged */
START_CRIT_SECTION();
fsm_truncate_avail(BufferGetPage(buf), first_removed_slot);
/*
* Truncation of a relation is WAL-logged at a higher-level, and we
* will be called at WAL replay. But if checksums are enabled, we need
* to still write a WAL record to protect against a torn page, if the
* page is flushed to disk before the truncation WAL record. We cannot
* use MarkBufferDirtyHint here, because that will not dirty the page
* during recovery.
*/
MarkBufferDirty(buf);
if (!InRecovery && RelationNeedsWAL(rel) && XLogHintBitIsNeeded())
log_newpage_buffer(buf, false);
END_CRIT_SECTION();
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;
/*
* Update upper-level FSM pages to account for the truncation. This is
* important because the just-truncated pages were likely marked as
* all-free, and would be preferentially selected.
*/
FreeSpaceMapVacuumRange(rel, nblocks, InvalidBlockNumber);
}
/*
* Insert value (stored in GinBtree) to tree described by stack
*
* During an index build, buildStats is non-null and the counters
* it contains should be incremented as needed.
*
* NB: the passed-in stack is freed, as though by freeGinBtreeStack.
*/
void
ginInsertValue(GinBtree btree, GinBtreeStack *stack, GinStatsData *buildStats)
{
GinBtreeStack *parent = stack;
BlockNumber rootBlkno = InvalidBuffer;
Page page,
rpage,
lpage;
/* remember root BlockNumber */
while (parent)
{
rootBlkno = parent->blkno;
parent = parent->parent;
}
while (stack)
{
XLogRecData *rdata;
BlockNumber savedRightLink;
page = BufferGetPage(stack->buffer);
savedRightLink = GinPageGetOpaque(page)->rightlink;
if (btree->isEnoughSpace(btree, stack->buffer, stack->off))
{
START_CRIT_SECTION();
btree->placeToPage(btree, stack->buffer, stack->off, &rdata);
MarkBufferDirty(stack->buffer);
if (RelationNeedsWAL(btree->index))
{
XLogRecPtr recptr;
recptr = XLogInsert(RM_GIN_ID, XLOG_GIN_INSERT, rdata);
PageSetLSN(page, recptr);
PageSetTLI(page, ThisTimeLineID);
}
LockBuffer(stack->buffer, GIN_UNLOCK);
END_CRIT_SECTION();
freeGinBtreeStack(stack);
return;
}
else
{
Buffer rbuffer = GinNewBuffer(btree->index);
Page newlpage;
/*
* newlpage is a pointer to memory page, it doesn't associate with
* buffer, stack->buffer should be untouched
*/
newlpage = btree->splitPage(btree, stack->buffer, rbuffer, stack->off, &rdata);
((ginxlogSplit *) (rdata->data))->rootBlkno = rootBlkno;
/* During index build, count the newly-split page */
if (buildStats)
{
if (btree->isData)
buildStats->nDataPages++;
else
buildStats->nEntryPages++;
}
parent = stack->parent;
if (parent == NULL)
{
/*
* split root, so we need to allocate new left page and place
* pointer on root to left and right page
*/
Buffer lbuffer = GinNewBuffer(btree->index);
((ginxlogSplit *) (rdata->data))->isRootSplit = TRUE;
((ginxlogSplit *) (rdata->data))->rrlink = InvalidBlockNumber;
page = BufferGetPage(stack->buffer);
lpage = BufferGetPage(lbuffer);
rpage = BufferGetPage(rbuffer);
GinPageGetOpaque(rpage)->rightlink = InvalidBlockNumber;
GinPageGetOpaque(newlpage)->rightlink = BufferGetBlockNumber(rbuffer);
((ginxlogSplit *) (rdata->data))->lblkno = BufferGetBlockNumber(lbuffer);
START_CRIT_SECTION();
GinInitBuffer(stack->buffer, GinPageGetOpaque(newlpage)->flags & ~GIN_LEAF);
PageRestoreTempPage(newlpage, lpage);
btree->fillRoot(btree, stack->buffer, lbuffer, rbuffer);
MarkBufferDirty(rbuffer);
MarkBufferDirty(lbuffer);
MarkBufferDirty(stack->buffer);
if (RelationNeedsWAL(btree->index))
{
XLogRecPtr recptr;
recptr = XLogInsert(RM_GIN_ID, XLOG_GIN_SPLIT, rdata);
PageSetLSN(page, recptr);
PageSetTLI(page, ThisTimeLineID);
PageSetLSN(lpage, recptr);
PageSetTLI(lpage, ThisTimeLineID);
PageSetLSN(rpage, recptr);
PageSetTLI(rpage, ThisTimeLineID);
}
UnlockReleaseBuffer(rbuffer);
UnlockReleaseBuffer(lbuffer);
LockBuffer(stack->buffer, GIN_UNLOCK);
END_CRIT_SECTION();
freeGinBtreeStack(stack);
/* During index build, count the newly-added root page */
if (buildStats)
{
if (btree->isData)
buildStats->nDataPages++;
else
buildStats->nEntryPages++;
}
return;
}
else
{
/* split non-root page */
((ginxlogSplit *) (rdata->data))->isRootSplit = FALSE;
((ginxlogSplit *) (rdata->data))->rrlink = savedRightLink;
lpage = BufferGetPage(stack->buffer);
rpage = BufferGetPage(rbuffer);
GinPageGetOpaque(rpage)->rightlink = savedRightLink;
GinPageGetOpaque(newlpage)->rightlink = BufferGetBlockNumber(rbuffer);
START_CRIT_SECTION();
PageRestoreTempPage(newlpage, lpage);
MarkBufferDirty(rbuffer);
MarkBufferDirty(stack->buffer);
if (RelationNeedsWAL(btree->index))
{
XLogRecPtr recptr;
recptr = XLogInsert(RM_GIN_ID, XLOG_GIN_SPLIT, rdata);
PageSetLSN(lpage, recptr);
PageSetTLI(lpage, ThisTimeLineID);
PageSetLSN(rpage, recptr);
PageSetTLI(rpage, ThisTimeLineID);
}
UnlockReleaseBuffer(rbuffer);
END_CRIT_SECTION();
}
}
btree->isDelete = FALSE;
/* search parent to lock */
LockBuffer(parent->buffer, GIN_EXCLUSIVE);
/* move right if it's needed */
page = BufferGetPage(parent->buffer);
while ((parent->off = btree->findChildPtr(btree, page, stack->blkno, parent->off)) == InvalidOffsetNumber)
{
BlockNumber rightlink = GinPageGetOpaque(page)->rightlink;
LockBuffer(parent->buffer, GIN_UNLOCK);
if (rightlink == InvalidBlockNumber)
{
/*
* rightmost page, but we don't find parent, we should use
* plain search...
*/
ginFindParents(btree, stack, rootBlkno);
parent = stack->parent;
page = BufferGetPage(parent->buffer);
break;
}
parent->blkno = rightlink;
parent->buffer = ReleaseAndReadBuffer(parent->buffer, btree->index, parent->blkno);
LockBuffer(parent->buffer, GIN_EXCLUSIVE);
page = BufferGetPage(parent->buffer);
}
UnlockReleaseBuffer(stack->buffer);
pfree(stack);
stack = parent;
}
}
/*
* Bulk deletion of all index entries pointing to a set of heap tuples and
* check invalid tuples after crash recovery.
* The set of target tuples is specified via a callback routine that tells
* whether any given heap tuple (identified by ItemPointer) is being deleted.
*
* Result: a palloc'd struct containing statistical info for VACUUM displays.
*/
Datum
gistbulkdelete(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 rel = info->index;
GistBDItem *stack,
*ptr;
/* first time through? */
if (stats == NULL)
stats = (IndexBulkDeleteResult *) palloc0(sizeof(IndexBulkDeleteResult));
/* we'll re-count the tuples each time */
stats->estimated_count = false;
stats->num_index_tuples = 0;
stack = (GistBDItem *) palloc0(sizeof(GistBDItem));
stack->blkno = GIST_ROOT_BLKNO;
while (stack)
{
Buffer buffer;
Page page;
OffsetNumber i,
maxoff;
IndexTuple idxtuple;
ItemId iid;
buffer = ReadBufferExtended(rel, MAIN_FORKNUM, stack->blkno,
RBM_NORMAL, info->strategy);
LockBuffer(buffer, GIST_SHARE);
gistcheckpage(rel, buffer);
page = (Page) BufferGetPage(buffer);
if (GistPageIsLeaf(page))
{
OffsetNumber todelete[MaxOffsetNumber];
int ntodelete = 0;
LockBuffer(buffer, GIST_UNLOCK);
LockBuffer(buffer, GIST_EXCLUSIVE);
page = (Page) BufferGetPage(buffer);
if (stack->blkno == GIST_ROOT_BLKNO && !GistPageIsLeaf(page))
{
/* only the root can become non-leaf during relock */
UnlockReleaseBuffer(buffer);
/* one more check */
continue;
}
/*
* check for split proceeded after look at parent, we should check
* it after relock
*/
pushStackIfSplited(page, stack);
/*
* Remove deletable tuples from page
*/
maxoff = PageGetMaxOffsetNumber(page);
for (i = FirstOffsetNumber; i <= maxoff; i = OffsetNumberNext(i))
{
iid = PageGetItemId(page, i);
idxtuple = (IndexTuple) PageGetItem(page, iid);
if (callback(&(idxtuple->t_tid), callback_state))
{
todelete[ntodelete] = i - ntodelete;
ntodelete++;
stats->tuples_removed += 1;
}
else
stats->num_index_tuples += 1;
}
if (ntodelete)
{
START_CRIT_SECTION();
MarkBufferDirty(buffer);
for (i = 0; i < ntodelete; i++)
PageIndexTupleDelete(page, todelete[i]);
GistMarkTuplesDeleted(page);
if (RelationNeedsWAL(rel))
{
XLogRecPtr recptr;
recptr = gistXLogUpdate(rel->rd_node, buffer,
todelete, ntodelete,
NULL, 0, InvalidBuffer);
PageSetLSN(page, recptr);
PageSetTLI(page, ThisTimeLineID);
}
else
PageSetLSN(page, GetXLogRecPtrForTemp());
END_CRIT_SECTION();
}
}
else
{
/* check for split proceeded after look at parent */
pushStackIfSplited(page, stack);
maxoff = PageGetMaxOffsetNumber(page);
for (i = FirstOffsetNumber; i <= maxoff; i = OffsetNumberNext(i))
{
iid = PageGetItemId(page, i);
idxtuple = (IndexTuple) PageGetItem(page, iid);
ptr = (GistBDItem *) palloc(sizeof(GistBDItem));
ptr->blkno = ItemPointerGetBlockNumber(&(idxtuple->t_tid));
ptr->parentlsn = PageGetLSN(page);
ptr->next = stack->next;
stack->next = ptr;
if (GistTupleIsInvalid(idxtuple))
ereport(LOG,
(errmsg("index \"%s\" contains an inner tuple marked as invalid",
RelationGetRelationName(rel)),
errdetail("This is caused by an incomplete page split at crash recovery before upgrading to PostgreSQL 9.1."),
errhint("Please REINDEX it.")));
}
}
UnlockReleaseBuffer(buffer);
ptr = stack->next;
pfree(stack);
stack = ptr;
vacuum_delay_point();
}
PG_RETURN_POINTER(stats);
}
/*
* Workhouse routine for doing insertion into a GiST index. Note that
* this routine assumes it is invoked in a short-lived memory context,
* so it does not bother releasing palloc'd allocations.
*/
void
gistdoinsert(Relation r, IndexTuple itup, Size freespace,
GISTSTATE *giststate, Relation heapRel, bool is_build)
{
ItemId iid;
IndexTuple idxtuple;
GISTInsertStack firststack;
GISTInsertStack *stack;
GISTInsertState state;
bool xlocked = false;
memset(&state, 0, sizeof(GISTInsertState));
state.freespace = freespace;
state.r = r;
state.heapRel = heapRel;
state.is_build = is_build;
/* Start from the root */
firststack.blkno = GIST_ROOT_BLKNO;
firststack.lsn = 0;
firststack.parent = NULL;
firststack.downlinkoffnum = InvalidOffsetNumber;
state.stack = stack = &firststack;
/*
* Walk down along the path of smallest penalty, updating the parent
* pointers with the key we're inserting as we go. If we crash in the
* middle, the tree is consistent, although the possible parent updates
* were a waste.
*/
for (;;)
{
if (XLogRecPtrIsInvalid(stack->lsn))
stack->buffer = ReadBuffer(state.r, stack->blkno);
/*
* Be optimistic and grab shared lock first. Swap it for an exclusive
* lock later if we need to update the page.
*/
if (!xlocked)
{
LockBuffer(stack->buffer, GIST_SHARE);
gistcheckpage(state.r, stack->buffer);
}
stack->page = (Page) BufferGetPage(stack->buffer);
stack->lsn = xlocked ?
PageGetLSN(stack->page) : BufferGetLSNAtomic(stack->buffer);
Assert(!RelationNeedsWAL(state.r) || !XLogRecPtrIsInvalid(stack->lsn));
/*
* If this page was split but the downlink was never inserted to the
* parent because the inserting backend crashed before doing that, fix
* that now.
*/
if (GistFollowRight(stack->page))
{
if (!xlocked)
{
LockBuffer(stack->buffer, GIST_UNLOCK);
LockBuffer(stack->buffer, GIST_EXCLUSIVE);
xlocked = true;
/* someone might've completed the split when we unlocked */
if (!GistFollowRight(stack->page))
continue;
}
gistfixsplit(&state, giststate);
UnlockReleaseBuffer(stack->buffer);
xlocked = false;
state.stack = stack = stack->parent;
continue;
}
if (stack->blkno != GIST_ROOT_BLKNO &&
stack->parent->lsn < GistPageGetNSN(stack->page))
{
/*
* Concurrent split detected. There's no guarantee that the
* downlink for this page is consistent with the tuple we're
* inserting anymore, so go back to parent and rechoose the best
* child.
*/
UnlockReleaseBuffer(stack->buffer);
xlocked = false;
state.stack = stack = stack->parent;
continue;
}
if (!GistPageIsLeaf(stack->page))
{
/*
* This is an internal page so continue to walk down the tree.
* Find the child node that has the minimum insertion penalty.
*/
BlockNumber childblkno;
IndexTuple newtup;
GISTInsertStack *item;
OffsetNumber downlinkoffnum;
/* currently, internal pages are never deleted */
Assert(!GistPageIsDeleted(stack->page));
downlinkoffnum = gistchoose(state.r, stack->page, itup, giststate);
iid = PageGetItemId(stack->page, downlinkoffnum);
idxtuple = (IndexTuple) PageGetItem(stack->page, iid);
childblkno = ItemPointerGetBlockNumber(&(idxtuple->t_tid));
/*
* Check that it's not a leftover invalid tuple from pre-9.1
*/
if (GistTupleIsInvalid(idxtuple))
ereport(ERROR,
(errmsg("index \"%s\" contains an inner tuple marked as invalid",
RelationGetRelationName(r)),
errdetail("This is caused by an incomplete page split at crash recovery before upgrading to PostgreSQL 9.1."),
errhint("Please REINDEX it.")));
/*
* Check that the key representing the target child node is
* consistent with the key we're inserting. Update it if it's not.
*/
newtup = gistgetadjusted(state.r, idxtuple, itup, giststate);
if (newtup)
{
/*
* Swap shared lock for an exclusive one. Beware, the page may
* change while we unlock/lock the page...
*/
if (!xlocked)
{
LockBuffer(stack->buffer, GIST_UNLOCK);
LockBuffer(stack->buffer, GIST_EXCLUSIVE);
xlocked = true;
stack->page = (Page) BufferGetPage(stack->buffer);
if (PageGetLSN(stack->page) != stack->lsn)
{
/* the page was changed while we unlocked it, retry */
continue;
}
}
/*
* Update the tuple.
*
* We still hold the lock after gistinserttuple(), but it
* might have to split the page to make the updated tuple fit.
* In that case the updated tuple might migrate to the other
* half of the split, so we have to go back to the parent and
* descend back to the half that's a better fit for the new
* tuple.
*/
if (gistinserttuple(&state, stack, giststate, newtup,
downlinkoffnum))
{
/*
* If this was a root split, the root page continues to be
* the parent and the updated tuple went to one of the
* child pages, so we just need to retry from the root
* page.
*/
if (stack->blkno != GIST_ROOT_BLKNO)
{
UnlockReleaseBuffer(stack->buffer);
xlocked = false;
state.stack = stack = stack->parent;
}
continue;
}
}
LockBuffer(stack->buffer, GIST_UNLOCK);
xlocked = false;
/* descend to the chosen child */
item = (GISTInsertStack *) palloc0(sizeof(GISTInsertStack));
item->blkno = childblkno;
item->parent = stack;
item->downlinkoffnum = downlinkoffnum;
state.stack = stack = item;
}
else
{
/*
* Leaf page. Insert the new key. We've already updated all the
* parents on the way down, but we might have to split the page if
* it doesn't fit. gistinserthere() will take care of that.
*/
/*
* Swap shared lock for an exclusive one. Be careful, the page may
* change while we unlock/lock the page...
*/
if (!xlocked)
{
LockBuffer(stack->buffer, GIST_UNLOCK);
LockBuffer(stack->buffer, GIST_EXCLUSIVE);
xlocked = true;
stack->page = (Page) BufferGetPage(stack->buffer);
stack->lsn = PageGetLSN(stack->page);
if (stack->blkno == GIST_ROOT_BLKNO)
{
/*
* the only page that can become inner instead of leaf is
* the root page, so for root we should recheck it
*/
if (!GistPageIsLeaf(stack->page))
{
/*
* very rare situation: during unlock/lock index with
* number of pages = 1 was increased
*/
LockBuffer(stack->buffer, GIST_UNLOCK);
xlocked = false;
continue;
}
/*
* we don't need to check root split, because checking
* leaf/inner is enough to recognize split for root
*/
}
else if (GistFollowRight(stack->page) ||
stack->parent->lsn < GistPageGetNSN(stack->page))
{
/*
* The page was split while we momentarily unlocked the
* page. Go back to parent.
*/
UnlockReleaseBuffer(stack->buffer);
xlocked = false;
state.stack = stack = stack->parent;
continue;
}
}
/*
* The page might have been deleted after we scanned the parent
* and saw the downlink.
*/
if (GistPageIsDeleted(stack->page))
{
UnlockReleaseBuffer(stack->buffer);
xlocked = false;
state.stack = stack = stack->parent;
continue;
}
/* now state.stack->(page, buffer and blkno) points to leaf page */
gistinserttuple(&state, stack, giststate, itup,
InvalidOffsetNumber);
LockBuffer(stack->buffer, GIST_UNLOCK);
/* Release any pins we might still hold before exiting */
for (; stack; stack = stack->parent)
ReleaseBuffer(stack->buffer);
break;
}
}
}
/*
* Main entry point to GiST index build. Initially calls insert over and over,
* but switches to more efficient buffering build algorithm after a certain
* number of tuples (unless buffering mode is disabled).
*/
Datum
gistbuild(PG_FUNCTION_ARGS)
{
Relation heap = (Relation) PG_GETARG_POINTER(0);
Relation index = (Relation) PG_GETARG_POINTER(1);
IndexInfo *indexInfo = (IndexInfo *) PG_GETARG_POINTER(2);
IndexBuildResult *result;
double reltuples;
GISTBuildState buildstate;
Buffer buffer;
Page page;
MemoryContext oldcxt = CurrentMemoryContext;
int fillfactor;
buildstate.indexrel = index;
if (index->rd_options)
{
/* Get buffering mode from the options string */
GiSTOptions *options = (GiSTOptions *) index->rd_options;
char *bufferingMode = (char *) options + options->bufferingModeOffset;
if (strcmp(bufferingMode, "on") == 0)
buildstate.bufferingMode = GIST_BUFFERING_STATS;
else if (strcmp(bufferingMode, "off") == 0)
buildstate.bufferingMode = GIST_BUFFERING_DISABLED;
else
buildstate.bufferingMode = GIST_BUFFERING_AUTO;
fillfactor = options->fillfactor;
}
else
{
/*
* By default, switch to buffering mode when the index grows too large
* to fit in cache.
*/
buildstate.bufferingMode = GIST_BUFFERING_AUTO;
fillfactor = GIST_DEFAULT_FILLFACTOR;
}
/* Calculate target amount of free space to leave on pages */
buildstate.freespace = BLCKSZ * (100 - fillfactor) / 100;
/*
* We expect to be called exactly once for any index relation. If that's
* not the case, big trouble's what we have.
*/
if (RelationGetNumberOfBlocks(index) != 0)
elog(ERROR, "index \"%s\" already contains data",
RelationGetRelationName(index));
/*
* We can't yet handle unlogged GiST indexes, because we depend on LSNs.
* This is duplicative of an error in gistbuildempty, but we want to check
* here so as to throw error before doing all the index-build work.
*/
if (heap->rd_rel->relpersistence == RELPERSISTENCE_UNLOGGED)
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
errmsg("unlogged GiST indexes are not supported")));
/* no locking is needed */
buildstate.giststate = initGISTstate(index);
/*
* Create a temporary memory context that is reset once for each tuple
* processed. (Note: we don't bother to make this a child of the
* giststate's scanCxt, so we have to delete it separately at the end.)
*/
buildstate.giststate->tempCxt = createTempGistContext();
/* initialize the root page */
buffer = gistNewBuffer(index);
Assert(BufferGetBlockNumber(buffer) == GIST_ROOT_BLKNO);
page = BufferGetPage(buffer);
START_CRIT_SECTION();
GISTInitBuffer(buffer, F_LEAF);
MarkBufferDirty(buffer);
if (RelationNeedsWAL(index))
{
XLogRecPtr recptr;
XLogRecData rdata;
rdata.data = (char *) &(index->rd_node);
rdata.len = sizeof(RelFileNode);
rdata.buffer = InvalidBuffer;
rdata.next = NULL;
recptr = XLogInsert(RM_GIST_ID, XLOG_GIST_CREATE_INDEX, &rdata);
PageSetLSN(page, recptr);
PageSetTLI(page, ThisTimeLineID);
}
else
PageSetLSN(page, GetXLogRecPtrForTemp());
UnlockReleaseBuffer(buffer);
END_CRIT_SECTION();
/* build the index */
buildstate.indtuples = 0;
buildstate.indtuplesSize = 0;
/*
* Do the heap scan.
*/
reltuples = IndexBuildHeapScan(heap, index, indexInfo, true,
gistBuildCallback, (void *) &buildstate);
/*
* If buffering was used, flush out all the tuples that are still in the
* buffers.
*/
if (buildstate.bufferingMode == GIST_BUFFERING_ACTIVE)
{
elog(DEBUG1, "all tuples processed, emptying buffers");
gistEmptyAllBuffers(&buildstate);
}
/* okay, all heap tuples are indexed */
MemoryContextSwitchTo(oldcxt);
MemoryContextDelete(buildstate.giststate->tempCxt);
freeGISTstate(buildstate.giststate);
/*
* Return statistics
*/
result = (IndexBuildResult *) palloc(sizeof(IndexBuildResult));
result->heap_tuples = reltuples;
result->index_tuples = (double) buildstate.indtuples;
PG_RETURN_POINTER(result);
}
/*
* get_relation_info -
* Retrieves catalog information for a given relation.
*
* Given the Oid of the relation, return the following info into fields
* of the RelOptInfo struct:
*
* min_attr lowest valid AttrNumber
* max_attr highest valid AttrNumber
* indexlist list of IndexOptInfos for relation's indexes
* serverid if it's a foreign table, the server OID
* fdwroutine if it's a foreign table, the FDW function pointers
* pages number of pages
* tuples number of tuples
*
* Also, initialize the attr_needed[] and attr_widths[] arrays. In most
* cases these are left as zeroes, but sometimes we need to compute attr
* widths here, and we may as well cache the results for costsize.c.
*
* If inhparent is true, all we need to do is set up the attr arrays:
* the RelOptInfo actually represents the appendrel formed by an inheritance
* tree, and so the parent rel's physical size and index information isn't
* important for it.
*/
void
get_relation_info(PlannerInfo *root, Oid relationObjectId, bool inhparent,
RelOptInfo *rel)
{
Index varno = rel->relid;
Relation relation;
bool hasindex;
List *indexinfos = NIL;
/*
* We need not lock the relation since it was already locked, either by
* the rewriter or when expand_inherited_rtentry() added it to the query's
* rangetable.
*/
relation = heap_open(relationObjectId, NoLock);
/* Temporary and unlogged relations are inaccessible during recovery. */
if (!RelationNeedsWAL(relation) && RecoveryInProgress())
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
errmsg("cannot access temporary or unlogged relations during recovery")));
rel->min_attr = FirstLowInvalidHeapAttributeNumber + 1;
rel->max_attr = RelationGetNumberOfAttributes(relation);
rel->reltablespace = RelationGetForm(relation)->reltablespace;
Assert(rel->max_attr >= rel->min_attr);
rel->attr_needed = (Relids *)
palloc0((rel->max_attr - rel->min_attr + 1) * sizeof(Relids));
rel->attr_widths = (int32 *)
palloc0((rel->max_attr - rel->min_attr + 1) * sizeof(int32));
/*
* Estimate relation size --- unless it's an inheritance parent, in which
* case the size will be computed later in set_append_rel_pathlist, and we
* must leave it zero for now to avoid bollixing the total_table_pages
* calculation.
*/
if (!inhparent)
estimate_rel_size(relation, rel->attr_widths - rel->min_attr,
&rel->pages, &rel->tuples, &rel->allvisfrac);
/* Retrive the parallel_degree reloption, if set. */
rel->rel_parallel_degree = RelationGetParallelDegree(relation, -1);
/*
* Make list of indexes. Ignore indexes on system catalogs if told to.
* Don't bother with indexes for an inheritance parent, either.
*/
if (inhparent ||
(IgnoreSystemIndexes && IsSystemRelation(relation)))
hasindex = false;
else
hasindex = relation->rd_rel->relhasindex;
if (hasindex)
{
List *indexoidlist;
ListCell *l;
LOCKMODE lmode;
indexoidlist = RelationGetIndexList(relation);
/*
* For each index, we get the same type of lock that the executor will
* need, and do not release it. This saves a couple of trips to the
* shared lock manager while not creating any real loss of
* concurrency, because no schema changes could be happening on the
* index while we hold lock on the parent rel, and neither lock type
* blocks any other kind of index operation.
*/
if (rel->relid == root->parse->resultRelation)
lmode = RowExclusiveLock;
else
lmode = AccessShareLock;
foreach(l, indexoidlist)
{
Oid indexoid = lfirst_oid(l);
Relation indexRelation;
Form_pg_index index;
IndexAmRoutine *amroutine;
IndexOptInfo *info;
int ncolumns;
int i;
/*
* Extract info from the relation descriptor for the index.
*/
indexRelation = index_open(indexoid, lmode);
index = indexRelation->rd_index;
/*
* Ignore invalid indexes, since they can't safely be used for
* queries. Note that this is OK because the data structure we
* are constructing is only used by the planner --- the executor
* still needs to insert into "invalid" indexes, if they're marked
* IndexIsReady.
*/
if (!IndexIsValid(index))
{
index_close(indexRelation, NoLock);
continue;
}
/*
* If the index is valid, but cannot yet be used, ignore it; but
* mark the plan we are generating as transient. See
* src/backend/access/heap/README.HOT for discussion.
*/
if (index->indcheckxmin &&
!TransactionIdPrecedes(HeapTupleHeaderGetXmin(indexRelation->rd_indextuple->t_data),
TransactionXmin))
{
root->glob->transientPlan = true;
index_close(indexRelation, NoLock);
continue;
}
info = makeNode(IndexOptInfo);
info->indexoid = index->indexrelid;
info->reltablespace =
RelationGetForm(indexRelation)->reltablespace;
info->rel = rel;
info->ncolumns = ncolumns = index->indnatts;
info->indexkeys = (int *) palloc(sizeof(int) * ncolumns);
info->indexcollations = (Oid *) palloc(sizeof(Oid) * ncolumns);
info->opfamily = (Oid *) palloc(sizeof(Oid) * ncolumns);
info->opcintype = (Oid *) palloc(sizeof(Oid) * ncolumns);
info->canreturn = (bool *) palloc(sizeof(bool) * ncolumns);
for (i = 0; i < ncolumns; i++)
{
info->indexkeys[i] = index->indkey.values[i];
info->indexcollations[i] = indexRelation->rd_indcollation[i];
info->opfamily[i] = indexRelation->rd_opfamily[i];
info->opcintype[i] = indexRelation->rd_opcintype[i];
info->canreturn[i] = index_can_return(indexRelation, i + 1);
}
info->relam = indexRelation->rd_rel->relam;
/* We copy just the fields we need, not all of rd_amroutine */
amroutine = indexRelation->rd_amroutine;
info->amcanorderbyop = amroutine->amcanorderbyop;
info->amoptionalkey = amroutine->amoptionalkey;
info->amsearcharray = amroutine->amsearcharray;
info->amsearchnulls = amroutine->amsearchnulls;
info->amhasgettuple = (amroutine->amgettuple != NULL);
info->amhasgetbitmap = (amroutine->amgetbitmap != NULL);
info->amcostestimate = amroutine->amcostestimate;
Assert(info->amcostestimate != NULL);
/*
* Fetch the ordering information for the index, if any.
*/
if (info->relam == BTREE_AM_OID)
{
/*
* If it's a btree index, we can use its opfamily OIDs
* directly as the sort ordering opfamily OIDs.
*/
Assert(amroutine->amcanorder);
info->sortopfamily = info->opfamily;
info->reverse_sort = (bool *) palloc(sizeof(bool) * ncolumns);
info->nulls_first = (bool *) palloc(sizeof(bool) * ncolumns);
for (i = 0; i < ncolumns; i++)
{
int16 opt = indexRelation->rd_indoption[i];
info->reverse_sort[i] = (opt & INDOPTION_DESC) != 0;
info->nulls_first[i] = (opt & INDOPTION_NULLS_FIRST) != 0;
}
}
else if (amroutine->amcanorder)
{
/*
* Otherwise, identify the corresponding btree opfamilies by
* trying to map this index's "<" operators into btree. Since
* "<" uniquely defines the behavior of a sort order, this is
* a sufficient test.
*
* XXX This method is rather slow and also requires the
* undesirable assumption that the other index AM numbers its
* strategies the same as btree. It'd be better to have a way
* to explicitly declare the corresponding btree opfamily for
* each opfamily of the other index type. But given the lack
* of current or foreseeable amcanorder index types, it's not
* worth expending more effort on now.
*/
info->sortopfamily = (Oid *) palloc(sizeof(Oid) * ncolumns);
info->reverse_sort = (bool *) palloc(sizeof(bool) * ncolumns);
info->nulls_first = (bool *) palloc(sizeof(bool) * ncolumns);
for (i = 0; i < ncolumns; i++)
{
int16 opt = indexRelation->rd_indoption[i];
Oid ltopr;
Oid btopfamily;
Oid btopcintype;
int16 btstrategy;
info->reverse_sort[i] = (opt & INDOPTION_DESC) != 0;
info->nulls_first[i] = (opt & INDOPTION_NULLS_FIRST) != 0;
ltopr = get_opfamily_member(info->opfamily[i],
info->opcintype[i],
info->opcintype[i],
BTLessStrategyNumber);
if (OidIsValid(ltopr) &&
get_ordering_op_properties(ltopr,
&btopfamily,
&btopcintype,
&btstrategy) &&
btopcintype == info->opcintype[i] &&
btstrategy == BTLessStrategyNumber)
{
/* Successful mapping */
info->sortopfamily[i] = btopfamily;
}
else
{
/* Fail ... quietly treat index as unordered */
info->sortopfamily = NULL;
info->reverse_sort = NULL;
info->nulls_first = NULL;
break;
}
}
}
else
{
info->sortopfamily = NULL;
info->reverse_sort = NULL;
info->nulls_first = NULL;
}
/*
* Fetch the index expressions and predicate, if any. We must
* modify the copies we obtain from the relcache to have the
* correct varno for the parent relation, so that they match up
* correctly against qual clauses.
*/
info->indexprs = RelationGetIndexExpressions(indexRelation);
info->indpred = RelationGetIndexPredicate(indexRelation);
if (info->indexprs && varno != 1)
ChangeVarNodes((Node *) info->indexprs, 1, varno, 0);
if (info->indpred && varno != 1)
ChangeVarNodes((Node *) info->indpred, 1, varno, 0);
/* Build targetlist using the completed indexprs data */
info->indextlist = build_index_tlist(root, info, relation);
info->indrestrictinfo = NIL; /* set later, in indxpath.c */
info->predOK = false; /* set later, in indxpath.c */
info->unique = index->indisunique;
info->immediate = index->indimmediate;
info->hypothetical = false;
/*
* Estimate the index size. If it's not a partial index, we lock
* the number-of-tuples estimate to equal the parent table; if it
* is partial then we have to use the same methods as we would for
* a table, except we can be sure that the index is not larger
* than the table.
*/
if (info->indpred == NIL)
{
info->pages = RelationGetNumberOfBlocks(indexRelation);
info->tuples = rel->tuples;
}
else
{
double allvisfrac; /* dummy */
estimate_rel_size(indexRelation, NULL,
&info->pages, &info->tuples, &allvisfrac);
if (info->tuples > rel->tuples)
info->tuples = rel->tuples;
}
if (info->relam == BTREE_AM_OID)
{
/* For btrees, get tree height while we have the index open */
info->tree_height = _bt_getrootheight(indexRelation);
}
else
{
/* For other index types, just set it to "unknown" for now */
info->tree_height = -1;
}
index_close(indexRelation, NoLock);
indexinfos = lcons(info, indexinfos);
}
list_free(indexoidlist);
}
/*
* Insert an index tuple into the index relation. The revmap is updated to
* mark the range containing the given page as pointing to the inserted entry.
* A WAL record is written.
*
* The buffer, if valid, is first checked for free space to insert the new
* entry; if there isn't enough, a new buffer is obtained and pinned. No
* buffer lock must be held on entry, no buffer lock is held on exit.
*
* Return value is the offset number where the tuple was inserted.
*/
OffsetNumber
brin_doinsert(Relation idxrel, BlockNumber pagesPerRange,
BrinRevmap *revmap, Buffer *buffer, BlockNumber heapBlk,
BrinTuple *tup, Size itemsz)
{
Page page;
BlockNumber blk;
OffsetNumber off;
Buffer revmapbuf;
ItemPointerData tid;
bool extended = false;
itemsz = MAXALIGN(itemsz);
/* Make sure the revmap is long enough to contain the entry we need */
brinRevmapExtend(revmap, heapBlk);
/*
* Obtain a locked buffer to insert the new tuple. Note
* brin_getinsertbuffer ensures there's enough space in the returned
* buffer.
*/
if (BufferIsValid(*buffer))
{
/*
* It's possible that another backend (or ourselves!) extended the
* revmap over the page we held a pin on, so we cannot assume that
* it's still a regular page.
*/
LockBuffer(*buffer, BUFFER_LOCK_EXCLUSIVE);
if (br_page_get_freespace(BufferGetPage(*buffer)) < itemsz)
{
UnlockReleaseBuffer(*buffer);
*buffer = InvalidBuffer;
}
}
if (!BufferIsValid(*buffer))
{
*buffer = brin_getinsertbuffer(idxrel, InvalidBuffer, itemsz, &extended);
Assert(BufferIsValid(*buffer));
Assert(br_page_get_freespace(BufferGetPage(*buffer)) >= itemsz);
}
/* Now obtain lock on revmap buffer */
revmapbuf = brinLockRevmapPageForUpdate(revmap, heapBlk);
page = BufferGetPage(*buffer);
blk = BufferGetBlockNumber(*buffer);
START_CRIT_SECTION();
off = PageAddItem(page, (Item) tup, itemsz, InvalidOffsetNumber,
false, false);
if (off == InvalidOffsetNumber)
elog(ERROR, "could not insert new index tuple to page");
MarkBufferDirty(*buffer);
BRIN_elog((DEBUG2, "inserted tuple (%u,%u) for range starting at %u",
blk, off, heapBlk));
ItemPointerSet(&tid, blk, off);
brinSetHeapBlockItemptr(revmapbuf, pagesPerRange, heapBlk, tid);
MarkBufferDirty(revmapbuf);
/* XLOG stuff */
if (RelationNeedsWAL(idxrel))
{
xl_brin_insert xlrec;
XLogRecPtr recptr;
uint8 info;
info = XLOG_BRIN_INSERT | (extended ? XLOG_BRIN_INIT_PAGE : 0);
xlrec.heapBlk = heapBlk;
xlrec.pagesPerRange = pagesPerRange;
xlrec.offnum = off;
XLogBeginInsert();
XLogRegisterData((char *) &xlrec, SizeOfBrinInsert);
XLogRegisterBuffer(0, *buffer, REGBUF_STANDARD | (extended ? REGBUF_WILL_INIT : 0));
XLogRegisterBufData(0, (char *) tup, itemsz);
XLogRegisterBuffer(1, revmapbuf, 0);
recptr = XLogInsert(RM_BRIN_ID, info);
PageSetLSN(page, recptr);
PageSetLSN(BufferGetPage(revmapbuf), recptr);
}
END_CRIT_SECTION();
/* Tuple is firmly on buffer; we can release our locks */
LockBuffer(*buffer, BUFFER_LOCK_UNLOCK);
LockBuffer(revmapbuf, BUFFER_LOCK_UNLOCK);
if (extended)
FreeSpaceMapVacuum(idxrel);
return off;
}
/*
* Creates new posting tree containing the given TIDs. Returns the page
* number of the root of the new posting tree.
*
* items[] must be in sorted order with no duplicates.
*/
BlockNumber
createPostingTree(Relation index, ItemPointerData *items, uint32 nitems,
GinStatsData *buildStats)
{
BlockNumber blkno;
Buffer buffer;
Page page;
int nrootitems;
/* Calculate how many TIDs will fit on first page. */
nrootitems = Min(nitems, GinMaxLeafDataItems);
/*
* Create the root page.
*/
buffer = GinNewBuffer(index);
page = BufferGetPage(buffer);
blkno = BufferGetBlockNumber(buffer);
START_CRIT_SECTION();
GinInitBuffer(buffer, GIN_DATA | GIN_LEAF);
memcpy(GinDataPageGetData(page), items, sizeof(ItemPointerData) * nrootitems);
GinPageGetOpaque(page)->maxoff = nrootitems;
MarkBufferDirty(buffer);
if (RelationNeedsWAL(index))
{
XLogRecPtr recptr;
XLogRecData rdata[2];
ginxlogCreatePostingTree data;
data.node = index->rd_node;
data.blkno = blkno;
data.nitem = nrootitems;
rdata[0].buffer = InvalidBuffer;
rdata[0].data = (char *) &data;
rdata[0].len = sizeof(ginxlogCreatePostingTree);
rdata[0].next = &rdata[1];
rdata[1].buffer = InvalidBuffer;
rdata[1].data = (char *) items;
rdata[1].len = sizeof(ItemPointerData) * nrootitems;
rdata[1].next = NULL;
recptr = XLogInsert(RM_GIN_ID, XLOG_GIN_CREATE_PTREE, rdata);
PageSetLSN(page, recptr);
}
UnlockReleaseBuffer(buffer);
END_CRIT_SECTION();
/* During index build, count the newly-added data page */
if (buildStats)
buildStats->nDataPages++;
/*
* Add any remaining TIDs to the newly-created posting tree.
*/
if (nitems > nrootitems)
{
ginInsertItemPointers(index, blkno,
items + nrootitems,
nitems - nrootitems,
buildStats);
}
return blkno;
}
/*
* _bt_getroot() -- Get the root page of the btree.
*
* Since the root page can move around the btree file, we have to read
* its location from the metadata page, and then read the root page
* itself. If no root page exists yet, we have to create one. The
* standard class of race conditions exists here; I think I covered
* them all in the Hopi Indian rain dance of lock requests below.
*
* The access type parameter (BT_READ or BT_WRITE) controls whether
* a new root page will be created or not. If access = BT_READ,
* and no root page exists, we just return InvalidBuffer. For
* BT_WRITE, we try to create the root page if it doesn't exist.
* NOTE that the returned root page will have only a read lock set
* on it even if access = BT_WRITE!
*
* The returned page is not necessarily the true root --- it could be
* a "fast root" (a page that is alone in its level due to deletions).
* Also, if the root page is split while we are "in flight" to it,
* what we will return is the old root, which is now just the leftmost
* page on a probably-not-very-wide level. For most purposes this is
* as good as or better than the true root, so we do not bother to
* insist on finding the true root. We do, however, guarantee to
* return a live (not deleted or half-dead) page.
*
* On successful return, the root page is pinned and read-locked.
* The metadata page is not locked or pinned on exit.
*/
Buffer
_bt_getroot(Relation rel, int access)
{
Buffer metabuf;
Page metapg;
BTPageOpaque metaopaque;
Buffer rootbuf;
Page rootpage;
BTPageOpaque rootopaque;
BlockNumber rootblkno;
uint32 rootlevel;
BTMetaPageData *metad;
/*
* Try to use previously-cached metapage data to find the root. This
* normally saves one buffer access per index search, which is a very
* helpful savings in bufmgr traffic and hence contention.
*/
if (rel->rd_amcache != NULL)
{
metad = (BTMetaPageData *) rel->rd_amcache;
/* We shouldn't have cached it if any of these fail */
Assert(metad->btm_magic == BTREE_MAGIC);
Assert(metad->btm_version == BTREE_VERSION);
Assert(metad->btm_root != P_NONE);
rootblkno = metad->btm_fastroot;
Assert(rootblkno != P_NONE);
rootlevel = metad->btm_fastlevel;
rootbuf = _bt_getbuf(rel, rootblkno, BT_READ);
rootpage = BufferGetPage(rootbuf);
rootopaque = (BTPageOpaque) PageGetSpecialPointer(rootpage);
/*
* Since the cache might be stale, we check the page more carefully
* here than normal. We *must* check that it's not deleted. If it's
* not alone on its level, then we reject too --- this may be overly
* paranoid but better safe than sorry. Note we don't check P_ISROOT,
* because that's not set in a "fast root".
*/
if (!P_IGNORE(rootopaque) &&
rootopaque->btpo.level == rootlevel &&
P_LEFTMOST(rootopaque) &&
P_RIGHTMOST(rootopaque))
{
/* OK, accept cached page as the root */
return rootbuf;
}
_bt_relbuf(rel, rootbuf);
/* Cache is stale, throw it away */
if (rel->rd_amcache)
pfree(rel->rd_amcache);
rel->rd_amcache = NULL;
}
metabuf = _bt_getbuf(rel, BTREE_METAPAGE, BT_READ);
metapg = BufferGetPage(metabuf);
metaopaque = (BTPageOpaque) PageGetSpecialPointer(metapg);
metad = BTPageGetMeta(metapg);
/* sanity-check the metapage */
if (!(metaopaque->btpo_flags & BTP_META) ||
metad->btm_magic != BTREE_MAGIC)
ereport(ERROR,
(errcode(ERRCODE_INDEX_CORRUPTED),
errmsg("index \"%s\" is not a btree",
RelationGetRelationName(rel))));
if (metad->btm_version != BTREE_VERSION)
ereport(ERROR,
(errcode(ERRCODE_INDEX_CORRUPTED),
errmsg("version mismatch in index \"%s\": file version %d, code version %d",
RelationGetRelationName(rel),
metad->btm_version, BTREE_VERSION)));
/* if no root page initialized yet, do it */
if (metad->btm_root == P_NONE)
{
/* If access = BT_READ, caller doesn't want us to create root yet */
if (access == BT_READ)
{
_bt_relbuf(rel, metabuf);
return InvalidBuffer;
}
/* trade in our read lock for a write lock */
LockBuffer(metabuf, BUFFER_LOCK_UNLOCK);
LockBuffer(metabuf, BT_WRITE);
/*
* Race condition: if someone else initialized the metadata between
* the time we released the read lock and acquired the write lock, we
* must avoid doing it again.
*/
if (metad->btm_root != P_NONE)
{
/*
* Metadata initialized by someone else. In order to guarantee no
* deadlocks, we have to release the metadata page and start all
* over again. (Is that really true? But it's hardly worth trying
* to optimize this case.)
*/
_bt_relbuf(rel, metabuf);
return _bt_getroot(rel, access);
}
/*
* Get, initialize, write, and leave a lock of the appropriate type on
* the new root page. Since this is the first page in the tree, it's
* a leaf as well as the root.
*/
rootbuf = _bt_getbuf(rel, P_NEW, BT_WRITE);
rootblkno = BufferGetBlockNumber(rootbuf);
rootpage = BufferGetPage(rootbuf);
rootopaque = (BTPageOpaque) PageGetSpecialPointer(rootpage);
rootopaque->btpo_prev = rootopaque->btpo_next = P_NONE;
rootopaque->btpo_flags = (BTP_LEAF | BTP_ROOT);
rootopaque->btpo.level = 0;
rootopaque->btpo_cycleid = 0;
/* NO ELOG(ERROR) till meta is updated */
START_CRIT_SECTION();
metad->btm_root = rootblkno;
metad->btm_level = 0;
metad->btm_fastroot = rootblkno;
metad->btm_fastlevel = 0;
MarkBufferDirty(rootbuf);
MarkBufferDirty(metabuf);
/* XLOG stuff */
if (RelationNeedsWAL(rel))
{
xl_btree_newroot xlrec;
XLogRecPtr recptr;
XLogRecData rdata;
xlrec.node = rel->rd_node;
xlrec.rootblk = rootblkno;
xlrec.level = 0;
rdata.data = (char *) &xlrec;
rdata.len = SizeOfBtreeNewroot;
rdata.buffer = InvalidBuffer;
rdata.next = NULL;
recptr = XLogInsert(RM_BTREE_ID, XLOG_BTREE_NEWROOT, &rdata);
PageSetLSN(rootpage, recptr);
PageSetTLI(rootpage, ThisTimeLineID);
PageSetLSN(metapg, recptr);
PageSetTLI(metapg, ThisTimeLineID);
}
END_CRIT_SECTION();
/*
* Send out relcache inval for metapage change (probably unnecessary
* here, but let's be safe).
*/
CacheInvalidateRelcache(rel);
/*
* swap root write lock for read lock. There is no danger of anyone
* else accessing the new root page while it's unlocked, since no one
* else knows where it is yet.
*/
LockBuffer(rootbuf, BUFFER_LOCK_UNLOCK);
LockBuffer(rootbuf, BT_READ);
/* okay, metadata is correct, release lock on it */
_bt_relbuf(rel, metabuf);
}
else
{
rootblkno = metad->btm_fastroot;
Assert(rootblkno != P_NONE);
rootlevel = metad->btm_fastlevel;
/*
* Cache the metapage data for next time
*/
rel->rd_amcache = MemoryContextAlloc(rel->rd_indexcxt,
sizeof(BTMetaPageData));
memcpy(rel->rd_amcache, metad, sizeof(BTMetaPageData));
/*
* We are done with the metapage; arrange to release it via first
* _bt_relandgetbuf call
*/
rootbuf = metabuf;
for (;;)
{
rootbuf = _bt_relandgetbuf(rel, rootbuf, rootblkno, BT_READ);
rootpage = BufferGetPage(rootbuf);
rootopaque = (BTPageOpaque) PageGetSpecialPointer(rootpage);
if (!P_IGNORE(rootopaque))
break;
/* it's dead, Jim. step right one page */
if (P_RIGHTMOST(rootopaque))
elog(ERROR, "no live root page found in index \"%s\"",
RelationGetRelationName(rel));
rootblkno = rootopaque->btpo_next;
}
/* Note: can't check btpo.level on deleted pages */
if (rootopaque->btpo.level != rootlevel)
elog(ERROR, "root page %u of index \"%s\" has level %u, expected %u",
rootblkno, RelationGetRelationName(rel),
rootopaque->btpo.level, rootlevel);
}
/*
* By here, we have a pin and read lock on the root page, and no lock set
* on the metadata page. Return the root page's buffer.
*/
return rootbuf;
}