/*
* Return the offset of the next matching index entry. We begin the
* search at offset "n" and search for matches in the direction
* "dir". If no more matching entries are found on the page,
* InvalidOffsetNumber is returned.
*/
static OffsetNumber
findnext(IndexScanDesc s, OffsetNumber n, ScanDirection dir)
{
OffsetNumber maxoff;
IndexTuple it;
RTreePageOpaque po;
RTreeScanOpaque so;
Page p;
so = (RTreeScanOpaque) s->opaque;
p = BufferGetPage(so->curbuf);
maxoff = PageGetMaxOffsetNumber(p);
po = (RTreePageOpaque) PageGetSpecialPointer(p);
/*
* If we modified the index during the scan, we may have a pointer to a
* ghost tuple, before the scan. If this is the case, back up one.
*/
if (so->s_flags & RTS_CURBEFORE)
{
so->s_flags &= ~RTS_CURBEFORE;
n = OffsetNumberPrev(n);
}
while (n >= FirstOffsetNumber && n <= maxoff)
{
it = (IndexTuple) PageGetItem(p, PageGetItemId(p, n));
if (po->flags & F_LEAF)
{
if (index_keytest(it,
RelationGetDescr(s->indexRelation),
s->numberOfKeys, s->keyData))
break;
}
else
{
if (index_keytest(it,
RelationGetDescr(s->indexRelation),
so->s_internalNKey, so->s_internalKey))
break;
}
if (ScanDirectionIsBackward(dir))
n = OffsetNumberPrev(n);
else
n = OffsetNumberNext(n);
}
if (n >= FirstOffsetNumber && n <= maxoff)
return n; /* found a match on this page */
else
return InvalidOffsetNumber; /* no match, go to next page */
}
/*
* Return the offset of the first index entry that is consistent with
* the search key after offset 'n' in the current page. If there are
* no more consistent entries, return InvalidOffsetNumber.
* Page should be locked....
*/
static OffsetNumber
gistfindnext(IndexScanDesc scan, OffsetNumber n, ScanDirection dir)
{
OffsetNumber maxoff;
IndexTuple it;
GISTScanOpaque so;
MemoryContext oldcxt;
Page p;
so = (GISTScanOpaque) scan->opaque;
p = BufferGetPage(so->curbuf);
maxoff = PageGetMaxOffsetNumber(p);
/*
* Make sure we're in a short-lived memory context when we invoke a
* user-supplied GiST method in gistindex_keytest(), so we don't leak
* memory
*/
oldcxt = MemoryContextSwitchTo(so->tempCxt);
/*
* If we modified the index during the scan, we may have a pointer to a
* ghost tuple, before the scan. If this is the case, back up one.
*/
if (so->flags & GS_CURBEFORE)
{
so->flags &= ~GS_CURBEFORE;
n = OffsetNumberPrev(n);
}
while (n >= FirstOffsetNumber && n <= maxoff)
{
it = (IndexTuple) PageGetItem(p, PageGetItemId(p, n));
if (gistindex_keytest(it, scan, n))
break;
if (ScanDirectionIsBackward(dir))
n = OffsetNumberPrev(n);
else
n = OffsetNumberNext(n);
}
MemoryContextSwitchTo(oldcxt);
MemoryContextReset(so->tempCxt);
/*
* If we found a matching entry, return its offset; otherwise return
* InvalidOffsetNumber to inform the caller to go to the next page.
*/
if (n >= FirstOffsetNumber && n <= maxoff)
return n;
else
return InvalidOffsetNumber;
}
/*
* _bt_endpoint() -- Find the first or last page in the index, and scan
* from there to the first key satisfying all the quals.
*
* This is used by _bt_first() to set up a scan when we've determined
* that the scan must start at the beginning or end of the index (for
* a forward or backward scan respectively). Exit conditions are the
* same as for _bt_first().
*/
static bool
_bt_endpoint(IndexScanDesc scan, ScanDirection dir)
{
Relation rel = scan->indexRelation;
BTScanOpaque so = (BTScanOpaque) scan->opaque;
Buffer buf;
Page page;
BTPageOpaque opaque;
OffsetNumber start;
BTScanPosItem *currItem;
/*
* Scan down to the leftmost or rightmost leaf page. This is a simplified
* version of _bt_search(). We don't maintain a stack since we know we
* won't need it.
*/
buf = _bt_get_endpoint(rel, 0, ScanDirectionIsBackward(dir));
if (!BufferIsValid(buf))
{
/*
* Empty index. Lock the whole relation, as nothing finer to lock
* exists.
*/
PredicateLockRelation(rel, scan->xs_snapshot);
so->currPos.buf = InvalidBuffer;
return false;
}
PredicateLockPage(rel, BufferGetBlockNumber(buf), scan->xs_snapshot);
page = BufferGetPage(buf);
opaque = (BTPageOpaque) PageGetSpecialPointer(page);
Assert(P_ISLEAF(opaque));
if (ScanDirectionIsForward(dir))
{
/* There could be dead pages to the left, so not this: */
/* Assert(P_LEFTMOST(opaque)); */
start = P_FIRSTDATAKEY(opaque);
}
else if (ScanDirectionIsBackward(dir))
{
Assert(P_RIGHTMOST(opaque));
start = PageGetMaxOffsetNumber(page);
}
else
{
elog(ERROR, "invalid scan direction: %d", (int) dir);
start = 0; /* keep compiler quiet */
}
/* remember which buffer we have pinned */
so->currPos.buf = buf;
/* initialize moreLeft/moreRight appropriately for scan direction */
if (ScanDirectionIsForward(dir))
{
so->currPos.moreLeft = false;
so->currPos.moreRight = true;
}
else
{
so->currPos.moreLeft = true;
so->currPos.moreRight = false;
}
so->numKilled = 0; /* just paranoia */
so->markItemIndex = -1; /* ditto */
/*
* Now load data from the first page of the scan.
*/
if (!_bt_readpage(scan, dir, start))
{
/*
* There's no actually-matching data on this page. Try to advance to
* the next page. Return false if there's no matching data at all.
*/
if (!_bt_steppage(scan, dir))
return false;
}
/* Drop the lock, but not pin, on the current page */
LockBuffer(so->currPos.buf, BUFFER_LOCK_UNLOCK);
/* OK, itemIndex says what to return */
currItem = &so->currPos.items[so->currPos.itemIndex];
scan->xs_ctup.t_self = currItem->heapTid;
if (scan->xs_want_itup)
scan->xs_itup = (IndexTuple) (so->currTuples + currItem->tupleOffset);
return true;
}
/*
* hidden_getnext
*
* Returns the next tuple from the hidden data. This has to be called after
* exceeding heap/index scan.
*/
HeapTuple
hidden_getnext(HiddenScanDesc hscan, ScanDirection direction)
{
if (ScanDirectionIsForward(direction))
{
if (hscan->hdn_idx < 0)
hscan->hdn_idx = 0;
while (hscan->hdn_idx < hscan->hdn_len)
{
HeapTuple tuple;
bool valid = true;
/* fetch */
tuple = hscan->hdn_tuples[hscan->hdn_idx++];
/* key test */
if (hscan->hdn_nkeys > 0)
HeapKeyTest(tuple, RelationGetDescr(hscan->hdn_rel),
hscan->hdn_nkeys, hscan->hdn_key, valid);
if (valid)
{
/* save the last tuple for the sake of "no movement" */
hscan->hdn_lasttuple = tuple;
return tuple;
}
}
}
else if (ScanDirectionIsBackward(direction))
{
if (hscan->hdn_idx >= hscan->hdn_len)
hscan->hdn_idx = hscan->hdn_len - 1;
while (hscan->hdn_idx >= 0)
{
HeapTuple tuple;
bool valid = true;
/* fetch */
tuple = hscan->hdn_tuples[hscan->hdn_idx--];
/* key test */
if (hscan->hdn_nkeys > 0)
HeapKeyTest(tuple, RelationGetDescr(hscan->hdn_rel),
hscan->hdn_nkeys, hscan->hdn_key, valid);
if (valid)
{
/* save the last tuple for the sake of "no movement" */
hscan->hdn_lasttuple = tuple;
return tuple;
}
}
}
else
{
/*
* ``no movement'' scan direction: refetch prior tuple
*/
return hscan->hdn_lasttuple;
}
hscan->hdn_lasttuple = NULL;
return NULL;
}
/*
* _bt_first() -- Find the first item in a scan.
*
* We need to be clever about the type of scan, the operation it's
* performing, and the tree ordering. We find the
* first item in the tree that satisfies the qualification
* associated with the scan descriptor. On exit, the page containing
* the current index tuple is read locked and pinned, and the scan's
* opaque data entry is updated to include the buffer.
*/
bool
_bt_first(IndexScanDesc scan, ScanDirection dir)
{
Relation rel = scan->indexRelation;
BTScanOpaque so = (BTScanOpaque) scan->opaque;
Buffer buf;
Page page;
BTStack stack;
OffsetNumber offnum;
BTItem btitem;
IndexTuple itup;
ItemPointer current;
BlockNumber blkno;
StrategyNumber strat;
bool res;
int32 result;
bool scanFromEnd;
bool continuescan;
ScanKey scankeys = NULL;
int keysCount = 0;
int *nKeyIs = NULL;
int i,
j;
StrategyNumber strat_total;
/*
* Order the scan keys in our canonical fashion and eliminate any
* redundant keys.
*/
_bt_orderkeys(scan);
/*
* Quit now if _bt_orderkeys() discovered that the scan keys can never
* be satisfied (eg, x == 1 AND x > 2).
*/
if (!so->qual_ok)
return false;
/*
* Examine the scan keys to discover where we need to start the scan.
*/
scanFromEnd = false;
strat_total = BTEqualStrategyNumber;
if (so->numberOfKeys > 0)
{
nKeyIs = (int *) palloc(so->numberOfKeys * sizeof(int));
for (i = 0; i < so->numberOfKeys; i++)
{
AttrNumber attno = so->keyData[i].sk_attno;
/* ignore keys for already-determined attrs */
if (attno <= keysCount)
continue;
/* if we didn't find a boundary for the preceding attr, quit */
if (attno > keysCount + 1)
break;
strat = _bt_getstrat(rel, attno,
so->keyData[i].sk_procedure);
/*
* Can we use this key as a starting boundary for this attr?
*
* We can use multiple keys if they look like, say, = >= = but we
* have to stop after accepting a > or < boundary.
*/
if (strat == strat_total ||
strat == BTEqualStrategyNumber)
nKeyIs[keysCount++] = i;
else if (ScanDirectionIsBackward(dir) &&
(strat == BTLessStrategyNumber ||
strat == BTLessEqualStrategyNumber))
{
nKeyIs[keysCount++] = i;
strat_total = strat;
if (strat == BTLessStrategyNumber)
break;
}
else if (ScanDirectionIsForward(dir) &&
(strat == BTGreaterStrategyNumber ||
strat == BTGreaterEqualStrategyNumber))
{
nKeyIs[keysCount++] = i;
strat_total = strat;
if (strat == BTGreaterStrategyNumber)
break;
}
}
if (keysCount == 0)
scanFromEnd = true;
}
else
scanFromEnd = true;
/* if we just need to walk down one edge of the tree, do that */
if (scanFromEnd)
{
if (nKeyIs)
pfree(nKeyIs);
return _bt_endpoint(scan, dir);
}
/*
* We want to start the scan somewhere within the index. Set up a
* scankey we can use to search for the correct starting point.
*/
scankeys = (ScanKey) palloc(keysCount * sizeof(ScanKeyData));
for (i = 0; i < keysCount; i++)
{
FmgrInfo *procinfo;
j = nKeyIs[i];
/*
* _bt_orderkeys disallows it, but it's place to add some code
* later
*/
if (so->keyData[j].sk_flags & SK_ISNULL)
{
pfree(nKeyIs);
pfree(scankeys);
elog(ERROR, "btree doesn't support is(not)null, yet");
return false;
}
procinfo = index_getprocinfo(rel, i + 1, BTORDER_PROC);
ScanKeyEntryInitializeWithInfo(scankeys + i,
so->keyData[j].sk_flags,
i + 1,
procinfo,
CurrentMemoryContext,
so->keyData[j].sk_argument);
}
if (nKeyIs)
pfree(nKeyIs);
current = &(scan->currentItemData);
/*
* Use the manufactured scan key to descend the tree and position
* ourselves on the target leaf page.
*/
stack = _bt_search(rel, keysCount, scankeys, &buf, BT_READ);
/* don't need to keep the stack around... */
_bt_freestack(stack);
if (!BufferIsValid(buf))
{
/* Only get here if index is completely empty */
ItemPointerSetInvalid(current);
so->btso_curbuf = InvalidBuffer;
pfree(scankeys);
return false;
}
/* remember which buffer we have pinned */
so->btso_curbuf = buf;
blkno = BufferGetBlockNumber(buf);
page = BufferGetPage(buf);
/* position to the precise item on the page */
offnum = _bt_binsrch(rel, buf, keysCount, scankeys);
ItemPointerSet(current, blkno, offnum);
/*
* At this point we are positioned at the first item >= scan key, or
* possibly at the end of a page on which all the existing items are
* less than the scan key and we know that everything on later pages
* is greater than or equal to scan key.
*
* We could step forward in the latter case, but that'd be a waste of
* time if we want to scan backwards. So, it's now time to examine
* the scan strategy to find the exact place to start the scan.
*
* Note: if _bt_step fails (meaning we fell off the end of the index in
* one direction or the other), we either return false (no matches) or
* call _bt_endpoint() to set up a scan starting at that index
* endpoint, as appropriate for the desired scan type.
*
* it's yet other place to add some code later for is(not)null ...
*/
switch (strat_total)
{
case BTLessStrategyNumber:
/*
* Back up one to arrive at last item < scankey
*/
if (!_bt_step(scan, &buf, BackwardScanDirection))
{
pfree(scankeys);
return false;
}
break;
case BTLessEqualStrategyNumber:
/*
* We need to find the last item <= scankey, so step forward
* till we find one > scankey, then step back one.
*/
if (offnum > PageGetMaxOffsetNumber(page))
{
if (!_bt_step(scan, &buf, ForwardScanDirection))
{
pfree(scankeys);
return _bt_endpoint(scan, dir);
}
}
for (;;)
{
offnum = ItemPointerGetOffsetNumber(current);
page = BufferGetPage(buf);
result = _bt_compare(rel, keysCount, scankeys, page, offnum);
if (result < 0)
break;
if (!_bt_step(scan, &buf, ForwardScanDirection))
{
pfree(scankeys);
return _bt_endpoint(scan, dir);
}
}
if (!_bt_step(scan, &buf, BackwardScanDirection))
{
pfree(scankeys);
return false;
}
break;
case BTEqualStrategyNumber:
/*
* Make sure we are on the first equal item; might have to
* step forward if currently at end of page.
*/
if (offnum > PageGetMaxOffsetNumber(page))
{
if (!_bt_step(scan, &buf, ForwardScanDirection))
{
pfree(scankeys);
return false;
}
offnum = ItemPointerGetOffsetNumber(current);
page = BufferGetPage(buf);
}
result = _bt_compare(rel, keysCount, scankeys, page, offnum);
if (result != 0)
goto nomatches; /* no equal items! */
/*
* If a backward scan was specified, need to start with last
* equal item not first one.
*/
if (ScanDirectionIsBackward(dir))
{
do
{
if (!_bt_step(scan, &buf, ForwardScanDirection))
{
pfree(scankeys);
return _bt_endpoint(scan, dir);
}
offnum = ItemPointerGetOffsetNumber(current);
page = BufferGetPage(buf);
result = _bt_compare(rel, keysCount, scankeys, page, offnum);
} while (result == 0);
if (!_bt_step(scan, &buf, BackwardScanDirection))
elog(ERROR, "equal items disappeared?");
}
break;
case BTGreaterEqualStrategyNumber:
/*
* We want the first item >= scankey, which is where we are...
* unless we're not anywhere at all...
*/
if (offnum > PageGetMaxOffsetNumber(page))
{
if (!_bt_step(scan, &buf, ForwardScanDirection))
{
pfree(scankeys);
return false;
}
}
break;
case BTGreaterStrategyNumber:
/*
* We want the first item > scankey, so make sure we are on an
* item and then step over any equal items.
*/
if (offnum > PageGetMaxOffsetNumber(page))
{
if (!_bt_step(scan, &buf, ForwardScanDirection))
{
pfree(scankeys);
return false;
}
offnum = ItemPointerGetOffsetNumber(current);
page = BufferGetPage(buf);
}
result = _bt_compare(rel, keysCount, scankeys, page, offnum);
while (result == 0)
{
if (!_bt_step(scan, &buf, ForwardScanDirection))
{
pfree(scankeys);
return false;
}
offnum = ItemPointerGetOffsetNumber(current);
page = BufferGetPage(buf);
result = _bt_compare(rel, keysCount, scankeys, page, offnum);
}
break;
}
/* okay, current item pointer for the scan is right */
offnum = ItemPointerGetOffsetNumber(current);
page = BufferGetPage(buf);
btitem = (BTItem) PageGetItem(page, PageGetItemId(page, offnum));
itup = &btitem->bti_itup;
/* is the first item actually acceptable? */
if (_bt_checkkeys(scan, itup, dir, &continuescan))
{
/* yes, return it */
scan->xs_ctup.t_self = itup->t_tid;
res = true;
}
else if (continuescan)
{
/* no, but there might be another one that is */
res = _bt_next(scan, dir);
}
else
{
/* no tuples in the index match this scan key */
nomatches:
ItemPointerSetInvalid(current);
so->btso_curbuf = InvalidBuffer;
_bt_relbuf(rel, buf);
res = false;
}
pfree(scankeys);
return res;
}
/*
* _bt_endpoint() -- Find the first or last key in the index.
*
* This is used by _bt_first() to set up a scan when we've determined
* that the scan must start at the beginning or end of the index (for
* a forward or backward scan respectively).
*/
static bool
_bt_endpoint(IndexScanDesc scan, ScanDirection dir)
{
Relation rel;
Buffer buf;
Page page;
BTPageOpaque opaque;
ItemPointer current;
OffsetNumber maxoff;
OffsetNumber start;
BlockNumber blkno;
BTItem btitem;
IndexTuple itup;
BTScanOpaque so;
bool res;
bool continuescan;
rel = scan->indexRelation;
current = &(scan->currentItemData);
so = (BTScanOpaque) scan->opaque;
/*
* Scan down to the leftmost or rightmost leaf page. This is a
* simplified version of _bt_search(). We don't maintain a stack
* since we know we won't need it.
*/
buf = _bt_get_endpoint(rel, 0, ScanDirectionIsBackward(dir));
if (!BufferIsValid(buf))
{
/* empty index... */
ItemPointerSetInvalid(current);
so->btso_curbuf = InvalidBuffer;
return false;
}
blkno = BufferGetBlockNumber(buf);
page = BufferGetPage(buf);
opaque = (BTPageOpaque) PageGetSpecialPointer(page);
Assert(P_ISLEAF(opaque));
maxoff = PageGetMaxOffsetNumber(page);
if (ScanDirectionIsForward(dir))
{
/* There could be dead pages to the left, so not this: */
/* Assert(P_LEFTMOST(opaque)); */
start = P_FIRSTDATAKEY(opaque);
}
else if (ScanDirectionIsBackward(dir))
{
Assert(P_RIGHTMOST(opaque));
start = PageGetMaxOffsetNumber(page);
if (start < P_FIRSTDATAKEY(opaque)) /* watch out for empty
* page */
start = P_FIRSTDATAKEY(opaque);
}
else
{
elog(ERROR, "invalid scan direction: %d", (int) dir);
start = 0; /* keep compiler quiet */
}
ItemPointerSet(current, blkno, start);
/* remember which buffer we have pinned */
so->btso_curbuf = buf;
/*
* Left/rightmost page could be empty due to deletions, if so step
* till we find a nonempty page.
*/
if (start > maxoff)
{
if (!_bt_step(scan, &buf, dir))
return false;
start = ItemPointerGetOffsetNumber(current);
page = BufferGetPage(buf);
}
btitem = (BTItem) PageGetItem(page, PageGetItemId(page, start));
itup = &(btitem->bti_itup);
/* see if we picked a winner */
if (_bt_checkkeys(scan, itup, dir, &continuescan))
{
/* yes, return it */
scan->xs_ctup.t_self = itup->t_tid;
res = true;
}
else if (continuescan)
{
/* no, but there might be another one that is */
res = _bt_next(scan, dir);
}
else
{
/* no tuples in the index match this scan key */
ItemPointerSetInvalid(current);
so->btso_curbuf = InvalidBuffer;
_bt_relbuf(rel, buf);
res = false;
}
return res;
}
/*
* _hash_first() -- Find the first item in a scan.
*
* Find the first item in the index that
* satisfies the qualification associated with the scan descriptor. On
* success, the page containing the current index tuple is read locked
* and pinned, and the scan's opaque data entry is updated to
* include the buffer.
*/
bool
_hash_first(IndexScanDesc scan, ScanDirection dir)
{
Relation rel = scan->indexRelation;
HashScanOpaque so = (HashScanOpaque) scan->opaque;
uint32 hashkey;
Bucket bucket;
BlockNumber blkno;
Buffer buf;
Buffer metabuf;
Page page;
HashPageOpaque opaque;
HashMetaPage metap;
IndexTuple itup;
ItemPointer current;
OffsetNumber offnum;
MIRROREDLOCK_BUFMGR_MUST_ALREADY_BE_HELD;
pgstat_count_index_scan(rel);
current = &(scan->currentItemData);
ItemPointerSetInvalid(current);
/*
* We do not support hash scans with no index qualification, because we
* would have to read the whole index rather than just one bucket. That
* creates a whole raft of problems, since we haven't got a practical way
* to lock all the buckets against splits or compactions.
*/
if (scan->numberOfKeys < 1)
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
errmsg("hash indexes do not support whole-index scans")));
/*
* If the constant in the index qual is NULL, assume it cannot match any
* items in the index.
*/
if (scan->keyData[0].sk_flags & SK_ISNULL)
return false;
/*
* Okay to compute the hash key. We want to do this before acquiring any
* locks, in case a user-defined hash function happens to be slow.
*/
hashkey = _hash_datum2hashkey(rel, scan->keyData[0].sk_argument);
/*
* Acquire shared split lock so we can compute the target bucket safely
* (see README).
*/
_hash_getlock(rel, 0, HASH_SHARE);
/* Read the metapage */
metabuf = _hash_getbuf(rel, HASH_METAPAGE, HASH_READ);
_hash_checkpage(rel, metabuf, LH_META_PAGE);
metap = (HashMetaPage) BufferGetPage(metabuf);
/*
* Compute the target bucket number, and convert to block number.
*/
bucket = _hash_hashkey2bucket(hashkey,
metap->hashm_maxbucket,
metap->hashm_highmask,
metap->hashm_lowmask);
blkno = BUCKET_TO_BLKNO(metap, bucket);
/* done with the metapage */
_hash_relbuf(rel, metabuf);
/*
* Acquire share lock on target bucket; then we can release split lock.
*/
_hash_getlock(rel, blkno, HASH_SHARE);
_hash_droplock(rel, 0, HASH_SHARE);
/* Update scan opaque state to show we have lock on the bucket */
so->hashso_bucket = bucket;
so->hashso_bucket_valid = true;
so->hashso_bucket_blkno = blkno;
/* Fetch the primary bucket page for the bucket */
buf = _hash_getbuf(rel, blkno, HASH_READ);
_hash_checkpage(rel, buf, LH_BUCKET_PAGE);
page = BufferGetPage(buf);
opaque = (HashPageOpaque) PageGetSpecialPointer(page);
Assert(opaque->hasho_bucket == bucket);
/* If a backwards scan is requested, move to the end of the chain */
if (ScanDirectionIsBackward(dir))
{
while (BlockNumberIsValid(opaque->hasho_nextblkno))
_hash_readnext(rel, &buf, &page, &opaque);
}
/* Now find the first tuple satisfying the qualification */
if (!_hash_step(scan, &buf, dir))
return false;
/* if we're here, _hash_step found a valid tuple */
offnum = ItemPointerGetOffsetNumber(current);
_hash_checkpage(rel, buf, LH_BUCKET_PAGE | LH_OVERFLOW_PAGE);
page = BufferGetPage(buf);
itup = (IndexTuple) PageGetItem(page, PageGetItemId(page, offnum));
scan->xs_ctup.t_self = itup->t_tid;
return true;
}
/*
* Test whether an indextuple satisfies a row-comparison scan condition.
*
* Return true if so, false if not. If not, also clear *continuescan if
* it's not possible for any future tuples in the current scan direction
* to pass the qual.
*
* This is a subroutine for _bt_checkkeys, which see for more info.
*/
static bool
_bt_check_rowcompare(ScanKey skey, IndexTuple tuple, TupleDesc tupdesc,
ScanDirection dir, bool *continuescan)
{
ScanKey subkey = (ScanKey) DatumGetPointer(skey->sk_argument);
int32 cmpresult = 0;
bool result;
/* First subkey should be same as the header says */
Assert(subkey->sk_attno == skey->sk_attno);
/* Loop over columns of the row condition */
for (;;)
{
Datum datum;
bool isNull;
Assert(subkey->sk_flags & SK_ROW_MEMBER);
datum = index_getattr(tuple,
subkey->sk_attno,
tupdesc,
&isNull);
if (isNull)
{
if (subkey->sk_flags & SK_BT_NULLS_FIRST)
{
/*
* Since NULLs are sorted before non-NULLs, we know we have
* reached the lower limit of the range of values for this
* index attr. On a backward scan, we can stop if this qual is
* one of the "must match" subset. On a forward scan,
* however, we should keep going.
*/
if ((subkey->sk_flags & SK_BT_REQBKWD) &&
ScanDirectionIsBackward(dir))
*continuescan = false;
}
else
{
/*
* Since NULLs are sorted after non-NULLs, we know we have
* reached the upper limit of the range of values for this
* index attr. On a forward scan, we can stop if this qual is
* one of the "must match" subset. On a backward scan,
* however, we should keep going.
*/
if ((subkey->sk_flags & SK_BT_REQFWD) &&
ScanDirectionIsForward(dir))
*continuescan = false;
}
/*
* In any case, this indextuple doesn't match the qual.
*/
return false;
}
if (subkey->sk_flags & SK_ISNULL)
{
/*
* Unlike the simple-scankey case, this isn't a disallowed case.
* But it can never match. If all the earlier row comparison
* columns are required for the scan direction, we can stop the
* scan, because there can't be another tuple that will succeed.
*/
if (subkey != (ScanKey) DatumGetPointer(skey->sk_argument))
subkey--;
if ((subkey->sk_flags & SK_BT_REQFWD) &&
ScanDirectionIsForward(dir))
*continuescan = false;
else if ((subkey->sk_flags & SK_BT_REQBKWD) &&
ScanDirectionIsBackward(dir))
*continuescan = false;
return false;
}
/* Perform the test --- three-way comparison not bool operator */
cmpresult = DatumGetInt32(FunctionCall2(&subkey->sk_func,
datum,
subkey->sk_argument));
if (subkey->sk_flags & SK_BT_DESC)
cmpresult = -cmpresult;
/* Done comparing if unequal, else advance to next column */
if (cmpresult != 0)
break;
if (subkey->sk_flags & SK_ROW_END)
break;
subkey++;
}
/*
* At this point cmpresult indicates the overall result of the row
* comparison, and subkey points to the deciding column (or the last
* column if the result is "=").
*/
switch (subkey->sk_strategy)
{
/* EQ and NE cases aren't allowed here */
case BTLessStrategyNumber:
result = (cmpresult < 0);
break;
case BTLessEqualStrategyNumber:
result = (cmpresult <= 0);
break;
case BTGreaterEqualStrategyNumber:
result = (cmpresult >= 0);
break;
case BTGreaterStrategyNumber:
result = (cmpresult > 0);
break;
default:
elog(ERROR, "unrecognized RowCompareType: %d",
(int) subkey->sk_strategy);
result = 0; /* keep compiler quiet */
break;
}
if (!result)
{
/*
* Tuple fails this qual. If it's a required qual for the current
* scan direction, then we can conclude no further tuples will pass,
* either. Note we have to look at the deciding column, not
* necessarily the first or last column of the row condition.
*/
if ((subkey->sk_flags & SK_BT_REQFWD) &&
ScanDirectionIsForward(dir))
*continuescan = false;
else if ((subkey->sk_flags & SK_BT_REQBKWD) &&
ScanDirectionIsBackward(dir))
*continuescan = false;
}
return result;
}
/* ----------------------------------------------------------------
* IndexOnlyNext
*
* Retrieve a tuple from the IndexOnlyScan node's index.
* ----------------------------------------------------------------
*/
static TupleTableSlot *
IndexOnlyNext(IndexOnlyScanState *node)
{
EState *estate;
ExprContext *econtext;
ScanDirection direction;
IndexScanDesc scandesc;
TupleTableSlot *slot;
ItemPointer tid;
/*
* extract necessary information from index scan node
*/
estate = node->ss.ps.state;
direction = estate->es_direction;
/* flip direction if this is an overall backward scan */
if (ScanDirectionIsBackward(((IndexOnlyScan *) node->ss.ps.plan)->indexorderdir))
{
if (ScanDirectionIsForward(direction))
direction = BackwardScanDirection;
else if (ScanDirectionIsBackward(direction))
direction = ForwardScanDirection;
}
scandesc = node->ioss_ScanDesc;
econtext = node->ss.ps.ps_ExprContext;
slot = node->ss.ss_ScanTupleSlot;
/*
* OK, now that we have what we need, fetch the next tuple.
*/
while ((tid = index_getnext_tid(scandesc, direction)) != NULL)
{
HeapTuple tuple = NULL;
/*
* We can skip the heap fetch if the TID references a heap page on
* which all tuples are known visible to everybody. In any case,
* we'll use the index tuple not the heap tuple as the data source.
*
* Note on Memory Ordering Effects: visibilitymap_test does not lock
* the visibility map buffer, and therefore the result we read here
* could be slightly stale. However, it can't be stale enough to
* matter.
*
* We need to detect clearing a VM bit due to an insert right away,
* because the tuple is present in the index page but not visible. The
* reading of the TID by this scan (using a shared lock on the index
* buffer) is serialized with the insert of the TID into the index
* (using an exclusive lock on the index buffer). Because the VM bit
* is cleared before updating the index, and locking/unlocking of the
* index page acts as a full memory barrier, we are sure to see the
* cleared bit if we see a recently-inserted TID.
*
* Deletes do not update the index page (only VACUUM will clear out
* the TID), so the clearing of the VM bit by a delete is not
* serialized with this test below, and we may see a value that is
* significantly stale. However, we don't care about the delete right
* away, because the tuple is still visible until the deleting
* transaction commits or the statement ends (if it's our
* transaction). In either case, the lock on the VM buffer will have
* been released (acting as a write barrier) after clearing the
* bit. And for us to have a snapshot that includes the deleting
* transaction (making the tuple invisible), we must have acquired
* ProcArrayLock after that time, acting as a read barrier.
*
* It's worth going through this complexity to avoid needing to lock
* the VM buffer, which could cause significant contention.
*/
if (!visibilitymap_test(scandesc->heapRelation,
ItemPointerGetBlockNumber(tid),
&node->ioss_VMBuffer))
{
/*
* Rats, we have to visit the heap to check visibility.
*/
node->ioss_HeapFetches++;
tuple = index_fetch_heap(scandesc);
if (tuple == NULL)
continue; /* no visible tuple, try next index entry */
/*
* Only MVCC snapshots are supported here, so there should be no
* need to keep following the HOT chain once a visible entry has
* been found. If we did want to allow that, we'd need to keep
* more state to remember not to call index_getnext_tid next time.
*/
if (scandesc->xs_continue_hot)
elog(ERROR, "non-MVCC snapshots are not supported in index-only scans");
/*
* Note: at this point we are holding a pin on the heap page, as
* recorded in scandesc->xs_cbuf. We could release that pin now,
* but it's not clear whether it's a win to do so. The next index
* entry might require a visit to the same heap page.
*/
}
/*
* Fill the scan tuple slot with data from the index.
*/
StoreIndexTuple(slot, scandesc->xs_itup, scandesc->xs_itupdesc);
/*
* If the index was lossy, we have to recheck the index quals.
* (Currently, this can never happen, but we should support the case
* for possible future use, eg with GiST indexes.)
*/
if (scandesc->xs_recheck)
{
econtext->ecxt_scantuple = slot;
ResetExprContext(econtext);
if (!ExecQual(node->indexqual, econtext, false))
{
/* Fails recheck, so drop it and loop back for another */
InstrCountFiltered2(node, 1);
continue;
}
}
/*
* We don't currently support rechecking ORDER BY distances. (In
* principle, if the index can support retrieval of the originally
* indexed value, it should be able to produce an exact distance
* calculation too. So it's not clear that adding code here for
* recheck/re-sort would be worth the trouble. But we should at least
* throw an error if someone tries it.)
*/
if (scandesc->numberOfOrderBys > 0 && scandesc->xs_recheckorderby)
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
errmsg("lossy distance functions are not supported in index-only scans")));
/*
* Predicate locks for index-only scans must be acquired at the page
* level when the heap is not accessed, since tuple-level predicate
* locks need the tuple's xmin value. If we had to visit the tuple
* anyway, then we already have the tuple-level lock and can skip the
* page lock.
*/
if (tuple == NULL)
PredicateLockPage(scandesc->heapRelation,
ItemPointerGetBlockNumber(tid),
estate->es_snapshot);
return slot;
}
/*
* if we get here it means the index scan failed so we are at the end of
* the scan..
*/
return ExecClearTuple(slot);
}
/* ----------------------------------------------------------------
* IndexNext
*
* Retrieve a tuple from the IndexScan node's currentRelation
* using the index specified in the IndexScanState information.
* ----------------------------------------------------------------
*/
TupleTableSlot *
IndexNext(IndexScanState *node)
{
EState *estate;
ExprContext *econtext;
ScanDirection direction;
IndexScanDesc scandesc;
Index scanrelid;
HeapTuple tuple;
TupleTableSlot *slot;
/*
* extract necessary information from index scan node
*/
estate = node->ss.ps.state;
direction = estate->es_direction;
initScanDesc(node);
/* flip direction if this is an overall backward scan */
if (ScanDirectionIsBackward(((IndexScan *) node->ss.ps.plan)->indexorderdir))
{
if (ScanDirectionIsForward(direction))
direction = BackwardScanDirection;
else if (ScanDirectionIsBackward(direction))
direction = ForwardScanDirection;
}
scandesc = node->iss_ScanDesc;
econtext = node->ss.ps.ps_ExprContext;
slot = node->ss.ss_ScanTupleSlot;
scanrelid = ((IndexScan *) node->ss.ps.plan)->scan.scanrelid;
/*
* Check if we are evaluating PlanQual for tuple of this relation.
* Additional checking is not good, but no other way for now. We could
* introduce new nodes for this case and handle IndexScan --> NewNode
* switching in Init/ReScan plan...
*/
if (estate->es_evTuple != NULL &&
estate->es_evTuple[scanrelid - 1] != NULL)
{
if (estate->es_evTupleNull[scanrelid - 1])
{
if (!node->ss.ps.delayEagerFree)
{
ExecEagerFreeIndexScan(node);
}
return ExecClearTuple(slot);
}
ExecStoreGenericTuple(estate->es_evTuple[scanrelid - 1], slot, false);
/* Does the tuple meet the indexqual condition? */
econtext->ecxt_scantuple = slot;
ResetExprContext(econtext);
if (!ExecQual(node->indexqualorig, econtext, false))
{
if (!node->ss.ps.delayEagerFree)
{
ExecEagerFreeIndexScan(node);
}
ExecClearTuple(slot); /* would not be returned by scan */
}
/* Flag for the next call that no more tuples */
estate->es_evTupleNull[scanrelid - 1] = true;
Gpmon_M_Incr_Rows_Out(GpmonPktFromIndexScanState(node));
CheckSendPlanStateGpmonPkt(&node->ss.ps);
return slot;
}
/*
* ok, now that we have what we need, fetch the next tuple.
*/
if ((tuple = index_getnext(scandesc, direction)) != NULL)
{
/*
* Store the scanned tuple in the scan tuple slot of the scan state.
* Note: we pass 'false' because tuples returned by amgetnext are
* pointers onto disk pages and must not be pfree()'d.
*/
ExecStoreHeapTuple(tuple, /* tuple to store */
slot, /* slot to store in */
scandesc->xs_cbuf, /* buffer containing tuple */
false); /* don't pfree */
Gpmon_M_Incr_Rows_Out(GpmonPktFromIndexScanState(node));
CheckSendPlanStateGpmonPkt(&node->ss.ps);
return slot;
}
if (!node->ss.ps.delayEagerFree)
{
ExecEagerFreeIndexScan(node);
}
/*
* if we get here it means the index scan failed so we are at the end of
* the scan..
*/
return ExecClearTuple(slot);
}
/*
* _hash_first() -- Find the first item in a scan.
*
* Find the first item in the index that
* satisfies the qualification associated with the scan descriptor. On
* success, the page containing the current index tuple is read locked
* and pinned, and the scan's opaque data entry is updated to
* include the buffer.
*/
bool
_hash_first(IndexScanDesc scan, ScanDirection dir)
{
Relation rel = scan->indexRelation;
HashScanOpaque so = (HashScanOpaque) scan->opaque;
ScanKey cur;
uint32 hashkey;
Bucket bucket;
BlockNumber blkno;
Buffer buf;
Buffer metabuf;
Page page;
HashPageOpaque opaque;
HashMetaPage metap;
IndexTuple itup;
ItemPointer current;
OffsetNumber offnum;
pgstat_count_index_scan(rel);
current = &(so->hashso_curpos);
ItemPointerSetInvalid(current);
/*
* We do not support hash scans with no index qualification, because we
* would have to read the whole index rather than just one bucket. That
* creates a whole raft of problems, since we haven't got a practical way
* to lock all the buckets against splits or compactions.
*/
if (scan->numberOfKeys < 1)
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
errmsg("hash indexes do not support whole-index scans")));
/* There may be more than one index qual, but we hash only the first */
cur = &scan->keyData[0];
/* We support only single-column hash indexes */
Assert(cur->sk_attno == 1);
/* And there's only one operator strategy, too */
Assert(cur->sk_strategy == HTEqualStrategyNumber);
/*
* If the constant in the index qual is NULL, assume it cannot match any
* items in the index.
*/
if (cur->sk_flags & SK_ISNULL)
return false;
/*
* Okay to compute the hash key. We want to do this before acquiring any
* locks, in case a user-defined hash function happens to be slow.
*
* If scankey operator is not a cross-type comparison, we can use the
* cached hash function; otherwise gotta look it up in the catalogs.
*
* We support the convention that sk_subtype == InvalidOid means the
* opclass input type; this is a hack to simplify life for ScanKeyInit().
*/
if (cur->sk_subtype == rel->rd_opcintype[0] ||
cur->sk_subtype == InvalidOid)
hashkey = _hash_datum2hashkey(rel, cur->sk_argument);
else
hashkey = _hash_datum2hashkey_type(rel, cur->sk_argument,
cur->sk_subtype);
so->hashso_sk_hash = hashkey;
/*
* Acquire shared split lock so we can compute the target bucket safely
* (see README).
*/
_hash_getlock(rel, 0, HASH_SHARE);
/* Read the metapage */
metabuf = _hash_getbuf(rel, HASH_METAPAGE, HASH_READ, LH_META_PAGE);
metap = HashPageGetMeta(BufferGetPage(metabuf));
/*
* Compute the target bucket number, and convert to block number.
*/
bucket = _hash_hashkey2bucket(hashkey,
metap->hashm_maxbucket,
metap->hashm_highmask,
metap->hashm_lowmask);
blkno = BUCKET_TO_BLKNO(metap, bucket);
/* done with the metapage */
_hash_relbuf(rel, metabuf);
/*
* Acquire share lock on target bucket; then we can release split lock.
*/
_hash_getlock(rel, blkno, HASH_SHARE);
_hash_droplock(rel, 0, HASH_SHARE);
/* Update scan opaque state to show we have lock on the bucket */
so->hashso_bucket = bucket;
so->hashso_bucket_valid = true;
so->hashso_bucket_blkno = blkno;
/* Fetch the primary bucket page for the bucket */
buf = _hash_getbuf(rel, blkno, HASH_READ, LH_BUCKET_PAGE);
page = BufferGetPage(buf);
opaque = (HashPageOpaque) PageGetSpecialPointer(page);
Assert(opaque->hasho_bucket == bucket);
/* If a backwards scan is requested, move to the end of the chain */
if (ScanDirectionIsBackward(dir))
{
while (BlockNumberIsValid(opaque->hasho_nextblkno))
_hash_readnext(rel, &buf, &page, &opaque);
}
/* Now find the first tuple satisfying the qualification */
if (!_hash_step(scan, &buf, dir))
return false;
/* if we're here, _hash_step found a valid tuple */
offnum = ItemPointerGetOffsetNumber(current);
_hash_checkpage(rel, buf, LH_BUCKET_PAGE | LH_OVERFLOW_PAGE);
page = BufferGetPage(buf);
itup = (IndexTuple) PageGetItem(page, PageGetItemId(page, offnum));
so->hashso_heappos = itup->t_tid;
return true;
}
static bool
rtnext(IndexScanDesc s, ScanDirection dir)
{
Page p;
OffsetNumber n;
RTreePageOpaque po;
RTreeScanOpaque so;
so = (RTreeScanOpaque) s->opaque;
if (!ItemPointerIsValid(&(s->currentItemData)))
{
/* first call: start at the root */
Assert(BufferIsValid(so->curbuf) == false);
so->curbuf = ReadBuffer(s->indexRelation, P_ROOT);
pgstat_count_index_scan(&s->xs_pgstat_info);
}
p = BufferGetPage(so->curbuf);
po = (RTreePageOpaque) PageGetSpecialPointer(p);
if (!ItemPointerIsValid(&(s->currentItemData)))
{
/* first call: start at first/last offset */
if (ScanDirectionIsForward(dir))
n = FirstOffsetNumber;
else
n = PageGetMaxOffsetNumber(p);
}
else
{
/* go on to the next offset */
n = ItemPointerGetOffsetNumber(&(s->currentItemData));
if (ScanDirectionIsForward(dir))
n = OffsetNumberNext(n);
else
n = OffsetNumberPrev(n);
}
for (;;)
{
IndexTuple it;
RTSTACK *stk;
n = findnext(s, n, dir);
/* no match on this page, so read in the next stack entry */
if (n == InvalidOffsetNumber)
{
/* if out of stack entries, we're done */
if (so->s_stack == NULL)
{
ReleaseBuffer(so->curbuf);
so->curbuf = InvalidBuffer;
return false;
}
stk = so->s_stack;
so->curbuf = ReleaseAndReadBuffer(so->curbuf, s->indexRelation,
stk->rts_blk);
p = BufferGetPage(so->curbuf);
po = (RTreePageOpaque) PageGetSpecialPointer(p);
if (ScanDirectionIsBackward(dir))
n = OffsetNumberPrev(stk->rts_child);
else
n = OffsetNumberNext(stk->rts_child);
so->s_stack = stk->rts_parent;
pfree(stk);
continue;
}
if (po->flags & F_LEAF)
{
ItemPointerSet(&(s->currentItemData),
BufferGetBlockNumber(so->curbuf),
n);
it = (IndexTuple) PageGetItem(p, PageGetItemId(p, n));
s->xs_ctup.t_self = it->t_tid;
return true;
}
else
{
BlockNumber blk;
stk = (RTSTACK *) palloc(sizeof(RTSTACK));
stk->rts_child = n;
stk->rts_blk = BufferGetBlockNumber(so->curbuf);
stk->rts_parent = so->s_stack;
so->s_stack = stk;
it = (IndexTuple) PageGetItem(p, PageGetItemId(p, n));
blk = ItemPointerGetBlockNumber(&(it->t_tid));
/*
* Note that we release the pin on the page as we descend down the
* tree, even though there's a good chance we'll eventually need
* to re-read the buffer later in this scan. This may or may not
* be optimal, but it doesn't seem likely to make a huge
* performance difference either way.
*/
so->curbuf = ReleaseAndReadBuffer(so->curbuf, s->indexRelation, blk);
p = BufferGetPage(so->curbuf);
po = (RTreePageOpaque) PageGetSpecialPointer(p);
if (ScanDirectionIsBackward(dir))
n = PageGetMaxOffsetNumber(p);
else
n = FirstOffsetNumber;
}
}
}
/* ----------------------------------------------------------------
* IndexOnlyNext
*
* Retrieve a tuple from the IndexOnlyScan node's index.
* ----------------------------------------------------------------
*/
static TupleTableSlot *
IndexOnlyNext(IndexOnlyScanState *node)
{
EState *estate;
ExprContext *econtext;
ScanDirection direction;
IndexScanDesc scandesc;
TupleTableSlot *slot;
ItemPointer tid;
/*
* extract necessary information from index scan node
*/
estate = node->ss.ps.state;
direction = estate->es_direction;
/* flip direction if this is an overall backward scan */
if (ScanDirectionIsBackward(((IndexOnlyScan *) node->ss.ps.plan)->indexorderdir))
{
if (ScanDirectionIsForward(direction))
direction = BackwardScanDirection;
else if (ScanDirectionIsBackward(direction))
direction = ForwardScanDirection;
}
scandesc = node->ioss_ScanDesc;
econtext = node->ss.ps.ps_ExprContext;
slot = node->ss.ss_ScanTupleSlot;
/*
* OK, now that we have what we need, fetch the next tuple.
*/
while ((tid = index_getnext_tid(scandesc, direction)) != NULL)
{
HeapTuple tuple = NULL;
/*
* We can skip the heap fetch if the TID references a heap page on
* which all tuples are known visible to everybody. In any case,
* we'll use the index tuple not the heap tuple as the data source.
*
* Note on Memory Ordering Effects: visibilitymap_test does not lock
* the visibility map buffer, and therefore the result we read here
* could be slightly stale. However, it can't be stale enough to
* matter. It suffices to show that (1) there is a read barrier
* between the time we read the index TID and the time we test the
* visibility map; and (2) there is a write barrier between the time
* some other concurrent process clears the visibility map bit and the
* time it inserts the index TID. Since acquiring or releasing a
* LWLock interposes a full barrier, this is easy to show: (1) is
* satisfied by the release of the index buffer content lock after
* reading the TID; and (2) is satisfied by the acquisition of the
* buffer content lock in order to insert the TID.
*/
if (!visibilitymap_test(scandesc->heapRelation,
ItemPointerGetBlockNumber(tid),
&node->ioss_VMBuffer))
{
/*
* Rats, we have to visit the heap to check visibility.
*/
node->ioss_HeapFetches++;
tuple = index_fetch_heap(scandesc);
if (tuple == NULL)
continue; /* no visible tuple, try next index entry */
/*
* Only MVCC snapshots are supported here, so there should be no
* need to keep following the HOT chain once a visible entry has
* been found. If we did want to allow that, we'd need to keep
* more state to remember not to call index_getnext_tid next time.
*/
if (scandesc->xs_continue_hot)
elog(ERROR, "non-MVCC snapshots are not supported in index-only scans");
/*
* Note: at this point we are holding a pin on the heap page, as
* recorded in scandesc->xs_cbuf. We could release that pin now,
* but it's not clear whether it's a win to do so. The next index
* entry might require a visit to the same heap page.
*/
}
/*
* Fill the scan tuple slot with data from the index.
*/
StoreIndexTuple(slot, scandesc->xs_itup, scandesc->xs_itupdesc);
/*
* If the index was lossy, we have to recheck the index quals.
* (Currently, this can never happen, but we should support the case
* for possible future use, eg with GiST indexes.)
*/
if (scandesc->xs_recheck)
{
econtext->ecxt_scantuple = slot;
ResetExprContext(econtext);
if (!ExecQual(node->indexqual, econtext, false))
{
/* Fails recheck, so drop it and loop back for another */
InstrCountFiltered2(node, 1);
continue;
}
}
/*
* Predicate locks for index-only scans must be acquired at the page
* level when the heap is not accessed, since tuple-level predicate
* locks need the tuple's xmin value. If we had to visit the tuple
* anyway, then we already have the tuple-level lock and can skip the
* page lock.
*/
if (tuple == NULL)
PredicateLockPage(scandesc->heapRelation,
ItemPointerGetBlockNumber(tid),
estate->es_snapshot);
return slot;
}
/*
* if we get here it means the index scan failed so we are at the end of
* the scan..
*/
return ExecClearTuple(slot);
}
/*
* _bt_first() -- Find the first item in a scan.
*
* We need to be clever about the direction of scan, the search
* conditions, and the tree ordering. We find the first item (or,
* if backwards scan, the last item) in the tree that satisfies the
* qualifications in the scan key. On success exit, the page containing
* the current index tuple is pinned but not locked, and data about
* the matching tuple(s) on the page has been loaded into so->currPos.
* scan->xs_ctup.t_self is set to the heap TID of the current tuple,
* and if requested, scan->xs_itup points to a copy of the index tuple.
*
* If there are no matching items in the index, we return FALSE, with no
* pins or locks held.
*
* Note that scan->keyData[], and the so->keyData[] scankey built from it,
* are both search-type scankeys (see nbtree/README for more about this).
* Within this routine, we build a temporary insertion-type scankey to use
* in locating the scan start position.
*/
bool
_bt_first(IndexScanDesc scan, ScanDirection dir)
{
Relation rel = scan->indexRelation;
BTScanOpaque so = (BTScanOpaque) scan->opaque;
Buffer buf;
BTStack stack;
OffsetNumber offnum;
StrategyNumber strat;
bool nextkey;
bool goback;
ScanKey startKeys[INDEX_MAX_KEYS];
ScanKeyData scankeys[INDEX_MAX_KEYS];
ScanKeyData notnullkeys[INDEX_MAX_KEYS];
int keysCount = 0;
int i;
StrategyNumber strat_total;
BTScanPosItem *currItem;
pgstat_count_index_scan(rel);
/*
* Examine the scan keys and eliminate any redundant keys; also mark the
* keys that must be matched to continue the scan.
*/
_bt_preprocess_keys(scan);
/*
* Quit now if _bt_preprocess_keys() discovered that the scan keys can
* never be satisfied (eg, x == 1 AND x > 2).
*/
if (!so->qual_ok)
return false;
/*----------
* Examine the scan keys to discover where we need to start the scan.
*
* We want to identify the keys that can be used as starting boundaries;
* these are =, >, or >= keys for a forward scan or =, <, <= keys for
* a backwards scan. We can use keys for multiple attributes so long as
* the prior attributes had only =, >= (resp. =, <=) keys. Once we accept
* a > or < boundary or find an attribute with no boundary (which can be
* thought of as the same as "> -infinity"), we can't use keys for any
* attributes to its right, because it would break our simplistic notion
* of what initial positioning strategy to use.
*
* When the scan keys include cross-type operators, _bt_preprocess_keys
* may not be able to eliminate redundant keys; in such cases we will
* arbitrarily pick a usable one for each attribute. This is correct
* but possibly not optimal behavior. (For example, with keys like
* "x >= 4 AND x >= 5" we would elect to scan starting at x=4 when
* x=5 would be more efficient.) Since the situation only arises given
* a poorly-worded query plus an incomplete opfamily, live with it.
*
* When both equality and inequality keys appear for a single attribute
* (again, only possible when cross-type operators appear), we *must*
* select one of the equality keys for the starting point, because
* _bt_checkkeys() will stop the scan as soon as an equality qual fails.
* For example, if we have keys like "x >= 4 AND x = 10" and we elect to
* start at x=4, we will fail and stop before reaching x=10. If multiple
* equality quals survive preprocessing, however, it doesn't matter which
* one we use --- by definition, they are either redundant or
* contradictory.
*
* Any regular (not SK_SEARCHNULL) key implies a NOT NULL qualifier.
* If the index stores nulls at the end of the index we'll be starting
* from, and we have no boundary key for the column (which means the key
* we deduced NOT NULL from is an inequality key that constrains the other
* end of the index), then we cons up an explicit SK_SEARCHNOTNULL key to
* use as a boundary key. If we didn't do this, we might find ourselves
* traversing a lot of null entries at the start of the scan.
*
* In this loop, row-comparison keys are treated the same as keys on their
* first (leftmost) columns. We'll add on lower-order columns of the row
* comparison below, if possible.
*
* The selected scan keys (at most one per index column) are remembered by
* storing their addresses into the local startKeys[] array.
*----------
*/
strat_total = BTEqualStrategyNumber;
if (so->numberOfKeys > 0)
{
AttrNumber curattr;
ScanKey chosen;
ScanKey impliesNN;
ScanKey cur;
/*
* chosen is the so-far-chosen key for the current attribute, if any.
* We don't cast the decision in stone until we reach keys for the
* next attribute.
*/
curattr = 1;
chosen = NULL;
/* Also remember any scankey that implies a NOT NULL constraint */
impliesNN = NULL;
/*
* Loop iterates from 0 to numberOfKeys inclusive; we use the last
* pass to handle after-last-key processing. Actual exit from the
* loop is at one of the "break" statements below.
*/
for (cur = so->keyData, i = 0;; cur++, i++)
{
if (i >= so->numberOfKeys || cur->sk_attno != curattr)
{
/*
* Done looking at keys for curattr. If we didn't find a
* usable boundary key, see if we can deduce a NOT NULL key.
*/
if (chosen == NULL && impliesNN != NULL &&
((impliesNN->sk_flags & SK_BT_NULLS_FIRST) ?
ScanDirectionIsForward(dir) :
ScanDirectionIsBackward(dir)))
{
/* Yes, so build the key in notnullkeys[keysCount] */
chosen = ¬nullkeys[keysCount];
ScanKeyEntryInitialize(chosen,
(SK_SEARCHNOTNULL | SK_ISNULL |
(impliesNN->sk_flags &
(SK_BT_DESC | SK_BT_NULLS_FIRST))),
curattr,
((impliesNN->sk_flags & SK_BT_NULLS_FIRST) ?
BTGreaterStrategyNumber :
BTLessStrategyNumber),
InvalidOid,
InvalidOid,
InvalidOid,
(Datum) 0);
}
/*
* If we still didn't find a usable boundary key, quit; else
* save the boundary key pointer in startKeys.
*/
if (chosen == NULL)
break;
startKeys[keysCount++] = chosen;
/*
* Adjust strat_total, and quit if we have stored a > or <
* key.
*/
strat = chosen->sk_strategy;
if (strat != BTEqualStrategyNumber)
{
strat_total = strat;
if (strat == BTGreaterStrategyNumber ||
strat == BTLessStrategyNumber)
break;
}
/*
* Done if that was the last attribute, or if next key is not
* in sequence (implying no boundary key is available for the
* next attribute).
*/
if (i >= so->numberOfKeys ||
cur->sk_attno != curattr + 1)
break;
/*
* Reset for next attr.
*/
curattr = cur->sk_attno;
chosen = NULL;
impliesNN = NULL;
}
/*
* Can we use this key as a starting boundary for this attr?
*
* If not, does it imply a NOT NULL constraint? (Because
* SK_SEARCHNULL keys are always assigned BTEqualStrategyNumber,
* *any* inequality key works for that; we need not test.)
*/
switch (cur->sk_strategy)
{
case BTLessStrategyNumber:
case BTLessEqualStrategyNumber:
if (chosen == NULL)
{
if (ScanDirectionIsBackward(dir))
chosen = cur;
else
impliesNN = cur;
}
break;
case BTEqualStrategyNumber:
/* override any non-equality choice */
chosen = cur;
break;
case BTGreaterEqualStrategyNumber:
case BTGreaterStrategyNumber:
if (chosen == NULL)
{
if (ScanDirectionIsForward(dir))
chosen = cur;
else
impliesNN = cur;
}
break;
}
}
}
/*
* If we found no usable boundary keys, we have to start from one end of
* the tree. Walk down that edge to the first or last key, and scan from
* there.
*/
if (keysCount == 0)
return _bt_endpoint(scan, dir);
/*
* We want to start the scan somewhere within the index. Set up an
* insertion scankey we can use to search for the boundary point we
* identified above. The insertion scankey is built in the local
* scankeys[] array, using the keys identified by startKeys[].
*/
Assert(keysCount <= INDEX_MAX_KEYS);
for (i = 0; i < keysCount; i++)
{
ScanKey cur = startKeys[i];
Assert(cur->sk_attno == i + 1);
if (cur->sk_flags & SK_ROW_HEADER)
{
/*
* Row comparison header: look to the first row member instead.
*
* The member scankeys are already in insertion format (ie, they
* have sk_func = 3-way-comparison function), but we have to watch
* out for nulls, which _bt_preprocess_keys didn't check. A null
* in the first row member makes the condition unmatchable, just
* like qual_ok = false.
*/
ScanKey subkey = (ScanKey) DatumGetPointer(cur->sk_argument);
Assert(subkey->sk_flags & SK_ROW_MEMBER);
if (subkey->sk_flags & SK_ISNULL)
return false;
memcpy(scankeys + i, subkey, sizeof(ScanKeyData));
/*
* If the row comparison is the last positioning key we accepted,
* try to add additional keys from the lower-order row members.
* (If we accepted independent conditions on additional index
* columns, we use those instead --- doesn't seem worth trying to
* determine which is more restrictive.) Note that this is OK
* even if the row comparison is of ">" or "<" type, because the
* condition applied to all but the last row member is effectively
* ">=" or "<=", and so the extra keys don't break the positioning
* scheme. But, by the same token, if we aren't able to use all
* the row members, then the part of the row comparison that we
* did use has to be treated as just a ">=" or "<=" condition, and
* so we'd better adjust strat_total accordingly.
*/
if (i == keysCount - 1)
{
bool used_all_subkeys = false;
Assert(!(subkey->sk_flags & SK_ROW_END));
for (;;)
{
subkey++;
Assert(subkey->sk_flags & SK_ROW_MEMBER);
if (subkey->sk_attno != keysCount + 1)
break; /* out-of-sequence, can't use it */
if (subkey->sk_strategy != cur->sk_strategy)
break; /* wrong direction, can't use it */
if (subkey->sk_flags & SK_ISNULL)
break; /* can't use null keys */
Assert(keysCount < INDEX_MAX_KEYS);
memcpy(scankeys + keysCount, subkey, sizeof(ScanKeyData));
keysCount++;
if (subkey->sk_flags & SK_ROW_END)
{
used_all_subkeys = true;
break;
}
}
if (!used_all_subkeys)
{
switch (strat_total)
{
case BTLessStrategyNumber:
strat_total = BTLessEqualStrategyNumber;
break;
case BTGreaterStrategyNumber:
strat_total = BTGreaterEqualStrategyNumber;
break;
}
}
break; /* done with outer loop */
}
}
else
{
/*
* Ordinary comparison key. Transform the search-style scan key
* to an insertion scan key by replacing the sk_func with the
* appropriate btree comparison function.
*
* If scankey operator is not a cross-type comparison, we can use
* the cached comparison function; otherwise gotta look it up in
* the catalogs. (That can't lead to infinite recursion, since no
* indexscan initiated by syscache lookup will use cross-data-type
* operators.)
*
* We support the convention that sk_subtype == InvalidOid means
* the opclass input type; this is a hack to simplify life for
* ScanKeyInit().
*/
if (cur->sk_subtype == rel->rd_opcintype[i] ||
cur->sk_subtype == InvalidOid)
{
FmgrInfo *procinfo;
procinfo = index_getprocinfo(rel, cur->sk_attno, BTORDER_PROC);
ScanKeyEntryInitializeWithInfo(scankeys + i,
cur->sk_flags,
cur->sk_attno,
InvalidStrategy,
cur->sk_subtype,
cur->sk_collation,
procinfo,
cur->sk_argument);
}
else
{
RegProcedure cmp_proc;
cmp_proc = get_opfamily_proc(rel->rd_opfamily[i],
rel->rd_opcintype[i],
cur->sk_subtype,
BTORDER_PROC);
if (!RegProcedureIsValid(cmp_proc))
elog(ERROR, "missing support function %d(%u,%u) for attribute %d of index \"%s\"",
BTORDER_PROC, rel->rd_opcintype[i], cur->sk_subtype,
cur->sk_attno, RelationGetRelationName(rel));
ScanKeyEntryInitialize(scankeys + i,
cur->sk_flags,
cur->sk_attno,
InvalidStrategy,
cur->sk_subtype,
cur->sk_collation,
cmp_proc,
cur->sk_argument);
}
}
}
/*----------
* Examine the selected initial-positioning strategy to determine exactly
* where we need to start the scan, and set flag variables to control the
* code below.
*
* If nextkey = false, _bt_search and _bt_binsrch will locate the first
* item >= scan key. If nextkey = true, they will locate the first
* item > scan key.
*
* If goback = true, we will then step back one item, while if
* goback = false, we will start the scan on the located item.
*----------
*/
switch (strat_total)
{
case BTLessStrategyNumber:
/*
* Find first item >= scankey, then back up one to arrive at last
* item < scankey. (Note: this positioning strategy is only used
* for a backward scan, so that is always the correct starting
* position.)
*/
nextkey = false;
goback = true;
break;
case BTLessEqualStrategyNumber:
/*
* Find first item > scankey, then back up one to arrive at last
* item <= scankey. (Note: this positioning strategy is only used
* for a backward scan, so that is always the correct starting
* position.)
*/
nextkey = true;
goback = true;
break;
case BTEqualStrategyNumber:
/*
* If a backward scan was specified, need to start with last equal
* item not first one.
*/
if (ScanDirectionIsBackward(dir))
{
/*
* This is the same as the <= strategy. We will check at the
* end whether the found item is actually =.
*/
nextkey = true;
goback = true;
}
else
{
/*
* This is the same as the >= strategy. We will check at the
* end whether the found item is actually =.
*/
nextkey = false;
goback = false;
}
break;
case BTGreaterEqualStrategyNumber:
/*
* Find first item >= scankey. (This is only used for forward
* scans.)
*/
nextkey = false;
goback = false;
break;
case BTGreaterStrategyNumber:
/*
* Find first item > scankey. (This is only used for forward
* scans.)
*/
nextkey = true;
goback = false;
break;
default:
/* can't get here, but keep compiler quiet */
elog(ERROR, "unrecognized strat_total: %d", (int) strat_total);
return false;
}
/*
* Use the manufactured insertion scan key to descend the tree and
* position ourselves on the target leaf page.
*/
stack = _bt_search(rel, keysCount, scankeys, nextkey, &buf, BT_READ);
/* don't need to keep the stack around... */
_bt_freestack(stack);
/* remember which buffer we have pinned, if any */
so->currPos.buf = buf;
if (!BufferIsValid(buf))
{
/*
* We only get here if the index is completely empty. Lock relation
* because nothing finer to lock exists.
*/
PredicateLockRelation(rel, scan->xs_snapshot);
return false;
}
else
PredicateLockPage(rel, BufferGetBlockNumber(buf),
scan->xs_snapshot);
/* initialize moreLeft/moreRight appropriately for scan direction */
if (ScanDirectionIsForward(dir))
{
so->currPos.moreLeft = false;
so->currPos.moreRight = true;
}
else
{
so->currPos.moreLeft = true;
so->currPos.moreRight = false;
}
so->numKilled = 0; /* just paranoia */
so->markItemIndex = -1; /* ditto */
/* position to the precise item on the page */
offnum = _bt_binsrch(rel, buf, keysCount, scankeys, nextkey);
/*
* If nextkey = false, we are positioned at the first item >= scan key, or
* possibly at the end of a page on which all the existing items are less
* than the scan key and we know that everything on later pages is greater
* than or equal to scan key.
*
* If nextkey = true, we are positioned at the first item > scan key, or
* possibly at the end of a page on which all the existing items are less
* than or equal to the scan key and we know that everything on later
* pages is greater than scan key.
*
* The actually desired starting point is either this item or the prior
* one, or in the end-of-page case it's the first item on the next page or
* the last item on this page. Adjust the starting offset if needed. (If
* this results in an offset before the first item or after the last one,
* _bt_readpage will report no items found, and then we'll step to the
* next page as needed.)
*/
if (goback)
offnum = OffsetNumberPrev(offnum);
/*
* Now load data from the first page of the scan.
*/
if (!_bt_readpage(scan, dir, offnum))
{
/*
* There's no actually-matching data on this page. Try to advance to
* the next page. Return false if there's no matching data at all.
*/
if (!_bt_steppage(scan, dir))
return false;
}
/* Drop the lock, but not pin, on the current page */
LockBuffer(so->currPos.buf, BUFFER_LOCK_UNLOCK);
/* OK, itemIndex says what to return */
currItem = &so->currPos.items[so->currPos.itemIndex];
scan->xs_ctup.t_self = currItem->heapTid;
if (scan->xs_want_itup)
scan->xs_itup = (IndexTuple) (so->currTuples + currItem->tupleOffset);
return true;
}
/*
* Test whether an indextuple satisfies all the scankey conditions.
*
* If so, copy its TID into scan->xs_ctup.t_self, and return TRUE.
* If not, return FALSE (xs_ctup is not changed).
*
* If the tuple fails to pass the qual, we also determine whether there's
* any need to continue the scan beyond this tuple, and set *continuescan
* accordingly. See comments for _bt_preprocess_keys(), above, about how
* this is done.
*
* scan: index scan descriptor (containing a search-type scankey)
* page: buffer page containing index tuple
* offnum: offset number of index tuple (must be a valid item!)
* dir: direction we are scanning in
* continuescan: output parameter (will be set correctly in all cases)
*/
bool
_bt_checkkeys(IndexScanDesc scan,
Page page, OffsetNumber offnum,
ScanDirection dir, bool *continuescan)
{
ItemId iid = PageGetItemId(page, offnum);
bool tuple_valid;
IndexTuple tuple;
TupleDesc tupdesc;
BTScanOpaque so;
int keysz;
int ikey;
ScanKey key;
*continuescan = true; /* default assumption */
/*
* If the scan specifies not to return killed tuples, then we treat a
* killed tuple as not passing the qual. Most of the time, it's a win to
* not bother examining the tuple's index keys, but just return
* immediately with continuescan = true to proceed to the next tuple.
* However, if this is the last tuple on the page, we should check the
* index keys to prevent uselessly advancing to the next page.
*/
if (scan->ignore_killed_tuples && ItemIdIsDead(iid))
{
/* return immediately if there are more tuples on the page */
if (ScanDirectionIsForward(dir))
{
if (offnum < PageGetMaxOffsetNumber(page))
return false;
}
else
{
BTPageOpaque opaque = (BTPageOpaque) PageGetSpecialPointer(page);
if (offnum > P_FIRSTDATAKEY(opaque))
return false;
}
/*
* OK, we want to check the keys, but we'll return FALSE even if the
* tuple passes the key tests.
*/
tuple_valid = false;
}
else
tuple_valid = true;
tuple = (IndexTuple) PageGetItem(page, iid);
IncrIndexProcessed();
tupdesc = RelationGetDescr(scan->indexRelation);
so = (BTScanOpaque) scan->opaque;
keysz = so->numberOfKeys;
for (key = so->keyData, ikey = 0; ikey < keysz; key++, ikey++)
{
Datum datum;
bool isNull;
Datum test;
/* row-comparison keys need special processing */
if (key->sk_flags & SK_ROW_HEADER)
{
if (_bt_check_rowcompare(key, tuple, tupdesc, dir, continuescan))
continue;
return false;
}
datum = index_getattr(tuple,
key->sk_attno,
tupdesc,
&isNull);
if (key->sk_flags & SK_ISNULL)
{
/* Handle IS NULL tests */
Assert(key->sk_flags & SK_SEARCHNULL);
if (isNull)
continue; /* tuple satisfies this qual */
/*
* Tuple fails this qual. If it's a required qual for the current
* scan direction, then we can conclude no further tuples will
* pass, either.
*/
if ((key->sk_flags & SK_BT_REQFWD) &&
ScanDirectionIsForward(dir))
*continuescan = false;
else if ((key->sk_flags & SK_BT_REQBKWD) &&
ScanDirectionIsBackward(dir))
*continuescan = false;
/*
* In any case, this indextuple doesn't match the qual.
*/
return false;
}
if (isNull)
{
if (key->sk_flags & SK_BT_NULLS_FIRST)
{
/*
* Since NULLs are sorted before non-NULLs, we know we have
* reached the lower limit of the range of values for this
* index attr. On a backward scan, we can stop if this qual
* is one of the "must match" subset. On a forward scan,
* however, we should keep going.
*/
if ((key->sk_flags & SK_BT_REQBKWD) &&
ScanDirectionIsBackward(dir))
*continuescan = false;
}
else
{
/*
* Since NULLs are sorted after non-NULLs, we know we have
* reached the upper limit of the range of values for this
* index attr. On a forward scan, we can stop if this qual is
* one of the "must match" subset. On a backward scan,
* however, we should keep going.
*/
if ((key->sk_flags & SK_BT_REQFWD) &&
ScanDirectionIsForward(dir))
*continuescan = false;
}
/*
* In any case, this indextuple doesn't match the qual.
*/
return false;
}
test = FunctionCall2(&key->sk_func, datum, key->sk_argument);
if (!DatumGetBool(test))
{
/*
* Tuple fails this qual. If it's a required qual for the current
* scan direction, then we can conclude no further tuples will
* pass, either.
*
* Note: because we stop the scan as soon as any required equality
* qual fails, it is critical that equality quals be used for the
* initial positioning in _bt_first() when they are available. See
* comments in _bt_first().
*/
if ((key->sk_flags & SK_BT_REQFWD) &&
ScanDirectionIsForward(dir))
*continuescan = false;
else if ((key->sk_flags & SK_BT_REQBKWD) &&
ScanDirectionIsBackward(dir))
*continuescan = false;
/*
* In any case, this indextuple doesn't match the qual.
*/
return false;
}
}
/* If we get here, the tuple passes all index quals. */
if (tuple_valid)
scan->xs_ctup.t_self = tuple->t_tid;
return tuple_valid;
}
/* ----------------------------------------------------------------
* IndexNext
*
* Retrieve a tuple from the IndexScan node's currentRelation
* using the index specified in the IndexScanState information.
* ----------------------------------------------------------------
*/
static TupleTableSlot *
IndexNext(IndexScanState *node)
{
EState *estate;
ExprContext *econtext;
ScanDirection direction;
IndexScanDesc scandesc;
HeapTuple tuple;
TupleTableSlot *slot;
/*
* extract necessary information from index scan node
*/
estate = node->ss.ps.state;
direction = estate->es_direction;
/* flip direction if this is an overall backward scan */
if (ScanDirectionIsBackward(((IndexScan *) node->ss.ps.plan)->indexorderdir))
{
if (ScanDirectionIsForward(direction))
direction = BackwardScanDirection;
else if (ScanDirectionIsBackward(direction))
direction = ForwardScanDirection;
}
scandesc = node->iss_ScanDesc;
econtext = node->ss.ps.ps_ExprContext;
slot = node->ss.ss_ScanTupleSlot;
/*
* ok, now that we have what we need, fetch the next tuple.
*/
while ((tuple = index_getnext(scandesc, direction)) != NULL)
{
/*
* Store the scanned tuple in the scan tuple slot of the scan state.
* Note: we pass 'false' because tuples returned by amgetnext are
* pointers onto disk pages and must not be pfree()'d.
*/
ExecStoreTuple(tuple, /* tuple to store */
slot, /* slot to store in */
scandesc->xs_cbuf, /* buffer containing tuple */
false); /* don't pfree */
/*
* If the index was lossy, we have to recheck the index quals using
* the real tuple.
*/
if (scandesc->xs_recheck)
{
econtext->ecxt_scantuple = slot;
ResetExprContext(econtext);
if (!ExecQual(node->indexqualorig, econtext, false))
continue; /* nope, so ask index for another one */
}
return slot;
}
/*
* if we get here it means the index scan failed so we are at the end of
* the scan..
*/
return ExecClearTuple(slot);
}
/*
* Fetch a tuples that matchs the search key; this can be invoked
* either to fetch the first such tuple or subsequent matching
* tuples. Returns true iff a matching tuple was found.
*/
static int
gistnext(IndexScanDesc scan, ScanDirection dir, ItemPointer tids,
int maxtids, bool ignore_killed_tuples)
{
MIRROREDLOCK_BUFMGR_DECLARE;
Page p;
OffsetNumber n;
GISTScanOpaque so;
GISTSearchStack *stk;
IndexTuple it;
GISTPageOpaque opaque;
int ntids = 0;
so = (GISTScanOpaque) scan->opaque;
// -------- MirroredLock ----------
MIRROREDLOCK_BUFMGR_LOCK;
if ( so->qual_ok == false )
return 0;
if (ItemPointerIsValid(&so->curpos) == false)
{
/* Being asked to fetch the first entry, so start at the root */
Assert(so->curbuf == InvalidBuffer);
Assert(so->stack == NULL);
so->curbuf = ReadBuffer(scan->indexRelation, GIST_ROOT_BLKNO);
stk = so->stack = (GISTSearchStack *) palloc0(sizeof(GISTSearchStack));
stk->next = NULL;
stk->block = GIST_ROOT_BLKNO;
pgstat_count_index_scan(scan->indexRelation);
}
else if (so->curbuf == InvalidBuffer)
{
MIRROREDLOCK_BUFMGR_UNLOCK;
// -------- MirroredLock ----------
return 0;
}
/*
* check stored pointers from last visit
*/
if ( so->nPageData > 0 )
{
while( ntids < maxtids && so->curPageData < so->nPageData )
{
tids[ ntids ] = scan->xs_ctup.t_self = so->pageData[ so->curPageData ].heapPtr;
ItemPointerSet(&(so->curpos),
BufferGetBlockNumber(so->curbuf),
so->pageData[ so->curPageData ].pageOffset);
so->curPageData ++;
ntids++;
}
if ( ntids == maxtids )
{
MIRROREDLOCK_BUFMGR_UNLOCK;
// -------- MirroredLock ----------
return ntids;
}
/*
* Go to the next page
*/
stk = so->stack->next;
pfree(so->stack);
so->stack = stk;
/* If we're out of stack entries, we're done */
if (so->stack == NULL)
{
ReleaseBuffer(so->curbuf);
so->curbuf = InvalidBuffer;
MIRROREDLOCK_BUFMGR_UNLOCK;
// -------- MirroredLock ----------
return ntids;
}
so->curbuf = ReleaseAndReadBuffer(so->curbuf,
scan->indexRelation,
stk->block);
}
for (;;)
{
/* First of all, we need lock buffer */
Assert(so->curbuf != InvalidBuffer);
LockBuffer(so->curbuf, GIST_SHARE);
gistcheckpage(scan->indexRelation, so->curbuf);
p = BufferGetPage(so->curbuf);
opaque = GistPageGetOpaque(p);
/* remember lsn to identify page changed for tuple's killing */
so->stack->lsn = PageGetLSN(p);
/* check page split, occured from last visit or visit to parent */
if (!XLogRecPtrIsInvalid(so->stack->parentlsn) &&
XLByteLT(so->stack->parentlsn, opaque->nsn) &&
opaque->rightlink != InvalidBlockNumber /* sanity check */ &&
(so->stack->next == NULL || so->stack->next->block != opaque->rightlink) /* check if already
added */ )
{
/* detect page split, follow right link to add pages */
stk = (GISTSearchStack *) palloc(sizeof(GISTSearchStack));
stk->next = so->stack->next;
stk->block = opaque->rightlink;
stk->parentlsn = so->stack->parentlsn;
memset(&(stk->lsn), 0, sizeof(GistNSN));
so->stack->next = stk;
}
/* if page is empty, then just skip it */
if (PageIsEmpty(p))
{
LockBuffer(so->curbuf, GIST_UNLOCK);
stk = so->stack->next;
pfree(so->stack);
so->stack = stk;
if (so->stack == NULL)
{
ReleaseBuffer(so->curbuf);
so->curbuf = InvalidBuffer;
MIRROREDLOCK_BUFMGR_UNLOCK;
// -------- MirroredLock ----------
return ntids;
}
so->curbuf = ReleaseAndReadBuffer(so->curbuf, scan->indexRelation,
stk->block);
continue;
}
if (ScanDirectionIsBackward(dir))
n = PageGetMaxOffsetNumber(p);
else
n = FirstOffsetNumber;
/* wonderful, we can look at page */
so->nPageData = so->curPageData = 0;
for (;;)
{
n = gistfindnext(scan, n, dir);
if (!OffsetNumberIsValid(n))
{
while( ntids < maxtids && so->curPageData < so->nPageData )
{
tids[ ntids ] = scan->xs_ctup.t_self =
so->pageData[ so->curPageData ].heapPtr;
ItemPointerSet(&(so->curpos),
BufferGetBlockNumber(so->curbuf),
so->pageData[ so->curPageData ].pageOffset);
so->curPageData ++;
ntids++;
}
if ( ntids == maxtids )
{
LockBuffer(so->curbuf, GIST_UNLOCK);
MIRROREDLOCK_BUFMGR_UNLOCK;
// -------- MirroredLock ----------
return ntids;
}
/*
* We ran out of matching index entries on the current page,
* so pop the top stack entry and use it to continue the
* search.
*/
LockBuffer(so->curbuf, GIST_UNLOCK);
stk = so->stack->next;
pfree(so->stack);
so->stack = stk;
/* If we're out of stack entries, we're done */
if (so->stack == NULL)
{
ReleaseBuffer(so->curbuf);
so->curbuf = InvalidBuffer;
MIRROREDLOCK_BUFMGR_UNLOCK;
// -------- MirroredLock ----------
return ntids;
}
so->curbuf = ReleaseAndReadBuffer(so->curbuf,
scan->indexRelation,
stk->block);
/* XXX go up */
break;
}
if (GistPageIsLeaf(p))
{
/*
* We've found a matching index entry in a leaf page, so
* return success. Note that we keep "curbuf" pinned so that
* we can efficiently resume the index scan later.
*/
if (!(ignore_killed_tuples && ItemIdIsDead(PageGetItemId(p, n))))
{
it = (IndexTuple) PageGetItem(p, PageGetItemId(p, n));
so->pageData[ so->nPageData ].heapPtr = it->t_tid;
so->pageData[ so->nPageData ].pageOffset = n;
so->nPageData ++;
}
}
else
{
/*
* We've found an entry in an internal node whose key is
* consistent with the search key, so push it to stack
*/
stk = (GISTSearchStack *) palloc(sizeof(GISTSearchStack));
it = (IndexTuple) PageGetItem(p, PageGetItemId(p, n));
stk->block = ItemPointerGetBlockNumber(&(it->t_tid));
memset(&(stk->lsn), 0, sizeof(GistNSN));
stk->parentlsn = so->stack->lsn;
stk->next = so->stack->next;
so->stack->next = stk;
}
if (ScanDirectionIsBackward(dir))
n = OffsetNumberPrev(n);
else
n = OffsetNumberNext(n);
}
}
MIRROREDLOCK_BUFMGR_UNLOCK;
// -------- MirroredLock ----------
return ntids;
}