static void
remove_duplicate(Spooler *self, Relation heap, IndexTuple itup, const char *relname)
{
HeapTupleData tuple;
BlockNumber blknum;
BlockNumber offnum;
Buffer buffer;
Page page;
ItemId itemid;
blknum = ItemPointerGetBlockNumber(&itup->t_tid);
offnum = ItemPointerGetOffsetNumber(&itup->t_tid);
buffer = ReadBuffer(heap, blknum);
LockBuffer(buffer, BUFFER_LOCK_SHARE);
page = BufferGetPage(buffer);
itemid = PageGetItemId(page, offnum);
tuple.t_data = ItemIdIsNormal(itemid)
? (HeapTupleHeader) PageGetItem(page, itemid)
: NULL;
LockBuffer(buffer, BUFFER_LOCK_UNLOCK);
if (tuple.t_data != NULL)
{
char *str;
TupleDesc tupdesc;
simple_heap_delete(heap, &itup->t_tid);
/* output duplicate bad file. */
if (self->dup_fp == NULL)
if ((self->dup_fp = AllocateFile(self->dup_badfile, "w")) == NULL)
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not open duplicate bad file \"%s\": %m",
self->dup_badfile)));
tupdesc = RelationGetDescr(heap);
tuple.t_len = ItemIdGetLength(itemid);
tuple.t_self = itup->t_tid;
str = tuple_to_cstring(RelationGetDescr(heap), &tuple);
if (fprintf(self->dup_fp, "%s\n", str) < 0 || fflush(self->dup_fp))
ereport(ERROR,
(errcode_for_file_access(),
errmsg("could not write parse badfile \"%s\": %m",
self->dup_badfile)));
pfree(str);
}
ReleaseBuffer(buffer);
LoggerLog(WARNING, "Duplicate error Record " int64_FMT
": Rejected - duplicate key value violates unique constraint \"%s\"\n",
self->dup_old + self->dup_new, relname);
}
/*
* Given an opened sequence relation, lock the page buffer and find the tuple
*
* *buf receives the reference to the pinned-and-ex-locked buffer
* *seqtuple receives the reference to the sequence tuple proper
* (this arg should point to a local variable of type HeapTupleData)
*
* Function's return value points to the data payload of the tuple
*/
static Form_pg_sequence
read_seq_tuple(SeqTable elm, Relation rel, Buffer *buf, HeapTuple seqtuple)
{
Page page;
ItemId lp;
sequence_magic *sm;
Form_pg_sequence seq;
*buf = ReadBuffer(rel, 0);
LockBuffer(*buf, BUFFER_LOCK_EXCLUSIVE);
page = BufferGetPage(*buf);
sm = (sequence_magic *) PageGetSpecialPointer(page);
if (sm->magic != SEQ_MAGIC)
elog(ERROR, "bad magic number in sequence \"%s\": %08X",
RelationGetRelationName(rel), sm->magic);
lp = PageGetItemId(page, FirstOffsetNumber);
Assert(ItemIdIsNormal(lp));
/* Note we currently only bother to set these two fields of *seqtuple */
seqtuple->t_data = (HeapTupleHeader) PageGetItem(page, lp);
seqtuple->t_len = ItemIdGetLength(lp);
/*
* Previous releases of Postgres neglected to prevent SELECT FOR UPDATE on
* a sequence, which would leave a non-frozen XID in the sequence tuple's
* xmax, which eventually leads to clog access failures or worse. If we
* see this has happened, clean up after it. We treat this like a hint
* bit update, ie, don't bother to WAL-log it, since we can certainly do
* this again if the update gets lost.
*/
Assert(!(seqtuple->t_data->t_infomask & HEAP_XMAX_IS_MULTI));
if (HeapTupleHeaderGetRawXmax(seqtuple->t_data) != InvalidTransactionId)
{
HeapTupleHeaderSetXmax(seqtuple->t_data, InvalidTransactionId);
seqtuple->t_data->t_infomask &= ~HEAP_XMAX_COMMITTED;
seqtuple->t_data->t_infomask |= HEAP_XMAX_INVALID;
MarkBufferDirtyHint(*buf, true);
}
seq = (Form_pg_sequence) GETSTRUCT(seqtuple);
/* this is a handy place to update our copy of the increment */
elm->increment = seq->increment_by;
return seq;
}
static text *
istatus_text(ItemId itemid)
{
StringInfoData buf;
initStringInfo(&buf);
if (ItemIdDeleted(itemid))
appendStringInfoString(&buf, "DELETED ");
if (ItemIdIsNormal(itemid))
appendStringInfoString(&buf, "USED ");
if (ItemIdIsDead(itemid))
appendStringInfoString(&buf, "DEAD ");
if (buf.len == 0)
appendStringInfoString(&buf, "UNUSED ");
buf.data[buf.len - 1] = '\0';
return cstring_to_text(buf.data);
}
/*
* lazy_check_needs_freeze() -- scan page to see if any tuples
* need to be cleaned to avoid wraparound
*
* Returns true if the page needs to be vacuumed using cleanup lock.
*/
static bool
lazy_check_needs_freeze(Buffer buf)
{
Page page;
OffsetNumber offnum,
maxoff;
HeapTupleHeader tupleheader;
page = BufferGetPage(buf);
if (PageIsNew(page) || PageIsEmpty(page))
{
/* PageIsNew probably shouldn't happen... */
return false;
}
maxoff = PageGetMaxOffsetNumber(page);
for (offnum = FirstOffsetNumber;
offnum <= maxoff;
offnum = OffsetNumberNext(offnum))
{
ItemId itemid;
itemid = PageGetItemId(page, offnum);
if (!ItemIdIsNormal(itemid))
continue;
tupleheader = (HeapTupleHeader) PageGetItem(page, itemid);
if (heap_tuple_needs_freeze(tupleheader, FreezeLimit, buf))
return true;
} /* scan along page */
return false;
}
/* ----------------------------------------------------------------
* BitmapHeapNext
*
* Retrieve next tuple from the BitmapHeapScan node's currentRelation
* ----------------------------------------------------------------
*/
static TupleTableSlot *
BitmapHeapNext(BitmapHeapScanState *node)
{
ExprContext *econtext;
HeapScanDesc scan;
TIDBitmap *tbm;
TBMIterator *tbmiterator;
TBMIterateResult *tbmres;
TBMIterator *prefetch_iterator;
OffsetNumber targoffset;
TupleTableSlot *slot;
/*
* extract necessary information from index scan node
*/
econtext = node->ss.ps.ps_ExprContext;
slot = node->ss.ss_ScanTupleSlot;
scan = node->ss.ss_currentScanDesc;
tbm = node->tbm;
tbmiterator = node->tbmiterator;
tbmres = node->tbmres;
prefetch_iterator = node->prefetch_iterator;
/*
* If we haven't yet performed the underlying index scan, do it, and begin
* the iteration over the bitmap.
*
* For prefetching, we use *two* iterators, one for the pages we are
* actually scanning and another that runs ahead of the first for
* prefetching. node->prefetch_pages tracks exactly how many pages ahead
* the prefetch iterator is. Also, node->prefetch_target tracks the
* desired prefetch distance, which starts small and increases up to the
* GUC-controlled maximum, target_prefetch_pages. This is to avoid doing
* a lot of prefetching in a scan that stops after a few tuples because of
* a LIMIT.
*/
if (tbm == NULL)
{
tbm = (TIDBitmap *) MultiExecProcNode(outerPlanState(node));
if (!tbm || !IsA(tbm, TIDBitmap))
elog(ERROR, "unrecognized result from subplan");
node->tbm = tbm;
node->tbmiterator = tbmiterator = tbm_begin_iterate(tbm);
node->tbmres = tbmres = NULL;
#ifdef USE_PREFETCH
if (target_prefetch_pages > 0)
{
node->prefetch_iterator = prefetch_iterator = tbm_begin_iterate(tbm);
node->prefetch_pages = 0;
node->prefetch_target = -1;
}
#endif /* USE_PREFETCH */
}
for (;;)
{
Page dp;
ItemId lp;
/*
* Get next page of results if needed
*/
if (tbmres == NULL)
{
node->tbmres = tbmres = tbm_iterate(tbmiterator);
if (tbmres == NULL)
{
/* no more entries in the bitmap */
break;
}
#ifdef USE_PREFETCH
if (node->prefetch_pages > 0)
{
/* The main iterator has closed the distance by one page */
node->prefetch_pages--;
}
else if (prefetch_iterator)
{
/* Do not let the prefetch iterator get behind the main one */
TBMIterateResult *tbmpre = tbm_iterate(prefetch_iterator);
if (tbmpre == NULL || tbmpre->blockno != tbmres->blockno)
elog(ERROR, "prefetch and main iterators are out of sync");
}
#endif /* USE_PREFETCH */
/*
* Ignore any claimed entries past what we think is the end of the
* relation. (This is probably not necessary given that we got at
* least AccessShareLock on the table before performing any of the
* indexscans, but let's be safe.)
*/
if (tbmres->blockno >= scan->rs_nblocks)
{
node->tbmres = tbmres = NULL;
continue;
}
/*
* Fetch the current heap page and identify candidate tuples.
*/
bitgetpage(scan, tbmres);
/*
* Set rs_cindex to first slot to examine
*/
scan->rs_cindex = 0;
#ifdef USE_PREFETCH
/*
* Increase prefetch target if it's not yet at the max. Note that
* we will increase it to zero after fetching the very first
* page/tuple, then to one after the second tuple is fetched, then
* it doubles as later pages are fetched.
*/
if (node->prefetch_target >= target_prefetch_pages)
/* don't increase any further */ ;
else if (node->prefetch_target >= target_prefetch_pages / 2)
node->prefetch_target = target_prefetch_pages;
else if (node->prefetch_target > 0)
node->prefetch_target *= 2;
else
node->prefetch_target++;
#endif /* USE_PREFETCH */
}
else
{
/*
* Continuing in previously obtained page; advance rs_cindex
*/
scan->rs_cindex++;
#ifdef USE_PREFETCH
/*
* Try to prefetch at least a few pages even before we get to the
* second page if we don't stop reading after the first tuple.
*/
if (node->prefetch_target < target_prefetch_pages)
node->prefetch_target++;
#endif /* USE_PREFETCH */
}
/*
* Out of range? If so, nothing more to look at on this page
*/
if (scan->rs_cindex < 0 || scan->rs_cindex >= scan->rs_ntuples)
{
node->tbmres = tbmres = NULL;
continue;
}
#ifdef USE_PREFETCH
/*
* We issue prefetch requests *after* fetching the current page to try
* to avoid having prefetching interfere with the main I/O. Also, this
* should happen only when we have determined there is still something
* to do on the current page, else we may uselessly prefetch the same
* page we are just about to request for real.
*/
if (prefetch_iterator)
{
while (node->prefetch_pages < node->prefetch_target)
{
TBMIterateResult *tbmpre = tbm_iterate(prefetch_iterator);
if (tbmpre == NULL)
{
/* No more pages to prefetch */
tbm_end_iterate(prefetch_iterator);
node->prefetch_iterator = prefetch_iterator = NULL;
break;
}
node->prefetch_pages++;
PrefetchBuffer(scan->rs_rd, MAIN_FORKNUM, tbmpre->blockno);
}
}
#endif /* USE_PREFETCH */
/*
* Okay to fetch the tuple
*/
targoffset = scan->rs_vistuples[scan->rs_cindex];
dp = (Page) BufferGetPage(scan->rs_cbuf);
lp = PageGetItemId(dp, targoffset);
Assert(ItemIdIsNormal(lp));
scan->rs_ctup.t_data = (HeapTupleHeader) PageGetItem((Page) dp, lp);
scan->rs_ctup.t_len = ItemIdGetLength(lp);
ItemPointerSet(&scan->rs_ctup.t_self, tbmres->blockno, targoffset);
pgstat_count_heap_fetch(scan->rs_rd);
/*
* Set up the result slot to point to this tuple. Note that the slot
* acquires a pin on the buffer.
*/
ExecStoreTuple(&scan->rs_ctup,
slot,
scan->rs_cbuf,
false);
/*
* If we are using lossy info, we have to recheck the qual conditions
* at every tuple.
*/
if (tbmres->recheck)
{
econtext->ecxt_scantuple = slot;
ResetExprContext(econtext);
if (!ExecQual(node->bitmapqualorig, econtext, false))
{
/* Fails recheck, so drop it and loop back for another */
ExecClearTuple(slot);
continue;
}
}
/* OK to return this tuple */
return slot;
}
/*
* if we get here it means we are at the end of the scan..
*/
return ExecClearTuple(slot);
}
/*
* bitgetpage - subroutine for BitmapHeapNext()
*
* This routine reads and pins the specified page of the relation, then
* builds an array indicating which tuples on the page are both potentially
* interesting according to the bitmap, and visible according to the snapshot.
*/
static void
bitgetpage(HeapScanDesc scan, TBMIterateResult *tbmres)
{
BlockNumber page = tbmres->blockno;
Buffer buffer;
Snapshot snapshot;
int ntup;
/*
* Acquire pin on the target heap page, trading in any pin we held before.
*/
Assert(page < scan->rs_nblocks);
scan->rs_cbuf = ReleaseAndReadBuffer(scan->rs_cbuf,
scan->rs_rd,
page);
buffer = scan->rs_cbuf;
snapshot = scan->rs_snapshot;
ntup = 0;
/*
* Prune and repair fragmentation for the whole page, if possible.
*/
Assert(TransactionIdIsValid(RecentGlobalXmin));
heap_page_prune_opt(scan->rs_rd, buffer, RecentGlobalXmin);
/*
* We must hold share lock on the buffer content while examining tuple
* visibility. Afterwards, however, the tuples we have found to be
* visible are guaranteed good as long as we hold the buffer pin.
*/
LockBuffer(buffer, BUFFER_LOCK_SHARE);
/*
* We need two separate strategies for lossy and non-lossy cases.
*/
if (tbmres->ntuples >= 0)
{
/*
* Bitmap is non-lossy, so we just look through the offsets listed in
* tbmres; but we have to follow any HOT chain starting at each such
* offset.
*/
int curslot;
for (curslot = 0; curslot < tbmres->ntuples; curslot++)
{
OffsetNumber offnum = tbmres->offsets[curslot];
ItemPointerData tid;
ItemPointerSet(&tid, page, offnum);
if (heap_hot_search_buffer(&tid, scan->rs_rd, buffer, snapshot, NULL))
scan->rs_vistuples[ntup++] = ItemPointerGetOffsetNumber(&tid);
}
}
else
{
/*
* Bitmap is lossy, so we must examine each item pointer on the page.
* But we can ignore HOT chains, since we'll check each tuple anyway.
*/
Page dp = (Page) BufferGetPage(buffer);
OffsetNumber maxoff = PageGetMaxOffsetNumber(dp);
OffsetNumber offnum;
for (offnum = FirstOffsetNumber; offnum <= maxoff; offnum = OffsetNumberNext(offnum))
{
ItemId lp;
HeapTupleData loctup;
bool valid;
lp = PageGetItemId(dp, offnum);
if (!ItemIdIsNormal(lp))
continue;
loctup.t_data = (HeapTupleHeader) PageGetItem((Page) dp, lp);
loctup.t_len = ItemIdGetLength(lp);
loctup.t_tableOid = scan->rs_rd->rd_id;
ItemPointerSet(&loctup.t_self, page, offnum);
valid = HeapTupleSatisfiesVisibility(&loctup, snapshot, buffer);
if (valid)
{
scan->rs_vistuples[ntup++] = offnum;
PredicateLockTuple(scan->rs_rd, &loctup, snapshot);
}
CheckForSerializableConflictOut(valid, scan->rs_rd, &loctup,
buffer, snapshot);
}
}
LockBuffer(buffer, BUFFER_LOCK_UNLOCK);
Assert(ntup <= MaxHeapTuplesPerPage);
scan->rs_ntuples = ntup;
}
/*
* For all items in this page, find their respective root line pointers.
* If item k is part of a HOT-chain with root at item j, then we set
* root_offsets[k - 1] = j.
*
* The passed-in root_offsets array must have MaxHeapTuplesPerPage entries.
* We zero out all unused entries.
*
* The function must be called with at least share lock on the buffer, to
* prevent concurrent prune operations.
*
* Note: The information collected here is valid only as long as the caller
* holds a pin on the buffer. Once pin is released, a tuple might be pruned
* and reused by a completely unrelated tuple.
*/
void
heap_get_root_tuples(Page page, OffsetNumber *root_offsets)
{
OffsetNumber offnum,
maxoff;
MemSet(root_offsets, 0, MaxHeapTuplesPerPage * sizeof(OffsetNumber));
maxoff = PageGetMaxOffsetNumber(page);
for (offnum = FirstOffsetNumber; offnum <= maxoff; offnum = OffsetNumberNext(offnum))
{
ItemId lp = PageGetItemId(page, offnum);
HeapTupleHeader htup;
OffsetNumber nextoffnum;
TransactionId priorXmax;
/* skip unused and dead items */
if (!ItemIdIsUsed(lp) || ItemIdIsDead(lp))
continue;
if (ItemIdIsNormal(lp))
{
htup = (HeapTupleHeader) PageGetItem(page, lp);
/*
* Check if this tuple is part of a HOT-chain rooted at some other
* tuple. If so, skip it for now; we'll process it when we find
* its root.
*/
if (HeapTupleHeaderIsHeapOnly(htup))
continue;
/*
* This is either a plain tuple or the root of a HOT-chain.
* Remember it in the mapping.
*/
root_offsets[offnum - 1] = offnum;
/* If it's not the start of a HOT-chain, we're done with it */
if (!HeapTupleHeaderIsHotUpdated(htup))
continue;
/* Set up to scan the HOT-chain */
nextoffnum = ItemPointerGetOffsetNumber(&htup->t_ctid);
priorXmax = HeapTupleHeaderGetXmax(htup);
}
else
{
/* Must be a redirect item. We do not set its root_offsets entry */
Assert(ItemIdIsRedirected(lp));
/* Set up to scan the HOT-chain */
nextoffnum = ItemIdGetRedirect(lp);
priorXmax = InvalidTransactionId;
}
/*
* Now follow the HOT-chain and collect other tuples in the chain.
*
* Note: Even though this is a nested loop, the complexity of the
* function is O(N) because a tuple in the page should be visited not
* more than twice, once in the outer loop and once in HOT-chain
* chases.
*/
for (;;)
{
lp = PageGetItemId(page, nextoffnum);
/* Check for broken chains */
if (!ItemIdIsNormal(lp))
break;
htup = (HeapTupleHeader) PageGetItem(page, lp);
if (TransactionIdIsValid(priorXmax) &&
!TransactionIdEquals(priorXmax, HeapTupleHeaderGetXmin(htup)))
break;
/* Remember the root line pointer for this item */
root_offsets[nextoffnum - 1] = offnum;
/* Advance to next chain member, if any */
if (!HeapTupleHeaderIsHotUpdated(htup))
break;
nextoffnum = ItemPointerGetOffsetNumber(&htup->t_ctid);
priorXmax = HeapTupleHeaderGetXmax(htup);
}
}
}
/*
* Perform the actual page changes needed by heap_page_prune.
* It is expected that the caller has suitable pin and lock on the
* buffer, and is inside a critical section.
*
* This is split out because it is also used by heap_xlog_clean()
* to replay the WAL record when needed after a crash. Note that the
* arguments are identical to those of log_heap_clean().
*/
void
heap_page_prune_execute(Buffer buffer,
OffsetNumber *redirected, int nredirected,
OffsetNumber *nowdead, int ndead,
OffsetNumber *nowunused, int nunused,
bool redirect_move)
{
Page page = (Page) BufferGetPage(buffer);
OffsetNumber *offnum;
int i;
/* Update all redirected or moved line pointers */
offnum = redirected;
for (i = 0; i < nredirected; i++)
{
OffsetNumber fromoff = *offnum++;
OffsetNumber tooff = *offnum++;
ItemId fromlp = PageGetItemId(page, fromoff);
if (redirect_move)
{
/* Physically move the "to" item to the "from" slot */
ItemId tolp = PageGetItemId(page, tooff);
HeapTupleHeader htup;
*fromlp = *tolp;
ItemIdSetUnused(tolp);
/*
* Change heap-only status of the tuple because after the line
* pointer manipulation, it's no longer a heap-only tuple, but is
* directly pointed to by index entries.
*/
Assert(ItemIdIsNormal(fromlp));
htup = (HeapTupleHeader) PageGetItem(page, fromlp);
Assert(HeapTupleHeaderIsHeapOnly(htup));
HeapTupleHeaderClearHeapOnly(htup);
}
else
{
/* Just insert a REDIRECT link at fromoff */
ItemIdSetRedirect(fromlp, tooff);
}
}
/* Update all now-dead line pointers */
offnum = nowdead;
for (i = 0; i < ndead; i++)
{
OffsetNumber off = *offnum++;
ItemId lp = PageGetItemId(page, off);
ItemIdSetDead(lp);
}
/* Update all now-unused line pointers */
offnum = nowunused;
for (i = 0; i < nunused; i++)
{
OffsetNumber off = *offnum++;
ItemId lp = PageGetItemId(page, off);
ItemIdSetUnused(lp);
}
/*
* Finally, repair any fragmentation, and update the page's hint bit about
* whether it has free pointers.
*/
PageRepairFragmentation(page);
}
/*
* Prune specified item pointer or a HOT chain originating at that item.
*
* If the item is an index-referenced tuple (i.e. not a heap-only tuple),
* the HOT chain is pruned by removing all DEAD tuples at the start of the HOT
* chain. We also prune any RECENTLY_DEAD tuples preceding a DEAD tuple.
* This is OK because a RECENTLY_DEAD tuple preceding a DEAD tuple is really
* DEAD, the OldestXmin test is just too coarse to detect it.
*
* The root line pointer is redirected to the tuple immediately after the
* latest DEAD tuple. If all tuples in the chain are DEAD, the root line
* pointer is marked LP_DEAD. (This includes the case of a DEAD simple
* tuple, which we treat as a chain of length 1.)
*
* OldestXmin is the cutoff XID used to identify dead tuples.
*
* We don't actually change the page here, except perhaps for hint-bit updates
* caused by HeapTupleSatisfiesVacuum. We just add entries to the arrays in
* prstate showing the changes to be made. Items to be redirected are added
* to the redirected[] array (two entries per redirection); items to be set to
* LP_DEAD state are added to nowdead[]; and items to be set to LP_UNUSED
* state are added to nowunused[].
*
* If redirect_move is true, we intend to get rid of redirecting line pointers,
* not just make redirection entries.
*
* Returns the number of tuples (to be) deleted from the page.
*/
static int
heap_prune_chain(Relation relation, Buffer buffer, OffsetNumber rootoffnum,
TransactionId OldestXmin,
PruneState *prstate,
bool redirect_move)
{
int ndeleted = 0;
Page dp = (Page) BufferGetPage(buffer);
TransactionId priorXmax = InvalidTransactionId;
ItemId rootlp;
HeapTupleHeader htup;
OffsetNumber latestdead = InvalidOffsetNumber,
redirect_target = InvalidOffsetNumber,
maxoff = PageGetMaxOffsetNumber(dp),
offnum;
OffsetNumber chainitems[MaxHeapTuplesPerPage];
int nchain = 0,
i;
rootlp = PageGetItemId(dp, rootoffnum);
/*
* If it's a heap-only tuple, then it is not the start of a HOT chain.
*/
if (ItemIdIsNormal(rootlp))
{
htup = (HeapTupleHeader) PageGetItem(dp, rootlp);
if (HeapTupleHeaderIsHeapOnly(htup))
{
/*
* If the tuple is DEAD and doesn't chain to anything else, mark
* it unused immediately. (If it does chain, we can only remove
* it as part of pruning its chain.)
*
* We need this primarily to handle aborted HOT updates, that is,
* XMIN_INVALID heap-only tuples. Those might not be linked to by
* any chain, since the parent tuple might be re-updated before
* any pruning occurs. So we have to be able to reap them
* separately from chain-pruning. (Note that
* HeapTupleHeaderIsHotUpdated will never return true for an
* XMIN_INVALID tuple, so this code will work even when there were
* sequential updates within the aborted transaction.)
*
* Note that we might first arrive at a dead heap-only tuple
* either here or while following a chain below. Whichever path
* gets there first will mark the tuple unused.
*/
if (HeapTupleSatisfiesVacuum(relation, htup, OldestXmin, buffer)
== HEAPTUPLE_DEAD && !HeapTupleHeaderIsHotUpdated(htup))
{
heap_prune_record_unused(prstate, rootoffnum);
ndeleted++;
}
/* Nothing more to do */
return ndeleted;
}
}
/* Start from the root tuple */
offnum = rootoffnum;
/* while not end of the chain */
for (;;)
{
ItemId lp;
bool tupdead,
recent_dead;
/* Some sanity checks */
if (offnum < FirstOffsetNumber || offnum > maxoff)
break;
/* If item is already processed, stop --- it must not be same chain */
if (prstate->marked[offnum])
break;
lp = PageGetItemId(dp, offnum);
/* Unused item obviously isn't part of the chain */
if (!ItemIdIsUsed(lp))
break;
/*
* If we are looking at the redirected root line pointer, jump to the
* first normal tuple in the chain. If we find a redirect somewhere
* else, stop --- it must not be same chain.
*/
if (ItemIdIsRedirected(lp))
{
if (nchain > 0)
break; /* not at start of chain */
chainitems[nchain++] = offnum;
offnum = ItemIdGetRedirect(rootlp);
continue;
}
/*
* Likewise, a dead item pointer can't be part of the chain. (We
* already eliminated the case of dead root tuple outside this
* function.)
*/
if (ItemIdIsDead(lp))
break;
Assert(ItemIdIsNormal(lp));
htup = (HeapTupleHeader) PageGetItem(dp, lp);
/*
* Check the tuple XMIN against prior XMAX, if any
*/
if (TransactionIdIsValid(priorXmax) &&
!TransactionIdEquals(HeapTupleHeaderGetXmin(htup), priorXmax))
break;
/*
* OK, this tuple is indeed a member of the chain.
*/
chainitems[nchain++] = offnum;
/*
* Check tuple's visibility status.
*/
tupdead = recent_dead = false;
switch (HeapTupleSatisfiesVacuum(relation, htup, OldestXmin, buffer))
{
case HEAPTUPLE_DEAD:
tupdead = true;
break;
case HEAPTUPLE_RECENTLY_DEAD:
recent_dead = true;
/*
* This tuple may soon become DEAD. Update the hint field so
* that the page is reconsidered for pruning in future.
*/
heap_prune_record_prunable(prstate,
HeapTupleHeaderGetXmax(htup));
break;
case HEAPTUPLE_DELETE_IN_PROGRESS:
/*
* This tuple may soon become DEAD. Update the hint field so
* that the page is reconsidered for pruning in future.
*/
heap_prune_record_prunable(prstate,
HeapTupleHeaderGetXmax(htup));
break;
case HEAPTUPLE_LIVE:
case HEAPTUPLE_INSERT_IN_PROGRESS:
/*
* If we wanted to optimize for aborts, we might consider
* marking the page prunable when we see INSERT_IN_PROGRESS.
* But we don't. See related decisions about when to mark the
* page prunable in heapam.c.
*/
break;
default:
elog(ERROR, "unexpected HeapTupleSatisfiesVacuum result");
break;
}
/*
* Remember the last DEAD tuple seen. We will advance past
* RECENTLY_DEAD tuples just in case there's a DEAD one after them;
* but we can't advance past anything else. (XXX is it really worth
* continuing to scan beyond RECENTLY_DEAD? The case where we will
* find another DEAD tuple is a fairly unusual corner case.)
*/
if (tupdead)
latestdead = offnum;
else if (!recent_dead)
break;
/*
* If the tuple is not HOT-updated, then we are at the end of this
* HOT-update chain.
*/
if (!HeapTupleHeaderIsHotUpdated(htup))
break;
/*
* Advance to next chain member.
*/
Assert(ItemPointerGetBlockNumber(&htup->t_ctid) ==
BufferGetBlockNumber(buffer));
offnum = ItemPointerGetOffsetNumber(&htup->t_ctid);
priorXmax = HeapTupleHeaderGetXmax(htup);
}
/*
* If we found a DEAD tuple in the chain, adjust the HOT chain so that all
* the DEAD tuples at the start of the chain are removed and the root line
* pointer is appropriately redirected.
*/
if (OffsetNumberIsValid(latestdead))
{
/*
* Mark as unused each intermediate item that we are able to remove
* from the chain.
*
* When the previous item is the last dead tuple seen, we are at the
* right candidate for redirection.
*/
for (i = 1; (i < nchain) && (chainitems[i - 1] != latestdead); i++)
{
heap_prune_record_unused(prstate, chainitems[i]);
ndeleted++;
}
/*
* If the root entry had been a normal tuple, we are deleting it, so
* count it in the result. But changing a redirect (even to DEAD
* state) doesn't count.
*/
if (ItemIdIsNormal(rootlp))
ndeleted++;
/*
* If the DEAD tuple is at the end of the chain, the entire chain is
* dead and the root line pointer can be marked dead. Otherwise just
* redirect the root to the correct chain member.
*/
if (i >= nchain)
heap_prune_record_dead(prstate, rootoffnum);
else
{
heap_prune_record_redirect(prstate, rootoffnum, chainitems[i]);
/* If the redirection will be a move, need more processing */
if (redirect_move)
redirect_target = chainitems[i];
}
}
else if (nchain < 2 && ItemIdIsRedirected(rootlp))
{
/*
* We found a redirect item that doesn't point to a valid follow-on
* item. This can happen if the loop in heap_page_prune caused us to
* visit the dead successor of a redirect item before visiting the
* redirect item. We can clean up by setting the redirect item to
* DEAD state.
*/
heap_prune_record_dead(prstate, rootoffnum);
}
else if (redirect_move && ItemIdIsRedirected(rootlp))
{
/*
* If we desire to eliminate LP_REDIRECT items by moving tuples, make
* a redirection entry for each redirected root item; this will cause
* heap_page_prune_execute to actually do the move. (We get here only
* when there are no DEAD tuples in the chain; otherwise the
* redirection entry was made above.)
*/
heap_prune_record_redirect(prstate, rootoffnum, chainitems[1]);
redirect_target = chainitems[1];
}
/*
* If we are going to implement a redirect by moving tuples, we have to
* issue a cache invalidation against the redirection target tuple,
* because its CTID will be effectively changed by the move. Note that
* CacheInvalidateHeapTuple only queues the request, it doesn't send it;
* if we fail before reaching EndNonTransactionalInvalidation, nothing
* happens and no harm is done.
*/
if (OffsetNumberIsValid(redirect_target))
{
ItemId firstlp = PageGetItemId(dp, redirect_target);
HeapTupleData firsttup;
Assert(ItemIdIsNormal(firstlp));
/* Set up firsttup to reference the tuple at its existing CTID */
firsttup.t_data = (HeapTupleHeader) PageGetItem(dp, firstlp);
firsttup.t_len = ItemIdGetLength(firstlp);
ItemPointerSet(&firsttup.t_self,
BufferGetBlockNumber(buffer),
redirect_target);
CacheInvalidateHeapTuple(relation, &firsttup);
}
return ndeleted;
}
/*
* 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;
}
}
/*
* 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;
bool extended;
Assert(newsz == MAXALIGN(newsz));
/* If the item is oversized, don't bother. */
if (newsz > BrinMaxItemSize)
{
ereport(ERROR,
(errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED),
errmsg("index row size %lu exceeds maximum %lu for index \"%s\"",
(unsigned long) newsz,
(unsigned long) BrinMaxItemSize,
RelationGetRelationName(idxrel))));
return false; /* keep compiler quiet */
}
/* 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);
if (!BufferIsValid(newbuf))
{
Assert(!extended);
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)
{
Assert(!extended);
newbuf = InvalidBuffer;
}
}
else
{
LockBuffer(oldbuf, BUFFER_LOCK_EXCLUSIVE);
newbuf = InvalidBuffer;
extended = false;
}
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 this happens, and the new buffer was obtained by extending the
* relation, then we need to ensure we don't leave it uninitialized or
* forget about it.
*/
if (BufferIsValid(newbuf))
{
if (extended)
brin_initialize_empty_new_buffer(idxrel, newbuf);
UnlockReleaseBuffer(newbuf);
if (extended)
FreeSpaceMapVacuum(idxrel);
}
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))
{
if (extended)
brin_initialize_empty_new_buffer(idxrel, newbuf);
UnlockReleaseBuffer(newbuf);
if (extended)
FreeSpaceMapVacuum(idxrel);
}
return false;
}
/*
* 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 (((BrinPageFlags(oldpage) & BRIN_EVACUATE_PAGE) == 0) &&
brin_can_do_samepage_update(oldbuf, origsz, newsz))
{
if (BufferIsValid(newbuf))
{
/* as above */
if (extended)
brin_initialize_empty_new_buffer(idxrel, newbuf);
UnlockReleaseBuffer(newbuf);
}
START_CRIT_SECTION();
if (!PageIndexTupleOverwrite(oldpage, oldoff, (Item) newtup, newsz))
elog(ERROR, "failed to replace 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);
if (extended)
FreeSpaceMapVacuum(idxrel);
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;
BlockNumber newblk = InvalidBlockNumber;
Size freespace = 0;
revmapbuf = brinLockRevmapPageForUpdate(revmap, heapBlk);
START_CRIT_SECTION();
/*
* We need to initialize the page if it's newly obtained. Note we
* will WAL-log the initialization as part of the update, so we don't
* need to do that here.
*/
if (extended)
brin_page_init(BufferGetPage(newbuf), BRIN_PAGETYPE_REGULAR);
PageIndexTupleDeleteNoCompact(oldpage, oldoff);
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);
/* needed to update FSM below */
if (extended)
{
newblk = BufferGetBlockNumber(newbuf);
freespace = br_page_get_freespace(newpage);
}
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, 0);
/* 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);
if (extended)
{
Assert(BlockNumberIsValid(newblk));
RecordPageWithFreeSpace(idxrel, newblk, freespace);
FreeSpaceMapVacuum(idxrel);
}
return true;
}
}
static const char *
convert_gpdb4_heap_page(char *page)
{
VERSION4_PageHeaderData *oldhdr;
PageHeader newhdr;
OffsetNumber off;
OffsetNumber maxoff;
if (PageGetPageSize(page) != BLCKSZ)
return "invalid block size on page";
/* Can only convert from GPDB4 format */
if (PageGetPageLayoutVersion(page) != 4)
return "invalid page version";
oldhdr = (VERSION4_PageHeaderData *) page;
/* Other checks that PageHeaderIsValid() normally performs */
if (!((oldhdr->pd_flags & ~PD_VALID_FLAG_BITS) == 0 &&
oldhdr->pd_lower >= SizeOfPageHeaderData &&
oldhdr->pd_lower <= oldhdr->pd_upper &&
oldhdr->pd_upper <= oldhdr->pd_special &&
oldhdr->pd_special <= BLCKSZ &&
oldhdr->pd_special == MAXALIGN(oldhdr->pd_special)))
return "invalid header";
/*
* Ok, it's a valid heap page, from GPDB4. We know how to convert that!
*/
/* 1. pd_prune_xid was added to page header */
/* 2. Line pointer flags were changed */
/* 3. HEAP_COMPRESSED flag was removed */
/* 4. On-disk representation of numerics was changed */
/*
* Also, money datatype was widened from 32 to 64 bits. (pg_upgrade
* should've refused the upgrade)
*/
/*
* First, check if there is enough space on the page, after we expand the header.
*/
oldhdr = (VERSION4_PageHeaderData *) page;
newhdr = (PageHeader) page;
/*
* If there isn't enough space on this page for the new header field, relocate a tuple
*/
make_room(page);
/*
* There is space. Move the line pointers. We also convert the line pointer flags
* while we're at it. Begin from end to beginning, so that we don't overwrite items
* we haven't processed yet.
*/
maxoff = (oldhdr->pd_lower - VERSION4_SizeOfPageHeaderData) / sizeof(ItemIdData); /* PageGetMaxOffsetNumber */
for (off = maxoff; off >= 1; off--)
{
ItemIdData iid = oldhdr->pd_linp[off - 1]; /* PageGetItemId */
if (iid.lp_flags == VERSION4_LP_UNUSED)
iid.lp_flags = LP_UNUSED;
else if (iid.lp_flags == VERSION4_LP_USED)
iid.lp_flags = LP_NORMAL;
else
{
/* LP_DELETE and LP_USED were never used on heap pages. */
return "unexpected LP_DELETE or LP_DEAD line pointer on old-format heap page";
}
newhdr->pd_linp[off - 1] = iid;
}
newhdr->pd_lower = (char *) &newhdr->pd_linp[maxoff] - (char *) page;
/* Initialize the field that was added after version 4 format */
newhdr->pd_prune_xid = 0;
/*
* Ok, the page header and line pointers are in the new format now. Mangle
* the tuples themselves
*/
for (off = 1; off <= maxoff; off++)
{
ItemId iid = PageGetItemId(page, off); /* we can use PageGetItemId now */
HeapTupleHeader htup;
if (!ItemIdIsNormal(iid))
continue;
htup = (HeapTupleHeader) PageGetItem(page, iid);
convert_heaptuple(htup);
}
/*
* Finally, change the version number.
*/
PageSetPageSizeAndVersion(page, BLCKSZ, TARGET_PAGE_VERSION);
return NULL;
}
/* ----------------
* index_getnext - get the next heap tuple from a scan
*
* The result is the next heap tuple satisfying the scan keys and the
* snapshot, or NULL if no more matching tuples exist. On success,
* the buffer containing the heap tuple is pinned (the pin will be dropped
* at the next index_getnext or index_endscan).
*
* Note: caller must check scan->xs_recheck, and perform rechecking of the
* scan keys if required. We do not do that here because we don't have
* enough information to do it efficiently in the general case.
* ----------------
*/
HeapTuple
index_getnext(IndexScanDesc scan, ScanDirection direction)
{
HeapTuple heapTuple = &scan->xs_ctup;
ItemPointer tid = &heapTuple->t_self;
FmgrInfo *procedure;
SCAN_CHECKS;
GET_SCAN_PROCEDURE(amgettuple);
Assert(TransactionIdIsValid(RecentGlobalXmin));
/*
* We always reset xs_hot_dead; if we are here then either we are just
* starting the scan, or we previously returned a visible tuple, and in
* either case it's inappropriate to kill the prior index entry.
*/
scan->xs_hot_dead = false;
for (;;)
{
OffsetNumber offnum;
bool at_chain_start;
Page dp;
if (scan->xs_next_hot != InvalidOffsetNumber)
{
/*
* We are resuming scan of a HOT chain after having returned an
* earlier member. Must still hold pin on current heap page.
*/
Assert(BufferIsValid(scan->xs_cbuf));
Assert(ItemPointerGetBlockNumber(tid) ==
BufferGetBlockNumber(scan->xs_cbuf));
Assert(TransactionIdIsValid(scan->xs_prev_xmax));
offnum = scan->xs_next_hot;
at_chain_start = false;
scan->xs_next_hot = InvalidOffsetNumber;
}
else
{
bool found;
Buffer prev_buf;
/*
* If we scanned a whole HOT chain and found only dead tuples,
* tell index AM to kill its entry for that TID. We do not do this
* when in recovery because it may violate MVCC to do so. see
* comments in RelationGetIndexScan().
*/
if (!scan->xactStartedInRecovery)
scan->kill_prior_tuple = scan->xs_hot_dead;
/*
* The AM's gettuple proc finds the next index entry matching the
* scan keys, and puts the TID in xs_ctup.t_self (ie, *tid). It
* should also set scan->xs_recheck, though we pay no attention to
* that here.
*/
found = DatumGetBool(FunctionCall2(procedure,
PointerGetDatum(scan),
Int32GetDatum(direction)));
/* Reset kill flag immediately for safety */
scan->kill_prior_tuple = false;
/* If we're out of index entries, break out of outer loop */
if (!found)
break;
pgstat_count_index_tuples(scan->indexRelation, 1);
/* Switch to correct buffer if we don't have it already */
prev_buf = scan->xs_cbuf;
scan->xs_cbuf = ReleaseAndReadBuffer(scan->xs_cbuf,
scan->heapRelation,
ItemPointerGetBlockNumber(tid));
/*
* Prune page, but only if we weren't already on this page
*/
if (prev_buf != scan->xs_cbuf)
heap_page_prune_opt(scan->heapRelation, scan->xs_cbuf,
RecentGlobalXmin);
/* Prepare to scan HOT chain starting at index-referenced offnum */
offnum = ItemPointerGetOffsetNumber(tid);
at_chain_start = true;
/* We don't know what the first tuple's xmin should be */
scan->xs_prev_xmax = InvalidTransactionId;
/* Initialize flag to detect if all entries are dead */
scan->xs_hot_dead = true;
}
/* Obtain share-lock on the buffer so we can examine visibility */
LockBuffer(scan->xs_cbuf, BUFFER_LOCK_SHARE);
dp = (Page) BufferGetPage(scan->xs_cbuf);
/* Scan through possible multiple members of HOT-chain */
for (;;)
{
ItemId lp;
ItemPointer ctid;
bool valid;
/* check for bogus TID */
if (offnum < FirstOffsetNumber ||
offnum > PageGetMaxOffsetNumber(dp))
break;
lp = PageGetItemId(dp, offnum);
/* check for unused, dead, or redirected items */
if (!ItemIdIsNormal(lp))
{
/* We should only see a redirect at start of chain */
if (ItemIdIsRedirected(lp) && at_chain_start)
{
/* Follow the redirect */
offnum = ItemIdGetRedirect(lp);
at_chain_start = false;
continue;
}
/* else must be end of chain */
break;
}
/*
* We must initialize all of *heapTuple (ie, scan->xs_ctup) since
* it is returned to the executor on success.
*/
heapTuple->t_data = (HeapTupleHeader) PageGetItem(dp, lp);
heapTuple->t_len = ItemIdGetLength(lp);
ItemPointerSetOffsetNumber(tid, offnum);
heapTuple->t_tableOid = RelationGetRelid(scan->heapRelation);
ctid = &heapTuple->t_data->t_ctid;
/*
* Shouldn't see a HEAP_ONLY tuple at chain start. (This test
* should be unnecessary, since the chain root can't be removed
* while we have pin on the index entry, but let's make it
* anyway.)
*/
if (at_chain_start && HeapTupleIsHeapOnly(heapTuple))
break;
/*
* The xmin should match the previous xmax value, else chain is
* broken. (Note: this test is not optional because it protects
* us against the case where the prior chain member's xmax aborted
* since we looked at it.)
*/
if (TransactionIdIsValid(scan->xs_prev_xmax) &&
!TransactionIdEquals(scan->xs_prev_xmax,
HeapTupleHeaderGetXmin(heapTuple->t_data)))
break;
/* If it's visible per the snapshot, we must return it */
valid = HeapTupleSatisfiesVisibility(heapTuple, scan->xs_snapshot,
scan->xs_cbuf);
CheckForSerializableConflictOut(valid, scan->heapRelation,
heapTuple, scan->xs_cbuf);
if (valid)
{
/*
* If the snapshot is MVCC, we know that it could accept at
* most one member of the HOT chain, so we can skip examining
* any more members. Otherwise, check for continuation of the
* HOT-chain, and set state for next time.
*/
if (IsMVCCSnapshot(scan->xs_snapshot)
&& !IsolationIsSerializable())
scan->xs_next_hot = InvalidOffsetNumber;
else if (HeapTupleIsHotUpdated(heapTuple))
{
Assert(ItemPointerGetBlockNumber(ctid) ==
ItemPointerGetBlockNumber(tid));
scan->xs_next_hot = ItemPointerGetOffsetNumber(ctid);
scan->xs_prev_xmax = HeapTupleHeaderGetXmax(heapTuple->t_data);
}
else
scan->xs_next_hot = InvalidOffsetNumber;
PredicateLockTuple(scan->heapRelation, heapTuple);
LockBuffer(scan->xs_cbuf, BUFFER_LOCK_UNLOCK);
pgstat_count_heap_fetch(scan->indexRelation);
return heapTuple;
}
/*
* If we can't see it, maybe no one else can either. Check to see
* if the tuple is dead to all transactions. If we find that all
* the tuples in the HOT chain are dead, we'll signal the index AM
* to not return that TID on future indexscans.
*/
if (scan->xs_hot_dead &&
HeapTupleSatisfiesVacuum(heapTuple->t_data, RecentGlobalXmin,
scan->xs_cbuf) != HEAPTUPLE_DEAD)
scan->xs_hot_dead = false;
/*
* Check to see if HOT chain continues past this tuple; if so
* fetch the next offnum (we don't bother storing it into
* xs_next_hot, but must store xs_prev_xmax), and loop around.
*/
if (HeapTupleIsHotUpdated(heapTuple))
{
Assert(ItemPointerGetBlockNumber(ctid) ==
ItemPointerGetBlockNumber(tid));
offnum = ItemPointerGetOffsetNumber(ctid);
at_chain_start = false;
scan->xs_prev_xmax = HeapTupleHeaderGetXmax(heapTuple->t_data);
}
else
break; /* end of chain */
} /* loop over a single HOT chain */
LockBuffer(scan->xs_cbuf, BUFFER_LOCK_UNLOCK);
/* Loop around to ask index AM for another TID */
scan->xs_next_hot = InvalidOffsetNumber;
}
/* Release any held pin on a heap page */
if (BufferIsValid(scan->xs_cbuf))
{
ReleaseBuffer(scan->xs_cbuf);
scan->xs_cbuf = InvalidBuffer;
}
return NULL; /* failure exit */
}
/**
* This method estimates the number of tuples and pages in a heaptable relation. Getting the number of blocks is straightforward.
* Estimating the number of tuples is a little trickier. There are two factors that complicate this:
* 1. Tuples may be of variable length.
* 2. There may be dead tuples lying around.
* To do this, it chooses a certain number of blocks (as determined by a guc) randomly. The process of choosing is not strictly
* uniformly random since we have a target number of blocks in mind. We start processing blocks in order and choose an block
* with a probability p determined by the ratio of target to total blocks. It is possible that we get really unlucky and reject
* a large number of blocks up front. We compensate for this by increasing p dynamically. Thus, we are guaranteed to choose the target number
* of blocks. We read all heaptuples from these blocks and keep count of number of live tuples. We scale up this count to
* estimate reltuples. Relpages is an exact value.
*
* Input:
* rel - Relation. Must be a heaptable.
*
* Output:
* reltuples - estimated number of tuples in relation.
* relpages - exact number of pages.
*/
static void gp_statistics_estimate_reltuples_relpages_heap(Relation rel, float4 *reltuples, float4 *relpages)
{
MIRROREDLOCK_BUFMGR_DECLARE;
float4 nrowsseen = 0; /* # rows seen (including dead rows) */
float4 nrowsdead = 0; /* # rows dead */
float4 totalEmptyPages = 0; /* # of empty pages with only dead rows */
float4 totalSamplePages = 0; /* # of pages sampled */
BlockNumber nblockstotal = 0; /* nblocks in relation */
BlockNumber nblockstarget = (BlockNumber) gp_statistics_blocks_target;
BlockNumber nblocksseen = 0;
int j = 0; /* counter */
/**
* Ensure that the right kind of relation with the right kind of storage is passed to us.
*/
Assert(rel->rd_rel->relkind == RELKIND_RELATION);
Assert(RelationIsHeap(rel));
nblockstotal = RelationGetNumberOfBlocks(rel);
if (nblockstotal == 0 || nblockstarget == 0)
{
/**
* If there are no blocks, there cannot be tuples.
*/
*reltuples = 0.0;
*relpages = 0.0;
return;
}
for (j=0 ; j<nblockstotal; j++)
{
/**
* Threshold is dynamically adjusted based on how many blocks we need to examine and how many blocks
* are left.
*/
double threshold = ((double) nblockstarget - nblocksseen)/((double) nblockstotal - j);
/**
* Random dice thrown to determine if current block is chosen.
*/
double diceValue = ((double) random()) / ((double) MAX_RANDOM_VALUE);
if (threshold >= 1.0 || diceValue <= threshold)
{
totalSamplePages++;
/**
* Block j shall be examined!
*/
BlockNumber targblock = j;
Buffer targbuffer;
Page targpage;
OffsetNumber targoffset,
maxoffset;
/**
* Check for cancellations.
*/
CHECK_FOR_INTERRUPTS();
/*
* We must maintain a pin on the target page's buffer to ensure that
* the maxoffset value stays good (else concurrent VACUUM might delete
* tuples out from under us). Hence, pin the page until we are done
* looking at it. We don't maintain a lock on the page, so tuples
* could get added to it, but we ignore such tuples.
*/
// -------- MirroredLock ----------
MIRROREDLOCK_BUFMGR_LOCK;
targbuffer = ReadBuffer(rel, targblock);
LockBuffer(targbuffer, BUFFER_LOCK_SHARE);
targpage = BufferGetPage(targbuffer);
maxoffset = PageGetMaxOffsetNumber(targpage);
/* Figure out overall nrowsdead/nrowsseen ratio */
/* Figure out # of empty pages based on page level #rowsseen and #rowsdead.*/
float4 pageRowsSeen = 0.0;
float4 pageRowsDead = 0.0;
/* Inner loop over all tuples on the selected block. */
for (targoffset = FirstOffsetNumber; targoffset <= maxoffset; targoffset++)
{
ItemId itemid;
itemid = PageGetItemId(targpage, targoffset);
nrowsseen++;
pageRowsSeen++;
if(!ItemIdIsNormal(itemid))
{
nrowsdead += 1;
pageRowsDead++;
}
else
{
HeapTupleData targtuple;
ItemPointerSet(&targtuple.t_self, targblock, targoffset);
targtuple.t_data = (HeapTupleHeader) PageGetItem(targpage, itemid);
targtuple.t_len = ItemIdGetLength(itemid);
if(!HeapTupleSatisfiesVisibility(rel, &targtuple, SnapshotNow, targbuffer))
{
nrowsdead += 1;
pageRowsDead++;
}
}
}
/* Now release the pin on the page */
UnlockReleaseBuffer(targbuffer);
MIRROREDLOCK_BUFMGR_UNLOCK;
// -------- MirroredLock ----------
/* detect empty pages: pageRowsSeen == pageRowsDead, also log the nrowsseen (total) and nrowsdead (total) */
if (pageRowsSeen == pageRowsDead && pageRowsSeen > 0)
{
totalEmptyPages++;
}
nblocksseen++;
}
}
Assert(nblocksseen > 0);
/**
* To calculate reltuples, scale up the number of live rows per block seen to the total number
* of blocks.
*/
*reltuples = ceil((nrowsseen - nrowsdead) * nblockstotal / nblocksseen);
*relpages = nblockstotal;
if (totalSamplePages * 0.5 <= totalEmptyPages && totalSamplePages != 0)
{
/*
* LOG empty pages of bloated table for each segments.
*/
elog(DEBUG1, "ANALYZE detected 50%% or more empty pages (%f empty out of %f pages), please run VACUUM FULL for accurate estimation.", totalEmptyPages, totalSamplePages);
}
return;
}
/* ----------------------------------------------------------------
* BitmapHeapNext
*
* Retrieve next tuple from the BitmapHeapScan node's currentRelation
* ----------------------------------------------------------------
*/
static TupleTableSlot *
BitmapHeapNext(BitmapHeapScanState *node)
{
EState *estate;
ExprContext *econtext;
HeapScanDesc scan;
Index scanrelid;
TIDBitmap *tbm;
TBMIterateResult *tbmres;
OffsetNumber targoffset;
TupleTableSlot *slot;
/*
* extract necessary information from index scan node
*/
estate = node->ss.ps.state;
econtext = node->ss.ps.ps_ExprContext;
slot = node->ss.ss_ScanTupleSlot;
scan = node->ss.ss_currentScanDesc;
scanrelid = ((BitmapHeapScan *) node->ss.ps.plan)->scan.scanrelid;
tbm = node->tbm;
tbmres = node->tbmres;
/*
* 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])
return ExecClearTuple(slot);
ExecStoreTuple(estate->es_evTuple[scanrelid - 1],
slot, InvalidBuffer, false);
/* Does the tuple meet the original qual conditions? */
econtext->ecxt_scantuple = slot;
ResetExprContext(econtext);
if (!ExecQual(node->bitmapqualorig, econtext, false))
ExecClearTuple(slot); /* would not be returned by scan */
/* Flag for the next call that no more tuples */
estate->es_evTupleNull[scanrelid - 1] = true;
return slot;
}
/*
* If we haven't yet performed the underlying index scan, do it, and
* prepare the bitmap to be iterated over.
*/
if (tbm == NULL)
{
tbm = (TIDBitmap *) MultiExecProcNode(outerPlanState(node));
if (!tbm || !IsA(tbm, TIDBitmap))
elog(ERROR, "unrecognized result from subplan");
node->tbm = tbm;
node->tbmres = tbmres = NULL;
tbm_begin_iterate(tbm);
}
for (;;)
{
Page dp;
ItemId lp;
/*
* Get next page of results if needed
*/
if (tbmres == NULL)
{
node->tbmres = tbmres = tbm_iterate(tbm);
if (tbmres == NULL)
{
/* no more entries in the bitmap */
break;
}
/*
* Ignore any claimed entries past what we think is the end of the
* relation. (This is probably not necessary given that we got at
* least AccessShareLock on the table before performing any of the
* indexscans, but let's be safe.)
*/
if (tbmres->blockno >= scan->rs_nblocks)
{
node->tbmres = tbmres = NULL;
continue;
}
/*
* Fetch the current heap page and identify candidate tuples.
*/
bitgetpage(scan, tbmres);
/*
* Set rs_cindex to first slot to examine
*/
scan->rs_cindex = 0;
}
else
{
/*
* Continuing in previously obtained page; advance rs_cindex
*/
scan->rs_cindex++;
}
/*
* Out of range? If so, nothing more to look at on this page
*/
if (scan->rs_cindex < 0 || scan->rs_cindex >= scan->rs_ntuples)
{
node->tbmres = tbmres = NULL;
continue;
}
/*
* Okay to fetch the tuple
*/
targoffset = scan->rs_vistuples[scan->rs_cindex];
dp = (Page) BufferGetPage(scan->rs_cbuf);
lp = PageGetItemId(dp, targoffset);
Assert(ItemIdIsNormal(lp));
scan->rs_ctup.t_data = (HeapTupleHeader) PageGetItem((Page) dp, lp);
scan->rs_ctup.t_len = ItemIdGetLength(lp);
ItemPointerSet(&scan->rs_ctup.t_self, tbmres->blockno, targoffset);
pgstat_count_heap_fetch(scan->rs_rd);
/*
* Set up the result slot to point to this tuple. Note that the slot
* acquires a pin on the buffer.
*/
ExecStoreTuple(&scan->rs_ctup,
slot,
scan->rs_cbuf,
false);
/*
* If we are using lossy info, we have to recheck the qual conditions
* at every tuple.
*/
if (tbmres->ntuples < 0)
{
econtext->ecxt_scantuple = slot;
ResetExprContext(econtext);
if (!ExecQual(node->bitmapqualorig, econtext, false))
{
/* Fails recheck, so drop it and loop back for another */
ExecClearTuple(slot);
continue;
}
}
/* OK to return this tuple */
return slot;
}
/*
* if we get here it means we are at the end of the scan..
*/
return ExecClearTuple(slot);
}
/*
* This function takes an already open relation and scans its pages,
* skipping those that have the corresponding visibility map bit set.
* For pages we skip, we find the free space from the free space map
* and approximate tuple_len on that basis. For the others, we count
* the exact number of dead tuples etc.
*
* This scan is loosely based on vacuumlazy.c:lazy_scan_heap(), but
* we do not try to avoid skipping single pages.
*/
static void
statapprox_heap(Relation rel, output_type *stat)
{
BlockNumber scanned,
nblocks,
blkno;
Buffer vmbuffer = InvalidBuffer;
BufferAccessStrategy bstrategy;
TransactionId OldestXmin;
uint64 misc_count = 0;
OldestXmin = GetOldestXmin(rel, PROCARRAY_FLAGS_VACUUM);
bstrategy = GetAccessStrategy(BAS_BULKREAD);
nblocks = RelationGetNumberOfBlocks(rel);
scanned = 0;
for (blkno = 0; blkno < nblocks; blkno++)
{
Buffer buf;
Page page;
OffsetNumber offnum,
maxoff;
Size freespace;
CHECK_FOR_INTERRUPTS();
/*
* If the page has only visible tuples, then we can find out the free
* space from the FSM and move on.
*/
if (VM_ALL_VISIBLE(rel, blkno, &vmbuffer))
{
freespace = GetRecordedFreeSpace(rel, blkno);
stat->tuple_len += BLCKSZ - freespace;
stat->free_space += freespace;
continue;
}
buf = ReadBufferExtended(rel, MAIN_FORKNUM, blkno,
RBM_NORMAL, bstrategy);
LockBuffer(buf, BUFFER_LOCK_SHARE);
page = BufferGetPage(buf);
/*
* It's not safe to call PageGetHeapFreeSpace() on new pages, so we
* treat them as being free space for our purposes.
*/
if (!PageIsNew(page))
stat->free_space += PageGetHeapFreeSpace(page);
else
stat->free_space += BLCKSZ - SizeOfPageHeaderData;
if (PageIsNew(page) || PageIsEmpty(page))
{
UnlockReleaseBuffer(buf);
continue;
}
scanned++;
/*
* Look at each tuple on the page and decide whether it's live or
* dead, then count it and its size. Unlike lazy_scan_heap, we can
* afford to ignore problems and special cases.
*/
maxoff = PageGetMaxOffsetNumber(page);
for (offnum = FirstOffsetNumber;
offnum <= maxoff;
offnum = OffsetNumberNext(offnum))
{
ItemId itemid;
HeapTupleData tuple;
itemid = PageGetItemId(page, offnum);
if (!ItemIdIsUsed(itemid) || ItemIdIsRedirected(itemid) ||
ItemIdIsDead(itemid))
{
continue;
}
Assert(ItemIdIsNormal(itemid));
ItemPointerSet(&(tuple.t_self), blkno, offnum);
tuple.t_data = (HeapTupleHeader) PageGetItem(page, itemid);
tuple.t_len = ItemIdGetLength(itemid);
tuple.t_tableOid = RelationGetRelid(rel);
/*
* We count live and dead tuples, but we also need to add up
* others in order to feed vac_estimate_reltuples.
*/
switch (HeapTupleSatisfiesVacuum(&tuple, OldestXmin, buf))
{
case HEAPTUPLE_RECENTLY_DEAD:
misc_count++;
/* Fall through */
case HEAPTUPLE_DEAD:
stat->dead_tuple_len += tuple.t_len;
stat->dead_tuple_count++;
break;
case HEAPTUPLE_LIVE:
stat->tuple_len += tuple.t_len;
stat->tuple_count++;
break;
case HEAPTUPLE_INSERT_IN_PROGRESS:
case HEAPTUPLE_DELETE_IN_PROGRESS:
misc_count++;
break;
default:
elog(ERROR, "unexpected HeapTupleSatisfiesVacuum result");
break;
}
}
UnlockReleaseBuffer(buf);
}
stat->table_len = (uint64) nblocks *BLCKSZ;
stat->tuple_count = vac_estimate_reltuples(rel, false, nblocks, scanned,
stat->tuple_count + misc_count);
/*
* Calculate percentages if the relation has one or more pages.
*/
if (nblocks != 0)
{
stat->scanned_percent = 100 * scanned / nblocks;
stat->tuple_percent = 100.0 * stat->tuple_len / stat->table_len;
stat->dead_tuple_percent = 100.0 * stat->dead_tuple_len / stat->table_len;
stat->free_percent = 100.0 * stat->free_space / stat->table_len;
}
if (BufferIsValid(vmbuffer))
{
ReleaseBuffer(vmbuffer);
vmbuffer = InvalidBuffer;
}
}
/*
* 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, List *updated_stats)
{
MIRROREDLOCK_BUFMGR_DECLARE;
BlockNumber nblocks,
blkno;
HeapTupleData tuple;
char *relname;
BlockNumber empty_pages,
vacuumed_pages;
double num_tuples,
tups_vacuumed,
nkeep,
nunused;
IndexBulkDeleteResult **indstats;
int i;
int reindex_count = 1;
PGRUsage ru0;
/* Fetch gp_persistent_relation_node information that will be added to XLOG record. */
RelationFetchGpRelationNodeForXLog(onerel);
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->nonempty_pages = 0;
lazy_space_alloc(vacrelstats, nblocks);
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;
vacuum_delay_point();
/*
* 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)
{
/* Remove index entries */
for (i = 0; i < nindexes; i++)
lazy_vacuum_index(Irel[i], &indstats[i], vacrelstats);
reindex_count++;
/* Remove tuples from heap */
lazy_vacuum_heap(onerel, vacrelstats);
/* Forget the now-vacuumed tuples, and press on */
vacrelstats->num_dead_tuples = 0;
vacrelstats->num_index_scans++;
}
/* -------- MirroredLock ---------- */
MIRROREDLOCK_BUFMGR_LOCK;
buf = ReadBufferWithStrategy(onerel, blkno, 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);
MIRROREDLOCK_BUFMGR_UNLOCK;
/* -------- MirroredLock ---------- */
LockRelationForExtension(onerel, ExclusiveLock);
UnlockRelationForExtension(onerel, ExclusiveLock);
/* -------- MirroredLock ---------- */
MIRROREDLOCK_BUFMGR_LOCK;
LockBufferForCleanup(buf);
if (PageIsNew(page))
{
ereport(WARNING,
(errmsg("relation \"%s\" page %u is uninitialized --- fixing",
relname, blkno)));
PageInit(page, BufferGetPageSize(buf), 0);
/* must record in xlog so that changetracking will know about this change */
log_heap_newpage(onerel, page, blkno);
empty_pages++;
lazy_record_free_space(vacrelstats, blkno,
PageGetHeapFreeSpace(page));
}
MarkBufferDirty(buf);
UnlockReleaseBuffer(buf);
MIRROREDLOCK_BUFMGR_UNLOCK;
/* -------- MirroredLock ---------- */
continue;
}
if (PageIsEmpty(page))
{
empty_pages++;
lazy_record_free_space(vacrelstats, blkno,
PageGetHeapFreeSpace(page));
UnlockReleaseBuffer(buf);
MIRROREDLOCK_BUFMGR_UNLOCK;
/* -------- MirroredLock ---------- */
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, false);
/*
* Now scan the page to collect vacuumable items and check for tuples
* requiring freezing.
*/
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));
continue;
}
Assert(ItemIdIsNormal(itemid));
tuple.t_data = (HeapTupleHeader) PageGetItem(page, itemid);
tuple.t_len = ItemIdGetLength(itemid);
tupgone = false;
switch (HeapTupleSatisfiesVacuum(onerel, 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 */
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);
break;
case HEAPTUPLE_RECENTLY_DEAD:
/*
* If tuple is recently deleted then we must not remove it
* from relation.
*/
nkeep += 1;
break;
case HEAPTUPLE_INSERT_IN_PROGRESS:
/* This is an expected case during concurrent vacuum */
break;
case HEAPTUPLE_DELETE_IN_PROGRESS:
/* This is an expected case during concurrent vacuum */
break;
default:
elog(ERROR, "unexpected HeapTupleSatisfiesVacuum result");
break;
}
if (tupgone)
{
lazy_record_dead_tuple(vacrelstats, &(tuple.t_self));
tups_vacuumed += 1;
}
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, false))
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);
/* no XLOG for temp tables, though */
if (!onerel->rd_istemp)
{
XLogRecPtr recptr;
recptr = log_heap_freeze(onerel, buf, FreezeLimit,
frozen, nfrozen);
PageSetLSN(page, recptr);
}
}
/*
* 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);
/* Forget the now-vacuumed tuples, and press on */
vacrelstats->num_dead_tuples = 0;
vacuumed_pages++;
}
/*
* 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)
{
lazy_record_free_space(vacrelstats, blkno,
PageGetHeapFreeSpace(page));
}
/* Remember the location of the last page with nonremovable tuples */
if (hastup)
vacrelstats->nonempty_pages = blkno + 1;
UnlockReleaseBuffer(buf);
MIRROREDLOCK_BUFMGR_UNLOCK;
/* -------- MirroredLock ---------- */
}
/* save stats for use later */
vacrelstats->rel_tuples = num_tuples;
vacrelstats->tuples_deleted = tups_vacuumed;
/* 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)
{
/* Remove index entries */
for (i = 0; i < nindexes; i++)
lazy_vacuum_index(Irel[i], &indstats[i], vacrelstats);
reindex_count++;
/* 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, updated_stats);
/* 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 pages",
RelationGetRelationName(onerel),
tups_vacuumed, num_tuples, nblocks),
errdetail("%.0f dead row versions cannot be removed yet.\n"
"There were %.0f unused item pointers.\n"
"%u pages contain useful free space.\n"
"%u pages are entirely empty.\n"
"%s.",
nkeep,
nunused,
vacrelstats->tot_free_pages,
empty_pages,
pg_rusage_show(&ru0))));
}
/*
* 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);
/*
* 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);
/* 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++;
}
/* Release the pin on the visibility map page */
if (BufferIsValid(vmbuffer))
{
ReleaseBuffer(vmbuffer);
vmbuffer = InvalidBuffer;
}
/* 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))));
}
/* ----------------------------------------------------------------
* BitmapHeapNext
*
* Retrieve next tuple from the BitmapHeapScan node's currentRelation
* ----------------------------------------------------------------
*/
static TupleTableSlot *
BitmapHeapNext(BitmapHeapScanState *node)
{
ExprContext *econtext;
HeapScanDesc scan;
TIDBitmap *tbm;
TBMIterator *tbmiterator = NULL;
TBMSharedIterator *shared_tbmiterator = NULL;
TBMIterateResult *tbmres;
OffsetNumber targoffset;
TupleTableSlot *slot;
ParallelBitmapHeapState *pstate = node->pstate;
dsa_area *dsa = node->ss.ps.state->es_query_dsa;
/*
* extract necessary information from index scan node
*/
econtext = node->ss.ps.ps_ExprContext;
slot = node->ss.ss_ScanTupleSlot;
scan = node->ss.ss_currentScanDesc;
tbm = node->tbm;
if (pstate == NULL)
tbmiterator = node->tbmiterator;
else
shared_tbmiterator = node->shared_tbmiterator;
tbmres = node->tbmres;
/*
* If we haven't yet performed the underlying index scan, do it, and begin
* the iteration over the bitmap.
*
* For prefetching, we use *two* iterators, one for the pages we are
* actually scanning and another that runs ahead of the first for
* prefetching. node->prefetch_pages tracks exactly how many pages ahead
* the prefetch iterator is. Also, node->prefetch_target tracks the
* desired prefetch distance, which starts small and increases up to the
* node->prefetch_maximum. This is to avoid doing a lot of prefetching in
* a scan that stops after a few tuples because of a LIMIT.
*/
if (!node->initialized)
{
if (!pstate)
{
tbm = (TIDBitmap *) MultiExecProcNode(outerPlanState(node));
if (!tbm || !IsA(tbm, TIDBitmap))
elog(ERROR, "unrecognized result from subplan");
node->tbm = tbm;
node->tbmiterator = tbmiterator = tbm_begin_iterate(tbm);
node->tbmres = tbmres = NULL;
#ifdef USE_PREFETCH
if (node->prefetch_maximum > 0)
{
node->prefetch_iterator = tbm_begin_iterate(tbm);
node->prefetch_pages = 0;
node->prefetch_target = -1;
}
#endif /* USE_PREFETCH */
}
else
{
/*
* The leader will immediately come out of the function, but
* others will be blocked until leader populates the TBM and wakes
* them up.
*/
if (BitmapShouldInitializeSharedState(pstate))
{
tbm = (TIDBitmap *) MultiExecProcNode(outerPlanState(node));
if (!tbm || !IsA(tbm, TIDBitmap))
elog(ERROR, "unrecognized result from subplan");
node->tbm = tbm;
/*
* Prepare to iterate over the TBM. This will return the
* dsa_pointer of the iterator state which will be used by
* multiple processes to iterate jointly.
*/
pstate->tbmiterator = tbm_prepare_shared_iterate(tbm);
#ifdef USE_PREFETCH
if (node->prefetch_maximum > 0)
{
pstate->prefetch_iterator =
tbm_prepare_shared_iterate(tbm);
/*
* We don't need the mutex here as we haven't yet woke up
* others.
*/
pstate->prefetch_pages = 0;
pstate->prefetch_target = -1;
}
#endif
/* We have initialized the shared state so wake up others. */
BitmapDoneInitializingSharedState(pstate);
}
/* Allocate a private iterator and attach the shared state to it */
node->shared_tbmiterator = shared_tbmiterator =
tbm_attach_shared_iterate(dsa, pstate->tbmiterator);
node->tbmres = tbmres = NULL;
#ifdef USE_PREFETCH
if (node->prefetch_maximum > 0)
{
node->shared_prefetch_iterator =
tbm_attach_shared_iterate(dsa, pstate->prefetch_iterator);
}
#endif /* USE_PREFETCH */
}
node->initialized = true;
}
for (;;)
{
Page dp;
ItemId lp;
CHECK_FOR_INTERRUPTS();
/*
* Get next page of results if needed
*/
if (tbmres == NULL)
{
if (!pstate)
node->tbmres = tbmres = tbm_iterate(tbmiterator);
else
node->tbmres = tbmres = tbm_shared_iterate(shared_tbmiterator);
if (tbmres == NULL)
{
/* no more entries in the bitmap */
break;
}
BitmapAdjustPrefetchIterator(node, tbmres);
/*
* Ignore any claimed entries past what we think is the end of the
* relation. (This is probably not necessary given that we got at
* least AccessShareLock on the table before performing any of the
* indexscans, but let's be safe.)
*/
if (tbmres->blockno >= scan->rs_nblocks)
{
node->tbmres = tbmres = NULL;
continue;
}
/*
* Fetch the current heap page and identify candidate tuples.
*/
bitgetpage(scan, tbmres);
if (tbmres->ntuples >= 0)
node->exact_pages++;
else
node->lossy_pages++;
/*
* Set rs_cindex to first slot to examine
*/
scan->rs_cindex = 0;
/* Adjust the prefetch target */
BitmapAdjustPrefetchTarget(node);
}
else
{
/*
* Continuing in previously obtained page; advance rs_cindex
*/
scan->rs_cindex++;
#ifdef USE_PREFETCH
/*
* Try to prefetch at least a few pages even before we get to the
* second page if we don't stop reading after the first tuple.
*/
if (!pstate)
{
if (node->prefetch_target < node->prefetch_maximum)
node->prefetch_target++;
}
else if (pstate->prefetch_target < node->prefetch_maximum)
{
/* take spinlock while updating shared state */
SpinLockAcquire(&pstate->mutex);
if (pstate->prefetch_target < node->prefetch_maximum)
pstate->prefetch_target++;
SpinLockRelease(&pstate->mutex);
}
#endif /* USE_PREFETCH */
}
/*
* Out of range? If so, nothing more to look at on this page
*/
if (scan->rs_cindex < 0 || scan->rs_cindex >= scan->rs_ntuples)
{
node->tbmres = tbmres = NULL;
continue;
}
/*
* We issue prefetch requests *after* fetching the current page to try
* to avoid having prefetching interfere with the main I/O. Also, this
* should happen only when we have determined there is still something
* to do on the current page, else we may uselessly prefetch the same
* page we are just about to request for real.
*/
BitmapPrefetch(node, scan);
/*
* Okay to fetch the tuple
*/
targoffset = scan->rs_vistuples[scan->rs_cindex];
dp = (Page) BufferGetPage(scan->rs_cbuf);
lp = PageGetItemId(dp, targoffset);
Assert(ItemIdIsNormal(lp));
scan->rs_ctup.t_data = (HeapTupleHeader) PageGetItem((Page) dp, lp);
scan->rs_ctup.t_len = ItemIdGetLength(lp);
scan->rs_ctup.t_tableOid = scan->rs_rd->rd_id;
ItemPointerSet(&scan->rs_ctup.t_self, tbmres->blockno, targoffset);
pgstat_count_heap_fetch(scan->rs_rd);
/*
* Set up the result slot to point to this tuple. Note that the slot
* acquires a pin on the buffer.
*/
ExecStoreTuple(&scan->rs_ctup,
slot,
scan->rs_cbuf,
false);
/*
* If we are using lossy info, we have to recheck the qual conditions
* at every tuple.
*/
if (tbmres->recheck)
{
econtext->ecxt_scantuple = slot;
ResetExprContext(econtext);
if (!ExecQual(node->bitmapqualorig, econtext))
{
/* Fails recheck, so drop it and loop back for another */
InstrCountFiltered2(node, 1);
ExecClearTuple(slot);
continue;
}
}
/* OK to return this tuple */
return slot;
}
/*
* if we get here it means we are at the end of the scan..
*/
return ExecClearTuple(slot);
}
