/*
* Set value in given FSM page and slot.
*
* If minValue > 0, the updated page is also searched for a page with at
* least minValue of free space. If one is found, its slot number is
* returned, -1 otherwise.
*/
static int
fsm_set_and_search(Relation rel, FSMAddress addr, uint16 slot,
uint8 newValue, uint8 minValue)
{
Buffer buf;
Page page;
int newslot = -1;
buf = fsm_readbuf(rel, addr, true);
LockBuffer(buf, BUFFER_LOCK_EXCLUSIVE);
page = BufferGetPage(buf);
if (fsm_set_avail(page, slot, newValue))
MarkBufferDirtyHint(buf, false);
if (minValue != 0)
{
/* Search while we still hold the lock */
newslot = fsm_search_avail(buf, minValue,
addr.level == FSM_BOTTOM_LEVEL,
true);
}
UnlockReleaseBuffer(buf);
return newslot;
}
/*
* XLogRecordPageWithFreeSpace - like RecordPageWithFreeSpace, for use in
* WAL replay
*/
void
XLogRecordPageWithFreeSpace(RelFileNode rnode, BlockNumber heapBlk,
Size spaceAvail)
{
int new_cat = fsm_space_avail_to_cat(spaceAvail);
FSMAddress addr;
uint16 slot;
BlockNumber blkno;
Buffer buf;
Page page;
/* Get the location of the FSM byte representing the heap block */
addr = fsm_get_location(heapBlk, &slot);
blkno = fsm_logical_to_physical(addr);
/* If the page doesn't exist already, extend */
buf = XLogReadBufferExtended(rnode, FSM_FORKNUM, blkno, RBM_ZERO_ON_ERROR);
LockBuffer(buf, BUFFER_LOCK_EXCLUSIVE);
page = BufferGetPage(buf);
if (PageIsNew(page))
PageInit(page, BLCKSZ, 0);
if (fsm_set_avail(page, slot, new_cat))
MarkBufferDirtyHint(buf, false);
UnlockReleaseBuffer(buf);
}
/*
* Initiate page evacuation protocol.
*
* The page must be locked in exclusive mode by the caller.
*
* If the page is not yet initialized or empty, return false without doing
* anything; it can be used for revmap without any further changes. If it
* contains tuples, mark it for evacuation and return true.
*/
bool
brin_start_evacuating_page(Relation idxRel, Buffer buf)
{
OffsetNumber off;
OffsetNumber maxoff;
Page page;
page = BufferGetPage(buf);
if (PageIsNew(page))
return false;
maxoff = PageGetMaxOffsetNumber(page);
for (off = FirstOffsetNumber; off <= maxoff; off++)
{
ItemId lp;
lp = PageGetItemId(page, off);
if (ItemIdIsUsed(lp))
{
/* prevent other backends from adding more stuff to this page */
BrinPageFlags(page) |= BRIN_EVACUATE_PAGE;
MarkBufferDirtyHint(buf, true);
return true;
}
}
return false;
}
/*
* FreeSpaceMapTruncateRel - adjust for truncation of a relation.
*
* The caller must hold AccessExclusiveLock on the relation, to ensure that
* other backends receive the smgr invalidation event that this function sends
* before they access the FSM again.
*
* nblocks is the new___ size of the heap.
*/
void
FreeSpaceMapTruncateRel(Relation rel, BlockNumber nblocks)
{
BlockNumber new_nfsmblocks;
FSMAddress first_removed_address;
uint16 first_removed_slot;
Buffer buf;
RelationOpenSmgr(rel);
/*
* If no FSM has been created yet for this relation, there's nothing to
* truncate.
*/
if (!smgrexists(rel->rd_smgr, FSM_FORKNUM))
return;
/* Get the location in the FSM of the first removed heap block */
first_removed_address = fsm_get_location(nblocks, &first_removed_slot);
/*
* Zero out the tail of the last remaining FSM page. If the slot
* representing the first removed heap block is at a page boundary, as the
* first slot on the FSM page that first_removed_address points to, we can
* just truncate that page altogether.
*/
if (first_removed_slot > 0)
{
buf = fsm_readbuf(rel, first_removed_address, false);
if (!BufferIsValid(buf))
return; /* nothing to do; the FSM was already smaller */
LockBuffer(buf, BUFFER_LOCK_EXCLUSIVE);
fsm_truncate_avail(BufferGetPage(buf), first_removed_slot);
MarkBufferDirtyHint(buf, false);
UnlockReleaseBuffer(buf);
new_nfsmblocks = fsm_logical_to_physical(first_removed_address) + 1;
}
else
{
new_nfsmblocks = fsm_logical_to_physical(first_removed_address);
if (smgrnblocks(rel->rd_smgr, FSM_FORKNUM) <= new_nfsmblocks)
return; /* nothing to do; the FSM was already smaller */
}
/* Truncate the unused FSM pages, and send smgr inval message */
smgrtruncate(rel->rd_smgr, FSM_FORKNUM, new_nfsmblocks);
/*
* We might as well update the local smgr_fsm_nblocks setting.
* smgrtruncate sent an smgr cache inval message, which will cause other
* backends to invalidate their copy of smgr_fsm_nblocks, and this one too
* at the next command boundary. But this ensures it isn't outright wrong
* until then.
*/
if (rel->rd_smgr)
rel->rd_smgr->smgr_fsm_nblocks = new_nfsmblocks;
}
/*
* XLogRecordPageWithFreeSpace - like RecordPageWithFreeSpace, for use in
* WAL replay
*/
void
XLogRecordPageWithFreeSpace(RelFileNode rnode, BlockNumber heapBlk,
Size spaceAvail)
{
int new_cat = fsm_space_avail_to_cat(spaceAvail);
FSMAddress addr;
uint16 slot;
BlockNumber blkno;
Buffer buf;
Page page;
bool write_to_fsm;
/* This is meant to mirror the logic in fsm_allow_writes() */
if (heapBlk >= HEAP_FSM_CREATION_THRESHOLD)
write_to_fsm = true;
else
{
/* Open the relation at smgr level */
SMgrRelation smgr = smgropen(rnode, InvalidBackendId);
if (smgrexists(smgr, FSM_FORKNUM))
write_to_fsm = true;
else
{
BlockNumber heap_nblocks = smgrnblocks(smgr, MAIN_FORKNUM);
if (heap_nblocks > HEAP_FSM_CREATION_THRESHOLD)
write_to_fsm = true;
else
write_to_fsm = false;
}
}
if (!write_to_fsm)
return;
/* Get the location of the FSM byte representing the heap block */
addr = fsm_get_location(heapBlk, &slot);
blkno = fsm_logical_to_physical(addr);
/* If the page doesn't exist already, extend */
buf = XLogReadBufferExtended(rnode, FSM_FORKNUM, blkno, RBM_ZERO_ON_ERROR);
LockBuffer(buf, BUFFER_LOCK_EXCLUSIVE);
page = BufferGetPage(buf);
if (PageIsNew(page))
PageInit(page, BLCKSZ, 0);
if (fsm_set_avail(page, slot, new_cat))
MarkBufferDirtyHint(buf, false);
UnlockReleaseBuffer(buf);
}
/*
* 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;
}
/*
* SetHintBits()
*
* Set commit/abort hint bits on a tuple, if appropriate at this time.
*
* It is only safe to set a transaction-committed hint bit if we know the
* transaction's commit record has been flushed to disk, or if the table is
* temporary or unlogged and will be obliterated by a crash anyway. We
* cannot change the LSN of the page here because we may hold only a share
* lock on the buffer, so we can't use the LSN to interlock this; we have to
* just refrain from setting the hint bit until some future re-examination
* of the tuple.
*
* We can always set hint bits when marking a transaction aborted. (Some
* code in heapam.c relies on that!)
*
* Also, if we are cleaning up HEAP_MOVED_IN or HEAP_MOVED_OFF entries, then
* we can always set the hint bits, since pre-9.0 VACUUM FULL always used
* synchronous commits and didn't move tuples that weren't previously
* hinted. (This is not known by this subroutine, but is applied by its
* callers.) Note: old-style VACUUM FULL is gone, but we have to keep this
* module's support for MOVED_OFF/MOVED_IN flag bits for as long as we
* support in-place update from pre-9.0 databases.
*
* Normal commits may be asynchronous, so for those we need to get the LSN
* of the transaction and then check whether this is flushed.
*
* The caller should pass xid as the XID of the transaction to check, or
* InvalidTransactionId if no check is needed.
*/
static inline void
SetHintBits(HeapTupleHeader tuple, Buffer buffer,
uint16 infomask, TransactionId xid)
{
if (TransactionIdIsValid(xid))
{
/* NB: xid must be known committed here! */
XLogRecPtr commitLSN = TransactionIdGetCommitLSN(xid);
if (XLogNeedsFlush(commitLSN) && BufferIsPermanent(buffer))
return; /* not flushed yet, so don't set hint */
}
tuple->t_infomask |= infomask;
MarkBufferDirtyHint(buffer, true);
}
/*
* btvacuumpage --- VACUUM one page
*
* This processes a single page for btvacuumscan(). In some cases we
* must go back and re-examine previously-scanned pages; this routine
* recurses when necessary to handle that case.
*
* blkno is the page to process. orig_blkno is the highest block number
* reached by the outer btvacuumscan loop (the same as blkno, unless we
* are recursing to re-examine a previous page).
*/
static void
btvacuumpage(BTVacState *vstate, BlockNumber blkno, BlockNumber orig_blkno)
{
IndexVacuumInfo *info = vstate->info;
IndexBulkDeleteResult *stats = vstate->stats;
IndexBulkDeleteCallback callback = vstate->callback;
void *callback_state = vstate->callback_state;
Relation rel = info->index;
bool delete_now;
BlockNumber recurse_to;
Buffer buf;
Page page;
BTPageOpaque opaque = NULL;
restart:
delete_now = false;
recurse_to = P_NONE;
/* call vacuum_delay_point while not holding any buffer lock */
vacuum_delay_point();
/*
* We can't use _bt_getbuf() here because it always applies
* _bt_checkpage(), which will barf on an all-zero page. We want to
* recycle all-zero pages, not fail. Also, we want to use a nondefault
* buffer access strategy.
*/
buf = ReadBufferExtended(rel, MAIN_FORKNUM, blkno, RBM_NORMAL,
info->strategy);
LockBuffer(buf, BT_READ);
page = BufferGetPage(buf);
if (!PageIsNew(page))
{
_bt_checkpage(rel, buf);
opaque = (BTPageOpaque) PageGetSpecialPointer(page);
}
/*
* If we are recursing, the only case we want to do anything with is a
* live leaf page having the current vacuum cycle ID. Any other state
* implies we already saw the page (eg, deleted it as being empty).
*/
if (blkno != orig_blkno)
{
if (_bt_page_recyclable(page) ||
P_IGNORE(opaque) ||
!P_ISLEAF(opaque) ||
opaque->btpo_cycleid != vstate->cycleid)
{
_bt_relbuf(rel, buf);
return;
}
}
/* Page is valid, see what to do with it */
if (_bt_page_recyclable(page))
{
/* Okay to recycle this page */
RecordFreeIndexPage(rel, blkno);
vstate->totFreePages++;
stats->pages_deleted++;
}
else if (P_ISDELETED(opaque))
{
/* Already deleted, but can't recycle yet */
stats->pages_deleted++;
}
else if (P_ISHALFDEAD(opaque))
{
/* Half-dead, try to delete */
delete_now = true;
}
else if (P_ISLEAF(opaque))
{
OffsetNumber deletable[MaxOffsetNumber];
int ndeletable;
OffsetNumber offnum,
minoff,
maxoff;
/*
* Trade in the initial read lock for a super-exclusive write lock on
* this page. We must get such a lock on every leaf page over the
* course of the vacuum scan, whether or not it actually contains any
* deletable tuples --- see nbtree/README.
*/
LockBuffer(buf, BUFFER_LOCK_UNLOCK);
LockBufferForCleanup(buf);
/*
* Remember highest leaf page number we've taken cleanup lock on; see
* notes in btvacuumscan
*/
if (blkno > vstate->lastBlockLocked)
vstate->lastBlockLocked = blkno;
/*
* Check whether we need to recurse back to earlier pages. What we
* are concerned about is a page split that happened since we started
* the vacuum scan. If the split moved some tuples to a lower page
* then we might have missed 'em. If so, set up for tail recursion.
* (Must do this before possibly clearing btpo_cycleid below!)
*/
if (vstate->cycleid != 0 &&
opaque->btpo_cycleid == vstate->cycleid &&
!(opaque->btpo_flags & BTP_SPLIT_END) &&
!P_RIGHTMOST(opaque) &&
opaque->btpo_next < orig_blkno)
recurse_to = opaque->btpo_next;
/*
* Scan over all items to see which ones need deleted according to the
* callback function.
*/
ndeletable = 0;
minoff = P_FIRSTDATAKEY(opaque);
maxoff = PageGetMaxOffsetNumber(page);
if (callback)
{
for (offnum = minoff;
offnum <= maxoff;
offnum = OffsetNumberNext(offnum))
{
IndexTuple itup;
ItemPointer htup;
itup = (IndexTuple) PageGetItem(page,
PageGetItemId(page, offnum));
htup = &(itup->t_tid);
/*
* During Hot Standby we currently assume that
* XLOG_BTREE_VACUUM records do not produce conflicts. That is
* only true as long as the callback function depends only
* upon whether the index tuple refers to heap tuples removed
* in the initial heap scan. When vacuum starts it derives a
* value of OldestXmin. Backends taking later snapshots could
* have a RecentGlobalXmin with a later xid than the vacuum's
* OldestXmin, so it is possible that row versions deleted
* after OldestXmin could be marked as killed by other
* backends. The callback function *could* look at the index
* tuple state in isolation and decide to delete the index
* tuple, though currently it does not. If it ever did, we
* would need to reconsider whether XLOG_BTREE_VACUUM records
* should cause conflicts. If they did cause conflicts they
* would be fairly harsh conflicts, since we haven't yet
* worked out a way to pass a useful value for
* latestRemovedXid on the XLOG_BTREE_VACUUM records. This
* applies to *any* type of index that marks index tuples as
* killed.
*/
if (callback(htup, callback_state))
deletable[ndeletable++] = offnum;
}
}
/*
* Apply any needed deletes. We issue just one _bt_delitems_vacuum()
* call per page, so as to minimize WAL traffic.
*/
if (ndeletable > 0)
{
/*
* Notice that the issued XLOG_BTREE_VACUUM WAL record includes
* all information to the replay code to allow it to get a cleanup
* lock on all pages between the previous lastBlockVacuumed and
* this page. This ensures that WAL replay locks all leaf pages at
* some point, which is important should non-MVCC scans be
* requested. This is currently unused on standby, but we record
* it anyway, so that the WAL contains the required information.
*
* Since we can visit leaf pages out-of-order when recursing,
* replay might end up locking such pages an extra time, but it
* doesn't seem worth the amount of bookkeeping it'd take to avoid
* that.
*/
_bt_delitems_vacuum(rel, buf, deletable, ndeletable,
vstate->lastBlockVacuumed);
/*
* Remember highest leaf page number we've issued a
* XLOG_BTREE_VACUUM WAL record for.
*/
if (blkno > vstate->lastBlockVacuumed)
vstate->lastBlockVacuumed = blkno;
stats->tuples_removed += ndeletable;
/* must recompute maxoff */
maxoff = PageGetMaxOffsetNumber(page);
}
else
{
/*
* If the page has been split during this vacuum cycle, it seems
* worth expending a write to clear btpo_cycleid even if we don't
* have any deletions to do. (If we do, _bt_delitems_vacuum takes
* care of this.) This ensures we won't process the page again.
*
* We treat this like a hint-bit update because there's no need to
* WAL-log it.
*/
if (vstate->cycleid != 0 &&
opaque->btpo_cycleid == vstate->cycleid)
{
opaque->btpo_cycleid = 0;
MarkBufferDirtyHint(buf, true);
}
}
/*
* If it's now empty, try to delete; else count the live tuples. We
* don't delete when recursing, though, to avoid putting entries into
* freePages out-of-order (doesn't seem worth any extra code to handle
* the case).
*/
if (minoff > maxoff)
delete_now = (blkno == orig_blkno);
else
stats->num_index_tuples += maxoff - minoff + 1;
}
if (delete_now)
{
MemoryContext oldcontext;
int ndel;
/* Run pagedel in a temp context to avoid memory leakage */
MemoryContextReset(vstate->pagedelcontext);
oldcontext = MemoryContextSwitchTo(vstate->pagedelcontext);
ndel = _bt_pagedel(rel, buf);
/* count only this page, else may double-count parent */
if (ndel)
stats->pages_deleted++;
MemoryContextSwitchTo(oldcontext);
/* pagedel released buffer, so we shouldn't */
}
else
_bt_relbuf(rel, buf);
/*
* This is really tail recursion, but if the compiler is too stupid to
* optimize it as such, we'd eat an uncomfortably large amount of stack
* space per recursion level (due to the deletable[] array). A failure is
* improbable since the number of levels isn't likely to be large ... but
* just in case, let's hand-optimize into a loop.
*/
if (recurse_to != P_NONE)
{
blkno = recurse_to;
goto restart;
}
}
/*
* hashgettuple() -- Get the next tuple in the scan.
*/
bool
hashgettuple(IndexScanDesc scan, ScanDirection dir)
{
HashScanOpaque so = (HashScanOpaque) scan->opaque;
Relation rel = scan->indexRelation;
Buffer buf;
Page page;
OffsetNumber offnum;
ItemPointer current;
bool res;
/* Hash indexes are always lossy since we store only the hash code */
scan->xs_recheck = true;
/*
* We hold pin but not lock on current buffer while outside the hash AM.
* Reacquire the read lock here.
*/
if (BufferIsValid(so->hashso_curbuf))
_hash_chgbufaccess(rel, so->hashso_curbuf, HASH_NOLOCK, HASH_READ);
/*
* If we've already initialized this scan, we can just advance it in the
* appropriate direction. If we haven't done so yet, we call a routine to
* get the first item in the scan.
*/
current = &(so->hashso_curpos);
if (ItemPointerIsValid(current))
{
/*
* An insertion into the current index page could have happened while
* we didn't have read lock on it. Re-find our position by looking
* for the TID we previously returned. (Because we hold share lock on
* the bucket, no deletions or splits could have occurred; therefore
* we can expect that the TID still exists in the current index page,
* at an offset >= where we were.)
*/
OffsetNumber maxoffnum;
buf = so->hashso_curbuf;
Assert(BufferIsValid(buf));
page = BufferGetPage(buf);
TestForOldSnapshot(scan->xs_snapshot, rel, page);
maxoffnum = PageGetMaxOffsetNumber(page);
for (offnum = ItemPointerGetOffsetNumber(current);
offnum <= maxoffnum;
offnum = OffsetNumberNext(offnum))
{
IndexTuple itup;
itup = (IndexTuple) PageGetItem(page, PageGetItemId(page, offnum));
if (ItemPointerEquals(&(so->hashso_heappos), &(itup->t_tid)))
break;
}
if (offnum > maxoffnum)
elog(ERROR, "failed to re-find scan position within index \"%s\"",
RelationGetRelationName(rel));
ItemPointerSetOffsetNumber(current, offnum);
/*
* Check to see if we should kill the previously-fetched tuple.
*/
if (scan->kill_prior_tuple)
{
/*
* Yes, so mark it by setting the LP_DEAD state in the item flags.
*/
ItemIdMarkDead(PageGetItemId(page, offnum));
/*
* Since this can be redone later if needed, mark as a hint.
*/
MarkBufferDirtyHint(buf, true);
}
/*
* Now continue the scan.
*/
res = _hash_next(scan, dir);
}
else
res = _hash_first(scan, dir);
/*
* Skip killed tuples if asked to.
*/
if (scan->ignore_killed_tuples)
{
while (res)
{
offnum = ItemPointerGetOffsetNumber(current);
page = BufferGetPage(so->hashso_curbuf);
if (!ItemIdIsDead(PageGetItemId(page, offnum)))
break;
res = _hash_next(scan, dir);
}
}
/* Release read lock on current buffer, but keep it pinned */
if (BufferIsValid(so->hashso_curbuf))
_hash_chgbufaccess(rel, so->hashso_curbuf, HASH_READ, HASH_NOLOCK);
/* Return current heap TID on success */
scan->xs_ctup.t_self = so->hashso_heappos;
return res;
}
/*
* Recursive guts of FreeSpaceMapVacuum
*/
static uint8
fsm_vacuum_page(Relation rel, FSMAddress addr, bool *eof_p)
{
Buffer buf;
Page page;
uint8 max_avail;
/* Read the page if it exists, or return EOF */
buf = fsm_readbuf(rel, addr, false);
if (!BufferIsValid(buf))
{
*eof_p = true;
return 0;
}
else
*eof_p = false;
page = BufferGetPage(buf);
/*
* Recurse into children, and fix the information stored about them at
* this level.
*/
if (addr.level > FSM_BOTTOM_LEVEL)
{
int slot;
bool eof = false;
for (slot = 0; slot < SlotsPerFSMPage; slot++)
{
int child_avail;
CHECK_FOR_INTERRUPTS();
/* After we hit end-of-file, just clear the rest of the slots */
if (!eof)
child_avail = fsm_vacuum_page(rel, fsm_get_child(addr, slot), &eof);
else
child_avail = 0;
/* Update information about the child */
if (fsm_get_avail(page, slot) != child_avail)
{
LockBuffer(buf, BUFFER_LOCK_EXCLUSIVE);
fsm_set_avail(BufferGetPage(buf), slot, child_avail);
MarkBufferDirtyHint(buf, false);
LockBuffer(buf, BUFFER_LOCK_UNLOCK);
}
}
}
max_avail = fsm_get_max_avail(BufferGetPage(buf));
/*
* Reset the next slot pointer. This encourages the use of low-numbered
* pages, increasing the chances that a later vacuum can truncate the
* relation.
*/
((FSMPage) PageGetContents(page))->fp_next_slot = 0;
ReleaseBuffer(buf);
return max_avail;
}
/*
* Prune and repair fragmentation in the specified page.
*
* Caller must have pin and buffer cleanup lock on the page.
*
* OldestXmin is the cutoff XID used to distinguish whether tuples are DEAD
* or RECENTLY_DEAD (see HeapTupleSatisfiesVacuum).
*
* If redirect_move is set, we remove redirecting line pointers by
* updating the root line pointer to point directly to the first non-dead
* tuple in the chain. NOTE: eliminating the redirect changes the first
* tuple's effective CTID, and is therefore unsafe except within VACUUM FULL.
* The only reason we support this capability at all is that by using it,
* VACUUM FULL need not cope with LP_REDIRECT items at all; which seems a
* good thing since VACUUM FULL is overly complicated already.
*
* If report_stats is true then we send the number of reclaimed heap-only
* tuples to pgstats. (This must be FALSE during vacuum, since vacuum will
* send its own new total to pgstats, and we don't want this delta applied
* on top of that.)
*
* Returns the number of tuples deleted from the page.
*/
int
heap_page_prune(Relation relation, Buffer buffer, TransactionId OldestXmin,
bool redirect_move, bool report_stats)
{
int ndeleted = 0;
Page page = BufferGetPage(buffer);
OffsetNumber offnum,
maxoff;
PruneState prstate;
/*
* Our strategy is to scan the page and make lists of items to change,
* then apply the changes within a critical section. This keeps as much
* logic as possible out of the critical section, and also ensures that
* WAL replay will work the same as the normal case.
*
* First, inform inval.c that upcoming CacheInvalidateHeapTuple calls are
* nontransactional.
*/
if (redirect_move)
BeginNonTransactionalInvalidation();
/*
* Initialize the new pd_prune_xid value to zero (indicating no prunable
* tuples). If we find any tuples which may soon become prunable, we will
* save the lowest relevant XID in new_prune_xid. Also initialize the rest
* of our working state.
*/
prstate.new_prune_xid = InvalidTransactionId;
prstate.nredirected = prstate.ndead = prstate.nunused = 0;
memset(prstate.marked, 0, sizeof(prstate.marked));
/* Scan the page */
maxoff = PageGetMaxOffsetNumber(page);
for (offnum = FirstOffsetNumber;
offnum <= maxoff;
offnum = OffsetNumberNext(offnum))
{
ItemId itemid;
/* Ignore items already processed as part of an earlier chain */
if (prstate.marked[offnum])
continue;
/* Nothing to do if slot is empty or already dead */
itemid = PageGetItemId(page, offnum);
if (!ItemIdIsUsed(itemid) || ItemIdIsDead(itemid))
continue;
/* Process this item or chain of items */
ndeleted += heap_prune_chain(relation, buffer, offnum,
OldestXmin,
&prstate,
redirect_move);
}
/*
* Send invalidation messages for any tuples we are about to move. It is
* safe to do this now, even though we could theoretically still fail
* before making the actual page update, because a useless cache
* invalidation doesn't hurt anything. Also, no one else can reload the
* tuples while we have exclusive buffer lock, so it's not too early to
* send the invals. This avoids sending the invals while inside the
* critical section, which is a good thing for robustness.
*/
if (redirect_move)
EndNonTransactionalInvalidation();
/* Any error while applying the changes is critical */
START_CRIT_SECTION();
/* Have we found any prunable items? */
if (prstate.nredirected > 0 || prstate.ndead > 0 || prstate.nunused > 0)
{
/*
* Apply the planned item changes, then repair page fragmentation, and
* update the page's hint bit about whether it has free line pointers.
*/
heap_page_prune_execute(buffer,
prstate.redirected, prstate.nredirected,
prstate.nowdead, prstate.ndead,
prstate.nowunused, prstate.nunused,
redirect_move);
/*
* Update the page's pd_prune_xid field to either zero, or the lowest
* XID of any soon-prunable tuple.
*/
((PageHeader) page)->pd_prune_xid = prstate.new_prune_xid;
/*
* Also clear the "page is full" flag, since there's no point in
* repeating the prune/defrag process until something else happens to
* the page.
*/
PageClearFull(page);
MarkBufferDirty(buffer);
/*
* Emit a WAL HEAP_CLEAN or HEAP_CLEAN_MOVE record showing what we did
*/
if (!relation->rd_istemp)
{
XLogRecPtr recptr;
recptr = log_heap_clean(relation, buffer,
prstate.redirected, prstate.nredirected,
prstate.nowdead, prstate.ndead,
prstate.nowunused, prstate.nunused,
redirect_move);
PageSetLSN(BufferGetPage(buffer), recptr);
}
}
else
{
/*
* If we didn't prune anything, but have found a new value for the
* pd_prune_xid field, update it and mark the buffer dirty. This is
* treated as a non-WAL-logged hint.
*
* Also clear the "page is full" flag if it is set, since there's no
* point in repeating the prune/defrag process until something else
* happens to the page.
*/
if (((PageHeader) page)->pd_prune_xid != prstate.new_prune_xid ||
PageIsFull(page))
{
((PageHeader) page)->pd_prune_xid = prstate.new_prune_xid;
PageClearFull(page);
MarkBufferDirtyHint(buffer, relation);
}
}
END_CRIT_SECTION();
/*
* If requested, report the number of tuples reclaimed to pgstats. This is
* ndeleted minus ndead, because we don't want to count a now-DEAD root
* item as a deletion for this purpose.
*/
if (report_stats && ndeleted > prstate.ndead)
pgstat_update_heap_dead_tuples(relation, ndeleted - prstate.ndead);
/*
* XXX Should we update the FSM information of this page ?
*
* There are two schools of thought here. We may not want to update FSM
* information so that the page is not used for unrelated UPDATEs/INSERTs
* and any free space in this page will remain available for further
* UPDATEs in *this* page, thus improving chances for doing HOT updates.
*
* But for a large table and where a page does not receive further UPDATEs
* for a long time, we might waste this space by not updating the FSM
* information. The relation may get extended and fragmented further.
*
* One possibility is to leave "fillfactor" worth of space in this page
* and update FSM with the remaining space.
*
* In any case, the current FSM implementation doesn't accept
* one-page-at-a-time updates, so this is all academic for now.
*/
return ndeleted;
}
/*
* gistkillitems() -- set LP_DEAD state for items an indexscan caller has
* told us were killed.
*
* We re-read page here, so it's important to check page LSN. If the page
* has been modified since the last read (as determined by LSN), we cannot
* flag any entries because it is possible that the old entry was vacuumed
* away and the TID was re-used by a completely different heap tuple.
*/
static void
gistkillitems(IndexScanDesc scan)
{
GISTScanOpaque so = (GISTScanOpaque) scan->opaque;
Buffer buffer;
Page page;
OffsetNumber offnum;
ItemId iid;
int i;
bool killedsomething = false;
Assert(so->curBlkno != InvalidBlockNumber);
Assert(!XLogRecPtrIsInvalid(so->curPageLSN));
Assert(so->killedItems != NULL);
buffer = ReadBuffer(scan->indexRelation, so->curBlkno);
if (!BufferIsValid(buffer))
return;
LockBuffer(buffer, GIST_SHARE);
gistcheckpage(scan->indexRelation, buffer);
page = BufferGetPage(buffer);
/*
* If page LSN differs it means that the page was modified since the last
* read. killedItems could be not valid so LP_DEAD hints applying is not
* safe.
*/
if (BufferGetLSNAtomic(buffer) != so->curPageLSN)
{
UnlockReleaseBuffer(buffer);
so->numKilled = 0; /* reset counter */
return;
}
Assert(GistPageIsLeaf(page));
/*
* Mark all killedItems as dead. We need no additional recheck, because,
* if page was modified, pageLSN must have changed.
*/
for (i = 0; i < so->numKilled; i++)
{
offnum = so->killedItems[i];
iid = PageGetItemId(page, offnum);
ItemIdMarkDead(iid);
killedsomething = true;
}
if (killedsomething)
{
GistMarkPageHasGarbage(page);
MarkBufferDirtyHint(buffer, true);
}
UnlockReleaseBuffer(buffer);
/*
* Always reset the scan state, so we don't look for same items on other
* pages.
*/
so->numKilled = 0;
}
/*
* Recursive guts of FreeSpaceMapVacuum
*
* Examine the FSM page indicated by addr, as well as its children, updating
* upper-level nodes that cover the heap block range from start to end-1.
* (It's okay if end is beyond the actual end of the map.)
* Return the maximum freespace value on this page.
*
* If addr is past the end of the FSM, set *eof_p to true and return 0.
*
* This traverses the tree in depth-first order. The tree is stored
* physically in depth-first order, so this should be pretty I/O efficient.
*/
static uint8
fsm_vacuum_page(Relation rel, FSMAddress addr,
BlockNumber start, BlockNumber end,
bool *eof_p)
{
Buffer buf;
Page page;
uint8 max_avail;
/* Read the page if it exists, or return EOF */
buf = fsm_readbuf(rel, addr, false);
if (!BufferIsValid(buf))
{
*eof_p = true;
return 0;
}
else
*eof_p = false;
page = BufferGetPage(buf);
/*
* If we're above the bottom level, recurse into children, and fix the
* information stored about them at this level.
*/
if (addr.level > FSM_BOTTOM_LEVEL)
{
FSMAddress fsm_start,
fsm_end;
uint16 fsm_start_slot,
fsm_end_slot;
int slot,
start_slot,
end_slot;
bool eof = false;
/*
* Compute the range of slots we need to update on this page, given
* the requested range of heap blocks to consider. The first slot to
* update is the one covering the "start" block, and the last slot is
* the one covering "end - 1". (Some of this work will be duplicated
* in each recursive call, but it's cheap enough to not worry about.)
*/
fsm_start = fsm_get_location(start, &fsm_start_slot);
fsm_end = fsm_get_location(end - 1, &fsm_end_slot);
while (fsm_start.level < addr.level)
{
fsm_start = fsm_get_parent(fsm_start, &fsm_start_slot);
fsm_end = fsm_get_parent(fsm_end, &fsm_end_slot);
}
Assert(fsm_start.level == addr.level);
if (fsm_start.logpageno == addr.logpageno)
start_slot = fsm_start_slot;
else if (fsm_start.logpageno > addr.logpageno)
start_slot = SlotsPerFSMPage; /* shouldn't get here... */
else
start_slot = 0;
if (fsm_end.logpageno == addr.logpageno)
end_slot = fsm_end_slot;
else if (fsm_end.logpageno > addr.logpageno)
end_slot = SlotsPerFSMPage - 1;
else
end_slot = -1; /* shouldn't get here... */
for (slot = start_slot; slot <= end_slot; slot++)
{
int child_avail;
CHECK_FOR_INTERRUPTS();
/* After we hit end-of-file, just clear the rest of the slots */
if (!eof)
child_avail = fsm_vacuum_page(rel, fsm_get_child(addr, slot),
start, end,
&eof);
else
child_avail = 0;
/* Update information about the child */
if (fsm_get_avail(page, slot) != child_avail)
{
LockBuffer(buf, BUFFER_LOCK_EXCLUSIVE);
fsm_set_avail(page, slot, child_avail);
MarkBufferDirtyHint(buf, false);
LockBuffer(buf, BUFFER_LOCK_UNLOCK);
}
}
}
/* Now get the maximum value on the page, to return to caller */
max_avail = fsm_get_max_avail(page);
/*
* Reset the next slot pointer. This encourages the use of low-numbered
* pages, increasing the chances that a later vacuum can truncate the
* relation. We don't bother with a lock here, nor with marking the page
* dirty if it wasn't already, since this is just a hint.
*/
((FSMPage) PageGetContents(page))->fp_next_slot = 0;
ReleaseBuffer(buf);
return max_avail;
}
/*
* Searches for a slot with category at least minvalue.
* Returns slot number, or -1 if none found.
*
* The caller must hold at least a shared lock on the page, and this
* function can unlock and lock the page again in exclusive mode if it
* needs to be updated. exclusive_lock_held should be set to true if the
* caller is already holding an exclusive lock, to avoid extra work.
*
* If advancenext is false, fp_next_slot is set to point to the returned
* slot, and if it's true, to the slot after the returned slot.
*/
int
fsm_search_avail(Buffer buf, uint8 minvalue, bool advancenext,
bool exclusive_lock_held)
{
Page page = BufferGetPage(buf);
FSMPage fsmpage = (FSMPage) PageGetContents(page);
int nodeno;
int target;
uint16 slot;
restart:
/*
* Check the root first, and exit quickly if there's no leaf with enough
* free space
*/
if (fsmpage->fp_nodes[0] < minvalue)
return -1;
/*
* Start search using fp_next_slot. It's just a hint, so check that it's
* sane. (This also handles wrapping around when the prior call returned
* the last slot on the page.)
*/
target = fsmpage->fp_next_slot;
if (target < 0 || target >= LeafNodesPerPage)
target = 0;
target += NonLeafNodesPerPage;
/*----------
* Start the search from the target slot. At every step, move one
* node to the right, then climb up to the parent. Stop when we reach
* a node with enough free space (as we must, since the root has enough
* space).
*
* The idea is to gradually expand our "search triangle", that is, all
* nodes covered by the current node, and to be sure we search to the
* right from the start point. At the first step, only the target slot
* is examined. When we move up from a left child to its parent, we are
* adding the right-hand subtree of that parent to the search triangle.
* When we move right then up from a right child, we are dropping the
* current search triangle (which we know doesn't contain any suitable
* page) and instead looking at the next-larger-size triangle to its
* right. So we never look left from our original start point, and at
* each step the size of the search triangle doubles, ensuring it takes
* only log2(N) work to search N pages.
*
* The "move right" operation will wrap around if it hits the right edge
* of the tree, so the behavior is still good if we start near the right.
* Note also that the move-and-climb behavior ensures that we can't end
* up on one of the missing nodes at the right of the leaf level.
*
* For example, consider this tree:
*
* 7
* 7 6
* 5 7 6 5
* 4 5 5 7 2 6 5 2
* T
*
* Assume that the target node is the node indicated by the letter T,
* and we're searching for a node with value of 6 or higher. The search
* begins at T. At the first iteration, we move to the right, then to the
* parent, arriving at the rightmost 5. At the second iteration, we move
* to the right, wrapping around, then climb up, arriving at the 7 on the
* third level. 7 satisfies our search, so we descend down to the bottom,
* following the path of sevens. This is in fact the first suitable page
* to the right of (allowing for wraparound) our start point.
*----------
*/
nodeno = target;
while (nodeno > 0)
{
if (fsmpage->fp_nodes[nodeno] >= minvalue)
break;
/*
* Move to the right, wrapping around on same level if necessary, then
* climb up.
*/
nodeno = parentof(rightneighbor(nodeno));
}
/*
* We're now at a node with enough free space, somewhere in the middle of
* the tree. Descend to the bottom, following a path with enough free
* space, preferring to move left if there's a choice.
*/
while (nodeno < NonLeafNodesPerPage)
{
int childnodeno = leftchild(nodeno);
if (childnodeno < NodesPerPage &&
fsmpage->fp_nodes[childnodeno] >= minvalue)
{
nodeno = childnodeno;
continue;
}
childnodeno++; /* point to right child */
if (childnodeno < NodesPerPage &&
fsmpage->fp_nodes[childnodeno] >= minvalue)
{
nodeno = childnodeno;
}
else
{
/*
* Oops. The parent node promised that either left or right child
* has enough space, but neither actually did. This can happen in
* case of a "torn page", IOW if we crashed earlier while writing
* the page to disk, and only part of the page made it to disk.
*
* Fix the corruption and restart.
*/
RelFileNode rnode;
ForkNumber forknum;
BlockNumber blknum;
BufferGetTag(buf, &rnode, &forknum, &blknum);
elog(DEBUG1, "fixing corrupt FSM block %u, relation %u/%u/%u",
blknum, rnode.spcNode, rnode.dbNode, rnode.relNode);
/* make sure we hold an exclusive lock */
if (!exclusive_lock_held)
{
LockBuffer(buf, BUFFER_LOCK_UNLOCK);
LockBuffer(buf, BUFFER_LOCK_EXCLUSIVE);
exclusive_lock_held = true;
}
fsm_rebuild_page(page);
MarkBufferDirtyHint(buf, false);
goto restart;
}
}
/* We're now at the bottom level, at a node with enough space. */
slot = nodeno - NonLeafNodesPerPage;
/*
* Update the next-target pointer. Note that we do this even if we're only
* holding a shared lock, on the grounds that it's better to use a shared
* lock and get a garbled next pointer every now and then, than take the
* concurrency hit of an exclusive lock.
*
* Wrap-around is handled at the beginning of this function.
*/
fsmpage->fp_next_slot = slot + (advancenext ? 1 : 0);
return slot;
}