/* * allocate and initialize a new BTPageState. the returned structure * is suitable for immediate use by _bt_buildadd. */ static BTPageState * _bt_pagestate(BTWriteState *wstate, uint32 level) { BTPageState *state = (BTPageState *) palloc0(sizeof(BTPageState)); /* create initial page for level */ state->btps_page = _bt_blnewpage(level); /* and assign it a page position */ state->btps_blkno = wstate->btws_pages_alloced++; state->btps_minkey = NULL; /* initialize lastoff so first item goes into P_FIRSTKEY */ state->btps_lastoff = P_HIKEY; state->btps_level = level; /* set "full" threshold based on level. See notes at head of file. */ if (level > 0) state->btps_full = (BLCKSZ * (100 - BTREE_NONLEAF_FILLFACTOR) / 100); else state->btps_full = RelationGetTargetPageFreeSpace(wstate->index, BTREE_DEFAULT_FILLFACTOR); /* no parent level, yet */ state->btps_next = NULL; return state; }
/* * Per-tuple callback from IndexBuildHeapScan */ static void gistbuildCallback(Relation index, ItemPointer tupleId, Datum *values, bool *isnull, bool tupleIsAlive __attribute__((unused)), void *state) { GISTBuildState *buildstate = (GISTBuildState *) state; IndexTuple itup; MemoryContext oldCtx; oldCtx = MemoryContextSwitchTo(buildstate->tmpCtx); /* form an index tuple and point it at the heap tuple */ itup = gistFormTuple(&buildstate->giststate, index, values, isnull, true /* size is currently bogus */ ); itup->t_tid = *tupleId; /* * Since we already have the index relation locked, we call gistdoinsert * directly. Normal access method calls dispatch through gistinsert, * which locks the relation for write. This is the right thing to do if * you're inserting single tups, but not when you're initializing the * whole index at once. * * In this path we respect the fillfactor setting, whereas insertions * after initial build do not. */ gistdoinsert(index, itup, RelationGetTargetPageFreeSpace(index, GIST_DEFAULT_FILLFACTOR), &buildstate->giststate); buildstate->indtuples += 1; MemoryContextSwitchTo(oldCtx); MemoryContextReset(buildstate->tmpCtx); }
/* * Insert a tuple to the new relation. This has to track heap_insert * and its subsidiary functions! * * t_self of the tuple is set to the new TID of the tuple. If t_ctid of the * tuple is invalid on entry, it's replaced with the new TID as well (in * the inserted data only, not in the caller's copy). */ static void raw_heap_insert(RewriteState state, HeapTuple tup) { Page page = state->rs_buffer; Size pageFreeSpace, saveFreeSpace; Size len; OffsetNumber newoff; HeapTuple heaptup; /* * If the new tuple is too big for storage or contains already toasted * out-of-line attributes from some other relation, invoke the toaster. * * Note: below this point, heaptup is the data we actually intend to store * into the relation; tup is the caller's original untoasted data. */ if (state->rs_new_rel->rd_rel->relkind == RELKIND_TOASTVALUE) { /* toast table entries should never be recursively toasted */ Assert(!HeapTupleHasExternal(tup)); heaptup = tup; } else if (HeapTupleHasExternal(tup) || tup->t_len > TOAST_TUPLE_THRESHOLD) heaptup = toast_insert_or_update(state->rs_new_rel, tup, NULL, HEAP_INSERT_SKIP_FSM | (state->rs_use_wal ? 0 : HEAP_INSERT_SKIP_WAL)); else heaptup = tup; len = MAXALIGN(heaptup->t_len); /* be conservative */ /* * If we're gonna fail for oversize tuple, do it right away */ if (len > MaxHeapTupleSize) ereport(ERROR, (errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED), errmsg("row is too big: size %zu, maximum size %zu", len, MaxHeapTupleSize))); /* Compute desired extra freespace due to fillfactor option */ saveFreeSpace = RelationGetTargetPageFreeSpace(state->rs_new_rel, HEAP_DEFAULT_FILLFACTOR); /* Now we can check to see if there's enough free space already. */ if (state->rs_buffer_valid) { pageFreeSpace = PageGetHeapFreeSpace(page); if (len + saveFreeSpace > pageFreeSpace) { /* Doesn't fit, so write out the existing page */ /* XLOG stuff */ if (state->rs_use_wal) log_newpage(&state->rs_new_rel->rd_node, MAIN_FORKNUM, state->rs_blockno, page, true); /* * Now write the page. We say isTemp = true even if it's not a * temp table, because there's no need for smgr to schedule an * fsync for this write; we'll do it ourselves in * end_heap_rewrite. */ RelationOpenSmgr(state->rs_new_rel); PageSetChecksumInplace(page, state->rs_blockno); smgrextend(state->rs_new_rel->rd_smgr, MAIN_FORKNUM, state->rs_blockno, (char *) page, true); state->rs_blockno++; state->rs_buffer_valid = false; } } if (!state->rs_buffer_valid) { /* Initialize a new empty page */ PageInit(page, BLCKSZ, 0); state->rs_buffer_valid = true; } /* And now we can insert the tuple into the page */ newoff = PageAddItem(page, (Item) heaptup->t_data, heaptup->t_len, InvalidOffsetNumber, false, true); if (newoff == InvalidOffsetNumber) elog(ERROR, "failed to add tuple"); /* Update caller's t_self to the actual position where it was stored */ ItemPointerSet(&(tup->t_self), state->rs_blockno, newoff); /* * Insert the correct position into CTID of the stored tuple, too, if the * caller didn't supply a valid CTID. */ if (!ItemPointerIsValid(&tup->t_data->t_ctid)) { ItemId newitemid; HeapTupleHeader onpage_tup; newitemid = PageGetItemId(page, newoff); onpage_tup = (HeapTupleHeader) PageGetItem(page, newitemid); onpage_tup->t_ctid = tup->t_self; } /* If heaptup is a private copy, release it. */ if (heaptup != tup) heap_freetuple(heaptup); }
/* * RelationGetBufferForTuple * * Returns pinned and exclusive-locked buffer of a page in given relation * with free space >= given len. * * If otherBuffer is not InvalidBuffer, then it references a previously * pinned buffer of another page in the same relation; on return, this * buffer will also be exclusive-locked. (This case is used by heap_update; * the otherBuffer contains the tuple being updated.) * * The reason for passing otherBuffer is that if two backends are doing * concurrent heap_update operations, a deadlock could occur if they try * to lock the same two buffers in opposite orders. To ensure that this * can't happen, we impose the rule that buffers of a relation must be * locked in increasing page number order. This is most conveniently done * by having RelationGetBufferForTuple lock them both, with suitable care * for ordering. * * NOTE: it is unlikely, but not quite impossible, for otherBuffer to be the * same buffer we select for insertion of the new tuple (this could only * happen if space is freed in that page after heap_update finds there's not * enough there). In that case, the page will be pinned and locked only once. * * For the vmbuffer and vmbuffer_other arguments, we avoid deadlock by * locking them only after locking the corresponding heap page, and taking * no further lwlocks while they are locked. * * We normally use FSM to help us find free space. However, * if HEAP_INSERT_SKIP_FSM is specified, we just append a new empty page to * the end of the relation if the tuple won't fit on the current target page. * This can save some cycles when we know the relation is new and doesn't * contain useful amounts of free space. * * HEAP_INSERT_SKIP_FSM is also useful for non-WAL-logged additions to a * relation, if the caller holds exclusive lock and is careful to invalidate * relation's smgr_targblock before the first insertion --- that ensures that * all insertions will occur into newly added pages and not be intermixed * with tuples from other transactions. That way, a crash can't risk losing * any committed data of other transactions. (See heap_insert's comments * for additional constraints needed for safe usage of this behavior.) * * The caller can also provide a BulkInsertState object to optimize many * insertions into the same relation. This keeps a pin on the current * insertion target page (to save pin/unpin cycles) and also passes a * BULKWRITE buffer selection strategy object to the buffer manager. * Passing NULL for bistate selects the default behavior. * * We always try to avoid filling existing pages further than the fillfactor. * This is OK since this routine is not consulted when updating a tuple and * keeping it on the same page, which is the scenario fillfactor is meant * to reserve space for. * * ereport(ERROR) is allowed here, so this routine *must* be called * before any (unlogged) changes are made in buffer pool. */ Buffer RelationGetBufferForTuple(Relation relation, Size len, Buffer otherBuffer, int options, BulkInsertState bistate, Buffer *vmbuffer, Buffer *vmbuffer_other) { bool use_fsm = !(options & HEAP_INSERT_SKIP_FSM); Buffer buffer = InvalidBuffer; Page page; Size pageFreeSpace, saveFreeSpace; BlockNumber targetBlock, otherBlock; bool needLock; len = MAXALIGN(len); /* be conservative */ /* Bulk insert is not supported for updates, only inserts. */ Assert(otherBuffer == InvalidBuffer || !bistate); /* * If we're gonna fail for oversize tuple, do it right away */ if (len > MaxHeapTupleSize) ereport(ERROR, (errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED), errmsg("row is too big: size %lu, maximum size %lu", (unsigned long) len, (unsigned long) MaxHeapTupleSize))); /* Compute desired extra freespace due to fillfactor option */ saveFreeSpace = RelationGetTargetPageFreeSpace(relation, HEAP_DEFAULT_FILLFACTOR); if (otherBuffer != InvalidBuffer) otherBlock = BufferGetBlockNumber(otherBuffer); else otherBlock = InvalidBlockNumber; /* just to keep compiler quiet */ /* * We first try to put the tuple on the same page we last inserted a tuple * on, as cached in the BulkInsertState or relcache entry. If that * doesn't work, we ask the Free Space Map to locate a suitable page. * Since the FSM's info might be out of date, we have to be prepared to * loop around and retry multiple times. (To insure this isn't an infinite * loop, we must update the FSM with the correct amount of free space on * each page that proves not to be suitable.) If the FSM has no record of * a page with enough free space, we give up and extend the relation. * * When use_fsm is false, we either put the tuple onto the existing target * page or extend the relation. */ if (len + saveFreeSpace > MaxHeapTupleSize) { /* can't fit, don't bother asking FSM */ targetBlock = InvalidBlockNumber; use_fsm = false; } else if (bistate && bistate->current_buf != InvalidBuffer) targetBlock = BufferGetBlockNumber(bistate->current_buf); else targetBlock = RelationGetTargetBlock(relation); if (targetBlock == InvalidBlockNumber && use_fsm) { /* * We have no cached target page, so ask the FSM for an initial * target. */ targetBlock = GetPageWithFreeSpace(relation, len + saveFreeSpace); /* * If the FSM knows nothing of the rel, try the last page before we * give up and extend. This avoids one-tuple-per-page syndrome during * bootstrapping or in a recently-started system. */ if (targetBlock == InvalidBlockNumber) { BlockNumber nblocks = RelationGetNumberOfBlocks(relation); if (nblocks > 0) targetBlock = nblocks - 1; } } while (targetBlock != InvalidBlockNumber) { /* * Read and exclusive-lock the target block, as well as the other * block if one was given, taking suitable care with lock ordering and * the possibility they are the same block. * * If the page-level all-visible flag is set, caller will need to * clear both that and the corresponding visibility map bit. However, * by the time we return, we'll have x-locked the buffer, and we don't * want to do any I/O while in that state. So we check the bit here * before taking the lock, and pin the page if it appears necessary. * Checking without the lock creates a risk of getting the wrong * answer, so we'll have to recheck after acquiring the lock. */ if (otherBuffer == InvalidBuffer) { /* easy case */ buffer = ReadBufferBI(relation, targetBlock, bistate); if (PageIsAllVisible(BufferGetPage(buffer))) visibilitymap_pin(relation, targetBlock, vmbuffer); LockBuffer(buffer, BUFFER_LOCK_EXCLUSIVE); } else if (otherBlock == targetBlock) { /* also easy case */ buffer = otherBuffer; if (PageIsAllVisible(BufferGetPage(buffer))) visibilitymap_pin(relation, targetBlock, vmbuffer); LockBuffer(buffer, BUFFER_LOCK_EXCLUSIVE); } else if (otherBlock < targetBlock) { /* lock other buffer first */ buffer = ReadBuffer(relation, targetBlock); if (PageIsAllVisible(BufferGetPage(buffer))) visibilitymap_pin(relation, targetBlock, vmbuffer); LockBuffer(otherBuffer, BUFFER_LOCK_EXCLUSIVE); LockBuffer(buffer, BUFFER_LOCK_EXCLUSIVE); } else { /* lock target buffer first */ buffer = ReadBuffer(relation, targetBlock); if (PageIsAllVisible(BufferGetPage(buffer))) visibilitymap_pin(relation, targetBlock, vmbuffer); LockBuffer(buffer, BUFFER_LOCK_EXCLUSIVE); LockBuffer(otherBuffer, BUFFER_LOCK_EXCLUSIVE); } /* * We now have the target page (and the other buffer, if any) pinned * and locked. However, since our initial PageIsAllVisible checks * were performed before acquiring the lock, the results might now be * out of date, either for the selected victim buffer, or for the * other buffer passed by the caller. In that case, we'll need to * give up our locks, go get the pin(s) we failed to get earlier, and * re-lock. That's pretty painful, but hopefully shouldn't happen * often. * * Note that there's a small possibility that we didn't pin the page * above but still have the correct page pinned anyway, either because * we've already made a previous pass through this loop, or because * caller passed us the right page anyway. * * Note also that it's possible that by the time we get the pin and * retake the buffer locks, the visibility map bit will have been * cleared by some other backend anyway. In that case, we'll have * done a bit of extra work for no gain, but there's no real harm * done. */ if (otherBuffer == InvalidBuffer || buffer <= otherBuffer) GetVisibilityMapPins(relation, buffer, otherBuffer, targetBlock, otherBlock, vmbuffer, vmbuffer_other); else GetVisibilityMapPins(relation, otherBuffer, buffer, otherBlock, targetBlock, vmbuffer_other, vmbuffer); /* * Now we can check to see if there's enough free space here. If so, * we're done. */ page = BufferGetPage(buffer); pageFreeSpace = PageGetHeapFreeSpace(page); if (len + saveFreeSpace <= pageFreeSpace) { /* use this page as future insert target, too */ RelationSetTargetBlock(relation, targetBlock); return buffer; } /* * Not enough space, so we must give up our page locks and pin (if * any) and prepare to look elsewhere. We don't care which order we * unlock the two buffers in, so this can be slightly simpler than the * code above. */ LockBuffer(buffer, BUFFER_LOCK_UNLOCK); if (otherBuffer == InvalidBuffer) ReleaseBuffer(buffer); else if (otherBlock != targetBlock) { LockBuffer(otherBuffer, BUFFER_LOCK_UNLOCK); ReleaseBuffer(buffer); } /* Without FSM, always fall out of the loop and extend */ if (!use_fsm) break; /* * Update FSM as to condition of this page, and ask for another page * to try. */ targetBlock = RecordAndGetPageWithFreeSpace(relation, targetBlock, pageFreeSpace, len + saveFreeSpace); } /* * Have to extend the relation. * * We have to use a lock to ensure no one else is extending the rel at the * same time, else we will both try to initialize the same new page. We * can skip locking for new or temp relations, however, since no one else * could be accessing them. */ needLock = !RELATION_IS_LOCAL(relation); if (needLock) LockRelationForExtension(relation, ExclusiveLock); /* * XXX This does an lseek - rather expensive - but at the moment it is the * only way to accurately determine how many blocks are in a relation. Is * it worth keeping an accurate file length in shared memory someplace, * rather than relying on the kernel to do it for us? */ buffer = ReadBufferBI(relation, P_NEW, bistate); /* * We can be certain that locking the otherBuffer first is OK, since it * must have a lower page number. */ if (otherBuffer != InvalidBuffer) LockBuffer(otherBuffer, BUFFER_LOCK_EXCLUSIVE); /* * Now acquire lock on the new page. */ LockBuffer(buffer, BUFFER_LOCK_EXCLUSIVE); /* * Release the file-extension lock; it's now OK for someone else to extend * the relation some more. Note that we cannot release this lock before * we have buffer lock on the new page, or we risk a race condition * against vacuumlazy.c --- see comments therein. */ if (needLock) UnlockRelationForExtension(relation, ExclusiveLock); /* * We need to initialize the empty new page. Double-check that it really * is empty (this should never happen, but if it does we don't want to * risk wiping out valid data). */ page = BufferGetPage(buffer); if (!PageIsNew(page)) elog(ERROR, "page %u of relation \"%s\" should be empty but is not", BufferGetBlockNumber(buffer), RelationGetRelationName(relation)); PageInit(page, BufferGetPageSize(buffer), 0); if (len > PageGetHeapFreeSpace(page)) { /* We should not get here given the test at the top */ elog(PANIC, "tuple is too big: size %lu", (unsigned long) len); } /* * Remember the new page as our target for future insertions. * * XXX should we enter the new page into the free space map immediately, * or just keep it for this backend's exclusive use in the short run * (until VACUUM sees it)? Seems to depend on whether you expect the * current backend to make more insertions or not, which is probably a * good bet most of the time. So for now, don't add it to FSM yet. */ RelationSetTargetBlock(relation, BufferGetBlockNumber(buffer)); return buffer; }
/* * Optionally prune and repair fragmentation in the specified page. * * This is an opportunistic function. It will perform housekeeping * only if the page heuristically looks like a candidate for pruning and we * can acquire buffer cleanup lock without blocking. * * Note: this is called quite often. It's important that it fall out quickly * if there's not any use in pruning. * * Caller must have pin on the buffer, and must *not* have a lock on it. * * OldestXmin is the cutoff XID used to distinguish whether tuples are DEAD * or RECENTLY_DEAD (see HeapTupleSatisfiesVacuum). */ void heap_page_prune_opt(Relation relation, Buffer buffer, TransactionId OldestXmin) { Page page = BufferGetPage(buffer); Size minfree; /* * In GPDB we may call into here without having a local snapshot and thus * no valid OldestXmin transaction id. Exit early if so. */ if (!TransactionIdIsValid(OldestXmin)) return; /* * Let's see if we really need pruning. * * Forget it if page is not hinted to contain something prunable that's * older than OldestXmin. */ if (!PageIsPrunable(page, OldestXmin)) return; /* * We prune when a previous UPDATE failed to find enough space on the page * for a new tuple version, or when free space falls below the relation's * fill-factor target (but not less than 10%). * * Checking free space here is questionable since we aren't holding any * lock on the buffer; in the worst case we could get a bogus answer. It's * unlikely to be *seriously* wrong, though, since reading either pd_lower * or pd_upper is probably atomic. Avoiding taking a lock seems more * important than sometimes getting a wrong answer in what is after all * just a heuristic estimate. */ minfree = RelationGetTargetPageFreeSpace(relation, HEAP_DEFAULT_FILLFACTOR); minfree = Max(minfree, BLCKSZ / 10); if (PageIsFull(page) || PageGetHeapFreeSpace(page) < minfree) { /* * Check if we have gp_persistent_relation_node information, to be * added to the XLOG record. As in some cases it maybe too late to * fetch the same and hence for such cases just give-up. */ if (!RelationAllowedToGenerateXLogRecord(relation)) return; /* OK, try to get exclusive buffer lock */ if (!ConditionalLockBufferForCleanup(buffer)) return; /* * Now that we have buffer lock, get accurate information about the * page's free space, and recheck the heuristic about whether to * prune. (We needn't recheck PageIsPrunable, since no one else could * have pruned while we hold pin.) */ if (PageIsFull(page) || PageGetHeapFreeSpace(page) < minfree) { /* OK to prune (though not to remove redirects) */ (void) heap_page_prune(relation, buffer, OldestXmin, false, true); } /* And release buffer lock */ LockBuffer(buffer, BUFFER_LOCK_UNLOCK); } }
/* * Get a buffer of the type and parity specified by flags, having at least * as much free space as indicated by needSpace. We use the lastUsedPages * cache to assign the same buffer previously requested when possible. * The returned buffer is already pinned and exclusive-locked. * * *isNew is set true if the page was initialized here, false if it was * already valid. */ Buffer SpGistGetBuffer(Relation index, int flags, int needSpace, bool *isNew) { SpGistCache *cache = spgGetCache(index); SpGistLastUsedPage *lup; /* Bail out if even an empty page wouldn't meet the demand */ if (needSpace > SPGIST_PAGE_CAPACITY) elog(ERROR, "desired SPGiST tuple size is too big"); /* * If possible, increase the space request to include relation's * fillfactor. This ensures that when we add unrelated tuples to a page, * we try to keep 100-fillfactor% available for adding tuples that are * related to the ones already on it. But fillfactor mustn't cause an * error for requests that would otherwise be legal. */ needSpace += RelationGetTargetPageFreeSpace(index, SPGIST_DEFAULT_FILLFACTOR); needSpace = Min(needSpace, SPGIST_PAGE_CAPACITY); /* Get the cache entry for this flags setting */ lup = GET_LUP(cache, flags); /* If we have nothing cached, just turn it over to allocNewBuffer */ if (lup->blkno == InvalidBlockNumber) { *isNew = true; return allocNewBuffer(index, flags); } /* fixed pages should never be in cache */ Assert(!SpGistBlockIsFixed(lup->blkno)); /* If cached freeSpace isn't enough, don't bother looking at the page */ if (lup->freeSpace >= needSpace) { Buffer buffer; Page page; buffer = ReadBuffer(index, lup->blkno); if (!ConditionalLockBuffer(buffer)) { /* * buffer is locked by another process, so return a new buffer */ ReleaseBuffer(buffer); *isNew = true; return allocNewBuffer(index, flags); } page = BufferGetPage(buffer); if (PageIsNew(page) || SpGistPageIsDeleted(page) || PageIsEmpty(page)) { /* OK to initialize the page */ uint16 pageflags = 0; if (GBUF_REQ_LEAF(flags)) pageflags |= SPGIST_LEAF; if (GBUF_REQ_NULLS(flags)) pageflags |= SPGIST_NULLS; SpGistInitBuffer(buffer, pageflags); lup->freeSpace = PageGetExactFreeSpace(page) - needSpace; *isNew = true; return buffer; } /* * Check that page is of right type and has enough space. We must * recheck this since our cache isn't necessarily up to date. */ if ((GBUF_REQ_LEAF(flags) ? SpGistPageIsLeaf(page) : !SpGistPageIsLeaf(page)) && (GBUF_REQ_NULLS(flags) ? SpGistPageStoresNulls(page) : !SpGistPageStoresNulls(page))) { int freeSpace = PageGetExactFreeSpace(page); if (freeSpace >= needSpace) { /* Success, update freespace info and return the buffer */ lup->freeSpace = freeSpace - needSpace; *isNew = false; return buffer; } } /* * fallback to allocation of new buffer */ UnlockReleaseBuffer(buffer); } /* No success with cache, so return a new buffer */ *isNew = true; return allocNewBuffer(index, flags); }
/* * Optionally prune and repair fragmentation in the specified page. * * This is an opportunistic function. It will perform housekeeping * only if the page heuristically looks like a candidate for pruning and we * can acquire buffer cleanup lock without blocking. * * Note: this is called quite often. It's important that it fall out quickly * if there's not any use in pruning. * * Caller must have pin on the buffer, and must *not* have a lock on it. * * OldestXmin is the cutoff XID used to distinguish whether tuples are DEAD * or RECENTLY_DEAD (see HeapTupleSatisfiesVacuum). */ void heap_page_prune_opt(Relation relation, Buffer buffer, TransactionId OldestXmin) { Page page = BufferGetPage(buffer); Size minfree; /* * Let's see if we really need pruning. * * Forget it if page is not hinted to contain something prunable that's * older than OldestXmin. */ if (!PageIsPrunable(page, OldestXmin)) return; /* * We can't write WAL in recovery mode, so there's no point trying to * clean the page. The master will likely issue a cleaning WAL record soon * anyway, so this is no particular loss. */ if (RecoveryInProgress()) return; /* * We prune when a previous UPDATE failed to find enough space on the page * for a new tuple version, or when free space falls below the relation's * fill-factor target (but not less than 10%). * * Checking free space here is questionable since we aren't holding any * lock on the buffer; in the worst case we could get a bogus answer. It's * unlikely to be *seriously* wrong, though, since reading either pd_lower * or pd_upper is probably atomic. Avoiding taking a lock seems more * important than sometimes getting a wrong answer in what is after all * just a heuristic estimate. */ minfree = RelationGetTargetPageFreeSpace(relation, HEAP_DEFAULT_FILLFACTOR); minfree = Max(minfree, BLCKSZ / 10); if (PageIsFull(page) || PageGetHeapFreeSpace(page) < minfree) { /* OK, try to get exclusive buffer lock */ if (!ConditionalLockBufferForCleanup(buffer)) return; /* * Now that we have buffer lock, get accurate information about the * page's free space, and recheck the heuristic about whether to * prune. (We needn't recheck PageIsPrunable, since no one else could * have pruned while we hold pin.) */ if (PageIsFull(page) || PageGetHeapFreeSpace(page) < minfree) { TransactionId ignore = InvalidTransactionId; /* return value not * needed */ /* OK to prune */ (void) heap_page_prune(relation, buffer, OldestXmin, true, &ignore); } /* And release buffer lock */ LockBuffer(buffer, BUFFER_LOCK_UNLOCK); } }
/* * Optionally prune and repair fragmentation in the specified page. * * This is an opportunistic function. It will perform housekeeping * only if the page heuristically looks like a candidate for pruning and we * can acquire buffer cleanup lock without blocking. * * Note: this is called quite often. It's important that it fall out quickly * if there's not any use in pruning. * * Caller must have pin on the buffer, and must *not* have a lock on it. * * OldestXmin is the cutoff XID used to distinguish whether tuples are DEAD * or RECENTLY_DEAD (see HeapTupleSatisfiesVacuum). */ void heap_page_prune_opt(Relation relation, Buffer buffer) { Page page = BufferGetPage(buffer); Size minfree; TransactionId OldestXmin; /* * We can't write WAL in recovery mode, so there's no point trying to * clean the page. The master will likely issue a cleaning WAL record soon * anyway, so this is no particular loss. */ if (RecoveryInProgress()) return; /* * Use the appropriate xmin horizon for this relation. If it's a proper * catalog relation or a user defined, additional, catalog relation, we * need to use the horizon that includes slots, otherwise the data-only * horizon can be used. Note that the toast relation of user defined * relations are *not* considered catalog relations. * * It is OK to apply the old snapshot limit before acquiring the cleanup * lock because the worst that can happen is that we are not quite as * aggressive about the cleanup (by however many transaction IDs are * consumed between this point and acquiring the lock). This allows us to * save significant overhead in the case where the page is found not to be * prunable. */ if (IsCatalogRelation(relation) || RelationIsAccessibleInLogicalDecoding(relation)) OldestXmin = RecentGlobalXmin; else OldestXmin = TransactionIdLimitedForOldSnapshots(RecentGlobalDataXmin, relation); Assert(TransactionIdIsValid(OldestXmin)); /* * Let's see if we really need pruning. * * Forget it if page is not hinted to contain something prunable that's * older than OldestXmin. */ if (!PageIsPrunable(page, OldestXmin)) return; /* * We prune when a previous UPDATE failed to find enough space on the page * for a new tuple version, or when free space falls below the relation's * fill-factor target (but not less than 10%). * * Checking free space here is questionable since we aren't holding any * lock on the buffer; in the worst case we could get a bogus answer. It's * unlikely to be *seriously* wrong, though, since reading either pd_lower * or pd_upper is probably atomic. Avoiding taking a lock seems more * important than sometimes getting a wrong answer in what is after all * just a heuristic estimate. */ minfree = RelationGetTargetPageFreeSpace(relation, HEAP_DEFAULT_FILLFACTOR); minfree = Max(minfree, BLCKSZ / 10); if (PageIsFull(page) || PageGetHeapFreeSpace(page) < minfree) { /* OK, try to get exclusive buffer lock */ if (!ConditionalLockBufferForCleanup(buffer)) return; /* * Now that we have buffer lock, get accurate information about the * page's free space, and recheck the heuristic about whether to * prune. (We needn't recheck PageIsPrunable, since no one else could * have pruned while we hold pin.) */ if (PageIsFull(page) || PageGetHeapFreeSpace(page) < minfree) { TransactionId ignore = InvalidTransactionId; /* return value not * needed */ /* OK to prune */ (void) heap_page_prune(relation, buffer, OldestXmin, true, &ignore); } /* And release buffer lock */ LockBuffer(buffer, BUFFER_LOCK_UNLOCK); } }
/* * RelationGetBufferForTuple * * Returns pinned and exclusive-locked buffer of a page in given relation * with free space >= given len. * * If otherBuffer is not InvalidBuffer, then it references a previously * pinned buffer of another page in the same relation; on return, this * buffer will also be exclusive-locked. (This case is used by heap_update; * the otherBuffer contains the tuple being updated.) * * The reason for passing otherBuffer is that if two backends are doing * concurrent heap_update operations, a deadlock could occur if they try * to lock the same two buffers in opposite orders. To ensure that this * can't happen, we impose the rule that buffers of a relation must be * locked in increasing page number order. This is most conveniently done * by having RelationGetBufferForTuple lock them both, with suitable care * for ordering. * * NOTE: it is unlikely, but not quite impossible, for otherBuffer to be the * same buffer we select for insertion of the new tuple (this could only * happen if space is freed in that page after heap_update finds there's not * enough there). In that case, the page will be pinned and locked only once. * * If use_fsm is true (the normal case), we use FSM to help us find free * space. If use_fsm is false, we always append a new empty page to the * end of the relation if the tuple won't fit on the current target page. * This can save some cycles when we know the relation is new and doesn't * contain useful amounts of free space. * * The use_fsm = false case is also useful for non-WAL-logged additions to a * relation, if the caller holds exclusive lock and is careful to invalidate * relation->rd_targblock before the first insertion --- that ensures that * all insertions will occur into newly added pages and not be intermixed * with tuples from other transactions. That way, a crash can't risk losing * any committed data of other transactions. (See heap_insert's comments * for additional constraints needed for safe usage of this behavior.) * * We always try to avoid filling existing pages further than the fillfactor. * This is OK since this routine is not consulted when updating a tuple and * keeping it on the same page, which is the scenario fillfactor is meant * to reserve space for. * * ereport(ERROR) is allowed here, so this routine *must* be called * before any (unlogged) changes are made in buffer pool. */ Buffer RelationGetBufferForTuple(Relation relation, Size len, Buffer otherBuffer, bool use_fsm) { Buffer buffer = InvalidBuffer; Page pageHeader; Size pageFreeSpace, saveFreeSpace; BlockNumber targetBlock, otherBlock; bool needLock; MIRROREDLOCK_BUFMGR_MUST_ALREADY_BE_HELD; len = MAXALIGN(len); /* be conservative */ /* * If we're gonna fail for oversize tuple, do it right away */ if (len > MaxHeapTupleSize) ereport(ERROR, (errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED), errmsg("row is too big: size %lu, maximum size %lu", (unsigned long) len, (unsigned long) MaxHeapTupleSize))); /* Compute desired extra freespace due to fillfactor option */ saveFreeSpace = RelationGetTargetPageFreeSpace(relation, HEAP_DEFAULT_FILLFACTOR); if (otherBuffer != InvalidBuffer) otherBlock = BufferGetBlockNumber(otherBuffer); else otherBlock = InvalidBlockNumber; /* just to keep compiler quiet */ /* * We first try to put the tuple on the same page we last inserted a tuple * on, as cached in the relcache entry. If that doesn't work, we ask the * shared Free Space Map to locate a suitable page. Since the FSM's info * might be out of date, we have to be prepared to loop around and retry * multiple times. (To insure this isn't an infinite loop, we must update * the FSM with the correct amount of free space on each page that proves * not to be suitable.) If the FSM has no record of a page with enough * free space, we give up and extend the relation. * * When use_fsm is false, we either put the tuple onto the existing target * page or extend the relation. */ if (len + saveFreeSpace <= MaxHeapTupleSize) targetBlock = relation->rd_targblock; else { /* can't fit, don't screw up FSM request tracking by trying */ targetBlock = InvalidBlockNumber; use_fsm = false; } if (targetBlock == InvalidBlockNumber && use_fsm) { /* * We have no cached target page, so ask the FSM for an initial * target. */ targetBlock = GetPageWithFreeSpace(&relation->rd_node, len + saveFreeSpace); /* * If the FSM knows nothing of the rel, try the last page before we * give up and extend. This avoids one-tuple-per-page syndrome during * bootstrapping or in a recently-started system. */ if (targetBlock == InvalidBlockNumber) { BlockNumber nblocks = RelationGetNumberOfBlocks(relation); if (nblocks > 0) targetBlock = nblocks - 1; } } while (targetBlock != InvalidBlockNumber) { /* * Read and exclusive-lock the target block, as well as the other * block if one was given, taking suitable care with lock ordering and * the possibility they are the same block. */ if (otherBuffer == InvalidBuffer) { /* easy case */ buffer = ReadBuffer(relation, targetBlock); LockBuffer(buffer, BUFFER_LOCK_EXCLUSIVE); } else if (otherBlock == targetBlock) { /* also easy case */ buffer = otherBuffer; LockBuffer(buffer, BUFFER_LOCK_EXCLUSIVE); } else if (otherBlock < targetBlock) { /* lock other buffer first */ buffer = ReadBuffer(relation, targetBlock); LockBuffer(otherBuffer, BUFFER_LOCK_EXCLUSIVE); LockBuffer(buffer, BUFFER_LOCK_EXCLUSIVE); } else { /* lock target buffer first */ buffer = ReadBuffer(relation, targetBlock); LockBuffer(buffer, BUFFER_LOCK_EXCLUSIVE); LockBuffer(otherBuffer, BUFFER_LOCK_EXCLUSIVE); } /* * Now we can check to see if there's enough free space here. If so, * we're done. */ pageHeader = (Page) BufferGetPage(buffer); pageFreeSpace = PageGetFreeSpace(pageHeader); if (len + saveFreeSpace <= pageFreeSpace) { /* use this page as future insert target, too */ relation->rd_targblock = targetBlock; return buffer; } /* * Not enough space, so we must give up our page locks and pin (if * any) and prepare to look elsewhere. We don't care which order we * unlock the two buffers in, so this can be slightly simpler than the * code above. */ LockBuffer(buffer, BUFFER_LOCK_UNLOCK); if (otherBuffer == InvalidBuffer) ReleaseBuffer(buffer); else if (otherBlock != targetBlock) { LockBuffer(otherBuffer, BUFFER_LOCK_UNLOCK); ReleaseBuffer(buffer); } /* Without FSM, always fall out of the loop and extend */ if (!use_fsm) break; /* * Update FSM as to condition of this page, and ask for another page * to try. */ targetBlock = RecordAndGetPageWithFreeSpace(&relation->rd_node, targetBlock, pageFreeSpace, len + saveFreeSpace); } /* * Have to extend the relation. * * We have to use a lock to ensure no one else is extending the rel at the * same time, else we will both try to initialize the same new page. We * can skip locking for new or temp relations, however, since no one else * could be accessing them. */ needLock = !RELATION_IS_LOCAL(relation); if (needLock) LockRelationForExtension(relation, ExclusiveLock); /* * XXX This does an lseek - rather expensive - but at the moment it is the * only way to accurately determine how many blocks are in a relation. Is * it worth keeping an accurate file length in shared memory someplace, * rather than relying on the kernel to do it for us? */ buffer = ReadBuffer(relation, P_NEW); /* * We can be certain that locking the otherBuffer first is OK, since it * must have a lower page number. */ if (otherBuffer != InvalidBuffer) LockBuffer(otherBuffer, BUFFER_LOCK_EXCLUSIVE); /* * Now acquire lock on the new page. */ LockBuffer(buffer, BUFFER_LOCK_EXCLUSIVE); /* * Release the file-extension lock; it's now OK for someone else to extend * the relation some more. Note that we cannot release this lock before * we have buffer lock on the new page, or we risk a race condition * against vacuumlazy.c --- see comments therein. */ if (needLock) UnlockRelationForExtension(relation, ExclusiveLock); /* * We need to initialize the empty new page. Double-check that it really * is empty (this should never happen, but if it does we don't want to * risk wiping out valid data). */ pageHeader = (Page) BufferGetPage(buffer); if (!PageIsNew((PageHeader) pageHeader)) elog(ERROR, "page %u of relation \"%s\" should be empty but is not", BufferGetBlockNumber(buffer), RelationGetRelationName(relation)); PageInit(pageHeader, BufferGetPageSize(buffer), 0); if (len > PageGetFreeSpace(pageHeader)) { /* We should not get here given the test at the top */ elog(PANIC, "tuple is too big: size %lu", (unsigned long) len); } /* * Remember the new page as our target for future insertions. * * XXX should we enter the new page into the free space map immediately, * or just keep it for this backend's exclusive use in the short run * (until VACUUM sees it)? Seems to depend on whether you expect the * current backend to make more insertions or not, which is probably a * good bet most of the time. So for now, don't add it to FSM yet. */ relation->rd_targblock = BufferGetBlockNumber(buffer); return buffer; }
/** * @brief Create LoadStatus file and load heap tuples directly. * @return void */ static void DirectWriterInsert(DirectWriter *self, HeapTuple tuple) { Page page; OffsetNumber offnum; ItemId itemId; Item item; LoadStatus *ls = &self->ls; /* Compress the tuple data if needed. */ if (tuple->t_len > TOAST_TUPLE_THRESHOLD) tuple = toast_insert_or_update(self->base.rel, tuple, NULL, 0); BULKLOAD_PROFILE(&prof_writer_toast); /* Assign oids if needed. */ if (self->base.rel->rd_rel->relhasoids) { Assert(!OidIsValid(HeapTupleGetOid(tuple))); HeapTupleSetOid(tuple, GetNewOid(self->base.rel)); } /* Assume the tuple has been toasted already. */ if (MAXALIGN(tuple->t_len) > MaxHeapTupleSize) ereport(ERROR, (errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED), errmsg("row is too big: size %lu, maximum size %lu", (unsigned long) tuple->t_len, (unsigned long) MaxHeapTupleSize))); /* Fill current page, or go to next page if the page is full. */ page = GetCurrentPage(self); if (PageGetFreeSpace(page) < MAXALIGN(tuple->t_len) + RelationGetTargetPageFreeSpace(self->base.rel, HEAP_DEFAULT_FILLFACTOR)) { if (self->curblk < BLOCK_BUF_NUM - 1) self->curblk++; else { flush_pages(self); self->curblk = 0; /* recycle from first block */ } page = GetCurrentPage(self); /* Initialize current block */ PageInit(page, BLCKSZ, 0); PageSetTLI(page, ThisTimeLineID); } tuple->t_data->t_infomask &= ~(HEAP_XACT_MASK); tuple->t_data->t_infomask2 &= ~(HEAP2_XACT_MASK); tuple->t_data->t_infomask |= HEAP_XMAX_INVALID; HeapTupleHeaderSetXmin(tuple->t_data, self->xid); HeapTupleHeaderSetCmin(tuple->t_data, self->cid); HeapTupleHeaderSetXmax(tuple->t_data, 0); /* put the tuple on local page. */ offnum = PageAddItem(page, (Item) tuple->t_data, tuple->t_len, InvalidOffsetNumber, false, true); ItemPointerSet(&(tuple->t_self), LS_TOTAL_CNT(ls) + self->curblk, offnum); itemId = PageGetItemId(page, offnum); item = PageGetItem(page, itemId); ((HeapTupleHeader) item)->t_ctid = tuple->t_self; BULKLOAD_PROFILE(&prof_writer_table); SpoolerInsert(&self->spooler, tuple); BULKLOAD_PROFILE(&prof_writer_index); }