Line data Source code
1 : /*-------------------------------------------------------------------------
2 : *
3 : * spgvacuum.c
4 : * vacuum for SP-GiST
5 : *
6 : *
7 : * Portions Copyright (c) 1996-2017, PostgreSQL Global Development Group
8 : * Portions Copyright (c) 1994, Regents of the University of California
9 : *
10 : * IDENTIFICATION
11 : * src/backend/access/spgist/spgvacuum.c
12 : *
13 : *-------------------------------------------------------------------------
14 : */
15 :
16 : #include "postgres.h"
17 :
18 : #include "access/genam.h"
19 : #include "access/spgist_private.h"
20 : #include "access/spgxlog.h"
21 : #include "access/transam.h"
22 : #include "access/xloginsert.h"
23 : #include "catalog/storage_xlog.h"
24 : #include "commands/vacuum.h"
25 : #include "miscadmin.h"
26 : #include "storage/bufmgr.h"
27 : #include "storage/indexfsm.h"
28 : #include "storage/lmgr.h"
29 : #include "utils/snapmgr.h"
30 :
31 :
32 : /* Entry in pending-list of TIDs we need to revisit */
33 : typedef struct spgVacPendingItem
34 : {
35 : ItemPointerData tid; /* redirection target to visit */
36 : bool done; /* have we dealt with this? */
37 : struct spgVacPendingItem *next; /* list link */
38 : } spgVacPendingItem;
39 :
40 : /* Local state for vacuum operations */
41 : typedef struct spgBulkDeleteState
42 : {
43 : /* Parameters passed in to spgvacuumscan */
44 : IndexVacuumInfo *info;
45 : IndexBulkDeleteResult *stats;
46 : IndexBulkDeleteCallback callback;
47 : void *callback_state;
48 :
49 : /* Additional working state */
50 : SpGistState spgstate; /* for SPGiST operations that need one */
51 : spgVacPendingItem *pendingList; /* TIDs we need to (re)visit */
52 : TransactionId myXmin; /* for detecting newly-added redirects */
53 : BlockNumber lastFilledBlock; /* last non-deletable block */
54 : } spgBulkDeleteState;
55 :
56 :
57 : /*
58 : * Add TID to pendingList, but only if not already present.
59 : *
60 : * Note that new items are always appended at the end of the list; this
61 : * ensures that scans of the list don't miss items added during the scan.
62 : */
63 : static void
64 0 : spgAddPendingTID(spgBulkDeleteState *bds, ItemPointer tid)
65 : {
66 : spgVacPendingItem *pitem;
67 : spgVacPendingItem **listLink;
68 :
69 : /* search the list for pre-existing entry */
70 0 : listLink = &bds->pendingList;
71 0 : while (*listLink != NULL)
72 : {
73 0 : pitem = *listLink;
74 0 : if (ItemPointerEquals(tid, &pitem->tid))
75 0 : return; /* already in list, do nothing */
76 0 : listLink = &pitem->next;
77 : }
78 : /* not there, so append new entry */
79 0 : pitem = (spgVacPendingItem *) palloc(sizeof(spgVacPendingItem));
80 0 : pitem->tid = *tid;
81 0 : pitem->done = false;
82 0 : pitem->next = NULL;
83 0 : *listLink = pitem;
84 : }
85 :
86 : /*
87 : * Clear pendingList
88 : */
89 : static void
90 0 : spgClearPendingList(spgBulkDeleteState *bds)
91 : {
92 : spgVacPendingItem *pitem;
93 : spgVacPendingItem *nitem;
94 :
95 0 : for (pitem = bds->pendingList; pitem != NULL; pitem = nitem)
96 : {
97 0 : nitem = pitem->next;
98 : /* All items in list should have been dealt with */
99 0 : Assert(pitem->done);
100 0 : pfree(pitem);
101 : }
102 0 : bds->pendingList = NULL;
103 0 : }
104 :
105 : /*
106 : * Vacuum a regular (non-root) leaf page
107 : *
108 : * We must delete tuples that are targeted for deletion by the VACUUM,
109 : * but not move any tuples that are referenced by outside links; we assume
110 : * those are the ones that are heads of chains.
111 : *
112 : * If we find a REDIRECT that was made by a concurrently-running transaction,
113 : * we must add its target TID to pendingList. (We don't try to visit the
114 : * target immediately, first because we don't want VACUUM locking more than
115 : * one buffer at a time, and second because the duplicate-filtering logic
116 : * in spgAddPendingTID is useful to ensure we can't get caught in an infinite
117 : * loop in the face of continuous concurrent insertions.)
118 : *
119 : * If forPending is true, we are examining the page as a consequence of
120 : * chasing a redirect link, not as part of the normal sequential scan.
121 : * We still vacuum the page normally, but we don't increment the stats
122 : * about live tuples; else we'd double-count those tuples, since the page
123 : * has been or will be visited in the sequential scan as well.
124 : */
125 : static void
126 554 : vacuumLeafPage(spgBulkDeleteState *bds, Relation index, Buffer buffer,
127 : bool forPending)
128 : {
129 554 : Page page = BufferGetPage(buffer);
130 : spgxlogVacuumLeaf xlrec;
131 : OffsetNumber toDead[MaxIndexTuplesPerPage];
132 : OffsetNumber toPlaceholder[MaxIndexTuplesPerPage];
133 : OffsetNumber moveSrc[MaxIndexTuplesPerPage];
134 : OffsetNumber moveDest[MaxIndexTuplesPerPage];
135 : OffsetNumber chainSrc[MaxIndexTuplesPerPage];
136 : OffsetNumber chainDest[MaxIndexTuplesPerPage];
137 : OffsetNumber predecessor[MaxIndexTuplesPerPage + 1];
138 : bool deletable[MaxIndexTuplesPerPage + 1];
139 : int nDeletable;
140 : OffsetNumber i,
141 554 : max = PageGetMaxOffsetNumber(page);
142 :
143 554 : memset(predecessor, 0, sizeof(predecessor));
144 554 : memset(deletable, 0, sizeof(deletable));
145 554 : nDeletable = 0;
146 :
147 : /* Scan page, identify tuples to delete, accumulate stats */
148 113312 : for (i = FirstOffsetNumber; i <= max; i++)
149 : {
150 : SpGistLeafTuple lt;
151 :
152 112758 : lt = (SpGistLeafTuple) PageGetItem(page,
153 : PageGetItemId(page, i));
154 112758 : if (lt->tupstate == SPGIST_LIVE)
155 : {
156 85470 : Assert(ItemPointerIsValid(<->heapPtr));
157 :
158 85470 : if (bds->callback(<->heapPtr, bds->callback_state))
159 : {
160 15009 : bds->stats->tuples_removed += 1;
161 15009 : deletable[i] = true;
162 15009 : nDeletable++;
163 : }
164 : else
165 : {
166 70461 : if (!forPending)
167 70461 : bds->stats->num_index_tuples += 1;
168 : }
169 :
170 : /* Form predecessor map, too */
171 85470 : if (lt->nextOffset != InvalidOffsetNumber)
172 : {
173 : /* paranoia about corrupted chain links */
174 168512 : if (lt->nextOffset < FirstOffsetNumber ||
175 168512 : lt->nextOffset > max ||
176 84256 : predecessor[lt->nextOffset] != InvalidOffsetNumber)
177 0 : elog(ERROR, "inconsistent tuple chain links in page %u of index \"%s\"",
178 : BufferGetBlockNumber(buffer),
179 : RelationGetRelationName(index));
180 84256 : predecessor[lt->nextOffset] = i;
181 : }
182 : }
183 27288 : else if (lt->tupstate == SPGIST_REDIRECT)
184 : {
185 116 : SpGistDeadTuple dt = (SpGistDeadTuple) lt;
186 :
187 116 : Assert(dt->nextOffset == InvalidOffsetNumber);
188 116 : Assert(ItemPointerIsValid(&dt->pointer));
189 :
190 : /*
191 : * Add target TID to pending list if the redirection could have
192 : * happened since VACUUM started.
193 : *
194 : * Note: we could make a tighter test by seeing if the xid is
195 : * "running" according to the active snapshot; but tqual.c doesn't
196 : * currently export a suitable API, and it's not entirely clear
197 : * that a tighter test is worth the cycles anyway.
198 : */
199 116 : if (TransactionIdFollowsOrEquals(dt->xid, bds->myXmin))
200 0 : spgAddPendingTID(bds, &dt->pointer);
201 : }
202 : else
203 : {
204 27172 : Assert(lt->nextOffset == InvalidOffsetNumber);
205 : }
206 : }
207 :
208 554 : if (nDeletable == 0)
209 990 : return; /* nothing more to do */
210 :
211 : /*----------
212 : * Figure out exactly what we have to do. We do this separately from
213 : * actually modifying the page, mainly so that we have a representation
214 : * that can be dumped into WAL and then the replay code can do exactly
215 : * the same thing. The output of this step consists of six arrays
216 : * describing four kinds of operations, to be performed in this order:
217 : *
218 : * toDead[]: tuple numbers to be replaced with DEAD tuples
219 : * toPlaceholder[]: tuple numbers to be replaced with PLACEHOLDER tuples
220 : * moveSrc[]: tuple numbers that need to be relocated to another offset
221 : * (replacing the tuple there) and then replaced with PLACEHOLDER tuples
222 : * moveDest[]: new locations for moveSrc tuples
223 : * chainSrc[]: tuple numbers whose chain links (nextOffset) need updates
224 : * chainDest[]: new values of nextOffset for chainSrc members
225 : *
226 : * It's easiest to figure out what we have to do by processing tuple
227 : * chains, so we iterate over all the tuples (not just the deletable
228 : * ones!) to identify chain heads, then chase down each chain and make
229 : * work item entries for deletable tuples within the chain.
230 : *----------
231 : */
232 118 : xlrec.nDead = xlrec.nPlaceholder = xlrec.nMove = xlrec.nChain = 0;
233 :
234 20391 : for (i = FirstOffsetNumber; i <= max; i++)
235 : {
236 : SpGistLeafTuple head;
237 : bool interveningDeletable;
238 : OffsetNumber prevLive;
239 : OffsetNumber j;
240 :
241 20273 : head = (SpGistLeafTuple) PageGetItem(page,
242 : PageGetItemId(page, i));
243 20273 : if (head->tupstate != SPGIST_LIVE)
244 21 : continue; /* can't be a chain member */
245 20252 : if (predecessor[i] != 0)
246 20133 : continue; /* not a chain head */
247 :
248 : /* initialize ... */
249 119 : interveningDeletable = false;
250 119 : prevLive = deletable[i] ? InvalidOffsetNumber : i;
251 :
252 : /* scan down the chain ... */
253 119 : j = head->nextOffset;
254 20371 : while (j != InvalidOffsetNumber)
255 : {
256 : SpGistLeafTuple lt;
257 :
258 20133 : lt = (SpGistLeafTuple) PageGetItem(page,
259 : PageGetItemId(page, j));
260 20133 : if (lt->tupstate != SPGIST_LIVE)
261 : {
262 : /* all tuples in chain should be live */
263 0 : elog(ERROR, "unexpected SPGiST tuple state: %d",
264 : lt->tupstate);
265 : }
266 :
267 20133 : if (deletable[j])
268 : {
269 : /* This tuple should be replaced by a placeholder */
270 14892 : toPlaceholder[xlrec.nPlaceholder] = j;
271 14892 : xlrec.nPlaceholder++;
272 : /* previous live tuple's chain link will need an update */
273 14892 : interveningDeletable = true;
274 : }
275 5241 : else if (prevLive == InvalidOffsetNumber)
276 : {
277 : /*
278 : * This is the first live tuple in the chain. It has to move
279 : * to the head position.
280 : */
281 117 : moveSrc[xlrec.nMove] = j;
282 117 : moveDest[xlrec.nMove] = i;
283 117 : xlrec.nMove++;
284 : /* Chain updates will be applied after the move */
285 117 : prevLive = i;
286 117 : interveningDeletable = false;
287 : }
288 : else
289 : {
290 : /*
291 : * Second or later live tuple. Arrange to re-chain it to the
292 : * previous live one, if there was a gap.
293 : */
294 5124 : if (interveningDeletable)
295 : {
296 4882 : chainSrc[xlrec.nChain] = prevLive;
297 4882 : chainDest[xlrec.nChain] = j;
298 4882 : xlrec.nChain++;
299 : }
300 5124 : prevLive = j;
301 5124 : interveningDeletable = false;
302 : }
303 :
304 20133 : j = lt->nextOffset;
305 : }
306 :
307 119 : if (prevLive == InvalidOffsetNumber)
308 : {
309 : /* The chain is entirely removable, so we need a DEAD tuple */
310 0 : toDead[xlrec.nDead] = i;
311 0 : xlrec.nDead++;
312 : }
313 119 : else if (interveningDeletable)
314 : {
315 : /* One or more deletions at end of chain, so close it off */
316 119 : chainSrc[xlrec.nChain] = prevLive;
317 119 : chainDest[xlrec.nChain] = InvalidOffsetNumber;
318 119 : xlrec.nChain++;
319 : }
320 : }
321 :
322 : /* sanity check ... */
323 118 : if (nDeletable != xlrec.nDead + xlrec.nPlaceholder + xlrec.nMove)
324 0 : elog(ERROR, "inconsistent counts of deletable tuples");
325 :
326 : /* Do the updates */
327 118 : START_CRIT_SECTION();
328 :
329 118 : spgPageIndexMultiDelete(&bds->spgstate, page,
330 118 : toDead, xlrec.nDead,
331 : SPGIST_DEAD, SPGIST_DEAD,
332 : InvalidBlockNumber, InvalidOffsetNumber);
333 :
334 118 : spgPageIndexMultiDelete(&bds->spgstate, page,
335 118 : toPlaceholder, xlrec.nPlaceholder,
336 : SPGIST_PLACEHOLDER, SPGIST_PLACEHOLDER,
337 : InvalidBlockNumber, InvalidOffsetNumber);
338 :
339 : /*
340 : * We implement the move step by swapping the item pointers of the source
341 : * and target tuples, then replacing the newly-source tuples with
342 : * placeholders. This is perhaps unduly friendly with the page data
343 : * representation, but it's fast and doesn't risk page overflow when a
344 : * tuple to be relocated is large.
345 : */
346 235 : for (i = 0; i < xlrec.nMove; i++)
347 : {
348 117 : ItemId idSrc = PageGetItemId(page, moveSrc[i]);
349 117 : ItemId idDest = PageGetItemId(page, moveDest[i]);
350 : ItemIdData tmp;
351 :
352 117 : tmp = *idSrc;
353 117 : *idSrc = *idDest;
354 117 : *idDest = tmp;
355 : }
356 :
357 118 : spgPageIndexMultiDelete(&bds->spgstate, page,
358 118 : moveSrc, xlrec.nMove,
359 : SPGIST_PLACEHOLDER, SPGIST_PLACEHOLDER,
360 : InvalidBlockNumber, InvalidOffsetNumber);
361 :
362 5119 : for (i = 0; i < xlrec.nChain; i++)
363 : {
364 : SpGistLeafTuple lt;
365 :
366 5001 : lt = (SpGistLeafTuple) PageGetItem(page,
367 : PageGetItemId(page, chainSrc[i]));
368 5001 : Assert(lt->tupstate == SPGIST_LIVE);
369 5001 : lt->nextOffset = chainDest[i];
370 : }
371 :
372 118 : MarkBufferDirty(buffer);
373 :
374 118 : if (RelationNeedsWAL(index))
375 : {
376 : XLogRecPtr recptr;
377 :
378 118 : XLogBeginInsert();
379 :
380 118 : STORE_STATE(&bds->spgstate, xlrec.stateSrc);
381 :
382 118 : XLogRegisterData((char *) &xlrec, SizeOfSpgxlogVacuumLeaf);
383 : /* sizeof(xlrec) should be a multiple of sizeof(OffsetNumber) */
384 118 : XLogRegisterData((char *) toDead, sizeof(OffsetNumber) * xlrec.nDead);
385 118 : XLogRegisterData((char *) toPlaceholder, sizeof(OffsetNumber) * xlrec.nPlaceholder);
386 118 : XLogRegisterData((char *) moveSrc, sizeof(OffsetNumber) * xlrec.nMove);
387 118 : XLogRegisterData((char *) moveDest, sizeof(OffsetNumber) * xlrec.nMove);
388 118 : XLogRegisterData((char *) chainSrc, sizeof(OffsetNumber) * xlrec.nChain);
389 118 : XLogRegisterData((char *) chainDest, sizeof(OffsetNumber) * xlrec.nChain);
390 :
391 118 : XLogRegisterBuffer(0, buffer, REGBUF_STANDARD);
392 :
393 118 : recptr = XLogInsert(RM_SPGIST_ID, XLOG_SPGIST_VACUUM_LEAF);
394 :
395 118 : PageSetLSN(page, recptr);
396 : }
397 :
398 118 : END_CRIT_SECTION();
399 : }
400 :
401 : /*
402 : * Vacuum a root page when it is also a leaf
403 : *
404 : * On the root, we just delete any dead leaf tuples; no fancy business
405 : */
406 : static void
407 4 : vacuumLeafRoot(spgBulkDeleteState *bds, Relation index, Buffer buffer)
408 : {
409 4 : Page page = BufferGetPage(buffer);
410 : spgxlogVacuumRoot xlrec;
411 : OffsetNumber toDelete[MaxIndexTuplesPerPage];
412 : OffsetNumber i,
413 4 : max = PageGetMaxOffsetNumber(page);
414 :
415 4 : xlrec.nDelete = 0;
416 :
417 : /* Scan page, identify tuples to delete, accumulate stats */
418 22 : for (i = FirstOffsetNumber; i <= max; i++)
419 : {
420 : SpGistLeafTuple lt;
421 :
422 18 : lt = (SpGistLeafTuple) PageGetItem(page,
423 : PageGetItemId(page, i));
424 18 : if (lt->tupstate == SPGIST_LIVE)
425 : {
426 18 : Assert(ItemPointerIsValid(<->heapPtr));
427 :
428 18 : if (bds->callback(<->heapPtr, bds->callback_state))
429 : {
430 0 : bds->stats->tuples_removed += 1;
431 0 : toDelete[xlrec.nDelete] = i;
432 0 : xlrec.nDelete++;
433 : }
434 : else
435 : {
436 18 : bds->stats->num_index_tuples += 1;
437 : }
438 : }
439 : else
440 : {
441 : /* all tuples on root should be live */
442 0 : elog(ERROR, "unexpected SPGiST tuple state: %d",
443 : lt->tupstate);
444 : }
445 : }
446 :
447 4 : if (xlrec.nDelete == 0)
448 8 : return; /* nothing more to do */
449 :
450 : /* Do the update */
451 0 : START_CRIT_SECTION();
452 :
453 : /* The tuple numbers are in order, so we can use PageIndexMultiDelete */
454 0 : PageIndexMultiDelete(page, toDelete, xlrec.nDelete);
455 :
456 0 : MarkBufferDirty(buffer);
457 :
458 0 : if (RelationNeedsWAL(index))
459 : {
460 : XLogRecPtr recptr;
461 :
462 0 : XLogBeginInsert();
463 :
464 : /* Prepare WAL record */
465 0 : STORE_STATE(&bds->spgstate, xlrec.stateSrc);
466 :
467 0 : XLogRegisterData((char *) &xlrec, SizeOfSpgxlogVacuumRoot);
468 : /* sizeof(xlrec) should be a multiple of sizeof(OffsetNumber) */
469 0 : XLogRegisterData((char *) toDelete,
470 0 : sizeof(OffsetNumber) * xlrec.nDelete);
471 :
472 0 : XLogRegisterBuffer(0, buffer, REGBUF_STANDARD);
473 :
474 0 : recptr = XLogInsert(RM_SPGIST_ID, XLOG_SPGIST_VACUUM_ROOT);
475 :
476 0 : PageSetLSN(page, recptr);
477 : }
478 :
479 0 : END_CRIT_SECTION();
480 : }
481 :
482 : /*
483 : * Clean up redirect and placeholder tuples on the given page
484 : *
485 : * Redirect tuples can be marked placeholder once they're old enough.
486 : * Placeholder tuples can be removed if it won't change the offsets of
487 : * non-placeholder ones.
488 : *
489 : * Unlike the routines above, this works on both leaf and inner pages.
490 : */
491 : static void
492 570 : vacuumRedirectAndPlaceholder(Relation index, Buffer buffer)
493 : {
494 570 : Page page = BufferGetPage(buffer);
495 570 : SpGistPageOpaque opaque = SpGistPageGetOpaque(page);
496 : OffsetNumber i,
497 570 : max = PageGetMaxOffsetNumber(page),
498 570 : firstPlaceholder = InvalidOffsetNumber;
499 570 : bool hasNonPlaceholder = false;
500 570 : bool hasUpdate = false;
501 : OffsetNumber itemToPlaceholder[MaxIndexTuplesPerPage];
502 : OffsetNumber itemnos[MaxIndexTuplesPerPage];
503 : spgxlogVacuumRedirect xlrec;
504 :
505 570 : xlrec.nToPlaceholder = 0;
506 570 : xlrec.newestRedirectXid = InvalidTransactionId;
507 :
508 570 : START_CRIT_SECTION();
509 :
510 : /*
511 : * Scan backwards to convert old redirection tuples to placeholder tuples,
512 : * and identify location of last non-placeholder tuple while at it.
513 : */
514 23585 : for (i = max;
515 23008 : i >= FirstOffsetNumber &&
516 36208 : (opaque->nRedirection > 0 || !hasNonPlaceholder);
517 22445 : i--)
518 : {
519 : SpGistDeadTuple dt;
520 :
521 22445 : dt = (SpGistDeadTuple) PageGetItem(page, PageGetItemId(page, i));
522 :
523 22561 : if (dt->tupstate == SPGIST_REDIRECT &&
524 116 : TransactionIdPrecedes(dt->xid, RecentGlobalXmin))
525 : {
526 116 : dt->tupstate = SPGIST_PLACEHOLDER;
527 116 : Assert(opaque->nRedirection > 0);
528 116 : opaque->nRedirection--;
529 116 : opaque->nPlaceholder++;
530 :
531 : /* remember newest XID among the removed redirects */
532 116 : if (!TransactionIdIsValid(xlrec.newestRedirectXid) ||
533 0 : TransactionIdPrecedes(xlrec.newestRedirectXid, dt->xid))
534 116 : xlrec.newestRedirectXid = dt->xid;
535 :
536 116 : ItemPointerSetInvalid(&dt->pointer);
537 :
538 116 : itemToPlaceholder[xlrec.nToPlaceholder] = i;
539 116 : xlrec.nToPlaceholder++;
540 :
541 116 : hasUpdate = true;
542 : }
543 :
544 22445 : if (dt->tupstate == SPGIST_PLACEHOLDER)
545 : {
546 21807 : if (!hasNonPlaceholder)
547 21799 : firstPlaceholder = i;
548 : }
549 : else
550 : {
551 638 : hasNonPlaceholder = true;
552 : }
553 : }
554 :
555 : /*
556 : * Any placeholder tuples at the end of page can safely be removed. We
557 : * can't remove ones before the last non-placeholder, though, because we
558 : * can't alter the offset numbers of non-placeholder tuples.
559 : */
560 570 : if (firstPlaceholder != InvalidOffsetNumber)
561 : {
562 : /*
563 : * We do not store this array to rdata because it's easy to recreate.
564 : */
565 22088 : for (i = firstPlaceholder; i <= max; i++)
566 21799 : itemnos[i - firstPlaceholder] = i;
567 :
568 289 : i = max - firstPlaceholder + 1;
569 289 : Assert(opaque->nPlaceholder >= i);
570 289 : opaque->nPlaceholder -= i;
571 :
572 : /* The array is surely sorted, so can use PageIndexMultiDelete */
573 289 : PageIndexMultiDelete(page, itemnos, i);
574 :
575 289 : hasUpdate = true;
576 : }
577 :
578 570 : xlrec.firstPlaceholder = firstPlaceholder;
579 :
580 570 : if (hasUpdate)
581 289 : MarkBufferDirty(buffer);
582 :
583 570 : if (hasUpdate && RelationNeedsWAL(index))
584 : {
585 : XLogRecPtr recptr;
586 :
587 289 : XLogBeginInsert();
588 :
589 289 : XLogRegisterData((char *) &xlrec, SizeOfSpgxlogVacuumRedirect);
590 289 : XLogRegisterData((char *) itemToPlaceholder,
591 289 : sizeof(OffsetNumber) * xlrec.nToPlaceholder);
592 :
593 289 : XLogRegisterBuffer(0, buffer, REGBUF_STANDARD);
594 :
595 289 : recptr = XLogInsert(RM_SPGIST_ID, XLOG_SPGIST_VACUUM_REDIRECT);
596 :
597 289 : PageSetLSN(page, recptr);
598 : }
599 :
600 570 : END_CRIT_SECTION();
601 570 : }
602 :
603 : /*
604 : * Process one page during a bulkdelete scan
605 : */
606 : static void
607 584 : spgvacuumpage(spgBulkDeleteState *bds, BlockNumber blkno)
608 : {
609 584 : Relation index = bds->info->index;
610 : Buffer buffer;
611 : Page page;
612 :
613 : /* call vacuum_delay_point while not holding any buffer lock */
614 584 : vacuum_delay_point();
615 :
616 584 : buffer = ReadBufferExtended(index, MAIN_FORKNUM, blkno,
617 584 : RBM_NORMAL, bds->info->strategy);
618 584 : LockBuffer(buffer, BUFFER_LOCK_EXCLUSIVE);
619 584 : page = (Page) BufferGetPage(buffer);
620 :
621 584 : if (PageIsNew(page))
622 : {
623 : /*
624 : * We found an all-zero page, which could happen if the database
625 : * crashed just after extending the file. Recycle it.
626 : */
627 : }
628 584 : else if (PageIsEmpty(page))
629 : {
630 : /* nothing to do */
631 : }
632 574 : else if (SpGistPageIsLeaf(page))
633 : {
634 558 : if (SpGistBlockIsRoot(blkno))
635 : {
636 4 : vacuumLeafRoot(bds, index, buffer);
637 : /* no need for vacuumRedirectAndPlaceholder */
638 : }
639 : else
640 : {
641 554 : vacuumLeafPage(bds, index, buffer, false);
642 554 : vacuumRedirectAndPlaceholder(index, buffer);
643 : }
644 : }
645 : else
646 : {
647 : /* inner page */
648 16 : vacuumRedirectAndPlaceholder(index, buffer);
649 : }
650 :
651 : /*
652 : * The root pages must never be deleted, nor marked as available in FSM,
653 : * because we don't want them ever returned by a search for a place to put
654 : * a new tuple. Otherwise, check for empty page, and make sure the FSM
655 : * knows about it.
656 : */
657 584 : if (!SpGistBlockIsRoot(blkno))
658 : {
659 568 : if (PageIsNew(page) || PageIsEmpty(page))
660 : {
661 5 : RecordFreeIndexPage(index, blkno);
662 5 : bds->stats->pages_deleted++;
663 : }
664 : else
665 : {
666 563 : SpGistSetLastUsedPage(index, buffer);
667 563 : bds->lastFilledBlock = blkno;
668 : }
669 : }
670 :
671 584 : UnlockReleaseBuffer(buffer);
672 584 : }
673 :
674 : /*
675 : * Process the pending-TID list between pages of the main scan
676 : */
677 : static void
678 0 : spgprocesspending(spgBulkDeleteState *bds)
679 : {
680 0 : Relation index = bds->info->index;
681 : spgVacPendingItem *pitem;
682 : spgVacPendingItem *nitem;
683 : BlockNumber blkno;
684 : Buffer buffer;
685 : Page page;
686 :
687 0 : for (pitem = bds->pendingList; pitem != NULL; pitem = pitem->next)
688 : {
689 0 : if (pitem->done)
690 0 : continue; /* ignore already-done items */
691 :
692 : /* call vacuum_delay_point while not holding any buffer lock */
693 0 : vacuum_delay_point();
694 :
695 : /* examine the referenced page */
696 0 : blkno = ItemPointerGetBlockNumber(&pitem->tid);
697 0 : buffer = ReadBufferExtended(index, MAIN_FORKNUM, blkno,
698 0 : RBM_NORMAL, bds->info->strategy);
699 0 : LockBuffer(buffer, BUFFER_LOCK_EXCLUSIVE);
700 0 : page = (Page) BufferGetPage(buffer);
701 :
702 0 : if (PageIsNew(page) || SpGistPageIsDeleted(page))
703 : {
704 : /* Probably shouldn't happen, but ignore it */
705 : }
706 0 : else if (SpGistPageIsLeaf(page))
707 : {
708 0 : if (SpGistBlockIsRoot(blkno))
709 : {
710 : /* this should definitely not happen */
711 0 : elog(ERROR, "redirection leads to root page of index \"%s\"",
712 : RelationGetRelationName(index));
713 : }
714 :
715 : /* deal with any deletable tuples */
716 0 : vacuumLeafPage(bds, index, buffer, true);
717 : /* might as well do this while we are here */
718 0 : vacuumRedirectAndPlaceholder(index, buffer);
719 :
720 0 : SpGistSetLastUsedPage(index, buffer);
721 :
722 : /*
723 : * We can mark as done not only this item, but any later ones
724 : * pointing at the same page, since we vacuumed the whole page.
725 : */
726 0 : pitem->done = true;
727 0 : for (nitem = pitem->next; nitem != NULL; nitem = nitem->next)
728 : {
729 0 : if (ItemPointerGetBlockNumber(&nitem->tid) == blkno)
730 0 : nitem->done = true;
731 : }
732 : }
733 : else
734 : {
735 : /*
736 : * On an inner page, visit the referenced inner tuple and add all
737 : * its downlinks to the pending list. We might have pending items
738 : * for more than one inner tuple on the same page (in fact this is
739 : * pretty likely given the way space allocation works), so get
740 : * them all while we are here.
741 : */
742 0 : for (nitem = pitem; nitem != NULL; nitem = nitem->next)
743 : {
744 0 : if (nitem->done)
745 0 : continue;
746 0 : if (ItemPointerGetBlockNumber(&nitem->tid) == blkno)
747 : {
748 : OffsetNumber offset;
749 : SpGistInnerTuple innerTuple;
750 :
751 0 : offset = ItemPointerGetOffsetNumber(&nitem->tid);
752 0 : innerTuple = (SpGistInnerTuple) PageGetItem(page,
753 : PageGetItemId(page, offset));
754 0 : if (innerTuple->tupstate == SPGIST_LIVE)
755 : {
756 : SpGistNodeTuple node;
757 : int i;
758 :
759 0 : SGITITERATE(innerTuple, i, node)
760 : {
761 0 : if (ItemPointerIsValid(&node->t_tid))
762 0 : spgAddPendingTID(bds, &node->t_tid);
763 : }
764 : }
765 0 : else if (innerTuple->tupstate == SPGIST_REDIRECT)
766 : {
767 : /* transfer attention to redirect point */
768 0 : spgAddPendingTID(bds,
769 : &((SpGistDeadTuple) innerTuple)->pointer);
770 : }
771 : else
772 0 : elog(ERROR, "unexpected SPGiST tuple state: %d",
773 : innerTuple->tupstate);
774 :
775 0 : nitem->done = true;
776 : }
777 : }
778 : }
779 :
780 0 : UnlockReleaseBuffer(buffer);
781 : }
782 :
783 0 : spgClearPendingList(bds);
784 0 : }
785 :
786 : /*
787 : * Perform a bulkdelete scan
788 : */
789 : static void
790 8 : spgvacuumscan(spgBulkDeleteState *bds)
791 : {
792 8 : Relation index = bds->info->index;
793 : bool needLock;
794 : BlockNumber num_pages,
795 : blkno;
796 :
797 : /* Finish setting up spgBulkDeleteState */
798 8 : initSpGistState(&bds->spgstate, index);
799 8 : bds->pendingList = NULL;
800 8 : bds->myXmin = GetActiveSnapshot()->xmin;
801 8 : bds->lastFilledBlock = SPGIST_LAST_FIXED_BLKNO;
802 :
803 : /*
804 : * Reset counts that will be incremented during the scan; needed in case
805 : * of multiple scans during a single VACUUM command
806 : */
807 8 : bds->stats->estimated_count = false;
808 8 : bds->stats->num_index_tuples = 0;
809 8 : bds->stats->pages_deleted = 0;
810 :
811 : /* We can skip locking for new or temp relations */
812 8 : needLock = !RELATION_IS_LOCAL(index);
813 :
814 : /*
815 : * The outer loop iterates over all index pages except the metapage, in
816 : * physical order (we hope the kernel will cooperate in providing
817 : * read-ahead for speed). It is critical that we visit all leaf pages,
818 : * including ones added after we start the scan, else we might fail to
819 : * delete some deletable tuples. See more extensive comments about this
820 : * in btvacuumscan().
821 : */
822 8 : blkno = SPGIST_METAPAGE_BLKNO + 1;
823 : for (;;)
824 : {
825 : /* Get the current relation length */
826 16 : if (needLock)
827 16 : LockRelationForExtension(index, ExclusiveLock);
828 16 : num_pages = RelationGetNumberOfBlocks(index);
829 16 : if (needLock)
830 16 : UnlockRelationForExtension(index, ExclusiveLock);
831 :
832 : /* Quit if we've scanned the whole relation */
833 16 : if (blkno >= num_pages)
834 8 : break;
835 : /* Iterate over pages, then loop back to recheck length */
836 592 : for (; blkno < num_pages; blkno++)
837 : {
838 584 : spgvacuumpage(bds, blkno);
839 : /* empty the pending-list after each page */
840 584 : if (bds->pendingList != NULL)
841 0 : spgprocesspending(bds);
842 : }
843 8 : }
844 :
845 : /* Propagate local lastUsedPage cache to metablock */
846 8 : SpGistUpdateMetaPage(index);
847 :
848 : /*
849 : * Truncate index if possible
850 : *
851 : * XXX disabled because it's unsafe due to possible concurrent inserts.
852 : * We'd have to rescan the pages to make sure they're still empty, and it
853 : * doesn't seem worth it. Note that btree doesn't do this either.
854 : *
855 : * Another reason not to truncate is that it could invalidate the cached
856 : * pages-with-freespace pointers in the metapage and other backends'
857 : * relation caches, that is leave them pointing to nonexistent pages.
858 : * Adding RelationGetNumberOfBlocks calls to protect the places that use
859 : * those pointers would be unduly expensive.
860 : */
861 : #ifdef NOT_USED
862 : if (num_pages > bds->lastFilledBlock + 1)
863 : {
864 : BlockNumber lastBlock = num_pages - 1;
865 :
866 : num_pages = bds->lastFilledBlock + 1;
867 : RelationTruncate(index, num_pages);
868 : bds->stats->pages_removed += lastBlock - bds->lastFilledBlock;
869 : bds->stats->pages_deleted -= lastBlock - bds->lastFilledBlock;
870 : }
871 : #endif
872 :
873 : /* Report final stats */
874 8 : bds->stats->num_pages = num_pages;
875 8 : bds->stats->pages_free = bds->stats->pages_deleted;
876 8 : }
877 :
878 : /*
879 : * Bulk deletion of all index entries pointing to a set of heap tuples.
880 : * The set of target tuples is specified via a callback routine that tells
881 : * whether any given heap tuple (identified by ItemPointer) is being deleted.
882 : *
883 : * Result: a palloc'd struct containing statistical info for VACUUM displays.
884 : */
885 : IndexBulkDeleteResult *
886 2 : spgbulkdelete(IndexVacuumInfo *info, IndexBulkDeleteResult *stats,
887 : IndexBulkDeleteCallback callback, void *callback_state)
888 : {
889 : spgBulkDeleteState bds;
890 :
891 : /* allocate stats if first time through, else re-use existing struct */
892 2 : if (stats == NULL)
893 2 : stats = (IndexBulkDeleteResult *) palloc0(sizeof(IndexBulkDeleteResult));
894 2 : bds.info = info;
895 2 : bds.stats = stats;
896 2 : bds.callback = callback;
897 2 : bds.callback_state = callback_state;
898 :
899 2 : spgvacuumscan(&bds);
900 :
901 2 : return stats;
902 : }
903 :
904 : /* Dummy callback to delete no tuples during spgvacuumcleanup */
905 : static bool
906 45472 : dummy_callback(ItemPointer itemptr, void *state)
907 : {
908 45472 : return false;
909 : }
910 :
911 : /*
912 : * Post-VACUUM cleanup.
913 : *
914 : * Result: a palloc'd struct containing statistical info for VACUUM displays.
915 : */
916 : IndexBulkDeleteResult *
917 13 : spgvacuumcleanup(IndexVacuumInfo *info, IndexBulkDeleteResult *stats)
918 : {
919 13 : Relation index = info->index;
920 : spgBulkDeleteState bds;
921 :
922 : /* No-op in ANALYZE ONLY mode */
923 13 : if (info->analyze_only)
924 5 : return stats;
925 :
926 : /*
927 : * We don't need to scan the index if there was a preceding bulkdelete
928 : * pass. Otherwise, make a pass that won't delete any live tuples, but
929 : * might still accomplish useful stuff with redirect/placeholder cleanup,
930 : * and in any case will provide stats.
931 : */
932 8 : if (stats == NULL)
933 : {
934 6 : stats = (IndexBulkDeleteResult *) palloc0(sizeof(IndexBulkDeleteResult));
935 6 : bds.info = info;
936 6 : bds.stats = stats;
937 6 : bds.callback = dummy_callback;
938 6 : bds.callback_state = NULL;
939 :
940 6 : spgvacuumscan(&bds);
941 : }
942 :
943 : /* Finally, vacuum the FSM */
944 8 : IndexFreeSpaceMapVacuum(index);
945 :
946 : /*
947 : * It's quite possible for us to be fooled by concurrent tuple moves into
948 : * double-counting some index tuples, so disbelieve any total that exceeds
949 : * the underlying heap's count ... if we know that accurately. Otherwise
950 : * this might just make matters worse.
951 : */
952 8 : if (!info->estimated_count)
953 : {
954 8 : if (stats->num_index_tuples > info->num_heap_tuples)
955 0 : stats->num_index_tuples = info->num_heap_tuples;
956 : }
957 :
958 8 : return stats;
959 : }
|
