Line data Source code
1 : /*-------------------------------------------------------------------------
2 : *
3 : * hashinsert.c
4 : * Item insertion in hash tables for Postgres.
5 : *
6 : * Portions Copyright (c) 1996-2026, PostgreSQL Global Development Group
7 : * Portions Copyright (c) 1994, Regents of the University of California
8 : *
9 : *
10 : * IDENTIFICATION
11 : * src/backend/access/hash/hashinsert.c
12 : *
13 : *-------------------------------------------------------------------------
14 : */
15 :
16 : #include "postgres.h"
17 :
18 : #include "access/hash.h"
19 : #include "access/hash_xlog.h"
20 : #include "access/xloginsert.h"
21 : #include "miscadmin.h"
22 : #include "storage/predicate.h"
23 : #include "utils/rel.h"
24 :
25 : static void _hash_vacuum_one_page(Relation rel, Relation hrel,
26 : Buffer metabuf, Buffer buf);
27 :
28 : /*
29 : * _hash_doinsert() -- Handle insertion of a single index tuple.
30 : *
31 : * This routine is called by the public interface routines, hashbuild
32 : * and hashinsert. By here, itup is completely filled in.
33 : *
34 : * 'sorted' must only be passed as 'true' when inserts are done in hashkey
35 : * order.
36 : */
37 : void
38 357419 : _hash_doinsert(Relation rel, IndexTuple itup, Relation heapRel, bool sorted)
39 : {
40 357419 : Buffer buf = InvalidBuffer;
41 : Buffer bucket_buf;
42 : Buffer metabuf;
43 : HashMetaPage metap;
44 357419 : HashMetaPage usedmetap = NULL;
45 : Page metapage;
46 : Page page;
47 : HashPageOpaque pageopaque;
48 : Size itemsz;
49 : bool do_expand;
50 : uint32 hashkey;
51 : Bucket bucket;
52 : OffsetNumber itup_off;
53 :
54 : /*
55 : * Get the hash key for the item (it's stored in the index tuple itself).
56 : */
57 357419 : hashkey = _hash_get_indextuple_hashkey(itup);
58 :
59 : /* compute item size too */
60 357419 : itemsz = IndexTupleSize(itup);
61 357419 : itemsz = MAXALIGN(itemsz); /* be safe, PageAddItem will do this but we
62 : * need to be consistent */
63 :
64 357419 : restart_insert:
65 :
66 : /*
67 : * Read the metapage. We don't lock it yet; HashMaxItemSize() will
68 : * examine pd_pagesize_version, but that can't change so we can examine it
69 : * without a lock.
70 : */
71 357419 : metabuf = _hash_getbuf(rel, HASH_METAPAGE, HASH_NOLOCK, LH_META_PAGE);
72 357419 : metapage = BufferGetPage(metabuf);
73 :
74 : /*
75 : * Check whether the item can fit on a hash page at all. (Eventually, we
76 : * ought to try to apply TOAST methods if not.) Note that at this point,
77 : * itemsz doesn't include the ItemId.
78 : *
79 : * XXX this is useless code if we are only storing hash keys.
80 : */
81 357419 : if (itemsz > HashMaxItemSize(metapage))
82 0 : ereport(ERROR,
83 : (errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED),
84 : errmsg("index row size %zu exceeds hash maximum %zu",
85 : itemsz, HashMaxItemSize(metapage)),
86 : errhint("Values larger than a buffer page cannot be indexed.")));
87 :
88 : /* Lock the primary bucket page for the target bucket. */
89 357419 : buf = _hash_getbucketbuf_from_hashkey(rel, hashkey, HASH_WRITE,
90 : &usedmetap);
91 : Assert(usedmetap != NULL);
92 :
93 357419 : CheckForSerializableConflictIn(rel, NULL, BufferGetBlockNumber(buf));
94 :
95 : /* remember the primary bucket buffer to release the pin on it at end. */
96 357413 : bucket_buf = buf;
97 :
98 357413 : page = BufferGetPage(buf);
99 357413 : pageopaque = HashPageGetOpaque(page);
100 357413 : bucket = pageopaque->hasho_bucket;
101 :
102 : /*
103 : * If this bucket is in the process of being split, try to finish the
104 : * split before inserting, because that might create room for the
105 : * insertion to proceed without allocating an additional overflow page.
106 : * It's only interesting to finish the split if we're trying to insert
107 : * into the bucket from which we're removing tuples (the "old" bucket),
108 : * not if we're trying to insert into the bucket into which tuples are
109 : * being moved (the "new" bucket).
110 : */
111 357413 : if (H_BUCKET_BEING_SPLIT(pageopaque) && IsBufferCleanupOK(buf))
112 : {
113 : /* release the lock on bucket buffer, before completing the split. */
114 0 : LockBuffer(buf, BUFFER_LOCK_UNLOCK);
115 :
116 0 : _hash_finish_split(rel, metabuf, buf, bucket,
117 0 : usedmetap->hashm_maxbucket,
118 0 : usedmetap->hashm_highmask,
119 0 : usedmetap->hashm_lowmask);
120 :
121 : /* release the pin on old and meta buffer. retry for insert. */
122 0 : _hash_dropbuf(rel, buf);
123 0 : _hash_dropbuf(rel, metabuf);
124 0 : goto restart_insert;
125 : }
126 :
127 : /* Do the insertion */
128 582126 : while (PageGetFreeSpace(page) < itemsz)
129 : {
130 : BlockNumber nextblkno;
131 :
132 : /*
133 : * Check if current page has any DEAD tuples. If yes, delete these
134 : * tuples and see if we can get a space for the new item to be
135 : * inserted before moving to the next page in the bucket chain.
136 : */
137 224713 : if (H_HAS_DEAD_TUPLES(pageopaque))
138 : {
139 :
140 0 : if (IsBufferCleanupOK(buf))
141 : {
142 0 : _hash_vacuum_one_page(rel, heapRel, metabuf, buf);
143 :
144 0 : if (PageGetFreeSpace(page) >= itemsz)
145 0 : break; /* OK, now we have enough space */
146 : }
147 : }
148 :
149 : /*
150 : * no space on this page; check for an overflow page
151 : */
152 224713 : nextblkno = pageopaque->hasho_nextblkno;
153 :
154 224713 : if (BlockNumberIsValid(nextblkno))
155 : {
156 : /*
157 : * ovfl page exists; go get it. if it doesn't have room, we'll
158 : * find out next pass through the loop test above. we always
159 : * release both the lock and pin if this is an overflow page, but
160 : * only the lock if this is the primary bucket page, since the pin
161 : * on the primary bucket must be retained throughout the scan.
162 : */
163 224566 : if (buf != bucket_buf)
164 183393 : _hash_relbuf(rel, buf);
165 : else
166 41173 : LockBuffer(buf, BUFFER_LOCK_UNLOCK);
167 224566 : buf = _hash_getbuf(rel, nextblkno, HASH_WRITE, LH_OVERFLOW_PAGE);
168 224566 : page = BufferGetPage(buf);
169 : }
170 : else
171 : {
172 : /*
173 : * we're at the end of the bucket chain and we haven't found a
174 : * page with enough room. allocate a new overflow page.
175 : */
176 :
177 : /* release our write lock without modifying buffer */
178 147 : LockBuffer(buf, BUFFER_LOCK_UNLOCK);
179 :
180 : /* chain to a new overflow page */
181 147 : buf = _hash_addovflpage(rel, metabuf, buf, (buf == bucket_buf));
182 147 : page = BufferGetPage(buf);
183 :
184 : /* should fit now, given test above */
185 : Assert(PageGetFreeSpace(page) >= itemsz);
186 : }
187 224713 : pageopaque = HashPageGetOpaque(page);
188 : Assert((pageopaque->hasho_flag & LH_PAGE_TYPE) == LH_OVERFLOW_PAGE);
189 : Assert(pageopaque->hasho_bucket == bucket);
190 : }
191 :
192 : /*
193 : * Write-lock the metapage so we can increment the tuple count. After
194 : * incrementing it, check to see if it's time for a split.
195 : */
196 357413 : LockBuffer(metabuf, BUFFER_LOCK_EXCLUSIVE);
197 :
198 : /* Do the update. No ereport(ERROR) until changes are logged */
199 357413 : START_CRIT_SECTION();
200 :
201 : /* found page with enough space, so add the item here */
202 357413 : itup_off = _hash_pgaddtup(rel, buf, itemsz, itup, sorted);
203 357413 : MarkBufferDirty(buf);
204 :
205 : /* metapage operations */
206 357413 : metap = HashPageGetMeta(metapage);
207 357413 : metap->hashm_ntuples += 1;
208 :
209 : /* Make sure this stays in sync with _hash_expandtable() */
210 357413 : do_expand = metap->hashm_ntuples >
211 357413 : (double) metap->hashm_ffactor * (metap->hashm_maxbucket + 1);
212 :
213 357413 : MarkBufferDirty(metabuf);
214 :
215 : /* XLOG stuff */
216 357413 : if (RelationNeedsWAL(rel))
217 : {
218 : xl_hash_insert xlrec;
219 : XLogRecPtr recptr;
220 :
221 356790 : xlrec.offnum = itup_off;
222 :
223 356790 : XLogBeginInsert();
224 356790 : XLogRegisterData(&xlrec, SizeOfHashInsert);
225 :
226 356790 : XLogRegisterBuffer(1, metabuf, REGBUF_STANDARD);
227 :
228 356790 : XLogRegisterBuffer(0, buf, REGBUF_STANDARD);
229 356790 : XLogRegisterBufData(0, itup, IndexTupleSize(itup));
230 :
231 356790 : recptr = XLogInsert(RM_HASH_ID, XLOG_HASH_INSERT);
232 :
233 356790 : PageSetLSN(BufferGetPage(buf), recptr);
234 356790 : PageSetLSN(BufferGetPage(metabuf), recptr);
235 : }
236 :
237 357413 : END_CRIT_SECTION();
238 :
239 : /* drop lock on metapage, but keep pin */
240 357413 : LockBuffer(metabuf, BUFFER_LOCK_UNLOCK);
241 :
242 : /*
243 : * Release the modified page and ensure to release the pin on primary
244 : * page.
245 : */
246 357413 : _hash_relbuf(rel, buf);
247 357413 : if (buf != bucket_buf)
248 41224 : _hash_dropbuf(rel, bucket_buf);
249 :
250 : /* Attempt to split if a split is needed */
251 357413 : if (do_expand)
252 546 : _hash_expandtable(rel, metabuf);
253 :
254 : /* Finally drop our pin on the metapage */
255 357413 : _hash_dropbuf(rel, metabuf);
256 357413 : }
257 :
258 : /*
259 : * _hash_pgaddtup() -- add a tuple to a particular page in the index.
260 : *
261 : * This routine adds the tuple to the page as requested; it does not write out
262 : * the page. It is an error to call this function without pin and write lock
263 : * on the target buffer.
264 : *
265 : * Returns the offset number at which the tuple was inserted. This function
266 : * is responsible for preserving the condition that tuples in a hash index
267 : * page are sorted by hashkey value, however, if the caller is certain that
268 : * the hashkey for the tuple being added is >= the hashkeys of all existing
269 : * tuples on the page, then the 'appendtup' flag may be passed as true. This
270 : * saves from having to binary search for the correct location to insert the
271 : * tuple.
272 : */
273 : OffsetNumber
274 357413 : _hash_pgaddtup(Relation rel, Buffer buf, Size itemsize, IndexTuple itup,
275 : bool appendtup)
276 : {
277 : OffsetNumber itup_off;
278 : Page page;
279 :
280 357413 : _hash_checkpage(rel, buf, LH_BUCKET_PAGE | LH_OVERFLOW_PAGE);
281 357413 : page = BufferGetPage(buf);
282 :
283 : /*
284 : * Find where to insert the tuple (preserving page's hashkey ordering). If
285 : * 'appendtup' is true then we just insert it at the end.
286 : */
287 357413 : if (appendtup)
288 : {
289 60500 : itup_off = PageGetMaxOffsetNumber(page) + 1;
290 :
291 : #ifdef USE_ASSERT_CHECKING
292 : /* ensure this tuple's hashkey is >= the final existing tuple */
293 : if (PageGetMaxOffsetNumber(page) > 0)
294 : {
295 : IndexTuple lasttup;
296 : ItemId itemid;
297 :
298 : itemid = PageGetItemId(page, PageGetMaxOffsetNumber(page));
299 : lasttup = (IndexTuple) PageGetItem(page, itemid);
300 :
301 : Assert(_hash_get_indextuple_hashkey(lasttup) <=
302 : _hash_get_indextuple_hashkey(itup));
303 : }
304 : #endif
305 : }
306 : else
307 : {
308 296913 : uint32 hashkey = _hash_get_indextuple_hashkey(itup);
309 :
310 296913 : itup_off = _hash_binsearch(page, hashkey);
311 : }
312 :
313 357413 : if (PageAddItem(page, itup, itemsize, itup_off, false, false) == InvalidOffsetNumber)
314 0 : elog(ERROR, "failed to add index item to \"%s\"", RelationGetRelationName(rel));
315 :
316 357413 : return itup_off;
317 : }
318 :
319 : /*
320 : * _hash_pgaddmultitup() -- add a tuple vector to a particular page in the
321 : * index.
322 : *
323 : * This routine has same requirements for locking and tuple ordering as
324 : * _hash_pgaddtup().
325 : *
326 : * Returns the offset number array at which the tuples were inserted.
327 : */
328 : void
329 631 : _hash_pgaddmultitup(Relation rel, Buffer buf, IndexTuple *itups,
330 : OffsetNumber *itup_offsets, uint16 nitups)
331 : {
332 : OffsetNumber itup_off;
333 : Page page;
334 : uint32 hashkey;
335 : int i;
336 :
337 631 : _hash_checkpage(rel, buf, LH_BUCKET_PAGE | LH_OVERFLOW_PAGE);
338 631 : page = BufferGetPage(buf);
339 :
340 63608 : for (i = 0; i < nitups; i++)
341 : {
342 : Size itemsize;
343 :
344 62977 : itemsize = IndexTupleSize(itups[i]);
345 62977 : itemsize = MAXALIGN(itemsize);
346 :
347 : /* Find where to insert the tuple (preserving page's hashkey ordering) */
348 62977 : hashkey = _hash_get_indextuple_hashkey(itups[i]);
349 62977 : itup_off = _hash_binsearch(page, hashkey);
350 :
351 62977 : itup_offsets[i] = itup_off;
352 :
353 62977 : if (PageAddItem(page, itups[i], itemsize, itup_off, false, false) == InvalidOffsetNumber)
354 0 : elog(ERROR, "failed to add index item to \"%s\"", RelationGetRelationName(rel));
355 : }
356 631 : }
357 :
358 : /*
359 : * _hash_vacuum_one_page - vacuum just one index page.
360 : *
361 : * Try to remove LP_DEAD items from the given page. We must acquire cleanup
362 : * lock on the page being modified before calling this function.
363 : */
364 :
365 : static void
366 0 : _hash_vacuum_one_page(Relation rel, Relation hrel, Buffer metabuf, Buffer buf)
367 : {
368 : OffsetNumber deletable[MaxOffsetNumber];
369 0 : int ndeletable = 0;
370 : OffsetNumber offnum,
371 : maxoff;
372 0 : Page page = BufferGetPage(buf);
373 : HashPageOpaque pageopaque;
374 : HashMetaPage metap;
375 :
376 : /* Scan each tuple in page to see if it is marked as LP_DEAD */
377 0 : maxoff = PageGetMaxOffsetNumber(page);
378 0 : for (offnum = FirstOffsetNumber;
379 0 : offnum <= maxoff;
380 0 : offnum = OffsetNumberNext(offnum))
381 : {
382 0 : ItemId itemId = PageGetItemId(page, offnum);
383 :
384 0 : if (ItemIdIsDead(itemId))
385 0 : deletable[ndeletable++] = offnum;
386 : }
387 :
388 0 : if (ndeletable > 0)
389 : {
390 : TransactionId snapshotConflictHorizon;
391 :
392 : snapshotConflictHorizon =
393 0 : index_compute_xid_horizon_for_tuples(rel, hrel, buf,
394 : deletable, ndeletable);
395 :
396 : /*
397 : * Write-lock the meta page so that we can decrement tuple count.
398 : */
399 0 : LockBuffer(metabuf, BUFFER_LOCK_EXCLUSIVE);
400 :
401 : /* No ereport(ERROR) until changes are logged */
402 0 : START_CRIT_SECTION();
403 :
404 0 : PageIndexMultiDelete(page, deletable, ndeletable);
405 :
406 : /*
407 : * Mark the page as not containing any LP_DEAD items. This is not
408 : * certainly true (there might be some that have recently been marked,
409 : * but weren't included in our target-item list), but it will almost
410 : * always be true and it doesn't seem worth an additional page scan to
411 : * check it. Remember that LH_PAGE_HAS_DEAD_TUPLES is only a hint
412 : * anyway.
413 : */
414 0 : pageopaque = HashPageGetOpaque(page);
415 0 : pageopaque->hasho_flag &= ~LH_PAGE_HAS_DEAD_TUPLES;
416 :
417 0 : metap = HashPageGetMeta(BufferGetPage(metabuf));
418 0 : metap->hashm_ntuples -= ndeletable;
419 :
420 0 : MarkBufferDirty(buf);
421 0 : MarkBufferDirty(metabuf);
422 :
423 : /* XLOG stuff */
424 0 : if (RelationNeedsWAL(rel))
425 : {
426 : xl_hash_vacuum_one_page xlrec;
427 : XLogRecPtr recptr;
428 :
429 0 : xlrec.isCatalogRel = RelationIsAccessibleInLogicalDecoding(hrel);
430 0 : xlrec.snapshotConflictHorizon = snapshotConflictHorizon;
431 0 : xlrec.ntuples = ndeletable;
432 :
433 0 : XLogBeginInsert();
434 0 : XLogRegisterBuffer(0, buf, REGBUF_STANDARD);
435 0 : XLogRegisterData(&xlrec, SizeOfHashVacuumOnePage);
436 :
437 : /*
438 : * We need the target-offsets array whether or not we store the
439 : * whole buffer, to allow us to find the snapshotConflictHorizon
440 : * on a standby server.
441 : */
442 0 : XLogRegisterData(deletable,
443 : ndeletable * sizeof(OffsetNumber));
444 :
445 0 : XLogRegisterBuffer(1, metabuf, REGBUF_STANDARD);
446 :
447 0 : recptr = XLogInsert(RM_HASH_ID, XLOG_HASH_VACUUM_ONE_PAGE);
448 :
449 0 : PageSetLSN(BufferGetPage(buf), recptr);
450 0 : PageSetLSN(BufferGetPage(metabuf), recptr);
451 : }
452 :
453 0 : END_CRIT_SECTION();
454 :
455 : /*
456 : * Releasing write lock on meta page as we have updated the tuple
457 : * count.
458 : */
459 0 : LockBuffer(metabuf, BUFFER_LOCK_UNLOCK);
460 : }
461 0 : }
|
