| 1 |
/* |
| 2 |
* brin_revmap.c |
| 3 |
* Range map for BRIN indexes |
| 4 |
* |
| 5 |
* The range map (revmap) is a translation structure for BRIN indexes: for each |
| 6 |
* page range there is one summary tuple, and its location is tracked by the |
| 7 |
* revmap. Whenever a new tuple is inserted into a table that violates the |
| 8 |
* previously recorded summary values, a new tuple is inserted into the index |
| 9 |
* and the revmap is updated to point to it. |
| 10 |
* |
| 11 |
* The revmap is stored in the first pages of the index, immediately following |
| 12 |
* the metapage. When the revmap needs to be expanded, all tuples on the |
| 13 |
* regular BRIN page at that block (if any) are moved out of the way. |
| 14 |
* |
| 15 |
* Portions Copyright (c) 1996-2025, PostgreSQL Global Development Group |
| 16 |
* Portions Copyright (c) 1994, Regents of the University of California |
| 17 |
* |
| 18 |
* IDENTIFICATION |
| 19 |
* src/backend/access/brin/brin_revmap.c |
| 20 |
*/ |
| 21 |
#include "postgres.h" |
| 22 |
|
| 23 |
#include "access/brin_page.h" |
| 24 |
#include "access/brin_pageops.h" |
| 25 |
#include "access/brin_revmap.h" |
| 26 |
#include "access/brin_tuple.h" |
| 27 |
#include "access/brin_xlog.h" |
| 28 |
#include "access/rmgr.h" |
| 29 |
#include "access/xloginsert.h" |
| 30 |
#include "miscadmin.h" |
| 31 |
#include "storage/bufmgr.h" |
| 32 |
#include "utils/rel.h" |
| 33 |
|
| 34 |
|
| 35 |
/* |
| 36 |
* In revmap pages, each item stores an ItemPointerData. These defines let one |
| 37 |
* find the logical revmap page number and index number of the revmap item for |
| 38 |
* the given heap block number. |
| 39 |
*/ |
| 40 |
#define HEAPBLK_TO_REVMAP_BLK(pagesPerRange, heapBlk) \ |
| 41 |
((heapBlk / pagesPerRange) / REVMAP_PAGE_MAXITEMS) |
| 42 |
#define HEAPBLK_TO_REVMAP_INDEX(pagesPerRange, heapBlk) \ |
| 43 |
((heapBlk / pagesPerRange) % REVMAP_PAGE_MAXITEMS) |
| 44 |
|
| 45 |
|
| 46 |
struct BrinRevmap |
| 47 |
{ |
| 48 |
Relation rm_irel; |
| 49 |
BlockNumber rm_pagesPerRange; |
| 50 |
BlockNumber rm_lastRevmapPage; /* cached from the metapage */ |
| 51 |
Buffer rm_metaBuf; |
| 52 |
Buffer rm_currBuf; |
| 53 |
}; |
| 54 |
|
| 55 |
/* typedef appears in brin_revmap.h */ |
| 56 |
|
| 57 |
|
| 58 |
static BlockNumber revmap_get_blkno(BrinRevmap *revmap, |
| 59 |
BlockNumber heapBlk); |
| 60 |
static Buffer revmap_get_buffer(BrinRevmap *revmap, BlockNumber heapBlk); |
| 61 |
static BlockNumber revmap_extend_and_get_blkno(BrinRevmap *revmap, |
| 62 |
BlockNumber heapBlk); |
| 63 |
static void revmap_physical_extend(BrinRevmap *revmap); |
| 64 |
|
| 65 |
/* |
| 66 |
* Initialize an access object for a range map. This must be freed by |
| 67 |
* brinRevmapTerminate when caller is done with it. |
| 68 |
*/ |
| 69 |
BrinRevmap * |
| 70 |
brinRevmapInitialize(Relation idxrel, BlockNumber *pagesPerRange) |
| 71 |
{ |
| 72 |
BrinRevmap *revmap; |
| 73 |
Buffer meta; |
| 74 |
BrinMetaPageData *metadata; |
| 75 |
Page page; |
| 76 |
|
| 77 |
meta = ReadBuffer(idxrel, BRIN_METAPAGE_BLKNO); |
| 78 |
LockBuffer(meta, BUFFER_LOCK_SHARE); |
| 79 |
page = BufferGetPage(meta); |
| 80 |
metadata = (BrinMetaPageData *) PageGetContents(page); |
| 81 |
|
| 82 |
revmap = palloc(sizeof(BrinRevmap)); |
| 83 |
revmap->rm_irel = idxrel; |
| 84 |
revmap->rm_pagesPerRange = metadata->pagesPerRange; |
| 85 |
revmap->rm_lastRevmapPage = metadata->lastRevmapPage; |
| 86 |
revmap->rm_metaBuf = meta; |
| 87 |
revmap->rm_currBuf = InvalidBuffer; |
| 88 |
|
| 89 |
*pagesPerRange = metadata->pagesPerRange; |
| 90 |
|
| 91 |
LockBuffer(meta, BUFFER_LOCK_UNLOCK); |
| 92 |
|
| 93 |
return revmap; |
| 94 |
} |
| 95 |
|
| 96 |
/* |
| 97 |
* Release resources associated with a revmap access object. |
| 98 |
*/ |
| 99 |
void |
| 100 |
brinRevmapTerminate(BrinRevmap *revmap) |
| 101 |
{ |
| 102 |
ReleaseBuffer(revmap->rm_metaBuf); |
| 103 |
if (revmap->rm_currBuf != InvalidBuffer) |
| 104 |
ReleaseBuffer(revmap->rm_currBuf); |
| 105 |
pfree(revmap); |
| 106 |
} |
| 107 |
|
| 108 |
/* |
| 109 |
* Extend the revmap to cover the given heap block number. |
| 110 |
*/ |
| 111 |
void |
| 112 |
brinRevmapExtend(BrinRevmap *revmap, BlockNumber heapBlk) |
| 113 |
{ |
| 114 |
BlockNumber mapBlk PG_USED_FOR_ASSERTS_ONLY; |
| 115 |
|
| 116 |
mapBlk = revmap_extend_and_get_blkno(revmap, heapBlk); |
| 117 |
|
| 118 |
/* Ensure the buffer we got is in the expected range */ |
| 119 |
Assert(mapBlk != InvalidBlockNumber && |
| 120 |
mapBlk != BRIN_METAPAGE_BLKNO && |
| 121 |
mapBlk <= revmap->rm_lastRevmapPage); |
| 122 |
} |
| 123 |
|
| 124 |
/* |
| 125 |
* Prepare to insert an entry into the revmap; the revmap buffer in which the |
| 126 |
* entry is to reside is locked and returned. Most callers should call |
| 127 |
* brinRevmapExtend beforehand, as this routine does not extend the revmap if |
| 128 |
* it's not long enough. |
| 129 |
* |
| 130 |
* The returned buffer is also recorded in the revmap struct; finishing that |
| 131 |
* releases the buffer, therefore the caller needn't do it explicitly. |
| 132 |
*/ |
| 133 |
Buffer |
| 134 |
brinLockRevmapPageForUpdate(BrinRevmap *revmap, BlockNumber heapBlk) |
| 135 |
{ |
| 136 |
Buffer rmBuf; |
| 137 |
|
| 138 |
rmBuf = revmap_get_buffer(revmap, heapBlk); |
| 139 |
LockBuffer(rmBuf, BUFFER_LOCK_EXCLUSIVE); |
| 140 |
|
| 141 |
return rmBuf; |
| 142 |
} |
| 143 |
|
| 144 |
/* |
| 145 |
* In the given revmap buffer (locked appropriately by caller), which is used |
| 146 |
* in a BRIN index of pagesPerRange pages per range, set the element |
| 147 |
* corresponding to heap block number heapBlk to the given TID. |
| 148 |
* |
| 149 |
* Once the operation is complete, the caller must update the LSN on the |
| 150 |
* returned buffer. |
| 151 |
* |
| 152 |
* This is used both in regular operation and during WAL replay. |
| 153 |
*/ |
| 154 |
void |
| 155 |
brinSetHeapBlockItemptr(Buffer buf, BlockNumber pagesPerRange, |
| 156 |
BlockNumber heapBlk, ItemPointerData tid) |
| 157 |
{ |
| 158 |
RevmapContents *contents; |
| 159 |
ItemPointerData *iptr; |
| 160 |
Page page; |
| 161 |
|
| 162 |
/* The correct page should already be pinned and locked */ |
| 163 |
page = BufferGetPage(buf); |
| 164 |
contents = (RevmapContents *) PageGetContents(page); |
| 165 |
iptr = (ItemPointerData *) contents->rm_tids; |
| 166 |
iptr += HEAPBLK_TO_REVMAP_INDEX(pagesPerRange, heapBlk); |
| 167 |
|
| 168 |
if (ItemPointerIsValid(&tid)) |
| 169 |
ItemPointerSet(iptr, |
| 170 |
ItemPointerGetBlockNumber(&tid), |
| 171 |
ItemPointerGetOffsetNumber(&tid)); |
| 172 |
else |
| 173 |
ItemPointerSetInvalid(iptr); |
| 174 |
} |
| 175 |
|
| 176 |
/* |
| 177 |
* Fetch the BrinTuple for a given heap block. |
| 178 |
* |
| 179 |
* The buffer containing the tuple is locked, and returned in *buf. The |
| 180 |
* returned tuple points to the shared buffer and must not be freed; if caller |
| 181 |
* wants to use it after releasing the buffer lock, it must create its own |
| 182 |
* palloc'ed copy. As an optimization, the caller can pass a pinned buffer |
| 183 |
* *buf on entry, which will avoid a pin-unpin cycle when the next tuple is on |
| 184 |
* the same page as a previous one. |
| 185 |
* |
| 186 |
* If no tuple is found for the given heap range, returns NULL. In that case, |
| 187 |
* *buf might still be updated (and pin must be released by caller), but it's |
| 188 |
* not locked. |
| 189 |
* |
| 190 |
* The output tuple offset within the buffer is returned in *off, and its size |
| 191 |
* is returned in *size. |
| 192 |
*/ |
| 193 |
BrinTuple * |
| 194 |
brinGetTupleForHeapBlock(BrinRevmap *revmap, BlockNumber heapBlk, |
| 195 |
Buffer *buf, OffsetNumber *off, Size *size, int mode) |
| 196 |
{ |
| 197 |
Relation idxRel = revmap->rm_irel; |
| 198 |
BlockNumber mapBlk; |
| 199 |
RevmapContents *contents; |
| 200 |
ItemPointerData *iptr; |
| 201 |
BlockNumber blk; |
| 202 |
Page page; |
| 203 |
ItemId lp; |
| 204 |
BrinTuple *tup; |
| 205 |
ItemPointerData previptr; |
| 206 |
|
| 207 |
/* normalize the heap block number to be the first page in the range */ |
| 208 |
heapBlk = (heapBlk / revmap->rm_pagesPerRange) * revmap->rm_pagesPerRange; |
| 209 |
|
| 210 |
/* |
| 211 |
* Compute the revmap page number we need. If Invalid is returned (i.e., |
| 212 |
* the revmap page hasn't been created yet), the requested page range is |
| 213 |
* not summarized. |
| 214 |
*/ |
| 215 |
mapBlk = revmap_get_blkno(revmap, heapBlk); |
| 216 |
if (mapBlk == InvalidBlockNumber) |
| 217 |
{ |
| 218 |
*off = InvalidOffsetNumber; |
| 219 |
return NULL; |
| 220 |
} |
| 221 |
|
| 222 |
ItemPointerSetInvalid(&previptr); |
| 223 |
for (;;) |
| 224 |
{ |
| 225 |
CHECK_FOR_INTERRUPTS(); |
| 226 |
|
| 227 |
if (revmap->rm_currBuf == InvalidBuffer || |
| 228 |
BufferGetBlockNumber(revmap->rm_currBuf) != mapBlk) |
| 229 |
{ |
| 230 |
if (revmap->rm_currBuf != InvalidBuffer) |
| 231 |
ReleaseBuffer(revmap->rm_currBuf); |
| 232 |
|
| 233 |
Assert(mapBlk != InvalidBlockNumber); |
| 234 |
revmap->rm_currBuf = ReadBuffer(revmap->rm_irel, mapBlk); |
| 235 |
} |
| 236 |
|
| 237 |
LockBuffer(revmap->rm_currBuf, BUFFER_LOCK_SHARE); |
| 238 |
|
| 239 |
contents = (RevmapContents *) |
| 240 |
PageGetContents(BufferGetPage(revmap->rm_currBuf)); |
| 241 |
iptr = contents->rm_tids; |
| 242 |
iptr += HEAPBLK_TO_REVMAP_INDEX(revmap->rm_pagesPerRange, heapBlk); |
| 243 |
|
| 244 |
if (!ItemPointerIsValid(iptr)) |
| 245 |
{ |
| 246 |
LockBuffer(revmap->rm_currBuf, BUFFER_LOCK_UNLOCK); |
| 247 |
return NULL; |
| 248 |
} |
| 249 |
|
| 250 |
/* |
| 251 |
* Check the TID we got in a previous iteration, if any, and save the |
| 252 |
* current TID we got from the revmap; if we loop, we can sanity-check |
| 253 |
* that the next one we get is different. Otherwise we might be stuck |
| 254 |
* looping forever if the revmap is somehow badly broken. |
| 255 |
*/ |
| 256 |
if (ItemPointerIsValid(&previptr) && ItemPointerEquals(&previptr, iptr)) |
| 257 |
ereport(ERROR, |
| 258 |
(errcode(ERRCODE_INDEX_CORRUPTED), |
| 259 |
errmsg_internal("corrupted BRIN index: inconsistent range map"))); |
| 260 |
previptr = *iptr; |
| 261 |
|
| 262 |
blk = ItemPointerGetBlockNumber(iptr); |
| 263 |
*off = ItemPointerGetOffsetNumber(iptr); |
| 264 |
|
| 265 |
LockBuffer(revmap->rm_currBuf, BUFFER_LOCK_UNLOCK); |
| 266 |
|
| 267 |
/* Ok, got a pointer to where the BrinTuple should be. Fetch it. */ |
| 268 |
if (!BufferIsValid(*buf) || BufferGetBlockNumber(*buf) != blk) |
| 269 |
{ |
| 270 |
if (BufferIsValid(*buf)) |
| 271 |
ReleaseBuffer(*buf); |
| 272 |
*buf = ReadBuffer(idxRel, blk); |
| 273 |
} |
| 274 |
LockBuffer(*buf, mode); |
| 275 |
page = BufferGetPage(*buf); |
| 276 |
|
| 277 |
/* If we land on a revmap page, start over */ |
| 278 |
if (BRIN_IS_REGULAR_PAGE(page)) |
| 279 |
{ |
| 280 |
/* |
| 281 |
* If the offset number is greater than what's in the page, it's |
| 282 |
* possible that the range was desummarized concurrently. Just |
| 283 |
* return NULL to handle that case. |
| 284 |
*/ |
| 285 |
if (*off > PageGetMaxOffsetNumber(page)) |
| 286 |
{ |
| 287 |
LockBuffer(*buf, BUFFER_LOCK_UNLOCK); |
| 288 |
return NULL; |
| 289 |
} |
| 290 |
|
| 291 |
lp = PageGetItemId(page, *off); |
| 292 |
if (ItemIdIsUsed(lp)) |
| 293 |
{ |
| 294 |
tup = (BrinTuple *) PageGetItem(page, lp); |
| 295 |
|
| 296 |
if (tup->bt_blkno == heapBlk) |
| 297 |
{ |
| 298 |
if (size) |
| 299 |
*size = ItemIdGetLength(lp); |
| 300 |
/* found it! */ |
| 301 |
return tup; |
| 302 |
} |
| 303 |
} |
| 304 |
} |
| 305 |
|
| 306 |
/* |
| 307 |
* No luck. Assume that the revmap was updated concurrently. |
| 308 |
*/ |
| 309 |
LockBuffer(*buf, BUFFER_LOCK_UNLOCK); |
| 310 |
} |
| 311 |
/* not reached, but keep compiler quiet */ |
| 312 |
return NULL; |
| 313 |
} |
| 314 |
|
| 315 |
/* |
| 316 |
* Delete an index tuple, marking a page range as unsummarized. |
| 317 |
* |
| 318 |
* Index must be locked in ShareUpdateExclusiveLock mode. |
| 319 |
* |
| 320 |
* Return false if caller should retry. |
| 321 |
*/ |
| 322 |
bool |
| 323 |
brinRevmapDesummarizeRange(Relation idxrel, BlockNumber heapBlk) |
| 324 |
{ |
| 325 |
BrinRevmap *revmap; |
| 326 |
BlockNumber pagesPerRange; |
| 327 |
RevmapContents *contents; |
| 328 |
ItemPointerData *iptr; |
| 329 |
ItemPointerData invalidIptr; |
| 330 |
BlockNumber revmapBlk; |
| 331 |
Buffer revmapBuf; |
| 332 |
Buffer regBuf; |
| 333 |
Page revmapPg; |
| 334 |
Page regPg; |
| 335 |
OffsetNumber revmapOffset; |
| 336 |
OffsetNumber regOffset; |
| 337 |
ItemId lp; |
| 338 |
|
| 339 |
revmap = brinRevmapInitialize(idxrel, &pagesPerRange); |
| 340 |
|
| 341 |
revmapBlk = revmap_get_blkno(revmap, heapBlk); |
| 342 |
if (!BlockNumberIsValid(revmapBlk)) |
| 343 |
{ |
| 344 |
/* revmap page doesn't exist: range not summarized, we're done */ |
| 345 |
brinRevmapTerminate(revmap); |
| 346 |
return true; |
| 347 |
} |
| 348 |
|
| 349 |
/* Lock the revmap page, obtain the index tuple pointer from it */ |
| 350 |
revmapBuf = brinLockRevmapPageForUpdate(revmap, heapBlk); |
| 351 |
revmapPg = BufferGetPage(revmapBuf); |
| 352 |
revmapOffset = HEAPBLK_TO_REVMAP_INDEX(revmap->rm_pagesPerRange, heapBlk); |
| 353 |
|
| 354 |
contents = (RevmapContents *) PageGetContents(revmapPg); |
| 355 |
iptr = contents->rm_tids; |
| 356 |
iptr += revmapOffset; |
| 357 |
|
| 358 |
if (!ItemPointerIsValid(iptr)) |
| 359 |
{ |
| 360 |
/* no index tuple: range not summarized, we're done */ |
| 361 |
LockBuffer(revmapBuf, BUFFER_LOCK_UNLOCK); |
| 362 |
brinRevmapTerminate(revmap); |
| 363 |
return true; |
| 364 |
} |
| 365 |
|
| 366 |
regBuf = ReadBuffer(idxrel, ItemPointerGetBlockNumber(iptr)); |
| 367 |
LockBuffer(regBuf, BUFFER_LOCK_EXCLUSIVE); |
| 368 |
regPg = BufferGetPage(regBuf); |
| 369 |
|
| 370 |
/* if this is no longer a regular page, tell caller to start over */ |
| 371 |
if (!BRIN_IS_REGULAR_PAGE(regPg)) |
| 372 |
{ |
| 373 |
LockBuffer(revmapBuf, BUFFER_LOCK_UNLOCK); |
| 374 |
LockBuffer(regBuf, BUFFER_LOCK_UNLOCK); |
| 375 |
brinRevmapTerminate(revmap); |
| 376 |
return false; |
| 377 |
} |
| 378 |
|
| 379 |
regOffset = ItemPointerGetOffsetNumber(iptr); |
| 380 |
if (regOffset > PageGetMaxOffsetNumber(regPg)) |
| 381 |
ereport(ERROR, |
| 382 |
(errcode(ERRCODE_INDEX_CORRUPTED), |
| 383 |
errmsg("corrupted BRIN index: inconsistent range map"))); |
| 384 |
|
| 385 |
lp = PageGetItemId(regPg, regOffset); |
| 386 |
if (!ItemIdIsUsed(lp)) |
| 387 |
ereport(ERROR, |
| 388 |
(errcode(ERRCODE_INDEX_CORRUPTED), |
| 389 |
errmsg("corrupted BRIN index: inconsistent range map"))); |
| 390 |
|
| 391 |
/* |
| 392 |
* Placeholder tuples only appear during unfinished summarization, and we |
| 393 |
* hold ShareUpdateExclusiveLock, so this function cannot run concurrently |
| 394 |
* with that. So any placeholder tuples that exist are leftovers from a |
| 395 |
* crashed or aborted summarization; remove them silently. |
| 396 |
*/ |
| 397 |
|
| 398 |
START_CRIT_SECTION(); |
| 399 |
|
| 400 |
ItemPointerSetInvalid(&invalidIptr); |
| 401 |
brinSetHeapBlockItemptr(revmapBuf, revmap->rm_pagesPerRange, heapBlk, |
| 402 |
invalidIptr); |
| 403 |
PageIndexTupleDeleteNoCompact(regPg, regOffset); |
| 404 |
/* XXX record free space in FSM? */ |
| 405 |
|
| 406 |
MarkBufferDirty(regBuf); |
| 407 |
MarkBufferDirty(revmapBuf); |
| 408 |
|
| 409 |
if (RelationNeedsWAL(idxrel)) |
| 410 |
{ |
| 411 |
xl_brin_desummarize xlrec; |
| 412 |
XLogRecPtr recptr; |
| 413 |
|
| 414 |
xlrec.pagesPerRange = revmap->rm_pagesPerRange; |
| 415 |
xlrec.heapBlk = heapBlk; |
| 416 |
xlrec.regOffset = regOffset; |
| 417 |
|
| 418 |
XLogBeginInsert(); |
| 419 |
XLogRegisterData(&xlrec, SizeOfBrinDesummarize); |
| 420 |
XLogRegisterBuffer(0, revmapBuf, 0); |
| 421 |
XLogRegisterBuffer(1, regBuf, REGBUF_STANDARD); |
| 422 |
recptr = XLogInsert(RM_BRIN_ID, XLOG_BRIN_DESUMMARIZE); |
| 423 |
PageSetLSN(revmapPg, recptr); |
| 424 |
PageSetLSN(regPg, recptr); |
| 425 |
} |
| 426 |
|
| 427 |
END_CRIT_SECTION(); |
| 428 |
|
| 429 |
UnlockReleaseBuffer(regBuf); |
| 430 |
LockBuffer(revmapBuf, BUFFER_LOCK_UNLOCK); |
| 431 |
brinRevmapTerminate(revmap); |
| 432 |
|
| 433 |
return true; |
| 434 |
} |
| 435 |
|
| 436 |
/* |
| 437 |
* Given a heap block number, find the corresponding physical revmap block |
| 438 |
* number and return it. If the revmap page hasn't been allocated yet, return |
| 439 |
* InvalidBlockNumber. |
| 440 |
*/ |
| 441 |
static BlockNumber |
| 442 |
revmap_get_blkno(BrinRevmap *revmap, BlockNumber heapBlk) |
| 443 |
{ |
| 444 |
BlockNumber targetblk; |
| 445 |
|
| 446 |
/* obtain revmap block number, skip 1 for metapage block */ |
| 447 |
targetblk = HEAPBLK_TO_REVMAP_BLK(revmap->rm_pagesPerRange, heapBlk) + 1; |
| 448 |
|
| 449 |
/* Normal case: the revmap page is already allocated */ |
| 450 |
if (targetblk <= revmap->rm_lastRevmapPage) |
| 451 |
return targetblk; |
| 452 |
|
| 453 |
return InvalidBlockNumber; |
| 454 |
} |
| 455 |
|
| 456 |
/* |
| 457 |
* Obtain and return a buffer containing the revmap page for the given heap |
| 458 |
* page. The revmap must have been previously extended to cover that page. |
| 459 |
* The returned buffer is also recorded in the revmap struct; finishing that |
| 460 |
* releases the buffer, therefore the caller needn't do it explicitly. |
| 461 |
*/ |
| 462 |
static Buffer |
| 463 |
revmap_get_buffer(BrinRevmap *revmap, BlockNumber heapBlk) |
| 464 |
{ |
| 465 |
BlockNumber mapBlk; |
| 466 |
|
| 467 |
/* Translate the heap block number to physical index location. */ |
| 468 |
mapBlk = revmap_get_blkno(revmap, heapBlk); |
| 469 |
|
| 470 |
if (mapBlk == InvalidBlockNumber) |
| 471 |
elog(ERROR, "revmap does not cover heap block %u", heapBlk); |
| 472 |
|
| 473 |
/* Ensure the buffer we got is in the expected range */ |
| 474 |
Assert(mapBlk != BRIN_METAPAGE_BLKNO && |
| 475 |
mapBlk <= revmap->rm_lastRevmapPage); |
| 476 |
|
| 477 |
/* |
| 478 |
* Obtain the buffer from which we need to read. If we already have the |
| 479 |
* correct buffer in our access struct, use that; otherwise, release that, |
| 480 |
* (if valid) and read the one we need. |
| 481 |
*/ |
| 482 |
if (revmap->rm_currBuf == InvalidBuffer || |
| 483 |
mapBlk != BufferGetBlockNumber(revmap->rm_currBuf)) |
| 484 |
{ |
| 485 |
if (revmap->rm_currBuf != InvalidBuffer) |
| 486 |
ReleaseBuffer(revmap->rm_currBuf); |
| 487 |
|
| 488 |
revmap->rm_currBuf = ReadBuffer(revmap->rm_irel, mapBlk); |
| 489 |
} |
| 490 |
|
| 491 |
return revmap->rm_currBuf; |
| 492 |
} |
| 493 |
|
| 494 |
/* |
| 495 |
* Given a heap block number, find the corresponding physical revmap block |
| 496 |
* number and return it. If the revmap page hasn't been allocated yet, extend |
| 497 |
* the revmap until it is. |
| 498 |
*/ |
| 499 |
static BlockNumber |
| 500 |
revmap_extend_and_get_blkno(BrinRevmap *revmap, BlockNumber heapBlk) |
| 501 |
{ |
| 502 |
BlockNumber targetblk; |
| 503 |
|
| 504 |
/* obtain revmap block number, skip 1 for metapage block */ |
| 505 |
targetblk = HEAPBLK_TO_REVMAP_BLK(revmap->rm_pagesPerRange, heapBlk) + 1; |
| 506 |
|
| 507 |
/* Extend the revmap, if necessary */ |
| 508 |
while (targetblk > revmap->rm_lastRevmapPage) |
| 509 |
{ |
| 510 |
CHECK_FOR_INTERRUPTS(); |
| 511 |
revmap_physical_extend(revmap); |
| 512 |
} |
| 513 |
|
| 514 |
return targetblk; |
| 515 |
} |
| 516 |
|
| 517 |
/* |
| 518 |
* Try to extend the revmap by one page. This might not happen for a number of |
| 519 |
* reasons; caller is expected to retry until the expected outcome is obtained. |
| 520 |
*/ |
| 521 |
static void |
| 522 |
revmap_physical_extend(BrinRevmap *revmap) |
| 523 |
{ |
| 524 |
Buffer buf; |
| 525 |
Page page; |
| 526 |
Page metapage; |
| 527 |
BrinMetaPageData *metadata; |
| 528 |
BlockNumber mapBlk; |
| 529 |
BlockNumber nblocks; |
| 530 |
Relation irel = revmap->rm_irel; |
| 531 |
|
| 532 |
/* |
| 533 |
* Lock the metapage. This locks out concurrent extensions of the revmap, |
| 534 |
* but note that we still need to grab the relation extension lock because |
| 535 |
* another backend can extend the index with regular BRIN pages. |
| 536 |
*/ |
| 537 |
LockBuffer(revmap->rm_metaBuf, BUFFER_LOCK_EXCLUSIVE); |
| 538 |
metapage = BufferGetPage(revmap->rm_metaBuf); |
| 539 |
metadata = (BrinMetaPageData *) PageGetContents(metapage); |
| 540 |
|
| 541 |
/* |
| 542 |
* Check that our cached lastRevmapPage value was up-to-date; if it |
| 543 |
* wasn't, update the cached copy and have caller start over. |
| 544 |
*/ |
| 545 |
if (metadata->lastRevmapPage != revmap->rm_lastRevmapPage) |
| 546 |
{ |
| 547 |
revmap->rm_lastRevmapPage = metadata->lastRevmapPage; |
| 548 |
LockBuffer(revmap->rm_metaBuf, BUFFER_LOCK_UNLOCK); |
| 549 |
return; |
| 550 |
} |
| 551 |
mapBlk = metadata->lastRevmapPage + 1; |
| 552 |
|
| 553 |
nblocks = RelationGetNumberOfBlocks(irel); |
| 554 |
if (mapBlk < nblocks) |
| 555 |
{ |
| 556 |
buf = ReadBuffer(irel, mapBlk); |
| 557 |
LockBuffer(buf, BUFFER_LOCK_EXCLUSIVE); |
| 558 |
page = BufferGetPage(buf); |
| 559 |
} |
| 560 |
else |
| 561 |
{ |
| 562 |
buf = ExtendBufferedRel(BMR_REL(irel), MAIN_FORKNUM, NULL, |
| 563 |
EB_LOCK_FIRST); |
| 564 |
if (BufferGetBlockNumber(buf) != mapBlk) |
| 565 |
{ |
| 566 |
/* |
| 567 |
* Very rare corner case: somebody extended the relation |
| 568 |
* concurrently after we read its length. If this happens, give |
| 569 |
* up and have caller start over. We will have to evacuate that |
| 570 |
* page from under whoever is using it. |
| 571 |
*/ |
| 572 |
LockBuffer(revmap->rm_metaBuf, BUFFER_LOCK_UNLOCK); |
| 573 |
UnlockReleaseBuffer(buf); |
| 574 |
return; |
| 575 |
} |
| 576 |
page = BufferGetPage(buf); |
| 577 |
} |
| 578 |
|
| 579 |
/* Check that it's a regular block (or an empty page) */ |
| 580 |
if (!PageIsNew(page) && !BRIN_IS_REGULAR_PAGE(page)) |
| 581 |
ereport(ERROR, |
| 582 |
(errcode(ERRCODE_INDEX_CORRUPTED), |
| 583 |
errmsg("unexpected page type 0x%04X in BRIN index \"%s\" block %u", |
| 584 |
BrinPageType(page), |
| 585 |
RelationGetRelationName(irel), |
| 586 |
BufferGetBlockNumber(buf)))); |
| 587 |
|
| 588 |
/* If the page is in use, evacuate it and restart */ |
| 589 |
if (brin_start_evacuating_page(irel, buf)) |
| 590 |
{ |
| 591 |
LockBuffer(revmap->rm_metaBuf, BUFFER_LOCK_UNLOCK); |
| 592 |
brin_evacuate_page(irel, revmap->rm_pagesPerRange, revmap, buf); |
| 593 |
|
| 594 |
/* have caller start over */ |
| 595 |
return; |
| 596 |
} |
| 597 |
|
| 598 |
/* |
| 599 |
* Ok, we have now locked the metapage and the target block. Re-initialize |
| 600 |
* the target block as a revmap page, and update the metapage. |
| 601 |
*/ |
| 602 |
START_CRIT_SECTION(); |
| 603 |
|
| 604 |
/* the rm_tids array is initialized to all invalid by PageInit */ |
| 605 |
brin_page_init(page, BRIN_PAGETYPE_REVMAP); |
| 606 |
MarkBufferDirty(buf); |
| 607 |
|
| 608 |
metadata->lastRevmapPage = mapBlk; |
| 609 |
|
| 610 |
/* |
| 611 |
* Set pd_lower just past the end of the metadata. This is essential, |
| 612 |
* because without doing so, metadata will be lost if xlog.c compresses |
| 613 |
* the page. (We must do this here because pre-v11 versions of PG did not |
| 614 |
* set the metapage's pd_lower correctly, so a pg_upgraded index might |
| 615 |
* contain the wrong value.) |
| 616 |
*/ |
| 617 |
((PageHeader) metapage)->pd_lower = |
| 618 |
((char *) metadata + sizeof(BrinMetaPageData)) - (char *) metapage; |
| 619 |
|
| 620 |
MarkBufferDirty(revmap->rm_metaBuf); |
| 621 |
|
| 622 |
if (RelationNeedsWAL(revmap->rm_irel)) |
| 623 |
{ |
| 624 |
xl_brin_revmap_extend xlrec; |
| 625 |
XLogRecPtr recptr; |
| 626 |
|
| 627 |
xlrec.targetBlk = mapBlk; |
| 628 |
|
| 629 |
XLogBeginInsert(); |
| 630 |
XLogRegisterData(&xlrec, SizeOfBrinRevmapExtend); |
| 631 |
XLogRegisterBuffer(0, revmap->rm_metaBuf, REGBUF_STANDARD); |
| 632 |
|
| 633 |
XLogRegisterBuffer(1, buf, REGBUF_WILL_INIT); |
| 634 |
|
| 635 |
recptr = XLogInsert(RM_BRIN_ID, XLOG_BRIN_REVMAP_EXTEND); |
| 636 |
PageSetLSN(metapage, recptr); |
| 637 |
PageSetLSN(page, recptr); |
| 638 |
} |
| 639 |
|
| 640 |
END_CRIT_SECTION(); |
| 641 |
|
| 642 |
LockBuffer(revmap->rm_metaBuf, BUFFER_LOCK_UNLOCK); |
| 643 |
|
| 644 |
UnlockReleaseBuffer(buf); |
| 645 |
} |