1 /*-------------------------------------------------------------------------
4 * POSTGRES free space map for quickly finding free space in relations
7 * Portions Copyright (c) 1996-2004, PostgreSQL Global Development Group
8 * Portions Copyright (c) 1994, Regents of the University of California
11 * $PostgreSQL: pgsql/src/backend/storage/freespace/freespace.c,v 1.33 2004/08/29 04:12:47 momjian Exp $
16 * The only really interesting aspect of this code is the heuristics for
17 * deciding how much information we can afford to keep about each relation,
18 * given that we have a limited amount of workspace in shared memory.
19 * These currently work as follows:
21 * The number of distinct relations tracked is limited by a configuration
22 * variable (MaxFSMRelations). When this would be exceeded, we discard the
23 * least recently used relation. A doubly-linked list with move-to-front
24 * behavior keeps track of which relation is least recently used.
26 * For each known relation, we track the average request size given to
27 * GetPageWithFreeSpace() as well as the most recent number of pages given
28 * to RecordRelationFreeSpace(). The average request size is not directly
29 * used in this module, but we expect VACUUM to use it to filter out
30 * uninteresting amounts of space before calling RecordRelationFreeSpace().
31 * The sum of the RRFS page counts is thus the total number of "interesting"
32 * pages that we would like to track; this is called DesiredFSMPages.
34 * The number of pages actually tracked is limited by a configuration variable
35 * (MaxFSMPages). When this is less than DesiredFSMPages, each relation
36 * gets to keep a fraction MaxFSMPages/DesiredFSMPages of its free pages.
37 * We discard pages with less free space to reach this target.
39 * Actually, our space allocation is done in "chunks" of CHUNKPAGES pages,
40 * with each relation guaranteed at least one chunk. This reduces thrashing
41 * of the storage allocations when there are small changes in the RRFS page
42 * counts from one VACUUM to the next. (XXX it might also be worthwhile to
43 * impose some kind of moving-average smoothing on the RRFS page counts?)
45 * So the actual arithmetic is: for each relation compute myRequest as the
46 * number of chunks needed to hold its RRFS page count (not counting the
47 * first, guaranteed chunk); compute sumRequests as the sum of these values
48 * over all relations; then for each relation figure its target allocation
50 * 1 + round(spareChunks * myRequest / sumRequests)
51 * where spareChunks = totalChunks - numRels is the number of chunks we have
52 * a choice what to do with. We round off these numbers because truncating
53 * all of them would waste significant space. But because of roundoff, it's
54 * possible for the last few relations to get less space than they should;
55 * the target allocation must be checked against remaining available space.
57 *-------------------------------------------------------------------------
66 #include "miscadmin.h"
67 #include "storage/fd.h"
68 #include "storage/freespace.h"
69 #include "storage/itemptr.h"
70 #include "storage/lwlock.h"
71 #include "storage/shmem.h"
74 /* Initial value for average-request moving average */
75 #define INITIAL_AVERAGE ((Size) (BLCKSZ / 32))
78 * Number of pages and bytes per allocation chunk. Indexes can squeeze 50%
79 * more pages into the same space because they don't need to remember how much
80 * free space on each page. The nominal number of pages, CHUNKPAGES, is for
81 * regular rels, and INDEXCHUNKPAGES is for indexes. CHUNKPAGES should be
82 * even so that no space is wasted in the index case.
85 #define CHUNKBYTES (CHUNKPAGES * sizeof(FSMPageData))
86 #define INDEXCHUNKPAGES ((int) (CHUNKBYTES / sizeof(IndexFSMPageData)))
90 * Typedefs and macros for items in the page-storage arena. We use the
91 * existing ItemPointer and BlockId data structures, which are designed
92 * to pack well (they should be 6 and 4 bytes apiece regardless of machine
93 * alignment issues). Unfortunately we can't use the ItemPointer access
94 * macros, because they include Asserts insisting that ip_posid != 0.
96 typedef ItemPointerData FSMPageData;
97 typedef BlockIdData IndexFSMPageData;
99 #define FSMPageGetPageNum(ptr) \
100 BlockIdGetBlockNumber(&(ptr)->ip_blkid)
101 #define FSMPageGetSpace(ptr) \
102 ((Size) (ptr)->ip_posid)
103 #define FSMPageSetPageNum(ptr, pg) \
104 BlockIdSet(&(ptr)->ip_blkid, pg)
105 #define FSMPageSetSpace(ptr, sz) \
106 ((ptr)->ip_posid = (OffsetNumber) (sz))
107 #define IndexFSMPageGetPageNum(ptr) \
108 BlockIdGetBlockNumber(ptr)
109 #define IndexFSMPageSetPageNum(ptr, pg) \
113 * During database shutdown, we store the contents of FSM into a disk file,
114 * which is re-read during startup. This way we don't have a startup
115 * transient condition where FSM isn't really functioning.
117 * The file format is:
119 * endian constant 0x01020304 for detecting endianness problems
122 * -- for each rel, in *reverse* usage order:
128 * arena data array of storedPages FSMPageData or IndexFSMPageData
132 /* Name of FSM cache file (relative to $PGDATA) */
133 #define FSM_CACHE_FILENAME "global/pg_fsm.cache"
135 /* Fixed values in header */
136 #define FSM_CACHE_LABEL "FSM"
137 #define FSM_CACHE_ENDIAN 0x01020304
138 #define FSM_CACHE_VERSION 20030305
140 /* File header layout */
141 typedef struct FsmCacheFileHeader
147 } FsmCacheFileHeader;
149 /* Per-relation header */
150 typedef struct FsmCacheRelHeader
152 RelFileNode key; /* hash key (must be first) */
153 bool isIndex; /* if true, we store only page numbers */
154 uint32 avgRequest; /* moving average of space requests */
155 int32 lastPageCount; /* pages passed to RecordRelationFreeSpace */
156 int32 storedPages; /* # of pages stored in arena */
161 * Shared free-space-map objects
163 * The per-relation objects are indexed by a hash table, and are also members
164 * of two linked lists: one ordered by recency of usage (most recent first),
165 * and the other ordered by physical location of the associated storage in
166 * the page-info arena.
168 * Each relation owns one or more chunks of per-page storage in the "arena".
169 * The chunks for each relation are always consecutive, so that it can treat
170 * its page storage as a simple array. We further insist that its page data
171 * be ordered by block number, so that binary search is possible.
173 * Note: we handle pointers to these items as pointers, not as SHMEM_OFFSETs.
174 * This assumes that all processes accessing the map will have the shared
175 * memory segment mapped at the same place in their address space.
177 typedef struct FSMHeader FSMHeader;
178 typedef struct FSMRelation FSMRelation;
180 /* Header for whole map */
183 FSMRelation *usageList; /* FSMRelations in usage-recency order */
184 FSMRelation *usageListTail; /* tail of usage-recency list */
185 FSMRelation *firstRel; /* FSMRelations in arena storage order */
186 FSMRelation *lastRel; /* tail of storage-order list */
187 int numRels; /* number of FSMRelations now in use */
188 double sumRequests; /* sum of requested chunks over all rels */
189 char *arena; /* arena for page-info storage */
190 int totalChunks; /* total size of arena, in chunks */
191 int usedChunks; /* # of chunks assigned */
192 /* NB: there are totalChunks - usedChunks free chunks at end of arena */
196 * Per-relation struct --- this is an entry in the shared hash table.
197 * The hash key is the RelFileNode value (hence, we look at the physical
198 * relation ID, not the logical ID, which is appropriate).
202 RelFileNode key; /* hash key (must be first) */
203 FSMRelation *nextUsage; /* next rel in usage-recency order */
204 FSMRelation *priorUsage; /* prior rel in usage-recency order */
205 FSMRelation *nextPhysical; /* next rel in arena-storage order */
206 FSMRelation *priorPhysical; /* prior rel in arena-storage order */
207 bool isIndex; /* if true, we store only page numbers */
208 Size avgRequest; /* moving average of space requests */
209 int lastPageCount; /* pages passed to RecordRelationFreeSpace */
210 int firstChunk; /* chunk # of my first chunk in arena */
211 int storedPages; /* # of pages stored in arena */
212 int nextPage; /* index (from 0) to start next search at */
216 int MaxFSMRelations; /* these are set by guc.c */
219 static FSMHeader *FreeSpaceMap; /* points to FSMHeader in shared memory */
220 static HTAB *FreeSpaceMapRelHash; /* points to (what used to be) FSMHeader->relHash */
223 static FSMRelation *lookup_fsm_rel(RelFileNode *rel);
224 static FSMRelation *create_fsm_rel(RelFileNode *rel);
225 static void delete_fsm_rel(FSMRelation *fsmrel);
226 static int realloc_fsm_rel(FSMRelation *fsmrel, int nPages, bool isIndex);
227 static void link_fsm_rel_usage(FSMRelation *fsmrel);
228 static void unlink_fsm_rel_usage(FSMRelation *fsmrel);
229 static void link_fsm_rel_storage(FSMRelation *fsmrel);
230 static void unlink_fsm_rel_storage(FSMRelation *fsmrel);
231 static BlockNumber find_free_space(FSMRelation *fsmrel, Size spaceNeeded);
232 static BlockNumber find_index_free_space(FSMRelation *fsmrel);
233 static void fsm_record_free_space(FSMRelation *fsmrel, BlockNumber page,
235 static bool lookup_fsm_page_entry(FSMRelation *fsmrel, BlockNumber page,
237 static void compact_fsm_storage(void);
238 static void push_fsm_rels_after(FSMRelation *afterRel);
239 static void pack_incoming_pages(FSMPageData *newLocation, int newPages,
240 PageFreeSpaceInfo *pageSpaces, int nPages);
241 static void pack_existing_pages(FSMPageData *newLocation, int newPages,
242 FSMPageData *oldLocation, int oldPages);
243 static int fsm_calc_request(FSMRelation *fsmrel);
244 static int fsm_calc_target_allocation(int myRequest);
245 static int fsm_current_chunks(FSMRelation *fsmrel);
246 static int fsm_current_allocation(FSMRelation *fsmrel);
255 * InitFreeSpaceMap -- Initialize the freespace module.
257 * This must be called once during shared memory initialization.
258 * It builds the empty free space map table. FreeSpaceLock must also be
259 * initialized at some point, but is not touched here --- we assume there is
260 * no need for locking, since only the calling process can be accessing shared
264 InitFreeSpaceMap(void)
270 /* Create table header */
271 FreeSpaceMap = (FSMHeader *) ShmemInitStruct("Free Space Map Header",sizeof(FSMHeader),&found);
272 if (FreeSpaceMap == NULL)
274 (errcode(ERRCODE_OUT_OF_MEMORY),
275 errmsg("insufficient shared memory for free space map")));
277 MemSet(FreeSpaceMap, 0, sizeof(FSMHeader));
279 /* Create hashtable for FSMRelations */
280 info.keysize = sizeof(RelFileNode);
281 info.entrysize = sizeof(FSMRelation);
282 info.hash = tag_hash;
284 FreeSpaceMapRelHash = ShmemInitHash("Free Space Map Hash",
285 MaxFSMRelations / 10,
288 (HASH_ELEM | HASH_FUNCTION));
290 if (!FreeSpaceMapRelHash)
292 (errcode(ERRCODE_OUT_OF_MEMORY),
293 errmsg("insufficient shared memory for free space map")));
299 /* Allocate page-storage arena */
300 nchunks = (MaxFSMPages - 1) / CHUNKPAGES + 1;
301 /* This check ensures spareChunks will be greater than zero */
302 if (nchunks <= MaxFSMRelations)
304 (errcode(ERRCODE_INVALID_PARAMETER_VALUE),
305 errmsg("max_fsm_pages must exceed max_fsm_relations * %d",
308 FreeSpaceMap->arena = (char *) ShmemAlloc(nchunks * CHUNKBYTES);
309 if (FreeSpaceMap->arena == NULL)
311 (errcode(ERRCODE_OUT_OF_MEMORY),
312 errmsg("insufficient shared memory for free space map")));
314 FreeSpaceMap->totalChunks = nchunks;
315 FreeSpaceMap->usedChunks = 0;
316 FreeSpaceMap->sumRequests = 0;
320 * Estimate amount of shmem space needed for FSM.
323 FreeSpaceShmemSize(void)
329 size = MAXALIGN(sizeof(FSMHeader));
331 /* hash table, including the FSMRelation objects */
332 size += hash_estimate_size(MaxFSMRelations, sizeof(FSMRelation));
334 /* page-storage arena */
335 nchunks = (MaxFSMPages - 1) / CHUNKPAGES + 1;
337 if (nchunks >= (INT_MAX / CHUNKBYTES))
339 (errcode(ERRCODE_INVALID_PARAMETER_VALUE),
340 errmsg("max_fsm_pages is too large")));
342 size += MAXALIGN(nchunks * CHUNKBYTES);
348 * GetPageWithFreeSpace - try to find a page in the given relation with
349 * at least the specified amount of free space.
351 * If successful, return the block number; if not, return InvalidBlockNumber.
353 * The caller must be prepared for the possibility that the returned page
354 * will turn out to have too little space available by the time the caller
355 * gets a lock on it. In that case, the caller should report the actual
356 * amount of free space available on that page and then try again (see
357 * RecordAndGetPageWithFreeSpace). If InvalidBlockNumber is returned,
358 * extend the relation.
361 GetPageWithFreeSpace(RelFileNode *rel, Size spaceNeeded)
364 BlockNumber freepage;
366 LWLockAcquire(FreeSpaceLock, LW_EXCLUSIVE);
369 * We always add a rel to the hashtable when it is inquired about.
371 fsmrel = create_fsm_rel(rel);
374 * Update the moving average of space requests. This code implements
375 * an exponential moving average with an equivalent period of about 63
376 * requests. Ignore silly requests, however, to ensure that the
377 * average stays sane.
379 if (spaceNeeded > 0 && spaceNeeded < BLCKSZ)
381 int cur_avg = (int) fsmrel->avgRequest;
383 cur_avg += ((int) spaceNeeded - cur_avg) / 32;
384 fsmrel->avgRequest = (Size) cur_avg;
386 freepage = find_free_space(fsmrel, spaceNeeded);
387 LWLockRelease(FreeSpaceLock);
392 * RecordAndGetPageWithFreeSpace - update info about a page and try again.
394 * We provide this combo form, instead of a separate Record operation,
395 * to save one lock and hash table lookup cycle.
398 RecordAndGetPageWithFreeSpace(RelFileNode *rel,
404 BlockNumber freepage;
406 /* Sanity check: ensure spaceAvail will fit into OffsetNumber */
407 AssertArg(oldSpaceAvail < BLCKSZ);
409 LWLockAcquire(FreeSpaceLock, LW_EXCLUSIVE);
412 * We always add a rel to the hashtable when it is inquired about.
414 fsmrel = create_fsm_rel(rel);
417 fsm_record_free_space(fsmrel, oldPage, oldSpaceAvail);
420 * Update the moving average of space requests, same as in
421 * GetPageWithFreeSpace.
423 if (spaceNeeded > 0 && spaceNeeded < BLCKSZ)
425 int cur_avg = (int) fsmrel->avgRequest;
427 cur_avg += ((int) spaceNeeded - cur_avg) / 32;
428 fsmrel->avgRequest = (Size) cur_avg;
431 freepage = find_free_space(fsmrel, spaceNeeded);
432 LWLockRelease(FreeSpaceLock);
437 * GetAvgFSMRequestSize - get average FSM request size for a relation.
439 * If the relation is not known to FSM, return a default value.
442 GetAvgFSMRequestSize(RelFileNode *rel)
447 LWLockAcquire(FreeSpaceLock, LW_EXCLUSIVE);
448 fsmrel = lookup_fsm_rel(rel);
450 result = fsmrel->avgRequest;
452 result = INITIAL_AVERAGE;
453 LWLockRelease(FreeSpaceLock);
458 * RecordRelationFreeSpace - record available-space info about a relation.
460 * Any pre-existing info about the relation is assumed obsolete and discarded.
462 * The given pageSpaces[] array must be sorted in order by blkno. Note that
463 * the FSM is at liberty to discard some or all of the data.
466 RecordRelationFreeSpace(RelFileNode *rel,
468 PageFreeSpaceInfo *pageSpaces)
472 /* Limit nPages to something sane */
475 else if (nPages > MaxFSMPages)
476 nPages = MaxFSMPages;
478 LWLockAcquire(FreeSpaceLock, LW_EXCLUSIVE);
481 * Note we don't record info about a relation unless there's already
482 * an FSM entry for it, implying someone has done GetPageWithFreeSpace
483 * for it. Inactive rels thus will not clutter the map simply by
486 fsmrel = lookup_fsm_rel(rel);
491 FSMPageData *newLocation;
493 curAlloc = realloc_fsm_rel(fsmrel, nPages, false);
494 curAllocPages = curAlloc * CHUNKPAGES;
497 * If the data fits in our current allocation, just copy it;
498 * otherwise must compress.
500 newLocation = (FSMPageData *)
501 (FreeSpaceMap->arena + fsmrel->firstChunk * CHUNKBYTES);
502 if (nPages <= curAllocPages)
506 for (i = 0; i < nPages; i++)
508 BlockNumber page = pageSpaces[i].blkno;
509 Size avail = pageSpaces[i].avail;
511 /* Check caller provides sorted data */
512 if (i > 0 && page <= pageSpaces[i - 1].blkno)
513 elog(ERROR, "free-space data is not in page order");
514 FSMPageSetPageNum(newLocation, page);
515 FSMPageSetSpace(newLocation, avail);
518 fsmrel->storedPages = nPages;
522 pack_incoming_pages(newLocation, curAllocPages,
524 fsmrel->storedPages = curAllocPages;
527 LWLockRelease(FreeSpaceLock);
531 * GetFreeIndexPage - like GetPageWithFreeSpace, but for indexes
534 GetFreeIndexPage(RelFileNode *rel)
537 BlockNumber freepage;
539 LWLockAcquire(FreeSpaceLock, LW_EXCLUSIVE);
542 * We always add a rel to the hashtable when it is inquired about.
544 fsmrel = create_fsm_rel(rel);
546 freepage = find_index_free_space(fsmrel);
547 LWLockRelease(FreeSpaceLock);
552 * RecordIndexFreeSpace - like RecordRelationFreeSpace, but for indexes
555 RecordIndexFreeSpace(RelFileNode *rel,
561 /* Limit nPages to something sane */
564 else if (nPages > MaxFSMPages)
565 nPages = MaxFSMPages;
567 LWLockAcquire(FreeSpaceLock, LW_EXCLUSIVE);
570 * Note we don't record info about a relation unless there's already
571 * an FSM entry for it, implying someone has done GetFreeIndexPage for
572 * it. Inactive rels thus will not clutter the map simply by being
575 fsmrel = lookup_fsm_rel(rel);
581 IndexFSMPageData *newLocation;
583 curAlloc = realloc_fsm_rel(fsmrel, nPages, true);
584 curAllocPages = curAlloc * INDEXCHUNKPAGES;
587 * If the data fits in our current allocation, just copy it;
588 * otherwise must compress. But compression is easy: we merely
589 * forget extra pages.
591 newLocation = (IndexFSMPageData *)
592 (FreeSpaceMap->arena + fsmrel->firstChunk * CHUNKBYTES);
593 if (nPages > curAllocPages)
594 nPages = curAllocPages;
596 for (i = 0; i < nPages; i++)
598 BlockNumber page = pages[i];
600 /* Check caller provides sorted data */
601 if (i > 0 && page <= pages[i - 1])
602 elog(ERROR, "free-space data is not in page order");
603 IndexFSMPageSetPageNum(newLocation, page);
606 fsmrel->storedPages = nPages;
608 LWLockRelease(FreeSpaceLock);
612 * FreeSpaceMapTruncateRel - adjust for truncation of a relation.
614 * We need to delete any stored data past the new relation length, so that
615 * we don't bogusly return removed block numbers.
618 FreeSpaceMapTruncateRel(RelFileNode *rel, BlockNumber nblocks)
622 LWLockAcquire(FreeSpaceLock, LW_EXCLUSIVE);
623 fsmrel = lookup_fsm_rel(rel);
628 /* Use lookup to locate first entry >= nblocks */
629 (void) lookup_fsm_page_entry(fsmrel, nblocks, &pageIndex);
630 /* Delete all such entries */
631 fsmrel->storedPages = pageIndex;
632 /* XXX should we adjust rel's lastPageCount and sumRequests? */
634 LWLockRelease(FreeSpaceLock);
638 * FreeSpaceMapForgetRel - forget all about a relation.
640 * This is called when a relation is deleted. Although we could just let
641 * the rel age out of the map, it's better to reclaim and reuse the space
645 FreeSpaceMapForgetRel(RelFileNode *rel)
649 LWLockAcquire(FreeSpaceLock, LW_EXCLUSIVE);
650 fsmrel = lookup_fsm_rel(rel);
652 delete_fsm_rel(fsmrel);
653 LWLockRelease(FreeSpaceLock);
657 * FreeSpaceMapForgetDatabase - forget all relations of a database.
659 * This is called during DROP DATABASE. As above, might as well reclaim
660 * map space sooner instead of later.
663 FreeSpaceMapForgetDatabase(Oid dbid)
668 LWLockAcquire(FreeSpaceLock, LW_EXCLUSIVE);
669 for (fsmrel = FreeSpaceMap->usageList; fsmrel; fsmrel = nextrel)
671 nextrel = fsmrel->nextUsage; /* in case we delete it */
672 if (fsmrel->key.dbNode == dbid)
673 delete_fsm_rel(fsmrel);
675 LWLockRelease(FreeSpaceLock);
679 * PrintFreeSpaceMapStatistics - print statistics about FSM contents
681 * The info is sent to ereport() with the specified message level. This is
682 * intended for use during VACUUM.
685 PrintFreeSpaceMapStatistics(int elevel)
693 LWLockAcquire(FreeSpaceLock, LW_EXCLUSIVE);
694 /* Count total space used --- tedious, but seems useful */
695 for (fsmrel = FreeSpaceMap->firstRel;
697 fsmrel = fsmrel->nextPhysical)
698 storedPages += fsmrel->storedPages;
699 /* Copy other stats before dropping lock */
700 numRels = FreeSpaceMap->numRels;
701 sumRequests = FreeSpaceMap->sumRequests;
702 LWLockRelease(FreeSpaceLock);
704 /* Convert stats to actual number of page slots needed */
705 needed = (sumRequests + numRels) * CHUNKPAGES;
708 (errmsg("free space map: %d relations, %d pages stored; %.0f total pages needed",
709 numRels, storedPages, needed),
710 errdetail("Allocated FSM size: %d relations + %d pages = %.0f kB shared memory.",
711 MaxFSMRelations, MaxFSMPages,
712 (double) FreeSpaceShmemSize() / 1024.0)));
716 * DumpFreeSpaceMap - dump contents of FSM into a disk file for later reload
718 * This is expected to be called during database shutdown, after updates to
719 * the FSM have stopped. We lock the FreeSpaceLock but that's purely pro
720 * forma --- if anyone else is still accessing FSM, there's a problem.
723 DumpFreeSpaceMap(int code, Datum arg)
726 char cachefilename[MAXPGPATH];
727 FsmCacheFileHeader header;
730 /* Try to create file */
731 snprintf(cachefilename, sizeof(cachefilename), "%s/%s",
732 DataDir, FSM_CACHE_FILENAME);
734 unlink(cachefilename); /* in case it exists w/wrong permissions */
736 fp = AllocateFile(cachefilename, PG_BINARY_W);
739 elog(LOG, "could not write \"%s\": %m", cachefilename);
743 LWLockAcquire(FreeSpaceLock, LW_EXCLUSIVE);
745 /* Write file header */
746 MemSet(&header, 0, sizeof(header));
747 strcpy(header.label, FSM_CACHE_LABEL);
748 header.endian = FSM_CACHE_ENDIAN;
749 header.version = FSM_CACHE_VERSION;
750 header.numRels = FreeSpaceMap->numRels;
751 if (fwrite(&header, 1, sizeof(header), fp) != sizeof(header))
754 /* For each relation, in order from least to most recently used... */
755 for (fsmrel = FreeSpaceMap->usageListTail;
757 fsmrel = fsmrel->priorUsage)
759 FsmCacheRelHeader relheader;
762 /* Write relation header */
763 MemSet(&relheader, 0, sizeof(relheader));
764 relheader.key = fsmrel->key;
765 relheader.isIndex = fsmrel->isIndex;
766 relheader.avgRequest = fsmrel->avgRequest;
767 relheader.lastPageCount = fsmrel->lastPageCount;
768 relheader.storedPages = fsmrel->storedPages;
769 if (fwrite(&relheader, 1, sizeof(relheader), fp) != sizeof(relheader))
772 /* Write the per-page data directly from the arena */
773 nPages = fsmrel->storedPages;
780 len = nPages * sizeof(IndexFSMPageData);
782 len = nPages * sizeof(FSMPageData);
784 (FreeSpaceMap->arena + fsmrel->firstChunk * CHUNKBYTES);
785 if (fwrite(data, 1, len, fp) != len)
791 LWLockRelease(FreeSpaceLock);
795 elog(LOG, "could not write \"%s\": %m", cachefilename);
796 /* Remove busted cache file */
797 unlink(cachefilename);
803 elog(LOG, "could not write \"%s\": %m", cachefilename);
806 LWLockRelease(FreeSpaceLock);
810 /* Remove busted cache file */
811 unlink(cachefilename);
815 * LoadFreeSpaceMap - load contents of FSM from a disk file
817 * This is expected to be called during database startup, before any FSM
818 * updates begin. We lock the FreeSpaceLock but that's purely pro
819 * forma --- if anyone else is accessing FSM yet, there's a problem.
821 * Notes: no complaint is issued if no cache file is found. If the file is
822 * found, it is deleted after reading. Thus, if we crash without a clean
823 * shutdown, the next cycle of life starts with no FSM data. To do otherwise,
824 * we'd need to do significantly more validation in this routine, because of
825 * the likelihood that what is in the dump file would be out-of-date, eg
826 * there might be entries for deleted or truncated rels.
829 LoadFreeSpaceMap(void)
832 char cachefilename[MAXPGPATH];
833 FsmCacheFileHeader header;
836 /* Try to open file */
837 snprintf(cachefilename, sizeof(cachefilename), "%s/%s",
838 DataDir, FSM_CACHE_FILENAME);
840 fp = AllocateFile(cachefilename, PG_BINARY_R);
844 elog(LOG, "could not read \"%s\": %m", cachefilename);
848 LWLockAcquire(FreeSpaceLock, LW_EXCLUSIVE);
850 /* Read and verify file header */
851 if (fread(&header, 1, sizeof(header), fp) != sizeof(header) ||
852 strcmp(header.label, FSM_CACHE_LABEL) != 0 ||
853 header.endian != FSM_CACHE_ENDIAN ||
854 header.version != FSM_CACHE_VERSION ||
857 elog(LOG, "bogus file header in \"%s\"", cachefilename);
861 /* For each relation, in order from least to most recently used... */
862 for (relno = 0; relno < header.numRels; relno++)
864 FsmCacheRelHeader relheader;
872 /* Read and verify relation header, as best we can */
873 if (fread(&relheader, 1, sizeof(relheader), fp) != sizeof(relheader) ||
874 (relheader.isIndex != false && relheader.isIndex != true) ||
875 relheader.avgRequest >= BLCKSZ ||
876 relheader.lastPageCount < 0 ||
877 relheader.storedPages < 0)
879 elog(LOG, "bogus rel header in \"%s\"", cachefilename);
883 /* Make sure lastPageCount doesn't exceed current MaxFSMPages */
884 if (relheader.lastPageCount > MaxFSMPages)
885 relheader.lastPageCount = MaxFSMPages;
887 /* Read the per-page data */
888 nPages = relheader.storedPages;
889 if (relheader.isIndex)
890 len = nPages * sizeof(IndexFSMPageData);
892 len = nPages * sizeof(FSMPageData);
893 data = (char *) palloc(len);
894 if (fread(data, 1, len, fp) != len)
896 elog(LOG, "premature EOF in \"%s\"", cachefilename);
902 * Okay, create the FSM entry and insert data into it. Since the
903 * rels were stored in reverse usage order, at the end of the loop
904 * they will be correctly usage-ordered in memory; and if
905 * MaxFSMRelations is less than it used to be, we will correctly
906 * drop the least recently used ones.
908 fsmrel = create_fsm_rel(&relheader.key);
909 fsmrel->avgRequest = relheader.avgRequest;
911 curAlloc = realloc_fsm_rel(fsmrel, relheader.lastPageCount,
913 if (relheader.isIndex)
915 IndexFSMPageData *newLocation;
917 curAllocPages = curAlloc * INDEXCHUNKPAGES;
920 * If the data fits in our current allocation, just copy it;
921 * otherwise must compress. But compression is easy: we
922 * merely forget extra pages.
924 newLocation = (IndexFSMPageData *)
925 (FreeSpaceMap->arena + fsmrel->firstChunk * CHUNKBYTES);
926 if (nPages > curAllocPages)
927 nPages = curAllocPages;
928 memcpy(newLocation, data, nPages * sizeof(IndexFSMPageData));
929 fsmrel->storedPages = nPages;
933 FSMPageData *newLocation;
935 curAllocPages = curAlloc * CHUNKPAGES;
938 * If the data fits in our current allocation, just copy it;
939 * otherwise must compress.
941 newLocation = (FSMPageData *)
942 (FreeSpaceMap->arena + fsmrel->firstChunk * CHUNKBYTES);
943 if (nPages <= curAllocPages)
945 memcpy(newLocation, data, nPages * sizeof(FSMPageData));
946 fsmrel->storedPages = nPages;
950 pack_existing_pages(newLocation, curAllocPages,
951 (FSMPageData *) data, nPages);
952 fsmrel->storedPages = curAllocPages;
962 LWLockRelease(FreeSpaceLock);
966 /* Remove cache file before it can become stale; see notes above */
967 unlink(cachefilename);
972 * Internal routines. These all assume the caller holds the FreeSpaceLock.
976 * Lookup a relation in the hash table. If not present, return NULL.
978 * The relation's position in the LRU list is not changed.
981 lookup_fsm_rel(RelFileNode *rel)
985 fsmrel = (FSMRelation *) hash_search(FreeSpaceMapRelHash,
996 * Lookup a relation in the hash table, creating an entry if not present.
998 * On successful lookup, the relation is moved to the front of the LRU list.
1000 static FSMRelation *
1001 create_fsm_rel(RelFileNode *rel)
1003 FSMRelation *fsmrel;
1006 fsmrel = (FSMRelation *) hash_search(FreeSpaceMapRelHash,
1012 (errcode(ERRCODE_OUT_OF_MEMORY),
1013 errmsg("out of shared memory")));
1017 /* New hashtable entry, initialize it (hash_search set the key) */
1018 fsmrel->isIndex = false; /* until we learn different */
1019 fsmrel->avgRequest = INITIAL_AVERAGE;
1020 fsmrel->lastPageCount = 0;
1021 fsmrel->firstChunk = -1; /* no space allocated */
1022 fsmrel->storedPages = 0;
1023 fsmrel->nextPage = 0;
1025 /* Discard lowest-priority existing rel, if we are over limit */
1026 if (FreeSpaceMap->numRels >= MaxFSMRelations)
1027 delete_fsm_rel(FreeSpaceMap->usageListTail);
1029 /* Add new entry at front of LRU list */
1030 link_fsm_rel_usage(fsmrel);
1031 fsmrel->nextPhysical = NULL; /* not in physical-storage list */
1032 fsmrel->priorPhysical = NULL;
1033 FreeSpaceMap->numRels++;
1034 /* sumRequests is unchanged because request must be zero */
1038 /* Existing entry, move to front of LRU list */
1039 if (fsmrel->priorUsage != NULL)
1041 unlink_fsm_rel_usage(fsmrel);
1042 link_fsm_rel_usage(fsmrel);
1050 * Remove an existing FSMRelation entry.
1053 delete_fsm_rel(FSMRelation *fsmrel)
1055 FSMRelation *result;
1057 FreeSpaceMap->sumRequests -= fsm_calc_request(fsmrel);
1058 unlink_fsm_rel_usage(fsmrel);
1059 unlink_fsm_rel_storage(fsmrel);
1060 FreeSpaceMap->numRels--;
1061 result = (FSMRelation *) hash_search(FreeSpaceMapRelHash,
1062 (void *) &(fsmrel->key),
1066 elog(ERROR, "FreeSpaceMap hashtable corrupted");
1070 * Reallocate space for a FSMRelation.
1072 * This is shared code for RecordRelationFreeSpace and RecordIndexFreeSpace.
1073 * The return value is the actual new allocation, in chunks.
1076 realloc_fsm_rel(FSMRelation *fsmrel, int nPages, bool isIndex)
1083 * Delete any existing entries, and update request status.
1085 fsmrel->storedPages = 0;
1086 FreeSpaceMap->sumRequests -= fsm_calc_request(fsmrel);
1087 fsmrel->lastPageCount = nPages;
1088 fsmrel->isIndex = isIndex;
1089 myRequest = fsm_calc_request(fsmrel);
1090 FreeSpaceMap->sumRequests += myRequest;
1091 myAlloc = fsm_calc_target_allocation(myRequest);
1094 * Need to reallocate space if (a) my target allocation is more than
1095 * my current allocation, AND (b) my actual immediate need
1096 * (myRequest+1 chunks) is more than my current allocation. Otherwise
1097 * just store the new data in-place.
1099 curAlloc = fsm_current_allocation(fsmrel);
1100 if (myAlloc > curAlloc && (myRequest + 1) > curAlloc && nPages > 0)
1102 /* Remove entry from storage list, and compact */
1103 unlink_fsm_rel_storage(fsmrel);
1104 compact_fsm_storage();
1105 /* Reattach to end of storage list */
1106 link_fsm_rel_storage(fsmrel);
1107 /* And allocate storage */
1108 fsmrel->firstChunk = FreeSpaceMap->usedChunks;
1109 FreeSpaceMap->usedChunks += myAlloc;
1111 /* Watch out for roundoff error */
1112 if (FreeSpaceMap->usedChunks > FreeSpaceMap->totalChunks)
1114 FreeSpaceMap->usedChunks = FreeSpaceMap->totalChunks;
1115 curAlloc = FreeSpaceMap->totalChunks - fsmrel->firstChunk;
1122 * Link a FSMRelation into the LRU list (always at the head).
1125 link_fsm_rel_usage(FSMRelation *fsmrel)
1127 fsmrel->priorUsage = NULL;
1128 fsmrel->nextUsage = FreeSpaceMap->usageList;
1129 FreeSpaceMap->usageList = fsmrel;
1130 if (fsmrel->nextUsage != NULL)
1131 fsmrel->nextUsage->priorUsage = fsmrel;
1133 FreeSpaceMap->usageListTail = fsmrel;
1137 * Delink a FSMRelation from the LRU list.
1140 unlink_fsm_rel_usage(FSMRelation *fsmrel)
1142 if (fsmrel->priorUsage != NULL)
1143 fsmrel->priorUsage->nextUsage = fsmrel->nextUsage;
1145 FreeSpaceMap->usageList = fsmrel->nextUsage;
1146 if (fsmrel->nextUsage != NULL)
1147 fsmrel->nextUsage->priorUsage = fsmrel->priorUsage;
1149 FreeSpaceMap->usageListTail = fsmrel->priorUsage;
1152 * We don't bother resetting fsmrel's links, since it's about to be
1153 * deleted or relinked at the head.
1158 * Link a FSMRelation into the storage-order list (always at the tail).
1161 link_fsm_rel_storage(FSMRelation *fsmrel)
1163 fsmrel->nextPhysical = NULL;
1164 fsmrel->priorPhysical = FreeSpaceMap->lastRel;
1165 if (FreeSpaceMap->lastRel != NULL)
1166 FreeSpaceMap->lastRel->nextPhysical = fsmrel;
1168 FreeSpaceMap->firstRel = fsmrel;
1169 FreeSpaceMap->lastRel = fsmrel;
1173 * Delink a FSMRelation from the storage-order list, if it's in it.
1176 unlink_fsm_rel_storage(FSMRelation *fsmrel)
1178 if (fsmrel->priorPhysical != NULL || FreeSpaceMap->firstRel == fsmrel)
1180 if (fsmrel->priorPhysical != NULL)
1181 fsmrel->priorPhysical->nextPhysical = fsmrel->nextPhysical;
1183 FreeSpaceMap->firstRel = fsmrel->nextPhysical;
1184 if (fsmrel->nextPhysical != NULL)
1185 fsmrel->nextPhysical->priorPhysical = fsmrel->priorPhysical;
1187 FreeSpaceMap->lastRel = fsmrel->priorPhysical;
1189 /* mark as not in list, since we may not put it back immediately */
1190 fsmrel->nextPhysical = NULL;
1191 fsmrel->priorPhysical = NULL;
1192 /* Also mark it as having no storage */
1193 fsmrel->firstChunk = -1;
1194 fsmrel->storedPages = 0;
1198 * Look to see if a page with at least the specified amount of space is
1199 * available in the given FSMRelation. If so, return its page number,
1200 * and advance the nextPage counter so that the next inquiry will return
1201 * a different page if possible; also update the entry to show that the
1202 * requested space is not available anymore. Return InvalidBlockNumber
1206 find_free_space(FSMRelation *fsmrel, Size spaceNeeded)
1209 int pagesToCheck, /* outer loop counter */
1210 pageIndex; /* current page index */
1212 if (fsmrel->isIndex)
1213 elog(ERROR, "find_free_space called for an index relation");
1214 info = (FSMPageData *)
1215 (FreeSpaceMap->arena + fsmrel->firstChunk * CHUNKBYTES);
1216 pageIndex = fsmrel->nextPage;
1217 /* Last operation may have left nextPage pointing past end */
1218 if (pageIndex >= fsmrel->storedPages)
1221 for (pagesToCheck = fsmrel->storedPages; pagesToCheck > 0; pagesToCheck--)
1223 FSMPageData *page = info + pageIndex;
1224 Size spaceAvail = FSMPageGetSpace(page);
1226 /* Check this page */
1227 if (spaceAvail >= spaceNeeded)
1230 * Found what we want --- adjust the entry, and update
1233 FSMPageSetSpace(page, spaceAvail - spaceNeeded);
1234 fsmrel->nextPage = pageIndex + 1;
1235 return FSMPageGetPageNum(page);
1237 /* Advance pageIndex, wrapping around if needed */
1238 if (++pageIndex >= fsmrel->storedPages)
1242 return InvalidBlockNumber; /* nothing found */
1246 * As above, but for index case --- we only deal in whole pages.
1249 find_index_free_space(FSMRelation *fsmrel)
1251 IndexFSMPageData *info;
1255 * If isIndex isn't set, it could be that RecordIndexFreeSpace() has
1256 * never yet been called on this relation, and we're still looking at
1257 * the default setting from create_fsm_rel(). If so, just act as
1258 * though there's no space.
1260 if (!fsmrel->isIndex)
1262 if (fsmrel->storedPages == 0)
1263 return InvalidBlockNumber;
1264 elog(ERROR, "find_index_free_space called for a non-index relation");
1268 * For indexes, there's no need for the nextPage state variable; we
1269 * just remove and return the first available page. (We could save
1270 * cycles here by returning the last page, but it seems better to
1271 * encourage re-use of lower-numbered pages.)
1273 if (fsmrel->storedPages <= 0)
1274 return InvalidBlockNumber; /* no pages available */
1275 info = (IndexFSMPageData *)
1276 (FreeSpaceMap->arena + fsmrel->firstChunk * CHUNKBYTES);
1277 result = IndexFSMPageGetPageNum(info);
1278 fsmrel->storedPages--;
1279 memmove(info, info + 1, fsmrel->storedPages * sizeof(IndexFSMPageData));
1284 * fsm_record_free_space - guts of RecordFreeSpace operation (now only
1285 * provided as part of RecordAndGetPageWithFreeSpace).
1288 fsm_record_free_space(FSMRelation *fsmrel, BlockNumber page, Size spaceAvail)
1292 if (fsmrel->isIndex)
1293 elog(ERROR, "fsm_record_free_space called for an index relation");
1294 if (lookup_fsm_page_entry(fsmrel, page, &pageIndex))
1296 /* Found an existing entry for page; update it */
1299 info = (FSMPageData *)
1300 (FreeSpaceMap->arena + fsmrel->firstChunk * CHUNKBYTES);
1302 FSMPageSetSpace(info, spaceAvail);
1307 * No existing entry; ignore the call. We used to add the page to
1308 * the FSM --- but in practice, if the page hasn't got enough
1309 * space to satisfy the caller who's kicking it back to us, then
1310 * it's probably uninteresting to everyone else as well.
1316 * Look for an entry for a specific page (block number) in a FSMRelation.
1317 * Returns TRUE if a matching entry exists, else FALSE.
1319 * The output argument *outPageIndex is set to indicate where the entry exists
1320 * (if TRUE result) or could be inserted (if FALSE result).
1323 lookup_fsm_page_entry(FSMRelation *fsmrel, BlockNumber page,
1326 /* Check for empty relation */
1327 if (fsmrel->storedPages <= 0)
1333 /* Do binary search */
1334 if (fsmrel->isIndex)
1336 IndexFSMPageData *info;
1340 info = (IndexFSMPageData *)
1341 (FreeSpaceMap->arena + fsmrel->firstChunk * CHUNKBYTES);
1343 high = fsmrel->storedPages - 1;
1349 middle = low + (high - low) / 2;
1350 probe = IndexFSMPageGetPageNum(info + middle);
1353 *outPageIndex = middle;
1356 else if (probe < page)
1361 *outPageIndex = low;
1370 info = (FSMPageData *)
1371 (FreeSpaceMap->arena + fsmrel->firstChunk * CHUNKBYTES);
1373 high = fsmrel->storedPages - 1;
1379 middle = low + (high - low) / 2;
1380 probe = FSMPageGetPageNum(info + middle);
1383 *outPageIndex = middle;
1386 else if (probe < page)
1391 *outPageIndex = low;
1397 * Re-pack the FSM storage arena, dropping data if necessary to meet the
1398 * current allocation target for each relation. At conclusion, all available
1399 * space in the arena will be coalesced at the end.
1402 compact_fsm_storage(void)
1404 int nextChunkIndex = 0;
1405 bool did_push = false;
1406 FSMRelation *fsmrel;
1408 for (fsmrel = FreeSpaceMap->firstRel;
1410 fsmrel = fsmrel->nextPhysical)
1421 * Calculate target allocation, make sure we don't overrun due to
1424 newAlloc = fsm_calc_target_allocation(fsm_calc_request(fsmrel));
1425 if (newAlloc > FreeSpaceMap->totalChunks - nextChunkIndex)
1426 newAlloc = FreeSpaceMap->totalChunks - nextChunkIndex;
1427 if (fsmrel->isIndex)
1428 newAllocPages = newAlloc * INDEXCHUNKPAGES;
1430 newAllocPages = newAlloc * CHUNKPAGES;
1433 * Determine current size, current and new locations
1435 curChunks = fsm_current_chunks(fsmrel);
1436 oldChunkIndex = fsmrel->firstChunk;
1437 oldLocation = FreeSpaceMap->arena + oldChunkIndex * CHUNKBYTES;
1438 newChunkIndex = nextChunkIndex;
1439 newLocation = FreeSpaceMap->arena + newChunkIndex * CHUNKBYTES;
1442 * It's possible that we have to move data down, not up, if the
1443 * allocations of previous rels expanded. This normally means that
1444 * our allocation expanded too (or at least got no worse), and
1445 * ditto for later rels. So there should be room to move all our
1446 * data down without dropping any --- but we might have to push down
1447 * following rels to acquire the room. We don't want to do the push
1448 * more than once, so pack everything against the end of the arena
1451 * In corner cases where we are on the short end of a roundoff choice
1452 * that we were formerly on the long end of, it's possible that we
1453 * have to move down and compress our data too. In fact, even after
1454 * pushing down the following rels, there might not be as much space
1455 * as we computed for this rel above --- that would imply that some
1456 * following rel(s) are also on the losing end of roundoff choices.
1457 * We could handle this fairly by doing the per-rel compactions
1458 * out-of-order, but that seems like way too much complexity to deal
1459 * with a very infrequent corner case. Instead, we simply drop pages
1460 * from the end of the current rel's data until it fits.
1462 if (newChunkIndex > oldChunkIndex)
1464 int limitChunkIndex;
1466 if (newAllocPages < fsmrel->storedPages)
1468 /* move and compress --- just drop excess pages */
1469 fsmrel->storedPages = newAllocPages;
1470 curChunks = fsm_current_chunks(fsmrel);
1472 /* is there enough space? */
1473 if (fsmrel->nextPhysical != NULL)
1474 limitChunkIndex = fsmrel->nextPhysical->firstChunk;
1476 limitChunkIndex = FreeSpaceMap->totalChunks;
1477 if (newChunkIndex + curChunks > limitChunkIndex)
1479 /* not enough space, push down following rels */
1482 push_fsm_rels_after(fsmrel);
1485 /* now is there enough space? */
1486 if (fsmrel->nextPhysical != NULL)
1487 limitChunkIndex = fsmrel->nextPhysical->firstChunk;
1489 limitChunkIndex = FreeSpaceMap->totalChunks;
1490 if (newChunkIndex + curChunks > limitChunkIndex)
1492 /* uh-oh, forcibly cut the allocation to fit */
1493 newAlloc = limitChunkIndex - newChunkIndex;
1495 * If newAlloc < 0 at this point, we are moving the rel's
1496 * firstChunk into territory currently assigned to a later
1497 * rel. This is okay so long as we do not copy any data.
1498 * The rels will be back in nondecreasing firstChunk order
1499 * at completion of the compaction pass.
1503 if (fsmrel->isIndex)
1504 newAllocPages = newAlloc * INDEXCHUNKPAGES;
1506 newAllocPages = newAlloc * CHUNKPAGES;
1507 fsmrel->storedPages = newAllocPages;
1508 curChunks = fsm_current_chunks(fsmrel);
1511 memmove(newLocation, oldLocation, curChunks * CHUNKBYTES);
1513 else if (newAllocPages < fsmrel->storedPages)
1516 * Need to compress the page data. For an index,
1517 * "compression" just means dropping excess pages; otherwise
1518 * we try to keep the ones with the most space.
1520 if (fsmrel->isIndex)
1522 fsmrel->storedPages = newAllocPages;
1523 /* may need to move data */
1524 if (newChunkIndex != oldChunkIndex)
1525 memmove(newLocation, oldLocation, newAlloc * CHUNKBYTES);
1529 pack_existing_pages((FSMPageData *) newLocation,
1531 (FSMPageData *) oldLocation,
1532 fsmrel->storedPages);
1533 fsmrel->storedPages = newAllocPages;
1536 else if (newChunkIndex != oldChunkIndex)
1539 * No compression needed, but must copy the data up
1541 memmove(newLocation, oldLocation, curChunks * CHUNKBYTES);
1543 fsmrel->firstChunk = newChunkIndex;
1544 nextChunkIndex += newAlloc;
1546 Assert(nextChunkIndex <= FreeSpaceMap->totalChunks);
1547 FreeSpaceMap->usedChunks = nextChunkIndex;
1551 * Push all FSMRels physically after afterRel to the end of the storage arena.
1553 * We sometimes have to do this when deletion or truncation of a relation
1554 * causes the allocations of remaining rels to expand markedly. We must
1555 * temporarily push existing data down to the end so that we can move it
1556 * back up in an orderly fashion.
1559 push_fsm_rels_after(FSMRelation *afterRel)
1561 int nextChunkIndex = FreeSpaceMap->totalChunks;
1562 FSMRelation *fsmrel;
1564 FreeSpaceMap->usedChunks = FreeSpaceMap->totalChunks;
1566 for (fsmrel = FreeSpaceMap->lastRel;
1568 fsmrel = fsmrel->priorPhysical)
1576 if (fsmrel == afterRel)
1579 chunkCount = fsm_current_chunks(fsmrel);
1580 nextChunkIndex -= chunkCount;
1581 newChunkIndex = nextChunkIndex;
1582 oldChunkIndex = fsmrel->firstChunk;
1583 if (newChunkIndex < oldChunkIndex)
1585 /* we're pushing down, how can it move up? */
1586 elog(PANIC, "inconsistent entry sizes in FSM");
1588 else if (newChunkIndex > oldChunkIndex)
1590 /* need to move it */
1591 newLocation = FreeSpaceMap->arena + newChunkIndex * CHUNKBYTES;
1592 oldLocation = FreeSpaceMap->arena + oldChunkIndex * CHUNKBYTES;
1593 memmove(newLocation, oldLocation, chunkCount * CHUNKBYTES);
1594 fsmrel->firstChunk = newChunkIndex;
1597 Assert(nextChunkIndex >= 0);
1601 * Pack a set of per-page freespace data into a smaller amount of space.
1603 * The method is to compute a low-resolution histogram of the free space
1604 * amounts, then determine which histogram bin contains the break point.
1605 * We then keep all pages above that bin, none below it, and just enough
1606 * of the pages in that bin to fill the output area exactly.
1608 #define HISTOGRAM_BINS 64
1611 pack_incoming_pages(FSMPageData *newLocation, int newPages,
1612 PageFreeSpaceInfo *pageSpaces, int nPages)
1614 int histogram[HISTOGRAM_BINS];
1621 Assert(newPages < nPages); /* else I shouldn't have been called */
1622 /* Build histogram */
1623 MemSet(histogram, 0, sizeof(histogram));
1624 for (i = 0; i < nPages; i++)
1626 Size avail = pageSpaces[i].avail;
1628 if (avail >= BLCKSZ)
1629 elog(ERROR, "bogus freespace amount");
1630 avail /= (BLCKSZ / HISTOGRAM_BINS);
1633 /* Find the breakpoint bin */
1635 for (i = HISTOGRAM_BINS - 1; i >= 0; i--)
1637 int sum = above + histogram[i];
1644 thresholdL = i * BLCKSZ / HISTOGRAM_BINS; /* low bound of bp bin */
1645 thresholdU = (i + 1) * BLCKSZ / HISTOGRAM_BINS; /* hi bound */
1646 binct = newPages - above; /* number to take from bp bin */
1647 /* And copy the appropriate data */
1648 for (i = 0; i < nPages; i++)
1650 BlockNumber page = pageSpaces[i].blkno;
1651 Size avail = pageSpaces[i].avail;
1653 /* Check caller provides sorted data */
1654 if (i > 0 && page <= pageSpaces[i - 1].blkno)
1655 elog(ERROR, "free-space data is not in page order");
1656 /* Save this page? */
1657 if (avail >= thresholdU ||
1658 (avail >= thresholdL && (--binct >= 0)))
1660 FSMPageSetPageNum(newLocation, page);
1661 FSMPageSetSpace(newLocation, avail);
1666 Assert(newPages == 0);
1670 * Pack a set of per-page freespace data into a smaller amount of space.
1672 * This is algorithmically identical to pack_incoming_pages(), but accepts
1673 * a different input representation. Also, we assume the input data has
1674 * previously been checked for validity (size in bounds, pages in order).
1676 * Note: it is possible for the source and destination arrays to overlap.
1677 * The caller is responsible for making sure newLocation is at lower addresses
1678 * so that we can copy data moving forward in the arrays without problem.
1681 pack_existing_pages(FSMPageData *newLocation, int newPages,
1682 FSMPageData *oldLocation, int oldPages)
1684 int histogram[HISTOGRAM_BINS];
1691 Assert(newPages < oldPages); /* else I shouldn't have been called */
1692 /* Build histogram */
1693 MemSet(histogram, 0, sizeof(histogram));
1694 for (i = 0; i < oldPages; i++)
1696 Size avail = FSMPageGetSpace(oldLocation + i);
1698 /* Shouldn't happen, but test to protect against stack clobber */
1699 if (avail >= BLCKSZ)
1700 elog(ERROR, "bogus freespace amount");
1701 avail /= (BLCKSZ / HISTOGRAM_BINS);
1704 /* Find the breakpoint bin */
1706 for (i = HISTOGRAM_BINS - 1; i >= 0; i--)
1708 int sum = above + histogram[i];
1715 thresholdL = i * BLCKSZ / HISTOGRAM_BINS; /* low bound of bp bin */
1716 thresholdU = (i + 1) * BLCKSZ / HISTOGRAM_BINS; /* hi bound */
1717 binct = newPages - above; /* number to take from bp bin */
1718 /* And copy the appropriate data */
1719 for (i = 0; i < oldPages; i++)
1721 BlockNumber page = FSMPageGetPageNum(oldLocation + i);
1722 Size avail = FSMPageGetSpace(oldLocation + i);
1724 /* Save this page? */
1725 if (avail >= thresholdU ||
1726 (avail >= thresholdL && (--binct >= 0)))
1728 FSMPageSetPageNum(newLocation, page);
1729 FSMPageSetSpace(newLocation, avail);
1734 Assert(newPages == 0);
1738 * Calculate number of chunks "requested" by a rel.
1740 * Rel's lastPageCount and isIndex settings must be up-to-date when called.
1742 * See notes at top of file for details.
1745 fsm_calc_request(FSMRelation *fsmrel)
1749 /* Convert page count to chunk count */
1750 if (fsmrel->isIndex)
1751 chunkCount = (fsmrel->lastPageCount - 1) / INDEXCHUNKPAGES + 1;
1753 chunkCount = (fsmrel->lastPageCount - 1) / CHUNKPAGES + 1;
1754 /* "Request" is anything beyond our one guaranteed chunk */
1755 if (chunkCount <= 0)
1758 return chunkCount - 1;
1762 * Calculate target allocation (number of chunks) for a rel
1764 * Parameter is the result from fsm_calc_request(). The global sumRequests
1765 * and numRels totals must be up-to-date already.
1767 * See notes at top of file for details.
1770 fsm_calc_target_allocation(int myRequest)
1775 spareChunks = FreeSpaceMap->totalChunks - FreeSpaceMap->numRels;
1776 Assert(spareChunks > 0);
1777 if (spareChunks >= FreeSpaceMap->sumRequests)
1779 /* We aren't oversubscribed, so allocate exactly the request */
1784 extra = (int) rint(spareChunks * myRequest / FreeSpaceMap->sumRequests);
1785 if (extra < 0) /* shouldn't happen, but make sure */
1792 * Calculate number of chunks actually used to store current data
1795 fsm_current_chunks(FSMRelation *fsmrel)
1799 /* Make sure storedPages==0 produces right answer */
1800 if (fsmrel->storedPages <= 0)
1802 /* Convert page count to chunk count */
1803 if (fsmrel->isIndex)
1804 chunkCount = (fsmrel->storedPages - 1) / INDEXCHUNKPAGES + 1;
1806 chunkCount = (fsmrel->storedPages - 1) / CHUNKPAGES + 1;
1811 * Calculate current actual allocation (number of chunks) for a rel
1814 fsm_current_allocation(FSMRelation *fsmrel)
1816 if (fsmrel->nextPhysical != NULL)
1817 return fsmrel->nextPhysical->firstChunk - fsmrel->firstChunk;
1818 else if (fsmrel == FreeSpaceMap->lastRel)
1819 return FreeSpaceMap->usedChunks - fsmrel->firstChunk;
1822 /* it's not in the storage-order list */
1823 Assert(fsmrel->firstChunk < 0 && fsmrel->storedPages == 0);
1829 #ifdef FREESPACE_DEBUG
1831 * Dump contents of freespace map for debugging.
1833 * We assume caller holds the FreeSpaceLock, or is otherwise unconcerned
1834 * about other processes.
1839 FSMRelation *fsmrel;
1840 FSMRelation *prevrel = NULL;
1844 for (fsmrel = FreeSpaceMap->usageList; fsmrel; fsmrel = fsmrel->nextUsage)
1847 fprintf(stderr, "Map %d: rel %u/%u/%u isIndex %d avgRequest %u lastPageCount %d nextPage %d\nMap= ",
1849 fsmrel->key.spcNode, fsmrel->key.dbNode, fsmrel->key.relNode,
1850 (int) fsmrel->isIndex, fsmrel->avgRequest,
1851 fsmrel->lastPageCount, fsmrel->nextPage);
1852 if (fsmrel->isIndex)
1854 IndexFSMPageData *page;
1856 page = (IndexFSMPageData *)
1857 (FreeSpaceMap->arena + fsmrel->firstChunk * CHUNKBYTES);
1858 for (nPages = 0; nPages < fsmrel->storedPages; nPages++)
1860 fprintf(stderr, " %u",
1861 IndexFSMPageGetPageNum(page));
1869 page = (FSMPageData *)
1870 (FreeSpaceMap->arena + fsmrel->firstChunk * CHUNKBYTES);
1871 for (nPages = 0; nPages < fsmrel->storedPages; nPages++)
1873 fprintf(stderr, " %u:%u",
1874 FSMPageGetPageNum(page),
1875 FSMPageGetSpace(page));
1879 fprintf(stderr, "\n");
1880 /* Cross-check list links */
1881 if (prevrel != fsmrel->priorUsage)
1882 fprintf(stderr, "DumpFreeSpace: broken list links\n");
1885 if (prevrel != FreeSpaceMap->usageListTail)
1886 fprintf(stderr, "DumpFreeSpace: broken list links\n");
1887 /* Cross-check global counters */
1888 if (relNum != FreeSpaceMap->numRels)
1889 fprintf(stderr, "DumpFreeSpace: %d rels in list, but numRels = %d\n",
1890 relNum, FreeSpaceMap->numRels);
1893 #endif /* FREESPACE_DEBUG */