/ Check-in [638a39bb]
Login

Many hyperlinks are disabled.
Use anonymous login to enable hyperlinks.

Overview
Comment:Further work on mmap(). Still does not work right - autovacuum tests are the first to fail.
Downloads: Tarball | ZIP archive | SQL archive
Timelines: family | ancestors | mmap-experimental
Files: files | file ages | folders
SHA1:638a39bbaa034a08d73459a36bbe14b76b34189f
User & Date: drh 2011-11-14 12:34:15
Context
2011-11-14
12:34
Further work on mmap(). Still does not work right - autovacuum tests are the first to fail. Closed-Leaf check-in: 638a39bb user: drh tags: mmap-experimental
01:55
Begin making experimental changes to use mmap() for reading content from a database. The code compiles, but crashes on the test suite. check-in: 09be42d5 user: drh tags: mmap-experimental
Changes
Hide Diffs Unified Diffs Ignore Whitespace Patch

Changes to src/btree.c.

723
724
725
726
727
728
729










730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
...
806
807
808
809
810
811
812

813
814
815
816
817
818
819
....
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
....
2319
2320
2321
2322
2323
2324
2325

2326
2327
2328
2329
2330
2331
2332
2333
2334
2335
2336

2337
2338
2339
2340
2341
2342
2343
2344
2345
2346
2347
2348
2349
2350
2351
2352
....
2493
2494
2495
2496
2497
2498
2499
2500
2501
2502
2503
2504
2505
2506
2507
....
2663
2664
2665
2666
2667
2668
2669

2670
2671
2672
2673
2674
2675
2676
....
2909
2910
2911
2912
2913
2914
2915

2916
2917
2918
2919
2920
2921
2922
....
3061
3062
3063
3064
3065
3066
3067

3068
3069
3070
3071
3072
3073
3074
....
3420
3421
3422
3423
3424
3425
3426
3427
3428
3429
3430
3431
3432
3433
3434
....
3959
3960
3961
3962
3963
3964
3965
3966
3967
3968
3969
3970
3971
3972
3973
....
4821
4822
4823
4824
4825
4826
4827
4828
4829
4830
4831
4832
4833
4834
4835
....
4996
4997
4998
4999
5000
5001
5002

5003
5004
5005
5006
5007
5008
5009
....
7366
7367
7368
7369
7370
7371
7372
7373
7374
7375
7376
7377
7378
7379
7380
....
7926
7927
7928
7929
7930
7931
7932

7933
7934
7935
7936
7937
7938
7939
....
8213
8214
8215
8216
8217
8218
8219
8220
8221
8222
8223
8224
8225
8226
8227
8228
8229
8230
8231
8232
8233
8234
8235
8236
  if( pCur->eState!=CURSOR_VALID || pCur->skipNext!=0 ){
    *pHasMoved = 1;
  }else{
    *pHasMoved = 0;
  }
  return SQLITE_OK;
}











/*
** Make a btree page is writable.
*/
static int btreeMakePageWriteable(MemPage *pPage){
  int rc;
  if( sqlite3PagerIswriteable(pPage->pDbPage) ) return SQLITE_OK;
  rc = sqlite3PagerWrite(pPage->pDbPage);
  pPage->aData = sqlite3PagerGetData(pPage->pDbPage);
  pPage->aDataEnd = &pPage->aData[pPage->pBt->usableSize];
  pPage->aCellIdx = &pPage->aData[pPage->cellOffset];
  return rc;
}

#ifndef SQLITE_OMIT_AUTOVACUUM
/*
** Given a page number of a regular database page, return the page
** number for the pointer-map page that contains the entry for the
................................................................................
  }
  assert( offset <= (int)pBt->usableSize-5 );
  pPtrmap = (u8 *)sqlite3PagerGetData(pDbPage);

  if( eType!=pPtrmap[offset] || get4byte(&pPtrmap[offset+1])!=parent ){
    TRACE(("PTRMAP_UPDATE: %d->(%d,%d)\n", key, eType, parent));
    *pRC= rc = sqlite3PagerWrite(pDbPage);

    if( rc==SQLITE_OK ){
      pPtrmap[offset] = eType;
      put4byte(&pPtrmap[offset+1], parent);
    }
  }

ptrmap_exit:
................................................................................
** call to btreeGetPage.
*/
static void releasePage(MemPage *pPage){
  if( pPage ){
    assert( pPage->aData );
    assert( pPage->pBt );
    assert( sqlite3PagerGetExtra(pPage->pDbPage) == (void*)pPage );
    assert( sqlite3PagerGetData(pPage->pDbPage)==pPage->aData );
    assert( sqlite3_mutex_held(pPage->pBt->mutex) );
    sqlite3PagerUnref(pPage->pDbPage);
  }
}

/*
** During a rollback, when the pager reloads information into the cache
................................................................................
*/
static int lockBtree(BtShared *pBt){
  int rc;              /* Result code from subfunctions */
  MemPage *pPage1;     /* Page 1 of the database file */
  int nPage;           /* Number of pages in the database */
  int nPageFile = 0;   /* Number of pages in the database file */
  int nPageHeader;     /* Number of pages in the database according to hdr */


  assert( sqlite3_mutex_held(pBt->mutex) );
  assert( pBt->pPage1==0 );
  rc = sqlite3PagerSharedLock(pBt->pPager);
  if( rc!=SQLITE_OK ) return rc;
  rc = btreeGetPage(pBt, 1, &pPage1, 0);
  if( rc!=SQLITE_OK ) return rc;

  /* Do some checking to help insure the file we opened really is
  ** a valid database file. 
  */

  nPage = nPageHeader = get4byte(28+(u8*)pPage1->aData);
  sqlite3PagerPagecount(pBt->pPager, &nPageFile);
  if( nPage==0 || memcmp(24+(u8*)pPage1->aData, 92+(u8*)pPage1->aData,4)!=0 ){
    nPage = nPageFile;
  }
  if( nPage>0 ){
    u32 pageSize;
    u32 usableSize;
    u8 *page1 = pPage1->aData;
    rc = SQLITE_NOTADB;
    if( memcmp(page1, zMagicHeader, 16)!=0 ){
      goto page1_init_failed;
    }

#ifdef SQLITE_OMIT_WAL
    if( page1[18]>1 ){
................................................................................
  if( pBt->nPage>0 ){
    return SQLITE_OK;
  }
  pP1 = pBt->pPage1;
  assert( pP1!=0 );
  rc = btreeMakePageWriteable(pP1);
  if( rc ) return rc;
  data = pP1->aData;
  memcpy(data, zMagicHeader, sizeof(zMagicHeader));
  assert( sizeof(zMagicHeader)==16 );
  data[16] = (u8)((pBt->pageSize>>8)&0xff);
  data[17] = (u8)((pBt->pageSize>>16)&0xff);
  data[18] = 1;
  data[19] = 1;
  assert( pBt->usableSize<=pBt->pageSize && pBt->usableSize+255>=pBt->pageSize);
................................................................................

      /* If the db-size header field is incorrect (as it may be if an old
      ** client has been writing the database file), update it now. Doing
      ** this sooner rather than later means the database size can safely 
      ** re-read the database size from page 1 if a savepoint or transaction
      ** rollback occurs within the transaction.
      */

      if( pBt->nPage!=get4byte(&pPage1->aData[28]) ){
        rc = btreeMakePageWriteable(pPage1);
        if( rc==SQLITE_OK ){
          put4byte(&pPage1->aData[28], pBt->nPage);
        }
      }
    }
................................................................................
  assert( sqlite3_mutex_held(pBt->mutex) );
  assert( iLastPg>nFin );

  if( !PTRMAP_ISPAGE(pBt, iLastPg) && iLastPg!=PENDING_BYTE_PAGE(pBt) ){
    u8 eType;
    Pgno iPtrPage;


    nFreeList = get4byte(&pBt->pPage1->aData[36]);
    if( nFreeList==0 ){
      return SQLITE_DONE;
    }

    rc = ptrmapGet(pBt, iLastPg, &eType, &iPtrPage);
    if( rc!=SQLITE_OK ){
................................................................................
      /* It is not possible to create a database for which the final page
      ** is either a pointer-map page or the pending-byte page. If one
      ** is encountered, this indicates corruption.
      */
      return SQLITE_CORRUPT_BKPT;
    }


    nFree = get4byte(&pBt->pPage1->aData[36]);
    nEntry = pBt->usableSize/5;
    nPtrmap = (nFree-nOrig+PTRMAP_PAGENO(pBt, nOrig)+nEntry)/nEntry;
    nFin = nOrig - nFree - nPtrmap;
    if( nOrig>PENDING_BYTE_PAGE(pBt) && nFin<PENDING_BYTE_PAGE(pBt) ){
      nFin--;
    }
................................................................................
    assert( op==SAVEPOINT_RELEASE || op==SAVEPOINT_ROLLBACK );
    assert( iSavepoint>=0 || (iSavepoint==-1 && op==SAVEPOINT_ROLLBACK) );
    sqlite3BtreeEnter(p);
    rc = sqlite3PagerSavepoint(pBt->pPager, op, iSavepoint);
    if( rc==SQLITE_OK ){
      if( iSavepoint<0 && pBt->initiallyEmpty ) pBt->nPage = 0;
      rc = newDatabase(pBt);
      pBt->nPage = get4byte(28 + pBt->pPage1->aData);

      /* The database size was written into the offset 28 of the header
      ** when the transaction started, so we know that the value at offset
      ** 28 is nonzero. */
      assert( pBt->nPage>0 );
    }
    sqlite3BtreeLeave(p);
................................................................................
        ** output buffer, bypassing the page-cache altogether. This speeds
        ** up loading large records that span many overflow pages.
        */
        if( eOp==0                                             /* (1) */
         && offset==0                                          /* (2) */
         && pBt->inTransaction==TRANS_READ                     /* (4) */
         && (fd = sqlite3PagerFile(pBt->pPager))->pMethods     /* (3) */
         && pBt->pPage1->aData[19]==0x01                       /* (5) */
        ){
          u8 aSave[4];
          u8 *aWrite = &pBuf[-4];
          memcpy(aSave, aWrite, 4);
          rc = sqlite3OsRead(fd, aWrite, a+4, pBt->pageSize * (nextPage-1));
          nextPage = get4byte(aWrite);
          memcpy(aWrite, aSave, 4);
................................................................................
  MemPage *pTrunk = 0;
  MemPage *pPrevTrunk = 0;
  Pgno mxPage;     /* Total size of the database file */

  assert( sqlite3_mutex_held(pBt->mutex) );
  pPage1 = pBt->pPage1;
  mxPage = btreePagecount(pBt);
  n = get4byte(&pPage1->aData[36]);
  testcase( n==mxPage-1 );
  if( n>=mxPage ){
    return SQLITE_CORRUPT_BKPT;
  }
  if( n>0 ){
    /* There are pages on the freelist.  Reuse one of those pages. */
    Pgno iTrunk;
................................................................................
          int noContent;
          *pPgno = iPage;
          TRACE(("ALLOCATE: %d was leaf %d of %d on trunk %d"
                 ": %d more free pages\n",
                 *pPgno, closest+1, k, pTrunk->pgno, n-1));
          rc = btreeMakePageWriteable(pTrunk);
          if( rc ) goto end_allocate_page;

          if( closest<k-1 ){
            memcpy(&aData[8+closest*4], &aData[4+k*4], 4);
          }
          put4byte(&aData[4], k-1);
          noContent = !btreeGetHasContent(pBt, *pPgno);
          rc = btreeGetPage(pBt, *pPgno, ppPage, noContent);
          if( rc==SQLITE_OK ){
................................................................................

  sqlite3BtreeEnter(p);
  assert( p->inTrans>TRANS_NONE );
  assert( SQLITE_OK==querySharedCacheTableLock(p, MASTER_ROOT, READ_LOCK) );
  assert( pBt->pPage1 );
  assert( idx>=0 && idx<=15 );

  *pMeta = get4byte(&pBt->pPage1->aData[36 + idx*4]);

  /* If auto-vacuum is disabled in this build and this is an auto-vacuum
  ** database, mark the database as read-only.  */
#ifdef SQLITE_OMIT_AUTOVACUUM
  if( idx==BTREE_LARGEST_ROOT_PAGE && *pMeta>0 ) pBt->readOnly = 1;
#endif

................................................................................
    sCheck.anRef[i] = 1;
  }
  sqlite3StrAccumInit(&sCheck.errMsg, zErr, sizeof(zErr), 20000);
  sCheck.errMsg.useMalloc = 2;

  /* Check the integrity of the freelist
  */

  checkList(&sCheck, 1, get4byte(&pBt->pPage1->aData[32]),
            get4byte(&pBt->pPage1->aData[36]), "Main freelist: ");

  /* Check all the tables.
  */
  for(i=0; (int)i<nRoot && sCheck.mxErr; i++){
    if( aRoot[i]==0 ) continue;
................................................................................
  /* If setting the version fields to 1, do not automatically open the
  ** WAL connection, even if the version fields are currently set to 2.
  */
  pBt->doNotUseWAL = (u8)(iVersion==1);

  rc = sqlite3BtreeBeginTrans(pBtree, 0);
  if( rc==SQLITE_OK ){
    u8 *aData = pBt->pPage1->aData;
    if( aData[18]!=(u8)iVersion || aData[19]!=(u8)iVersion ){
      rc = sqlite3BtreeBeginTrans(pBtree, 2);
      if( rc==SQLITE_OK ){
        rc = btreeMakePageWriteable(pBt->pPage1);
        if( rc==SQLITE_OK ){
          aData = pBt->pPage1->aData;
          aData[18] = (u8)iVersion;
          aData[19] = (u8)iVersion;
        }
      }
    }
  }

  pBt->doNotUseWAL = 0;
  return rc;
}







>
>
>
>
>
>
>
>
>
>






<

|
<
<







 







>







 







|







 







>











>
|

|





<







 







|







 







>







 







>







 







>







 







|







 







|







 







|







 







>







 







|







 







>







 







|





|










723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745

746
747


748
749
750
751
752
753
754
...
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
....
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
....
2327
2328
2329
2330
2331
2332
2333
2334
2335
2336
2337
2338
2339
2340
2341
2342
2343
2344
2345
2346
2347
2348
2349
2350
2351
2352
2353
2354

2355
2356
2357
2358
2359
2360
2361
....
2502
2503
2504
2505
2506
2507
2508
2509
2510
2511
2512
2513
2514
2515
2516
....
2672
2673
2674
2675
2676
2677
2678
2679
2680
2681
2682
2683
2684
2685
2686
....
2919
2920
2921
2922
2923
2924
2925
2926
2927
2928
2929
2930
2931
2932
2933
....
3072
3073
3074
3075
3076
3077
3078
3079
3080
3081
3082
3083
3084
3085
3086
....
3432
3433
3434
3435
3436
3437
3438
3439
3440
3441
3442
3443
3444
3445
3446
....
3971
3972
3973
3974
3975
3976
3977
3978
3979
3980
3981
3982
3983
3984
3985
....
4833
4834
4835
4836
4837
4838
4839
4840
4841
4842
4843
4844
4845
4846
4847
....
5008
5009
5010
5011
5012
5013
5014
5015
5016
5017
5018
5019
5020
5021
5022
....
7379
7380
7381
7382
7383
7384
7385
7386
7387
7388
7389
7390
7391
7392
7393
....
7939
7940
7941
7942
7943
7944
7945
7946
7947
7948
7949
7950
7951
7952
7953
....
8227
8228
8229
8230
8231
8232
8233
8234
8235
8236
8237
8238
8239
8240
8241
8242
8243
8244
8245
8246
8247
8248
8249
8250
  if( pCur->eState!=CURSOR_VALID || pCur->skipNext!=0 ){
    *pHasMoved = 1;
  }else{
    *pHasMoved = 0;
  }
  return SQLITE_OK;
}

/*
** Set up the correct data pointers for a MemPage
*/
static u8 *btreeGetData(MemPage *pPage){
  pPage->aData = sqlite3PagerGetData(pPage->pDbPage);
  pPage->aDataEnd = &pPage->aData[pPage->pBt->usableSize];
  pPage->aCellIdx = &pPage->aData[pPage->cellOffset];
  return pPage->aData;
}

/*
** Make a btree page is writable.
*/
static int btreeMakePageWriteable(MemPage *pPage){
  int rc;

  rc = sqlite3PagerWrite(pPage->pDbPage);
  btreeGetData(pPage);


  return rc;
}

#ifndef SQLITE_OMIT_AUTOVACUUM
/*
** Given a page number of a regular database page, return the page
** number for the pointer-map page that contains the entry for the
................................................................................
  }
  assert( offset <= (int)pBt->usableSize-5 );
  pPtrmap = (u8 *)sqlite3PagerGetData(pDbPage);

  if( eType!=pPtrmap[offset] || get4byte(&pPtrmap[offset+1])!=parent ){
    TRACE(("PTRMAP_UPDATE: %d->(%d,%d)\n", key, eType, parent));
    *pRC= rc = sqlite3PagerWrite(pDbPage);
    pPtrmap = sqlite3PagerGetData(pDbPage);
    if( rc==SQLITE_OK ){
      pPtrmap[offset] = eType;
      put4byte(&pPtrmap[offset+1], parent);
    }
  }

ptrmap_exit:
................................................................................
** call to btreeGetPage.
*/
static void releasePage(MemPage *pPage){
  if( pPage ){
    assert( pPage->aData );
    assert( pPage->pBt );
    assert( sqlite3PagerGetExtra(pPage->pDbPage) == (void*)pPage );
    /* assert( sqlite3PagerGetData(pPage->pDbPage)==pPage->aData ); */
    assert( sqlite3_mutex_held(pPage->pBt->mutex) );
    sqlite3PagerUnref(pPage->pDbPage);
  }
}

/*
** During a rollback, when the pager reloads information into the cache
................................................................................
*/
static int lockBtree(BtShared *pBt){
  int rc;              /* Result code from subfunctions */
  MemPage *pPage1;     /* Page 1 of the database file */
  int nPage;           /* Number of pages in the database */
  int nPageFile = 0;   /* Number of pages in the database file */
  int nPageHeader;     /* Number of pages in the database according to hdr */
  u8 *page1;           /* Content of page 1 */

  assert( sqlite3_mutex_held(pBt->mutex) );
  assert( pBt->pPage1==0 );
  rc = sqlite3PagerSharedLock(pBt->pPager);
  if( rc!=SQLITE_OK ) return rc;
  rc = btreeGetPage(pBt, 1, &pPage1, 0);
  if( rc!=SQLITE_OK ) return rc;

  /* Do some checking to help insure the file we opened really is
  ** a valid database file. 
  */
  page1 = btreeGetData(pPage1);
  nPage = nPageHeader = get4byte(&page1[28]);
  sqlite3PagerPagecount(pBt->pPager, &nPageFile);
  if( nPage==0 || memcmp(24+page1, 92+page1, 4)!=0 ){
    nPage = nPageFile;
  }
  if( nPage>0 ){
    u32 pageSize;
    u32 usableSize;

    rc = SQLITE_NOTADB;
    if( memcmp(page1, zMagicHeader, 16)!=0 ){
      goto page1_init_failed;
    }

#ifdef SQLITE_OMIT_WAL
    if( page1[18]>1 ){
................................................................................
  if( pBt->nPage>0 ){
    return SQLITE_OK;
  }
  pP1 = pBt->pPage1;
  assert( pP1!=0 );
  rc = btreeMakePageWriteable(pP1);
  if( rc ) return rc;
  data = btreeGetData(pP1);
  memcpy(data, zMagicHeader, sizeof(zMagicHeader));
  assert( sizeof(zMagicHeader)==16 );
  data[16] = (u8)((pBt->pageSize>>8)&0xff);
  data[17] = (u8)((pBt->pageSize>>16)&0xff);
  data[18] = 1;
  data[19] = 1;
  assert( pBt->usableSize<=pBt->pageSize && pBt->usableSize+255>=pBt->pageSize);
................................................................................

      /* If the db-size header field is incorrect (as it may be if an old
      ** client has been writing the database file), update it now. Doing
      ** this sooner rather than later means the database size can safely 
      ** re-read the database size from page 1 if a savepoint or transaction
      ** rollback occurs within the transaction.
      */
      btreeGetData(pPage1);
      if( pBt->nPage!=get4byte(&pPage1->aData[28]) ){
        rc = btreeMakePageWriteable(pPage1);
        if( rc==SQLITE_OK ){
          put4byte(&pPage1->aData[28], pBt->nPage);
        }
      }
    }
................................................................................
  assert( sqlite3_mutex_held(pBt->mutex) );
  assert( iLastPg>nFin );

  if( !PTRMAP_ISPAGE(pBt, iLastPg) && iLastPg!=PENDING_BYTE_PAGE(pBt) ){
    u8 eType;
    Pgno iPtrPage;

    btreeGetData(pBt->pPage1);
    nFreeList = get4byte(&pBt->pPage1->aData[36]);
    if( nFreeList==0 ){
      return SQLITE_DONE;
    }

    rc = ptrmapGet(pBt, iLastPg, &eType, &iPtrPage);
    if( rc!=SQLITE_OK ){
................................................................................
      /* It is not possible to create a database for which the final page
      ** is either a pointer-map page or the pending-byte page. If one
      ** is encountered, this indicates corruption.
      */
      return SQLITE_CORRUPT_BKPT;
    }

    btreeGetData(pBt->pPage1);
    nFree = get4byte(&pBt->pPage1->aData[36]);
    nEntry = pBt->usableSize/5;
    nPtrmap = (nFree-nOrig+PTRMAP_PAGENO(pBt, nOrig)+nEntry)/nEntry;
    nFin = nOrig - nFree - nPtrmap;
    if( nOrig>PENDING_BYTE_PAGE(pBt) && nFin<PENDING_BYTE_PAGE(pBt) ){
      nFin--;
    }
................................................................................
    assert( op==SAVEPOINT_RELEASE || op==SAVEPOINT_ROLLBACK );
    assert( iSavepoint>=0 || (iSavepoint==-1 && op==SAVEPOINT_ROLLBACK) );
    sqlite3BtreeEnter(p);
    rc = sqlite3PagerSavepoint(pBt->pPager, op, iSavepoint);
    if( rc==SQLITE_OK ){
      if( iSavepoint<0 && pBt->initiallyEmpty ) pBt->nPage = 0;
      rc = newDatabase(pBt);
      pBt->nPage = get4byte(28 + btreeGetData(pBt->pPage1));

      /* The database size was written into the offset 28 of the header
      ** when the transaction started, so we know that the value at offset
      ** 28 is nonzero. */
      assert( pBt->nPage>0 );
    }
    sqlite3BtreeLeave(p);
................................................................................
        ** output buffer, bypassing the page-cache altogether. This speeds
        ** up loading large records that span many overflow pages.
        */
        if( eOp==0                                             /* (1) */
         && offset==0                                          /* (2) */
         && pBt->inTransaction==TRANS_READ                     /* (4) */
         && (fd = sqlite3PagerFile(pBt->pPager))->pMethods     /* (3) */
         && btreeGetData(pBt->pPage1)[19]==0x01                /* (5) */
        ){
          u8 aSave[4];
          u8 *aWrite = &pBuf[-4];
          memcpy(aSave, aWrite, 4);
          rc = sqlite3OsRead(fd, aWrite, a+4, pBt->pageSize * (nextPage-1));
          nextPage = get4byte(aWrite);
          memcpy(aWrite, aSave, 4);
................................................................................
  MemPage *pTrunk = 0;
  MemPage *pPrevTrunk = 0;
  Pgno mxPage;     /* Total size of the database file */

  assert( sqlite3_mutex_held(pBt->mutex) );
  pPage1 = pBt->pPage1;
  mxPage = btreePagecount(pBt);
  n = get4byte(&btreeGetData(pPage1)[36]);
  testcase( n==mxPage-1 );
  if( n>=mxPage ){
    return SQLITE_CORRUPT_BKPT;
  }
  if( n>0 ){
    /* There are pages on the freelist.  Reuse one of those pages. */
    Pgno iTrunk;
................................................................................
          int noContent;
          *pPgno = iPage;
          TRACE(("ALLOCATE: %d was leaf %d of %d on trunk %d"
                 ": %d more free pages\n",
                 *pPgno, closest+1, k, pTrunk->pgno, n-1));
          rc = btreeMakePageWriteable(pTrunk);
          if( rc ) goto end_allocate_page;
          aData = pTrunk->aData;
          if( closest<k-1 ){
            memcpy(&aData[8+closest*4], &aData[4+k*4], 4);
          }
          put4byte(&aData[4], k-1);
          noContent = !btreeGetHasContent(pBt, *pPgno);
          rc = btreeGetPage(pBt, *pPgno, ppPage, noContent);
          if( rc==SQLITE_OK ){
................................................................................

  sqlite3BtreeEnter(p);
  assert( p->inTrans>TRANS_NONE );
  assert( SQLITE_OK==querySharedCacheTableLock(p, MASTER_ROOT, READ_LOCK) );
  assert( pBt->pPage1 );
  assert( idx>=0 && idx<=15 );

  *pMeta = get4byte(&btreeGetData(pBt->pPage1)[36 + idx*4]);

  /* If auto-vacuum is disabled in this build and this is an auto-vacuum
  ** database, mark the database as read-only.  */
#ifdef SQLITE_OMIT_AUTOVACUUM
  if( idx==BTREE_LARGEST_ROOT_PAGE && *pMeta>0 ) pBt->readOnly = 1;
#endif

................................................................................
    sCheck.anRef[i] = 1;
  }
  sqlite3StrAccumInit(&sCheck.errMsg, zErr, sizeof(zErr), 20000);
  sCheck.errMsg.useMalloc = 2;

  /* Check the integrity of the freelist
  */
  btreeGetData(pBt->pPage1);
  checkList(&sCheck, 1, get4byte(&pBt->pPage1->aData[32]),
            get4byte(&pBt->pPage1->aData[36]), "Main freelist: ");

  /* Check all the tables.
  */
  for(i=0; (int)i<nRoot && sCheck.mxErr; i++){
    if( aRoot[i]==0 ) continue;
................................................................................
  /* If setting the version fields to 1, do not automatically open the
  ** WAL connection, even if the version fields are currently set to 2.
  */
  pBt->doNotUseWAL = (u8)(iVersion==1);

  rc = sqlite3BtreeBeginTrans(pBtree, 0);
  if( rc==SQLITE_OK ){
    u8 *aData = btreeGetData(pBt->pPage1);
    if( aData[18]!=(u8)iVersion || aData[19]!=(u8)iVersion ){
      rc = sqlite3BtreeBeginTrans(pBtree, 2);
      if( rc==SQLITE_OK ){
        rc = btreeMakePageWriteable(pBt->pPage1);
        if( rc==SQLITE_OK ){
          aData = btreeGetData(pBt->pPage1);
          aData[18] = (u8)iVersion;
          aData[19] = (u8)iVersion;
        }
      }
    }
  }

  pBt->doNotUseWAL = 0;
  return rc;
}

Changes to src/pager.c.

1722
1723
1724
1725
1726
1727
1728

1729
1730
1731
1732
1733
1734
1735
....
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803













1804
1805
1806
1807
1808
1809
1810
....
5286
5287
5288
5289
5290
5291
5292










5293
5294
5295
5296
5297
5298
5299
....
5336
5337
5338
5339
5340
5341
5342
5343
5344
5345
5346
5347
5348
5349
5350
5351
5352
5353
....
6433
6434
6435
6436
6437
6438
6439

6440
6441
6442
6443
6444
6445
6446
....
6482
6483
6484
6485
6486
6487
6488
6489
6490
6491
6492
6493
6494
6495
6496
6497
6498
** contents of the pager cache are discarded before switching back to 
** the OPEN state. Regardless of whether the pager is in exclusive-mode
** or not, any journal file left in the file-system will be treated
** as a hot-journal and rolled back the next time a read-transaction
** is opened (by this or by any other connection).
*/
static void pager_unlock(Pager *pPager){


  assert( pPager->eState==PAGER_READER 
       || pPager->eState==PAGER_OPEN 
       || pPager->eState==PAGER_ERROR 
  );

  sqlite3BitvecDestroy(pPager->pInJournal);
................................................................................
  }

  /* If Pager.errCode is set, the contents of the pager cache cannot be
  ** trusted. Now that there are no outstanding references to the pager,
  ** it can safely move back to PAGER_OPEN state. This happens in both
  ** normal and exclusive-locking mode.
  */
  if( pPager->errCode || pPager->pMapObject ){
    assert( !MEMDB );
    pager_reset(pPager);
    pPager->changeCountDone = pPager->tempFile;
    pPager->eState = PAGER_OPEN;
    pPager->errCode = SQLITE_OK;
    sqlite3OsUnmap(pPager->fd, pPager->pMapObject);
    pPager->pMapObject = 0;
    pPager->aFileContent = 0;
    pPager->nFileContent = 0;
  }

  pPager->journalOff = 0;
  pPager->journalHdr = 0;
  pPager->setMaster = 0;














}

/*
** This function is called whenever an IOERR or FULL error that requires
** the pager to transition into the ERROR state may ahve occurred.
** The first argument is a pointer to the pager structure, the second 
** the error-code about to be returned by a pager API function. The 
................................................................................
    assert( rc!=SQLITE_OK || pPager->eState==PAGER_WRITER_LOCKED );
    assert( assert_pager_state(pPager) );
  }

  PAGERTRACE(("TRANSACTION %d\n", PAGERID(pPager)));
  return rc;
}











/*
** Mark a single data page as writeable. The page is written into the 
** main journal or sub-journal as required. If the page is written into
** one of the journals, the corresponding bit is set in the 
** Pager.pInJournal bitvec and the PagerSavepoint.pInSavepoint bitvecs
** of any open savepoints as appropriate.
................................................................................
    rc = pager_open_journal(pPager);
    if( rc!=SQLITE_OK ) return rc;
  }
  assert( pPager->eState>=PAGER_WRITER_CACHEMOD );
  assert( assert_pager_state(pPager) );

  /* Make sure page content is held in malloced memory */
  if( pPg->pData!=pPg->pBuf ){
    memcpy(pPg->pBuf, pPg->pData, pPager->pageSize);
    pData = pPg->pData = pPg->pBuf;
  }

  /* Mark the page as dirty.  If the page has already been written
  ** to the journal then we can return right away.
  */
  sqlite3PcacheMakeDirty(pPg);
  if( pageInJournal(pPg) && !subjRequiresPage(pPg) ){
    assert( !pagerUseWal(pPager) );
................................................................................
      sqlite3PcacheMove(pPgOld, pPager->dbSize+1);
    }else{
      sqlite3PcacheDrop(pPgOld);
    }
  }

  origPgno = pPg->pgno;

  sqlite3PcacheMove(pPg, pgno);
  sqlite3PcacheMakeDirty(pPg);

  /* For an in-memory database, make sure the original page continues
  ** to exist, in case the transaction needs to roll back.  Use pPgOld
  ** as the original page since it has already been allocated.
  */
................................................................................
  return SQLITE_OK;
}
#endif

/*
** Return a pointer to the data for the specified page.
*/
void *sqlite3PagerGetData(DbPage *pPg){
  assert( pPg->nRef>0 || pPg->pPager->memDb );
  return pPg->pData;
}

/*
** Return a pointer to the Pager.nExtra bytes of "extra" space 
** allocated along with the specified page.
*/
void *sqlite3PagerGetExtra(DbPage *pPg){







>







 







|





<
<
<
<






>
>
>
>
>
>
>
>
>
>
>
>
>







 







>
>
>
>
>
>
>
>
>
>







 







|
<
|
<







 







>







 







|

|







1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
....
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794




1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
....
5296
5297
5298
5299
5300
5301
5302
5303
5304
5305
5306
5307
5308
5309
5310
5311
5312
5313
5314
5315
5316
5317
5318
5319
....
5356
5357
5358
5359
5360
5361
5362
5363

5364

5365
5366
5367
5368
5369
5370
5371
....
6451
6452
6453
6454
6455
6456
6457
6458
6459
6460
6461
6462
6463
6464
6465
....
6501
6502
6503
6504
6505
6506
6507
6508
6509
6510
6511
6512
6513
6514
6515
6516
6517
** contents of the pager cache are discarded before switching back to 
** the OPEN state. Regardless of whether the pager is in exclusive-mode
** or not, any journal file left in the file-system will be treated
** as a hot-journal and rolled back the next time a read-transaction
** is opened (by this or by any other connection).
*/
static void pager_unlock(Pager *pPager){
  PgHdr *pPg;

  assert( pPager->eState==PAGER_READER 
       || pPager->eState==PAGER_OPEN 
       || pPager->eState==PAGER_ERROR 
  );

  sqlite3BitvecDestroy(pPager->pInJournal);
................................................................................
  }

  /* If Pager.errCode is set, the contents of the pager cache cannot be
  ** trusted. Now that there are no outstanding references to the pager,
  ** it can safely move back to PAGER_OPEN state. This happens in both
  ** normal and exclusive-locking mode.
  */
  if( pPager->errCode ){
    assert( !MEMDB );
    pager_reset(pPager);
    pPager->changeCountDone = pPager->tempFile;
    pPager->eState = PAGER_OPEN;
    pPager->errCode = SQLITE_OK;




  }

  pPager->journalOff = 0;
  pPager->journalHdr = 0;
  pPager->setMaster = 0;

  pPg = 0;
  sqlite3PcacheFetch(pPager->pPCache, 1, 0, &pPg);
  if( pPg ){
    /* assert( sqlite3PcachePagecount(pPager->pPCache)==1 ); */
    pPg->pData = pPg->pBuf;
    sqlite3PcacheRelease(pPg);
  }else{
    /*assert( sqlite3PcachePagecount(pPager->pPCache)==0 );*/
  }
  sqlite3OsUnmap(pPager->fd, pPager->pMapObject);
  pPager->pMapObject = 0;
  pPager->aFileContent = 0;
  pPager->nFileContent = 0;
}

/*
** This function is called whenever an IOERR or FULL error that requires
** the pager to transition into the ERROR state may ahve occurred.
** The first argument is a pointer to the pager structure, the second 
** the error-code about to be returned by a pager API function. The 
................................................................................
    assert( rc!=SQLITE_OK || pPager->eState==PAGER_WRITER_LOCKED );
    assert( assert_pager_state(pPager) );
  }

  PAGERTRACE(("TRANSACTION %d\n", PAGERID(pPager)));
  return rc;
}

/*
** Make a copy of page content into malloced space.
*/
void makePageWriteable(Pager *pPager, PgHdr *pPg){
  if( pPg->pData!=pPg->pBuf ){
    memcpy(pPg->pBuf, pPg->pData, pPager->pageSize);
    pPg->pData = pPg->pBuf;
  }
}

/*
** Mark a single data page as writeable. The page is written into the 
** main journal or sub-journal as required. If the page is written into
** one of the journals, the corresponding bit is set in the 
** Pager.pInJournal bitvec and the PagerSavepoint.pInSavepoint bitvecs
** of any open savepoints as appropriate.
................................................................................
    rc = pager_open_journal(pPager);
    if( rc!=SQLITE_OK ) return rc;
  }
  assert( pPager->eState>=PAGER_WRITER_CACHEMOD );
  assert( assert_pager_state(pPager) );

  /* Make sure page content is held in malloced memory */
  makePageWriteable(pPager, pPg);

  pData = pPg->pData;


  /* Mark the page as dirty.  If the page has already been written
  ** to the journal then we can return right away.
  */
  sqlite3PcacheMakeDirty(pPg);
  if( pageInJournal(pPg) && !subjRequiresPage(pPg) ){
    assert( !pagerUseWal(pPager) );
................................................................................
      sqlite3PcacheMove(pPgOld, pPager->dbSize+1);
    }else{
      sqlite3PcacheDrop(pPgOld);
    }
  }

  origPgno = pPg->pgno;
  makePageWriteable(pPager, pPg);
  sqlite3PcacheMove(pPg, pgno);
  sqlite3PcacheMakeDirty(pPg);

  /* For an in-memory database, make sure the original page continues
  ** to exist, in case the transaction needs to roll back.  Use pPgOld
  ** as the original page since it has already been allocated.
  */
................................................................................
  return SQLITE_OK;
}
#endif

/*
** Return a pointer to the data for the specified page.
*/
u8 *sqlite3PagerGetData(DbPage *pPg){
  assert( pPg->nRef>0 || pPg->pPager->memDb );
  return (u8*)pPg->pData;
}

/*
** Return a pointer to the Pager.nExtra bytes of "extra" space 
** allocated along with the specified page.
*/
void *sqlite3PagerGetExtra(DbPage *pPg){

Changes to src/pager.h.

119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
void sqlite3PagerUnref(DbPage*);

/* Operations on page references. */
int sqlite3PagerWrite(DbPage*);
void sqlite3PagerDontWrite(DbPage*);
int sqlite3PagerMovepage(Pager*,DbPage*,Pgno,int);
int sqlite3PagerPageRefcount(DbPage*);
void *sqlite3PagerGetData(DbPage *); 
void *sqlite3PagerGetExtra(DbPage *); 

/* Functions used to manage pager transactions and savepoints. */
void sqlite3PagerPagecount(Pager*, int*);
int sqlite3PagerBegin(Pager*, int exFlag, int);
int sqlite3PagerCommitPhaseOne(Pager*,const char *zMaster, int);
int sqlite3PagerExclusiveLock(Pager*);







|







119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
void sqlite3PagerUnref(DbPage*);

/* Operations on page references. */
int sqlite3PagerWrite(DbPage*);
void sqlite3PagerDontWrite(DbPage*);
int sqlite3PagerMovepage(Pager*,DbPage*,Pgno,int);
int sqlite3PagerPageRefcount(DbPage*);
u8 *sqlite3PagerGetData(DbPage *); 
void *sqlite3PagerGetExtra(DbPage *); 

/* Functions used to manage pager transactions and savepoints. */
void sqlite3PagerPagecount(Pager*, int*);
int sqlite3PagerBegin(Pager*, int exFlag, int);
int sqlite3PagerCommitPhaseOne(Pager*,const char *zMaster, int);
int sqlite3PagerExclusiveLock(Pager*);

Changes to src/pcache.c.

127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
*/
static void pcacheUnpin(PgHdr *p){
  PCache *pCache = p->pCache;
  if( pCache->bPurgeable ){
    if( p->pgno==1 ){
      pCache->pPage1 = 0;
    }
    sqlite3GlobalConfig.pcache2.xUnpin(pCache->pCache, p->pPage, 0);
  }
}

/*************************************************** General Interfaces ******
**
** Initialize and shutdown the page cache subsystem. Neither of these 
** functions are threadsafe.







|







127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
*/
static void pcacheUnpin(PgHdr *p){
  PCache *pCache = p->pCache;
  if( pCache->bPurgeable ){
    if( p->pgno==1 ){
      pCache->pPage1 = 0;
    }
    sqlite3GlobalConfig.pcache2.xUnpin(pCache->pCache, p->pPage, 1);
  }
}

/*************************************************** General Interfaces ******
**
** Initialize and shutdown the page cache subsystem. Neither of these 
** functions are threadsafe.