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

Overview
Comment:Fix problems with redirected blocks in compressed databases.
Downloads: Tarball | ZIP archive
Timelines: family | ancestors | descendants | both | block-redirects
Files: files | file ages | folders
SHA1: 930b7e4507f1c6da6dca7b7edabd8180fe280da0
User & Date: dan 2013-01-26 19:17:51.756
Context
2013-01-26
20:18
Handle calls to lsm_work() with the nPage parameter set to not greater than zero. Remove some dead code. Leaf check-in: 520f3729b8 user: dan tags: block-redirects
19:17
Fix problems with redirected blocks in compressed databases. check-in: 930b7e4507 user: dan tags: block-redirects
17:24
Remove an inaccurate assert() from lsm_sorted.c. check-in: 259878e357 user: dan tags: block-redirects
Changes
Unified Diff Ignore Whitespace Patch
Changes to lsm-test/lsmtest_tdb3.c.
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
int test_lsm_zip_open(
  const char *zFilename, 
  int bClear, 
  TestDb **ppDb
){
  const char *zCfg = 
    "page_size=256 block_size=65536 autoflush=16384 "
    "max_freelist=2 autocheckpoint=32768 compression=1"
    "mmap=0 "
  ;
  return testLsmOpen(zCfg, zFilename, bClear, ppDb);
}

lsm_db *tdb_lsm(TestDb *pDb){
  if( pDb->pMethods->xClose==test_lsm_close ){
    return ((LsmDb *)pDb)->db;







|
<







1000
1001
1002
1003
1004
1005
1006
1007

1008
1009
1010
1011
1012
1013
1014
int test_lsm_zip_open(
  const char *zFilename, 
  int bClear, 
  TestDb **ppDb
){
  const char *zCfg = 
    "page_size=256 block_size=65536 autoflush=16384 "
    "autocheckpoint=32768 compression=1 mmap=0 "

  ;
  return testLsmOpen(zCfg, zFilename, bClear, ppDb);
}

lsm_db *tdb_lsm(TestDb *pDb){
  if( pDb->pMethods->xClose==test_lsm_close ){
    return ((LsmDb *)pDb)->db;
Changes to src/lsm_file.c.
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890

891
892
893
894
895
896
897
898


899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914

915
916
917
918
919
920
921
    i64 nMin = (i64)nBlock * (i64)pFS->nBlocksize;
    fsGrowMapping(pFS, nMin, &rc);
    if( rc!=LSM_OK ) return rc;
  }
  return lsmEnvSync(pFS->pEnv, pFS->fdDb);
}

static int fsPageGet(FileSystem *, Redirect *, Pgno, int, Page **, int *);

static int fsRedirectBlock(Redirect *p, int iBlk){
  if( p ){
    int i;
    for(i=0; i<p->n; i++){
      if( iBlk==p->a[i].iFrom ) return p->a[i].iTo;
    }
  }

  return iBlk;
}

Pgno lsmFsRedirectPage(FileSystem *pFS, Redirect *pRedir, Pgno iPg){
  Pgno iReal = iPg;

  if( pRedir ){
    const int nPagePerBlock = (pFS->nBlocksize / pFS->nPagesize);


    int iBlk = fsPageToBlock(pFS, iPg);
    int i;
    for(i=0; i<pRedir->n; i++){
      int iFrom = pRedir->a[i].iFrom;
      if( iFrom>iBlk ) break;
      if( iFrom==iBlk ){
        int iTo = pRedir->a[i].iTo;
        iReal = iPg - (Pgno)(iFrom - iTo) * nPagePerBlock;
        if( iTo==1 ){
          iReal += (fsFirstPageOnBlock(pFS, 1)-1);
        }
        break;
      }
    }
  }


  return iReal;
}

/*
** Parameter iBlock is a database file block. This function reads the value 
** stored in the blocks "next block" pointer and stores it in *piNext.
** LSM_OK is returned if everything is successful, or an LSM error code







|








>







|
>
>
















>







875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
    i64 nMin = (i64)nBlock * (i64)pFS->nBlocksize;
    fsGrowMapping(pFS, nMin, &rc);
    if( rc!=LSM_OK ) return rc;
  }
  return lsmEnvSync(pFS->pEnv, pFS->fdDb);
}

static int fsPageGet(FileSystem *, Segment *, Pgno, int, Page **, int *);

static int fsRedirectBlock(Redirect *p, int iBlk){
  if( p ){
    int i;
    for(i=0; i<p->n; i++){
      if( iBlk==p->a[i].iFrom ) return p->a[i].iTo;
    }
  }
  assert( iBlk!=0 );
  return iBlk;
}

Pgno lsmFsRedirectPage(FileSystem *pFS, Redirect *pRedir, Pgno iPg){
  Pgno iReal = iPg;

  if( pRedir ){
    const int nPagePerBlock = (
        pFS->pCompress ? pFS->nBlocksize : (pFS->nBlocksize / pFS->nPagesize)
    );
    int iBlk = fsPageToBlock(pFS, iPg);
    int i;
    for(i=0; i<pRedir->n; i++){
      int iFrom = pRedir->a[i].iFrom;
      if( iFrom>iBlk ) break;
      if( iFrom==iBlk ){
        int iTo = pRedir->a[i].iTo;
        iReal = iPg - (Pgno)(iFrom - iTo) * nPagePerBlock;
        if( iTo==1 ){
          iReal += (fsFirstPageOnBlock(pFS, 1)-1);
        }
        break;
      }
    }
  }

  assert( iReal!=0 );
  return iReal;
}

/*
** Parameter iBlock is a database file block. This function reads the value 
** stored in the blocks "next block" pointer and stores it in *piNext.
** LSM_OK is returned if everything is successful, or an LSM error code
970
971
972
973
974
975
976

977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013

1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026

1027
1028
1029
1030
1031
1032
1033
1034
}

/*
** This function is only called in compressed database mode.
*/
static int fsReadData(
  FileSystem *pFS,                /* File-system handle */

  i64 iOff,                       /* Read data from this offset */
  u8 *aData,                      /* Buffer to read data into */
  int nData                       /* Number of bytes to read */
){
  i64 iEob;                       /* End of block */
  int nRead;
  int rc;

  assert( pFS->pCompress );

  iEob = fsLastPageOnPagesBlock(pFS, iOff) + 1;
  nRead = LSM_MIN(iEob - iOff, nData);

  rc = lsmEnvRead(pFS->pEnv, pFS->fdDb, iOff, aData, nRead);
  if( rc==LSM_OK && nRead!=nData ){
    int iBlk;

    /* TODO: Fix the pSeg arg to this */
    assert( 0 );
    rc = fsBlockNext(pFS, 0, fsPageToBlock(pFS, iOff), &iBlk);
    if( rc==LSM_OK ){
      i64 iOff2 = fsFirstPageOnBlock(pFS, iBlk);
      rc = lsmEnvRead(pFS->pEnv, pFS->fdDb, iOff2, &aData[nRead], nData-nRead);
    }
  }

  return rc;
}

/*
** Parameter iBlock is a database file block. This function reads the value 
** stored in the blocks "previous block" pointer and stores it in *piPrev.
** LSM_OK is returned if everything is successful, or an LSM error code
** otherwise.
*/
static int fsBlockPrev(
  FileSystem *pFS,                /* File-system object handle */

  int iBlock,                     /* Read field from this block */
  int *piPrev                     /* OUT: Previous block in linked list */
){
  int rc = LSM_OK;                /* Return code */

  assert( pFS->bUseMmap==0 || pFS->pCompress==0 );
  assert( iBlock>0 );

  if( pFS->pCompress ){
    i64 iOff = fsFirstPageOnBlock(pFS, iBlock) - 4;
    u8 aPrev[4];                  /* 4-byte pointer read from db file */
    rc = lsmEnvRead(pFS->pEnv, pFS->fdDb, iOff, aPrev, sizeof(aPrev));
    if( rc==LSM_OK ){

      *piPrev = (int)lsmGetU32(aPrev);
    }
  }else{
    assert( 0 );
  }
  return rc;
}








>

















<
<
|

















>













>
|







974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998


999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
}

/*
** This function is only called in compressed database mode.
*/
static int fsReadData(
  FileSystem *pFS,                /* File-system handle */
  Segment *pSeg,                  /* Block redirection */
  i64 iOff,                       /* Read data from this offset */
  u8 *aData,                      /* Buffer to read data into */
  int nData                       /* Number of bytes to read */
){
  i64 iEob;                       /* End of block */
  int nRead;
  int rc;

  assert( pFS->pCompress );

  iEob = fsLastPageOnPagesBlock(pFS, iOff) + 1;
  nRead = LSM_MIN(iEob - iOff, nData);

  rc = lsmEnvRead(pFS->pEnv, pFS->fdDb, iOff, aData, nRead);
  if( rc==LSM_OK && nRead!=nData ){
    int iBlk;



    rc = fsBlockNext(pFS, pSeg, fsPageToBlock(pFS, iOff), &iBlk);
    if( rc==LSM_OK ){
      i64 iOff2 = fsFirstPageOnBlock(pFS, iBlk);
      rc = lsmEnvRead(pFS->pEnv, pFS->fdDb, iOff2, &aData[nRead], nData-nRead);
    }
  }

  return rc;
}

/*
** Parameter iBlock is a database file block. This function reads the value 
** stored in the blocks "previous block" pointer and stores it in *piPrev.
** LSM_OK is returned if everything is successful, or an LSM error code
** otherwise.
*/
static int fsBlockPrev(
  FileSystem *pFS,                /* File-system object handle */
  Segment *pSeg,                  /* Use this segment for block redirects */
  int iBlock,                     /* Read field from this block */
  int *piPrev                     /* OUT: Previous block in linked list */
){
  int rc = LSM_OK;                /* Return code */

  assert( pFS->bUseMmap==0 || pFS->pCompress==0 );
  assert( iBlock>0 );

  if( pFS->pCompress ){
    i64 iOff = fsFirstPageOnBlock(pFS, iBlock) - 4;
    u8 aPrev[4];                  /* 4-byte pointer read from db file */
    rc = lsmEnvRead(pFS->pEnv, pFS->fdDb, iOff, aPrev, sizeof(aPrev));
    if( rc==LSM_OK ){
      Redirect *pRedir = (pSeg ? pSeg->pRedirect : 0);
      *piPrev = fsRedirectBlock(pRedir, (int)lsmGetU32(aPrev));
    }
  }else{
    assert( 0 );
  }
  return rc;
}

1045
1046
1047
1048
1049
1050
1051
1052






1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070






1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
  nByte  = (aBuf[0] & 0x7F) << 14;
  nByte += (aBuf[1] & 0x7F) << 7;
  nByte += (aBuf[2] & 0x7F);
  *pbFree = !(aBuf[1] & 0x80);
  return nByte;
}

static int fsSubtractOffset(FileSystem *pFS, i64 iOff, int iSub, i64 *piRes){






  i64 iStart;
  int iBlk = 0;
  int rc;

  assert( pFS->pCompress );

  iStart = fsFirstPageOnBlock(pFS, fsPageToBlock(pFS, iOff));
  if( (iOff-iSub)>=iStart ){
    *piRes = (iOff-iSub);
    return LSM_OK;
  }

  rc = fsBlockPrev(pFS, fsPageToBlock(pFS, iOff), &iBlk);
  *piRes = fsLastPageOnBlock(pFS, iBlk) - iSub + (iOff - iStart + 1);
  return rc;
}

static int fsAddOffset(FileSystem *pFS, i64 iOff, int iAdd, i64 *piRes){






  i64 iEob;
  int iBlk;
  int rc;

  assert( pFS->pCompress );

  iEob = fsLastPageOnPagesBlock(pFS, iOff);
  if( (iOff+iAdd)<=iEob ){
    *piRes = (iOff+iAdd);
    return LSM_OK;
  }

  assert( 0 );
  rc = fsBlockNext(pFS, 0, fsPageToBlock(pFS, iOff), &iBlk);
  *piRes = fsFirstPageOnBlock(pFS, iBlk) + iAdd - (iEob - iOff + 1);
  return rc;
}

static int fsAllocateBuffer(FileSystem *pFS, int bWrite){
  u8 **pp;                        /* Pointer to either aIBuffer or aOBuffer */








|
>
>
>
>
>
>












|




|
>
>
>
>
>
>












<
|







1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099

1100
1101
1102
1103
1104
1105
1106
1107
  nByte  = (aBuf[0] & 0x7F) << 14;
  nByte += (aBuf[1] & 0x7F) << 7;
  nByte += (aBuf[2] & 0x7F);
  *pbFree = !(aBuf[1] & 0x80);
  return nByte;
}

static int fsSubtractOffset(
  FileSystem *pFS, 
  Segment *pSeg,
  i64 iOff, 
  int iSub, 
  i64 *piRes
){
  i64 iStart;
  int iBlk = 0;
  int rc;

  assert( pFS->pCompress );

  iStart = fsFirstPageOnBlock(pFS, fsPageToBlock(pFS, iOff));
  if( (iOff-iSub)>=iStart ){
    *piRes = (iOff-iSub);
    return LSM_OK;
  }

  rc = fsBlockPrev(pFS, pSeg, fsPageToBlock(pFS, iOff), &iBlk);
  *piRes = fsLastPageOnBlock(pFS, iBlk) - iSub + (iOff - iStart + 1);
  return rc;
}

static int fsAddOffset(
  FileSystem *pFS, 
  Segment *pSeg,
  i64 iOff, 
  int iAdd, 
  i64 *piRes
){
  i64 iEob;
  int iBlk;
  int rc;

  assert( pFS->pCompress );

  iEob = fsLastPageOnPagesBlock(pFS, iOff);
  if( (iOff+iAdd)<=iEob ){
    *piRes = (iOff+iAdd);
    return LSM_OK;
  }


  rc = fsBlockNext(pFS, pSeg, fsPageToBlock(pFS, iOff), &iBlk);
  *piRes = fsFirstPageOnBlock(pFS, iBlk) + iAdd - (iEob - iOff + 1);
  return rc;
}

static int fsAllocateBuffer(FileSystem *pFS, int bWrite){
  u8 **pp;                        /* Pointer to either aIBuffer or aOBuffer */

1115
1116
1117
1118
1119
1120
1121

1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
** uncompresses the compressed data for page pPg from the database and
** populates the pPg->aData[] buffer and pPg->nCompress field.
**
** LSM_OK is returned if successful, or an LSM error code otherwise.
*/
static int fsReadPagedata(
  FileSystem *pFS,                /* File-system handle */

  Page *pPg,                      /* Page to read and uncompress data for */
  int *pnSpace                    /* OUT: Total bytes of free space */
){
  lsm_compress *p = pFS->pCompress;
  i64 iOff = pPg->iPg;
  u8 aSz[3];
  int rc;

  assert( p && pPg->nCompress==0 );

  if( fsAllocateBuffer(pFS, 0) ) return LSM_NOMEM;

  rc = fsReadData(pFS, iOff, aSz, sizeof(aSz));

  if( rc==LSM_OK ){
    int bFree;
    if( aSz[0] & 0x80 ){
      pPg->nCompress = (int)getRecordSize(aSz, &bFree);
    }else{
      pPg->nCompress = (int)aSz[0] - sizeof(aSz)*2;
      bFree = 1;
    }
    if( bFree ){
      if( pnSpace ){
        *pnSpace = pPg->nCompress + sizeof(aSz)*2;
      }else{
        rc = LSM_CORRUPT_BKPT;
      }
    }else{
      rc = fsAddOffset(pFS, iOff, 3, &iOff);
      if( rc==LSM_OK ){
        if( pPg->nCompress>pFS->nBuffer ){
          rc = LSM_CORRUPT_BKPT;
        }else{
          rc = fsReadData(pFS, iOff, pFS->aIBuffer, pPg->nCompress);
        }
        if( rc==LSM_OK ){
          int n = pFS->nPagesize;
          rc = p->xUncompress(p->pCtx, 
              (char *)pPg->aData, &n, 
              (const char *)pFS->aIBuffer, pPg->nCompress
          );







>












|
















|




|







1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
** uncompresses the compressed data for page pPg from the database and
** populates the pPg->aData[] buffer and pPg->nCompress field.
**
** LSM_OK is returned if successful, or an LSM error code otherwise.
*/
static int fsReadPagedata(
  FileSystem *pFS,                /* File-system handle */
  Segment *pSeg,                  /* pPg is part of this segment */
  Page *pPg,                      /* Page to read and uncompress data for */
  int *pnSpace                    /* OUT: Total bytes of free space */
){
  lsm_compress *p = pFS->pCompress;
  i64 iOff = pPg->iPg;
  u8 aSz[3];
  int rc;

  assert( p && pPg->nCompress==0 );

  if( fsAllocateBuffer(pFS, 0) ) return LSM_NOMEM;

  rc = fsReadData(pFS, pSeg, iOff, aSz, sizeof(aSz));

  if( rc==LSM_OK ){
    int bFree;
    if( aSz[0] & 0x80 ){
      pPg->nCompress = (int)getRecordSize(aSz, &bFree);
    }else{
      pPg->nCompress = (int)aSz[0] - sizeof(aSz)*2;
      bFree = 1;
    }
    if( bFree ){
      if( pnSpace ){
        *pnSpace = pPg->nCompress + sizeof(aSz)*2;
      }else{
        rc = LSM_CORRUPT_BKPT;
      }
    }else{
      rc = fsAddOffset(pFS, pSeg, iOff, 3, &iOff);
      if( rc==LSM_OK ){
        if( pPg->nCompress>pFS->nBuffer ){
          rc = LSM_CORRUPT_BKPT;
        }else{
          rc = fsReadData(pFS, pSeg, iOff, pFS->aIBuffer, pPg->nCompress);
        }
        if( rc==LSM_OK ){
          int n = pFS->nPagesize;
          rc = p->xUncompress(p->pCtx, 
              (char *)pPg->aData, &n, 
              (const char *)pFS->aIBuffer, pPg->nCompress
          );
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
}

/*
** Return a handle for a database page.
*/
static int fsPageGet(
  FileSystem *pFS,                /* File-system handle */
  Redirect *pRedirect,            /* Block redirection to use (or NULL) */
  Pgno iPg,                       /* Page id */
  int noContent,                  /* True to not load content from disk */
  Page **ppPg,                    /* OUT: New page handle */
  int *pnSpace                    /* OUT: Bytes of free space */
){
  Page *p;
  int iHash;
  int rc = LSM_OK;

  /* In most cases iReal is the same as iPg. Except, if pRedirect is not
  ** NULL, and the block containing iPg has been redirected, then iReal
  ** is the page number after redirection.  */
  Pgno iReal = lsmFsRedirectPage(pFS, pRedirect, iPg);

  assert( iPg>=fsFirstPageOnBlock(pFS, 1) );
  assert( iReal>=fsFirstPageOnBlock(pFS, 1) );
  *ppPg = 0;

  assert( pFS->bUseMmap==0 || pFS->pCompress==0 );
  if( pFS->bUseMmap ){







|









|
|

|







1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
}

/*
** Return a handle for a database page.
*/
static int fsPageGet(
  FileSystem *pFS,                /* File-system handle */
  Segment *pSeg,                  /* Block redirection to use (or NULL) */
  Pgno iPg,                       /* Page id */
  int noContent,                  /* True to not load content from disk */
  Page **ppPg,                    /* OUT: New page handle */
  int *pnSpace                    /* OUT: Bytes of free space */
){
  Page *p;
  int iHash;
  int rc = LSM_OK;

  /* In most cases iReal is the same as iPg. Except, if pSeg->pRedirect is 
  ** not NULL, and the block containing iPg has been redirected, then iReal
  ** is the page number after redirection.  */
  Pgno iReal = lsmFsRedirectPage(pFS, (pSeg ? pSeg->pRedirect : 0), iPg);

  assert( iPg>=fsFirstPageOnBlock(pFS, 1) );
  assert( iReal>=fsFirstPageOnBlock(pFS, 1) );
  *ppPg = 0;

  assert( pFS->bUseMmap==0 || pFS->pCompress==0 );
  if( pFS->bUseMmap ){
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257

#ifdef LSM_DEBUG
        memset(p->aData, 0x56, pFS->nPagesize);
#endif
        assert( p->pLruNext==0 && p->pLruPrev==0 );
        if( noContent==0 ){
          if( pFS->pCompress ){
            rc = fsReadPagedata(pFS, p, &nSpace);
          }else{
            int nByte = pFS->nPagesize;
            i64 iOff = (i64)(iReal-1) * pFS->nPagesize;
            rc = lsmEnvRead(pFS->pEnv, pFS->fdDb, iOff, p->aData, nByte);
          }
          pFS->nRead++;
        }







|







1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274

#ifdef LSM_DEBUG
        memset(p->aData, 0x56, pFS->nPagesize);
#endif
        assert( p->pLruNext==0 && p->pLruPrev==0 );
        if( noContent==0 ){
          if( pFS->pCompress ){
            rc = fsReadPagedata(pFS, pSeg, p, &nSpace);
          }else{
            int nByte = pFS->nPagesize;
            i64 iOff = (i64)(iReal-1) * pFS->nPagesize;
            rc = lsmEnvRead(pFS->pEnv, pFS->fdDb, iOff, p->aData, nByte);
          }
          pFS->nRead++;
        }
1511
1512
1513
1514
1515
1516
1517














1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541





1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
    iBlk = iNext;
  }

  pRun->nSize -= (pRun->iFirst - fsFirstPageOnBlock(pFS, iBlk));
  assert( pRun->nSize>0 );
}















static int fsNextPageOffset(
  FileSystem *pFS,                /* File system object */
  Segment *pSeg,                  /* Segment to move within */
  Pgno iPg,                       /* Offset of current page */
  int nByte,                      /* Size of current page including headers */
  Pgno *piNext                    /* OUT: Offset of next page. Or zero (EOF) */
){
  Pgno iNext;
  int rc;

  assert( pFS->pCompress );

  rc = fsAddOffset(pFS, iPg, nByte-1, &iNext);
  if( pSeg && iNext==pSeg->iLastPg ){
    iNext = 0;
  }else if( rc==LSM_OK ){
    rc = fsAddOffset(pFS, iNext, 1, &iNext);
  }

  *piNext = iNext;
  return rc;
}

static int fsGetPageBefore(FileSystem *pFS, i64 iOff, Pgno *piPrev){





  u8 aSz[3];
  int rc;
  i64 iRead;

  rc = fsSubtractOffset(pFS, iOff, sizeof(aSz), &iRead);
  if( rc==LSM_OK ) rc = fsReadData(pFS, iRead, aSz, sizeof(aSz));

  if( rc==LSM_OK ){
    int bFree;
    int nSz;
    if( aSz[2] & 0x80 ){
      nSz = getRecordSize(aSz, &bFree) + sizeof(aSz)*2;
    }else{
      nSz = (int)(aSz[2] & 0x7F);
      bFree = 1;
    }
    rc = fsSubtractOffset(pFS, iOff, nSz, piPrev);
  }

  return rc;
}

/*
** The first argument to this function is a valid reference to a database







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












|



|






|
>
>
>
>
>




|
|










|







1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
    iBlk = iNext;
  }

  pRun->nSize -= (pRun->iFirst - fsFirstPageOnBlock(pFS, iBlk));
  assert( pRun->nSize>0 );
}

/*
** This function is only used in compressed database mode.
**
** Argument iPg is the page number (byte offset) of a page within segment
** pSeg. The page record, including all headers, is nByte bytes in size.
** Before returning, set *piNext to the page number of the next page in
** the segment, or to zero if iPg is the last.
**
** In other words, do:
**
**   *piNext = iPg + nByte;
**
** But take block overflow and redirection into account.
*/
static int fsNextPageOffset(
  FileSystem *pFS,                /* File system object */
  Segment *pSeg,                  /* Segment to move within */
  Pgno iPg,                       /* Offset of current page */
  int nByte,                      /* Size of current page including headers */
  Pgno *piNext                    /* OUT: Offset of next page. Or zero (EOF) */
){
  Pgno iNext;
  int rc;

  assert( pFS->pCompress );

  rc = fsAddOffset(pFS, pSeg, iPg, nByte-1, &iNext);
  if( pSeg && iNext==pSeg->iLastPg ){
    iNext = 0;
  }else if( rc==LSM_OK ){
    rc = fsAddOffset(pFS, pSeg, iNext, 1, &iNext);
  }

  *piNext = iNext;
  return rc;
}

static int fsGetPageBefore(
  FileSystem *pFS, 
  Segment *pSeg, 
  i64 iOff, 
  Pgno *piPrev
){
  u8 aSz[3];
  int rc;
  i64 iRead;

  rc = fsSubtractOffset(pFS, pSeg, iOff, sizeof(aSz), &iRead);
  if( rc==LSM_OK ) rc = fsReadData(pFS, pSeg, iRead, aSz, sizeof(aSz));

  if( rc==LSM_OK ){
    int bFree;
    int nSz;
    if( aSz[2] & 0x80 ){
      nSz = getRecordSize(aSz, &bFree) + sizeof(aSz)*2;
    }else{
      nSz = (int)(aSz[2] & 0x7F);
      bFree = 1;
    }
    rc = fsSubtractOffset(pFS, pSeg, iOff, nSz, piPrev);
  }

  return rc;
}

/*
** The first argument to this function is a valid reference to a database
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
  FileSystem *pFS = pPg->pFS;
  Pgno iPg = pPg->iPg;

  assert( 0==fsSegmentRedirects(pFS, pRun) );
  if( pFS->pCompress ){
    int nSpace = pPg->nCompress + 2*3;

    /* TODO: Fix redirects */
    assert( 0 );

    do {
      if( eDir>0 ){
        rc = fsNextPageOffset(pFS, pRun, iPg, nSpace, &iPg);
      }else{
        if( iPg==pRun->iFirst ){
          iPg = 0;
        }else{
          rc = fsGetPageBefore(pFS, iPg, &iPg);
        }
      }

      nSpace = 0;
      if( iPg!=0 ){
        rc = fsPageGet(pFS, pRun->pRedirect, iPg, 0, ppNext, &nSpace);
        assert( (*ppNext==0)==(rc!=LSM_OK || nSpace>0) );
      }else{
        *ppNext = 0;
      }
    }while( nSpace>0 && rc==LSM_OK );

  }else{







<
<
<







|





|







1623
1624
1625
1626
1627
1628
1629



1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
  FileSystem *pFS = pPg->pFS;
  Pgno iPg = pPg->iPg;

  assert( 0==fsSegmentRedirects(pFS, pRun) );
  if( pFS->pCompress ){
    int nSpace = pPg->nCompress + 2*3;




    do {
      if( eDir>0 ){
        rc = fsNextPageOffset(pFS, pRun, iPg, nSpace, &iPg);
      }else{
        if( iPg==pRun->iFirst ){
          iPg = 0;
        }else{
          rc = fsGetPageBefore(pFS, pRun, iPg, &iPg);
        }
      }

      nSpace = 0;
      if( iPg!=0 ){
        rc = fsPageGet(pFS, pRun, iPg, 0, ppNext, &nSpace);
        assert( (*ppNext==0)==(rc!=LSM_OK || nSpace>0) );
      }else{
        *ppNext = 0;
      }
    }while( nSpace>0 && rc==LSM_OK );

  }else{
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
            pRedir, lsmGetU32(&pPg->aData[pFS->nPagesize-4])
        );
        iPg = fsFirstPageOnBlock(pFS, iBlk);
      }else{
        iPg++;
      }
    }
    rc = fsPageGet(pFS, pRedir, iPg, 0, ppNext, 0);
  }

  return rc;
}

static Pgno findAppendPoint(FileSystem *pFS, Level *pLvl){
  int i;







|







1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
            pRedir, lsmGetU32(&pPg->aData[pFS->nPagesize-4])
        );
        iPg = fsFirstPageOnBlock(pFS, iBlk);
      }else{
        iPg++;
      }
    }
    rc = fsPageGet(pFS, pRun, iPg, 0, ppNext, 0);
  }

  return rc;
}

static Pgno findAppendPoint(FileSystem *pFS, Level *pLvl){
  int i;
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
}

/*
** Obtain a reference to page number iPg.
*/
int lsmFsDbPageGet(FileSystem *pFS, Segment *pSeg, Pgno iPg, Page **ppPg){
  assert( pFS );
  return fsPageGet(pFS, (pSeg ? pSeg->pRedirect : 0), iPg, 0, ppPg, 0);
}

/*
** Obtain a reference to the last page in the segment passed as the 
** second argument.
*/
int lsmFsDbPageLast(FileSystem *pFS, Segment *pSeg, Page **ppPg){
  int rc;
  Pgno iPg = pSeg->iLastPg;
  if( pFS->pCompress ){
    int nSpace;
    iPg++;
    do {
      nSpace = 0;
      rc = fsGetPageBefore(pFS, iPg, &iPg);
      if( rc==LSM_OK ){
        rc = fsPageGet(pFS, pSeg->pRedirect, iPg, 0, ppPg, &nSpace);
      }
    }while( rc==LSM_OK && nSpace>0 );

  }else{
    rc = fsPageGet(pFS, pSeg->pRedirect, iPg, 0, ppPg, 0);
  }
  return rc;
}

/*
** Return a reference to meta-page iPg. If successful, LSM_OK is returned
** and *ppPg populated with the new page reference. The reference should







|














|

|




|







1839
1840
1841
1842
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
}

/*
** Obtain a reference to page number iPg.
*/
int lsmFsDbPageGet(FileSystem *pFS, Segment *pSeg, Pgno iPg, Page **ppPg){
  assert( pFS );
  return fsPageGet(pFS, pSeg, iPg, 0, ppPg, 0);
}

/*
** Obtain a reference to the last page in the segment passed as the 
** second argument.
*/
int lsmFsDbPageLast(FileSystem *pFS, Segment *pSeg, Page **ppPg){
  int rc;
  Pgno iPg = pSeg->iLastPg;
  if( pFS->pCompress ){
    int nSpace;
    iPg++;
    do {
      nSpace = 0;
      rc = fsGetPageBefore(pFS, pSeg, iPg, &iPg);
      if( rc==LSM_OK ){
        rc = fsPageGet(pFS, pSeg, iPg, 0, ppPg, &nSpace);
      }
    }while( rc==LSM_OK && nSpace>0 );

  }else{
    rc = fsPageGet(pFS, pSeg, iPg, 0, ppPg, 0);
  }
  return rc;
}

/*
** Return a reference to meta-page iPg. If successful, LSM_OK is returned
** and *ppPg populated with the new page reference. The reference should
1938
1939
1940
1941
1942
1943
1944
1945


1946
1947
1948
1949
1950
1951
1952
  Segment *pSeg, 
  int iTo, 
  int iFrom, 
  Pgno *piPg
){
  Pgno iPg = *piPg;
  if( iFrom==fsPageToBlock(pFS, iPg) ){
    const int nPagePerBlock = (pFS->nBlocksize / pFS->nPagesize);


    *piPg = iPg - (Pgno)(iFrom - iTo) * nPagePerBlock;
  }
}

/*
** Copy the contents of block iFrom to block iTo. 
**







|
>
>







1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
  Segment *pSeg, 
  int iTo, 
  int iFrom, 
  Pgno *piPg
){
  Pgno iPg = *piPg;
  if( iFrom==fsPageToBlock(pFS, iPg) ){
    const int nPagePerBlock = (
        pFS->pCompress ? pFS ->nBlocksize : (pFS->nBlocksize / pFS->nPagesize)
    );
    *piPg = iPg - (Pgno)(iFrom - iTo) * nPagePerBlock;
  }
}

/*
** Copy the contents of block iFrom to block iTo. 
**
2647
2648
2649
2650
2651
2652
2653
2654
2655
2656
2657
2658
2659
2660
2661
  Segment *pSeg,
  int bExtra,                     /* If true, count the "next" block if any */
  int nUsed,
  u8 *aUsed
){
  if( pSeg ){
    if( pSeg && pSeg->nSize>0 ){
      Redirect *pRedir = pSeg->pRedirect;
      int rc;
      int iBlk;                   /* Current block (during iteration) */
      int iLastBlk;               /* Last block of segment */
      int iFirstBlk;              /* First block of segment */
      int bLastIsLastOnBlock;     /* True iLast is the last on its block */

      assert( 0==fsSegmentRedirects(pFS, pSeg) );







<







2682
2683
2684
2685
2686
2687
2688

2689
2690
2691
2692
2693
2694
2695
  Segment *pSeg,
  int bExtra,                     /* If true, count the "next" block if any */
  int nUsed,
  u8 *aUsed
){
  if( pSeg ){
    if( pSeg && pSeg->nSize>0 ){

      int rc;
      int iBlk;                   /* Current block (during iteration) */
      int iLastBlk;               /* Last block of segment */
      int iFirstBlk;              /* First block of segment */
      int bLastIsLastOnBlock;     /* True iLast is the last on its block */

      assert( 0==fsSegmentRedirects(pFS, pSeg) );
Changes to src/lsm_sorted.c.
4309
4310
4311
4312
4313
4314
4315
4316
4317

4318
4319
4320
4321
4322
4323
4324
    mergeworker.bFlush = 1;

    /* Do the work to create the new merged segment on disk */
    if( rc==LSM_OK ) rc = lsmMCursorFirst(pCsr);
    while( rc==LSM_OK && mergeWorkerDone(&mergeworker)==0 ){
      rc = mergeWorkerStep(&mergeworker);
    }
    assert( rc!=LSM_OK || mergeworker.nWork==0 || pNew->lhs.iFirst );
    mergeWorkerShutdown(&mergeworker, &rc);

    if( rc==LSM_OK && pNew->lhs.iFirst ){
      rc = lsmFsSortedFinish(pDb->pFS, &pNew->lhs);
    }
    nWrite = mergeworker.nWork;
    pNew->flags &= ~LEVEL_INCOMPLETE;
    if( eTree==TREE_NONE ){
      pNew->flags |= LEVEL_FREELIST_ONLY;







<

>







4309
4310
4311
4312
4313
4314
4315

4316
4317
4318
4319
4320
4321
4322
4323
4324
    mergeworker.bFlush = 1;

    /* Do the work to create the new merged segment on disk */
    if( rc==LSM_OK ) rc = lsmMCursorFirst(pCsr);
    while( rc==LSM_OK && mergeWorkerDone(&mergeworker)==0 ){
      rc = mergeWorkerStep(&mergeworker);
    }

    mergeWorkerShutdown(&mergeworker, &rc);
    assert( rc!=LSM_OK || mergeworker.nWork==0 || pNew->lhs.iFirst );
    if( rc==LSM_OK && pNew->lhs.iFirst ){
      rc = lsmFsSortedFinish(pDb->pFS, &pNew->lhs);
    }
    nWrite = mergeworker.nWork;
    pNew->flags &= ~LEVEL_INCOMPLETE;
    if( eTree==TREE_NONE ){
      pNew->flags |= LEVEL_FREELIST_ONLY;