/ Check-in [0bf1301a]
Login

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

Overview
Comment:Have the ota extension perform an incremental checkpoint after generating the wal file.
Downloads: Tarball | ZIP archive | SQL archive
Timelines: family | ancestors | descendants | both | ota-update
Files: files | file ages | folders
SHA1:0bf1301aacb3b717b4cc020fbda9fab0bae331c3
User & Date: dan 2014-10-20 16:24:23
Context
2014-10-20
16:34
Merge version-3.8.7 changes with this branch. check-in: d380a648 user: dan tags: ota-update
16:24
Have the ota extension perform an incremental checkpoint after generating the wal file. check-in: 0bf1301a user: dan tags: ota-update
2014-09-19
18:08
Add further tests to ota5.test. Add "ota.test", for running all ota tests. check-in: 95ffdaa5 user: dan tags: ota-update
Changes
Hide Diffs Unified Diffs Ignore Whitespace Patch

Changes to ext/ota/ota1.test.

120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
      CREATE INDEX i1 ON t1(b, c);
      CREATE INDEX i2 ON t1(c, b);
      CREATE INDEX i3 ON t1(a, b, c, a, b, c);
    }
  } {
    reset_db
    execsql $schema
  
    do_test 1.$tn2.$tn.1 {
      create_ota1 ota.db
      $cmd test.db ota.db
    } {SQLITE_DONE}
    
    do_execsql_test 1.$tn2.$tn.2 {
      SELECT * FROM t1 ORDER BY a ASC;
    } {
      1 2 3 
      2 two three 
      3 {} 8.2
    }
  
    do_execsql_test 1.$tn2.$tn.3 { PRAGMA integrity_check } ok
  }
}

#-------------------------------------------------------------------------
# Check that an OTA cannot be applied to a table that has no PK.
#







|




|







|







120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
      CREATE INDEX i1 ON t1(b, c);
      CREATE INDEX i2 ON t1(c, b);
      CREATE INDEX i3 ON t1(a, b, c, a, b, c);
    }
  } {
    reset_db
    execsql $schema

    do_test 1.$tn2.$tn.1 {
      create_ota1 ota.db
      $cmd test.db ota.db
    } {SQLITE_DONE}

    do_execsql_test 1.$tn2.$tn.2 {
      SELECT * FROM t1 ORDER BY a ASC;
    } {
      1 2 3 
      2 two three 
      3 {} 8.2
    }
 
    do_execsql_test 1.$tn2.$tn.3 { PRAGMA integrity_check } ok
  }
}

#-------------------------------------------------------------------------
# Check that an OTA cannot be applied to a table that has no PK.
#

Changes to ext/ota/sqlite3ota.c.

18
19
20
21
22
23
24
25
26
27
28






29


30
31
32
33




34
35
36












37













38
39

40
41
42
43
44
45
46
47

48
49


50
51
52
53
54
55
56
..
84
85
86
87
88
89
90

91
92

93
94
95
96
97

98
99
100
101
102
103
104
...
737
738
739
740
741
742
743














































































744
745
746
747
748
749
750
...
858
859
860
861
862
863
864
865
866






















867
868
869
870


871
872
873
874
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
926

927
928





929


930
931












932
933
934
935
936

937
938
939
940

941
942
943
944
945
946

947

948



949

950
951
952
953
954






955



956

957
958
959
960
961
962
963
964
965
966
967
968
969
970
...
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
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
....
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
1108
1109
1110
1111
1112
1113
1114
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
1164
1165
1166
1167
1168
1169
1170

#include "sqlite3.h"
#include "sqlite3ota.h"


/*
** The ota_state table is used to save the state of a partially applied
** update so that it can be resumed later. The table contains at most a
** single row:
**
**   "tbl"       -> Table currently being written (target database names).






**


**   "idx"       -> Index currently being written (target database names).
**                  Or, if the main table is being written, a NULL value.
**
**   "row"       -> Number of rows for this object already processed




**
**   "progress"  -> total number of key/value b-tree operations performed
**                  so far as part of this ota update.












*/













#define OTA_CREATE_STATE "CREATE TABLE IF NOT EXISTS ota.ota_state"        \
                             "(tbl, idx, row, progress)"


typedef struct OtaState OtaState;
typedef struct OtaObjIter OtaObjIter;

/*
** A structure to store values read from the ota_state table in memory.
*/
struct OtaState {

  char *zTbl;
  char *zIdx;


  int nRow;
  sqlite3_int64 nProgress;
};

/*
** An iterator of this type is used to iterate through all objects in
** the target database that require updating. For each such table, the
................................................................................
  sqlite3_stmt *pUpdate;          /* Last update statement (or NULL) */
};

/*
** OTA handle.
*/
struct sqlite3ota {

  sqlite3 *db;                    /* "main" -> target db, "ota" -> ota db */
  char *zTarget;                  /* Path to target db */

  int rc;                         /* Value returned by last ota_step() call */
  char *zErrmsg;                  /* Error message if rc!=SQLITE_OK */
  int nStep;                      /* Rows processed for current object */
  int nProgress;                  /* Rows processed for all objects */
  OtaObjIter objiter;

};

/*
** Prepare the SQL statement in buffer zSql against database handle db.
** If successful, set *ppStmt to point to the new statement and return
** SQLITE_OK. 
**
................................................................................
      memcpy(pIter->zMask, zMask, pIter->nTblCol);
    }
    sqlite3_free(zWhere);
    sqlite3_free(zSet);
  }
  return p->rc;
}















































































/*
** The SELECT statement iterating through the keys for the current object
** (p->objiter.pSelect) currently points to a valid row. This function
** determines the type of operation requested by this row and returns
** one of the following values to indicate the result:
**
................................................................................
      /* no-op */
      assert( eType==OTA_DELETE && pIter->zIdx );
    }
  }

  return p->rc;
}

/*






















** Step the OTA object.
*/
int sqlite3ota_step(sqlite3ota *p){
  if( p ){


    OtaObjIter *pIter = &p->objiter;
    while( p && p->rc==SQLITE_OK && pIter->zTbl ){

      if( pIter->bCleanup ){
        /* Clean up the ota_tmp_xxx table for the previous table. It 
        ** cannot be dropped as there are currently active SQL statements.
        ** But the contents can be deleted.  */
        otaMPrintfExec(p, "DELETE FROM ota.'ota_tmp_%q'", pIter->zTbl);
      }else{
        otaObjIterPrepareAll(p, pIter, 0);
        
        /* Advance to the next row to process. */
        if( p->rc==SQLITE_OK ){
          int rc = sqlite3_step(pIter->pSelect);
          if( rc==SQLITE_ROW ){
            p->nStep++;
            p->nProgress++;

            return otaStep(p);
          }
          p->rc = sqlite3_reset(pIter->pSelect);
          p->nStep = 0;
        }
      }

      otaObjIterNext(p, pIter);
    }

    if( p->rc==SQLITE_OK && pIter->zTbl==0 ){























      p->rc = SQLITE_DONE;










    }
  }
  return p->rc;
}

static void otaSaveTransactionState(sqlite3ota *p){
  otaMPrintfExec(p, 





    "INSERT OR REPLACE INTO ota.ota_state(rowid, tbl, idx, row, progress)"




    "VALUES(1, %Q, %Q, %d, %lld)",



    p->objiter.zTbl, p->objiter.zIdx, p->nStep, p->nProgress




  );














































}

/*
** Allocate an OtaState object and load the contents of the ota_state 
** table into it. Return a pointer to the new object. It is the 
** responsibility of the caller to eventually free the object using
** sqlite3_free().
**
** If an error occurs, leave an error code and message in the ota handle
** and return NULL.
*/
static OtaState *otaLoadState(sqlite3ota *p){
  const char *zSelect = "SELECT tbl, idx, row, progress FROM ota.ota_state";
  OtaState *pRet = 0;
  sqlite3_stmt *pStmt;
  int rc;


  assert( p->rc==SQLITE_OK );





  rc = prepareAndCollectError(p->db, &pStmt, &p->zErrmsg, zSelect);


  if( rc==SQLITE_OK ){
    if( sqlite3_step(pStmt)==SQLITE_ROW ){












      const char *zIdx = (const char*)sqlite3_column_text(pStmt, 1);
      const char *zTbl = (const char*)sqlite3_column_text(pStmt, 0);
      int nIdx = zIdx ? (strlen(zIdx) + 1) : 0;
      int nTbl = strlen(zTbl) + 1;
      int nByte = sizeof(OtaState) + nTbl + nIdx;


      pRet = (OtaState*)sqlite3_malloc(nByte);
      if( pRet ){
        pRet->zTbl = (char*)&pRet[1];

        memcpy(pRet->zTbl, sqlite3_column_text(pStmt, 0), nTbl);
        if( zIdx ){
          pRet->zIdx = &pRet->zTbl[nTbl];
          memcpy(pRet->zIdx, zIdx, nIdx);
        }else{
          pRet->zIdx = 0;

        }

        pRet->nRow = sqlite3_column_int(pStmt, 2);



        pRet->nProgress = sqlite3_column_int64(pStmt, 3);

      }
    }else{
      pRet = (OtaState*)sqlite3_malloc(sizeof(OtaState));
      if( pRet ){
        memset(pRet, 0, sizeof(*pRet));






      }



    }

    rc = sqlite3_finalize(pStmt);
    if( rc==SQLITE_OK && pRet==0 ) rc = SQLITE_NOMEM;
    if( rc!=SQLITE_OK ){
      sqlite3_free(pRet);
      pRet = 0;
    }
  }

  p->rc = rc;
  return pRet;
}

static int otaStrCompare(const char *z1, const char *z2){
  if( z1==0 && z2==0 ) return 0;
................................................................................
      rc = otaObjIterPrepareAll(p, &p->objiter, p->nStep);
    }

    p->rc = rc;
  }
}

/*
** This routine is a copy of the sqlite3FileSuffix3() routine from the core.
** It is a no-op unless SQLITE_ENABLE_8_3_NAMES is defined.
**
** If SQLITE_ENABLE_8_3_NAMES is set at compile-time and if the database
** filename in zBaseFilename is a URI with the "8_3_names=1" parameter and
** if filename in z[] has a suffix (a.k.a. "extension") that is longer than
** three characters, then shorten the suffix on z[] to be the last three
** characters of the original suffix.
**
** If SQLITE_ENABLE_8_3_NAMES is set to 2 at compile-time, then always
** do the suffix shortening regardless of URI parameter.
**
** Examples:
**
**     test.db-journal    =>   test.nal
**     test.db-wal        =>   test.wal
**     test.db-shm        =>   test.shm
**     test.db-mj7f3319fa =>   test.9fa
*/
static void otaFileSuffix3(const char *zBase, char *z){
#ifdef SQLITE_ENABLE_8_3_NAMES
#if SQLITE_ENABLE_8_3_NAMES<2
  if( sqlite3_uri_boolean(zBase, "8_3_names", 0) )
#endif
  {
    int i, sz;
    sz = sqlite3Strlen30(z);
    for(i=sz-1; i>0 && z[i]!='/' && z[i]!='.'; i--){}
    if( z[i]=='.' && ALWAYS(sz>i+4) ) memmove(&z[i+1], &z[sz-3], 4);
  }
#endif
}

/*
** Move the "*-oal" file corresponding to the target database to the
** "*-wal" location. If an error occurs, leave an error code and error 
** message in the ota handle.
*/
static void otaMoveOalFile(const char *zBase, sqlite3ota *p){
  char *zWal = sqlite3_mprintf("%s-wal", p->zTarget);
  char *zOal = sqlite3_mprintf("%s-oal", p->zTarget);

  assert( p->rc==SQLITE_DONE && p->zErrmsg==0 );
  if( zWal==0 || zOal==0 ){
    p->rc = SQLITE_NOMEM;
  }else{
    otaFileSuffix3(zBase, zWal);
    otaFileSuffix3(zBase, zOal);
    rename(zOal, zWal);
  }

  sqlite3_free(zWal);
  sqlite3_free(zOal);
}

/*
** If there is a "*-oal" file in the file-system corresponding to the
** target database in the file-system, delete it. If an error occurs,
** leave an error code and error message in the ota handle.
*/
static void otaDeleteOalFile(sqlite3ota *p){
  char *zOal = sqlite3_mprintf("%s-oal", p->zTarget);
................................................................................

/*
** Open and return a new OTA handle. 
*/
sqlite3ota *sqlite3ota_open(const char *zTarget, const char *zOta){
  sqlite3ota *p;
  int nTarget = strlen(zTarget);


  p = (sqlite3ota*)sqlite3_malloc(sizeof(sqlite3ota)+nTarget+1);
  if( p ){
    OtaState *pState = 0;

    /* Open the target database */
    memset(p, 0, sizeof(sqlite3ota));
    p->zTarget = (char*)&p[1];
    memcpy(p->zTarget, zTarget, nTarget+1);
    p->rc = sqlite3_open(zTarget, &p->db);
    if( p->rc ){
      p->zErrmsg = sqlite3_mprintf("%s", sqlite3_errmsg(p->db));
    }
    otaMPrintfExec(p, "ATTACH %Q AS ota", zOta);

    /* If it has not already been created, create the ota_state table */
    if( p->rc==SQLITE_OK ){
      p->rc = sqlite3_exec(p->db, OTA_CREATE_STATE, 0, 0, &p->zErrmsg);
    }

    if( p->rc==SQLITE_OK ){
      pState = otaLoadState(p);

      if( pState && pState->zTbl==0 ){

        otaDeleteOalFile(p);



      }

    }




    if( p->rc==SQLITE_OK ){
      const char *zScript =
        "PRAGMA journal_mode=off;"
        "PRAGMA pager_ota_mode=1;"
        "PRAGMA ota_mode=1;"
        "BEGIN IMMEDIATE;"
      ;
      p->rc = sqlite3_exec(p->db, zScript, 0, 0, &p->zErrmsg);
    }

    /* Point the object iterator at the first object */
    if( p->rc==SQLITE_OK ){
      p->rc = otaObjIterFirst(p, &p->objiter);
    }

    if( p->rc==SQLITE_OK ){
      p->nProgress = pState->nProgress;
      otaLoadTransactionState(p, pState);
    }






    sqlite3_free(pState);
  }

  return p;
}

/*
** Close the OTA handle.
*/
int sqlite3ota_close(sqlite3ota *p, char **pzErrmsg){
  int rc;
  if( p ){
    const char *zBase = sqlite3_db_filename(p->db, "main");

    /* If the update has not been fully applied, save the state in 
    ** the ota db. If successful, this call also commits the open 
    ** transaction on the ota db. */
    assert( p->rc!=SQLITE_ROW );
    if( p->rc==SQLITE_OK ){
      assert( p->zErrmsg==0 );
      otaSaveTransactionState(p);
    }

    /* Close all open statement handles. */
    otaObjIterFinalize(&p->objiter);

    /* Commit the transaction to the *-oal file. */
    if( p->rc==SQLITE_OK || p->rc==SQLITE_DONE ){
      rc = sqlite3_exec(p->db, "COMMIT", 0, 0, &p->zErrmsg);
      if( rc!=SQLITE_OK ) p->rc = rc;
    }


    /* Close the open database handles */
    sqlite3_close(p->db);

    /* If the OTA has been completely applied and no error occurred, move
    ** the *-oal file to *-wal. */
    if( p->rc==SQLITE_DONE ){
      otaMoveOalFile(zBase, p);
    }




    rc = p->rc;
    *pzErrmsg = p->zErrmsg;
    sqlite3_free(p);
  }else{
    rc = SQLITE_NOMEM;
    *pzErrmsg = 0;
  }







|
|

|
>
>
>
>
>
>

>
>
|
<

<
>
>
>
>

<
<
>
>
>
>
>
>
>
>
>
>
>
>

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

<
>








>


>
>







 







>


>




|
>







 







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







 









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




>
>
|
|

|
|
|
|
|
|
|
|
|
|
|
|
<
|
>
|
|
|
|
|
|

|
|

|
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
|
>
>
>
>
>
>
>
>
>
>






|
>
>
>
>
>
|
>
>
>
>
|
>
>
>
|
>
>
>
>

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












|



>


>
>
>
>
>
|
>
>
|
<
>
>
>
>
>
>
>
>
>
>
>
>
|
<
<
<
<
>

<
<
<
>
|
<
<
<
<
<
>
|
>
|
>
>
>
|
>
|
<
<
<
<
>
>
>
>
>
>
|
>
>
>

>
|
|
<
<
<
<
<







 







<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<







 







>

|







|
|
|
<
<








>
|
>
|
>
>
>
|
>
|
|
>
>
>
|
|
|
|
|
|
|
|
|

|
|
|
|

|
<
|
|
>
>
>
>
|
>
|











<





|




|



|
|
<


>
|
<
|
<
<
<
<
|
>
>
>








18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38

39

40
41
42
43
44


45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71

72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
...
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
...
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
...
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
1040
1041
1042
1043
1044
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
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
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
1164
1165
1166
1167
1168
1169
1170
1171
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
1200
1201




1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215





1216
1217
1218
1219
1220
1221
1222
....
1247
1248
1249
1250
1251
1252
1253
























































1254
1255
1256
1257
1258
1259
1260
....
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284


1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322

1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342

1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358

1359
1360
1361
1362

1363




1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375

#include "sqlite3.h"
#include "sqlite3ota.h"


/*
** The ota_state table is used to save the state of a partially applied
** update so that it can be resumed later. The table consists of integer
** keys mapped to values as follows:
**
** OTA_STATE_STAGE:
**   May be set to integer values 1, 2 or 3. As follows:
**       0: Nothing at all has been done.
**       1: the *-ota file is currently under construction.
**       2: the *-ota file has been constructed, but not yet moved 
**          to the *-wal path.
**       3: the checkpoint is underway.
**
** OTA_STATE_TBL:
**   Only valid if STAGE==1. The target database name of the table 
**   currently being written.

**

** OTA_STATE_IDX:
**   Only valid if STAGE==1. The target database name of the index 
**   currently being written, or NULL if the main table is currently being
**   updated.
**


** OTA_STATE_ROW:
**   Only valid if STAGE==1. Number of rows already processed for the current
**   table/index.
**
** OTA_STATE_PROGRESS:
**   Total number of sqlite3ota_step() calls made so far as part of this
**   ota update.
**
** OTA_STATE_CKPT:
**   Valid if STAGE==3. The blob to pass to sqlite3ckpt_start() to resume
**   the incremental checkpoint.
**
*/
#define OTA_STATE_STAGE       1
#define OTA_STATE_TBL         2
#define OTA_STATE_IDX         3
#define OTA_STATE_ROW         4
#define OTA_STATE_PROGRESS    5
#define OTA_STATE_CKPT        6

#define OTA_STAGE_OAL         1
#define OTA_STAGE_COPY        2
#define OTA_STAGE_CKPT        3
#define OTA_STAGE_DONE        4


#define OTA_CREATE_STATE "CREATE TABLE IF NOT EXISTS ota.ota_state"        \

                             "(k INTEGER PRIMARY KEY, v)"

typedef struct OtaState OtaState;
typedef struct OtaObjIter OtaObjIter;

/*
** A structure to store values read from the ota_state table in memory.
*/
struct OtaState {
  int eStage;
  char *zTbl;
  char *zIdx;
  unsigned char *pCkptState;
  int nCkptState;
  int nRow;
  sqlite3_int64 nProgress;
};

/*
** An iterator of this type is used to iterate through all objects in
** the target database that require updating. For each such table, the
................................................................................
  sqlite3_stmt *pUpdate;          /* Last update statement (or NULL) */
};

/*
** OTA handle.
*/
struct sqlite3ota {
  int eStage;                     /* Value of OTA_STATE_STAGE field */
  sqlite3 *db;                    /* "main" -> target db, "ota" -> ota db */
  char *zTarget;                  /* Path to target db */
  char *zOta;                     /* Path to ota db */
  int rc;                         /* Value returned by last ota_step() call */
  char *zErrmsg;                  /* Error message if rc!=SQLITE_OK */
  int nStep;                      /* Rows processed for current object */
  int nProgress;                  /* Rows processed for all objects */
  OtaObjIter objiter;             /* Iterator for skipping through tbl/idx */
  sqlite3_ckpt *pCkpt;            /* Incr-checkpoint handle */
};

/*
** Prepare the SQL statement in buffer zSql against database handle db.
** If successful, set *ppStmt to point to the new statement and return
** SQLITE_OK. 
**
................................................................................
      memcpy(pIter->zMask, zMask, pIter->nTblCol);
    }
    sqlite3_free(zWhere);
    sqlite3_free(zSet);
  }
  return p->rc;
}

static void otaOpenDatabase(sqlite3ota *p){
  assert( p->rc==SQLITE_OK );
  sqlite3_close(p->db);
  p->db = 0;

  p->rc = sqlite3_open(p->zTarget, &p->db);
  if( p->rc ){
    p->zErrmsg = sqlite3_mprintf("%s", sqlite3_errmsg(p->db));
  }
  otaMPrintfExec(p, "ATTACH %Q AS ota", p->zOta);
}

/*
** This routine is a copy of the sqlite3FileSuffix3() routine from the core.
** It is a no-op unless SQLITE_ENABLE_8_3_NAMES is defined.
**
** If SQLITE_ENABLE_8_3_NAMES is set at compile-time and if the database
** filename in zBaseFilename is a URI with the "8_3_names=1" parameter and
** if filename in z[] has a suffix (a.k.a. "extension") that is longer than
** three characters, then shorten the suffix on z[] to be the last three
** characters of the original suffix.
**
** If SQLITE_ENABLE_8_3_NAMES is set to 2 at compile-time, then always
** do the suffix shortening regardless of URI parameter.
**
** Examples:
**
**     test.db-journal    =>   test.nal
**     test.db-wal        =>   test.wal
**     test.db-shm        =>   test.shm
**     test.db-mj7f3319fa =>   test.9fa
*/
static void otaFileSuffix3(const char *zBase, char *z){
#ifdef SQLITE_ENABLE_8_3_NAMES
#if SQLITE_ENABLE_8_3_NAMES<2
  if( sqlite3_uri_boolean(zBase, "8_3_names", 0) )
#endif
  {
    int i, sz;
    sz = sqlite3Strlen30(z);
    for(i=sz-1; i>0 && z[i]!='/' && z[i]!='.'; i--){}
    if( z[i]=='.' && ALWAYS(sz>i+4) ) memmove(&z[i+1], &z[sz-3], 4);
  }
#endif
}

/*
** Move the "*-oal" file corresponding to the target database to the
** "*-wal" location. If an error occurs, leave an error code and error 
** message in the ota handle.
*/
static void otaMoveOalFile(sqlite3ota *p){
  const char *zBase = sqlite3_db_filename(p->db, "main");

  char *zWal = sqlite3_mprintf("%s-wal", zBase);
  char *zOal = sqlite3_mprintf("%s-oal", zBase);

  assert( p->rc==SQLITE_OK && p->zErrmsg==0 );
  if( zWal==0 || zOal==0 ){
    p->rc = SQLITE_NOMEM;
  }else{
    /* Move the *-oal file to *-wal. At this point connection p->db is
    ** holding a SHARED lock on the target database file (because it is
    ** in WAL mode). So no other connection may be writing the db.  */
    otaFileSuffix3(zBase, zWal);
    otaFileSuffix3(zBase, zOal);
    rename(zOal, zWal);

    /* Re-open the databases. */
    otaObjIterFinalize(&p->objiter);
    otaOpenDatabase(p);
    p->eStage = OTA_STAGE_CKPT;
  }

  sqlite3_free(zWal);
  sqlite3_free(zOal);
}

/*
** The SELECT statement iterating through the keys for the current object
** (p->objiter.pSelect) currently points to a valid row. This function
** determines the type of operation requested by this row and returns
** one of the following values to indicate the result:
**
................................................................................
      /* no-op */
      assert( eType==OTA_DELETE && pIter->zIdx );
    }
  }

  return p->rc;
}

/*
** Increment the schema cookie of the main database opened by p->db.
*/
static void otaIncrSchemaCookie(sqlite3ota *p){
  int iCookie = 1000000;
  sqlite3_stmt *pStmt;

  assert( p->rc==SQLITE_OK && p->zErrmsg==0 );
  p->rc = prepareAndCollectError(p->db, &pStmt, &p->zErrmsg, 
      "PRAGMA schema_version"
  );
  if( p->rc==SQLITE_OK ){
    if( SQLITE_ROW==sqlite3_step(pStmt) ){
      iCookie = sqlite3_column_int(pStmt, 0);
    }
    p->rc = sqlite3_finalize(pStmt);
  }
  if( p->rc==SQLITE_OK ){
    otaMPrintfExec(p, "PRAGMA schema_version = %d", iCookie+1);
  }
}

/*
** Step the OTA object.
*/
int sqlite3ota_step(sqlite3ota *p){
  if( p ){
    switch( p->eStage ){
      case OTA_STAGE_OAL: {
        OtaObjIter *pIter = &p->objiter;
        while( p && p->rc==SQLITE_OK && pIter->zTbl ){

          if( pIter->bCleanup ){
            /* Clean up the ota_tmp_xxx table for the previous table. It 
            ** cannot be dropped as there are currently active SQL statements.
            ** But the contents can be deleted.  */
            otaMPrintfExec(p, "DELETE FROM ota.'ota_tmp_%q'", pIter->zTbl);
          }else{
            otaObjIterPrepareAll(p, pIter, 0);

            /* Advance to the next row to process. */
            if( p->rc==SQLITE_OK ){
              int rc = sqlite3_step(pIter->pSelect);
              if( rc==SQLITE_ROW ){

                p->nProgress++;
                p->nStep++;
                return otaStep(p);
              }
              p->rc = sqlite3_reset(pIter->pSelect);
              p->nStep = 0;
            }
          }

          otaObjIterNext(p, pIter);
        }

        if( p->rc==SQLITE_OK && pIter->zTbl==0 ){
          p->nProgress++;
          otaIncrSchemaCookie(p);
          if( p->rc==SQLITE_OK ){
            p->rc = sqlite3_exec(p->db, "COMMIT", 0, 0, &p->zErrmsg);
          }
          if( p->rc==SQLITE_OK ){
            otaMoveOalFile(p);
          }
        }
        break;
      }

      case OTA_STAGE_CKPT: {

        if( p->rc==SQLITE_OK && p->pCkpt==0 ){
          p->rc = sqlite3_ckpt_open(p->db, 0, 0, &p->pCkpt);
        }
        if( p->rc==SQLITE_OK ){
          if( SQLITE_OK!=sqlite3_ckpt_step(p->pCkpt) ){
            p->rc = sqlite3_ckpt_close(p->pCkpt, 0, 0);
            p->pCkpt = 0;
            if( p->rc==SQLITE_OK ){
              p->eStage = OTA_STAGE_DONE;
              p->rc = SQLITE_DONE;
            }
          }
          p->nProgress++;
        }

        break;
      }

      default:
        break;
    }
  }
  return p->rc;
}

static void otaSaveTransactionState(sqlite3ota *p){
  sqlite3_stmt *pInsert;
  int rc;

  assert( (p->rc==SQLITE_OK || p->rc==SQLITE_DONE) && p->zErrmsg==0 );
  rc = prepareFreeAndCollectError(p->db, &pInsert, &p->zErrmsg, 
      sqlite3_mprintf(
        "INSERT OR REPLACE INTO ota.ota_state(k, v) VALUES "
        "(%d, %d), "
        "(%d, %Q), "
        "(%d, %Q), "
        "(%d, %d), "
        "(%d, %lld), "
        "(%d, ?) ",
        OTA_STATE_STAGE, p->eStage,
        OTA_STATE_TBL, p->objiter.zTbl, 
        OTA_STATE_IDX, p->objiter.zIdx, 
        OTA_STATE_ROW, p->nStep, 
        OTA_STATE_PROGRESS, p->nProgress,
        OTA_STATE_CKPT
      )
  );
  assert( pInsert==0 || rc==SQLITE_OK );
  if( rc==SQLITE_OK ){
    if( p->pCkpt ){
      unsigned char *pCkptState = 0;
      int nCkptState = 0;
      rc = sqlite3_ckpt_close(p->pCkpt, &pCkptState, &nCkptState);
      p->pCkpt = 0;
      sqlite3_bind_blob(pInsert, 1, pCkptState, nCkptState, SQLITE_TRANSIENT);
      sqlite3_free(pCkptState);
    }
  }
  if( rc==SQLITE_OK ){
    sqlite3_step(pInsert);
    rc = sqlite3_finalize(pInsert);
  }else{
    sqlite3_finalize(pInsert);
  }

  if( rc!=SQLITE_OK ){
    p->rc = rc;
  }
}

static char *otaStrndup(char *zStr, int nStr, int *pRc){
  char *zRet = 0;
  assert( *pRc==SQLITE_OK );

  if( zStr ){
    int nCopy = nStr;
    if( nCopy<0 ) nCopy = strlen(zStr) + 1;
    zRet = (char*)sqlite3_malloc(nCopy);
    if( zRet ){
      memcpy(zRet, zStr, nCopy);
    }else{
      *pRc = SQLITE_NOMEM;
    }
  }

  return zRet;
}

static void otaFreeState(OtaState *p){
  sqlite3_free(p->zTbl);
  sqlite3_free(p->zIdx);
  sqlite3_free(p->pCkptState);
  sqlite3_free(p);
}

/*
** Allocate an OtaState object and load the contents of the ota_state 
** table into it. Return a pointer to the new object. It is the 
** responsibility of the caller to eventually free the object using
** sqlite3_free().
**
** If an error occurs, leave an error code and message in the ota handle
** and return NULL.
*/
static OtaState *otaLoadState(sqlite3ota *p){
  const char *zSelect = "SELECT k, v FROM ota.ota_state";
  OtaState *pRet = 0;
  sqlite3_stmt *pStmt;
  int rc;
  int rc2;

  assert( p->rc==SQLITE_OK );
  pRet = (OtaState*)sqlite3_malloc(sizeof(OtaState));
  if( pRet==0 ){
    rc = SQLITE_NOMEM;
  }else{
    memset(pRet, 0, sizeof(OtaState));
    rc = prepareAndCollectError(p->db, &pStmt, &p->zErrmsg, zSelect);
  }

  while( rc==SQLITE_OK && SQLITE_ROW==sqlite3_step(pStmt) ){

    switch( sqlite3_column_int(pStmt, 0) ){
      case OTA_STATE_STAGE:
        pRet->eStage = sqlite3_column_int(pStmt, 1);
        if( pRet->eStage!=OTA_STAGE_OAL
         && pRet->eStage!=OTA_STAGE_COPY
         && pRet->eStage!=OTA_STAGE_CKPT
        ){
          p->rc = SQLITE_CORRUPT;
        }
        break;

      case OTA_STATE_TBL:
        pRet->zTbl = otaStrndup((char*)sqlite3_column_text(pStmt, 1), -1, &rc);




        break;




      case OTA_STATE_IDX:
        pRet->zIdx = otaStrndup((char*)sqlite3_column_text(pStmt, 1), -1, &rc);





        break;

      case OTA_STATE_ROW:
        pRet->nRow = sqlite3_column_int(pStmt, 1);
        break;

      case OTA_STATE_PROGRESS:
        pRet->nProgress = sqlite3_column_int64(pStmt, 1);
        break;





      case OTA_STATE_CKPT:
        pRet->nCkptState = sqlite3_column_bytes(pStmt, 1);
        pRet->pCkptState = (unsigned char*)otaStrndup(
            (char*)sqlite3_column_blob(pStmt, 1), pRet->nCkptState, &rc
        );
        break;

      default:
        rc = SQLITE_CORRUPT;
        break;
    }
  }
  rc2 = sqlite3_finalize(pStmt);
  if( rc==SQLITE_OK ) rc = rc2;






  p->rc = rc;
  return pRet;
}

static int otaStrCompare(const char *z1, const char *z2){
  if( z1==0 && z2==0 ) return 0;
................................................................................
      rc = otaObjIterPrepareAll(p, &p->objiter, p->nStep);
    }

    p->rc = rc;
  }
}

























































/*
** If there is a "*-oal" file in the file-system corresponding to the
** target database in the file-system, delete it. If an error occurs,
** leave an error code and error message in the ota handle.
*/
static void otaDeleteOalFile(sqlite3ota *p){
  char *zOal = sqlite3_mprintf("%s-oal", p->zTarget);
................................................................................

/*
** Open and return a new OTA handle. 
*/
sqlite3ota *sqlite3ota_open(const char *zTarget, const char *zOta){
  sqlite3ota *p;
  int nTarget = strlen(zTarget);
  int nOta = strlen(zOta);

  p = (sqlite3ota*)sqlite3_malloc(sizeof(sqlite3ota)+nTarget+1+nOta+1);
  if( p ){
    OtaState *pState = 0;

    /* Open the target database */
    memset(p, 0, sizeof(sqlite3ota));
    p->zTarget = (char*)&p[1];
    memcpy(p->zTarget, zTarget, nTarget+1);
    p->zOta = &p->zTarget[nTarget+1];
    memcpy(p->zOta, zOta, nOta+1);
    otaOpenDatabase(p);



    /* If it has not already been created, create the ota_state table */
    if( p->rc==SQLITE_OK ){
      p->rc = sqlite3_exec(p->db, OTA_CREATE_STATE, 0, 0, &p->zErrmsg);
    }

    if( p->rc==SQLITE_OK ){
      pState = otaLoadState(p);
      assert( pState || p->rc!=SQLITE_OK );
      if( pState ){
        if( pState->eStage==0 ){ 
          otaDeleteOalFile(p);
          p->eStage = 1;
        }else{
          p->eStage = pState->eStage;
        }
        p->nProgress = pState->nProgress;
      }
    }
    assert( p->rc!=SQLITE_OK || p->eStage!=0 );

    if( p->eStage==OTA_STAGE_OAL ){
      if( p->rc==SQLITE_OK ){
        const char *zScript =
          "PRAGMA journal_mode=off;"
          "PRAGMA pager_ota_mode=1;"
          "PRAGMA ota_mode=1;"
          "BEGIN IMMEDIATE;"
          ;
        p->rc = sqlite3_exec(p->db, zScript, 0, 0, &p->zErrmsg);
      }

      /* Point the object iterator at the first object */
      if( p->rc==SQLITE_OK ){
        p->rc = otaObjIterFirst(p, &p->objiter);
      }

      if( p->rc==SQLITE_OK ){

        otaLoadTransactionState(p, pState);
      }
    }else if( p->rc==SQLITE_OK && p->eStage==OTA_STAGE_CKPT ){
      p->rc = sqlite3_ckpt_open(
          p->db, pState->pCkptState, pState->nCkptState, &p->pCkpt
      );
    }

    otaFreeState(pState);
  }

  return p;
}

/*
** Close the OTA handle.
*/
int sqlite3ota_close(sqlite3ota *p, char **pzErrmsg){
  int rc;
  if( p ){


    /* If the update has not been fully applied, save the state in 
    ** the ota db. If successful, this call also commits the open 
    ** transaction on the ota db. */
    assert( p->rc!=SQLITE_ROW );
    if( p->rc==SQLITE_OK || p->rc==SQLITE_DONE ){
      assert( p->zErrmsg==0 );
      otaSaveTransactionState(p);
    }

    /* Close any open statement handles. */
    otaObjIterFinalize(&p->objiter);

    /* Commit the transaction to the *-oal file. */
    if( p->rc==SQLITE_OK && p->eStage==OTA_STAGE_OAL ){
      p->rc = sqlite3_exec(p->db, "COMMIT", 0, 0, &p->zErrmsg);

    }

    if( p->rc==SQLITE_OK && p->eStage==OTA_STAGE_CKPT ){
      p->rc = sqlite3_exec(p->db, "PRAGMA pager_ota_mode=2", 0, 0, &p->zErrmsg);

    }





    /* Close the open database handle */
    if( p->pCkpt ) sqlite3_ckpt_close(p->pCkpt, 0, 0);
    sqlite3_close(p->db);

    rc = p->rc;
    *pzErrmsg = p->zErrmsg;
    sqlite3_free(p);
  }else{
    rc = SQLITE_NOMEM;
    *pzErrmsg = 0;
  }

Changes to src/main.c.

1763
1764
1765
1766
1767
1768
1769


















1770
1771
1772
1773
1774
1775
1776
    sqlite3Error(db, rc);
  }
  rc = sqlite3ApiExit(db, rc);
  sqlite3_mutex_leave(db->mutex);
  return rc;
#endif
}




















/*
** Checkpoint database zDb. If zDb is NULL, or if the buffer zDb points
** to contains a zero-length string, all attached databases are 
** checkpointed.
*/







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







1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
    sqlite3Error(db, rc);
  }
  rc = sqlite3ApiExit(db, rc);
  sqlite3_mutex_leave(db->mutex);
  return rc;
#endif
}

int sqlite3_ckpt_open(
  sqlite3 *db, 
  unsigned char *a, int n, 
  sqlite3_ckpt **ppCkpt
){
  Pager *pPager;
  Btree *pBt;
  int rc;

  *ppCkpt = 0;
  sqlite3_mutex_enter(db->mutex);
  pBt = db->aDb[0].pBt;
  pPager = sqlite3BtreePager(pBt);
  rc = sqlite3PagerWalCheckpointStart(db, pPager, a, n, ppCkpt);
  sqlite3_mutex_leave(db->mutex);
  return rc;
}


/*
** Checkpoint database zDb. If zDb is NULL, or if the buffer zDb points
** to contains a zero-length string, all attached databases are 
** checkpointed.
*/

Changes to src/pager.c.

619
620
621
622
623
624
625


626
627
628
629
630
631
632
....
7284
7285
7286
7287
7288
7289
7290
7291
7292
7293
7294
7295
7296
7297
7298












7299
** otaMode
**   This variable is normally 0. It is set to 1 by the PagerSetOtaMode()
**   function - as a result of a "PRAGMA pager_ota_mode=1" command. Once 
**   the *-oal file has been opened and it has been determined that the 
**   database file has not been modified since it was created, this variable 
**   is set to 2.
**


**
*/
struct Pager {
  sqlite3_vfs *pVfs;          /* OS functions to use for IO */
  u8 exclusiveMode;           /* Boolean. True if locking_mode==EXCLUSIVE */
  u8 journalMode;             /* One of the PAGER_JOURNALMODE_* values */
  u8 useJournal;              /* Use a rollback journal on this file */
................................................................................
  return sqlite3WalFramesize(pPager->pWal);
}
#endif

/*
** Set or clear the "OTA mode" flag.
*/
int sqlite3PagerSetOtaMode(Pager *pPager, int bOta){
  if( pPager->pWal || pPager->eState!=PAGER_OPEN ){
    return SQLITE_ERROR;
  }
  pPager->otaMode = 1;
  return SQLITE_OK;
}













#endif /* SQLITE_OMIT_DISKIO */







>
>







 







|
|


|



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

619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
....
7286
7287
7288
7289
7290
7291
7292
7293
7294
7295
7296
7297
7298
7299
7300
7301
7302
7303
7304
7305
7306
7307
7308
7309
7310
7311
7312
7313
** otaMode
**   This variable is normally 0. It is set to 1 by the PagerSetOtaMode()
**   function - as a result of a "PRAGMA pager_ota_mode=1" command. Once 
**   the *-oal file has been opened and it has been determined that the 
**   database file has not been modified since it was created, this variable 
**   is set to 2.
**
** noCkptOnClose
**
**
*/
struct Pager {
  sqlite3_vfs *pVfs;          /* OS functions to use for IO */
  u8 exclusiveMode;           /* Boolean. True if locking_mode==EXCLUSIVE */
  u8 journalMode;             /* One of the PAGER_JOURNALMODE_* values */
  u8 useJournal;              /* Use a rollback journal on this file */
................................................................................
  return sqlite3WalFramesize(pPager->pWal);
}
#endif

/*
** Set or clear the "OTA mode" flag.
*/
int sqlite3PagerSetOtaMode(Pager *pPager, int iOta){
  if( iOta==1 && (pPager->pWal || pPager->eState!=PAGER_OPEN) ){
    return SQLITE_ERROR;
  }
  pPager->otaMode = iOta;
  return SQLITE_OK;
}

int sqlite3PagerWalCheckpointStart(
  sqlite3 *db, 
  Pager *pPager,
  u8 *a, int n, 
  sqlite3_ckpt **ppCkpt
){
  return sqlite3WalCheckpointStart(db, pPager->pWal, a, n,
      pPager->xBusyHandler, pPager->pBusyHandlerArg,
      pPager->ckptSyncFlags, ppCkpt
  );
}

#endif /* SQLITE_OMIT_DISKIO */

Changes to src/pager.h.

205
206
207
208
209
210
211

212
213
#else
# define disable_simulated_io_errors()
# define enable_simulated_io_errors()
#endif

int sqlite3PagerSetOtaMode(Pager *pPager, int bOta);
void sqlite3PagerWalSalt(Pager *pPager, u32 *aSalt);


#endif /* _PAGER_H_ */







>


205
206
207
208
209
210
211
212
213
214
#else
# define disable_simulated_io_errors()
# define enable_simulated_io_errors()
#endif

int sqlite3PagerSetOtaMode(Pager *pPager, int bOta);
void sqlite3PagerWalSalt(Pager *pPager, u32 *aSalt);
int sqlite3PagerWalCheckpointStart(sqlite3*, Pager*, u8*, int, sqlite3_ckpt**);

#endif /* _PAGER_H_ */

Changes to src/pragma.c.

896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
  ** Other clients see a rollback-mode database on which the pager_ota_mode
  ** client is holding a SHARED lock.
  */
  case PragTyp_PAGER_OTA_MODE: {
    Btree *pBt = pDb->pBt;
    assert( pBt!=0 );
    if( zRight ){
      int iArg = !!sqlite3Atoi(zRight);
      if( sqlite3BtreeIsInReadTrans(pBt) ){
        sqlite3ErrorMsg(pParse, 
            "cannot set pager_ota_mode with open transaction"
        );
      }else if( sqlite3PagerSetOtaMode(sqlite3BtreePager(pBt), iArg) ){
        sqlite3ErrorMsg(pParse, "cannot set pager_ota_mode in wal mode");
      }







|







896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
  ** Other clients see a rollback-mode database on which the pager_ota_mode
  ** client is holding a SHARED lock.
  */
  case PragTyp_PAGER_OTA_MODE: {
    Btree *pBt = pDb->pBt;
    assert( pBt!=0 );
    if( zRight ){
      int iArg = sqlite3Atoi(zRight);
      if( sqlite3BtreeIsInReadTrans(pBt) ){
        sqlite3ErrorMsg(pParse, 
            "cannot set pager_ota_mode with open transaction"
        );
      }else if( sqlite3PagerSetOtaMode(sqlite3BtreePager(pBt), iArg) ){
        sqlite3ErrorMsg(pParse, "cannot set pager_ota_mode in wal mode");
      }

Changes to src/sqlite.h.in.

7460
7461
7462
7463
7464
7465
7466


































7467
7468
7469
7470
7471
7472
7473
7474
7475
7476
7477
7478
  int bDelete,                    /* Zero for insert, non-zero for delete */
  const char *zIndex,             /* Index to write to */
  sqlite3_stmt**,                 /* OUT: New statement handle */
  const char ***pazColl,          /* OUT: Collation sequences for each column */
  int **paiCol, int *pnCol        /* OUT: See above */
);



































/*
** Undo the hack that converts floating point types to integer for
** builds on processors without floating point support.
*/
#ifdef SQLITE_OMIT_FLOATING_POINT
# undef double
#endif

#ifdef __cplusplus
}  /* End of the 'extern "C"' block */
#endif
#endif /* _SQLITE3_H_ */







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












7460
7461
7462
7463
7464
7465
7466
7467
7468
7469
7470
7471
7472
7473
7474
7475
7476
7477
7478
7479
7480
7481
7482
7483
7484
7485
7486
7487
7488
7489
7490
7491
7492
7493
7494
7495
7496
7497
7498
7499
7500
7501
7502
7503
7504
7505
7506
7507
7508
7509
7510
7511
7512
  int bDelete,                    /* Zero for insert, non-zero for delete */
  const char *zIndex,             /* Index to write to */
  sqlite3_stmt**,                 /* OUT: New statement handle */
  const char ***pazColl,          /* OUT: Collation sequences for each column */
  int **paiCol, int *pnCol        /* OUT: See above */
);

/*
** Incremental checkpoint API.
**
** An incremental checkpoint handle is opened using the sqlite3_ckpt_open()
** API. To begin a new checkpoint, the second and third arguments should both
** be passed zero. To resume an earlier checkpoint, the second and third
** arguments should specify a buffer returned by an earlier call to
** sqlite3_ckpt_close(). When resuming a checkpoint, if the database or WAL 
** file has been modified since the checkpoint was suspended, the 
** sqlite3_ckpt_open() call fails with SQLITE_MISMATCH.
**
** Each time sqlite3_ckpt_step() is called on an open checkpoint handle, a
** single page is copied from the WAL file to the database. If no error 
** occurs, but the checkpoint is not finished, SQLITE_OK is returned. If the
** checkpoint has been finished (and so sqlite3_ckpt_step() should not be
** called again), SQLITE_DONE is returned. Otherwise, if an error occurs,
** some other SQLite error code is returned.
**
** Calling sqlite3_ckpt_close() closes an open checkpoint handle. If the
** checkpoint has finished and no error has occurred, SQLITE_OK is returned
** and the two output parameters zeroed. Or, if an error has occurred, an
** error code is returned and the two output parameters are zeroed. Finally,
** if the checkpoint is not finished but no error has occurred, SQLITE_OK is
** returned and the first output variable set to point to a buffer allocated 
** using sqlite3_malloc() containing the serialized state of the checkpoint. 
** The contents of this buffer may be passed to a later call to
** sqlite3_ckpt_open() to restart the checkpoint. The second output variable 
** is set to the size of the buffer in bytes.
*/
typedef struct sqlite3_ckpt sqlite3_ckpt;
int sqlite3_ckpt_open(sqlite3*, unsigned char *a, int n, sqlite3_ckpt **ppCkpt);
int sqlite3_ckpt_step(sqlite3_ckpt*);
int sqlite3_ckpt_close(sqlite3_ckpt*, unsigned char **pa, int *pn);

/*
** Undo the hack that converts floating point types to integer for
** builds on processors without floating point support.
*/
#ifdef SQLITE_OMIT_FLOATING_POINT
# undef double
#endif

#ifdef __cplusplus
}  /* End of the 'extern "C"' block */
#endif
#endif /* _SQLITE3_H_ */

Changes to src/wal.c.

478
479
480
481
482
483
484


















485
486
487
488
489
490
491
....
1618
1619
1620
1621
1622
1623
1624





















































































































































1625
1626
1627
1628
1629
1630
1631
....
1656
1657
1658
1659
1660
1661
1662

1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674

1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709

1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753

1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774

1775

1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
....
2967
2968
2969
2970
2971
2972
2973
2974
2975
2976
2977
2978
2979
2980
2981
2982
2983
2984
2985
....
2997
2998
2999
3000
3001
3002
3003






















































































































3004
3005
3006
3007
3008
3009
3010
    ht_slot *aIndex;              /* i0, i1, i2... such that aPgno[iN] ascend */
    u32 *aPgno;                   /* Array of page numbers. */
    int nEntry;                   /* Nr. of entries in aPgno[] and aIndex[] */
    int iZero;                    /* Frame number associated with aPgno[0] */
  } aSegment[1];                  /* One for every 32KB page in the wal-index */
};



















/*
** Define the parameters of the hash tables in the wal-index file. There
** is a hash-table following every HASHTABLE_NPAGE page numbers in the
** wal-index.
**
** Changing any of these constants will alter the wal-index format and
** create incompatibilities.
................................................................................
/*
** The cache of the wal-index header must be valid to call this function.
** Return the page-size in bytes used by the database.
*/
static int walPagesize(Wal *pWal){
  return (pWal->hdr.szPage&0xfe00) + ((pWal->hdr.szPage&0x0001)<<16);
}






















































































































































/*
** Copy as much content as we can from the WAL back into the database file
** in response to an sqlite3_wal_checkpoint() request or the equivalent.
**
** The amount of information copies from WAL to database might be limited
** by active readers.  This routine will never overwrite a database page
................................................................................
*/
static int walCheckpoint(
  Wal *pWal,                      /* Wal connection */
  int eMode,                      /* One of PASSIVE, FULL or RESTART */
  int (*xBusyCall)(void*),        /* Function to call when busy */
  void *pBusyArg,                 /* Context argument for xBusyHandler */
  int sync_flags,                 /* Flags for OsSync() (or 0) */

  u8 *zBuf                        /* Temporary buffer to use */
){
  int rc;                         /* Return code */
  int szPage;                     /* Database page-size */
  WalIterator *pIter = 0;         /* Wal iterator context */
  u32 iDbpage = 0;                /* Next database page to write */
  u32 iFrame = 0;                 /* Wal frame containing data for iDbpage */
  u32 mxSafeFrame;                /* Max frame that can be backfilled */
  u32 mxPage;                     /* Max database page to write */
  int i;                          /* Loop counter */
  volatile WalCkptInfo *pInfo;    /* The checkpoint status information */
  int (*xBusy)(void*) = 0;        /* Function to call when waiting for locks */


  szPage = walPagesize(pWal);
  testcase( szPage<=32768 );
  testcase( szPage>=65536 );
  pInfo = walCkptInfo(pWal);
  if( pInfo->nBackfill>=pWal->hdr.mxFrame ) return SQLITE_OK;

  /* Allocate the iterator */
  rc = walIteratorInit(pWal, &pIter);
  if( rc!=SQLITE_OK ){
    return rc;
  }
  assert( pIter );

  if( eMode!=SQLITE_CHECKPOINT_PASSIVE ) xBusy = xBusyCall;

  /* Compute in mxSafeFrame the index of the last frame of the WAL that is
  ** safe to write into the database.  Frames beyond mxSafeFrame might
  ** overwrite database pages that are in use by active readers and thus
  ** cannot be backfilled from the WAL.
  */
  mxSafeFrame = pWal->hdr.mxFrame;
  mxPage = pWal->hdr.nPage;
  for(i=1; i<WAL_NREADER; i++){
    u32 y = pInfo->aReadMark[i];
    if( mxSafeFrame>y ){
      assert( y<=pWal->hdr.mxFrame );
      rc = walBusyLock(pWal, xBusy, pBusyArg, WAL_READ_LOCK(i), 1);
      if( rc==SQLITE_OK ){
        pInfo->aReadMark[i] = (i==1 ? mxSafeFrame : READMARK_NOT_USED);
        walUnlockExclusive(pWal, WAL_READ_LOCK(i), 1);
      }else if( rc==SQLITE_BUSY ){
        mxSafeFrame = y;
        xBusy = 0;
      }else{

        goto walcheckpoint_out;
      }
    }
  }

  if( pInfo->nBackfill<mxSafeFrame
   && (rc = walBusyLock(pWal, xBusy, pBusyArg, WAL_READ_LOCK(0), 1))==SQLITE_OK
  ){
    i64 nSize;                    /* Current size of database file */
    u32 nBackfill = pInfo->nBackfill;

    /* Sync the WAL to disk */
    if( sync_flags ){
      rc = sqlite3OsSync(pWal->pWalFd, sync_flags);
    }

    /* If the database may grow as a result of this checkpoint, hint
    ** about the eventual size of the db file to the VFS layer.
    */
    if( rc==SQLITE_OK ){
      i64 nReq = ((i64)mxPage * szPage);
      rc = sqlite3OsFileSize(pWal->pDbFd, &nSize);
      if( rc==SQLITE_OK && nSize<nReq ){
        sqlite3OsFileControlHint(pWal->pDbFd, SQLITE_FCNTL_SIZE_HINT, &nReq);
      }
    }


    /* Iterate through the contents of the WAL, copying data to the db file. */
    while( rc==SQLITE_OK && 0==walIteratorNext(pIter, &iDbpage, &iFrame) ){
      i64 iOffset;
      assert( walFramePgno(pWal, iFrame)==iDbpage );
      if( iFrame<=nBackfill || iFrame>mxSafeFrame || iDbpage>mxPage ) continue;
      iOffset = walFrameOffset(iFrame, szPage) + WAL_FRAME_HDRSIZE;
      /* testcase( IS_BIG_INT(iOffset) ); // requires a 4GiB WAL file */
      rc = sqlite3OsRead(pWal->pWalFd, zBuf, szPage, iOffset);
      if( rc!=SQLITE_OK ) break;
      iOffset = (iDbpage-1)*(i64)szPage;
      testcase( IS_BIG_INT(iOffset) );
      rc = sqlite3OsWrite(pWal->pDbFd, zBuf, szPage, iOffset);
      if( rc!=SQLITE_OK ) break;
    }

    /* If work was actually accomplished... */

    if( rc==SQLITE_OK ){
      if( mxSafeFrame==walIndexHdr(pWal)->mxFrame ){
        i64 szDb = pWal->hdr.nPage*(i64)szPage;
        testcase( IS_BIG_INT(szDb) );
        rc = sqlite3OsTruncate(pWal->pDbFd, szDb);
        if( rc==SQLITE_OK && sync_flags ){
          rc = sqlite3OsSync(pWal->pDbFd, sync_flags);
        }
      }
      if( rc==SQLITE_OK ){
        pInfo->nBackfill = mxSafeFrame;
      }
    }

    /* Release the reader lock held while backfilling */
    walUnlockExclusive(pWal, WAL_READ_LOCK(0), 1);
  }

  if( rc==SQLITE_BUSY ){
    /* Reset the return code so as not to report a checkpoint failure
    ** just because there are active readers.  */

    rc = SQLITE_OK;

  }

  /* If this is an SQLITE_CHECKPOINT_RESTART operation, and the entire wal
  ** file has been copied into the database file, then block until all
  ** readers have finished using the wal file. This ensures that the next
  ** process to write to the database restarts the wal file.
  */
  if( rc==SQLITE_OK && eMode!=SQLITE_CHECKPOINT_PASSIVE ){
    assert( pWal->writeLock );
    if( pInfo->nBackfill<pWal->hdr.mxFrame ){
      rc = SQLITE_BUSY;
    }else if( eMode==SQLITE_CHECKPOINT_RESTART ){
      assert( mxSafeFrame==pWal->hdr.mxFrame );
      rc = walBusyLock(pWal, xBusy, pBusyArg, WAL_READ_LOCK(1), WAL_NREADER-1);
      if( rc==SQLITE_OK ){
        walUnlockExclusive(pWal, WAL_READ_LOCK(1), WAL_NREADER-1);
      }
    }
  }

 walcheckpoint_out:
  walIteratorFree(pIter);
  return rc;
}

/*
** If the WAL file is currently larger than nMax bytes in size, truncate
** it to exactly nMax bytes. If an error occurs while doing so, ignore it.
*/
................................................................................
    if( isChanged && pWal->pDbFd->pMethods->iVersion>=3 ){
      sqlite3OsUnfetch(pWal->pDbFd, 0, 0);
    }
  }

  /* Copy data from the log to the database file. */
  if( rc==SQLITE_OK ){
    if( pWal->hdr.mxFrame && walPagesize(pWal)!=nBuf ){
      rc = SQLITE_CORRUPT_BKPT;
    }else{
      rc = walCheckpoint(pWal, eMode2, xBusy, pBusyArg, sync_flags, zBuf);
    }

    /* If no error occurred, set the output variables. */
    if( rc==SQLITE_OK || rc==SQLITE_BUSY ){
      if( pnLog ) *pnLog = (int)pWal->hdr.mxFrame;
      if( pnCkpt ) *pnCkpt = (int)(walCkptInfo(pWal)->nBackfill);
    }
  }
................................................................................
  /* Release the locks. */
  sqlite3WalEndWriteTransaction(pWal);
  walUnlockExclusive(pWal, WAL_CKPT_LOCK, 1);
  pWal->ckptLock = 0;
  WALTRACE(("WAL%p: checkpoint %s\n", pWal, rc ? "failed" : "ok"));
  return (rc==SQLITE_OK && eMode!=eMode2 ? SQLITE_BUSY : rc);
}























































































































/* Return the value to pass to a sqlite3_wal_hook callback, the
** number of frames in the WAL at the point of the last commit since
** sqlite3WalCallback() was called.  If no commits have occurred since
** the last call, then return 0.
*/
int sqlite3WalCallback(Wal *pWal){







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







 







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







 







>
|


<
<
<
<
<
<
<
<

>

<
<
<
<
<
<
<
<
<
<
<
<
<

<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
>
|
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
|
<
<
<
<
|
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
>
|
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
>
|
>
|
<







|


|








|







 







<
<
<
|
<







 







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







478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
....
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
....
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833








1834
1835
1836













1837




















1838
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
....
3037
3038
3039
3040
3041
3042
3043



3044

3045
3046
3047
3048
3049
3050
3051
....
3063
3064
3065
3066
3067
3068
3069
3070
3071
3072
3073
3074
3075
3076
3077
3078
3079
3080
3081
3082
3083
3084
3085
3086
3087
3088
3089
3090
3091
3092
3093
3094
3095
3096
3097
3098
3099
3100
3101
3102
3103
3104
3105
3106
3107
3108
3109
3110
3111
3112
3113
3114
3115
3116
3117
3118
3119
3120
3121
3122
3123
3124
3125
3126
3127
3128
3129
3130
3131
3132
3133
3134
3135
3136
3137
3138
3139
3140
3141
3142
3143
3144
3145
3146
3147
3148
3149
3150
3151
3152
3153
3154
3155
3156
3157
3158
3159
3160
3161
3162
3163
3164
3165
3166
3167
3168
3169
3170
3171
3172
3173
3174
3175
3176
3177
3178
3179
3180
3181
3182
3183
3184
3185
3186
3187
3188
3189
3190
3191
3192
3193
3194
    ht_slot *aIndex;              /* i0, i1, i2... such that aPgno[iN] ascend */
    u32 *aPgno;                   /* Array of page numbers. */
    int nEntry;                   /* Nr. of entries in aPgno[] and aIndex[] */
    int iZero;                    /* Frame number associated with aPgno[0] */
  } aSegment[1];                  /* One for every 32KB page in the wal-index */
};

/*
** walCheckpoint
*/
typedef struct WalCkpt WalCkpt;
struct WalCkpt {
  sqlite3 *db;                    /* Database pointer (incremental only) */
  int szPage;                     /* Database page-size */
  int sync_flags;                 /* Flags for OsSync() (or 0) */
  u32 mxSafeFrame;                /* Max frame that can be backfilled */
  u32 mxPage;                     /* Max database page to write */
  volatile WalCkptInfo *pInfo;    /* The checkpoint status information */
  WalIterator *pIter;             /* Wal iterator context */
  Wal *pWal;                      /* Pointer to owner object */                 
  u8 *aBuf;                       /* Temporary page-sized buffer to use */
  int rc;                         /* Error code. SQLITE_DONE -> finished */
  int nStep;                      /* Number of times pIter has been stepped */
};

/*
** Define the parameters of the hash tables in the wal-index file. There
** is a hash-table following every HASHTABLE_NPAGE page numbers in the
** wal-index.
**
** Changing any of these constants will alter the wal-index format and
** create incompatibilities.
................................................................................
/*
** The cache of the wal-index header must be valid to call this function.
** Return the page-size in bytes used by the database.
*/
static int walPagesize(Wal *pWal){
  return (pWal->hdr.szPage&0xfe00) + ((pWal->hdr.szPage&0x0001)<<16);
}

static int walCheckpointStart(
  Wal *pWal, 
  u8 *aBuf,                       /* Page-sized temporary buffer */
  int nBuf,                       /* Size of aBuf[] in bytes */
  int (*xBusy)(void*),            /* Function to call when busy (or NULL) */
  void *pBusyArg,                 /* Context argument for xBusyHandler */
  int sync_flags,                 /* Flags for OsSync() (or 0) */
  WalCkpt *p                      /* Allocated object to populate */
){
  int rc;                         /* Return code */
  int i;                          /* Iterator variable */

  memset(p, 0, sizeof(WalCkpt));

  if( pWal->hdr.mxFrame && walPagesize(pWal)!=nBuf ){
    return SQLITE_CORRUPT_BKPT;
  }

  p->szPage = walPagesize(pWal);
  p->pWal = pWal;
  p->aBuf = aBuf;
  p->sync_flags = sync_flags;
  testcase( p->szPage<=32768 );
  testcase( p->szPage>=65536 );
  p->pInfo = walCkptInfo(pWal);
  if( p->pInfo->nBackfill>=pWal->hdr.mxFrame ) return SQLITE_OK;

  /* Allocate the iterator */
  rc = walIteratorInit(pWal, &p->pIter);
  if( rc!=SQLITE_OK ) return rc;
  assert( p->pIter );

  /* Compute in mxSafeFrame the index of the last frame of the WAL that is
  ** safe to write into the database.  Frames beyond mxSafeFrame might
  ** overwrite database pages that are in use by active readers and thus
  ** cannot be backfilled from the WAL.
  */
  p->mxSafeFrame = pWal->hdr.mxFrame;
  p->mxPage = pWal->hdr.nPage;
  for(i=1; i<WAL_NREADER; i++){
    u32 y = p->pInfo->aReadMark[i];
    if( p->mxSafeFrame>y ){
      assert( y<=pWal->hdr.mxFrame );
      rc = walBusyLock(pWal, xBusy, pBusyArg, WAL_READ_LOCK(i), 1);
      if( rc==SQLITE_OK ){
        p->pInfo->aReadMark[i] = (i==1 ? p->mxSafeFrame : READMARK_NOT_USED);
        walUnlockExclusive(pWal, WAL_READ_LOCK(i), 1);
      }else if( rc==SQLITE_BUSY ){
        p->mxSafeFrame = y;
        xBusy = 0;
      }else{
        walIteratorFree(p->pIter);
        p->pIter = 0;
        return rc;
      }
    }
  }

  if( p->pInfo->nBackfill>=p->mxSafeFrame
   || (rc = walBusyLock(pWal, xBusy, pBusyArg, WAL_READ_LOCK(0), 1))!=SQLITE_OK
  ){
    walIteratorFree(p->pIter);
    p->pIter = 0;
  }
  if( rc==SQLITE_BUSY ) rc = SQLITE_OK;

  if( rc==SQLITE_OK && p->pIter ){
    /* Sync the WAL to disk */
    if( sync_flags ){
      rc = sqlite3OsSync(pWal->pWalFd, sync_flags);
    }

    /* If the database may grow as a result of this checkpoint, hint
    ** about the eventual size of the db file to the VFS layer.  */
    if( rc==SQLITE_OK ){
      i64 nSize;                  /* Current size of database file */
      i64 nReq = ((i64)p->mxPage * p->szPage);
      rc = sqlite3OsFileSize(pWal->pDbFd, &nSize);
      if( rc==SQLITE_OK && nSize<nReq ){
        sqlite3OsFileControlHint(pWal->pDbFd, SQLITE_FCNTL_SIZE_HINT, &nReq);
      }
    }
  }

  return rc;
}

static int walCheckpointStep(WalCkpt *p){
  u32 iDbpage = 0;                /* Next database page to write */
  u32 iFrame = 0;                 /* Wal frame containing data for iDbpage */
  int rc = SQLITE_DONE;

  assert( p->rc==SQLITE_OK );
  while( p->pIter && 0==walIteratorNext(p->pIter, &iDbpage, &iFrame) ){
    i64 iOffset;
    assert( walFramePgno(p->pWal, iFrame)==iDbpage );
    p->nStep++;
    if( iFrame<=p->pInfo->nBackfill 
     || iFrame>p->mxSafeFrame 
     || iDbpage>p->mxPage 
    ){
      continue;
    }

    iOffset = walFrameOffset(iFrame, p->szPage) + WAL_FRAME_HDRSIZE;
    /* testcase( IS_BIG_INT(iOffset) ); // requires a 4GiB WAL file */
    rc = sqlite3OsRead(p->pWal->pWalFd, p->aBuf, p->szPage, iOffset);
    if( rc!=SQLITE_OK ) break;
    iOffset = (iDbpage-1)*(i64)p->szPage;
    testcase( IS_BIG_INT(iOffset) );
    rc = sqlite3OsWrite(p->pWal->pDbFd, p->aBuf, p->szPage, iOffset);
    break;
  }

  p->rc = rc;
  return rc;
}

static int walCheckpointFinalize(WalCkpt *p){
  if( p->pIter ){
    int rc = p->rc;
    Wal *pWal = p->pWal;

    /* If work was completed */
    if( p->pIter && rc==SQLITE_DONE ){
      rc = SQLITE_OK;
      if( p->mxSafeFrame==walIndexHdr(pWal)->mxFrame ){
        i64 szDb = pWal->hdr.nPage*(i64)p->szPage;
        testcase( IS_BIG_INT(szDb) );
        rc = sqlite3OsTruncate(pWal->pDbFd, szDb);
        if( rc==SQLITE_OK && p->sync_flags ){
          rc = sqlite3OsSync(pWal->pDbFd, p->sync_flags);
        }
      }
      if( rc==SQLITE_OK ){
        p->pInfo->nBackfill = p->mxSafeFrame;
      }
      p->rc = rc;
    }

    /* Release the reader lock held while backfilling */
    walUnlockExclusive(pWal, WAL_READ_LOCK(0), 1);
    walIteratorFree(p->pIter);
    p->pIter = 0;
  }

  return p->rc;
}

/*
** Copy as much content as we can from the WAL back into the database file
** in response to an sqlite3_wal_checkpoint() request or the equivalent.
**
** The amount of information copies from WAL to database might be limited
** by active readers.  This routine will never overwrite a database page
................................................................................
*/
static int walCheckpoint(
  Wal *pWal,                      /* Wal connection */
  int eMode,                      /* One of PASSIVE, FULL or RESTART */
  int (*xBusyCall)(void*),        /* Function to call when busy */
  void *pBusyArg,                 /* Context argument for xBusyHandler */
  int sync_flags,                 /* Flags for OsSync() (or 0) */
  u8 *zBuf,                       /* Temporary buffer to use */
  int nBuf                        /* Size of zBuf in bytes */
){
  int rc;                         /* Return code */








  int (*xBusy)(void*) = 0;        /* Function to call when waiting for locks */
  WalCkpt sC;














  if( eMode!=SQLITE_CHECKPOINT_PASSIVE ) xBusy = xBusyCall;




















  rc = walCheckpointStart(pWal, zBuf, nBuf, xBusy, pBusyArg, sync_flags, &sC);
  if( sC.pIter==0 ) goto walcheckpoint_out;


















  assert( rc==SQLITE_OK );
























  /* Step the checkpoint object until it reports something other than 
  ** SQLITE_OK.  */




















  while( SQLITE_OK==(rc = walCheckpointStep(&sC)) );
  if( rc==SQLITE_DONE ) rc = SQLITE_OK;
  rc = walCheckpointFinalize(&sC);


  /* If this is an SQLITE_CHECKPOINT_RESTART operation, and the entire wal
  ** file has been copied into the database file, then block until all
  ** readers have finished using the wal file. This ensures that the next
  ** process to write to the database restarts the wal file.
  */
  if( rc==SQLITE_OK && eMode!=SQLITE_CHECKPOINT_PASSIVE ){
    assert( pWal->writeLock );
    if( sC.pInfo->nBackfill<pWal->hdr.mxFrame ){
      rc = SQLITE_BUSY;
    }else if( eMode==SQLITE_CHECKPOINT_RESTART ){
      assert( sC.mxSafeFrame==pWal->hdr.mxFrame );
      rc = walBusyLock(pWal, xBusy, pBusyArg, WAL_READ_LOCK(1), WAL_NREADER-1);
      if( rc==SQLITE_OK ){
        walUnlockExclusive(pWal, WAL_READ_LOCK(1), WAL_NREADER-1);
      }
    }
  }

 walcheckpoint_out:
  walIteratorFree(sC.pIter);
  return rc;
}

/*
** If the WAL file is currently larger than nMax bytes in size, truncate
** it to exactly nMax bytes. If an error occurs while doing so, ignore it.
*/
................................................................................
    if( isChanged && pWal->pDbFd->pMethods->iVersion>=3 ){
      sqlite3OsUnfetch(pWal->pDbFd, 0, 0);
    }
  }

  /* Copy data from the log to the database file. */
  if( rc==SQLITE_OK ){



    rc = walCheckpoint(pWal, eMode2, xBusy, pBusyArg, sync_flags, zBuf, nBuf);


    /* If no error occurred, set the output variables. */
    if( rc==SQLITE_OK || rc==SQLITE_BUSY ){
      if( pnLog ) *pnLog = (int)pWal->hdr.mxFrame;
      if( pnCkpt ) *pnCkpt = (int)(walCkptInfo(pWal)->nBackfill);
    }
  }
................................................................................
  /* Release the locks. */
  sqlite3WalEndWriteTransaction(pWal);
  walUnlockExclusive(pWal, WAL_CKPT_LOCK, 1);
  pWal->ckptLock = 0;
  WALTRACE(("WAL%p: checkpoint %s\n", pWal, rc ? "failed" : "ok"));
  return (rc==SQLITE_OK && eMode!=eMode2 ? SQLITE_BUSY : rc);
}

int sqlite3_ckpt_step(sqlite3_ckpt *pCkpt){
  int rc;
  WalCkpt *p = (WalCkpt*)pCkpt;
  sqlite3_mutex_enter(p->db->mutex);
  rc = walCheckpointStep(p);
  sqlite3_mutex_leave(p->db->mutex);
  return rc;
}

int sqlite3_ckpt_close(sqlite3_ckpt *pCkpt, u8 **paState, int *pnState){
  int rc;
  WalCkpt *p = (WalCkpt*)pCkpt;
  sqlite3 *db = p->db;
  Wal *pWal = p->pWal;
  sqlite3_mutex_enter(db->mutex);
  if( paState ){
    *paState = 0;
    *pnState = 0;
    if( p->rc==SQLITE_OK ){
      u8 *aState = sqlite3_malloc(sizeof(u32) * 3);
      if( aState==0 ){
        p->rc = SQLITE_NOMEM;
      }else{
        *pnState = sizeof(u32)*3;
        sqlite3Put4byte(&aState[0], p->nStep);
        sqlite3Put4byte(&aState[4], p->pWal->hdr.aCksum[0]);
        sqlite3Put4byte(&aState[8], p->pWal->hdr.aCksum[1]);
        *paState = aState;
      }
    }
  }
  rc = walCheckpointFinalize(p);
  walUnlockExclusive(pWal, WAL_CKPT_LOCK, 1);
  pWal->ckptLock = 0;
  sqlite3_free(p);
  memset(&pWal->hdr, 0, sizeof(WalIndexHdr));
  sqlite3_mutex_leave(db->mutex);
  return rc;
}

int sqlite3WalCheckpointStart(
  sqlite3 *db,                    /* Database connection */
  Wal *pWal,                      /* Wal connection */
  u8 *aState, int nState,         /* Checkpoint state to restore */
  int (*xBusy)(void*),            /* Function to call when busy */
  void *pBusyArg,                 /* Context argument for xBusyHandler */
  int sync_flags,                 /* Flags to sync db file with (or 0) */
  sqlite3_ckpt **ppCkpt           /* OUT: Incremental checkpoint object */
){
  WalCkpt *p = 0;
  int isChanged = 0;
  int rc;
  int pgsz;

  *ppCkpt = 0;
  if( pWal->readOnly ) return SQLITE_READONLY;
  WALTRACE(("WAL%p: checkpoint begins\n", pWal));
  rc = walLockExclusive(pWal, WAL_CKPT_LOCK, 1);
  if( rc ){
    /* Usually this is SQLITE_BUSY meaning that another thread or process
    ** is already running a checkpoint, or maybe a recovery.  But it might
    ** also be SQLITE_IOERR. */
    return rc;
  }
  pWal->ckptLock = 1;

  /* Read the wal-index header. */
  rc = walIndexReadHdr(pWal, &isChanged);
  if( rc!=SQLITE_OK ) goto ckptstart_out;
  if( isChanged && pWal->pDbFd->pMethods->iVersion>=3 ){
    sqlite3OsUnfetch(pWal->pDbFd, 0, 0);
  }

  pgsz = walPagesize(pWal);
  p = sqlite3_malloc(sizeof(WalCkpt) + pgsz);
  if( p==0 ){
    rc = SQLITE_NOMEM;
    goto ckptstart_out;
  }

  rc = walCheckpointStart(
      pWal, (u8*)&p[1], pgsz, xBusy, pBusyArg, sync_flags, p
  );
  p->db = db;

  if( rc==SQLITE_OK && aState ){
    if( nState!=sizeof(u32)*3 ){
      rc = SQLITE_CORRUPT_BKPT;
    }else{
      int i;
      if( pWal->hdr.aCksum[0]!=sqlite3Get4byte(&aState[4])
       || pWal->hdr.aCksum[1]!=sqlite3Get4byte(&aState[8])
      ){
        rc = SQLITE_MISMATCH;
      }else{
        p->nStep = (int)sqlite3Get4byte(aState);
        sqlite3Put4byte(&aState[4], pWal->hdr.aCksum[0]);
        sqlite3Put4byte(&aState[8], pWal->hdr.aCksum[1]);
        for(i=0; rc==SQLITE_OK && i<p->nStep; i++){
          u32 dummy1, dummy2; 
          rc = walIteratorNext(p->pIter, &dummy1, &dummy2);
        }
      }
    }
  }

 ckptstart_out:
  if( rc!=SQLITE_OK ){
    if( p ) walIteratorFree(p->pIter);
    walUnlockExclusive(pWal, WAL_CKPT_LOCK, 1);
    pWal->ckptLock = 0;
    sqlite3_free(p);
    p = 0;
  }
  *ppCkpt = (sqlite3_ckpt*)p;
  return rc;
}

/* Return the value to pass to a sqlite3_wal_hook callback, the
** number of frames in the WAL at the point of the last commit since
** sqlite3WalCallback() was called.  If no commits have occurred since
** the last call, then return 0.
*/
int sqlite3WalCallback(Wal *pWal){

Changes to src/wal.h.

123
124
125
126
127
128
129









130
131
132
133
134
135
136
137
138
139
/* Return true if the argument is non-NULL and the WAL module is using
** heap-memory for the wal-index. Otherwise, if the argument is NULL or the
** WAL module is using shared-memory, return false. 
*/
int sqlite3WalHeapMemory(Wal *pWal);

int sqlite3WalCheckSalt(Wal *pWal, sqlite3_file*);










#ifdef SQLITE_ENABLE_ZIPVFS
/* If the WAL file is not empty, return the number of bytes of content
** stored in each frame (i.e. the db page-size when the WAL was created).
*/
int sqlite3WalFramesize(Wal *pWal);
#endif

#endif /* ifndef SQLITE_OMIT_WAL */
#endif /* _WAL_H_ */







>
>
>
>
>
>
>
>
>










123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
/* Return true if the argument is non-NULL and the WAL module is using
** heap-memory for the wal-index. Otherwise, if the argument is NULL or the
** WAL module is using shared-memory, return false. 
*/
int sqlite3WalHeapMemory(Wal *pWal);

int sqlite3WalCheckSalt(Wal *pWal, sqlite3_file*);

int sqlite3WalCheckpointStart(sqlite3 *,
  Wal *pWal,                      /* Wal connection */
  u8 *aState, int nState,         /* Checkpoint state to restore */
  int (*xBusy)(void*),            /* Function to call when busy */
  void *pBusyArg,                 /* Context argument for xBusyHandler */
  int sync_flags,                 /* Flags to sync db file with (or 0) */
  sqlite3_ckpt **ppCkpt           /* OUT: Incremental checkpoint object */
);

#ifdef SQLITE_ENABLE_ZIPVFS
/* If the WAL file is not empty, return the number of bytes of content
** stored in each frame (i.e. the db page-size when the WAL was created).
*/
int sqlite3WalFramesize(Wal *pWal);
#endif

#endif /* ifndef SQLITE_OMIT_WAL */
#endif /* _WAL_H_ */

Changes to test/wal.test.

371
372
373
374
375
376
377

378
379
380
381
382
383
384
  list [file size test.db] [file size test.db-wal]
} [list 2048 [wal_file_size 3 1024]]

# Execute some transactions in auto-vacuum mode to test database file
# truncation.
#
do_test wal-8.1 {

  reopen_db
  catch { db close }
  forcedelete test.db test.db-wal

  sqlite3 db test.db
  db function blob blob
  execsql {







>







371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
  list [file size test.db] [file size test.db-wal]
} [list 2048 [wal_file_size 3 1024]]

# Execute some transactions in auto-vacuum mode to test database file
# truncation.
#
do_test wal-8.1 {
breakpoint
  reopen_db
  catch { db close }
  forcedelete test.db test.db-wal

  sqlite3 db test.db
  db function blob blob
  execsql {