/ Changes On Branch readonly-wal-recovery
Login
SQLite training in Houston TX on 2019-11-05 (details)
Part of the 2019 Tcl Conference

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

Changes In Branch readonly-wal-recovery Excluding Merge-Ins

This is equivalent to a diff from bb39744f to 486949fc

2017-11-14
19:34
Add the ability to read from read-only WAL-mode database files as long as the -wal and -shm files are present on disk. check-in: 00ec95fc user: drh tags: trunk
03:42
Improvement to a comment. No changes to code. Closed-Leaf check-in: 486949fc user: drh tags: readonly-wal-recovery
2017-11-13
05:51
Remove some branches in walTryBeginRead() that were added by check-in [ce5d13c2de] but became unreachable with the addition of logic in check-in [18b26843] that enabled read-only clients to parse the WAL file into a heap-memory WAL-index, thus guaranteeing that the WAL-index header is always available. check-in: 9c6b38b9 user: drh tags: readonly-wal-recovery
2017-11-01
19:44
Add the checkindex.c extension and the sqlite3_checker utility program used for doing live validation of large databases. check-in: 0c5d18a0 user: drh tags: trunk
07:06
Merge latest trunk changes into this branch. check-in: 985bfc99 user: dan tags: readonly-wal-recovery
2017-10-31
18:15
Merge all recent enhancements from trunk. check-in: 24adf90f user: drh tags: checkindex
14:56
Update the built procedures for the sqlite3_analyzer utility to allow it to be linked with an external sqlite3.o library. Automatically detect a missing dbstat extension and report the compile-time error. check-in: bb39744f user: drh tags: trunk
12:20
Fix redundancies in the makefiles that resulted from moving shell.c over to shell.c.in. check-in: bf09fa68 user: drh tags: trunk

Changes to src/main.c.

1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
      case SQLITE_BUSY_RECOVERY:      zName = "SQLITE_BUSY_RECOVERY";     break;
      case SQLITE_BUSY_SNAPSHOT:      zName = "SQLITE_BUSY_SNAPSHOT";     break;
      case SQLITE_LOCKED:             zName = "SQLITE_LOCKED";            break;
      case SQLITE_LOCKED_SHAREDCACHE: zName = "SQLITE_LOCKED_SHAREDCACHE";break;
      case SQLITE_NOMEM:              zName = "SQLITE_NOMEM";             break;
      case SQLITE_READONLY:           zName = "SQLITE_READONLY";          break;
      case SQLITE_READONLY_RECOVERY:  zName = "SQLITE_READONLY_RECOVERY"; break;
      case SQLITE_READONLY_CANTLOCK:  zName = "SQLITE_READONLY_CANTLOCK"; break;
      case SQLITE_READONLY_ROLLBACK:  zName = "SQLITE_READONLY_ROLLBACK"; break;
      case SQLITE_READONLY_DBMOVED:   zName = "SQLITE_READONLY_DBMOVED";  break;
      case SQLITE_INTERRUPT:          zName = "SQLITE_INTERRUPT";         break;
      case SQLITE_IOERR:              zName = "SQLITE_IOERR";             break;
      case SQLITE_IOERR_READ:         zName = "SQLITE_IOERR_READ";        break;
      case SQLITE_IOERR_SHORT_READ:   zName = "SQLITE_IOERR_SHORT_READ";  break;
      case SQLITE_IOERR_WRITE:        zName = "SQLITE_IOERR_WRITE";       break;







|







1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
      case SQLITE_BUSY_RECOVERY:      zName = "SQLITE_BUSY_RECOVERY";     break;
      case SQLITE_BUSY_SNAPSHOT:      zName = "SQLITE_BUSY_SNAPSHOT";     break;
      case SQLITE_LOCKED:             zName = "SQLITE_LOCKED";            break;
      case SQLITE_LOCKED_SHAREDCACHE: zName = "SQLITE_LOCKED_SHAREDCACHE";break;
      case SQLITE_NOMEM:              zName = "SQLITE_NOMEM";             break;
      case SQLITE_READONLY:           zName = "SQLITE_READONLY";          break;
      case SQLITE_READONLY_RECOVERY:  zName = "SQLITE_READONLY_RECOVERY"; break;
      case SQLITE_READONLY_CANTINIT:  zName = "SQLITE_READONLY_CANTINIT"; break;
      case SQLITE_READONLY_ROLLBACK:  zName = "SQLITE_READONLY_ROLLBACK"; break;
      case SQLITE_READONLY_DBMOVED:   zName = "SQLITE_READONLY_DBMOVED";  break;
      case SQLITE_INTERRUPT:          zName = "SQLITE_INTERRUPT";         break;
      case SQLITE_IOERR:              zName = "SQLITE_IOERR";             break;
      case SQLITE_IOERR_READ:         zName = "SQLITE_IOERR_READ";        break;
      case SQLITE_IOERR_SHORT_READ:   zName = "SQLITE_IOERR_SHORT_READ";  break;
      case SQLITE_IOERR_WRITE:        zName = "SQLITE_IOERR_WRITE";       break;

Changes to src/os_unix.c.

4104
4105
4106
4107
4108
4109
4110

4111
4112
4113
4114
4115
4116
4117
....
4265
4266
4267
4268
4269
4270
4271


























































4272
4273
4274
4275
4276
4277
4278
....
4304
4305
4306
4307
4308
4309
4310
4311
4312
4313
4314
4315
4316
4317
4318
4319
4320
....
4347
4348
4349
4350
4351
4352
4353
4354
4355
4356
4357
4358
4359
4360
4361
4362
4363
4364
4365
4366
4367
4368
4369
4370
4371
4372
4373
4374
4375
4376
4377
4378

4379

4380
4381
4382
4383


4384
4385
4386
4387
4388
4389
4390
4391
4392
4393
4394
4395
4396
4397
4398
4399
4400
4401
4402
4403
4404
4405
4406
4407
4408
4409
4410
4411
....
4421
4422
4423
4424
4425
4426
4427
4428
4429
4430
4431
4432
4433
4434
4435
....
4473
4474
4475
4476
4477
4478
4479





4480
4481
4482
4483
4484
4485
4486
  unixInodeInfo *pInode;     /* unixInodeInfo that owns this SHM node */
  sqlite3_mutex *mutex;      /* Mutex to access this object */
  char *zFilename;           /* Name of the mmapped file */
  int h;                     /* Open file descriptor */
  int szRegion;              /* Size of shared-memory regions */
  u16 nRegion;               /* Size of array apRegion */
  u8 isReadonly;             /* True if read-only */

  char **apRegion;           /* Array of mapped shared-memory regions */
  int nRef;                  /* Number of unixShm objects pointing to this */
  unixShm *pFirst;           /* All unixShm objects pointing to this */
#ifdef SQLITE_DEBUG
  u8 exclMask;               /* Mask of exclusive locks held */
  u8 sharedMask;             /* Mask of shared locks held */
  u8 nextShmId;              /* Next available unixShm.id value */
................................................................................
      robust_close(pFd, p->h, __LINE__);
      p->h = -1;
    }
    p->pInode->pShmNode = 0;
    sqlite3_free(p);
  }
}



























































/*
** Open a shared-memory area associated with open database file pDbFd.  
** This particular implementation uses mmapped files.
**
** The file used to implement shared-memory is in the same directory
** as the open database file and has the same name as the open database
................................................................................
** that no other processes are able to read or write the database.  In
** that case, we do not really need shared memory.  No shared memory
** file is created.  The shared memory will be simulated with heap memory.
*/
static int unixOpenSharedMemory(unixFile *pDbFd){
  struct unixShm *p = 0;          /* The connection to be opened */
  struct unixShmNode *pShmNode;   /* The underlying mmapped file */
  int rc;                         /* Result code */
  unixInodeInfo *pInode;          /* The inode of fd */
  char *zShmFilename;             /* Name of the file used for SHM */
  int nShmFilename;               /* Size of the SHM filename in bytes */

  /* Allocate space for the new unixShm object. */
  p = sqlite3_malloc64( sizeof(*p) );
  if( p==0 ) return SQLITE_NOMEM_BKPT;
  memset(p, 0, sizeof(*p));
  assert( pDbFd->pShm==0 );
................................................................................
#endif
    pShmNode = sqlite3_malloc64( sizeof(*pShmNode) + nShmFilename );
    if( pShmNode==0 ){
      rc = SQLITE_NOMEM_BKPT;
      goto shm_open_err;
    }
    memset(pShmNode, 0, sizeof(*pShmNode)+nShmFilename);
    zShmFilename = pShmNode->zFilename = (char*)&pShmNode[1];
#ifdef SQLITE_SHM_DIRECTORY
    sqlite3_snprintf(nShmFilename, zShmFilename, 
                     SQLITE_SHM_DIRECTORY "/sqlite-shm-%x-%x",
                     (u32)sStat.st_ino, (u32)sStat.st_dev);
#else
    sqlite3_snprintf(nShmFilename, zShmFilename, "%s-shm", zBasePath);
    sqlite3FileSuffix3(pDbFd->zPath, zShmFilename);
#endif
    pShmNode->h = -1;
    pDbFd->pInode->pShmNode = pShmNode;
    pShmNode->pInode = pDbFd->pInode;
    if( sqlite3GlobalConfig.bCoreMutex ){
      pShmNode->mutex = sqlite3_mutex_alloc(SQLITE_MUTEX_FAST);
      if( pShmNode->mutex==0 ){
        rc = SQLITE_NOMEM_BKPT;
        goto shm_open_err;
      }
    }

    if( pInode->bProcessLock==0 ){
      int openFlags = O_RDWR | O_CREAT;
      if( sqlite3_uri_boolean(pDbFd->zPath, "readonly_shm", 0) ){
        openFlags = O_RDONLY;
        pShmNode->isReadonly = 1;

      }

      pShmNode->h = robust_open(zShmFilename, openFlags, (sStat.st_mode&0777));
      if( pShmNode->h<0 ){
        rc = unixLogError(SQLITE_CANTOPEN_BKPT, "open", zShmFilename);
        goto shm_open_err;


      }

      /* If this process is running as root, make sure that the SHM file
      ** is owned by the same user that owns the original database.  Otherwise,
      ** the original owner will not be able to connect.
      */
      robustFchown(pShmNode->h, sStat.st_uid, sStat.st_gid);
  
      /* Check to see if another process is holding the dead-man switch.
      ** If not, truncate the file to zero length. 
      */
      rc = SQLITE_OK;
      if( unixShmSystemLock(pDbFd, F_WRLCK, UNIX_SHM_DMS, 1)==SQLITE_OK ){
        if( robust_ftruncate(pShmNode->h, 0) ){
          rc = unixLogError(SQLITE_IOERR_SHMOPEN, "ftruncate", zShmFilename);
        }
      }
      if( rc==SQLITE_OK ){
        rc = unixShmSystemLock(pDbFd, F_RDLCK, UNIX_SHM_DMS, 1);
      }
      if( rc ) goto shm_open_err;
    }
  }

  /* Make the new connection a child of the unixShmNode */
  p->pShmNode = pShmNode;
#ifdef SQLITE_DEBUG
  p->id = pShmNode->nextShmId++;
................................................................................
  ** at pShmNode->pFirst. This must be done while holding the pShmNode->mutex 
  ** mutex.
  */
  sqlite3_mutex_enter(pShmNode->mutex);
  p->pNext = pShmNode->pFirst;
  pShmNode->pFirst = p;
  sqlite3_mutex_leave(pShmNode->mutex);
  return SQLITE_OK;

  /* Jump here on any error */
shm_open_err:
  unixShmPurge(pDbFd);       /* This call frees pShmNode if required */
  sqlite3_free(p);
  unixLeaveMutex();
  return rc;
................................................................................
    rc = unixOpenSharedMemory(pDbFd);
    if( rc!=SQLITE_OK ) return rc;
  }

  p = pDbFd->pShm;
  pShmNode = p->pShmNode;
  sqlite3_mutex_enter(pShmNode->mutex);





  assert( szRegion==pShmNode->szRegion || pShmNode->nRegion==0 );
  assert( pShmNode->pInode==pDbFd->pInode );
  assert( pShmNode->h>=0 || pDbFd->pInode->bProcessLock==1 );
  assert( pShmNode->h<0 || pDbFd->pInode->bProcessLock==0 );

  /* Minimum number of regions required to be mapped. */
  nReqRegion = ((iRegion+nShmPerMap) / nShmPerMap) * nShmPerMap;







>







 







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







 







|

|







 







|

|



|
|













<
|
<
<
>

>
|
|
|
|
>
>







|
|
<
<
<
<
<
<
<
<
<
<
<
|







 







|







 







>
>
>
>
>







4104
4105
4106
4107
4108
4109
4110
4111
4112
4113
4114
4115
4116
4117
4118
....
4266
4267
4268
4269
4270
4271
4272
4273
4274
4275
4276
4277
4278
4279
4280
4281
4282
4283
4284
4285
4286
4287
4288
4289
4290
4291
4292
4293
4294
4295
4296
4297
4298
4299
4300
4301
4302
4303
4304
4305
4306
4307
4308
4309
4310
4311
4312
4313
4314
4315
4316
4317
4318
4319
4320
4321
4322
4323
4324
4325
4326
4327
4328
4329
4330
4331
4332
4333
4334
4335
4336
4337
....
4363
4364
4365
4366
4367
4368
4369
4370
4371
4372
4373
4374
4375
4376
4377
4378
4379
....
4406
4407
4408
4409
4410
4411
4412
4413
4414
4415
4416
4417
4418
4419
4420
4421
4422
4423
4424
4425
4426
4427
4428
4429
4430
4431
4432
4433

4434


4435
4436
4437
4438
4439
4440
4441
4442
4443
4444
4445
4446
4447
4448
4449
4450
4451
4452











4453
4454
4455
4456
4457
4458
4459
4460
....
4470
4471
4472
4473
4474
4475
4476
4477
4478
4479
4480
4481
4482
4483
4484
....
4522
4523
4524
4525
4526
4527
4528
4529
4530
4531
4532
4533
4534
4535
4536
4537
4538
4539
4540
  unixInodeInfo *pInode;     /* unixInodeInfo that owns this SHM node */
  sqlite3_mutex *mutex;      /* Mutex to access this object */
  char *zFilename;           /* Name of the mmapped file */
  int h;                     /* Open file descriptor */
  int szRegion;              /* Size of shared-memory regions */
  u16 nRegion;               /* Size of array apRegion */
  u8 isReadonly;             /* True if read-only */
  u8 isUnlocked;             /* True if no DMS lock held */
  char **apRegion;           /* Array of mapped shared-memory regions */
  int nRef;                  /* Number of unixShm objects pointing to this */
  unixShm *pFirst;           /* All unixShm objects pointing to this */
#ifdef SQLITE_DEBUG
  u8 exclMask;               /* Mask of exclusive locks held */
  u8 sharedMask;             /* Mask of shared locks held */
  u8 nextShmId;              /* Next available unixShm.id value */
................................................................................
      robust_close(pFd, p->h, __LINE__);
      p->h = -1;
    }
    p->pInode->pShmNode = 0;
    sqlite3_free(p);
  }
}

/*
** The DMS lock has not yet been taken on shm file pShmNode. Attempt to
** take it now. Return SQLITE_OK if successful, or an SQLite error
** code otherwise.
**
** If the DMS cannot be locked because this is a readonly_shm=1 
** connection and no other process already holds a lock, return
** SQLITE_READONLY_CANTINIT and set pShmNode->isUnlocked=1.
*/
static int unixLockSharedMemory(unixFile *pDbFd, unixShmNode *pShmNode){
  struct flock lock;
  int rc = SQLITE_OK;

  /* Use F_GETLK to determine the locks other processes are holding
  ** on the DMS byte. If it indicates that another process is holding
  ** a SHARED lock, then this process may also take a SHARED lock
  ** and proceed with opening the *-shm file. 
  **
  ** Or, if no other process is holding any lock, then this process
  ** is the first to open it. In this case take an EXCLUSIVE lock on the
  ** DMS byte and truncate the *-shm file to zero bytes in size. Then
  ** downgrade to a SHARED lock on the DMS byte.
  **
  ** If another process is holding an EXCLUSIVE lock on the DMS byte,
  ** return SQLITE_BUSY to the caller (it will try again). An earlier
  ** version of this code attempted the SHARED lock at this point. But
  ** this introduced a subtle race condition: if the process holding
  ** EXCLUSIVE failed just before truncating the *-shm file, then this
  ** process might open and use the *-shm file without truncating it.
  ** And if the *-shm file has been corrupted by a power failure or
  ** system crash, the database itself may also become corrupt.  */
  lock.l_whence = SEEK_SET;
  lock.l_start = UNIX_SHM_DMS;
  lock.l_len = 1;
  lock.l_type = F_WRLCK;
  if( osFcntl(pShmNode->h, F_GETLK, &lock)!=0 ) {
    rc = SQLITE_IOERR_LOCK;
  }else if( lock.l_type==F_UNLCK ){
    if( pShmNode->isReadonly ){
      pShmNode->isUnlocked = 1;
      rc = SQLITE_READONLY_CANTINIT;
    }else{
      rc = unixShmSystemLock(pDbFd, F_WRLCK, UNIX_SHM_DMS, 1);
      if( rc==SQLITE_OK && robust_ftruncate(pShmNode->h, 0) ){
        rc = unixLogError(SQLITE_IOERR_SHMOPEN,"ftruncate",pShmNode->zFilename);
      }
    }
  }else if( lock.l_type==F_WRLCK ){
    rc = SQLITE_BUSY;
  }

  if( rc==SQLITE_OK ){
    assert( lock.l_type==F_UNLCK || lock.l_type==F_RDLCK );
    rc = unixShmSystemLock(pDbFd, F_RDLCK, UNIX_SHM_DMS, 1);
  }
  return rc;
}

/*
** Open a shared-memory area associated with open database file pDbFd.  
** This particular implementation uses mmapped files.
**
** The file used to implement shared-memory is in the same directory
** as the open database file and has the same name as the open database
................................................................................
** that no other processes are able to read or write the database.  In
** that case, we do not really need shared memory.  No shared memory
** file is created.  The shared memory will be simulated with heap memory.
*/
static int unixOpenSharedMemory(unixFile *pDbFd){
  struct unixShm *p = 0;          /* The connection to be opened */
  struct unixShmNode *pShmNode;   /* The underlying mmapped file */
  int rc = SQLITE_OK;             /* Result code */
  unixInodeInfo *pInode;          /* The inode of fd */
  char *zShm;             /* Name of the file used for SHM */
  int nShmFilename;               /* Size of the SHM filename in bytes */

  /* Allocate space for the new unixShm object. */
  p = sqlite3_malloc64( sizeof(*p) );
  if( p==0 ) return SQLITE_NOMEM_BKPT;
  memset(p, 0, sizeof(*p));
  assert( pDbFd->pShm==0 );
................................................................................
#endif
    pShmNode = sqlite3_malloc64( sizeof(*pShmNode) + nShmFilename );
    if( pShmNode==0 ){
      rc = SQLITE_NOMEM_BKPT;
      goto shm_open_err;
    }
    memset(pShmNode, 0, sizeof(*pShmNode)+nShmFilename);
    zShm = pShmNode->zFilename = (char*)&pShmNode[1];
#ifdef SQLITE_SHM_DIRECTORY
    sqlite3_snprintf(nShmFilename, zShm, 
                     SQLITE_SHM_DIRECTORY "/sqlite-shm-%x-%x",
                     (u32)sStat.st_ino, (u32)sStat.st_dev);
#else
    sqlite3_snprintf(nShmFilename, zShm, "%s-shm", zBasePath);
    sqlite3FileSuffix3(pDbFd->zPath, zShm);
#endif
    pShmNode->h = -1;
    pDbFd->pInode->pShmNode = pShmNode;
    pShmNode->pInode = pDbFd->pInode;
    if( sqlite3GlobalConfig.bCoreMutex ){
      pShmNode->mutex = sqlite3_mutex_alloc(SQLITE_MUTEX_FAST);
      if( pShmNode->mutex==0 ){
        rc = SQLITE_NOMEM_BKPT;
        goto shm_open_err;
      }
    }

    if( pInode->bProcessLock==0 ){

      if( 0==sqlite3_uri_boolean(pDbFd->zPath, "readonly_shm", 0) ){


        pShmNode->h = robust_open(zShm, O_RDWR|O_CREAT, (sStat.st_mode&0777));
      }
      if( pShmNode->h<0 ){
        pShmNode->h = robust_open(zShm, O_RDONLY, (sStat.st_mode&0777));
        if( pShmNode->h<0 ){
          rc = unixLogError(SQLITE_CANTOPEN_BKPT, "open", zShm);
          goto shm_open_err;
        }
        pShmNode->isReadonly = 1;
      }

      /* If this process is running as root, make sure that the SHM file
      ** is owned by the same user that owns the original database.  Otherwise,
      ** the original owner will not be able to connect.
      */
      robustFchown(pShmNode->h, sStat.st_uid, sStat.st_gid);

      rc = unixLockSharedMemory(pDbFd, pShmNode);











      if( rc!=SQLITE_OK && rc!=SQLITE_READONLY_CANTINIT ) goto shm_open_err;
    }
  }

  /* Make the new connection a child of the unixShmNode */
  p->pShmNode = pShmNode;
#ifdef SQLITE_DEBUG
  p->id = pShmNode->nextShmId++;
................................................................................
  ** at pShmNode->pFirst. This must be done while holding the pShmNode->mutex 
  ** mutex.
  */
  sqlite3_mutex_enter(pShmNode->mutex);
  p->pNext = pShmNode->pFirst;
  pShmNode->pFirst = p;
  sqlite3_mutex_leave(pShmNode->mutex);
  return rc;

  /* Jump here on any error */
shm_open_err:
  unixShmPurge(pDbFd);       /* This call frees pShmNode if required */
  sqlite3_free(p);
  unixLeaveMutex();
  return rc;
................................................................................
    rc = unixOpenSharedMemory(pDbFd);
    if( rc!=SQLITE_OK ) return rc;
  }

  p = pDbFd->pShm;
  pShmNode = p->pShmNode;
  sqlite3_mutex_enter(pShmNode->mutex);
  if( pShmNode->isUnlocked ){
    rc = unixLockSharedMemory(pDbFd, pShmNode);
    if( rc!=SQLITE_OK ) goto shmpage_out;
    pShmNode->isUnlocked = 0;
  }
  assert( szRegion==pShmNode->szRegion || pShmNode->nRegion==0 );
  assert( pShmNode->pInode==pDbFd->pInode );
  assert( pShmNode->h>=0 || pDbFd->pInode->bProcessLock==1 );
  assert( pShmNode->h<0 || pDbFd->pInode->bProcessLock==0 );

  /* Minimum number of regions required to be mapped. */
  nReqRegion = ((iRegion+nShmPerMap) / nShmPerMap) * nShmPerMap;

Changes to src/os_win.c.

3669
3670
3671
3672
3673
3674
3675



3676
3677
3678
3679
3680
3681
3682
....
3815
3816
3817
3818
3819
3820
3821































3822
3823
3824
3825
3826
3827
3828
3829
3830
3831
3832
3833

3834
3835
3836
3837
3838
3839
3840
3841
....
3874
3875
3876
3877
3878
3879
3880
3881
3882
3883
3884
3885
3886
3887
3888
3889
3890
3891
3892
3893
3894
3895
3896
3897
3898
3899
3900
3901
3902
3903
3904
3905
3906
3907
3908
3909
3910
3911
....
3920
3921
3922
3923
3924
3925
3926
3927
3928
3929
3930
3931
3932
3933
3934
....
4124
4125
4126
4127
4128
4129
4130


4131
4132
4133
4134
4135
4136
4137
4138
4139
4140





4141
4142
4143
4144
4145
4146
4147
....
4179
4180
4181
4182
4183
4184
4185





4186
4187
4188
4189
4190
4191
4192
4193
4194
4195
4196
4197
4198
4199
4200
4201
4202
4203
4204
4205
4206
4207
4208
4209
4210
4211
4212
4213
4214
4215
4216
4217
4218
4219
4220
4221
4222
....
4239
4240
4241
4242
4243
4244
4245

4246
4247
4248
4249
4250
4251
4252
struct winShmNode {
  sqlite3_mutex *mutex;      /* Mutex to access this object */
  char *zFilename;           /* Name of the file */
  winFile hFile;             /* File handle from winOpen */

  int szRegion;              /* Size of shared-memory regions */
  int nRegion;               /* Size of array apRegion */



  struct ShmRegion {
    HANDLE hMap;             /* File handle from CreateFileMapping */
    void *pMap;
  } *aRegion;
  DWORD lastErrno;           /* The Windows errno from the last I/O error */

  int nRef;                  /* Number of winShm objects pointing to this */
................................................................................
      sqlite3_free(p->aRegion);
      sqlite3_free(p);
    }else{
      pp = &p->pNext;
    }
  }
}
































/*
** Open the shared-memory area associated with database file pDbFd.
**
** When opening a new shared-memory file, if no other instances of that
** file are currently open, in this process or in other processes, then
** the file must be truncated to zero length or have its header cleared.
*/
static int winOpenSharedMemory(winFile *pDbFd){
  struct winShm *p;                  /* The connection to be opened */
  struct winShmNode *pShmNode = 0;   /* The underlying mmapped file */
  int rc;                            /* Result code */

  struct winShmNode *pNew;           /* Newly allocated winShmNode */
  int nName;                         /* Size of zName in bytes */

  assert( pDbFd->pShm==0 );    /* Not previously opened */

  /* Allocate space for the new sqlite3_shm object.  Also speculatively
  ** allocate space for a new winShmNode and filename.
  */
................................................................................
      pShmNode->mutex = sqlite3_mutex_alloc(SQLITE_MUTEX_FAST);
      if( pShmNode->mutex==0 ){
        rc = SQLITE_IOERR_NOMEM_BKPT;
        goto shm_open_err;
      }
    }

    rc = winOpen(pDbFd->pVfs,
                 pShmNode->zFilename,             /* Name of the file (UTF-8) */
                 (sqlite3_file*)&pShmNode->hFile,  /* File handle here */
                 SQLITE_OPEN_WAL | SQLITE_OPEN_READWRITE | SQLITE_OPEN_CREATE,
                 0);
    if( SQLITE_OK!=rc ){
      goto shm_open_err;
    }

    /* Check to see if another process is holding the dead-man switch.
    ** If not, truncate the file to zero length.
    */
    if( winShmSystemLock(pShmNode, WINSHM_WRLCK, WIN_SHM_DMS, 1)==SQLITE_OK ){
      rc = winTruncate((sqlite3_file *)&pShmNode->hFile, 0);
      if( rc!=SQLITE_OK ){
        rc = winLogError(SQLITE_IOERR_SHMOPEN, osGetLastError(),
                         "winOpenShm", pDbFd->zPath);
      }
    }
    if( rc==SQLITE_OK ){
      winShmSystemLock(pShmNode, WINSHM_UNLCK, WIN_SHM_DMS, 1);
      rc = winShmSystemLock(pShmNode, WINSHM_RDLCK, WIN_SHM_DMS, 1);
    }
    if( rc ) goto shm_open_err;
  }

  /* Make the new connection a child of the winShmNode */
  p->pShmNode = pShmNode;
#if defined(SQLITE_DEBUG) || defined(SQLITE_HAVE_OS_TRACE)
  p->id = pShmNode->nextShmId++;
#endif
................................................................................
  ** at pShmNode->pFirst. This must be done while holding the pShmNode->mutex
  ** mutex.
  */
  sqlite3_mutex_enter(pShmNode->mutex);
  p->pNext = pShmNode->pFirst;
  pShmNode->pFirst = p;
  sqlite3_mutex_leave(pShmNode->mutex);
  return SQLITE_OK;

  /* Jump here on any error */
shm_open_err:
  winShmSystemLock(pShmNode, WINSHM_UNLCK, WIN_SHM_DMS, 1);
  winShmPurge(pDbFd->pVfs, 0);      /* This call frees pShmNode if required */
  sqlite3_free(p);
  sqlite3_free(pNew);
................................................................................
  int szRegion,                   /* Size of regions */
  int isWrite,                    /* True to extend file if necessary */
  void volatile **pp              /* OUT: Mapped memory */
){
  winFile *pDbFd = (winFile*)fd;
  winShm *pShm = pDbFd->pShm;
  winShmNode *pShmNode;


  int rc = SQLITE_OK;

  if( !pShm ){
    rc = winOpenSharedMemory(pDbFd);
    if( rc!=SQLITE_OK ) return rc;
    pShm = pDbFd->pShm;
  }
  pShmNode = pShm->pShmNode;

  sqlite3_mutex_enter(pShmNode->mutex);





  assert( szRegion==pShmNode->szRegion || pShmNode->nRegion==0 );

  if( pShmNode->nRegion<=iRegion ){
    struct ShmRegion *apNew;           /* New aRegion[] array */
    int nByte = (iRegion+1)*szRegion;  /* Minimum required file size */
    sqlite3_int64 sz;                  /* Current size of wal-index file */

................................................................................
        pShmNode->aRegion, (iRegion+1)*sizeof(apNew[0])
    );
    if( !apNew ){
      rc = SQLITE_IOERR_NOMEM_BKPT;
      goto shmpage_out;
    }
    pShmNode->aRegion = apNew;






    while( pShmNode->nRegion<=iRegion ){
      HANDLE hMap = NULL;         /* file-mapping handle */
      void *pMap = 0;             /* Mapped memory region */

#if SQLITE_OS_WINRT
      hMap = osCreateFileMappingFromApp(pShmNode->hFile.h,
          NULL, PAGE_READWRITE, nByte, NULL
      );
#elif defined(SQLITE_WIN32_HAS_WIDE)
      hMap = osCreateFileMappingW(pShmNode->hFile.h,
          NULL, PAGE_READWRITE, 0, nByte, NULL
      );
#elif defined(SQLITE_WIN32_HAS_ANSI) && SQLITE_WIN32_CREATEFILEMAPPINGA
      hMap = osCreateFileMappingA(pShmNode->hFile.h,
          NULL, PAGE_READWRITE, 0, nByte, NULL
      );
#endif
      OSTRACE(("SHM-MAP-CREATE pid=%lu, region=%d, size=%d, rc=%s\n",
               osGetCurrentProcessId(), pShmNode->nRegion, nByte,
               hMap ? "ok" : "failed"));
      if( hMap ){
        int iOffset = pShmNode->nRegion*szRegion;
        int iOffsetShift = iOffset % winSysInfo.dwAllocationGranularity;
#if SQLITE_OS_WINRT
        pMap = osMapViewOfFileFromApp(hMap, FILE_MAP_WRITE | FILE_MAP_READ,
            iOffset - iOffsetShift, szRegion + iOffsetShift
        );
#else
        pMap = osMapViewOfFile(hMap, FILE_MAP_WRITE | FILE_MAP_READ,
            0, iOffset - iOffsetShift, szRegion + iOffsetShift
        );
#endif
        OSTRACE(("SHM-MAP-MAP pid=%lu, region=%d, offset=%d, size=%d, rc=%s\n",
                 osGetCurrentProcessId(), pShmNode->nRegion, iOffset,
                 szRegion, pMap ? "ok" : "failed"));
      }
................................................................................
    int iOffset = iRegion*szRegion;
    int iOffsetShift = iOffset % winSysInfo.dwAllocationGranularity;
    char *p = (char *)pShmNode->aRegion[iRegion].pMap;
    *pp = (void *)&p[iOffsetShift];
  }else{
    *pp = 0;
  }

  sqlite3_mutex_leave(pShmNode->mutex);
  return rc;
}

#else
# define winShmMap     0
# define winShmLock    0







>
>
>







 







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










|
|
>
|







 







|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
<







 







|







 







>
>










>
>
>
>
>







 







>
>
>
>
>







|



|



|









|



|







 







>







3669
3670
3671
3672
3673
3674
3675
3676
3677
3678
3679
3680
3681
3682
3683
3684
3685
....
3818
3819
3820
3821
3822
3823
3824
3825
3826
3827
3828
3829
3830
3831
3832
3833
3834
3835
3836
3837
3838
3839
3840
3841
3842
3843
3844
3845
3846
3847
3848
3849
3850
3851
3852
3853
3854
3855
3856
3857
3858
3859
3860
3861
3862
3863
3864
3865
3866
3867
3868
3869
3870
3871
3872
3873
3874
3875
3876
....
3909
3910
3911
3912
3913
3914
3915
3916
3917
3918
3919
3920
3921
3922
3923
3924
3925
3926
3927
3928
3929
3930
3931
3932
3933
3934
3935
3936
3937
3938

3939
3940
3941
3942
3943
3944
3945
....
3954
3955
3956
3957
3958
3959
3960
3961
3962
3963
3964
3965
3966
3967
3968
....
4158
4159
4160
4161
4162
4163
4164
4165
4166
4167
4168
4169
4170
4171
4172
4173
4174
4175
4176
4177
4178
4179
4180
4181
4182
4183
4184
4185
4186
4187
4188
....
4220
4221
4222
4223
4224
4225
4226
4227
4228
4229
4230
4231
4232
4233
4234
4235
4236
4237
4238
4239
4240
4241
4242
4243
4244
4245
4246
4247
4248
4249
4250
4251
4252
4253
4254
4255
4256
4257
4258
4259
4260
4261
4262
4263
4264
4265
4266
4267
4268
....
4285
4286
4287
4288
4289
4290
4291
4292
4293
4294
4295
4296
4297
4298
4299
struct winShmNode {
  sqlite3_mutex *mutex;      /* Mutex to access this object */
  char *zFilename;           /* Name of the file */
  winFile hFile;             /* File handle from winOpen */

  int szRegion;              /* Size of shared-memory regions */
  int nRegion;               /* Size of array apRegion */
  u8 isReadonly;             /* True if read-only */
  u8 isUnlocked;             /* True if no DMS lock held */

  struct ShmRegion {
    HANDLE hMap;             /* File handle from CreateFileMapping */
    void *pMap;
  } *aRegion;
  DWORD lastErrno;           /* The Windows errno from the last I/O error */

  int nRef;                  /* Number of winShm objects pointing to this */
................................................................................
      sqlite3_free(p->aRegion);
      sqlite3_free(p);
    }else{
      pp = &p->pNext;
    }
  }
}

/*
** The DMS lock has not yet been taken on shm file pShmNode. Attempt to
** take it now. Return SQLITE_OK if successful, or an SQLite error
** code otherwise.
**
** If the DMS cannot be locked because this is a readonly_shm=1
** connection and no other process already holds a lock, return
** SQLITE_READONLY_CANTINIT and set pShmNode->isUnlocked=1.
*/
static int winLockSharedMemory(winShmNode *pShmNode){
  int rc = winShmSystemLock(pShmNode, WINSHM_WRLCK, WIN_SHM_DMS, 1);

  if( rc==SQLITE_OK ){
    if( pShmNode->isReadonly ){
      pShmNode->isUnlocked = 1;
      winShmSystemLock(pShmNode, WINSHM_UNLCK, WIN_SHM_DMS, 1);
      return SQLITE_READONLY_CANTINIT;
    }else if( winTruncate((sqlite3_file*)&pShmNode->hFile, 0) ){
      winShmSystemLock(pShmNode, WINSHM_UNLCK, WIN_SHM_DMS, 1);
      return winLogError(SQLITE_IOERR_SHMOPEN, osGetLastError(),
                         "winLockSharedMemory", pShmNode->zFilename);
    }
  }

  if( rc==SQLITE_OK ){
    winShmSystemLock(pShmNode, WINSHM_UNLCK, WIN_SHM_DMS, 1);
  }

  return winShmSystemLock(pShmNode, WINSHM_RDLCK, WIN_SHM_DMS, 1);
}

/*
** Open the shared-memory area associated with database file pDbFd.
**
** When opening a new shared-memory file, if no other instances of that
** file are currently open, in this process or in other processes, then
** the file must be truncated to zero length or have its header cleared.
*/
static int winOpenSharedMemory(winFile *pDbFd){
  struct winShm *p;                  /* The connection to be opened */
  winShmNode *pShmNode = 0;          /* The underlying mmapped file */
  int rc = SQLITE_OK;                /* Result code */
  int rc2 = SQLITE_ERROR;            /* winOpen result code */
  winShmNode *pNew;                  /* Newly allocated winShmNode */
  int nName;                         /* Size of zName in bytes */

  assert( pDbFd->pShm==0 );    /* Not previously opened */

  /* Allocate space for the new sqlite3_shm object.  Also speculatively
  ** allocate space for a new winShmNode and filename.
  */
................................................................................
      pShmNode->mutex = sqlite3_mutex_alloc(SQLITE_MUTEX_FAST);
      if( pShmNode->mutex==0 ){
        rc = SQLITE_IOERR_NOMEM_BKPT;
        goto shm_open_err;
      }
    }

    if( 0==sqlite3_uri_boolean(pDbFd->zPath, "readonly_shm", 0) ){
      rc2 = winOpen(pDbFd->pVfs,
                    pShmNode->zFilename,
                    (sqlite3_file*)&pShmNode->hFile,
                    SQLITE_OPEN_WAL|SQLITE_OPEN_READWRITE|SQLITE_OPEN_CREATE,
                    0);
    }
    if( rc2!=SQLITE_OK ){
      rc2 = winOpen(pDbFd->pVfs,
                    pShmNode->zFilename,
                    (sqlite3_file*)&pShmNode->hFile,
                    SQLITE_OPEN_WAL|SQLITE_OPEN_READONLY,
                    0);
      if( rc2!=SQLITE_OK ){
        rc = winLogError(rc2, osGetLastError(), "winOpenShm",
                         pShmNode->zFilename);
        goto shm_open_err;
      }
      pShmNode->isReadonly = 1;
    }

    rc = winLockSharedMemory(pShmNode);
    if( rc!=SQLITE_OK && rc!=SQLITE_READONLY_CANTINIT ) goto shm_open_err;

  }

  /* Make the new connection a child of the winShmNode */
  p->pShmNode = pShmNode;
#if defined(SQLITE_DEBUG) || defined(SQLITE_HAVE_OS_TRACE)
  p->id = pShmNode->nextShmId++;
#endif
................................................................................
  ** at pShmNode->pFirst. This must be done while holding the pShmNode->mutex
  ** mutex.
  */
  sqlite3_mutex_enter(pShmNode->mutex);
  p->pNext = pShmNode->pFirst;
  pShmNode->pFirst = p;
  sqlite3_mutex_leave(pShmNode->mutex);
  return rc;

  /* Jump here on any error */
shm_open_err:
  winShmSystemLock(pShmNode, WINSHM_UNLCK, WIN_SHM_DMS, 1);
  winShmPurge(pDbFd->pVfs, 0);      /* This call frees pShmNode if required */
  sqlite3_free(p);
  sqlite3_free(pNew);
................................................................................
  int szRegion,                   /* Size of regions */
  int isWrite,                    /* True to extend file if necessary */
  void volatile **pp              /* OUT: Mapped memory */
){
  winFile *pDbFd = (winFile*)fd;
  winShm *pShm = pDbFd->pShm;
  winShmNode *pShmNode;
  DWORD protect = PAGE_READWRITE;
  DWORD flags = FILE_MAP_WRITE | FILE_MAP_READ;
  int rc = SQLITE_OK;

  if( !pShm ){
    rc = winOpenSharedMemory(pDbFd);
    if( rc!=SQLITE_OK ) return rc;
    pShm = pDbFd->pShm;
  }
  pShmNode = pShm->pShmNode;

  sqlite3_mutex_enter(pShmNode->mutex);
  if( pShmNode->isUnlocked ){
    rc = winLockSharedMemory(pShmNode);
    if( rc!=SQLITE_OK ) goto shmpage_out;
    pShmNode->isUnlocked = 0;
  }
  assert( szRegion==pShmNode->szRegion || pShmNode->nRegion==0 );

  if( pShmNode->nRegion<=iRegion ){
    struct ShmRegion *apNew;           /* New aRegion[] array */
    int nByte = (iRegion+1)*szRegion;  /* Minimum required file size */
    sqlite3_int64 sz;                  /* Current size of wal-index file */

................................................................................
        pShmNode->aRegion, (iRegion+1)*sizeof(apNew[0])
    );
    if( !apNew ){
      rc = SQLITE_IOERR_NOMEM_BKPT;
      goto shmpage_out;
    }
    pShmNode->aRegion = apNew;

    if( pShmNode->isReadonly ){
      protect = PAGE_READONLY;
      flags = FILE_MAP_READ;
    }

    while( pShmNode->nRegion<=iRegion ){
      HANDLE hMap = NULL;         /* file-mapping handle */
      void *pMap = 0;             /* Mapped memory region */

#if SQLITE_OS_WINRT
      hMap = osCreateFileMappingFromApp(pShmNode->hFile.h,
          NULL, protect, nByte, NULL
      );
#elif defined(SQLITE_WIN32_HAS_WIDE)
      hMap = osCreateFileMappingW(pShmNode->hFile.h,
          NULL, protect, 0, nByte, NULL
      );
#elif defined(SQLITE_WIN32_HAS_ANSI) && SQLITE_WIN32_CREATEFILEMAPPINGA
      hMap = osCreateFileMappingA(pShmNode->hFile.h,
          NULL, protect, 0, nByte, NULL
      );
#endif
      OSTRACE(("SHM-MAP-CREATE pid=%lu, region=%d, size=%d, rc=%s\n",
               osGetCurrentProcessId(), pShmNode->nRegion, nByte,
               hMap ? "ok" : "failed"));
      if( hMap ){
        int iOffset = pShmNode->nRegion*szRegion;
        int iOffsetShift = iOffset % winSysInfo.dwAllocationGranularity;
#if SQLITE_OS_WINRT
        pMap = osMapViewOfFileFromApp(hMap, flags,
            iOffset - iOffsetShift, szRegion + iOffsetShift
        );
#else
        pMap = osMapViewOfFile(hMap, flags,
            0, iOffset - iOffsetShift, szRegion + iOffsetShift
        );
#endif
        OSTRACE(("SHM-MAP-MAP pid=%lu, region=%d, offset=%d, size=%d, rc=%s\n",
                 osGetCurrentProcessId(), pShmNode->nRegion, iOffset,
                 szRegion, pMap ? "ok" : "failed"));
      }
................................................................................
    int iOffset = iRegion*szRegion;
    int iOffsetShift = iOffset % winSysInfo.dwAllocationGranularity;
    char *p = (char *)pShmNode->aRegion[iRegion].pMap;
    *pp = (void *)&p[iOffsetShift];
  }else{
    *pp = 0;
  }
  if( pShmNode->isReadonly && rc==SQLITE_OK ) rc = SQLITE_READONLY;
  sqlite3_mutex_leave(pShmNode->mutex);
  return rc;
}

#else
# define winShmMap     0
# define winShmLock    0

Changes to src/printf.c.

1088
1089
1090
1091
1092
1093
1094






1095
1096

1097
1098
1099
1100
1101
1102
1103
  StrAccum acc;
  char zBuf[500];
  sqlite3StrAccumInit(&acc, 0, zBuf, sizeof(zBuf), 0);
  va_start(ap,zFormat);
  sqlite3VXPrintf(&acc, zFormat, ap);
  va_end(ap);
  sqlite3StrAccumFinish(&acc);






  fprintf(stdout,"%s", zBuf);
  fflush(stdout);

}
#endif


/*
** variable-argument wrapper around sqlite3VXPrintf().  The bFlags argument
** can contain the bit SQLITE_PRINTF_INTERNAL enable internal formats.







>
>
>
>
>
>


>







1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
  StrAccum acc;
  char zBuf[500];
  sqlite3StrAccumInit(&acc, 0, zBuf, sizeof(zBuf), 0);
  va_start(ap,zFormat);
  sqlite3VXPrintf(&acc, zFormat, ap);
  va_end(ap);
  sqlite3StrAccumFinish(&acc);
#ifdef SQLITE_OS_TRACE_PROC
  {
    extern void SQLITE_OS_TRACE_PROC(const char *zBuf, int nBuf);
    SQLITE_OS_TRACE_PROC(zBuf, sizeof(zBuf));
  }
#else
  fprintf(stdout,"%s", zBuf);
  fflush(stdout);
#endif
}
#endif


/*
** variable-argument wrapper around sqlite3VXPrintf().  The bFlags argument
** can contain the bit SQLITE_PRINTF_INTERNAL enable internal formats.

Changes to src/sqlite.h.in.

504
505
506
507
508
509
510

511
512
513
514
515

516
517
518
519
520
521
522
#define SQLITE_LOCKED_SHAREDCACHE      (SQLITE_LOCKED |  (1<<8))
#define SQLITE_BUSY_RECOVERY           (SQLITE_BUSY   |  (1<<8))
#define SQLITE_BUSY_SNAPSHOT           (SQLITE_BUSY   |  (2<<8))
#define SQLITE_CANTOPEN_NOTEMPDIR      (SQLITE_CANTOPEN | (1<<8))
#define SQLITE_CANTOPEN_ISDIR          (SQLITE_CANTOPEN | (2<<8))
#define SQLITE_CANTOPEN_FULLPATH       (SQLITE_CANTOPEN | (3<<8))
#define SQLITE_CANTOPEN_CONVPATH       (SQLITE_CANTOPEN | (4<<8))

#define SQLITE_CORRUPT_VTAB            (SQLITE_CORRUPT | (1<<8))
#define SQLITE_READONLY_RECOVERY       (SQLITE_READONLY | (1<<8))
#define SQLITE_READONLY_CANTLOCK       (SQLITE_READONLY | (2<<8))
#define SQLITE_READONLY_ROLLBACK       (SQLITE_READONLY | (3<<8))
#define SQLITE_READONLY_DBMOVED        (SQLITE_READONLY | (4<<8))

#define SQLITE_ABORT_ROLLBACK          (SQLITE_ABORT | (2<<8))
#define SQLITE_CONSTRAINT_CHECK        (SQLITE_CONSTRAINT | (1<<8))
#define SQLITE_CONSTRAINT_COMMITHOOK   (SQLITE_CONSTRAINT | (2<<8))
#define SQLITE_CONSTRAINT_FOREIGNKEY   (SQLITE_CONSTRAINT | (3<<8))
#define SQLITE_CONSTRAINT_FUNCTION     (SQLITE_CONSTRAINT | (4<<8))
#define SQLITE_CONSTRAINT_NOTNULL      (SQLITE_CONSTRAINT | (5<<8))
#define SQLITE_CONSTRAINT_PRIMARYKEY   (SQLITE_CONSTRAINT | (6<<8))







>





>







504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
#define SQLITE_LOCKED_SHAREDCACHE      (SQLITE_LOCKED |  (1<<8))
#define SQLITE_BUSY_RECOVERY           (SQLITE_BUSY   |  (1<<8))
#define SQLITE_BUSY_SNAPSHOT           (SQLITE_BUSY   |  (2<<8))
#define SQLITE_CANTOPEN_NOTEMPDIR      (SQLITE_CANTOPEN | (1<<8))
#define SQLITE_CANTOPEN_ISDIR          (SQLITE_CANTOPEN | (2<<8))
#define SQLITE_CANTOPEN_FULLPATH       (SQLITE_CANTOPEN | (3<<8))
#define SQLITE_CANTOPEN_CONVPATH       (SQLITE_CANTOPEN | (4<<8))
#define SQLITE_CANTOPEN_DIRTYWAL       (SQLITE_CANTOPEN | (5<<8))
#define SQLITE_CORRUPT_VTAB            (SQLITE_CORRUPT | (1<<8))
#define SQLITE_READONLY_RECOVERY       (SQLITE_READONLY | (1<<8))
#define SQLITE_READONLY_CANTLOCK       (SQLITE_READONLY | (2<<8))
#define SQLITE_READONLY_ROLLBACK       (SQLITE_READONLY | (3<<8))
#define SQLITE_READONLY_DBMOVED        (SQLITE_READONLY | (4<<8))
#define SQLITE_READONLY_CANTINIT       (SQLITE_READONLY | (5<<8))
#define SQLITE_ABORT_ROLLBACK          (SQLITE_ABORT | (2<<8))
#define SQLITE_CONSTRAINT_CHECK        (SQLITE_CONSTRAINT | (1<<8))
#define SQLITE_CONSTRAINT_COMMITHOOK   (SQLITE_CONSTRAINT | (2<<8))
#define SQLITE_CONSTRAINT_FOREIGNKEY   (SQLITE_CONSTRAINT | (3<<8))
#define SQLITE_CONSTRAINT_FUNCTION     (SQLITE_CONSTRAINT | (4<<8))
#define SQLITE_CONSTRAINT_NOTNULL      (SQLITE_CONSTRAINT | (5<<8))
#define SQLITE_CONSTRAINT_PRIMARYKEY   (SQLITE_CONSTRAINT | (6<<8))

Changes to src/wal.c.

451
452
453
454
455
456
457

458
459
460
461
462
463
464
...
539
540
541
542
543
544
545





546
547
548
549
550
551
552
...
571
572
573
574
575
576
577


578
579

580

581
582
583
584
585
586
587
....
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
....
1250
1251
1252
1253
1254
1255
1256
1257

1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271

1272
1273
1274
1275
1276
1277
1278
....
2057
2058
2059
2060
2061
2062
2063






2064
2065
2066
2067
2068
2069
2070
....
2080
2081
2082
2083
2084
2085
2086
2087
2088








2089













2090
2091
2092
2093
2094
2095
2096
2097
2098
2099
2100
2101
2102
2103
2104
2105
2106
2107
2108
2109
2110
....
2126
2127
2128
2129
2130
2131
2132









2133



2134
2135
2136
2137
2138
2139

























2140
























































































2141























































2142
2143
2144
2145
2146
2147
2148
....
2196
2197
2198
2199
2200
2201
2202



2203
2204
2205
2206
2207
2208
2209
....
2224
2225
2226
2227
2228
2229
2230


2231

2232
2233
2234
2235
2236
2237
2238
....
2253
2254
2255
2256
2257
2258
2259


2260
2261



2262
2263
2264
2265
2266
2267
2268
2269
2270
....
2330
2331
2332
2333
2334
2335
2336
2337
2338
2339
2340
2341
2342
2343
2344
....
2602
2603
2604
2605
2606
2607
2608
2609
2610
2611
2612
2613
2614
2615
2616
....
2665
2666
2667
2668
2669
2670
2671
2672
2673
2674
2675
2676
2677
2678
2679
2680
....
3442
3443
3444
3445
3446
3447
3448
3449
3450
3451
3452
3453
3454
3455
3456
3457
3458
3459
3460
3461
3462
3463
3464
3465
3466
3467
3468
3469
3470
3471
3472
3473
  u8 exclusiveMode;          /* Non-zero if connection is in exclusive mode */
  u8 writeLock;              /* True if in a write transaction */
  u8 ckptLock;               /* True if holding a checkpoint lock */
  u8 readOnly;               /* WAL_RDWR, WAL_RDONLY, or WAL_SHM_RDONLY */
  u8 truncateOnCommit;       /* True to truncate WAL file on commit */
  u8 syncHeader;             /* Fsync the WAL header if true */
  u8 padToSectorBoundary;    /* Pad transactions out to the next sector */

  WalIndexHdr hdr;           /* Wal-index header for current transaction */
  u32 minFrame;              /* Ignore wal frames before this one */
  u32 iReCksum;              /* On commit, recalculate checksums from here */
  const char *zWalName;      /* Name of WAL file */
  u32 nCkpt;                 /* Checkpoint sequence counter in the wal-header */
#ifdef SQLITE_DEBUG
  u8 lockError;              /* True if a locking error has occurred */
................................................................................
    sizeof(ht_slot)*HASHTABLE_NSLOT + HASHTABLE_NPAGE*sizeof(u32) \
)

/*
** Obtain a pointer to the iPage'th page of the wal-index. The wal-index
** is broken into pages of WALINDEX_PGSZ bytes. Wal-index pages are
** numbered from zero.





**
** If this call is successful, *ppPage is set to point to the wal-index
** page and SQLITE_OK is returned. If an error (an OOM or VFS error) occurs,
** then an SQLite error code is returned and *ppPage is set to 0.
*/
static int walIndexPage(Wal *pWal, int iPage, volatile u32 **ppPage){
  int rc = SQLITE_OK;
................................................................................
    if( pWal->exclusiveMode==WAL_HEAPMEMORY_MODE ){
      pWal->apWiData[iPage] = (u32 volatile *)sqlite3MallocZero(WALINDEX_PGSZ);
      if( !pWal->apWiData[iPage] ) rc = SQLITE_NOMEM_BKPT;
    }else{
      rc = sqlite3OsShmMap(pWal->pDbFd, iPage, WALINDEX_PGSZ, 
          pWal->writeLock, (void volatile **)&pWal->apWiData[iPage]
      );


      if( rc==SQLITE_READONLY ){
        pWal->readOnly |= WAL_SHM_RDONLY;

        rc = SQLITE_OK;

      }
    }
  }

  *ppPage = pWal->apWiData[iPage];
  assert( iPage==0 || *ppPage || rc!=SQLITE_OK );
  return rc;
................................................................................
** the necessary locks, this routine returns SQLITE_BUSY.
*/
static int walIndexRecover(Wal *pWal){
  int rc;                         /* Return Code */
  i64 nSize;                      /* Size of log file */
  u32 aFrameCksum[2] = {0, 0};
  int iLock;                      /* Lock offset to lock for checkpoint */
  int nLock;                      /* Number of locks to hold */

  /* Obtain an exclusive lock on all byte in the locking range not already
  ** locked by the caller. The caller is guaranteed to have locked the
  ** WAL_WRITE_LOCK byte, and may have also locked the WAL_CKPT_LOCK byte.
  ** If successful, the same bytes that are locked here are unlocked before
  ** this function returns.
  */
  assert( pWal->ckptLock==1 || pWal->ckptLock==0 );
  assert( WAL_ALL_BUT_WRITE==WAL_WRITE_LOCK+1 );
  assert( WAL_CKPT_LOCK==WAL_ALL_BUT_WRITE );
  assert( pWal->writeLock );
  iLock = WAL_ALL_BUT_WRITE + pWal->ckptLock;
  nLock = SQLITE_SHM_NLOCK - iLock;
  rc = walLockExclusive(pWal, iLock, nLock);






  if( rc ){
    return rc;
  }

  WALTRACE(("WAL%p: recovery begin...\n", pWal));

  memset(&pWal->hdr, 0, sizeof(WalIndexHdr));

  rc = sqlite3OsFileSize(pWal->pWalFd, &nSize);
  if( rc!=SQLITE_OK ){
    goto recovery_error;
................................................................................
          pWal->hdr.mxFrame, pWal->zWalName
      );
    }
  }

recovery_error:
  WALTRACE(("WAL%p: recovery %s\n", pWal, rc ? "failed" : "ok"));
  walUnlockExclusive(pWal, iLock, nLock);

  return rc;
}

/*
** Close an open wal-index.
*/
static void walIndexClose(Wal *pWal, int isDelete){
  if( pWal->exclusiveMode==WAL_HEAPMEMORY_MODE ){
    int i;
    for(i=0; i<pWal->nWiData; i++){
      sqlite3_free((void *)pWal->apWiData[i]);
      pWal->apWiData[i] = 0;
    }
  }else{

    sqlite3OsShmUnmap(pWal->pDbFd, isDelete);
  }
}

/* 
** Open a connection to the WAL file zWalName. The database file must 
** already be opened on connection pDbFd. The buffer that zWalName points
................................................................................
    testcase( pWal->szPage>=65536 );
  }

  /* The header was successfully read. Return zero. */
  return 0;
}







/*
** Read the wal-index header from the wal-index and into pWal->hdr.
** If the wal-header appears to be corrupt, try to reconstruct the
** wal-index from the WAL before returning.
**
** Set *pChanged to 1 if the wal-index header value in pWal->hdr is
** changed by this operation.  If pWal->hdr is unchanged, set *pChanged
................................................................................

  /* Ensure that page 0 of the wal-index (the page that contains the 
  ** wal-index header) is mapped. Return early if an error occurs here.
  */
  assert( pChanged );
  rc = walIndexPage(pWal, 0, &page0);
  if( rc!=SQLITE_OK ){
    return rc;
  };








  assert( page0 || pWal->writeLock==0 );














  /* If the first page of the wal-index has been mapped, try to read the
  ** wal-index header immediately, without holding any lock. This usually
  ** works, but may fail if the wal-index header is corrupt or currently 
  ** being modified by another thread or process.
  */
  badHdr = (page0 ? walIndexTryHdr(pWal, pChanged) : 1);

  /* If the first attempt failed, it might have been due to a race
  ** with a writer.  So get a WRITE lock and try again.
  */
  assert( badHdr==0 || pWal->writeLock==0 );
  if( badHdr ){
    if( pWal->readOnly & WAL_SHM_RDONLY ){
      if( SQLITE_OK==(rc = walLockShared(pWal, WAL_WRITE_LOCK)) ){
        walUnlockShared(pWal, WAL_WRITE_LOCK);
        rc = SQLITE_READONLY_RECOVERY;
      }
    }else if( SQLITE_OK==(rc = walLockExclusive(pWal, WAL_WRITE_LOCK, 1)) ){
      pWal->writeLock = 1;
      if( SQLITE_OK==(rc = walIndexPage(pWal, 0, &page0)) ){
................................................................................
  /* If the header is read successfully, check the version number to make
  ** sure the wal-index was not constructed with some future format that
  ** this version of SQLite cannot understand.
  */
  if( badHdr==0 && pWal->hdr.iVersion!=WALINDEX_MAX_VERSION ){
    rc = SQLITE_CANTOPEN_BKPT;
  }













  return rc;
}

/*
** This is the value that walTryBeginRead returns when it needs to
** be retried.

























*/
























































































#define WAL_RETRY  (-1)
























































/*
** Attempt to start a read transaction.  This might fail due to a race or
** other transient condition.  When that happens, it returns WAL_RETRY to
** indicate to the caller that it is safe to retry immediately.
**
** On success return SQLITE_OK.  On a permanent failure (such an
................................................................................
  int mxI;                        /* Index of largest aReadMark[] value */
  int i;                          /* Loop counter */
  int rc = SQLITE_OK;             /* Return code  */
  u32 mxFrame;                    /* Wal frame to lock to */

  assert( pWal->readLock<0 );     /* Not currently locked */




  /* Take steps to avoid spinning forever if there is a protocol error.
  **
  ** Circumstances that cause a RETRY should only last for the briefest
  ** instances of time.  No I/O or other system calls are done while the
  ** locks are held, so the locks should not be held for very long. But 
  ** if we are unlucky, another process that is holding a lock might get
  ** paged out or take a page-fault that is time-consuming to resolve, 
................................................................................
      return SQLITE_PROTOCOL;
    }
    if( cnt>=10 ) nDelay = (cnt-9)*(cnt-9)*39;
    sqlite3OsSleep(pWal->pVfs, nDelay);
  }

  if( !useWal ){


    rc = walIndexReadHdr(pWal, pChanged);

    if( rc==SQLITE_BUSY ){
      /* If there is not a recovery running in another thread or process
      ** then convert BUSY errors to WAL_RETRY.  If recovery is known to
      ** be running, convert BUSY to BUSY_RECOVERY.  There is a race here
      ** which might cause WAL_RETRY to be returned even if BUSY_RECOVERY
      ** would be technically correct.  But the race is benign since with
      ** WAL_RETRY this routine will be called again and will probably be
................................................................................
      }else if( rc==SQLITE_BUSY ){
        rc = SQLITE_BUSY_RECOVERY;
      }
    }
    if( rc!=SQLITE_OK ){
      return rc;
    }


  }




  pInfo = walCkptInfo(pWal);
  if( !useWal && pInfo->nBackfill==pWal->hdr.mxFrame 
#ifdef SQLITE_ENABLE_SNAPSHOT
   && (pWal->pSnapshot==0 || pWal->hdr.mxFrame==0
     || 0==memcmp(&pWal->hdr, pWal->pSnapshot, sizeof(WalIndexHdr)))
#endif
  ){
    /* The WAL has been completely backfilled (or it is empty).
    ** and can be safely ignored.
................................................................................
      }else if( rc!=SQLITE_BUSY ){
        return rc;
      }
    }
  }
  if( mxI==0 ){
    assert( rc==SQLITE_BUSY || (pWal->readOnly & WAL_SHM_RDONLY)!=0 );
    return rc==SQLITE_BUSY ? WAL_RETRY : SQLITE_READONLY_CANTLOCK;
  }

  rc = walLockShared(pWal, WAL_READ_LOCK(mxI));
  if( rc ){
    return rc==SQLITE_BUSY ? WAL_RETRY : rc;
  }
  /* Now that the read-lock has been obtained, check that neither the
................................................................................

  /* If the "last page" field of the wal-index header snapshot is 0, then
  ** no data will be read from the wal under any circumstances. Return early
  ** in this case as an optimization.  Likewise, if pWal->readLock==0, 
  ** then the WAL is ignored by the reader so return early, as if the 
  ** WAL were empty.
  */
  if( iLast==0 || pWal->readLock==0 ){
    *piRead = 0;
    return SQLITE_OK;
  }

  /* Search the hash table or tables for an entry matching page number
  ** pgno. Each iteration of the following for() loop searches one
  ** hash table (each hash table indexes up to HASHTABLE_NPAGE frames).
................................................................................
#ifdef SQLITE_ENABLE_EXPENSIVE_ASSERT
  /* If expensive assert() statements are available, do a linear search
  ** of the wal-index file content. Make sure the results agree with the
  ** result obtained using the hash indexes above.  */
  {
    u32 iRead2 = 0;
    u32 iTest;
    assert( pWal->minFrame>0 );
    for(iTest=iLast; iTest>=pWal->minFrame; iTest--){
      if( walFramePgno(pWal, iTest)==pgno ){
        iRead2 = iTest;
        break;
      }
    }
    assert( iRead==iRead2 );
  }
................................................................................
  ** locks are taken in this case). Nor should the pager attempt to
  ** upgrade to exclusive-mode following such an error.
  */
  assert( pWal->readLock>=0 || pWal->lockError );
  assert( pWal->readLock>=0 || (op<=0 && pWal->exclusiveMode==0) );

  if( op==0 ){
    if( pWal->exclusiveMode ){
      pWal->exclusiveMode = 0;
      if( walLockShared(pWal, WAL_READ_LOCK(pWal->readLock))!=SQLITE_OK ){
        pWal->exclusiveMode = 1;
      }
      rc = pWal->exclusiveMode==0;
    }else{
      /* Already in locking_mode=NORMAL */
      rc = 0;
    }
  }else if( op>0 ){
    assert( pWal->exclusiveMode==0 );
    assert( pWal->readLock>=0 );
    walUnlockShared(pWal, WAL_READ_LOCK(pWal->readLock));
    pWal->exclusiveMode = 1;
    rc = 1;
  }else{
    rc = pWal->exclusiveMode==0;
  }
  return rc;
}

/* 
** 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







>







 







>
>
>
>
>







 







>
>
|

>
|
>







 







<












<
|
>
>
>
>
>
>



>







 







|
>







|





|
>







 







>
>
>
>
>
>







 







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













|







 







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




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

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







 







>
>
>







 







>
>
|
>







 







>
>
|
|
>
>
>

|







 







|







 







|







 







|
|







 







|
|

|

|





|


|


|







451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
...
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
...
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
....
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
....
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
....
2074
2075
2076
2077
2078
2079
2080
2081
2082
2083
2084
2085
2086
2087
2088
2089
2090
2091
2092
2093
....
2103
2104
2105
2106
2107
2108
2109
2110

2111
2112
2113
2114
2115
2116
2117
2118
2119
2120
2121
2122
2123
2124
2125
2126
2127
2128
2129
2130
2131
2132
2133
2134
2135
2136
2137
2138
2139
2140
2141
2142
2143
2144
2145
2146
2147
2148
2149
2150
2151
2152
2153
....
2169
2170
2171
2172
2173
2174
2175
2176
2177
2178
2179
2180
2181
2182
2183
2184
2185
2186
2187
2188
2189
2190
2191
2192


2193
2194
2195
2196
2197
2198
2199
2200
2201
2202
2203
2204
2205
2206
2207
2208
2209
2210
2211
2212
2213
2214
2215
2216
2217
2218
2219
2220
2221
2222
2223
2224
2225
2226
2227
2228
2229
2230
2231
2232
2233
2234
2235
2236
2237
2238
2239
2240
2241
2242
2243
2244
2245
2246
2247
2248
2249
2250
2251
2252
2253
2254
2255
2256
2257
2258
2259
2260
2261
2262
2263
2264
2265
2266
2267
2268
2269
2270
2271
2272
2273
2274
2275
2276
2277
2278
2279
2280
2281
2282
2283
2284
2285
2286
2287
2288
2289
2290
2291
2292
2293
2294
2295
2296
2297
2298
2299
2300
2301
2302
2303
2304
2305
2306
2307
2308
2309
2310
2311
2312
2313
2314
2315
2316
2317
2318
2319
2320
2321
2322
2323
2324
2325
2326
2327
2328
2329
2330
2331
2332
2333
2334
2335
2336
2337
2338
2339
2340
2341
2342
2343
2344
2345
2346
2347
2348
2349
2350
2351
2352
2353
2354
2355
2356
2357
2358
2359
2360
2361
2362
2363
2364
2365
2366
2367
2368
2369
....
2417
2418
2419
2420
2421
2422
2423
2424
2425
2426
2427
2428
2429
2430
2431
2432
2433
....
2448
2449
2450
2451
2452
2453
2454
2455
2456
2457
2458
2459
2460
2461
2462
2463
2464
2465
....
2480
2481
2482
2483
2484
2485
2486
2487
2488
2489
2490
2491
2492
2493
2494
2495
2496
2497
2498
2499
2500
2501
2502
....
2562
2563
2564
2565
2566
2567
2568
2569
2570
2571
2572
2573
2574
2575
2576
....
2834
2835
2836
2837
2838
2839
2840
2841
2842
2843
2844
2845
2846
2847
2848
....
2897
2898
2899
2900
2901
2902
2903
2904
2905
2906
2907
2908
2909
2910
2911
2912
....
3674
3675
3676
3677
3678
3679
3680
3681
3682
3683
3684
3685
3686
3687
3688
3689
3690
3691
3692
3693
3694
3695
3696
3697
3698
3699
3700
3701
3702
3703
3704
3705
  u8 exclusiveMode;          /* Non-zero if connection is in exclusive mode */
  u8 writeLock;              /* True if in a write transaction */
  u8 ckptLock;               /* True if holding a checkpoint lock */
  u8 readOnly;               /* WAL_RDWR, WAL_RDONLY, or WAL_SHM_RDONLY */
  u8 truncateOnCommit;       /* True to truncate WAL file on commit */
  u8 syncHeader;             /* Fsync the WAL header if true */
  u8 padToSectorBoundary;    /* Pad transactions out to the next sector */
  u8 bShmUnreliable;         /* SHM content is read-only and unreliable */
  WalIndexHdr hdr;           /* Wal-index header for current transaction */
  u32 minFrame;              /* Ignore wal frames before this one */
  u32 iReCksum;              /* On commit, recalculate checksums from here */
  const char *zWalName;      /* Name of WAL file */
  u32 nCkpt;                 /* Checkpoint sequence counter in the wal-header */
#ifdef SQLITE_DEBUG
  u8 lockError;              /* True if a locking error has occurred */
................................................................................
    sizeof(ht_slot)*HASHTABLE_NSLOT + HASHTABLE_NPAGE*sizeof(u32) \
)

/*
** Obtain a pointer to the iPage'th page of the wal-index. The wal-index
** is broken into pages of WALINDEX_PGSZ bytes. Wal-index pages are
** numbered from zero.
**
** If the wal-index is currently smaller the iPage pages then the size
** of the wal-index might be increased, but only if it is safe to do
** so.  It is safe to enlarge the wal-index if pWal->writeLock is true
** or pWal->exclusiveMode==WAL_HEAPMEMORY_MODE.
**
** If this call is successful, *ppPage is set to point to the wal-index
** page and SQLITE_OK is returned. If an error (an OOM or VFS error) occurs,
** then an SQLite error code is returned and *ppPage is set to 0.
*/
static int walIndexPage(Wal *pWal, int iPage, volatile u32 **ppPage){
  int rc = SQLITE_OK;
................................................................................
    if( pWal->exclusiveMode==WAL_HEAPMEMORY_MODE ){
      pWal->apWiData[iPage] = (u32 volatile *)sqlite3MallocZero(WALINDEX_PGSZ);
      if( !pWal->apWiData[iPage] ) rc = SQLITE_NOMEM_BKPT;
    }else{
      rc = sqlite3OsShmMap(pWal->pDbFd, iPage, WALINDEX_PGSZ, 
          pWal->writeLock, (void volatile **)&pWal->apWiData[iPage]
      );
      assert( pWal->apWiData[iPage]!=0 || rc!=SQLITE_OK || pWal->writeLock==0 );
      testcase( pWal->apWiData[iPage]==0 && rc==SQLITE_OK );
      if( (rc&0xff)==SQLITE_READONLY ){
        pWal->readOnly |= WAL_SHM_RDONLY;
        if( rc==SQLITE_READONLY ){
          rc = SQLITE_OK;
        }
      }
    }
  }

  *ppPage = pWal->apWiData[iPage];
  assert( iPage==0 || *ppPage || rc!=SQLITE_OK );
  return rc;
................................................................................
** the necessary locks, this routine returns SQLITE_BUSY.
*/
static int walIndexRecover(Wal *pWal){
  int rc;                         /* Return Code */
  i64 nSize;                      /* Size of log file */
  u32 aFrameCksum[2] = {0, 0};
  int iLock;                      /* Lock offset to lock for checkpoint */


  /* Obtain an exclusive lock on all byte in the locking range not already
  ** locked by the caller. The caller is guaranteed to have locked the
  ** WAL_WRITE_LOCK byte, and may have also locked the WAL_CKPT_LOCK byte.
  ** If successful, the same bytes that are locked here are unlocked before
  ** this function returns.
  */
  assert( pWal->ckptLock==1 || pWal->ckptLock==0 );
  assert( WAL_ALL_BUT_WRITE==WAL_WRITE_LOCK+1 );
  assert( WAL_CKPT_LOCK==WAL_ALL_BUT_WRITE );
  assert( pWal->writeLock );
  iLock = WAL_ALL_BUT_WRITE + pWal->ckptLock;

  rc = walLockExclusive(pWal, iLock, WAL_READ_LOCK(0)-iLock);
  if( rc==SQLITE_OK ){
    rc = walLockExclusive(pWal, WAL_READ_LOCK(1), WAL_NREADER-1);
    if( rc!=SQLITE_OK ){
      walUnlockExclusive(pWal, iLock, WAL_READ_LOCK(0)-iLock);
    }
  }
  if( rc ){
    return rc;
  }

  WALTRACE(("WAL%p: recovery begin...\n", pWal));

  memset(&pWal->hdr, 0, sizeof(WalIndexHdr));

  rc = sqlite3OsFileSize(pWal->pWalFd, &nSize);
  if( rc!=SQLITE_OK ){
    goto recovery_error;
................................................................................
          pWal->hdr.mxFrame, pWal->zWalName
      );
    }
  }

recovery_error:
  WALTRACE(("WAL%p: recovery %s\n", pWal, rc ? "failed" : "ok"));
  walUnlockExclusive(pWal, iLock, WAL_READ_LOCK(0)-iLock);
  walUnlockExclusive(pWal, WAL_READ_LOCK(1), WAL_NREADER-1);
  return rc;
}

/*
** Close an open wal-index.
*/
static void walIndexClose(Wal *pWal, int isDelete){
  if( pWal->exclusiveMode==WAL_HEAPMEMORY_MODE || pWal->bShmUnreliable ){
    int i;
    for(i=0; i<pWal->nWiData; i++){
      sqlite3_free((void *)pWal->apWiData[i]);
      pWal->apWiData[i] = 0;
    }
  }
  if( pWal->exclusiveMode!=WAL_HEAPMEMORY_MODE ){
    sqlite3OsShmUnmap(pWal->pDbFd, isDelete);
  }
}

/* 
** Open a connection to the WAL file zWalName. The database file must 
** already be opened on connection pDbFd. The buffer that zWalName points
................................................................................
    testcase( pWal->szPage>=65536 );
  }

  /* The header was successfully read. Return zero. */
  return 0;
}

/*
** This is the value that walTryBeginRead returns when it needs to
** be retried.
*/
#define WAL_RETRY  (-1)

/*
** Read the wal-index header from the wal-index and into pWal->hdr.
** If the wal-header appears to be corrupt, try to reconstruct the
** wal-index from the WAL before returning.
**
** Set *pChanged to 1 if the wal-index header value in pWal->hdr is
** changed by this operation.  If pWal->hdr is unchanged, set *pChanged
................................................................................

  /* Ensure that page 0 of the wal-index (the page that contains the 
  ** wal-index header) is mapped. Return early if an error occurs here.
  */
  assert( pChanged );
  rc = walIndexPage(pWal, 0, &page0);
  if( rc!=SQLITE_OK ){
    assert( rc!=SQLITE_READONLY ); /* READONLY changed to OK in walIndexPage */

    if( rc==SQLITE_READONLY_CANTINIT ){
      /* The SQLITE_READONLY_CANTINIT return means that the shared-memory
      ** was openable but is not writable, and this thread is unable to
      ** confirm that another write-capable connection has the shared-memory
      ** open, and hence the content of the shared-memory is unreliable,
      ** since the shared-memory might be inconsistent with the WAL file
      ** and there is no writer on hand to fix it. */
      assert( page0==0 );
      assert( pWal->writeLock==0 );
      assert( pWal->readOnly & WAL_SHM_RDONLY );
      pWal->bShmUnreliable = 1;
      pWal->exclusiveMode = WAL_HEAPMEMORY_MODE;
      *pChanged = 1;
    }else{
      return rc; /* Any other non-OK return is just an error */
    }
  }else{
    /* page0 can be NULL if the SHM is zero bytes in size and pWal->writeLock
    ** is zero, which prevents the SHM from growing */
    testcase( page0!=0 );
  }
  assert( page0!=0 || pWal->writeLock==0 );

  /* If the first page of the wal-index has been mapped, try to read the
  ** wal-index header immediately, without holding any lock. This usually
  ** works, but may fail if the wal-index header is corrupt or currently 
  ** being modified by another thread or process.
  */
  badHdr = (page0 ? walIndexTryHdr(pWal, pChanged) : 1);

  /* If the first attempt failed, it might have been due to a race
  ** with a writer.  So get a WRITE lock and try again.
  */
  assert( badHdr==0 || pWal->writeLock==0 );
  if( badHdr ){
    if( pWal->bShmUnreliable==0 && (pWal->readOnly & WAL_SHM_RDONLY) ){
      if( SQLITE_OK==(rc = walLockShared(pWal, WAL_WRITE_LOCK)) ){
        walUnlockShared(pWal, WAL_WRITE_LOCK);
        rc = SQLITE_READONLY_RECOVERY;
      }
    }else if( SQLITE_OK==(rc = walLockExclusive(pWal, WAL_WRITE_LOCK, 1)) ){
      pWal->writeLock = 1;
      if( SQLITE_OK==(rc = walIndexPage(pWal, 0, &page0)) ){
................................................................................
  /* If the header is read successfully, check the version number to make
  ** sure the wal-index was not constructed with some future format that
  ** this version of SQLite cannot understand.
  */
  if( badHdr==0 && pWal->hdr.iVersion!=WALINDEX_MAX_VERSION ){
    rc = SQLITE_CANTOPEN_BKPT;
  }
  if( pWal->bShmUnreliable ){
    if( rc!=SQLITE_OK ){
      walIndexClose(pWal, 0);
      pWal->bShmUnreliable = 0;
      assert( pWal->nWiData>0 && pWal->apWiData[0]==0 );
      /* walIndexRecover() might have returned SHORT_READ if a concurrent
      ** writer truncated the WAL out from under it.  If that happens, it
      ** indicates that a writer has fixed the SHM file for us, so retry */
      if( rc==SQLITE_IOERR_SHORT_READ ) rc = WAL_RETRY;
    }
    pWal->exclusiveMode = WAL_NORMAL_MODE;
  }

  return rc;
}

/*


** Open a transaction in a connection where the shared-memory is read-only
** and where we cannot verify that there is a separate write-capable connection
** on hand to keep the shared-memory up-to-date with the WAL file.
**
** This can happen, for example, when the shared-memory is implemented by
** memory-mapping a *-shm file, where a prior writer has shut down and
** left the *-shm file on disk, and now the present connection is trying
** to use that database but lacks write permission on the *-shm file.
** Other scenarios are also possible, depending on the VFS implementation.
**
** Precondition:
**
**    The *-wal file has been read and an appropriate wal-index has been
**    constructed in pWal->apWiData[] using heap memory instead of shared
**    memory. 
**
** If this function returns SQLITE_OK, then the read transaction has
** been successfully opened. In this case output variable (*pChanged) 
** is set to true before returning if the caller should discard the
** contents of the page cache before proceeding. Or, if it returns 
** WAL_RETRY, then the heap memory wal-index has been discarded and 
** the caller should retry opening the read transaction from the 
** beginning (including attempting to map the *-shm file). 
**
** If an error occurs, an SQLite error code is returned.
*/
static int walBeginShmUnreliable(Wal *pWal, int *pChanged){
  i64 szWal;                      /* Size of wal file on disk in bytes */
  i64 iOffset;                    /* Current offset when reading wal file */
  u8 aBuf[WAL_HDRSIZE];           /* Buffer to load WAL header into */
  u8 *aFrame = 0;                 /* Malloc'd buffer to load entire frame */
  int szFrame;                    /* Number of bytes in buffer aFrame[] */
  u8 *aData;                      /* Pointer to data part of aFrame buffer */
  volatile void *pDummy;          /* Dummy argument for xShmMap */
  int rc;                         /* Return code */
  u32 aSaveCksum[2];              /* Saved copy of pWal->hdr.aFrameCksum */

  assert( pWal->bShmUnreliable );
  assert( pWal->readOnly & WAL_SHM_RDONLY );
  assert( pWal->nWiData>0 && pWal->apWiData[0] );

  /* Take WAL_READ_LOCK(0). This has the effect of preventing any
  ** writers from running a checkpoint, but does not stop them
  ** from running recovery.  */
  rc = walLockShared(pWal, WAL_READ_LOCK(0));
  if( rc!=SQLITE_OK ){
    if( rc==SQLITE_BUSY ) rc = WAL_RETRY;
    goto begin_unreliable_shm_out;
  }
  pWal->readLock = 0;

  /* Check to see if a separate writer has attached to the shared-memory area,
  ** thus making the shared-memory "reliable" again.  Do this by invoking
  ** the xShmMap() routine of the VFS and looking to see if the return
  ** is SQLITE_READONLY instead of SQLITE_READONLY_CANTINIT.
  **
  ** If the shared-memory is now "reliable" return WAL_RETRY, which will
  ** cause the heap-memory WAL-index to be discarded and the actual
  ** shared memory to be used in its place.
  **
  ** This step is important because, even though this connection is holding
  ** the WAL_READ_LOCK(0) which prevents a checkpoint, a writer might
  ** have already checkpointed the WAL file and, while the current
  ** is active, wrap the WAL and start overwriting frames that this
  ** process wants to use.
  **
  ** Once sqlite3OsShmMap() has been called for an sqlite3_file and has
  ** returned any SQLITE_READONLY value, it must return only SQLITE_READONLY
  ** or SQLITE_READONLY_CANTINIT or some error for all subsequent invocations,
  ** even if some external agent does a "chmod" to make the shared-memory
  ** writable by us, until sqlite3OsShmUnmap() has been called.
  ** This is a requirement on the VFS implementation.
   */
  rc = sqlite3OsShmMap(pWal->pDbFd, 0, WALINDEX_PGSZ, 0, &pDummy);
  assert( rc!=SQLITE_OK ); /* SQLITE_OK not possible for read-only connection */
  if( rc!=SQLITE_READONLY_CANTINIT ){
    rc = (rc==SQLITE_READONLY ? WAL_RETRY : rc);
    goto begin_unreliable_shm_out;
  }

  /* We reach this point only if the real shared-memory is still unreliable.
  ** Assume the in-memory WAL-index substitute is correct and load it
  ** into pWal->hdr.
  */
  memcpy(&pWal->hdr, (void*)walIndexHdr(pWal), sizeof(WalIndexHdr));

  /* Make sure some writer hasn't come in and changed the WAL file out
  ** from under us, then disconnected, while we were not looking.
  */
  rc = sqlite3OsFileSize(pWal->pWalFd, &szWal);
  if( rc!=SQLITE_OK ){
    goto begin_unreliable_shm_out;
  }
  if( szWal<WAL_HDRSIZE ){
    /* If the wal file is too small to contain a wal-header and the
    ** wal-index header has mxFrame==0, then it must be safe to proceed
    ** reading the database file only. However, the page cache cannot
    ** be trusted, as a read/write connection may have connected, written
    ** the db, run a checkpoint, truncated the wal file and disconnected
    ** since this client's last read transaction.  */
    *pChanged = 1;
    rc = (pWal->hdr.mxFrame==0 ? SQLITE_OK : WAL_RETRY);
    goto begin_unreliable_shm_out;
  }

  /* Check the salt keys at the start of the wal file still match. */
  rc = sqlite3OsRead(pWal->pWalFd, aBuf, WAL_HDRSIZE, 0);
  if( rc!=SQLITE_OK ){
    goto begin_unreliable_shm_out;
  }
  if( memcmp(&pWal->hdr.aSalt, &aBuf[16], 8) ){
    /* Some writer has wrapped the WAL file while we were not looking.
    ** Return WAL_RETRY which will cause the in-memory WAL-index to be
    ** rebuilt. */
    rc = WAL_RETRY;
    goto begin_unreliable_shm_out;
  }

  /* Allocate a buffer to read frames into */
  szFrame = pWal->hdr.szPage + WAL_FRAME_HDRSIZE;
  aFrame = (u8 *)sqlite3_malloc64(szFrame);
  if( aFrame==0 ){
    rc = SQLITE_NOMEM_BKPT;
    goto begin_unreliable_shm_out;
  }
  aData = &aFrame[WAL_FRAME_HDRSIZE];

  /* Check to see if a complete transaction has been appended to the
  ** wal file since the heap-memory wal-index was created. If so, the
  ** heap-memory wal-index is discarded and WAL_RETRY returned to
  ** the caller.  */
  aSaveCksum[0] = pWal->hdr.aFrameCksum[0];
  aSaveCksum[1] = pWal->hdr.aFrameCksum[1];
  for(iOffset=walFrameOffset(pWal->hdr.mxFrame+1, pWal->hdr.szPage); 
      iOffset+szFrame<=szWal; 
      iOffset+=szFrame
  ){
    u32 pgno;                   /* Database page number for frame */
    u32 nTruncate;              /* dbsize field from frame header */

    /* Read and decode the next log frame. */
    rc = sqlite3OsRead(pWal->pWalFd, aFrame, szFrame, iOffset);
    if( rc!=SQLITE_OK ) break;
    if( !walDecodeFrame(pWal, &pgno, &nTruncate, aData, aFrame) ) break;

    /* If nTruncate is non-zero, then a complete transaction has been
    ** appended to this wal file. Set rc to WAL_RETRY and break out of
    ** the loop.  */
    if( nTruncate ){
      rc = WAL_RETRY;
      break;
    }
  }
  pWal->hdr.aFrameCksum[0] = aSaveCksum[0];
  pWal->hdr.aFrameCksum[1] = aSaveCksum[1];

 begin_unreliable_shm_out:
  sqlite3_free(aFrame);
  if( rc!=SQLITE_OK ){
    int i;
    for(i=0; i<pWal->nWiData; i++){
      sqlite3_free((void*)pWal->apWiData[i]);
      pWal->apWiData[i] = 0;
    }
    pWal->bShmUnreliable = 0;
    sqlite3WalEndReadTransaction(pWal);
    *pChanged = 1;
  }
  return rc;
}

/*
** Attempt to start a read transaction.  This might fail due to a race or
** other transient condition.  When that happens, it returns WAL_RETRY to
** indicate to the caller that it is safe to retry immediately.
**
** On success return SQLITE_OK.  On a permanent failure (such an
................................................................................
  int mxI;                        /* Index of largest aReadMark[] value */
  int i;                          /* Loop counter */
  int rc = SQLITE_OK;             /* Return code  */
  u32 mxFrame;                    /* Wal frame to lock to */

  assert( pWal->readLock<0 );     /* Not currently locked */

  /* useWal may only be set for read/write connections */
  assert( (pWal->readOnly & WAL_SHM_RDONLY)==0 || useWal==0 );

  /* Take steps to avoid spinning forever if there is a protocol error.
  **
  ** Circumstances that cause a RETRY should only last for the briefest
  ** instances of time.  No I/O or other system calls are done while the
  ** locks are held, so the locks should not be held for very long. But 
  ** if we are unlucky, another process that is holding a lock might get
  ** paged out or take a page-fault that is time-consuming to resolve, 
................................................................................
      return SQLITE_PROTOCOL;
    }
    if( cnt>=10 ) nDelay = (cnt-9)*(cnt-9)*39;
    sqlite3OsSleep(pWal->pVfs, nDelay);
  }

  if( !useWal ){
    assert( rc==SQLITE_OK );
    if( pWal->bShmUnreliable==0 ){
      rc = walIndexReadHdr(pWal, pChanged);
    }
    if( rc==SQLITE_BUSY ){
      /* If there is not a recovery running in another thread or process
      ** then convert BUSY errors to WAL_RETRY.  If recovery is known to
      ** be running, convert BUSY to BUSY_RECOVERY.  There is a race here
      ** which might cause WAL_RETRY to be returned even if BUSY_RECOVERY
      ** would be technically correct.  But the race is benign since with
      ** WAL_RETRY this routine will be called again and will probably be
................................................................................
      }else if( rc==SQLITE_BUSY ){
        rc = SQLITE_BUSY_RECOVERY;
      }
    }
    if( rc!=SQLITE_OK ){
      return rc;
    }
    else if( pWal->bShmUnreliable ){
      return walBeginShmUnreliable(pWal, pChanged);
    }
  }

  assert( pWal->nWiData>0 );
  assert( pWal->apWiData[0]!=0 );
  pInfo = walCkptInfo(pWal);
  if( !useWal && pInfo->nBackfill==pWal->hdr.mxFrame
#ifdef SQLITE_ENABLE_SNAPSHOT
   && (pWal->pSnapshot==0 || pWal->hdr.mxFrame==0
     || 0==memcmp(&pWal->hdr, pWal->pSnapshot, sizeof(WalIndexHdr)))
#endif
  ){
    /* The WAL has been completely backfilled (or it is empty).
    ** and can be safely ignored.
................................................................................
      }else if( rc!=SQLITE_BUSY ){
        return rc;
      }
    }
  }
  if( mxI==0 ){
    assert( rc==SQLITE_BUSY || (pWal->readOnly & WAL_SHM_RDONLY)!=0 );
    return rc==SQLITE_BUSY ? WAL_RETRY : SQLITE_READONLY_CANTINIT;
  }

  rc = walLockShared(pWal, WAL_READ_LOCK(mxI));
  if( rc ){
    return rc==SQLITE_BUSY ? WAL_RETRY : rc;
  }
  /* Now that the read-lock has been obtained, check that neither the
................................................................................

  /* If the "last page" field of the wal-index header snapshot is 0, then
  ** no data will be read from the wal under any circumstances. Return early
  ** in this case as an optimization.  Likewise, if pWal->readLock==0, 
  ** then the WAL is ignored by the reader so return early, as if the 
  ** WAL were empty.
  */
  if( iLast==0 || (pWal->readLock==0 && pWal->bShmUnreliable==0) ){
    *piRead = 0;
    return SQLITE_OK;
  }

  /* Search the hash table or tables for an entry matching page number
  ** pgno. Each iteration of the following for() loop searches one
  ** hash table (each hash table indexes up to HASHTABLE_NPAGE frames).
................................................................................
#ifdef SQLITE_ENABLE_EXPENSIVE_ASSERT
  /* If expensive assert() statements are available, do a linear search
  ** of the wal-index file content. Make sure the results agree with the
  ** result obtained using the hash indexes above.  */
  {
    u32 iRead2 = 0;
    u32 iTest;
    assert( pWal->bShmUnreliable || pWal->minFrame>0 );
    for(iTest=iLast; iTest>=pWal->minFrame && iTest>0; iTest--){
      if( walFramePgno(pWal, iTest)==pgno ){
        iRead2 = iTest;
        break;
      }
    }
    assert( iRead==iRead2 );
  }
................................................................................
  ** locks are taken in this case). Nor should the pager attempt to
  ** upgrade to exclusive-mode following such an error.
  */
  assert( pWal->readLock>=0 || pWal->lockError );
  assert( pWal->readLock>=0 || (op<=0 && pWal->exclusiveMode==0) );

  if( op==0 ){
    if( pWal->exclusiveMode!=WAL_NORMAL_MODE ){
      pWal->exclusiveMode = WAL_NORMAL_MODE;
      if( walLockShared(pWal, WAL_READ_LOCK(pWal->readLock))!=SQLITE_OK ){
        pWal->exclusiveMode = WAL_EXCLUSIVE_MODE;
      }
      rc = pWal->exclusiveMode==WAL_NORMAL_MODE;
    }else{
      /* Already in locking_mode=NORMAL */
      rc = 0;
    }
  }else if( op>0 ){
    assert( pWal->exclusiveMode==WAL_NORMAL_MODE );
    assert( pWal->readLock>=0 );
    walUnlockShared(pWal, WAL_READ_LOCK(pWal->readLock));
    pWal->exclusiveMode = WAL_EXCLUSIVE_MODE;
    rc = 1;
  }else{
    rc = pWal->exclusiveMode==WAL_NORMAL_MODE;
  }
  return rc;
}

/* 
** 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

Changes to test/wal2.test.

118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
...
389
390
391
392
393
394
395
396

397

398
399
400
401
402
403
404
...
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
....
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
  }
} {4 10}
do_test wal2-1.1 {
  execsql { SELECT count(a), sum(a) FROM t1 } db2
} {4 10}

set RECOVER [list                                      \
  {0 1 lock exclusive}   {1 7 lock exclusive}          \
  {1 7 unlock exclusive} {0 1 unlock exclusive}        \
]
set READ [list                                         \
  {4 1 lock shared}    {4 1 unlock shared}             \
]
set INITSLOT [list                                     \
  {4 1 lock exclusive} {4 1 unlock exclusive}          \
]
................................................................................
# UPDATE: This has now changed. When running a checkpoint, if recovery is
# required the client grabs all exclusive locks (just as it would for a
# recovery performed as a pre-cursor to a normal database transaction).
#
set expected_locks [list]
lappend expected_locks {1 1 lock exclusive}   ;# Lock checkpoint
lappend expected_locks {0 1 lock exclusive}   ;# Lock writer
lappend expected_locks {2 6 lock exclusive}   ;# Lock recovery & all aReadMark[]

lappend expected_locks {2 6 unlock exclusive} ;# Unlock recovery & aReadMark[]

lappend expected_locks {0 1 unlock exclusive} ;# Unlock writer
lappend expected_locks {3 1 lock exclusive}   ;# Lock aReadMark[0]
lappend expected_locks {3 1 unlock exclusive} ;# Unlock aReadMark[0]
lappend expected_locks {1 1 unlock exclusive} ;# Unlock checkpoint
do_test wal2-5.1 {
  proc tvfs_cb {method args} {
    set ::shm_file [lindex $args 0]
................................................................................
  testvfs tvfs
  tvfs script tvfs_cb
  sqlite3 db test.db -vfs tvfs
  set {} {}
} {}

set RECOVERY {
  {0 1 lock exclusive} {1 7 lock exclusive} 
  {1 7 unlock exclusive} {0 1 unlock exclusive}
}
set READMARK0_READ {
  {3 1 lock shared} {3 1 unlock shared}
}
set READMARK0_WRITE {
  {3 1 lock shared} 
  {0 1 lock exclusive} {3 1 unlock shared} 
................................................................................
    list [file exists test.db-shm] [file exists test.db-wal]
  } {1 1}
  faultsim_save_and_close

  foreach {tn db_perm wal_perm shm_perm can_open can_read can_write} {
    2   00644   00644   00644   1   1   1
    3   00644   00400   00644   1   1   0
    4   00644   00644   00400   1   0   0
    5   00400   00644   00644   1   1   0

    7   00644   00000   00644   1   0   0
    8   00644   00644   00000   1   0   0
    9   00000   00644   00644   0   0   0
  } {
    faultsim_restore







|
|







 







|
>
|
>







 







|
|







 







|







118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
...
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
...
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
....
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
  }
} {4 10}
do_test wal2-1.1 {
  execsql { SELECT count(a), sum(a) FROM t1 } db2
} {4 10}

set RECOVER [list                                      \
  {0 1 lock exclusive}   {1 2 lock exclusive} {4 4 lock exclusive} \
  {1 2 unlock exclusive} {4 4 unlock exclusive} {0 1 unlock exclusive}  \
]
set READ [list                                         \
  {4 1 lock shared}    {4 1 unlock shared}             \
]
set INITSLOT [list                                     \
  {4 1 lock exclusive} {4 1 unlock exclusive}          \
]
................................................................................
# UPDATE: This has now changed. When running a checkpoint, if recovery is
# required the client grabs all exclusive locks (just as it would for a
# recovery performed as a pre-cursor to a normal database transaction).
#
set expected_locks [list]
lappend expected_locks {1 1 lock exclusive}   ;# Lock checkpoint
lappend expected_locks {0 1 lock exclusive}   ;# Lock writer
lappend expected_locks {2 1 lock exclusive}   ;# Lock recovery
lappend expected_locks {4 4 lock exclusive}   ;# Lock all aReadMark[]
lappend expected_locks {2 1 unlock exclusive} ;# Unlock recovery 
lappend expected_locks {4 4 unlock exclusive} ;# Unlock all aReadMark[] 
lappend expected_locks {0 1 unlock exclusive} ;# Unlock writer
lappend expected_locks {3 1 lock exclusive}   ;# Lock aReadMark[0]
lappend expected_locks {3 1 unlock exclusive} ;# Unlock aReadMark[0]
lappend expected_locks {1 1 unlock exclusive} ;# Unlock checkpoint
do_test wal2-5.1 {
  proc tvfs_cb {method args} {
    set ::shm_file [lindex $args 0]
................................................................................
  testvfs tvfs
  tvfs script tvfs_cb
  sqlite3 db test.db -vfs tvfs
  set {} {}
} {}

set RECOVERY {
  {0 1 lock exclusive} {1 2 lock exclusive} {4 4 lock exclusive}
  {1 2 unlock exclusive} {4 4 unlock exclusive} {0 1 unlock exclusive}
}
set READMARK0_READ {
  {3 1 lock shared} {3 1 unlock shared}
}
set READMARK0_WRITE {
  {3 1 lock shared} 
  {0 1 lock exclusive} {3 1 unlock shared} 
................................................................................
    list [file exists test.db-shm] [file exists test.db-wal]
  } {1 1}
  faultsim_save_and_close

  foreach {tn db_perm wal_perm shm_perm can_open can_read can_write} {
    2   00644   00644   00644   1   1   1
    3   00644   00400   00644   1   1   0
    4   00644   00644   00400   1   1   0
    5   00400   00644   00644   1   1   0

    7   00644   00000   00644   1   0   0
    8   00644   00644   00000   1   0   0
    9   00000   00644   00644   0   0   0
  } {
    faultsim_restore

Changes to test/walro.test.

97
98
99
100
101
102
103

104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
...
134
135
136
137
138
139
140




141
142
143
144
145
146
147
148
149
150
151
152
...
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
  do_test 1.1.13  { sql2 "INSERT INTO t1 VALUES('i', 'j')" } {}

  do_test 1.2.1 {
    code2 { db2 close }
    code1 { db close }
    list [file exists test.db-wal] [file exists test.db-shm]
  } {1 1}

  do_test 1.2.2 {
    code1 { sqlite3 db file:test.db?readonly_shm=1 }
    sql1 { SELECT * FROM t1 }
  } {a b c d e f g h i j}

  do_test 1.2.3 {
    code1 { db close }
    file attributes test.db-shm -permissions rw-r--r--
    hexio_write test.db-shm 0 01020304 
    file attributes test.db-shm -permissions r--r--r--
    code1 { sqlite3 db file:test.db?readonly_shm=1 }
    csql1 { SELECT * FROM t1 }
  } {1 {attempt to write a readonly database}}
  do_test 1.2.4 {
    code1 { sqlite3_extended_errcode db } 
  } {SQLITE_READONLY_RECOVERY}

  do_test 1.2.5 {
    file attributes test.db-shm -permissions rw-r--r--
    code2 { sqlite3 db2 test.db }
    sql2 "SELECT * FROM t1" 
  } {a b c d e f g h i j}
  file attributes test.db-shm -permissions r--r--r--
................................................................................
    set {} {}
  } {}
  do_test 1.2.8 { sql1 "SELECT * FROM t1" } {a b c d e f g h i j k l}

  # Now check that if the readonly_shm option is not supplied, or if it
  # is set to zero, it is not possible to connect to the database without
  # read-write access to the shm.




  do_test 1.3.1 {
    code1 { db close }
    code1 { sqlite3 db test.db }
    csql1 { SELECT * FROM t1 }
  } {1 {unable to open database file}}

  # Also test that if the -shm file can be opened for read/write access,
  # it is not if readonly_shm=1 is present in the URI.
  do_test 1.3.2.1 {
    code1 { db close }
    code2 { db2 close }
    file exists test.db-shm
................................................................................
  } {1 {unable to open database file}}
  do_test 1.3.2.3 {
    code1 { db close }
    close [open test.db-shm w]
    file attributes test.db-shm -permissions r--r--r--
    code1 { sqlite3 db file:test.db?readonly_shm=1 }
    csql1 { SELECT * FROM t1 }
  } {1 {attempt to write a readonly database}}
  do_test 1.3.2.4 {
    code1 { sqlite3_extended_errcode db } 
  } {SQLITE_READONLY_RECOVERY}

  #-----------------------------------------------------------------------
  # Test cases 1.4.* check that checkpoints and log wraps don't prevent
  # read-only connections from reading the database.
  do_test 1.4.1 {
    code1 { db close }
    forcedelete test.db-shm







>


|
|








|


|







 







>
>
>
>




|







 







|


|







97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
...
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
...
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
  do_test 1.1.13  { sql2 "INSERT INTO t1 VALUES('i', 'j')" } {}

  do_test 1.2.1 {
    code2 { db2 close }
    code1 { db close }
    list [file exists test.db-wal] [file exists test.db-shm]
  } {1 1}

  do_test 1.2.2 {
    code1 { sqlite3 db file:test.db?readonly_shm=1 }
    list [catch { sql1 { SELECT * FROM t1 } } msg] $msg
  } {0 {a b c d e f g h i j}}

  do_test 1.2.3 {
    code1 { db close }
    file attributes test.db-shm -permissions rw-r--r--
    hexio_write test.db-shm 0 01020304 
    file attributes test.db-shm -permissions r--r--r--
    code1 { sqlite3 db file:test.db?readonly_shm=1 }
    csql1 { SELECT * FROM t1 }
  } {0 {a b c d e f g h i j}}
  do_test 1.2.4 {
    code1 { sqlite3_extended_errcode db } 
  } {SQLITE_OK}

  do_test 1.2.5 {
    file attributes test.db-shm -permissions rw-r--r--
    code2 { sqlite3 db2 test.db }
    sql2 "SELECT * FROM t1" 
  } {a b c d e f g h i j}
  file attributes test.db-shm -permissions r--r--r--
................................................................................
    set {} {}
  } {}
  do_test 1.2.8 { sql1 "SELECT * FROM t1" } {a b c d e f g h i j k l}

  # Now check that if the readonly_shm option is not supplied, or if it
  # is set to zero, it is not possible to connect to the database without
  # read-write access to the shm.
  # 
  # UPDATE: os_unix.c now opens the *-shm file in readonly mode 
  # automatically.
  #
  do_test 1.3.1 {
    code1 { db close }
    code1 { sqlite3 db test.db }
    csql1 { SELECT * FROM t1 }
  } {0 {a b c d e f g h i j k l}}

  # Also test that if the -shm file can be opened for read/write access,
  # it is not if readonly_shm=1 is present in the URI.
  do_test 1.3.2.1 {
    code1 { db close }
    code2 { db2 close }
    file exists test.db-shm
................................................................................
  } {1 {unable to open database file}}
  do_test 1.3.2.3 {
    code1 { db close }
    close [open test.db-shm w]
    file attributes test.db-shm -permissions r--r--r--
    code1 { sqlite3 db file:test.db?readonly_shm=1 }
    csql1 { SELECT * FROM t1 }
  } {0 {a b c d e f g h i j k l}}
  do_test 1.3.2.4 {
    code1 { sqlite3_extended_errcode db } 
  } {SQLITE_OK}

  #-----------------------------------------------------------------------
  # Test cases 1.4.* check that checkpoints and log wraps don't prevent
  # read-only connections from reading the database.
  do_test 1.4.1 {
    code1 { db close }
    forcedelete test.db-shm

Added test/walro2.test.



















































































































































































































































































































































































































































































































































































































































































































































































































>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
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
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
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
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
# 2011 May 09
#
# The author disclaims copyright to this source code.  In place of
# a legal notice, here is a blessing:
#
#    May you do good and not evil.
#    May you find forgiveness for yourself and forgive others.
#    May you share freely, never taking more than you give.
#
#***********************************************************************
#
# This file contains tests for using WAL databases in read-only mode.
#

set testdir [file dirname $argv0]
source $testdir/tester.tcl
source $testdir/lock_common.tcl
source $testdir/wal_common.tcl
set ::testprefix walro2

# And only if the build is WAL-capable.
#
ifcapable !wal {
  finish_test
  return
}

proc copy_to_test2 {bZeroShm} {
  forcecopy test.db test.db2
  forcecopy test.db-wal test.db2-wal
  if {$bZeroShm} {
    forcedelete test.db2-shm
    set fd [open test.db2-shm w]
    seek $fd [expr [file size test.db-shm]-1]
    puts -nonewline $fd "\0"
    close $fd
  } else {
    forcecopy test.db-shm test.db2-shm
  }
}

foreach bZeroShm {0 1} {
set TN [expr $bZeroShm+1]
do_multiclient_test tn {
  
  # Close all connections and delete the database.
  #
  code1 { db close  }
  code2 { db2 close }
  code3 { db3 close }
  forcedelete test.db
  
  # Do not run tests with the connections in the same process.
  #
  if {$tn==2} continue

  foreach c {code1 code2 code3} {
    $c {
      sqlite3_shutdown
      sqlite3_config_uri 1
    }
  }

  do_test $TN.1.1 {
    code2 { sqlite3 db2 test.db }
    sql2 { 
      CREATE TABLE t1(x, y);
      PRAGMA journal_mode = WAL;
      INSERT INTO t1 VALUES('a', 'b');
      INSERT INTO t1 VALUES('c', 'd');
    }
    file exists test.db-shm
  } {1}

  do_test $TN.1.2.1 {
    copy_to_test2 $bZeroShm
    code1 {
      sqlite3 db file:test.db2?readonly_shm=1
    }

    sql1 { SELECT * FROM t1 }
  } {a b c d}
  do_test $TN.1.2.2 {
    sql1 { SELECT * FROM t1 }
  } {a b c d}

  do_test $TN.1.3.1 {
    code3 { sqlite3 db3 test.db2 }
    sql3 { SELECT * FROM t1 }
  } {a b c d}

  do_test $TN.1.3.2 {
    sql1 { SELECT * FROM t1 }
  } {a b c d}

  code1 { db close  }
  code2 { db2 close }
  code3 { db3 close }

  do_test $TN.2.1 {
    code2 { sqlite3 db2 test.db }
    sql2 { 
      INSERT INTO t1 VALUES('e', 'f');
      INSERT INTO t1 VALUES('g', 'h');
    }
    file exists test.db-shm
  } {1}

  do_test $TN.2.2 {
    copy_to_test2 $bZeroShm
    code1 {
      sqlite3 db file:test.db2?readonly_shm=1
    }
    sql1 { 
      BEGIN;
      SELECT * FROM t1;
    }
  } {a b c d e f g h}

  do_test $TN.2.3.1 {
    code3 { sqlite3 db3 test.db2 }
    sql3 { SELECT * FROM t1 }
  } {a b c d e f g h}
  do_test $TN.2.3.2 {
    sql3 { INSERT INTO t1 VALUES('i', 'j') }
    code3 { db3 close }
    sql1 { COMMIT } 
  } {}
  do_test $TN.2.3.3 {
    sql1 { SELECT * FROM t1 }
  } {a b c d e f g h i j}


  #-----------------------------------------------------------------------
  # 3.1.*: That a readonly_shm connection can read a database file if both
  #        the *-wal and *-shm files are zero bytes in size.
  #
  # 3.2.*: That it flushes the cache if, between transactions on a db with a
  #        zero byte *-wal file, some other connection modifies the db, then
  #        does "PRAGMA wal_checkpoint=truncate" to truncate the wal file
  #        back to zero bytes in size.
  #
  # 3.3.*: That, if between transactions some other process wraps the wal
  #        file, the readonly_shm client reruns recovery.
  #
  catch { code1 { db close } }
  catch { code2 { db2 close } }
  catch { code3 { db3 close } }
  do_test $TN.3.1.0 {
    list [file exists test.db-wal] [file exists test.db-shm]
  } {0 0}
  do_test $TN.3.1.1 {
    close [open test.db-wal w]
    close [open test.db-shm w]
    code1 {
      sqlite3 db file:test.db?readonly_shm=1
    }
    sql1 { SELECT * FROM t1 }
  } {a b c d e f g h}

  do_test $TN.3.2.0 {
    list [file size test.db-wal] [file size test.db-shm]
  } {0 0}
  do_test $TN.3.2.1 {
    code2 { sqlite3 db2 test.db }
    sql2 { INSERT INTO t1 VALUES(1, 2) ; PRAGMA wal_checkpoint=truncate }
    code2 { db2 close }
    sql1 { SELECT * FROM t1 }
  } {a b c d e f g h 1 2}
  do_test $TN.3.2.2 {
    list [file size test.db-wal] [file size test.db-shm]
  } {0 32768}

  do_test $TN.3.3.0 {
    code2 { sqlite3 db2 test.db }
    sql2 { 
      INSERT INTO t1 VALUES(3, 4);
      INSERT INTO t1 VALUES(5, 6);
      INSERT INTO t1 VALUES(7, 8);
      INSERT INTO t1 VALUES(9, 10);
    }
    code2 { db2 close }
    code1 { db close }
    list [file size test.db-wal] [file size test.db-shm]
  } [list [wal_file_size 4 1024] 32768]
  do_test $TN.3.3.1 {
    code1 { sqlite3 db file:test.db?readonly_shm=1 }
    sql1 { SELECT * FROM t1 }
  } {a b c d e f g h 1 2 3 4 5 6 7 8 9 10}
  do_test $TN.3.3.2 {
    code2 { sqlite3 db2 test.db }
    sql2 { 
      PRAGMA wal_checkpoint; 
      DELETE FROM t1;
      INSERT INTO t1 VALUES('i', 'ii');
    }
    code2 { db2 close }
    list [file size test.db-wal] [file size test.db-shm]
  } [list [wal_file_size 4 1024] 32768]
  do_test $TN.3.3.3 {
    sql1 { SELECT * FROM t1 }
  } {i ii}

  #-----------------------------------------------------------------------
  #
  #
  catch { code1 { db close } }
  catch { code2 { db2 close } }
  catch { code3 { db3 close } }

  do_test $TN.4.0 {
    code1 { forcedelete test.db }
    code1 { sqlite3 db test.db }
    sql1 {
      PRAGMA journal_mode = wal;
      CREATE TABLE t1(x);
      INSERT INTO t1 VALUES('hello');
      INSERT INTO t1 VALUES('world');
    }

    copy_to_test2 $bZeroShm

    code1 { db close }
  } {}

  do_test $TN.4.1.1 {
    code2 { sqlite3 db2 file:test.db2?readonly_shm=1 }
    sql2 { SELECT * FROM t1 }
  } {hello world}

  do_test $TN.4.1.2 {
    code3 { sqlite3 db3 test.db2 }
    sql3 {
      INSERT INTO t1 VALUES('!');
      PRAGMA wal_checkpoint = truncate;
    }
    code3 { db3 close }
  } {}
  do_test $TN.4.1.3 {
    sql2 { SELECT * FROM t1 }
  } {hello world !}

  catch { code1 { db close } }
  catch { code2 { db2 close } }
  catch { code3 { db3 close } }

  do_test $TN.4.2.1 {
    code1 { sqlite3 db test.db }
    sql1 {
      INSERT INTO t1 VALUES('!');
      INSERT INTO t1 VALUES('!');

      PRAGMA cache_size = 10;
      CREATE TABLE t2(x);

      BEGIN;
        WITH s(i) AS (
          SELECT 1 UNION ALL SELECT i+1 FROM s WHERE i<500
          )
        INSERT INTO t2 SELECT randomblob(500) FROM s;
        SELECT count(*) FROM t2;
    } 
  } {500}
  do_test $TN.4.2.2 {
    file size test.db-wal
  } {461152}
  do_test $TN.4.2.4 {
    file_control_persist_wal db 1; db close

    copy_to_test2 $bZeroShm
    code2 { sqlite3 db2 file:test.db2?readonly_shm=1 }
    sql2 {
      SELECT * FROM t1;
      SELECT count(*) FROM t2;
    }
  } {hello world ! ! 0}

  #-----------------------------------------------------------------------
  #
  #
  catch { code1 { db close } }
  catch { code2 { db2 close } }
  catch { code3 { db3 close } }

  do_test $TN.5.0 {
    code1 { forcedelete test.db }
    code1 { sqlite3 db test.db }
    sql1 {
      PRAGMA journal_mode = wal;
      CREATE TABLE t1(x);
      INSERT INTO t1 VALUES('hello');
      INSERT INTO t1 VALUES('world');
      INSERT INTO t1 VALUES('!');
      INSERT INTO t1 VALUES('world');
      INSERT INTO t1 VALUES('hello');
    }

    copy_to_test2 $bZeroShm
    
    code1 { db close }
  } {}

  do_test $TN.5.1 {
    code2 { sqlite3 db2 file:test.db2?readonly_shm=1 }
    sql2 {
      SELECT * FROM t1;
    }
  } {hello world ! world hello}

  do_test $TN.5.2 {
    code1 {
      proc handle_read {op args} {
        if {$op=="xRead" && [file tail [lindex $args 0]]=="test.db2-wal"} {
          set ::res2 [sql2 { SELECT * FROM t1 }]
        }
        puts "$msg xRead $args"
        return "SQLITE_OK"
      }
      testvfs tvfs -fullshm 1

      sqlite3 db file:test.db2?vfs=tvfs
      db eval { SELECT * FROM sqlite_master }

      tvfs filter xRead
      tvfs script handle_read
    }
    sql1 {
      PRAGMA wal_checkpoint = truncate;
    }
    code1 { set ::res2 }
  } {hello world ! world hello}

  do_test $TN.5.3 {
    code1 { db close }
    code1 { tvfs delete }
  } {}

  #-----------------------------------------------------------------------
  #
  #
  catch { code1 { db close } }
  catch { code2 { db2 close } }
  catch { code3 { db3 close } }

  do_test $TN.6.1 {
    code1 { forcedelete test.db }
    code1 { sqlite3 db test.db }
    sql1 {
      PRAGMA journal_mode = wal;
      CREATE TABLE t1(x);
      INSERT INTO t1 VALUES('hello');
      INSERT INTO t1 VALUES('world');
      INSERT INTO t1 VALUES('!');
      INSERT INTO t1 VALUES('world');
      INSERT INTO t1 VALUES('hello');
    }

    copy_to_test2 $bZeroShm
    
    code1 { db close }
  } {}

  do_test $TN.6.2 {
    code1 {
      set ::nRem 5
      proc handle_read {op args} {
        if {$op=="xRead" && [file tail [lindex $args 0]]=="test.db2-wal"} {
          incr ::nRem -1
          if {$::nRem==0} {
            code2 { sqlite3 db2 test.db2 }
            sql2  { PRAGMA wal_checkpoint = truncate }
          }
        }
        return "SQLITE_OK"
      }
      testvfs tvfs -fullshm 1

      tvfs filter xRead
      tvfs script handle_read

      sqlite3 db file:test.db2?readonly_shm=1&vfs=tvfs
      db eval { SELECT * FROM t1 }
    }
  } {hello world ! world hello}

  do_test $TN.6.3 {
    code1 { db close }
    code1 { tvfs delete }
  } {}
}
} ;# foreach bZeroShm

finish_test

Added test/walrofault.test.

























































































































>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
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
# 2011 May 09
#
# The author disclaims copyright to this source code.  In place of
# a legal notice, here is a blessing:
#
#    May you do good and not evil.
#    May you find forgiveness for yourself and forgive others.
#    May you share freely, never taking more than you give.
#
#***********************************************************************
#
# This file contains tests for using WAL databases in read-only mode.
#

set testdir [file dirname $argv0]
source $testdir/tester.tcl
source $testdir/malloc_common.tcl
set ::testprefix walro2

# And only if the build is WAL-capable.
#
ifcapable !wal {
  finish_test
  return
}

db close
sqlite3_shutdown
sqlite3_config_uri 1
sqlite3 db test.db

do_execsql_test 1.0 {
  CREATE TABLE t1(b);
  PRAGMA journal_mode = wal;
  INSERT INTO t1 VALUES('hello');
  INSERT INTO t1 VALUES('world');
  INSERT INTO t1 VALUES('!');
  INSERT INTO t1 VALUES('world');
  INSERT INTO t1 VALUES('hello');
  PRAGMA cache_size = 10;
  BEGIN;
    WITH s(i) AS ( SELECT 1 UNION ALL SELECT i+1 FROM s WHERE i<30 ) 
    INSERT INTO t1(b) SELECT randomblob(800) FROM s;
} {wal}
file_control_persist_wal db 1; db close
faultsim_save_and_close

do_faultsim_test 1 -faults oom* -prep {
  catch { db close }
  faultsim_restore
  sqlite3 db file:test.db?readonly_shm=1
} -body {
  execsql { SELECT * FROM t1 }
} -test {
  faultsim_test_result {0 {hello world ! world hello}}
}



finish_test