SQLite

Check-in [a038688c99]
Login

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

Overview
Comment:Support FCNTL_CHUNK_SIZE on windows too.
Downloads: Tarball | ZIP archive
Timelines: family | ancestors | descendants | both | experimental
Files: files | file ages | folders
SHA1: a038688c991435967b935946c2283707820bb5da
User & Date: dan 2010-07-28 14:26:18.000
Context
2010-07-28
15:10
Fix errors in wal3.test caused by recent modifications. (check-in: 0714aeccd8 user: dan tags: experimental)
14:26
Support FCNTL_CHUNK_SIZE on windows too. (check-in: a038688c99 user: dan tags: experimental)
2010-07-27
18:36
Merge trunk changes into experimental branch. (check-in: 621824092d user: dan tags: experimental)
Changes
Unified Diff Ignore Whitespace Patch
Changes to src/os_unix.c.
2761
2762
2763
2764
2765
2766
2767
2768
2769
2770
2771
2772
2773
2774
2775
2776
2777
2778
2779
2780
2781
2782
2783
2784
2785
2786
2787
2788
2789
2790
2791
2792
2793
2794
2795
2796
2797
2798
2799
2800
2801
2802
2803
2804
2805
2806
2807
2808
2809
2810
    amt -= wrote;
    offset += wrote;
    pBuf = &((char*)pBuf)[wrote];
  }
  SimulateIOError(( wrote=(-1), amt=1 ));
  SimulateDiskfullError(( wrote=0, amt=1 ));

  /* If the user has configured a chunk-size for this file, it could be
  ** that the file needs to be extended at this point. 
  */
  if( pFile->szChunk && amt==0 ){
    i64 nSize;                    /* Required file size */
    struct stat buf;              /* Used to hold return values of fstat() */
    int rc = fstat(pFile->h, &buf);
    if( rc!=0 ) return SQLITE_IOERR_FSTAT;
    nSize = ((offset+amt+pFile->szChunk-1) / pFile->szChunk) * pFile->szChunk;
    if( nSize>(i64)buf.st_size ){
#ifdef HAVE_POSIX_FALLOCATE
      if( posix_fallocate(pFile->h, buf.st_size, nSize-buf.st_size) ){
        return SQLITE_IOERR_WRITE;
      }
#else
      /* If the OS does not have posix_fallocate(), fake it. First use
      ** ftruncate() to set the file size, then write a single byte to
      ** the last byte in each block within the extended region.
      */
      int nBlk = buf.st_blksize;  /* File-system block size */
      i64 iWrite;                 /* Next offset to write to */

      if( ftruncate(pFile->h, nSize) ){
	pFile->lastErrno = errno;
	return SQLITE_IOERR_TRUNCATE;
      }
      iWrite = ((buf.st_size + 2*nBlk - 1)/nBlk)*nBlk-1;
      do {
	wrote = seekAndWrite(pFile, iWrite, "", 1);
	iWrite += nBlk;
      } while( wrote==1 && iWrite<nSize );
      if( wrote!=1 ) amt = 1;
#endif
    }
  }

  if( amt>0 ){
    if( wrote<0 ){
      /* lastErrno set by seekAndWrite */
      return SQLITE_IOERR_WRITE;
    }else{
      pFile->lastErrno = 0; /* not a system error */
      return SQLITE_FULL;







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







2761
2762
2763
2764
2765
2766
2767




































2768
2769
2770
2771
2772
2773
2774
    amt -= wrote;
    offset += wrote;
    pBuf = &((char*)pBuf)[wrote];
  }
  SimulateIOError(( wrote=(-1), amt=1 ));
  SimulateDiskfullError(( wrote=0, amt=1 ));





































  if( amt>0 ){
    if( wrote<0 ){
      /* lastErrno set by seekAndWrite */
      return SQLITE_IOERR_WRITE;
    }else{
      pFile->lastErrno = 0; /* not a system error */
      return SQLITE_FULL;
3079
3080
3081
3082
3083
3084
3085
















































3086
3087
3088
3089
3090
3091
3092
3093
3094
3095
3096
3097
3098
3099
3100
3101
3102
3103
3104
3105
3106
3107
3108
3109
3110
3111
3112
3113
3114
3115
3116
3117
/*
** Handler for proxy-locking file-control verbs.  Defined below in the
** proxying locking division.
*/
static int proxyFileControl(sqlite3_file*,int,void*);
#endif


















































/*
** Information and control of an open file handle.
*/
static int unixFileControl(sqlite3_file *id, int op, void *pArg){
  switch( op ){
    case SQLITE_FCNTL_LOCKSTATE: {
      *(int*)pArg = ((unixFile*)id)->eFileLock;
      return SQLITE_OK;
    }
    case SQLITE_LAST_ERRNO: {
      *(int*)pArg = ((unixFile*)id)->lastErrno;
      return SQLITE_OK;
    }
    case SQLITE_FCNTL_CHUNK_SIZE: {
      ((unixFile*)id)->szChunk = *(int *)pArg;
      return SQLITE_OK;				    
    }
    case SQLITE_FCNTL_SIZE_HINT: {
#if 0 /* No performance advantage seen on Linux */
      sqlite3_int64 szFile = *(sqlite3_int64*)pArg;
      unixFile *pFile = (unixFile*)id;
      ftruncate(pFile->h, szFile);
#endif
      return SQLITE_OK;
    }
#ifndef NDEBUG
    /* The pager calls this method to signal that it has done
    ** a rollback and that the database is therefore unchanged and
    ** it hence it is OK for the transaction change counter to be
    ** unchanged.
    */







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
















|


<
<
|
<
<
<







3043
3044
3045
3046
3047
3048
3049
3050
3051
3052
3053
3054
3055
3056
3057
3058
3059
3060
3061
3062
3063
3064
3065
3066
3067
3068
3069
3070
3071
3072
3073
3074
3075
3076
3077
3078
3079
3080
3081
3082
3083
3084
3085
3086
3087
3088
3089
3090
3091
3092
3093
3094
3095
3096
3097
3098
3099
3100
3101
3102
3103
3104
3105
3106
3107
3108
3109
3110
3111
3112
3113
3114
3115
3116


3117



3118
3119
3120
3121
3122
3123
3124
/*
** Handler for proxy-locking file-control verbs.  Defined below in the
** proxying locking division.
*/
static int proxyFileControl(sqlite3_file*,int,void*);
#endif

/* 
** This function is called to handle the SQLITE_FCNTL_SIZE_HINT 
** file-control operation.
**
** If the user has configured a chunk-size for this file, it could be
** that the file needs to be extended at this point. Otherwise, the
** SQLITE_FCNTL_SIZE_HINT operation is a no-op for Unix.
*/
static int fcntlSizeHint(unixFile *pFile, i64 nByte){
  if( pFile->szChunk ){
    i64 nSize;                    /* Required file size */
    struct stat buf;              /* Used to hold return values of fstat() */
   
    if( fstat(pFile->h, &buf) ) return SQLITE_IOERR_FSTAT;

    nSize = ((nByte+pFile->szChunk-1) / pFile->szChunk) * pFile->szChunk;
    if( nSize>(i64)buf.st_size ){
#if defined(HAVE_POSIX_FALLOCATE) && HAVE_POSIX_FALLOCATE
      if( posix_fallocate(pFile->h, buf.st_size, nSize-buf.st_size) ){
        return SQLITE_IOERR_WRITE;
      }
#else
      /* If the OS does not have posix_fallocate(), fake it. First use
      ** ftruncate() to set the file size, then write a single byte to
      ** the last byte in each block within the extended region. This
      ** is the same technique used by glibc to implement posix_fallocate()
      ** on systems that do not have a real fallocate() system call.
      */
      int nBlk = buf.st_blksize;  /* File-system block size */
      i64 iWrite;                 /* Next offset to write to */
      int nWrite;                 /* Return value from seekAndWrite() */

      if( ftruncate(pFile->h, nSize) ){
        pFile->lastErrno = errno;
        return SQLITE_IOERR_TRUNCATE;
      }
      iWrite = ((buf.st_size + 2*nBlk - 1)/nBlk)*nBlk-1;
      do {
        nWrite = seekAndWrite(pFile, iWrite, "", 1);
        iWrite += nBlk;
      } while( nWrite==1 && iWrite<nSize );
      if( nWrite!=1 ) return SQLITE_IOERR_WRITE;
#endif
    }
  }

  return SQLITE_OK;
}

/*
** Information and control of an open file handle.
*/
static int unixFileControl(sqlite3_file *id, int op, void *pArg){
  switch( op ){
    case SQLITE_FCNTL_LOCKSTATE: {
      *(int*)pArg = ((unixFile*)id)->eFileLock;
      return SQLITE_OK;
    }
    case SQLITE_LAST_ERRNO: {
      *(int*)pArg = ((unixFile*)id)->lastErrno;
      return SQLITE_OK;
    }
    case SQLITE_FCNTL_CHUNK_SIZE: {
      ((unixFile*)id)->szChunk = *(int *)pArg;
      return SQLITE_OK;
    }
    case SQLITE_FCNTL_SIZE_HINT: {


      return fcntlSizeHint((unixFile *)id, *(i64 *)pArg);



    }
#ifndef NDEBUG
    /* The pager calls this method to signal that it has done
    ** a rollback and that the database is therefore unchanged and
    ** it hence it is OK for the transaction change counter to be
    ** unchanged.
    */
Changes to src/os_win.c.
104
105
106
107
108
109
110

111
112
113
114
115
116
117
  HANDLE h;               /* Handle for accessing the file */
  unsigned char locktype; /* Type of lock currently held on this file */
  short sharedLockByte;   /* Randomly chosen byte used as a shared lock */
  DWORD lastErrno;        /* The Windows errno from the last I/O error */
  DWORD sectorSize;       /* Sector size of the device file is on */
  winShm *pShm;           /* Instance of shared memory on this file */
  const char *zPath;      /* Full pathname of this file */

#if SQLITE_OS_WINCE
  WCHAR *zDeleteOnClose;  /* Name of file to delete when closing */
  HANDLE hMutex;          /* Mutex used to control access to shared lock */  
  HANDLE hShared;         /* Shared memory segment used for locking */
  winceLock local;        /* Locks obtained by this instance of winFile */
  winceLock *shared;      /* Global shared lock memory for the file  */
#endif







>







104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
  HANDLE h;               /* Handle for accessing the file */
  unsigned char locktype; /* Type of lock currently held on this file */
  short sharedLockByte;   /* Randomly chosen byte used as a shared lock */
  DWORD lastErrno;        /* The Windows errno from the last I/O error */
  DWORD sectorSize;       /* Sector size of the device file is on */
  winShm *pShm;           /* Instance of shared memory on this file */
  const char *zPath;      /* Full pathname of this file */
  int szChunk;            /* Chunk size configured by FCNTL_CHUNK_SIZE */
#if SQLITE_OS_WINCE
  WCHAR *zDeleteOnClose;  /* Name of file to delete when closing */
  HANDLE hMutex;          /* Mutex used to control access to shared lock */  
  HANDLE hShared;         /* Shared memory segment used for locking */
  winceLock local;        /* Locks obtained by this instance of winFile */
  winceLock *shared;      /* Global shared lock memory for the file  */
#endif
615
616
617
618
619
620
621




































622
623
624
625
626
627
628
*****************************************************************************/
#endif /* SQLITE_OS_WINCE */

/*****************************************************************************
** The next group of routines implement the I/O methods specified
** by the sqlite3_io_methods object.
******************************************************************************/





































/*
** Close a file.
**
** It is reported that an attempt to close a handle might sometimes
** fail.  This is a very unreasonable result, but windows is notorious
** for being unreasonable so I do not doubt that it might happen.  If







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







616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
*****************************************************************************/
#endif /* SQLITE_OS_WINCE */

/*****************************************************************************
** The next group of routines implement the I/O methods specified
** by the sqlite3_io_methods object.
******************************************************************************/

/*
** Some microsoft compilers lack this definition.
*/
#ifndef INVALID_SET_FILE_POINTER
# define INVALID_SET_FILE_POINTER ((DWORD)-1)
#endif

/*
** Move the current position of the file handle passed as the first 
** argument to offset iOffset within the file. If successful, return 0. 
** Otherwise, set pFile->lastErrno and return non-zero.
*/
static int seekWinFile(winFile *pFile, sqlite3_int64 iOffset){
  LONG upperBits;                 /* Most sig. 32 bits of new offset */
  LONG lowerBits;                 /* Least sig. 32 bits of new offset */
  DWORD dwRet;                    /* Value returned by SetFilePointer() */

  upperBits = (LONG)((iOffset>>32) & 0x7fffffff);
  lowerBits = (LONG)(iOffset & 0xffffffff);

  /* API oddity: If successful, SetFilePointer() returns a dword 
  ** containing the lower 32-bits of the new file-offset. Or, if it fails,
  ** it returns INVALID_SET_FILE_POINTER. However according to MSDN, 
  ** INVALID_SET_FILE_POINTER may also be a valid new offset. So to determine 
  ** whether an error has actually occured, it is also necessary to call 
  ** GetLastError().
  */
  dwRet = SetFilePointer(pFile->h, lowerBits, &upperBits, FILE_BEGIN);
  if( (dwRet==INVALID_SET_FILE_POINTER && GetLastError()!=NO_ERROR) ){
    pFile->lastErrno = GetLastError();
    return 1;
  }

  return 0;
}

/*
** Close a file.
**
** It is reported that an attempt to close a handle might sometimes
** fail.  This is a very unreasonable result, but windows is notorious
** for being unreasonable so I do not doubt that it might happen.  If
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708


709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728

729
730

731
732
733
734
735
736
737
738
739





740
741
742
743
744
745
746
747
748
749
750
751





752
753
754
755
756

757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772

773
774
775








776

777
778
779
780
781
782
783

784
785
786
787
788
789
790
791
  }
#endif
  OSTRACE(("CLOSE %d %s\n", pFile->h, rc ? "ok" : "failed"));
  OpenCounter(-1);
  return rc ? SQLITE_OK : SQLITE_IOERR;
}

/*
** Some microsoft compilers lack this definition.
*/
#ifndef INVALID_SET_FILE_POINTER
# define INVALID_SET_FILE_POINTER ((DWORD)-1)
#endif

/*
** Read data from a file into a buffer.  Return SQLITE_OK if all
** bytes were read successfully and SQLITE_IOERR if anything goes
** wrong.
*/
static int winRead(
  sqlite3_file *id,          /* File to read from */
  void *pBuf,                /* Write content into this buffer */
  int amt,                   /* Number of bytes to read */
  sqlite3_int64 offset       /* Begin reading at this offset */
){
  LONG upperBits = (LONG)((offset>>32) & 0x7fffffff);
  LONG lowerBits = (LONG)(offset & 0xffffffff);
  DWORD rc;
  winFile *pFile = (winFile*)id;
  DWORD error;
  DWORD got;

  assert( id!=0 );
  SimulateIOError(return SQLITE_IOERR_READ);
  OSTRACE(("READ %d lock=%d\n", pFile->h, pFile->locktype));
  rc = SetFilePointer(pFile->h, lowerBits, &upperBits, FILE_BEGIN);
  if( rc==INVALID_SET_FILE_POINTER && (error=GetLastError())!=NO_ERROR ){
    pFile->lastErrno = error;
    return SQLITE_FULL;
  }
  if( !ReadFile(pFile->h, pBuf, amt, &got, 0) ){
    pFile->lastErrno = GetLastError();
    return SQLITE_IOERR_READ;
  }
  if( got==(DWORD)amt ){
    return SQLITE_OK;
  }else{
    /* Unread parts of the buffer must be zero-filled */
    memset(&((char*)pBuf)[got], 0, amt-got);
    return SQLITE_IOERR_SHORT_READ;
  }


}

/*
** Write data from a buffer into a file.  Return SQLITE_OK on success
** or some other error code on failure.
*/
static int winWrite(
  sqlite3_file *id,         /* File to write into */
  const void *pBuf,         /* The bytes to be written */
  int amt,                  /* Number of bytes to write */
  sqlite3_int64 offset      /* Offset into the file to begin writing at */
){
  LONG upperBits = (LONG)((offset>>32) & 0x7fffffff);
  LONG lowerBits = (LONG)(offset & 0xffffffff);
  DWORD rc;
  winFile *pFile = (winFile*)id;
  DWORD error;
  DWORD wrote = 0;

  assert( id!=0 );

  SimulateIOError(return SQLITE_IOERR_WRITE);
  SimulateDiskfullError(return SQLITE_FULL);

  OSTRACE(("WRITE %d lock=%d\n", pFile->h, pFile->locktype));
  rc = SetFilePointer(pFile->h, lowerBits, &upperBits, FILE_BEGIN);
  if( rc==INVALID_SET_FILE_POINTER && (error=GetLastError())!=NO_ERROR ){
    pFile->lastErrno = error;
    if( pFile->lastErrno==ERROR_HANDLE_DISK_FULL ){
      return SQLITE_FULL;
    }else{
      return SQLITE_IOERR_WRITE;
    }





  }
  assert( amt>0 );
  while(
     amt>0
     && (rc = WriteFile(pFile->h, pBuf, amt, &wrote, 0))!=0
     && wrote>0
  ){
    amt -= wrote;
    pBuf = &((char*)pBuf)[wrote];
  }
  if( !rc || amt>(int)wrote ){
    pFile->lastErrno = GetLastError();





    if( pFile->lastErrno==ERROR_HANDLE_DISK_FULL ){
      return SQLITE_FULL;
    }else{
      return SQLITE_IOERR_WRITE;
    }

  }
  return SQLITE_OK;
}

/*
** Truncate an open file to a specified size
*/
static int winTruncate(sqlite3_file *id, sqlite3_int64 nByte){
  LONG upperBits = (LONG)((nByte>>32) & 0x7fffffff);
  LONG lowerBits = (LONG)(nByte & 0xffffffff);
  DWORD dwRet;
  winFile *pFile = (winFile*)id;
  DWORD error;
  int rc = SQLITE_OK;

  assert( id!=0 );

  OSTRACE(("TRUNCATE %d %lld\n", pFile->h, nByte));
  SimulateIOError(return SQLITE_IOERR_TRUNCATE);
  dwRet = SetFilePointer(pFile->h, lowerBits, &upperBits, FILE_BEGIN);








  if( dwRet==INVALID_SET_FILE_POINTER && (error=GetLastError())!=NO_ERROR ){

    pFile->lastErrno = error;
    rc = SQLITE_IOERR_TRUNCATE;
  /* SetEndOfFile will fail if nByte is negative */
  }else if( !SetEndOfFile(pFile->h) ){
    pFile->lastErrno = GetLastError();
    rc = SQLITE_IOERR_TRUNCATE;
  }

  OSTRACE(("TRUNCATE %d %lld %s\n", pFile->h, nByte, rc==SQLITE_OK ? "ok" : "failed"));
  return rc;
}

#ifdef SQLITE_TEST
/*
** Count the number of fullsyncs and normal syncs.  This is used to test
** that syncs and fullsyncs are occuring at the right times.







<
<
<
<
<
<
<











<
<
<
|
<
|




|
<
|


|



|
<
<

|


>
>







|
|
|
|

<
<
|
|
<
<

|
>


>

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


<
<

>








<
<
<
|
<
|

|
>


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

<
|



>
|







695
696
697
698
699
700
701







702
703
704
705
706
707
708
709
710
711
712



713

714
715
716
717
718
719

720
721
722
723
724
725
726
727


728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745


746
747


748
749
750
751
752
753
754







755
756
757
758
759
760
761



762
763

764

765
766
767
768
769
770
771
772
773
774


775
776
777
778
779
780
781
782
783
784



785

786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804

805
806
807
808
809
810
811
812
813
814
815
816
817
  }
#endif
  OSTRACE(("CLOSE %d %s\n", pFile->h, rc ? "ok" : "failed"));
  OpenCounter(-1);
  return rc ? SQLITE_OK : SQLITE_IOERR;
}








/*
** Read data from a file into a buffer.  Return SQLITE_OK if all
** bytes were read successfully and SQLITE_IOERR if anything goes
** wrong.
*/
static int winRead(
  sqlite3_file *id,          /* File to read from */
  void *pBuf,                /* Write content into this buffer */
  int amt,                   /* Number of bytes to read */
  sqlite3_int64 offset       /* Begin reading at this offset */
){



  winFile *pFile = (winFile*)id;  /* file handle */

  DWORD nRead;                    /* Number of bytes actually read from file */

  assert( id!=0 );
  SimulateIOError(return SQLITE_IOERR_READ);
  OSTRACE(("READ %d lock=%d\n", pFile->h, pFile->locktype));


  if( seekWinFile(pFile, offset) ){
    return SQLITE_FULL;
  }
  if( !ReadFile(pFile->h, pBuf, amt, &nRead, 0) ){
    pFile->lastErrno = GetLastError();
    return SQLITE_IOERR_READ;
  }
  if( nRead<(DWORD)amt ){


    /* Unread parts of the buffer must be zero-filled */
    memset(&((char*)pBuf)[nRead], 0, amt-nRead);
    return SQLITE_IOERR_SHORT_READ;
  }

  return SQLITE_OK;
}

/*
** Write data from a buffer into a file.  Return SQLITE_OK on success
** or some other error code on failure.
*/
static int winWrite(
  sqlite3_file *id,               /* File to write into */
  const void *pBuf,               /* The bytes to be written */
  int amt,                        /* Number of bytes to write */
  sqlite3_int64 offset            /* Offset into the file to begin writing at */
){


  int rc;                         /* True if error has occured, else false */
  winFile *pFile = (winFile*)id;  /* File handle */



  assert( amt>0 );
  assert( pFile );
  SimulateIOError(return SQLITE_IOERR_WRITE);
  SimulateDiskfullError(return SQLITE_FULL);

  OSTRACE(("WRITE %d lock=%d\n", pFile->h, pFile->locktype));








  rc = seekWinFile(pFile, offset);
  if( rc==0 ){
    u8 *aRem = (u8 *)pBuf;        /* Data yet to be written */
    int nRem = amt;               /* Number of bytes yet to be written */
    DWORD nWrite;                 /* Bytes written by each WriteFile() call */




    while( nRem>0 && WriteFile(pFile->h, aRem, nRem, &nWrite, 0) && nWrite>0 ){
      aRem += nWrite;

      nRem -= nWrite;

    }
    if( nRem>0 ){
      pFile->lastErrno = GetLastError();
      rc = 1;
    }
  }

  if( rc ){
    if( pFile->lastErrno==ERROR_HANDLE_DISK_FULL ){
      return SQLITE_FULL;


    }
    return SQLITE_IOERR_WRITE;
  }
  return SQLITE_OK;
}

/*
** Truncate an open file to a specified size
*/
static int winTruncate(sqlite3_file *id, sqlite3_int64 nByte){



  winFile *pFile = (winFile*)id;  /* File handle object */

  int rc = SQLITE_OK;             /* Return code for this function */

  assert( pFile );

  OSTRACE(("TRUNCATE %d %lld\n", pFile->h, nByte));
  SimulateIOError(return SQLITE_IOERR_TRUNCATE);

  /* If the user has configured a chunk-size for this file, truncate the
  ** file so that it consists of an integer number of chunks (i.e. the
  ** actual file size after the operation may be larger than the requested
  ** size).
  */
  if( pFile->szChunk ){
    nByte = ((nByte + pFile->szChunk - 1)/pFile->szChunk) * pFile->szChunk;
  }

  /* SetEndOfFile() returns non-zero when successful, or zero when it fails. */
  if( seekWinFile(pFile, nByte) ){
    rc = SQLITE_IOERR_TRUNCATE;

  }else if( 0==SetEndOfFile(pFile->h) ){
    pFile->lastErrno = GetLastError();
    rc = SQLITE_IOERR_TRUNCATE;
  }

  OSTRACE(("TRUNCATE %d %lld %s\n", pFile->h, nByte, rc ? "failed" : "ok"));
  return rc;
}

#ifdef SQLITE_TEST
/*
** Count the number of fullsyncs and normal syncs.  This is used to test
** that syncs and fullsyncs are occuring at the right times.
1141
1142
1143
1144
1145
1146
1147




1148
1149
1150
1151
1152
1153
1154
    case SQLITE_FCNTL_LOCKSTATE: {
      *(int*)pArg = ((winFile*)id)->locktype;
      return SQLITE_OK;
    }
    case SQLITE_LAST_ERRNO: {
      *(int*)pArg = (int)((winFile*)id)->lastErrno;
      return SQLITE_OK;




    }
    case SQLITE_FCNTL_SIZE_HINT: {
      sqlite3_int64 sz = *(sqlite3_int64*)pArg;
      SimulateIOErrorBenign(1);
      winTruncate(id, sz);
      SimulateIOErrorBenign(0);
      return SQLITE_OK;







>
>
>
>







1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
    case SQLITE_FCNTL_LOCKSTATE: {
      *(int*)pArg = ((winFile*)id)->locktype;
      return SQLITE_OK;
    }
    case SQLITE_LAST_ERRNO: {
      *(int*)pArg = (int)((winFile*)id)->lastErrno;
      return SQLITE_OK;
    }
    case SQLITE_FCNTL_CHUNK_SIZE: {
      ((winFile*)id)->szChunk = *(int *)pArg;
      return SQLITE_OK;
    }
    case SQLITE_FCNTL_SIZE_HINT: {
      sqlite3_int64 sz = *(sqlite3_int64*)pArg;
      SimulateIOErrorBenign(1);
      winTruncate(id, sz);
      SimulateIOErrorBenign(0);
      return SQLITE_OK;
Changes to src/pager.c.
3361
3362
3363
3364
3365
3366
3367
3368
3369
3370
3371
3372
3373
3374
3375
    rc = pagerOpentemp(pPager, pPager->fd, pPager->vfsFlags);
  }

  /* Before the first write, give the VFS a hint of what the final
  ** file size will be.
  */
  assert( rc!=SQLITE_OK || isOpen(pPager->fd) );
  if( rc==SQLITE_OK && pPager->dbSize>(pPager->dbOrigSize+1) ){
    sqlite3_int64 szFile = pPager->pageSize * (sqlite3_int64)pPager->dbSize;
    sqlite3OsFileControl(pPager->fd, SQLITE_FCNTL_SIZE_HINT, &szFile);
  }

  while( rc==SQLITE_OK && pList ){
    Pgno pgno = pList->pgno;








|







3361
3362
3363
3364
3365
3366
3367
3368
3369
3370
3371
3372
3373
3374
3375
    rc = pagerOpentemp(pPager, pPager->fd, pPager->vfsFlags);
  }

  /* Before the first write, give the VFS a hint of what the final
  ** file size will be.
  */
  assert( rc!=SQLITE_OK || isOpen(pPager->fd) );
  if( rc==SQLITE_OK && pPager->dbSize>pPager->dbOrigSize ){
    sqlite3_int64 szFile = pPager->pageSize * (sqlite3_int64)pPager->dbSize;
    sqlite3OsFileControl(pPager->fd, SQLITE_FCNTL_SIZE_HINT, &szFile);
  }

  while( rc==SQLITE_OK && pList ){
    Pgno pgno = pList->pgno;

Changes to src/sqlite.h.in.
686
687
688
689
690
691
692








693
694
695
696
697
698
699
**
** The [SQLITE_FCNTL_SIZE_HINT] opcode is used by SQLite to give the VFS
** layer a hint of how large the database file will grow to be during the
** current transaction.  This hint is not guaranteed to be accurate but it
** is often close.  The underlying VFS might choose to preallocate database
** file space based on this hint in order to help writes to the database
** file run faster.








*/
#define SQLITE_FCNTL_LOCKSTATE        1
#define SQLITE_GET_LOCKPROXYFILE      2
#define SQLITE_SET_LOCKPROXYFILE      3
#define SQLITE_LAST_ERRNO             4
#define SQLITE_FCNTL_SIZE_HINT        5
#define SQLITE_FCNTL_CHUNK_SIZE       6







>
>
>
>
>
>
>
>







686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
**
** The [SQLITE_FCNTL_SIZE_HINT] opcode is used by SQLite to give the VFS
** layer a hint of how large the database file will grow to be during the
** current transaction.  This hint is not guaranteed to be accurate but it
** is often close.  The underlying VFS might choose to preallocate database
** file space based on this hint in order to help writes to the database
** file run faster.
**
** The [SQLITE_FCNTL_CHUNK_SIZE] opcode is used to request that the VFS
** extends and truncates the database file in chunks of a size specified
** by the user. The fourth argument to [sqlite3_file_control()] should 
** point to an integer (type int) containing the new chunk-size to use
** for the nominated database. Allocating database file space in large
** chunks (say 1MB at a time), may reduce file-system fragmentation and
** improve performance on some systems.
*/
#define SQLITE_FCNTL_LOCKSTATE        1
#define SQLITE_GET_LOCKPROXYFILE      2
#define SQLITE_SET_LOCKPROXYFILE      3
#define SQLITE_LAST_ERRNO             4
#define SQLITE_FCNTL_SIZE_HINT        5
#define SQLITE_FCNTL_CHUNK_SIZE       6
Changes to src/wal.c.
1517
1518
1519
1520
1521
1522
1523

1524
1525
1526
1527
1528
1529
1530
){
  int rc;                         /* Return code */
  int szPage = pWal->hdr.szPage;  /* Database page-size */
  WalIterator *pIter = 0;         /* Wal iterator context */
  u32 iDbpage = 0;                /* Next database page to write */
  u32 iFrame = 0;                 /* Wal frame containing data for iDbpage */
  u32 mxSafeFrame;                /* Max frame that can be backfilled */

  int i;                          /* Loop counter */
  volatile WalCkptInfo *pInfo;    /* The checkpoint status information */

  if( pWal->hdr.mxFrame==0 ) return SQLITE_OK;

  /* Allocate the iterator */
  rc = walIteratorInit(pWal, &pIter);







>







1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
){
  int rc;                         /* Return code */
  int szPage = pWal->hdr.szPage;  /* Database page-size */
  WalIterator *pIter = 0;         /* Wal iterator context */
  u32 iDbpage = 0;                /* Next database page to write */
  u32 iFrame = 0;                 /* Wal frame containing data for iDbpage */
  u32 mxSafeFrame;                /* Max frame that can be backfilled */
  u32 mxPage;                     /* Max database page to write */
  int i;                          /* Loop counter */
  volatile WalCkptInfo *pInfo;    /* The checkpoint status information */

  if( pWal->hdr.mxFrame==0 ) return SQLITE_OK;

  /* Allocate the iterator */
  rc = walIteratorInit(pWal, &pIter);
1541
1542
1543
1544
1545
1546
1547

1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567


1568
1569
1570
1571
1572
1573









1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586

  /* Compute in mxSafeFrame the index of the last frame of the WAL that is
  ** safe to write into the database.  Frames beyond mxSafeFrame might
  ** overwrite database pages that are in use by active readers and thus
  ** cannot be backfilled from the WAL.
  */
  mxSafeFrame = pWal->hdr.mxFrame;

  pInfo = walCkptInfo(pWal);
  for(i=1; i<WAL_NREADER; i++){
    u32 y = pInfo->aReadMark[i];
    if( mxSafeFrame>=y ){
      assert( y<=pWal->hdr.mxFrame );
      rc = walLockExclusive(pWal, WAL_READ_LOCK(i), 1);
      if( rc==SQLITE_OK ){
        pInfo->aReadMark[i] = READMARK_NOT_USED;
        walUnlockExclusive(pWal, WAL_READ_LOCK(i), 1);
      }else if( rc==SQLITE_BUSY ){
        mxSafeFrame = y;
      }else{
        goto walcheckpoint_out;
      }
    }
  }

  if( pInfo->nBackfill<mxSafeFrame
   && (rc = walLockExclusive(pWal, WAL_READ_LOCK(0), 1))==SQLITE_OK
  ){


    u32 nBackfill = pInfo->nBackfill;

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










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







>




















>
>






>
>
>
>
>
>
>
>
>





|







1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599

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

  if( pInfo->nBackfill<mxSafeFrame
   && (rc = walLockExclusive(pWal, WAL_READ_LOCK(0), 1))==SQLITE_OK
  ){
    i64 nReq;                     /* File size hint passed to VFS */
    i64 nSize;                    /* Current size of database file */
    u32 nBackfill = pInfo->nBackfill;

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

    /* If the database file may grow as a result of this checkpoint, hint
    ** about the eventual size of the db file to the VFS layer. 
    */
    nReq = ((i64)mxPage * szPage);
    rc = sqlite3OsFileSize(pWal->pDbFd, &nSize);
    if( rc==SQLITE_OK && nSize<nReq ){
      rc = sqlite3OsFileControl(pWal->pDbFd, SQLITE_FCNTL_SIZE_HINT, &nReq);
    }

    /* Iterate through the contents of the WAL, copying data to the db file. */
    while( rc==SQLITE_OK && 0==walIteratorNext(pIter, &iDbpage, &iFrame) ){
      i64 iOffset;
      assert( walFramePgno(pWal, iFrame)==iDbpage );
      if( iFrame<=nBackfill || iFrame>mxSafeFrame || iDbpage>mxPage ) continue;
      iOffset = walFrameOffset(iFrame, szPage) + WAL_FRAME_HDRSIZE;
      /* testcase( IS_BIG_INT(iOffset) ); // requires a 4GiB WAL file */
      rc = sqlite3OsRead(pWal->pWalFd, zBuf, szPage, iOffset);
      if( rc!=SQLITE_OK ) break;
      iOffset = (iDbpage-1)*(i64)szPage;
      testcase( IS_BIG_INT(iOffset) );
      rc = sqlite3OsWrite(pWal->pDbFd, zBuf, szPage, iOffset);
Changes to test/fallocate.test.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
#
#***********************************************************************
#

set testdir [file dirname $argv0]
source $testdir/tester.tcl

if {$::tcl_platform(platform)!="unix"} {
  finish_test
  return
}

file_control_chunksize_test db main [expr 1024*1024]

do_test fallocate-1.1 {
  execsql {
    PRAGMA page_size = 1024;
    PRAGMA auto_vacuum = 1;
    CREATE TABLE t1(a, b);







<
<
<
<
<







9
10
11
12
13
14
15





16
17
18
19
20
21
22
#
#***********************************************************************
#

set testdir [file dirname $argv0]
source $testdir/tester.tcl






file_control_chunksize_test db main [expr 1024*1024]

do_test fallocate-1.1 {
  execsql {
    PRAGMA page_size = 1024;
    PRAGMA auto_vacuum = 1;
    CREATE TABLE t1(a, b);
48
49
50
51
52
53
54
55



















































































56

  execsql { DELETE FROM t1 WHERE a = 2 }
  file size test.db
} [expr 1*1024*1024]

do_test fallocate-1.6 {
  execsql { PRAGMA freelist_count }
} {0}




















































































finish_test









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

>
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
  execsql { DELETE FROM t1 WHERE a = 2 }
  file size test.db
} [expr 1*1024*1024]

do_test fallocate-1.6 {
  execsql { PRAGMA freelist_count }
} {0}

# Start a write-transaction and read the "database file size" field from
# the journal file. This field should be set to the number of pages in
# the database file based on the size of the file on disk, not the actual
# logical size of the database within the file.
#
# We need to check this to verify that if in the unlikely event a rollback
# causes a database file to grow, the database grows to its previous size
# on disk, not to the minimum size required to hold the database image.
#
do_test fallocate-1.7 {
  execsql { BEGIN; INSERT INTO t1 VALUES(1, 2); }
  hexio_get_int [hexio_read test.db-journal 16 4]
} {1024}
do_test fallocate-1.8 { execsql { COMMIT } } {}


#-------------------------------------------------------------------------
# The following tests - fallocate-2.* - test that things work in WAL
# mode as well.
#
db close
file delete -force test.db
sqlite3 db test.db
file_control_chunksize_test db main [expr 32*1024]

do_test fallocate-2.1 {
  execsql {
    PRAGMA page_size = 1024;
    PRAGMA journal_mode = WAL;
    CREATE TABLE t1(a, b);
  }
  file size test.db
} [expr 32*1024]

do_test fallocate-2.2 {
  execsql { INSERT INTO t1 VALUES(1, zeroblob(35*1024)) }
  execsql { PRAGMA wal_checkpoint }
  file size test.db
} [expr 64*1024]

do_test fallocate-2.3 {
  execsql { DELETE FROM t1 }
  execsql { VACUUM }
  file size test.db
} [expr 64*1024]

do_test fallocate-2.4 {
  execsql { PRAGMA wal_checkpoint }
  file size test.db
} [expr 32*1024]

do_test fallocate-2.5 {
  execsql { 
    INSERT INTO t1 VALUES(2, randomblob(35*1024));
    PRAGMA wal_checkpoint;
    INSERT INTO t1 VALUES(3, randomblob(128));
    DELETE FROM t1 WHERE a = 2;
    VACUUM;
  }
  file size test.db
} [expr 64*1024]

do_test fallocate-2.6 {
  sqlite3 db2 test.db
  execsql { BEGIN ; SELECT count(a) FROM t1 } db2
  execsql {  
    INSERT INTO t1 VALUES(4, randomblob(128));
    PRAGMA wal_checkpoint;
  }
  file size test.db
} [expr 64*1024]

do_test fallocate-2.7 {
  execsql { SELECT count(b) FROM t1 } db2
} {1}

do_test fallocate-2.8 {
  execsql { COMMIT } db2
  execsql { PRAGMA wal_checkpoint }
  file size test.db
} [expr 32*1024]


finish_test