/ Check-in [581fadfe]
Login

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

Overview
Comment:Add code for the atomic-write optimisation. Disabled by default. (CVS 4268)
Downloads: Tarball | ZIP archive | SQL archive
Timelines: family | ancestors | descendants | both | trunk
Files: files | file ages | folders
SHA1: 581fadfe31757e3fb97b12f93c1e3c085e4b6009
User & Date: danielk1977 2007-08-22 11:22:04
Context
2007-08-22
11:41
Get the quick.test script running with SQLITE_THREADSAFE enabled. (CVS 4269) check-in: 1f28b7e4 user: drh tags: trunk
11:22
Add code for the atomic-write optimisation. Disabled by default. (CVS 4268) check-in: 581fadfe user: danielk1977 tags: trunk
02:57
Remove a debugging btree_breakpoint from vtab1.test. (CVS 4267) check-in: 6cc8641e user: drh tags: trunk
Changes
Hide Diffs Unified Diffs Ignore Whitespace Patch

Changes to main.mk.

53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
..
99
100
101
102
103
104
105

106
107
108
109
110
111
112
...
361
362
363
364
365
366
367



368
369
370
371
372
373
374
#
TCCX = $(TCC) $(OPTS) $(THREADSAFE) $(USLEEP) -I. -I$(TOP)/src

# Object files for the SQLite library.
#
LIBOBJ+= alter.o analyze.o attach.o auth.o btree.o build.o \
         callback.o complete.o date.o delete.o \
         expr.o func.o hash.o insert.o loadext.o \
         main.o malloc.o mem1.o mem2.o mutex.o \
         opcodes.o os.o os_os2.o os_unix.o os_win.o \
         pager.o parse.o pragma.o prepare.o printf.o random.o \
         select.o table.o tclsqlite.o tokenize.o trigger.o \
         update.o util.o vacuum.o \
         vdbe.o vdbeapi.o vdbeaux.o vdbeblob.o vdbefifo.o vdbemem.o \
         where.o utf.o legacy.o vtab.o
................................................................................
  $(TOP)/src/date.c \
  $(TOP)/src/delete.c \
  $(TOP)/src/expr.c \
  $(TOP)/src/func.c \
  $(TOP)/src/hash.c \
  $(TOP)/src/hash.h \
  $(TOP)/src/insert.c \

  $(TOP)/src/legacy.c \
  $(TOP)/src/loadext.c \
  $(TOP)/src/main.c \
  $(TOP)/src/malloc.c \
  $(TOP)/src/mem1.c \
  $(TOP)/src/mem2.c \
  $(TOP)/src/mutex.c \
................................................................................
	$(TCCX) -c $(TOP)/src/func.c

hash.o:	$(TOP)/src/hash.c $(HDR)
	$(TCCX) -c $(TOP)/src/hash.c

insert.o:	$(TOP)/src/insert.c $(HDR)
	$(TCCX) -c $(TOP)/src/insert.c




legacy.o:	$(TOP)/src/legacy.c $(HDR)
	$(TCCX) -c $(TOP)/src/legacy.c

loadext.o:	$(TOP)/src/loadext.c $(HDR)
	$(TCCX) -c $(TOP)/src/loadext.c








|







 







>







 







>
>
>







53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
..
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
...
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
#
TCCX = $(TCC) $(OPTS) $(THREADSAFE) $(USLEEP) -I. -I$(TOP)/src

# Object files for the SQLite library.
#
LIBOBJ+= alter.o analyze.o attach.o auth.o btree.o build.o \
         callback.o complete.o date.o delete.o \
         expr.o func.o hash.o insert.o journal.o loadext.o \
         main.o malloc.o mem1.o mem2.o mutex.o \
         opcodes.o os.o os_os2.o os_unix.o os_win.o \
         pager.o parse.o pragma.o prepare.o printf.o random.o \
         select.o table.o tclsqlite.o tokenize.o trigger.o \
         update.o util.o vacuum.o \
         vdbe.o vdbeapi.o vdbeaux.o vdbeblob.o vdbefifo.o vdbemem.o \
         where.o utf.o legacy.o vtab.o
................................................................................
  $(TOP)/src/date.c \
  $(TOP)/src/delete.c \
  $(TOP)/src/expr.c \
  $(TOP)/src/func.c \
  $(TOP)/src/hash.c \
  $(TOP)/src/hash.h \
  $(TOP)/src/insert.c \
  $(TOP)/src/journal.c \
  $(TOP)/src/legacy.c \
  $(TOP)/src/loadext.c \
  $(TOP)/src/main.c \
  $(TOP)/src/malloc.c \
  $(TOP)/src/mem1.c \
  $(TOP)/src/mem2.c \
  $(TOP)/src/mutex.c \
................................................................................
	$(TCCX) -c $(TOP)/src/func.c

hash.o:	$(TOP)/src/hash.c $(HDR)
	$(TCCX) -c $(TOP)/src/hash.c

insert.o:	$(TOP)/src/insert.c $(HDR)
	$(TCCX) -c $(TOP)/src/insert.c

journal.o:	$(TOP)/src/journal.c $(HDR)
	$(TCCX) -c $(TOP)/src/journal.c

legacy.o:	$(TOP)/src/legacy.c $(HDR)
	$(TCCX) -c $(TOP)/src/legacy.c

loadext.o:	$(TOP)/src/loadext.c $(HDR)
	$(TCCX) -c $(TOP)/src/loadext.c

Added src/journal.c.









































































































































































































































































































































































































































































>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
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
/*
** 2007 August 22
**
** 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.
**
*************************************************************************
**
** @(#) $Id: journal.c,v 1.1 2007/08/22 11:22:04 danielk1977 Exp $
*/

#ifdef SQLITE_ENABLE_ATOMIC_WRITE

/*
** This file implements a special kind of sqlite3_file object used
** by SQLite to create journal files if the atomic-write optimization
** is enabled.
**
** The distinctive characteristic of this sqlite3_file is that the
** actual on disk file is created lazily. When the file is created,
** the caller specifies a buffer size for an in-memory buffer to
** be used to service read() and write() requests. The actual file
** on disk is not created or populated until either:
**
**   1) The in-memory representation grows too large for the allocated 
**      buffer, or
**   2) The xSync() method is called.
*/

#include "sqliteInt.h"

struct JournalFile {
  sqlite3_io_methods *pMethod;

  int nBuf;
  char *zBuf;
  int iSize;

  int flags;
  sqlite3_vfs *pVfs;
  sqlite3_file *pReal;
  const char *zJournal;
};
typedef struct JournalFile JournalFile;

/*
** If it does not already exists, create and populate the on-disk file 
** for JournalFile p.
*/
static int createFile(JournalFile *p){
  int rc = SQLITE_OK;
  if( !p->pReal ){
    p->pReal = (sqlite3_file *)&p[1];
    rc = sqlite3OsOpen(p->pVfs, p->zJournal, p->pReal, p->flags, 0);
    if( rc==SQLITE_OK && p->iSize>0 ){
      assert(p->iSize<=p->nBuf);
      rc = sqlite3OsWrite(p->pReal, p->zBuf, p->iSize, 0);
    }
  }
  return rc;
}

/*
** Close the file.
*/
static int jrnlClose(sqlite3_file *pJfd){
  JournalFile *p = (JournalFile *)pJfd;
  if( p->pReal ){
    sqlite3OsClose(p->pReal);
  }
  sqlite3_free(p->zBuf);
  return SQLITE_OK;
}

/*
** Read data from the file.
*/
static int jrnlRead(
  sqlite3_file *pJfd, 
  void *zBuf, 
  int iAmt, 
  sqlite_int64 iOfst
){
  int rc = SQLITE_OK;
  JournalFile *p = (JournalFile *)pJfd;
  if( p->pReal ){
    rc = sqlite3OsRead(p->pReal, zBuf, iAmt, iOfst);
  }else{
    int n = iAmt;
    memset(zBuf, 0, n);
    if( n+iOfst>p->iSize ){
      rc = SQLITE_IOERR_SHORT_READ;
    }else{
      memcpy(zBuf, &p->zBuf[iOfst], n);
    }
  }
  return rc;
}

/*
** Write data to the file.
*/
static int jrnlWrite(
  sqlite3_file *pJfd, 
  const void *zBuf, 
  int iAmt, 
  sqlite_int64 iOfst
){
  int rc = SQLITE_OK;
  JournalFile *p = (JournalFile *)pJfd;
  if( !p->pReal && (iOfst+iAmt)>p->nBuf ){
    rc = createFile(p);
  }
  if( rc==SQLITE_OK ){
    if( p->pReal ){
      rc = sqlite3OsWrite(p->pReal, zBuf, iAmt, iOfst);
    }else{
      memcpy(&p->zBuf[iOfst], zBuf, iAmt);
      if( p->iSize<(iOfst+iAmt) ){
        p->iSize = (iOfst+iAmt);
      }
    }
  }
  return rc;
}

/*
** Truncate the file.
*/
static int jrnlTruncate(sqlite3_file *pJfd, sqlite_int64 size){
  int rc = SQLITE_OK;
  JournalFile *p = (JournalFile *)pJfd;
  if( p->pReal ){
    rc = sqlite3OsTruncate(p->pReal, size);
  }else if( size>p->iSize ){
    p->iSize = size;
  }
  return rc;
}

/*
** Sync the file.
*/
static int jrnlSync(sqlite3_file *pJfd, int flags){
  int rc;
  JournalFile *p = (JournalFile *)pJfd;
  rc = createFile(p);
  if( rc==SQLITE_OK ){
    rc = sqlite3OsSync(p->pReal, flags);
  }
  return rc;
}

/*
** Query the size of the file in bytes.
*/
static int jrnlFileSize(sqlite3_file *pJfd, sqlite_int64 *pSize){
  int rc = SQLITE_OK;
  JournalFile *p = (JournalFile *)pJfd;
  if( p->pReal ){
    rc = sqlite3OsFileSize(p->pReal, pSize);
  }else{
    *pSize = (sqlite_int64) p->iSize;
  }
  return rc;
}

/*
** Table of methods for JournalFile sqlite3_file object.
*/
static struct sqlite3_io_methods JournalFileMethods = {
  1,             /* iVersion */
  jrnlClose,     /* xClose */
  jrnlRead,      /* xRead */
  jrnlWrite,     /* xWrite */
  jrnlTruncate,  /* xTruncate */
  jrnlSync,      /* xSync */
  jrnlFileSize,  /* xFileSize */
  0,             /* xLock */
  0,             /* xUnlock */
  0,             /* xCheckReservedLock */
  0,             /* xBreakLock */
  0,             /* xLockState */
  0,             /* xSectorSize */
  0              /* xDeviceCharacteristics */
};

/* 
** Open a journal file.
*/
int sqlite3JournalOpen(
  sqlite3_vfs *pVfs, 
  const char *zName, 
  sqlite3_file *pJfd,
  int flags,
  int nBuf
){
  JournalFile *p = (JournalFile *)pJfd;
  memset(p, 0, sqlite3JournalSize(pVfs));
  if( nBuf>0 ){
    p->zBuf = sqlite3MallocZero(nBuf);
    if( !p->zBuf ){
      return SQLITE_NOMEM;
    }
  }else{
    return sqlite3OsOpen(pVfs, zName, pJfd, flags, 0);
  }
  p->pMethod = &JournalFileMethods;
  p->nBuf = nBuf;
  p->flags = flags;
  p->zJournal = zName;
  p->pVfs = pVfs;
  return SQLITE_OK;
}

/* 
** Return the number of bytes required to store a JournalFile that uses vfs
** pVfs to create the underlying on-disk files.
*/
int sqlite3JournalSize(sqlite3_vfs *pVfs){
  return (pVfs->szOsFile+sizeof(JournalFile));
}
#endif

Changes to src/os.c.

20
21
22
23
24
25
26

27
28



29
30
31
32
33
34
35
/*
** The following routines are convenience wrappers around methods
** of the sqlite3_file object.  This is mostly just syntactic sugar. All
** of this would be completely automatic if SQLite were coded using
** C++ instead of plain old C.
*/
int sqlite3OsClose(sqlite3_file *pId){

  if( !pId->pMethods ) return SQLITE_OK;
  return pId->pMethods->xClose(pId);



}
int sqlite3OsRead(sqlite3_file *id, void *pBuf, int amt, i64 offset){
  return id->pMethods->xRead(id, pBuf, amt, offset);
}
int sqlite3OsWrite(sqlite3_file *id, const void *pBuf, int amt, i64 offset){
  return id->pMethods->xWrite(id, pBuf, amt, offset);
}







>
|
|
>
>
>







20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
/*
** The following routines are convenience wrappers around methods
** of the sqlite3_file object.  This is mostly just syntactic sugar. All
** of this would be completely automatic if SQLite were coded using
** C++ instead of plain old C.
*/
int sqlite3OsClose(sqlite3_file *pId){
  int rc = SQLITE_OK;
  if( pId->pMethods ){
    rc = pId->pMethods->xClose(pId);
    pId->pMethods = 0;
  }
  return rc;
}
int sqlite3OsRead(sqlite3_file *id, void *pBuf, int amt, i64 offset){
  return id->pMethods->xRead(id, pBuf, amt, offset);
}
int sqlite3OsWrite(sqlite3_file *id, const void *pBuf, int amt, i64 offset){
  return id->pMethods->xWrite(id, pBuf, amt, offset);
}

Changes to src/pager.c.

14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
...
599
600
601
602
603
604
605




































606
607
608
609
610
611
612
....
1816
1817
1818
1819
1820
1821
1822

1823
1824
1825
1826
1827
1828
1829

1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
1846
1847
....
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
....
2994
2995
2996
2997
2998
2999
3000
3001
3002
3003
3004
3005
3006
3007
3008
....
3456
3457
3458
3459
3460
3461
3462





3463

3464
3465
3466
3467
3468
3469
3470
....
4006
4007
4008
4009
4010
4011
4012
4013
4014
4015
4016
4017
4018
4019
4020
4021


4022
4023
4024

4025
4026
4027
4028






4029
4030
4031
4032
4033
4034
4035
4036
4037
4038
4039
4040
....
4059
4060
4061
4062
4063
4064
4065






























4066
4067
4068
4069
4070
4071
4072
4073
4074
4075
4076
4077
4078
4079
4080
4081
....
4091
4092
4093
4094
4095
4096
4097
4098
4099

4100
4101
4102
4103
4104
4105
4106
** The pager is used to access a database disk file.  It implements
** atomic commit and rollback through the use of a journal file that
** is separate from the database file.  The pager also implements file
** locking to prevent two processes from writing the same database
** file simultaneously, or one process from reading the database while
** another is writing.
**
** @(#) $Id: pager.c,v 1.365 2007/08/22 00:39:20 drh Exp $
*/
#ifndef SQLITE_OMIT_DISKIO
#include "sqliteInt.h"
#include <assert.h>
#include <string.h>

/*
................................................................................
static int osUnlock(sqlite3_file *pFd, int eLock){
  if( !pFd->pMethods ){
    return SQLITE_OK;
  }
  return sqlite3OsUnlock(pFd, eLock);
}





































/*
** This function should be called when an error occurs within the pager
** code. The first argument is a pointer to the pager structure, the
** second the error-code about to be returned by a pager API function. 
** The value returned is a copy of the second argument to this function. 
**
** If the second argument is SQLITE_IOERR, SQLITE_CORRUPT, or SQLITE_FULL
................................................................................
  int rc = SQLITE_OK;
  int i;
  int tempFile = 0;
  int memDb = 0;
  int readOnly = 0;
  int useJournal = (flags & PAGER_OMIT_JOURNAL)==0;
  int noReadlock = (flags & PAGER_NO_READLOCK)!=0;


  /* The default return is a NULL pointer */
  *ppPager = 0;

  /* Allocate memory for the pager structure */
  pPager = sqlite3MallocZero(
    sizeof(*pPager) +           /* Pager structure */

    pVfs->szOsFile * 3 +        /* The db, journal and stmt journal files */ 
    pVfs->mxPathname * 3 + 30   /* zFilename, zDirectory, zJournal */
  );
  if( !pPager ){
    return SQLITE_NOMEM;
  }
  pPtr = (u8 *)&pPager[1];
  pPager->fd = (sqlite3_file*)&pPtr[pVfs->szOsFile*0];
  pPager->jfd = (sqlite3_file*)&pPtr[pVfs->szOsFile*1];
  pPager->stfd = (sqlite3_file*)&pPtr[pVfs->szOsFile*2];
  pPager->zFilename = (char*)&pPtr[pVfs->szOsFile*3];
  pPager->zDirectory = &pPager->zFilename[pVfs->mxPathname];
  pPager->zJournal = &pPager->zDirectory[pVfs->mxPathname];
  pPager->pVfs = pVfs;

  /* Open the pager file and set zFullPathname to point at malloc()ed 
  ** memory containing the complete filename (i.e. including the directory).
  */
................................................................................
      }
    }
  }else{
    /* If a temporary file is requested, it is not opened immediately.
    ** In this case we accept the default page size and delay actually
    ** opening the file until the first call to OsWrite().
    */ 
    /* rc = sqlite3PagerOpentemp(pVfs, pPager->fd, pPager->zFilename); */
    tempFile = 1;
    pPager->state = PAGER_EXCLUSIVE;
  }

  if( pPager && rc==SQLITE_OK ){
    pPager->pTmpSpace = (char *)sqlite3_malloc(SQLITE_DEFAULT_PAGE_SIZE);
  }
................................................................................
        ** a read/write file handle.
        */
        rc = SQLITE_BUSY;
        if( sqlite3OsAccess(pVfs, pPager->zJournal, SQLITE_ACCESS_EXISTS) ){
          int fout = 0;
          int flags = SQLITE_OPEN_READWRITE|SQLITE_OPEN_MAIN_JOURNAL;
          assert( !pPager->tempFile );
          rc = sqlite3OsOpen(pVfs, pPager->zJournal, pPager->jfd, flags, &fout);
          assert( rc!=SQLITE_OK || pPager->jfd->pMethods );
          if( fout&SQLITE_OPEN_READONLY ){
            rc = SQLITE_BUSY;
            sqlite3OsClose(pPager->jfd);
          }
        }
        if( rc!=SQLITE_OK ){
................................................................................
  }

  if( pPager->tempFile ){
    flags |= (SQLITE_OPEN_DELETEONCLOSE|SQLITE_OPEN_TEMP_JOURNAL);
  }else{
    flags |= (SQLITE_OPEN_MAIN_JOURNAL);
  }





  rc = sqlite3OsOpen(pVfs, pPager->zJournal, pPager->jfd, flags, 0);

  assert( rc!=SQLITE_OK || pPager->jfd->pMethods );
  pPager->journalOff = 0;
  pPager->setMaster = 0;
  pPager->journalHdr = 0;
  if( rc!=SQLITE_OK ){
    if( rc==SQLITE_NOMEM ){
      sqlite3OsDelete(pVfs, pPager->zJournal, 0);
................................................................................
}


/*
** This routine is called to increment the database file change-counter,
** stored at byte 24 of the pager file.
*/
static int pager_incr_changecounter(Pager *pPager){
  PgHdr *pPgHdr;
  u32 change_counter;
  int rc;

  if( !pPager->changeCountDone ){
    /* Open page 1 of the file for writing. */
    rc = sqlite3PagerGet(pPager, 1, &pPgHdr);
    if( rc!=SQLITE_OK ) return rc;


    rc = sqlite3PagerWrite(pPgHdr);
    if( rc!=SQLITE_OK ) return rc;
  

    /* Increment the value just read and write it back to byte 24. */
    change_counter = sqlite3Get4byte((u8*)pPager->dbFileVers);
    change_counter++;
    put32bits(((char*)PGHDR_TO_DATA(pPgHdr))+24, change_counter);






    /* Release the page reference. */
    sqlite3PagerUnref(pPgHdr);
    pPager->changeCountDone = 1;
  }
  return SQLITE_OK;
}

/*
** Sync the database file for the pager pPager. zMaster points to the name
** of a master journal file that should be written into the individual
** journal file. zMaster may be NULL, which is interpreted as no master
** journal (a single database transaction).
................................................................................

  /* If this is an in-memory db, or no pages have been written to, or this
  ** function has already been called, it is a no-op.
  */
  if( pPager->state!=PAGER_SYNCED && !MEMDB && pPager->dirtyCache ){
    PgHdr *pPg;
    assert( pPager->journalOpen );































    /* If a master journal file name has already been written to the
    ** journal file, then no sync is required. This happens when it is
    ** written, then the process fails to upgrade from a RESERVED to an
    ** EXCLUSIVE lock. The next time the process tries to commit the
    ** transaction the m-j name will have already been written.
    */
    if( !pPager->setMaster ){
      rc = pager_incr_changecounter(pPager);
      if( rc!=SQLITE_OK ) goto sync_exit;
#ifndef SQLITE_OMIT_AUTOVACUUM
      if( nTrunc!=0 ){
        /* If this transaction has made the database smaller, then all pages
        ** being discarded by the truncation must be written to the journal
        ** file.
        */
................................................................................
          }
        } 
      }
#endif
      rc = writeMasterJournal(pPager, zMaster);
      if( rc!=SQLITE_OK ) goto sync_exit;
      rc = syncJournal(pPager);
      if( rc!=SQLITE_OK ) goto sync_exit;
    }


#ifndef SQLITE_OMIT_AUTOVACUUM
    if( nTrunc!=0 ){
      rc = sqlite3PagerTruncate(pPager, nTrunc);
      if( rc!=SQLITE_OK ) goto sync_exit;
    }
#endif







|







 







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







 







>







>
|







|
|
|







 







<







 







|







 







>
>
>
>
>

>







 







|


|





>
>
|
|
|
>




>
>
>
>
>
>




|







 







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








|







 







<

>







14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
...
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
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
....
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
1882
1883
1884
1885
....
1905
1906
1907
1908
1909
1910
1911

1912
1913
1914
1915
1916
1917
1918
....
3031
3032
3033
3034
3035
3036
3037
3038
3039
3040
3041
3042
3043
3044
3045
....
3493
3494
3495
3496
3497
3498
3499
3500
3501
3502
3503
3504
3505
3506
3507
3508
3509
3510
3511
3512
3513
....
4049
4050
4051
4052
4053
4054
4055
4056
4057
4058
4059
4060
4061
4062
4063
4064
4065
4066
4067
4068
4069
4070
4071
4072
4073
4074
4075
4076
4077
4078
4079
4080
4081
4082
4083
4084
4085
4086
4087
4088
4089
4090
4091
4092
....
4111
4112
4113
4114
4115
4116
4117
4118
4119
4120
4121
4122
4123
4124
4125
4126
4127
4128
4129
4130
4131
4132
4133
4134
4135
4136
4137
4138
4139
4140
4141
4142
4143
4144
4145
4146
4147
4148
4149
4150
4151
4152
4153
4154
4155
4156
4157
4158
4159
4160
4161
4162
4163
....
4173
4174
4175
4176
4177
4178
4179

4180
4181
4182
4183
4184
4185
4186
4187
4188
** The pager is used to access a database disk file.  It implements
** atomic commit and rollback through the use of a journal file that
** is separate from the database file.  The pager also implements file
** locking to prevent two processes from writing the same database
** file simultaneously, or one process from reading the database while
** another is writing.
**
** @(#) $Id: pager.c,v 1.366 2007/08/22 11:22:04 danielk1977 Exp $
*/
#ifndef SQLITE_OMIT_DISKIO
#include "sqliteInt.h"
#include <assert.h>
#include <string.h>

/*
................................................................................
static int osUnlock(sqlite3_file *pFd, int eLock){
  if( !pFd->pMethods ){
    return SQLITE_OK;
  }
  return sqlite3OsUnlock(pFd, eLock);
}

/*
** This function determines whether or not the atomic-write optimization
** can be used with this pager. The optimization can be used if:
**
**  (a) the value returned by OsDeviceCharacteristics() indicates that
**      a database page may be written atomically, and
**  (b) the value returned by OsSectorSize() is less than or equal
**      to the page size.
**
** If the optimization cannot be used, 0 is returned. If it can be used,
** then the value returned is the size of the journal file when it
** contains rollback data for exactly one page.
*/
#ifdef SQLITE_ENABLE_ATOMIC_WRITE
static int jrnlBufferSize(Pager *pPager){
  int dc;           /* Device characteristics */
  int nSector;      /* Sector size */
  int nPage;        /* Page size */
  sqlite3_file *fd = pPager->fd;

  if( fd->pMethods ){
    dc = sqlite3OsDeviceCharacteristics(fd);
    nSector = sqlite3OsSectorSize(fd);
    nPage = pPager->pageSize;
  }

  assert(SQLITE_IOCAP_ATOMIC512==(512>>8));
  assert(SQLITE_IOCAP_ATOMIC64K==(65536>>8));

  if( !fd->pMethods || (dc&(SQLITE_IOCAP_ATOMIC|(nPage<<8))&&nSector<=nPage) ){
    return JOURNAL_HDR_SZ(pPager) + JOURNAL_PG_SZ(pPager);
  }
  return 0;
}
#endif

/*
** This function should be called when an error occurs within the pager
** code. The first argument is a pointer to the pager structure, the
** second the error-code about to be returned by a pager API function. 
** The value returned is a copy of the second argument to this function. 
**
** If the second argument is SQLITE_IOERR, SQLITE_CORRUPT, or SQLITE_FULL
................................................................................
  int rc = SQLITE_OK;
  int i;
  int tempFile = 0;
  int memDb = 0;
  int readOnly = 0;
  int useJournal = (flags & PAGER_OMIT_JOURNAL)==0;
  int noReadlock = (flags & PAGER_NO_READLOCK)!=0;
  int journalFileSize = sqlite3JournalSize(pVfs);

  /* The default return is a NULL pointer */
  *ppPager = 0;

  /* Allocate memory for the pager structure */
  pPager = sqlite3MallocZero(
    sizeof(*pPager) +           /* Pager structure */
    journalFileSize +           /* The journal file structure */ 
    pVfs->szOsFile * 2 +        /* The db and stmt journal files */ 
    pVfs->mxPathname * 3 + 30   /* zFilename, zDirectory, zJournal */
  );
  if( !pPager ){
    return SQLITE_NOMEM;
  }
  pPtr = (u8 *)&pPager[1];
  pPager->fd = (sqlite3_file*)&pPtr[pVfs->szOsFile*0];
  pPager->stfd = (sqlite3_file*)&pPtr[pVfs->szOsFile*1];
  pPager->jfd = (sqlite3_file*)&pPtr[pVfs->szOsFile*2];
  pPager->zFilename = (char*)&pPtr[pVfs->szOsFile*2+journalFileSize];
  pPager->zDirectory = &pPager->zFilename[pVfs->mxPathname];
  pPager->zJournal = &pPager->zDirectory[pVfs->mxPathname];
  pPager->pVfs = pVfs;

  /* Open the pager file and set zFullPathname to point at malloc()ed 
  ** memory containing the complete filename (i.e. including the directory).
  */
................................................................................
      }
    }
  }else{
    /* If a temporary file is requested, it is not opened immediately.
    ** In this case we accept the default page size and delay actually
    ** opening the file until the first call to OsWrite().
    */ 

    tempFile = 1;
    pPager->state = PAGER_EXCLUSIVE;
  }

  if( pPager && rc==SQLITE_OK ){
    pPager->pTmpSpace = (char *)sqlite3_malloc(SQLITE_DEFAULT_PAGE_SIZE);
  }
................................................................................
        ** a read/write file handle.
        */
        rc = SQLITE_BUSY;
        if( sqlite3OsAccess(pVfs, pPager->zJournal, SQLITE_ACCESS_EXISTS) ){
          int fout = 0;
          int flags = SQLITE_OPEN_READWRITE|SQLITE_OPEN_MAIN_JOURNAL;
          assert( !pPager->tempFile );
          rc = sqlite3OsOpen(pVfs, pPager->zJournal, pPager->jfd, flags,&fout);
          assert( rc!=SQLITE_OK || pPager->jfd->pMethods );
          if( fout&SQLITE_OPEN_READONLY ){
            rc = SQLITE_BUSY;
            sqlite3OsClose(pPager->jfd);
          }
        }
        if( rc!=SQLITE_OK ){
................................................................................
  }

  if( pPager->tempFile ){
    flags |= (SQLITE_OPEN_DELETEONCLOSE|SQLITE_OPEN_TEMP_JOURNAL);
  }else{
    flags |= (SQLITE_OPEN_MAIN_JOURNAL);
  }
#ifdef SQLITE_ENABLE_ATOMIC_WRITE
  rc = sqlite3JournalOpen(
      pVfs, pPager->zJournal, pPager->jfd, flags, jrnlBufferSize(pPager)
  );
#else
  rc = sqlite3OsOpen(pVfs, pPager->zJournal, pPager->jfd, flags, 0);
#endif
  assert( rc!=SQLITE_OK || pPager->jfd->pMethods );
  pPager->journalOff = 0;
  pPager->setMaster = 0;
  pPager->journalHdr = 0;
  if( rc!=SQLITE_OK ){
    if( rc==SQLITE_NOMEM ){
      sqlite3OsDelete(pVfs, pPager->zJournal, 0);
................................................................................
}


/*
** This routine is called to increment the database file change-counter,
** stored at byte 24 of the pager file.
*/
static int pager_incr_changecounter(Pager *pPager, int isDirect){
  PgHdr *pPgHdr;
  u32 change_counter;
  int rc = SQLITE_OK;

  if( !pPager->changeCountDone ){
    /* Open page 1 of the file for writing. */
    rc = sqlite3PagerGet(pPager, 1, &pPgHdr);
    if( rc!=SQLITE_OK ) return rc;

    if( !isDirect ){
      rc = sqlite3PagerWrite(pPgHdr);
      if( rc!=SQLITE_OK ) return rc;
    }

    /* Increment the value just read and write it back to byte 24. */
    change_counter = sqlite3Get4byte((u8*)pPager->dbFileVers);
    change_counter++;
    put32bits(((char*)PGHDR_TO_DATA(pPgHdr))+24, change_counter);

    if( isDirect && pPager->fd->pMethods ){
      const void *zBuf = PGHDR_TO_DATA(pPgHdr);
      rc = sqlite3OsWrite(pPager->fd, zBuf, pPager->pageSize, 0);
    }

    /* Release the page reference. */
    sqlite3PagerUnref(pPgHdr);
    pPager->changeCountDone = 1;
  }
  return rc;
}

/*
** Sync the database file for the pager pPager. zMaster points to the name
** of a master journal file that should be written into the individual
** journal file. zMaster may be NULL, which is interpreted as no master
** journal (a single database transaction).
................................................................................

  /* If this is an in-memory db, or no pages have been written to, or this
  ** function has already been called, it is a no-op.
  */
  if( pPager->state!=PAGER_SYNCED && !MEMDB && pPager->dirtyCache ){
    PgHdr *pPg;
    assert( pPager->journalOpen );

#ifdef SQLITE_ENABLE_ATOMIC_WRITE
    /* The atomic-write optimization can be used if all of the
    ** following are true:
    **
    **    + The file-system supports the atomic-write property for
    **      blocks of size page-size, and
    **    + This commit is not part of a multi-file transaction, and
    **    + Exactly one page has been modified and store in the journal file.
    **
    ** If the optimization can be used, then the journal file will never
    ** be created for this transaction.
    */
    if( !zMaster && pPager->journalOff==jrnlBufferSize(pPager) && nTrunc==0
      && (0==pPager->pDirty || 0==pPager->pDirty->pDirty)
    ){
      /* Update the nRec field in the journal file. */
      int offset = pPager->journalHdr + sizeof(aJournalMagic);
      assert(pPager->nRec==1);
      rc = write32bits(pPager->jfd, offset, pPager->nRec);

      /* Update the db file change counter. The following call will modify
      ** the in-memory representation of page 1 to include the updated
      ** change counter and then write page 1 directly to the database
      ** file. Because of the atomic-write property of the host file-system, 
      ** this is safe.
      */
      rc = pager_incr_changecounter(pPager, 1);
    }else 
#endif

    /* If a master journal file name has already been written to the
    ** journal file, then no sync is required. This happens when it is
    ** written, then the process fails to upgrade from a RESERVED to an
    ** EXCLUSIVE lock. The next time the process tries to commit the
    ** transaction the m-j name will have already been written.
    */
    if( !pPager->setMaster ){
      rc = pager_incr_changecounter(pPager, 0);
      if( rc!=SQLITE_OK ) goto sync_exit;
#ifndef SQLITE_OMIT_AUTOVACUUM
      if( nTrunc!=0 ){
        /* If this transaction has made the database smaller, then all pages
        ** being discarded by the truncation must be written to the journal
        ** file.
        */
................................................................................
          }
        } 
      }
#endif
      rc = writeMasterJournal(pPager, zMaster);
      if( rc!=SQLITE_OK ) goto sync_exit;
      rc = syncJournal(pPager);

    }
    if( rc!=SQLITE_OK ) goto sync_exit;

#ifndef SQLITE_OMIT_AUTOVACUUM
    if( nTrunc!=0 ){
      rc = sqlite3PagerTruncate(pPager, nTrunc);
      if( rc!=SQLITE_OK ) goto sync_exit;
    }
#endif

Changes to src/sqliteInt.h.

7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
....
1857
1858
1859
1860
1861
1862
1863







1864
1865
1866
1867
1868
1869
1870
**    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.
**
*************************************************************************
** Internal interface definitions for SQLite.
**
** @(#) $Id: sqliteInt.h,v 1.595 2007/08/21 19:33:56 drh Exp $
*/
#ifndef _SQLITEINT_H_
#define _SQLITEINT_H_
#include "sqliteLimit.h"


#if defined(SQLITE_TCL) || defined(TCLSH)
................................................................................
int sqlite3VtabCallDestroy(sqlite3*, int, const char *);
int sqlite3VtabBegin(sqlite3 *, sqlite3_vtab *);
FuncDef *sqlite3VtabOverloadFunction(sqlite3 *,FuncDef*, int nArg, Expr*);
void sqlite3InvalidFunction(sqlite3_context*,int,sqlite3_value**);
int sqlite3Reprepare(Vdbe*);
void sqlite3ExprListCheckLength(Parse*, ExprList*, int, const char*);
CollSeq *sqlite3BinaryCompareCollSeq(Parse *, Expr *, Expr *);








#if SQLITE_MAX_EXPR_DEPTH>0
  void sqlite3ExprSetHeight(Expr *);
  int sqlite3SelectExprHeight(Select *);
#else
  #define sqlite3ExprSetHeight(x)
#endif







|







 







>
>
>
>
>
>
>







7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
....
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
**    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.
**
*************************************************************************
** Internal interface definitions for SQLite.
**
** @(#) $Id: sqliteInt.h,v 1.596 2007/08/22 11:22:04 danielk1977 Exp $
*/
#ifndef _SQLITEINT_H_
#define _SQLITEINT_H_
#include "sqliteLimit.h"


#if defined(SQLITE_TCL) || defined(TCLSH)
................................................................................
int sqlite3VtabCallDestroy(sqlite3*, int, const char *);
int sqlite3VtabBegin(sqlite3 *, sqlite3_vtab *);
FuncDef *sqlite3VtabOverloadFunction(sqlite3 *,FuncDef*, int nArg, Expr*);
void sqlite3InvalidFunction(sqlite3_context*,int,sqlite3_value**);
int sqlite3Reprepare(Vdbe*);
void sqlite3ExprListCheckLength(Parse*, ExprList*, int, const char*);
CollSeq *sqlite3BinaryCompareCollSeq(Parse *, Expr *, Expr *);

#ifdef SQLITE_ENABLE_ATOMIC_WRITE
  int sqlite3JournalOpen(sqlite3_vfs *, const char *, sqlite3_file *, int, int);
  int sqlite3JournalSize(sqlite3_vfs *);
#else
  #define sqlite3JournalSize(pVfs) ((pVfs)->szOsFile)
#endif

#if SQLITE_MAX_EXPR_DEPTH>0
  void sqlite3ExprSetHeight(Expr *);
  int sqlite3SelectExprHeight(Select *);
#else
  #define sqlite3ExprSetHeight(x)
#endif