/ Check-in [41f2175b]
Login

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

Overview
Comment:Update the locking-style code in os_unix.c. The updates are as yet untested. (CVS 4244)
Downloads: Tarball | ZIP archive | SQL archive
Timelines: family | ancestors | descendants | both | trunk
Files: files | file ages | folders
SHA1: 41f2175b1ed7eccf271b687ee5c3ea262a0cd096
User & Date: danielk1977 2007-08-20 06:44:22
Context
2007-08-20
11:12
Get main.c to compile when HAVE_USLEEP is 0. (CVS 4245) check-in: b27f022f user: drh tags: trunk
06:44
Update the locking-style code in os_unix.c. The updates are as yet untested. (CVS 4244) check-in: 41f2175b user: danielk1977 tags: trunk
05:36
Fix full_fsync() related functionality broken as part of the migration to sqlite3_vfs. (CVS 4243) check-in: cb24cda1 user: danielk1977 tags: trunk
Changes
Hide Diffs Unified Diffs Ignore Whitespace Patch

Changes to src/os_unix.c.

77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
..
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
...
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
...
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
...
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
...
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
...
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
...
385
386
387
388
389
390
391



392
393
394
395
396
397
398
...
563
564
565
566
567
568
569
570

571

572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
...
590
591
592
593
594
595
596
597

598

599
600
601
602
603
604
605
...
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
....
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
....
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
....
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
....
1685
1686
1687
1688
1689
1690
1691

1692
1693
1694
1695




1696
1697
1698


1699
1700
1701
1702
1703
1704
1705
....
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
....
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
....
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
....
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
....
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
....
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
....
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
2062
2063
2064
2065
2066
2067
2068
2069
2070
2071
2072
2073
2074
2075
2076
2077
2078
2079
2080
2081
2082
....
2084
2085
2086
2087
2088
2089
2090
2091
2092
2093
2094
2095
2096
2097
2098
....
2102
2103
2104
2105
2106
2107
2108
2109
2110
2111
2112
2113
2114
2115
2116
....
2136
2137
2138
2139
2140
2141
2142
2143
2144
2145
2146
2147
2148
2149
2150
....
2163
2164
2165
2166
2167
2168
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
....
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
....
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
2370
2371
2372
2373
2374
2375
2376
2377


2378

2379
2380
2381
2382
2383
2384
2385

2386
2387
2388
2389
2390
2391
2392
2393
2394
2395
2396
2397
2398
2399


2400

2401
2402
2403
2404
2405
2406
2407
2408
2409
2410
2411
2412
2413
2414
2415
2416
2417
2418

2419
2420
2421
2422
2423
2424
2425
2426
2427
2428
2429
2430
2431
2432
2433
2434
2435
2436
2437
2438
2439
2440
2441
2442
2443
2444
2445
2446
2447
2448
2449
2450
2451
2452
2453
2454
2455
2456
2457
2458
2459
2460
2461
2462
2463
2464
2465
2466
2467
....
2498
2499
2500
2501
2502
2503
2504
2505
2506
2507
2508
2509
2510
2511
2512
....
2543
2544
2545
2546
2547
2548
2549










2550
2551
2552
2553
2554
2555
2556
....
2571
2572
2573
2574
2575
2576
2577


2578
2579
2580
2581
2582
2583
2584
/*
** Maximum supported path-length.
*/
#define MAX_PATHNAME 512


/*
** The unixFile structure is subclass of OsFile specific for the unix
** protability layer.
*/
typedef struct unixFile unixFile;
struct unixFile {
  sqlite3_io_methods const *pMethod;  /* Always the first entry */
  struct openCnt *pOpen;    /* Info about all open fd's on this inode */
  struct lockInfo *pLock;   /* Info about locks on this inode */
................................................................................
#endif /* SQLITE_ENABLE_LOCKING_STYLE */
  int h;                    /* The file descriptor */
  unsigned char locktype;   /* The type of lock held on this fd */
  unsigned char isOpen;     /* True if needs to be closed */
  int dirfd;                /* File descriptor for the directory */
  i64 offset;               /* Seek offset */
#ifdef SQLITE_UNIX_THREADS
  pthread_t tid;            /* The thread that "owns" this OsFile */
#endif
};


/*
** Provide the ability to override some OS-layer functions during
** testing.  This is used to simulate OS crashes to verify that 
................................................................................

/*
** Do not include any of the File I/O interface procedures if the
** SQLITE_OMIT_DISKIO macro is defined (indicating that the database
** will be in-memory only)
*/
#ifndef SQLITE_OMIT_DISKIO


/*
** Define various macros that are missing from some systems.
*/
#ifndef O_LARGEFILE
# define O_LARGEFILE 0
#endif
................................................................................
#ifdef SQLITE_UNIX_THREADS
#define threadid pthread_self()
#else
#define threadid 0
#endif

/*
** Set or check the OsFile.tid field.  This field is set when an OsFile
** is first opened.  All subsequent uses of the OsFile verify that the
** same thread is operating on the OsFile.  Some operating systems do
** not allow locks to be overridden by other threads and that restriction
** means that sqlite3* database handles cannot be moved from one thread
** to another.  This logic makes sure a user does not try to do that
** by mistake.
**
** Version 3.3.1 (2006-01-15):  OsFiles can be moved from one thread to
** another as long as we are running on a system that supports threads
** overriding each others locks (which now the most common behavior)
** or if no locks are held.  But the OsFile.pLock field needs to be
** recomputed because its key includes the thread-id.  See the 
** transferOwnership() function below for additional information
*/
#if defined(SQLITE_UNIX_THREADS)
# define SET_THREADID(X)   (X)->tid = pthread_self()
# define CHECK_THREADID(X) (threadsOverrideEachOthersLocks==0 && \
                            !pthread_equal((X)->tid, pthread_self()))
................................................................................
** specific inode of the database file (the inode is determined by the
** st_dev and st_ino fields of the stat structure that fstat() fills in)
** and check for locks already existing on that inode.  When locks are
** created or removed, we have to look at our own internal record of the
** locks to see if another thread has previously set a lock on that same
** inode.
**
** The OsFile structure for POSIX is no longer just an integer file
** descriptor.  It is now a structure that holds the integer file
** descriptor and a pointer to a structure that describes the internal
** locks on the corresponding inode.  There is one locking structure
** per inode, so if the same inode is opened twice, both OsFile structures
** point to the same locking structure.  The locking structure keeps
** a reference count (so we will know when to delete it) and a "cnt"
** field that tells us its internal lock status.  cnt==0 means the
** file is unlocked.  cnt==-1 means the file has an exclusive lock.
** cnt>0 means there are cnt shared locks on the file.
**
** Any attempt to lock or unlock a file first checks the locking
................................................................................
** 2004-Jan-11:
** More recent discoveries about POSIX advisory locks.  (The more
** I discover, the more I realize the a POSIX advisory locks are
** an abomination.)
**
** If you close a file descriptor that points to a file that has locks,
** all locks on that file that are owned by the current process are
** released.  To work around this problem, each OsFile structure contains
** a pointer to an openCnt structure.  There is one openCnt structure
** per open inode, which means that multiple OsFiles can point to a single
** openCnt.  When an attempt is made to close an OsFile, if there are
** other OsFiles open on the same inode that are holding locks, the call
** to close() the file descriptor is deferred until all of the locks clear.
** The openCnt structure keeps a list of file descriptors that need to
** be closed and that list is walked (and cleared) when the last lock
** clears.
**
** First, under Linux threads, because each thread has a separate
** process ID, lock operations in one thread do not override locks
................................................................................
};

/*
** An instance of the following structure is allocated for each open
** inode on each thread with a different process ID.  (Threads have
** different process IDs on linux, but not on most other unixes.)
**
** A single inode can have multiple file descriptors, so each OsFile
** structure contains a pointer to an instance of this object and this
** object keeps a count of the number of OsFiles pointing to it.
*/
struct lockInfo {
  struct lockKey key;  /* The lookup key */
  int cnt;             /* Number of SHARED locks held */
  int locktype;        /* One of SHARED_LOCK, RESERVED_LOCK etc. */
  int nRef;            /* Number of pointers to this structure */
};
................................................................................
        flockLockingStyle,           /* use flock() */
        dotlockLockingStyle,         /* use <file>.lock files */
        noLockingStyle,              /* useful for read-only file system */
        unsupportedLockingStyle      /* indicates unsupported file system */
} sqlite3LockingStyle;
#endif /* SQLITE_ENABLE_LOCKING_STYLE */




static void enterMutex(){
  sqlite3_mutex_enter(sqlite3_mutex_alloc(SQLITE_MUTEX_GLOBAL));
}
static void leaveMutex(){
  sqlite3_mutex_leave(sqlite3_mutex_alloc(SQLITE_MUTEX_GLOBAL));
}

................................................................................
}

#ifdef SQLITE_ENABLE_LOCKING_STYLE
/*
** Tests a byte-range locking query to see if byte range locks are 
** supported, if not we fall back to dotlockLockingStyle.
*/
static sqlite3LockingStyle sqlite3TestLockingStyle(const char *filePath, 

  int fd) {

  /* test byte-range lock using fcntl */
  struct flock lockInfo;
  
  lockInfo.l_len = 1;
  lockInfo.l_start = 0;
  lockInfo.l_whence = SEEK_SET;
  lockInfo.l_type = F_RDLCK;
  
  if (fcntl(fd, F_GETLK, &lockInfo) != -1) {
    return posixLockingStyle;
  } 
  
  /* testing for flock can give false positives.  So if if the above test
  ** fails, then we fall back to using dot-lock style locking.
  */  
  return dotlockLockingStyle;
................................................................................
/* 
** Examines the f_fstypename entry in the statfs structure as returned by 
** stat() for the file system hosting the database file, assigns the 
** appropriate locking style based on it's value.  These values and 
** assignments are based on Darwin/OSX behavior and have not been tested on 
** other systems.
*/
static sqlite3LockingStyle sqlite3DetectLockingStyle(const char *filePath, 

  int fd) {


#ifdef SQLITE_FIXED_LOCKING_STYLE
  return (sqlite3LockingStyle)SQLITE_FIXED_LOCKING_STYLE;
#else
  struct statfs fsInfo;

  if (statfs(filePath, &fsInfo) == -1)
................................................................................
  }
}
#else
  /* On single-threaded builds, ownership transfer is a no-op */
# define transferOwnership(X) SQLITE_OK
#endif

/*
** Delete the named file
*/
int sqlite3UnixDelete(const char *zFilename){
  SimulateIOError(return SQLITE_IOERR_DELETE);
  unlink(zFilename);
  return SQLITE_OK;
}

/*
** Return TRUE if the named file exists.
*/
int sqlite3UnixFileExists(const char *zFilename){
  return access(zFilename, 0)==0;
}

/*
** Attempt to open a file for both reading and writing.  If that
** fails, try opening it read-only.  If the file does not exist,
** try to create it.
**
** On success, a handle for the open file is written to *id
** and *pReadonly is set to 0 if the file was opened for reading and
** writing or 1 if the file was opened read-only.  The function returns
** SQLITE_OK.
**
** On failure, the function returns SQLITE_CANTOPEN and leaves
** *id and *pReadonly unchanged.
*/
#if 0
int sqlite3UnixOpenReadWrite(
  const char *zFilename,
  sqlite3_file **pId,
  int *pReadonly
){
  int h;
  
  assert( 0==*pId );
  h = open(zFilename, O_RDWR|O_CREAT|O_LARGEFILE|O_BINARY,
                        SQLITE_DEFAULT_FILE_PERMISSIONS);
  if( h<0 ){
#ifdef EISDIR
    if( errno==EISDIR ){
      return SQLITE_CANTOPEN;
    }
#endif
    h = open(zFilename, O_RDONLY|O_LARGEFILE|O_BINARY);
    if( h<0 ){
      return SQLITE_CANTOPEN; 
    }
    *pReadonly = 1;
  }else{
    *pReadonly = 0;
  }

  return CRASH_TEST_OVERRIDE(
    zFilename, pId, allocateUnixFile(h, pId, zFilename, 0)
  );
}


/*
** Attempt to open a new file for exclusive access by this process.
** The file will be opened for both reading and writing.  To avoid
** a potential security problem, we do not allow the file to have
** previously existed.  Nor do we allow the file to be a symbolic
** link.
**
** If delFlag is true, then make arrangements to automatically delete
** the file when it is closed.
**
** On success, write the file handle into *id and return SQLITE_OK.
**
** On failure, return SQLITE_CANTOPEN.
*/
int sqlite3UnixOpenExclusive(
  const char *zFilename, 
  sqlite3_file **pId, 
  int delFlag
){
  int h;

  assert( 0==*pId );
  h = open(zFilename,
                O_RDWR|O_CREAT|O_EXCL|O_NOFOLLOW|O_LARGEFILE|O_BINARY,
                delFlag ? 0600 : SQLITE_DEFAULT_FILE_PERMISSIONS);
  if( h<0 ){
    return SQLITE_CANTOPEN;
  }
  return CRASH_TEST_OVERRIDE(
    zFilename, pId, allocateUnixFile(h, pId, zFilename, delFlag)
  );
}

/*
** Attempt to open a new file for read-only access.
**
** On success, write the file handle into *id and return SQLITE_OK.
**
** On failure, return SQLITE_CANTOPEN.
*/
int sqlite3UnixOpenReadOnly(const char *zFilename, sqlite3_file **pId){
  int h;
  
  assert( 0==*pId );
  h = open(zFilename, O_RDONLY|O_LARGEFILE|O_BINARY);
  if( h<0 ){
    return SQLITE_CANTOPEN;
  }
  return CRASH_TEST_OVERRIDE(
    zFilename, pId, allocateUnixFile(h, pId, zFilename, 0)
  );
}
#endif

/*
** Attempt to open a file descriptor for the directory that contains a
** file.  This file descriptor can be used to fsync() the directory
** in order to make sure the creation of a new file is actually written
** to disk.
**
** This routine is only meaningful for Unix.  It is a no-op under
** windows since windows does not support hard links.
**
** If FULL_FSYNC is enabled, this function is not longer useful, 
** a FULL_FSYNC sync applies to all pending disk operations.
**
** On success, a handle for a previously open file at *id is
** updated with the new directory file descriptor and SQLITE_OK is
** returned.
**
** On failure, the function returns SQLITE_CANTOPEN and leaves
** *id unchanged.
*/
#if 0
static int unixOpenDirectory(
  OsFile *id,
  const char *zDirname
){
  int h;
  unixFile *pFile = (unixFile*)id;
  assert( pFile!=0 );
  SET_THREADID(pFile);
  assert( pFile->dirfd<0 );
  pFile->dirfd = h = open(zDirname, O_RDONLY|O_BINARY, 0);
  if( h<0 ){
    return SQLITE_CANTOPEN; 
  }
#ifdef FD_CLOEXEC
  fcntl(h, F_SETFD, fcntl(h, F_GETFD, 0) | FD_CLOEXEC);
#endif
  OSTRACE3("OPENDIR %-3d %s\n", h, zDirname);
  return SQLITE_OK;
}
#endif

/*
** Check that a given pathname is a directory and is writable 
**
*/
int sqlite3UnixIsDirWritable(char *zBuf){
#ifndef SQLITE_OMIT_PAGER_PRAGMAS
  struct stat buf;
  if( zBuf==0 ) return 0;
  if( zBuf[0]==0 ) return 0;
  if( stat(zBuf, &buf) ) return 0;
  if( !S_ISDIR(buf.st_mode) ) return 0;
  if( access(zBuf, 07) ) return 0;
#endif /* SQLITE_OMIT_PAGER_PRAGMAS */
  return 1;
}

/*
** Seek to the offset in id->offset then read cnt bytes into pBuf.
** Return the number of bytes actually read.  Update the offset.
*/
static int seekAndRead(unixFile *id, sqlite3_int64 offset, void *pBuf, int cnt){
  int got;
  i64 newOffset;
................................................................................
#endif
    close(pFile->dirfd);  /* Only need to sync once, so close the directory */
    pFile->dirfd = -1;    /* when we are done. */
  }
  return SQLITE_OK;
}

/*
** Sync the directory zDirname. This is a no-op on operating systems other
** than UNIX.
**
** This is used to make sure the master journal file has truely been deleted
** before making changes to individual journals on a multi-database commit.
** The F_FULLFSYNC option is not needed here.
*/
int sqlite3UnixSyncDirectory(const char *zDirname){
#ifdef SQLITE_DISABLE_DIRSYNC
  return SQLITE_OK;
#else
  int fd;
  int r;
  fd = open(zDirname, O_RDONLY|O_BINARY, 0);
  OSTRACE3("DIRSYNC %-3d (%s)\n", fd, zDirname);
  if( fd<0 ){
    return SQLITE_CANTOPEN; 
  }
  r = fsync(fd);
  close(fd);
  SimulateIOError( r=1 );
  if( r ){
    return SQLITE_IOERR_DIR_FSYNC;
  }else{
    return SQLITE_OK;
  }
#endif
}

/*
** Truncate an open file to a specified size
*/
static int unixTruncate(sqlite3_file *id, i64 nByte){
  int rc;
  assert( id );
  rc = ftruncate(((unixFile*)id)->h, (off_t)nByte);
................................................................................

  assert( pFile );
  OSTRACE7("LOCK    %d %s was %s(%s,%d) pid=%d\n", pFile->h,
      locktypeName(locktype), locktypeName(pFile->locktype),
      locktypeName(pLock->locktype), pLock->cnt , getpid());

  /* If there is already a lock of this type or more restrictive on the
  ** OsFile, do nothing. Don't use the end_lock: exit path, as
  ** enterMutex() hasn't been called yet.
  */
  if( pFile->locktype>=locktype ){
    OSTRACE3("LOCK    %d %s ok (already held)\n", pFile->h,
            locktypeName(locktype));
    return SQLITE_OK;
  }
................................................................................
  rc = transferOwnership(pFile);
  if( rc!=SQLITE_OK ){
    leaveMutex();
    return rc;
  }
  pLock = pFile->pLock;

  /* If some thread using this PID has a lock via a different OsFile*
  ** handle that precludes the requested lock, return BUSY.
  */
  if( (pFile->locktype!=pLock->locktype && 
          (pLock->locktype>=PENDING_LOCK || locktype>SHARED_LOCK))
  ){
    rc = SQLITE_BUSY;
    goto end_lock;
................................................................................
  unsigned char unLockFlag;         /* 1 = unlock, 0 = lock */
  unsigned char startEndFlag;       /* 1=rel to end of fork, 0=rel to start */
  int fd;                           /* file desc to assoc this lock with */
};

#define afpfsByteRangeLock2FSCTL        _IOWR('z', 23, struct ByteRangeLockPB2)


/* return 0 on success, 1 on failure.  To match the behavior of the 
  normal posix file locking (used in unixLock for example), we should 
  provide 'richer' return codes - specifically to differentiate between
  'file busy' and 'file system error' results */




static int _AFPFSSetLock(const char *path, int fd, unsigned long long offset, 
                         unsigned long long length, int setLockFlag)
{


  struct ByteRangeLockPB2       pb;
  int                     err;
  
  pb.unLockFlag = setLockFlag ? 0 : 1;
  pb.startEndFlag = 0;
  pb.offset = offset;
  pb.length = length; 
................................................................................

/*
 ** This routine checks if there is a RESERVED lock held on the specified
 ** file by this or any other process. If such a lock is held, return
 ** non-zero.  If the file is unlocked or holds only SHARED locks, then
 ** return zero.
 */
static int afpUnixCheckReservedLock(OsFile *id){
  int r = 0;
  unixFile *pFile = (unixFile*)id;
  
  assert( pFile ); 
  afpLockingContext *context = (afpLockingContext *) pFile->lockingContext;
  
  /* Check if a thread in this process holds such a lock */
................................................................................
  OSTRACE3("TEST WR-LOCK %d %d\n", pFile->h, r);
  
  return r;
}

/* AFP-style locking following the behavior of unixLock, see the unixLock 
** function comments for details of lock management. */
static int afpUnixLock(OsFile *id, int locktype)
{
  int rc = SQLITE_OK;
  unixFile *pFile = (unixFile*)id;
  afpLockingContext *context = (afpLockingContext *) pFile->lockingContext;
  int gotPendingLock = 0;
  
  assert( pFile );
  OSTRACE5("LOCK    %d %s was %s pid=%d\n", pFile->h,
         locktypeName(locktype), locktypeName(pFile->locktype), getpid());  
  /* If there is already a lock of this type or more restrictive on the
    ** OsFile, do nothing. Don't use the afp_end_lock: exit path, as
    ** enterMutex() hasn't been called yet.
    */
  if( pFile->locktype>=locktype ){
    OSTRACE3("LOCK    %d %s ok (already held)\n", pFile->h,
           locktypeName(locktype));
    return SQLITE_OK;
  }
................................................................................
/*
 ** Lower the locking level on file descriptor pFile to locktype.  locktype
 ** must be either NO_LOCK or SHARED_LOCK.
 **
 ** If the locking level of the file descriptor is already at or below
 ** the requested locking level, this routine is a no-op.
 */
static int afpUnixUnlock(OsFile *id, int locktype) {
  struct flock lock;
  int rc = SQLITE_OK;
  unixFile *pFile = (unixFile*)id;
  afpLockingContext *context = (afpLockingContext *) pFile->lockingContext;

  assert( pFile );
  OSTRACE5("UNLOCK  %d %d was %d pid=%d\n", pFile->h, locktype,
................................................................................
  leaveMutex();
  return rc;
}

/*
 ** Close a file & cleanup AFP specific locking context 
 */
static int afpUnixClose(OsFile **pId) {
  unixFile *id = (unixFile*)*pId;
  
  if( !id ) return SQLITE_OK;
  afpUnixUnlock(*pId, NO_LOCK);
  /* free the AFP locking structure */
  if (id->lockingContext != NULL) {
    if (((afpLockingContext *)id->lockingContext)->filePath != NULL)
      sqlite3_free(((afpLockingContext*)id->lockingContext)->filePath);
    sqlite3_free(id->lockingContext);
  }
  
  if( id->dirfd>=0 ) close(id->dirfd);
  id->dirfd = -1;
  close(id->h);
  id->isOpen = 0;
  OSTRACE2("CLOSE   %-3d\n", id->h);
  OpenCounter(-1);
  sqlite3_free(id);
  *pId = 0;
  return SQLITE_OK;
}


#pragma mark flock() style locking

/*
 ** The flockLockingContext is not used
 */
typedef void flockLockingContext;

static int flockUnixCheckReservedLock(OsFile *id) {
  unixFile *pFile = (unixFile*)id;
  
  if (pFile->locktype == RESERVED_LOCK) {
    return 1; /* already have a reserved lock */
  } else {
    /* attempt to get the lock */
    int rc = flock(pFile->h, LOCK_EX | LOCK_NB);
................................................................................
      flock(pFile->h, LOCK_UN);
      return 0;  /* no one has it reserved */
    }
    return 1; /* someone else might have it reserved */
  }
}

static int flockUnixLock(OsFile *id, int locktype) {
  unixFile *pFile = (unixFile*)id;
  
  /* if we already have a lock, it is exclusive.  
  ** Just adjust level and punt on outta here. */
  if (pFile->locktype > NO_LOCK) {
    pFile->locktype = locktype;
    return SQLITE_OK;
................................................................................
  } else {
    /* got it, set the type and return ok */
    pFile->locktype = locktype;
    return SQLITE_OK;
  }
}

static int flockUnixUnlock(OsFile *id, int locktype) {
  unixFile *pFile = (unixFile*)id;
  
  assert( locktype<=SHARED_LOCK );
  
  /* no-op if possible */
  if( pFile->locktype==locktype ){
    return SQLITE_OK;
................................................................................
    return SQLITE_OK;
  }
}

/*
 ** Close a file.
 */
static int flockUnixClose(OsFile **pId) {
  unixFile *id = (unixFile*)*pId;
  
  if( !id ) return SQLITE_OK;
  flockUnixUnlock(*pId, NO_LOCK);
  
  if( id->dirfd>=0 ) close(id->dirfd);
  id->dirfd = -1;
  enterMutex();
  
  close(id->h);  
  leaveMutex();
  id->isOpen = 0;
  OSTRACE2("CLOSE   %-3d\n", id->h);
  OpenCounter(-1);
  sqlite3_free(id);
  *pId = 0;
  return SQLITE_OK;
}

#pragma mark Old-School .lock file based locking

/*
 ** The dotlockLockingContext structure contains all dotlock (.lock) lock
................................................................................
 */
typedef struct dotlockLockingContext dotlockLockingContext;
struct dotlockLockingContext {
  char *lockPath;
};


static int dotlockUnixCheckReservedLock(OsFile *id) {
  unixFile *pFile = (unixFile*)id;
  dotlockLockingContext *context = 
    (dotlockLockingContext *) pFile->lockingContext;
  
  if (pFile->locktype == RESERVED_LOCK) {
    return 1; /* already have a reserved lock */
  } else {
................................................................................
      return 1;
    else
      /* file does not exist, we could have it if we want it */
      return 0;
  }
}

static int dotlockUnixLock(OsFile *id, int locktype) {
  unixFile *pFile = (unixFile*)id;
  dotlockLockingContext *context = 
    (dotlockLockingContext *) pFile->lockingContext;
  
  /* if we already have a lock, it is exclusive.  
  ** Just adjust level and punt on outta here. */
  if (pFile->locktype > NO_LOCK) {
................................................................................
  close(fd);
  
  /* got it, set the type and return ok */
  pFile->locktype = locktype;
  return SQLITE_OK;
}

static int dotlockUnixUnlock(OsFile *id, int locktype) {
  unixFile *pFile = (unixFile*)id;
  dotlockLockingContext *context = 
    (dotlockLockingContext *) pFile->lockingContext;
  
  assert( locktype<=SHARED_LOCK );
  
  /* no-op if possible */
................................................................................
  pFile->locktype = NO_LOCK;
  return SQLITE_OK;
}

/*
 ** Close a file.
 */
static int dotlockUnixClose(OsFile **pId) {
  unixFile *id = (unixFile*)*pId;
  
  if( !id ) return SQLITE_OK;
  dotlockUnixUnlock(*pId, NO_LOCK);
  /* free the dotlock locking structure */
  if (id->lockingContext != NULL) {
    if (((dotlockLockingContext *)id->lockingContext)->lockPath != NULL)
      sqlite3_free( ( (dotlockLockingContext *)
        id->lockingContext)->lockPath);
    sqlite3_free(id->lockingContext);
  }
  
  if( id->dirfd>=0 ) close(id->dirfd);
  id->dirfd = -1;
  enterMutex();
  
  close(id->h);
  
  leaveMutex();
  id->isOpen = 0;
  OSTRACE2("CLOSE   %-3d\n", id->h);
  OpenCounter(-1);
  sqlite3_free(id);
  *pId = 0;
  return SQLITE_OK;
}


#pragma mark No locking

/*
 ** The nolockLockingContext is void
 */
typedef void nolockLockingContext;

static int nolockUnixCheckReservedLock(OsFile *id) {
  return 0;
}

static int nolockUnixLock(OsFile *id, int locktype) {
  return SQLITE_OK;
}

static int nolockUnixUnlock(OsFile *id, int locktype) {
  return SQLITE_OK;
}

/*
 ** Close a file.
 */
static int nolockUnixClose(OsFile **pId) {
  unixFile *id = (unixFile*)*pId;
  
  if( !id ) return SQLITE_OK;
  if( id->dirfd>=0 ) close(id->dirfd);
  id->dirfd = -1;
  enterMutex();
  
  close(id->h);
  
  leaveMutex();
  id->isOpen = 0;
  OSTRACE2("CLOSE   %-3d\n", id->h);
  OpenCounter(-1);
  sqlite3_free(id);
  *pId = 0;
  return SQLITE_OK;
}

#endif /* SQLITE_ENABLE_LOCKING_STYLE */

#if 0

/*
** Change the value of the fullsync flag in the given file descriptor.
*/
static void unixSetFullSync(OsFile *id, int v){
  ((unixFile*)id)->fullSync = v;

}

/*
** Return the underlying file handle for an OsFile

*/
static int unixFileHandle(OsFile *id){
  return ((unixFile*)id)->h;
}

#endif

/*
** Return the sector size in bytes of the underlying block device for
** the specified file. This is almost always 512 bytes, but may be
** larger for some devices.
**
** SQLite code assumes this function cannot fail. It also assumes that
** if two files are created in the same file-system directory (i.e.
................................................................................
/*
** Return the device characteristics for the file. This is always 0.
*/
static int unixDeviceCharacteristics(sqlite3_file *id){
  return 0;
}

static int unixBreakLock(sqlite3_file *id){
  assert(!"TODO: unixBreakLock()");
  return 0;
}

/*
** Return an integer that indices the type of lock currently held
** by this handle.  (Used for testing and analysis only.)
*/
static int unixLockState(sqlite3_file *id){
  return ((unixFile*)id)->locktype;
}

/*
** This vector defines all the methods that can operate on an OsFile
** for unix.
*/
static const sqlite3_io_methods sqlite3UnixIoMethod = {
  1,                        /* iVersion */
  unixClose,
  unixRead,
  unixWrite,
................................................................................
  unixLockState,
  unixSectorSize,
  unixDeviceCharacteristics
};

#ifdef SQLITE_ENABLE_LOCKING_STYLE
/*
 ** This vector defines all the methods that can operate on an OsFile
 ** for unix with AFP style file locking.
 */
static const IoMethod sqlite3AFPLockingUnixIoMethod = {

    afpUnixClose,
    unixOpenDirectory,
    unixRead,
    unixWrite,
    unixSeek,
    unixTruncate,
    unixSync,
    unixSetFullSync,
    unixFileHandle,
    unixFileSize,
    afpUnixLock,
    afpUnixUnlock,
    unixLockState,
    afpUnixCheckReservedLock,


    unixSectorSize,

};

/*
 ** This vector defines all the methods that can operate on an OsFile
 ** for unix with flock() style file locking.
 */
static const IoMethod sqlite3FlockLockingUnixIoMethod = {

    flockUnixClose,
    unixOpenDirectory,
    unixRead,
    unixWrite,
    unixSeek,
    unixTruncate,
    unixSync,
    unixSetFullSync,
    unixFileHandle,
    unixFileSize,
    flockUnixLock,
    flockUnixUnlock,
    unixLockState,
    flockUnixCheckReservedLock,


    unixSectorSize,

};

/*
 ** This vector defines all the methods that can operate on an OsFile
 ** for unix with dotlock style file locking.
 */
static const IoMethod sqlite3DotlockLockingUnixIoMethod = {

    dotlockUnixClose,
    unixOpenDirectory,
    unixRead,
    unixWrite,
    unixSeek,
    unixTruncate,
    unixSync,
    unixSetFullSync,
    unixFileHandle,
    unixFileSize,
    dotlockUnixLock,
    dotlockUnixUnlock,
    unixLockState,
    dotlockUnixCheckReservedLock,


    unixSectorSize,

};

/*
 ** This vector defines all the methods that can operate on an OsFile
 ** for unix with dotlock style file locking.
 */
static const IoMethod sqlite3NolockLockingUnixIoMethod = {

  nolockUnixClose,
  unixOpenDirectory,
  unixRead,
  unixWrite,
  unixSeek,
  unixTruncate,
  unixSync,
  unixSetFullSync,
  unixFileHandle,
  unixFileSize,
  nolockUnixLock,
  nolockUnixUnlock,
  unixLockState,
  nolockUnixCheckReservedLock,


  unixSectorSize,

};

#endif /* SQLITE_ENABLE_LOCKING_STYLE */

/*
** Allocate memory for a new unixFile and initialize that unixFile.
** Write a pointer to the new unixFile into *pId.
** If we run out of memory, close the file and return an error.
*/
#ifdef SQLITE_ENABLE_LOCKING_STYLE
/* 
 ** When locking extensions are enabled, the filepath and locking style 
 ** are needed to determine the unixFile pMethod to use for locking operations.
 ** The locking-style specific lockingContext data structure is created 
 ** and assigned here also.
 */
static int allocateUnixFile(
  int h,                  /* Open file descriptor of file being opened */

  OsFile **pId,           /* Write completed initialization here */
  const char *zFilename,  /* Name of the file being opened */
  int delFlag             /* Delete-on-or-before-close flag */
){
  sqlite3LockingStyle lockingStyle;
  unixFile *pNew;
  unixFile f;
  int rc;

  memset(&f, 0, sizeof(f));
  lockingStyle = sqlite3DetectLockingStyle(zFilename, h);
  if ( lockingStyle == posixLockingStyle ) {
    enterMutex();
    rc = findLockInfo(h, &f.pLock, &f.pOpen);
    leaveMutex();
    if( rc ){
      close(h);
      unlink(zFilename);
      return SQLITE_NOMEM;
    }
  } else {
    /*  pLock and pOpen are only used for posix advisory locking */
    f.pLock = NULL;
    f.pOpen = NULL;
  }
  if( delFlag ){
    unlink(zFilename);
  }
  f.dirfd = -1;
  f.h = h;
  SET_THREADID(&f);
  pNew = sqlite3_malloc( sizeof(unixFile) );
  if( pNew==0 ){
    close(h);
    enterMutex();
    releaseLockInfo(f.pLock);
    releaseOpenCnt(f.pOpen);
    leaveMutex();
    *pId = 0;
    return SQLITE_NOMEM;
  }else{
    *pNew = f;
    switch(lockingStyle) {
      case afpLockingStyle: {
        /* afp locking uses the file path so it needs to be included in
        ** the afpLockingContext */
        int nFilename;
        pNew->pMethod = &sqlite3AFPLockingUnixIoMethod;
        pNew->lockingContext = 
................................................................................
        pNew->pMethod = &sqlite3UnixIoMethod;
        break;
      case noLockingStyle:
      case unsupportedLockingStyle:
      default: 
        pNew->pMethod = &sqlite3NolockLockingUnixIoMethod;
    }
    *pId = (OsFile*)pNew;
    OpenCounter(+1);
    return SQLITE_OK;
  }
}
#else /* SQLITE_ENABLE_LOCKING_STYLE */
static int fillInUnixFile(
  int h,                 /* Open file descriptor on file being opened */
................................................................................

#endif /* SQLITE_OMIT_DISKIO */
/***************************************************************************
** Everything above deals with file I/O.  Everything that follows deals
** with other miscellanous aspects of the operating system interface
****************************************************************************/











static int openDirectory(const char *zFilename, int *pFd){
  char *zDirname;
  int ii;
  int fd;

  zDirname = (char *)sqlite3_malloc(MAX_PATHNAME);
  if( !zDirname ){
................................................................................
  }
  sqlite3_free(zDirname);
  *pFd = fd;
  return (fd>0?SQLITE_OK:SQLITE_CANTOPEN);
}

/*


** Previously, the SQLite OS layer used three functions in place of this
** one:
**
**     sqlite3OsOpenReadWrite();
**     sqlite3OsOpenReadOnly();
**     sqlite3OsOpenExclusive();
**







|







 







|







 







<







 







|
|
|





|


|







 







|



|







 







|

|
|
|







 







|

|







 







>
>
>







 







|
>
|
>








|







 







|
>
|
>







 







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







 







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







 







|







 







|







 







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







 







|







 







|










|







 







|







 







|
|
|
|


|
|
|
|

|
|
|
|
|
|

<
<











|







 







|







 







|







 







|
|

|


|
|


|

|
|

<
<







 







|







 







|







 







|







 







|
|

|


|
|

|
|


|
|


|


|
|

<
<











|



|



|






|
|

|
|
|


|


|
|

<
<





<
>

|

|
|
>



|
>

|
|


<
<







 







<
<
<
<
<

<
<
<
<
<
<
<
<
|







 







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



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



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



|
|
|
|
>

<


<


<
<



<

>
>

>











|
|
|
|
|
|

>
|

<


|
<


|



|








|
|

<
<
<
|
|
|




|
|

<


<







 







<







 







>
>
>
>
>
>
>
>
>
>







 







>
>







77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
..
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
...
129
130
131
132
133
134
135

136
137
138
139
140
141
142
...
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
...
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
...
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
...
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
...
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
...
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
...
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
...
792
793
794
795
796
797
798












































































































































































799
800
801
802
803
804
805
....
1044
1045
1046
1047
1048
1049
1050






























1051
1052
1053
1054
1055
1056
1057
....
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
....
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
....
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506

1507
1508
1509
1510
1511
1512
1513
1514
1515
....
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
....
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
....
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
....
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787


1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
....
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
....
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
....
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
1886
1887
1888
....
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
....
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
....
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
....
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998


1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038


2039
2040
2041
2042
2043

2044
2045
2046
2047
2048
2049
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060


2061
2062
2063
2064
2065
2066
2067
....
2075
2076
2077
2078
2079
2080
2081





2082








2083
2084
2085
2086
2087
2088
2089
2090
....
2098
2099
2100
2101
2102
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
2154

2155
2156


2157
2158
2159

2160
2161
2162
2163
2164
2165
2166
2167
2168
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
....
2277
2278
2279
2280
2281
2282
2283

2284
2285
2286
2287
2288
2289
2290
....
2321
2322
2323
2324
2325
2326
2327
2328
2329
2330
2331
2332
2333
2334
2335
2336
2337
2338
2339
2340
2341
2342
2343
2344
....
2359
2360
2361
2362
2363
2364
2365
2366
2367
2368
2369
2370
2371
2372
2373
2374
/*
** Maximum supported path-length.
*/
#define MAX_PATHNAME 512


/*
** The unixFile structure is subclass of sqlite3_file specific for the unix
** protability layer.
*/
typedef struct unixFile unixFile;
struct unixFile {
  sqlite3_io_methods const *pMethod;  /* Always the first entry */
  struct openCnt *pOpen;    /* Info about all open fd's on this inode */
  struct lockInfo *pLock;   /* Info about locks on this inode */
................................................................................
#endif /* SQLITE_ENABLE_LOCKING_STYLE */
  int h;                    /* The file descriptor */
  unsigned char locktype;   /* The type of lock held on this fd */
  unsigned char isOpen;     /* True if needs to be closed */
  int dirfd;                /* File descriptor for the directory */
  i64 offset;               /* Seek offset */
#ifdef SQLITE_UNIX_THREADS
  pthread_t tid;            /* The thread that "owns" this unixFile */
#endif
};


/*
** Provide the ability to override some OS-layer functions during
** testing.  This is used to simulate OS crashes to verify that 
................................................................................

/*
** Do not include any of the File I/O interface procedures if the
** SQLITE_OMIT_DISKIO macro is defined (indicating that the database
** will be in-memory only)
*/
#ifndef SQLITE_OMIT_DISKIO


/*
** Define various macros that are missing from some systems.
*/
#ifndef O_LARGEFILE
# define O_LARGEFILE 0
#endif
................................................................................
#ifdef SQLITE_UNIX_THREADS
#define threadid pthread_self()
#else
#define threadid 0
#endif

/*
** Set or check the unixFile.tid field.  This field is set when an unixFile
** is first opened.  All subsequent uses of the unixFile verify that the
** same thread is operating on the unixFile.  Some operating systems do
** not allow locks to be overridden by other threads and that restriction
** means that sqlite3* database handles cannot be moved from one thread
** to another.  This logic makes sure a user does not try to do that
** by mistake.
**
** Version 3.3.1 (2006-01-15):  unixFile can be moved from one thread to
** another as long as we are running on a system that supports threads
** overriding each others locks (which now the most common behavior)
** or if no locks are held.  But the unixFile.pLock field needs to be
** recomputed because its key includes the thread-id.  See the 
** transferOwnership() function below for additional information
*/
#if defined(SQLITE_UNIX_THREADS)
# define SET_THREADID(X)   (X)->tid = pthread_self()
# define CHECK_THREADID(X) (threadsOverrideEachOthersLocks==0 && \
                            !pthread_equal((X)->tid, pthread_self()))
................................................................................
** specific inode of the database file (the inode is determined by the
** st_dev and st_ino fields of the stat structure that fstat() fills in)
** and check for locks already existing on that inode.  When locks are
** created or removed, we have to look at our own internal record of the
** locks to see if another thread has previously set a lock on that same
** inode.
**
** The sqlite3_file structure for POSIX is no longer just an integer file
** descriptor.  It is now a structure that holds the integer file
** descriptor and a pointer to a structure that describes the internal
** locks on the corresponding inode.  There is one locking structure
** per inode, so if the same inode is opened twice, both unixFile structures
** point to the same locking structure.  The locking structure keeps
** a reference count (so we will know when to delete it) and a "cnt"
** field that tells us its internal lock status.  cnt==0 means the
** file is unlocked.  cnt==-1 means the file has an exclusive lock.
** cnt>0 means there are cnt shared locks on the file.
**
** Any attempt to lock or unlock a file first checks the locking
................................................................................
** 2004-Jan-11:
** More recent discoveries about POSIX advisory locks.  (The more
** I discover, the more I realize the a POSIX advisory locks are
** an abomination.)
**
** If you close a file descriptor that points to a file that has locks,
** all locks on that file that are owned by the current process are
** released.  To work around this problem, each unixFile structure contains
** a pointer to an openCnt structure.  There is one openCnt structure
** per open inode, which means that multiple unixFile can point to a single
** openCnt.  When an attempt is made to close an unixFile, if there are
** other unixFile open on the same inode that are holding locks, the call
** to close() the file descriptor is deferred until all of the locks clear.
** The openCnt structure keeps a list of file descriptors that need to
** be closed and that list is walked (and cleared) when the last lock
** clears.
**
** First, under Linux threads, because each thread has a separate
** process ID, lock operations in one thread do not override locks
................................................................................
};

/*
** An instance of the following structure is allocated for each open
** inode on each thread with a different process ID.  (Threads have
** different process IDs on linux, but not on most other unixes.)
**
** A single inode can have multiple file descriptors, so each unixFile
** structure contains a pointer to an instance of this object and this
** object keeps a count of the number of unixFile pointing to it.
*/
struct lockInfo {
  struct lockKey key;  /* The lookup key */
  int cnt;             /* Number of SHARED locks held */
  int locktype;        /* One of SHARED_LOCK, RESERVED_LOCK etc. */
  int nRef;            /* Number of pointers to this structure */
};
................................................................................
        flockLockingStyle,           /* use flock() */
        dotlockLockingStyle,         /* use <file>.lock files */
        noLockingStyle,              /* useful for read-only file system */
        unsupportedLockingStyle      /* indicates unsupported file system */
} sqlite3LockingStyle;
#endif /* SQLITE_ENABLE_LOCKING_STYLE */

/*
** Helper functions to obtain and relinquish the global mutex.
*/
static void enterMutex(){
  sqlite3_mutex_enter(sqlite3_mutex_alloc(SQLITE_MUTEX_GLOBAL));
}
static void leaveMutex(){
  sqlite3_mutex_leave(sqlite3_mutex_alloc(SQLITE_MUTEX_GLOBAL));
}

................................................................................
}

#ifdef SQLITE_ENABLE_LOCKING_STYLE
/*
** Tests a byte-range locking query to see if byte range locks are 
** supported, if not we fall back to dotlockLockingStyle.
*/
static sqlite3LockingStyle sqlite3TestLockingStyle(
  const char *filePath, 
  int fd
){
  /* test byte-range lock using fcntl */
  struct flock lockInfo;
  
  lockInfo.l_len = 1;
  lockInfo.l_start = 0;
  lockInfo.l_whence = SEEK_SET;
  lockInfo.l_type = F_RDLCK;
  
  if( fcntl(fd, F_GETLK, &lockInfo)!=-1 ) {
    return posixLockingStyle;
  } 
  
  /* testing for flock can give false positives.  So if if the above test
  ** fails, then we fall back to using dot-lock style locking.
  */  
  return dotlockLockingStyle;
................................................................................
/* 
** Examines the f_fstypename entry in the statfs structure as returned by 
** stat() for the file system hosting the database file, assigns the 
** appropriate locking style based on it's value.  These values and 
** assignments are based on Darwin/OSX behavior and have not been tested on 
** other systems.
*/
static sqlite3LockingStyle sqlite3DetectLockingStyle(
  const char *filePath, 
  int fd
){

#ifdef SQLITE_FIXED_LOCKING_STYLE
  return (sqlite3LockingStyle)SQLITE_FIXED_LOCKING_STYLE;
#else
  struct statfs fsInfo;

  if (statfs(filePath, &fsInfo) == -1)
................................................................................
  }
}
#else
  /* On single-threaded builds, ownership transfer is a no-op */
# define transferOwnership(X) SQLITE_OK
#endif













































































































































































/*
** Seek to the offset in id->offset then read cnt bytes into pBuf.
** Return the number of bytes actually read.  Update the offset.
*/
static int seekAndRead(unixFile *id, sqlite3_int64 offset, void *pBuf, int cnt){
  int got;
  i64 newOffset;
................................................................................
#endif
    close(pFile->dirfd);  /* Only need to sync once, so close the directory */
    pFile->dirfd = -1;    /* when we are done. */
  }
  return SQLITE_OK;
}































/*
** Truncate an open file to a specified size
*/
static int unixTruncate(sqlite3_file *id, i64 nByte){
  int rc;
  assert( id );
  rc = ftruncate(((unixFile*)id)->h, (off_t)nByte);
................................................................................

  assert( pFile );
  OSTRACE7("LOCK    %d %s was %s(%s,%d) pid=%d\n", pFile->h,
      locktypeName(locktype), locktypeName(pFile->locktype),
      locktypeName(pLock->locktype), pLock->cnt , getpid());

  /* If there is already a lock of this type or more restrictive on the
  ** unixFile, do nothing. Don't use the end_lock: exit path, as
  ** enterMutex() hasn't been called yet.
  */
  if( pFile->locktype>=locktype ){
    OSTRACE3("LOCK    %d %s ok (already held)\n", pFile->h,
            locktypeName(locktype));
    return SQLITE_OK;
  }
................................................................................
  rc = transferOwnership(pFile);
  if( rc!=SQLITE_OK ){
    leaveMutex();
    return rc;
  }
  pLock = pFile->pLock;

  /* If some thread using this PID has a lock via a different unixFile*
  ** handle that precludes the requested lock, return BUSY.
  */
  if( (pFile->locktype!=pLock->locktype && 
          (pLock->locktype>=PENDING_LOCK || locktype>SHARED_LOCK))
  ){
    rc = SQLITE_BUSY;
    goto end_lock;
................................................................................
  unsigned char unLockFlag;         /* 1 = unlock, 0 = lock */
  unsigned char startEndFlag;       /* 1=rel to end of fork, 0=rel to start */
  int fd;                           /* file desc to assoc this lock with */
};

#define afpfsByteRangeLock2FSCTL        _IOWR('z', 23, struct ByteRangeLockPB2)

/* 
** Return 0 on success, 1 on failure.  To match the behavior of the 
** normal posix file locking (used in unixLock for example), we should 
** provide 'richer' return codes - specifically to differentiate between
** 'file busy' and 'file system error' results.
*/
static int _AFPFSSetLock(
  const char *path, 
  int fd, 
  unsigned long long offset, 
  unsigned long long length, 

  int setLockFlag
){
  struct ByteRangeLockPB2       pb;
  int                     err;
  
  pb.unLockFlag = setLockFlag ? 0 : 1;
  pb.startEndFlag = 0;
  pb.offset = offset;
  pb.length = length; 
................................................................................

/*
 ** This routine checks if there is a RESERVED lock held on the specified
 ** file by this or any other process. If such a lock is held, return
 ** non-zero.  If the file is unlocked or holds only SHARED locks, then
 ** return zero.
 */
static int afpUnixCheckReservedLock(sqlite3_file *id){
  int r = 0;
  unixFile *pFile = (unixFile*)id;
  
  assert( pFile ); 
  afpLockingContext *context = (afpLockingContext *) pFile->lockingContext;
  
  /* Check if a thread in this process holds such a lock */
................................................................................
  OSTRACE3("TEST WR-LOCK %d %d\n", pFile->h, r);
  
  return r;
}

/* AFP-style locking following the behavior of unixLock, see the unixLock 
** function comments for details of lock management. */
static int afpUnixLock(sqlite3_file *id, int locktype)
{
  int rc = SQLITE_OK;
  unixFile *pFile = (unixFile*)id;
  afpLockingContext *context = (afpLockingContext *) pFile->lockingContext;
  int gotPendingLock = 0;
  
  assert( pFile );
  OSTRACE5("LOCK    %d %s was %s pid=%d\n", pFile->h,
         locktypeName(locktype), locktypeName(pFile->locktype), getpid());  
  /* If there is already a lock of this type or more restrictive on the
    ** unixFile, do nothing. Don't use the afp_end_lock: exit path, as
    ** enterMutex() hasn't been called yet.
    */
  if( pFile->locktype>=locktype ){
    OSTRACE3("LOCK    %d %s ok (already held)\n", pFile->h,
           locktypeName(locktype));
    return SQLITE_OK;
  }
................................................................................
/*
 ** Lower the locking level on file descriptor pFile to locktype.  locktype
 ** must be either NO_LOCK or SHARED_LOCK.
 **
 ** If the locking level of the file descriptor is already at or below
 ** the requested locking level, this routine is a no-op.
 */
static int afpUnixUnlock(sqlite3_file *id, int locktype) {
  struct flock lock;
  int rc = SQLITE_OK;
  unixFile *pFile = (unixFile*)id;
  afpLockingContext *context = (afpLockingContext *) pFile->lockingContext;

  assert( pFile );
  OSTRACE5("UNLOCK  %d %d was %d pid=%d\n", pFile->h, locktype,
................................................................................
  leaveMutex();
  return rc;
}

/*
 ** Close a file & cleanup AFP specific locking context 
 */
static int afpUnixClose(sqlite3_file *id) {
  unixFile *pFile = (unixFile*)pId;

  if( !pFile ) return SQLITE_OK;
  afpUnixUnlock(*pId, NO_LOCK);
  /* free the AFP locking structure */
  if (pFile->lockingContext != NULL) {
    if (((afpLockingContext *)pFile->lockingContext)->filePath != NULL)
      sqlite3_free(((afpLockingContext*)pFile->lockingContext)->filePath);
    sqlite3_free(pFile->lockingContext);
  }

  if( pFile->dirfd>=0 ) close(pFile->dirfd);
  pFile->dirfd = -1;
  close(pFile->h);
  pFile->isOpen = 0;
  OSTRACE2("CLOSE   %-3d\n", pFile->h);
  OpenCounter(-1);


  return SQLITE_OK;
}


#pragma mark flock() style locking

/*
 ** The flockLockingContext is not used
 */
typedef void flockLockingContext;

static int flockUnixCheckReservedLock(sqlite3_file *id) {
  unixFile *pFile = (unixFile*)id;
  
  if (pFile->locktype == RESERVED_LOCK) {
    return 1; /* already have a reserved lock */
  } else {
    /* attempt to get the lock */
    int rc = flock(pFile->h, LOCK_EX | LOCK_NB);
................................................................................
      flock(pFile->h, LOCK_UN);
      return 0;  /* no one has it reserved */
    }
    return 1; /* someone else might have it reserved */
  }
}

static int flockUnixLock(sqlite3_file *id, int locktype) {
  unixFile *pFile = (unixFile*)id;
  
  /* if we already have a lock, it is exclusive.  
  ** Just adjust level and punt on outta here. */
  if (pFile->locktype > NO_LOCK) {
    pFile->locktype = locktype;
    return SQLITE_OK;
................................................................................
  } else {
    /* got it, set the type and return ok */
    pFile->locktype = locktype;
    return SQLITE_OK;
  }
}

static int flockUnixUnlock(sqlite3_file *id, int locktype) {
  unixFile *pFile = (unixFile*)id;
  
  assert( locktype<=SHARED_LOCK );
  
  /* no-op if possible */
  if( pFile->locktype==locktype ){
    return SQLITE_OK;
................................................................................
    return SQLITE_OK;
  }
}

/*
 ** Close a file.
 */
static int flockUnixClose(sqlite3_file *pId) {
  unixFile *pFile = (unixFile*)*pId;
  
  if( !pFile ) return SQLITE_OK;
  flockUnixUnlock(*pId, NO_LOCK);
  
  if( pFile->dirfd>=0 ) close(pFile->dirfd);
  pFile->dirfd = -1;
  enterMutex();
  
  close(pFile->h);  
  leaveMutex();
  pFile->isOpen = 0;
  OSTRACE2("CLOSE   %-3d\n", pFile->h);
  OpenCounter(-1);


  return SQLITE_OK;
}

#pragma mark Old-School .lock file based locking

/*
 ** The dotlockLockingContext structure contains all dotlock (.lock) lock
................................................................................
 */
typedef struct dotlockLockingContext dotlockLockingContext;
struct dotlockLockingContext {
  char *lockPath;
};


static int dotlockUnixCheckReservedLock(sqlite3_file *id) {
  unixFile *pFile = (unixFile*)id;
  dotlockLockingContext *context = 
    (dotlockLockingContext *) pFile->lockingContext;
  
  if (pFile->locktype == RESERVED_LOCK) {
    return 1; /* already have a reserved lock */
  } else {
................................................................................
      return 1;
    else
      /* file does not exist, we could have it if we want it */
      return 0;
  }
}

static int dotlockUnixLock(sqlite3_file *id, int locktype) {
  unixFile *pFile = (unixFile*)id;
  dotlockLockingContext *context = 
    (dotlockLockingContext *) pFile->lockingContext;
  
  /* if we already have a lock, it is exclusive.  
  ** Just adjust level and punt on outta here. */
  if (pFile->locktype > NO_LOCK) {
................................................................................
  close(fd);
  
  /* got it, set the type and return ok */
  pFile->locktype = locktype;
  return SQLITE_OK;
}

static int dotlockUnixUnlock(sqlite3_file *id, int locktype) {
  unixFile *pFile = (unixFile*)id;
  dotlockLockingContext *context = 
    (dotlockLockingContext *) pFile->lockingContext;
  
  assert( locktype<=SHARED_LOCK );
  
  /* no-op if possible */
................................................................................
  pFile->locktype = NO_LOCK;
  return SQLITE_OK;
}

/*
 ** Close a file.
 */
static int dotlockUnixClose(sqlite3_file *id) {
  unixFile *pFile = (unixFile*)id;
  
  if( !pFile ) return SQLITE_OK;
  dotlockUnixUnlock(*pId, NO_LOCK);
  /* free the dotlock locking structure */
  if (pFile->lockingContext != NULL) {
    if (((dotlockLockingContext *)pFile->lockingContext)->lockPath != NULL)
      sqlite3_free( ( (dotlockLockingContext *)
        pFile->lockingContext)->lockPath);
    sqlite3_free(pFile->lockingContext);
  }
  
  if( pFile->dirfd>=0 ) close(pFile->dirfd);
  pFile->dirfd = -1;
  enterMutex();
  
  close(pFile->h);
  
  leaveMutex();
  pFile->isOpen = 0;
  OSTRACE2("CLOSE   %-3d\n", pFile->h);
  OpenCounter(-1);


  return SQLITE_OK;
}


#pragma mark No locking

/*
 ** The nolockLockingContext is void
 */
typedef void nolockLockingContext;

static int nolockUnixCheckReservedLock(sqlite3_file *id) {
  return 0;
}

static int nolockUnixLock(sqlite3_file *id, int locktype) {
  return SQLITE_OK;
}

static int nolockUnixUnlock(sqlite3_file *id, int locktype) {
  return SQLITE_OK;
}

/*
 ** Close a file.
 */
static int nolockUnixClose(sqlite3_file *id) {
  unixFile *pFile = (unixFile*)id;
  
  if( !pFile ) return SQLITE_OK;
  if( pFile->dirfd>=0 ) close(pFile->dirfd);
  pFile->dirfd = -1;
  enterMutex();
  
  close(pFile->h);
  
  leaveMutex();
  pFile->isOpen = 0;
  OSTRACE2("CLOSE   %-3d\n", pFile->h);
  OpenCounter(-1);


  return SQLITE_OK;
}

#endif /* SQLITE_ENABLE_LOCKING_STYLE */



/*
** TODO: xBreakLock() for this vfs.
*/
static int unixBreakLock(sqlite3_file *id){
  assert(!"TODO: unixBreakLock()");
  return 0;
}

/*
** Return an integer that indices the type of lock currently held
** by this handle.  (Used for testing and analysis only.)
*/
static int unixLockState(sqlite3_file *id){
  return ((unixFile*)id)->locktype;
}



/*
** Return the sector size in bytes of the underlying block device for
** the specified file. This is almost always 512 bytes, but may be
** larger for some devices.
**
** SQLite code assumes this function cannot fail. It also assumes that
** if two files are created in the same file-system directory (i.e.
................................................................................
/*
** Return the device characteristics for the file. This is always 0.
*/
static int unixDeviceCharacteristics(sqlite3_file *id){
  return 0;
}






/*








** This vector defines all the methods that can operate on an sqlite3_file
** for unix.
*/
static const sqlite3_io_methods sqlite3UnixIoMethod = {
  1,                        /* iVersion */
  unixClose,
  unixRead,
  unixWrite,
................................................................................
  unixLockState,
  unixSectorSize,
  unixDeviceCharacteristics
};

#ifdef SQLITE_ENABLE_LOCKING_STYLE
/*
** This vector defines all the methods that can operate on an sqlite3_file
** for unix with AFP style file locking.
*/
static const sqlite3_io_methods sqlite3AFPLockingUnixIoMethod = {
  1,                        /* iVersion */
  unixClose,

  unixRead,
  unixWrite,

  unixTruncate,
  unixSync,


  unixFileSize,
  afpUnixLock,
  afpUnixUnlock,

  afpUnixCheckReservedLock,
  unixBreakLock,
  unixLockState,
  unixSectorSize,
  unixDeviceCharacteristics
};

/*
** This vector defines all the methods that can operate on an sqlite3_file
** for unix with flock() style file locking.
*/
static const sqlite3_io_methods sqlite3FlockLockingUnixIoMethod = {
  1,                        /* iVersion */
  flockUnixClose,

  unixRead,
  unixWrite,

  unixTruncate,
  unixSync,


  unixFileSize,
  flockUnixLock,
  flockUnixUnlock,

  flockUnixCheckReservedLock,
  unixBreakLock,
  unixLockState,
  unixSectorSize,
  unixDeviceCharacteristics
};

/*
** This vector defines all the methods that can operate on an sqlite3_file
** for unix with dotlock style file locking.
*/
static const sqlite3_io_methods sqlite3DotlockLockingUnixIoMethod = {
  1,                        /* iVersion */
  dotlockUnixClose,

  unixRead,
  unixWrite,

  unixTruncate,
  unixSync,


  unixFileSize,
  dotlockUnixLock,
  dotlockUnixUnlock,

  dotlockUnixCheckReservedLock,
  unixBreakLock,
  unixLockState,
  unixSectorSize,
  unixDeviceCharacteristics
};

/*
** This vector defines all the methods that can operate on an sqlite3_file
** for unix with dotlock style file locking.
*/
static const sqlite3_io_methods sqlite3NolockLockingUnixIoMethod = {
  1,                        /* iVersion */
  nolockUnixClose,

  unixRead,
  unixWrite,

  unixTruncate,
  unixSync,


  unixFileSize,
  nolockUnixLock,
  nolockUnixUnlock,

  nolockUnixCheckReservedLock,
  unixBreakLock,
  unixLockState,
  unixSectorSize,
  unixDeviceCharacteristics
};

#endif /* SQLITE_ENABLE_LOCKING_STYLE */

/*
** Allocate memory for a new unixFile and initialize that unixFile.
** Write a pointer to the new unixFile into *pId.
** If we run out of memory, close the file and return an error.
*/
#ifdef SQLITE_ENABLE_LOCKING_STYLE
/* 
** When locking extensions are enabled, the filepath and locking style 
** are needed to determine the unixFile pMethod to use for locking operations.
** The locking-style specific lockingContext data structure is created 
** and assigned here also.
*/
static int fillInUnixFile(
  int h,                  /* Open file descriptor of file being opened */
  int dirfd,              /* Directory file descriptor */
  sqlite3_file *pId,      /* Write completed initialization here */
  const char *zFilename,  /* Name of the file being opened */

){
  sqlite3LockingStyle lockingStyle;
  unixFile *pNew = (unixFile *)pId;

  int rc;

  memset(pNew, 0, sizeof(unixFile));
  lockingStyle = sqlite3DetectLockingStyle(zFilename, h);
  if ( lockingStyle == posixLockingStyle ) {
    enterMutex();
    rc = findLockInfo(h, &pNew->pLock, &pNew->pOpen);
    leaveMutex();
    if( rc ){
      close(h);
      unlink(zFilename);
      return SQLITE_NOMEM;
    }
  } else {
    /*  pLock and pOpen are only used for posix advisory locking */
    pNew->pLock = NULL;
    pNew->pOpen = NULL;
  }



  pNew->dirfd = -1;
  pNew->h = h;
  SET_THREADID(pNew);
  pNew = sqlite3_malloc( sizeof(unixFile) );
  if( pNew==0 ){
    close(h);
    enterMutex();
    releaseLockInfo(pNew->pLock);
    releaseOpenCnt(pNew->pOpen);
    leaveMutex();

    return SQLITE_NOMEM;
  }else{

    switch(lockingStyle) {
      case afpLockingStyle: {
        /* afp locking uses the file path so it needs to be included in
        ** the afpLockingContext */
        int nFilename;
        pNew->pMethod = &sqlite3AFPLockingUnixIoMethod;
        pNew->lockingContext = 
................................................................................
        pNew->pMethod = &sqlite3UnixIoMethod;
        break;
      case noLockingStyle:
      case unsupportedLockingStyle:
      default: 
        pNew->pMethod = &sqlite3NolockLockingUnixIoMethod;
    }

    OpenCounter(+1);
    return SQLITE_OK;
  }
}
#else /* SQLITE_ENABLE_LOCKING_STYLE */
static int fillInUnixFile(
  int h,                 /* Open file descriptor on file being opened */
................................................................................

#endif /* SQLITE_OMIT_DISKIO */
/***************************************************************************
** Everything above deals with file I/O.  Everything that follows deals
** with other miscellanous aspects of the operating system interface
****************************************************************************/

/*
** Open a file descriptor to the directory containing file zFilename.
** If successful, *pFd is set to the opened file descriptor and
** SQLITE_OK is returned. If an error occurs, either SQLITE_NOMEM
** or SQLITE_CANTOPEN is returned and *pFd is set to an undefined
** value.
**
** If SQLITE_OK is returned, the caller is responsible for closing
** the file descriptor *pFd using close().
*/
static int openDirectory(const char *zFilename, int *pFd){
  char *zDirname;
  int ii;
  int fd;

  zDirname = (char *)sqlite3_malloc(MAX_PATHNAME);
  if( !zDirname ){
................................................................................
  }
  sqlite3_free(zDirname);
  *pFd = fd;
  return (fd>0?SQLITE_OK:SQLITE_CANTOPEN);
}

/*
** Open the file zPath.
** 
** Previously, the SQLite OS layer used three functions in place of this
** one:
**
**     sqlite3OsOpenReadWrite();
**     sqlite3OsOpenReadOnly();
**     sqlite3OsOpenExclusive();
**