1FCNTL(3P) POSIX Programmer's Manual FCNTL(3P)
2
6 This manual page is part of the POSIX Programmer's Manual. The Linux
7 implementation of this interface may differ (consult the corresponding
8 Linux manual page for details of Linux behavior), or the interface may
9 not be implemented on Linux.
10
13 fcntl — file control
14
16 #include <fcntl.h>
17 int fcntl(int fildes, int cmd, ...);
19
21 The fcntl() function shall perform the operations described below on
22 open files. The fildes argument is a file descriptor.
23 The available values for cmd are defined in <fcntl.h> and are as fol‐
25 lows:
26 F_DUPFD Return a new file descriptor which shall be the lowest
28 numbered available (that is, not already open) file
29 descriptor greater than or equal to the third argument,
30 arg, taken as an integer of type int. The new file
31 descriptor shall refer to the same open file description
32 as the original file descriptor, and shall share any
33 locks. The FD_CLOEXEC flag associated with the new file
34 descriptor shall be cleared to keep the file open across
35 calls to one of the exec functions.
36 F_DUPFD_CLOEXEC
38 Like F_DUPFD, but the FD_CLOEXEC flag associated with the
39 new file descriptor shall be set.
40 F_GETFD Get the file descriptor flags defined in <fcntl.h> that
42 are associated with the file descriptor fildes. File
43 descriptor flags are associated with a single file
44 descriptor and do not affect other file descriptors that
45 refer to the same file.
46 F_SETFD Set the file descriptor flags defined in <fcntl.h>, that
48 are associated with fildes, to the third argument, arg,
49 taken as type int. If the FD_CLOEXEC flag in the third
50 argument is 0, the file descriptor shall remain open
51 across the exec functions; otherwise, the file descriptor
52 shall be closed upon successful execution of one of the
53 exec functions.
54 F_GETFL Get the file status flags and file access modes, defined
56 in <fcntl.h>, for the file description associated with
57 fildes. The file access modes can be extracted from the
58 return value using the mask O_ACCMODE, which is defined
59 in <fcntl.h>. File status flags and file access modes
60 are associated with the file description and do not
61 affect other file descriptors that refer to the same file
62 with different open file descriptions. The flags returned
63 may include non-standard file status flags which the
64 application did not set, provided that these additional
65 flags do not alter the behavior of a conforming applica‐
66 tion.
67 F_SETFL Set the file status flags, defined in <fcntl.h>, for the
69 file description associated with fildes from the corre‐
70 sponding bits in the third argument, arg, taken as type
71 int. Bits corresponding to the file access mode and the
72 file creation flags, as defined in <fcntl.h>, that are
73 set in arg shall be ignored. If any bits in arg other
74 than those mentioned here are changed by the application,
75 the result is unspecified. If fildes does not support
76 non-blocking operations, it is unspecified whether the
77 O_NONBLOCK flag will be ignored.
78 F_GETOWN If fildes refers to a socket, get the process or process
80 group ID specified to receive SIGURG signals when out-of-
81 band data is available. Positive values indicate a
82 process ID; negative values, other than −1, indicate a
83 process group ID. If fildes does not refer to a socket,
84 the results are unspecified.
85 F_SETOWN If fildes refers to a socket, set the process or process
87 group ID specified to receive SIGURG signals when out-of-
88 band data is available, using the value of the third
89 argument, arg, taken as type int. Positive values indi‐
90 cate a process ID; negative values, other than −1, indi‐
91 cate a process group ID. If fildes does not refer to a
92 socket, the results are unspecified.
93 The following values for cmd are available for advisory record locking.
95 Record locking shall be supported for regular files, and may be sup‐
96 ported for other files.
97 F_GETLK Get the first lock which blocks the lock description
99 pointed to by the third argument, arg, taken as a pointer
100 to type struct flock, defined in <fcntl.h>. The informa‐
101 tion retrieved shall overwrite the information passed to
102 fcntl() in the structure flock. If no lock is found that
103 would prevent this lock from being created, then the
104 structure shall be left unchanged except for the lock
105 type which shall be set to F_UNLCK.
106 F_SETLK Set or clear a file segment lock according to the lock
108 description pointed to by the third argument, arg, taken
109 as a pointer to type struct flock, defined in <fcntl.h>.
110 F_SETLK can establish shared (or read) locks (F_RDLCK) or
111 exclusive (or write) locks (F_WRLCK), as well as to
112 remove either type of lock (F_UNLCK). F_RDLCK, F_WRLCK,
113 and F_UNLCK are defined in <fcntl.h>. If a shared or
114 exclusive lock cannot be set, fcntl() shall return imme‐
115 diately with a return value of −1.
116 F_SETLKW This command shall be equivalent to F_SETLK except that
118 if a shared or exclusive lock is blocked by other locks,
119 the thread shall wait until the request can be satisfied.
120 If a signal that is to be caught is received while
121 fcntl() is waiting for a region, fcntl() shall be inter‐
122 rupted. Upon return from the signal handler, fcntl()
123 shall return −1 with errno set to [EINTR], and the lock
124 operation shall not be done.
125 Additional implementation-defined values for cmd may be defined in
127 <fcntl.h>. Their names shall start with F_.
128 When a shared lock is set on a segment of a file, other processes shall
130 be able to set shared locks on that segment or a portion of it. A
131 shared lock prevents any other process from setting an exclusive lock
132 on any portion of the protected area. A request for a shared lock shall
133 fail if the file descriptor was not opened with read access.
134 An exclusive lock shall prevent any other process from setting a shared
136 lock or an exclusive lock on any portion of the protected area. A
137 request for an exclusive lock shall fail if the file descriptor was not
138 opened with write access.
139 The structure flock describes the type (l_type), starting offset
141 (l_whence), relative offset (l_start), size (l_len), and process ID
142 (l_pid) of the segment of the file to be affected.
143 The value of l_whence is SEEK_SET, SEEK_CUR, or SEEK_END, to indicate
145 that the relative offset l_start bytes shall be measured from the start
146 of the file, current position, or end of the file, respectively. The
147 value of l_len is the number of consecutive bytes to be locked. The
148 value of l_len may be negative (where the definition of off_t permits
149 negative values of l_len). The l_pid field is only used with F_GETLK
150 to return the process ID of the process holding a blocking lock. After
151 a successful F_GETLK request, when a blocking lock is found, the values
152 returned in the flock structure shall be as follows:
153 l_type Type of blocking lock found.
155 l_whence SEEK_SET.
157 l_start Start of the blocking lock.
159 l_len Length of the blocking lock.
161 l_pid Process ID of the process that holds the blocking lock.
163 If the command is F_SETLKW and the process must wait for another
165 process to release a lock, then the range of bytes to be locked shall
166 be determined before the fcntl() function blocks. If the file size or
167 file descriptor seek offset change while fcntl() is blocked, this shall
168 not affect the range of bytes locked.
169 If l_len is positive, the area affected shall start at l_start and end
171 at l_start+l_len−1. If l_len is negative, the area affected shall
172 start at l_start+l_len and end at l_start−1. Locks may start and
173 extend beyond the current end of a file, but shall not extend before
174 the beginning of the file. A lock shall be set to extend to the largest
175 possible value of the file offset for that file by setting l_len to 0.
176 If such a lock also has l_start set to 0 and l_whence is set to
177 SEEK_SET, the whole file shall be locked.
178 There shall be at most one type of lock set for each byte in the file.
180 Before a successful return from an F_SETLK or an F_SETLKW request when
181 the calling process has previously existing locks on bytes in the
182 region specified by the request, the previous lock type for each byte
183 in the specified region shall be replaced by the new lock type. As
184 specified above under the descriptions of shared locks and exclusive
185 locks, an F_SETLK or an F_SETLKW request (respectively) shall fail or
186 block when another process has existing locks on bytes in the specified
187 region and the type of any of those locks conflicts with the type spec‐
188 ified in the request.
189 All locks associated with a file for a given process shall be removed
191 when a file descriptor for that file is closed by that process or the
192 process holding that file descriptor terminates. Locks are not inher‐
193 ited by a child process.
194 A potential for deadlock occurs if a process controlling a locked
196 region is put to sleep by attempting to lock the locked region of
197 another process. If the system detects that sleeping until a locked
198 region is unlocked would cause a deadlock, fcntl() shall fail with an
199 [EDEADLK] error.
200 An unlock (F_UNLCK) request in which l_len is non-zero and the offset
202 of the last byte of the requested segment is the maximum value for an
203 object of type off_t, when the process has an existing lock in which
204 l_len is 0 and which includes the last byte of the requested segment,
205 shall be treated as a request to unlock from the start of the requested
206 segment with an l_len equal to 0. Otherwise, an unlock (F_UNLCK)
207 request shall attempt to unlock only the requested segment.
208 When the file descriptor fildes refers to a shared memory object, the
210 behavior of fcntl() shall be the same as for a regular file except the
211 effect of the following values for the argument cmd shall be unspeci‐
212 fied: F_SETFL, F_GETLK, F_SETLK, and F_SETLKW.
213 If fildes refers to a typed memory object, the result of the fcntl()
215 function is unspecified.
216
218 Upon successful completion, the value returned shall depend on cmd as
219 follows:
220 F_DUPFD A new file descriptor.
222 F_DUPFD_CLOEXEC
224 A new file descriptor.
225 F_GETFD Value of flags defined in <fcntl.h>. The return value
227 shall not be negative.
228 F_SETFD Value other than −1.
230 F_GETFL Value of file status flags and access modes. The return
232 value is not negative.
233 F_SETFL Value other than −1.
235 F_GETLK Value other than −1.
237 F_SETLK Value other than −1.
239 F_SETLKW Value other than −1.
241 F_GETOWN Value of the socket owner process or process group; this
243 will not be −1.
244 F_SETOWN Value other than −1.
246 Otherwise, −1 shall be returned and errno set to indicate the error.
248
250 The fcntl() function shall fail if:
251 EACCES or EAGAIN
253 The cmd argument is F_SETLK; the type of lock (l_type) is a
254 shared (F_RDLCK) or exclusive (F_WRLCK) lock and the segment of
255 a file to be locked is already exclusive-locked by another
256 process, or the type is an exclusive lock and some portion of
257 the segment of a file to be locked is already shared-locked or
258 exclusive-locked by another process.
259 EBADF The fildes argument is not a valid open file descriptor, or the
261 argument cmd is F_SETLK or F_SETLKW, the type of lock, l_type,
262 is a shared lock (F_RDLCK), and fildes is not a valid file
263 descriptor open for reading, or the type of lock, l_type, is an
264 exclusive lock (F_WRLCK), and fildes is not a valid file
265 descriptor open for writing.
266 EINTR The cmd argument is F_SETLKW and the function was interrupted by
268 a signal.
269 EINVAL The cmd argument is invalid, or the cmd argument is F_DUPFD or
271 F_DUPFD_CLOEXEC and arg is negative or greater than or equal to
272 {OPEN_MAX}, or the cmd argument is F_GETLK, F_SETLK, or F_SETLKW
273 and the data pointed to by arg is not valid, or fildes refers to
274 a file that does not support locking.
275 EMFILE The argument cmd is F_DUPFD or F_DUPFD_CLOEXEC and all file
277 descriptors available to the process are currently open, or no
278 file descriptors greater than or equal to arg are available.
279 ENOLCK The argument cmd is F_SETLK or F_SETLKW and satisfying the lock
281 or unlock request would result in the number of locked regions
282 in the system exceeding a system-imposed limit.
283 EOVERFLOW
285 One of the values to be returned cannot be represented cor‐
286 rectly.
287 EOVERFLOW
289 The cmd argument is F_GETLK, F_SETLK, or F_SETLKW and the small‐
290 est or, if l_len is non-zero, the largest offset of any byte in
291 the requested segment cannot be represented correctly in an
292 object of type off_t.
293 The fcntl() function may fail if:
295 EDEADLK
297 The cmd argument is F_SETLKW, the lock is blocked by a lock from
298 another process, and putting the calling process to sleep to
299 wait for that lock to become free would cause a deadlock.
300 The following sections are informative.
302
304 Locking and Unlocking a File
305 The following example demonstrates how to place a lock on bytes 100 to
306 109 of a file and then later remove it. F_SETLK is used to perform a
307 non-blocking lock request so that the process does not have to wait if
308 an incompatible lock is held by another process; instead the process
309 can take some other action.
310 #include <stdlib.h>
312 #include <unistd.h>
313 #include <fcntl.h>
314 #include <errno.h>
315 #include <stdio.h>
316 int
318 main(int argc, char *argv[])
319 {
320 int fd;
321 struct flock fl;
322 fd = open("testfile", O_RDWR);
324 if (fd == -1)
325 /* Handle error */;
326 /* Make a non-blocking request to place a write lock
328 on bytes 100-109 of testfile */
329 fl.l_type = F_WRLCK;
331 fl.l_whence = SEEK_SET;
332 fl.l_start = 100;
333 fl.l_len = 10;
334 if (fcntl(fd, F_SETLK, &fl) == −1) {
336 if (errno == EACCES || errno == EAGAIN) {
337 printf("Already locked by another process\n");
338 /* We can't get the lock at the moment */
340 } else {
342 /* Handle unexpected error */;
343 }
344 } else { /* Lock was granted... */
345 /* Perform I/O on bytes 100 to 109 of file */
347 /* Unlock the locked bytes */
349 fl.l_type = F_UNLCK;
351 fl.l_whence = SEEK_SET;
352 fl.l_start = 100;
353 fl.l_len = 10;
354 if (fcntl(fd, F_SETLK, &fl) == −1)
355 /* Handle error */;
356 }
357 exit(EXIT_SUCCESS);
358 } /* main */
359 Setting the Close-on-Exec Flag
361 The following example demonstrates how to set the close-on-exec flag
362 for the file descriptor fd.
363 #include <unistd.h>
365 #include <fcntl.h>
366 ...
367 int flags;
368 flags = fcntl(fd, F_GETFD);
370 if (flags == −1)
371 /* Handle error */;
372 flags |= FD_CLOEXEC;
373 if (fcntl(fd, F_SETFD, flags) == −1)
374 /* Handle error */;"
375
377 The arg values to F_GETFD, F_SETFD, F_GETFL, and F_SETFL all represent
378 flag values to allow for future growth. Applications using these func‐
379 tions should do a read-modify-write operation on them, rather than
380 assuming that only the values defined by this volume of POSIX.1‐2008
381 are valid. It is a common error to forget this, particularly in the
382 case of F_SETFD. Some implementations set additional file status flags
383 to advise the application of default behavior, even though the applica‐
384 tion did not request these flags.
385
387 The ellipsis in the SYNOPSIS is the syntax specified by the ISO C stan‐
388 dard for a variable number of arguments. It is used because System V
389 uses pointers for the implementation of file locking functions.
390 This volume of POSIX.1‐2008 permits concurrent read and write access to
392 file data using the fcntl() function; this is a change from the 1984
393 /usr/group standard and early proposals. Without concurrency controls,
394 this feature may not be fully utilized without occasional loss of data.
395 Data losses occur in several ways. One case occurs when several pro‐
397 cesses try to update the same record, without sequencing controls; sev‐
398 eral updates may occur in parallel and the last writer ``wins''.
399 Another case is a bit-tree or other internal list-based database that
400 is undergoing reorganization. Without exclusive use to the tree segment
401 by the updating process, other reading processes chance getting lost in
402 the database when the index blocks are split, condensed, inserted, or
403 deleted. While fcntl() is useful for many applications, it is not
404 intended to be overly general and does not handle the bit-tree example
405 well.
406 This facility is only required for regular files because it is not
408 appropriate for many devices such as terminals and network connections.
409 Since fcntl() works with ``any file descriptor associated with that
411 file, however it is obtained'', the file descriptor may have been
412 inherited through a fork() or exec operation and thus may affect a file
413 that another process also has open.
414 The use of the open file description to identify what to lock requires
416 extra calls and presents problems if several processes are sharing an
417 open file description, but there are too many implementations of the
418 existing mechanism for this volume of POSIX.1‐2008 to use different
419 specifications.
420 Another consequence of this model is that closing any file descriptor
422 for a given file (whether or not it is the same open file description
423 that created the lock) causes the locks on that file to be relinquished
424 for that process. Equivalently, any close for any file/process pair
425 relinquishes the locks owned on that file for that process. But note
426 that while an open file description may be shared through fork(), locks
427 are not inherited through fork(). Yet locks may be inherited through
428 one of the exec functions.
429 The identification of a machine in a network environment is outside the
431 scope of this volume of POSIX.1‐2008. Thus, an l_sysid member, such as
432 found in System V, is not included in the locking structure.
433 Changing of lock types can result in a previously locked region being
435 split into smaller regions.
436 Mandatory locking was a major feature of the 1984 /usr/group standard.
438 For advisory file record locking to be effective, all processes that
440 have access to a file must cooperate and use the advisory mechanism
441 before doing I/O on the file. Enforcement-mode record locking is impor‐
442 tant when it cannot be assumed that all processes are cooperating. For
443 example, if one user uses an editor to update a file at the same time
444 that a second user executes another process that updates the same file
445 and if only one of the two processes is using advisory locking, the
446 processes are not cooperating. Enforcement-mode record locking would
447 protect against accidental collisions.
448 Secondly, advisory record locking requires a process using locking to
450 bracket each I/O operation with lock (or test) and unlock operations.
451 With enforcement-mode file and record locking, a process can lock the
452 file once and unlock when all I/O operations have been completed.
453 Enforcement-mode record locking provides a base that can be enhanced;
454 for example, with sharable locks. That is, the mechanism could be
455 enhanced to allow a process to lock a file so other processes could
456 read it, but none of them could write it.
457 Mandatory locks were omitted for several reasons:
459 1. Mandatory lock setting was done by multiplexing the set-group-ID
461 bit in most implementations; this was confusing, at best.
462 2. The relationship to file truncation as supported in 4.2 BSD was not
464 well specified.
465 3. Any publicly readable file could be locked by anyone. Many histori‐
467 cal implementations keep the password database in a publicly read‐
468 able file. A malicious user could thus prohibit logins. Another
469 possibility would be to hold open a long-distance telephone line.
470 4. Some demand-paged historical implementations offer memory mapped
472 files, and enforcement cannot be done on that type of file.
473 Since sleeping on a region is interrupted with any signal, alarm() may
475 be used to provide a timeout facility in applications requiring it.
476 This is useful in deadlock detection. Since implementation of full
477 deadlock detection is not always feasible, the [EDEADLK] error was made
478 optional.
479
481 None.
482
484 alarm(), close(), exec, open(), sigaction()
485 The Base Definitions volume of POSIX.1‐2008, <fcntl.h>, <signal.h>
487
489 Portions of this text are reprinted and reproduced in electronic form
490 from IEEE Std 1003.1, 2013 Edition, Standard for Information Technology
491 -- Portable Operating System Interface (POSIX), The Open Group Base
492 Specifications Issue 7, Copyright (C) 2013 by the Institute of Electri‐
493 cal and Electronics Engineers, Inc and The Open Group. (This is
494 POSIX.1-2008 with the 2013 Technical Corrigendum 1 applied.) In the
495 event of any discrepancy between this version and the original IEEE and
496 The Open Group Standard, the original IEEE and The Open Group Standard
497 is the referee document. The original Standard can be obtained online
498 at http://www.unix.org/online.html .
499 Any typographical or formatting errors that appear in this page are
501 most likely to have been introduced during the conversion of the source
502 files to man page format. To report such errors, see https://www.ker‐
503 nel.org/doc/man-pages/reporting_bugs.html .
504IEEE/The Open Group 2013 FCNTL(3P)