]> git.ipfire.org Git - thirdparty/man-pages.git/blob - man2/fcntl.2
fcntl.2: F_OFD_GETLK fails with EAGAIN when there is a lock conflict
[thirdparty/man-pages.git] / man2 / fcntl.2
1 '\" t
2 .\" This manpage is Copyright (C) 1992 Drew Eckhardt;
3 .\" and Copyright (C) 1993 Michael Haardt, Ian Jackson;
4 .\" and Copyright (C) 1998 Jamie Lokier;
5 .\" and Copyright (C) 2002-2010, 2014 Michael Kerrisk;
6 .\" and Copyright (C) 2014 Jeff Layton
7 .\"
8 .\" %%%LICENSE_START(VERBATIM)
9 .\" Permission is granted to make and distribute verbatim copies of this
10 .\" manual provided the copyright notice and this permission notice are
11 .\" preserved on all copies.
12 .\"
13 .\" Permission is granted to copy and distribute modified versions of this
14 .\" manual under the conditions for verbatim copying, provided that the
15 .\" entire resulting derived work is distributed under the terms of a
16 .\" permission notice identical to this one.
17 .\"
18 .\" Since the Linux kernel and libraries are constantly changing, this
19 .\" manual page may be incorrect or out-of-date. The author(s) assume no
20 .\" responsibility for errors or omissions, or for damages resulting from
21 .\" the use of the information contained herein. The author(s) may not
22 .\" have taken the same level of care in the production of this manual,
23 .\" which is licensed free of charge, as they might when working
24 .\" professionally.
25 .\"
26 .\" Formatted or processed versions of this manual, if unaccompanied by
27 .\" the source, must acknowledge the copyright and authors of this work.
28 .\" %%%LICENSE_END
29 .\"
30 .\" Modified 1993-07-24 by Rik Faith <faith@cs.unc.edu>
31 .\" Modified 1995-09-26 by Andries Brouwer <aeb@cwi.nl>
32 .\" and again on 960413 and 980804 and 981223.
33 .\" Modified 1998-12-11 by Jamie Lokier <jamie@imbolc.ucc.ie>
34 .\" Applied correction by Christian Ehrhardt - aeb, 990712
35 .\" Modified 2002-04-23 by Michael Kerrisk <mtk.manpages@gmail.com>
36 .\" Added note on F_SETFL and O_DIRECT
37 .\" Complete rewrite + expansion of material on file locking
38 .\" Incorporated description of F_NOTIFY, drawing on
39 .\" Stephen Rothwell's notes in Documentation/dnotify.txt.
40 .\" Added description of F_SETLEASE and F_GETLEASE
41 .\" Corrected and polished, aeb, 020527.
42 .\" Modified 2004-03-03 by Michael Kerrisk <mtk.manpages@gmail.com>
43 .\" Modified description of file leases: fixed some errors of detail
44 .\" Replaced the term "lease contestant" by "lease breaker"
45 .\" Modified, 27 May 2004, Michael Kerrisk <mtk.manpages@gmail.com>
46 .\" Added notes on capability requirements
47 .\" Modified 2004-12-08, added O_NOATIME after note from Martin Pool
48 .\" 2004-12-10, mtk, noted F_GETOWN bug after suggestion from aeb.
49 .\" 2005-04-08 Jamie Lokier <jamie@shareable.org>, mtk
50 .\" Described behavior of F_SETOWN/F_SETSIG in
51 .\" multithreaded processes, and generally cleaned
52 .\" up the discussion of F_SETOWN.
53 .\" 2005-05-20, Johannes Nicolai <johannes.nicolai@hpi.uni-potsdam.de>,
54 .\" mtk: Noted F_SETOWN bug for socket file descriptor in Linux 2.4
55 .\" and earlier. Added text on permissions required to send signal.
56 .\" 2009-09-30, Michael Kerrisk
57 .\" Note obsolete F_SETOWN behavior with threads.
58 .\" Document F_SETOWN_EX and F_GETOWN_EX
59 .\" 2010-06-17, Michael Kerrisk
60 .\" Document F_SETPIPE_SZ and F_GETPIPE_SZ.
61 .\"
62 .TH FCNTL 2 2014-05-21 "Linux" "Linux Programmer's Manual"
63 .SH NAME
64 fcntl \- manipulate file descriptor
65 .SH SYNOPSIS
66 .nf
67 .B #include <unistd.h>
68 .B #include <fcntl.h>
69 .sp
70 .BI "int fcntl(int " fd ", int " cmd ", ... /* " arg " */ );"
71 .fi
72 .SH DESCRIPTION
73 .BR fcntl ()
74 performs one of the operations described below on the open file descriptor
75 .IR fd .
76 The operation is determined by
77 .IR cmd .
78
79 .BR fcntl ()
80 can take an optional third argument.
81 Whether or not this argument is required is determined by
82 .IR cmd .
83 The required argument type is indicated in parentheses after each
84 .I cmd
85 name (in most cases, the required type is
86 .IR int ,
87 and we identify the argument using the name
88 .IR arg ),
89 or
90 .I void
91 is specified if the argument is not required.
92
93 Certain of the operations below are supported only since a particular
94 Linux kernel version.
95 The preferred method of checking whether the host kernel supports
96 a particular operation is to invoke
97 .BR fcntl ()
98 with the desired
99 .IR cmd
100 value and then test whether the call failed with
101 .BR EINVAL ,
102 indicating that the kernel does not recognize this value.
103 .SS Duplicating a file descriptor
104 .TP
105 .BR F_DUPFD " (\fIint\fP)"
106 Find the lowest numbered available file descriptor
107 greater than or equal to
108 .I arg
109 and make it be a copy of
110 .IR fd .
111 This is different from
112 .BR dup2 (2),
113 which uses exactly the descriptor specified.
114 .IP
115 On success, the new descriptor is returned.
116 .IP
117 See
118 .BR dup (2)
119 for further details.
120 .TP
121 .BR F_DUPFD_CLOEXEC " (\fIint\fP; since Linux 2.6.24)"
122 As for
123 .BR F_DUPFD ,
124 but additionally set the
125 close-on-exec flag for the duplicate descriptor.
126 Specifying this flag permits a program to avoid an additional
127 .BR fcntl ()
128 .B F_SETFD
129 operation to set the
130 .B FD_CLOEXEC
131 flag.
132 For an explanation of why this flag is useful,
133 see the description of
134 .B O_CLOEXEC
135 in
136 .BR open (2).
137 .SS File descriptor flags
138 The following commands manipulate the flags associated with
139 a file descriptor.
140 Currently, only one such flag is defined:
141 .BR FD_CLOEXEC ,
142 the close-on-exec flag.
143 If the
144 .B FD_CLOEXEC
145 bit is 0, the file descriptor will remain open across an
146 .BR execve (2),
147 otherwise it will be closed.
148 .TP
149 .BR F_GETFD " (\fIvoid\fP)"
150 Read the file descriptor flags;
151 .I arg
152 is ignored.
153 .TP
154 .BR F_SETFD " (\fIint\fP)"
155 Set the file descriptor flags to the value specified by
156 .IR arg .
157 .PP
158 In multithreaded programs, using
159 .BR fcntl ()
160 .B F_SETFD
161 to set the close-on-exec flag at the same time as another thread performs a
162 .BR fork (2)
163 plus
164 .BR execve (2)
165 is vulnerable to a race condition that may unintentionally leak
166 the file descriptor to the program executed in the child process.
167 See the discussion of the
168 .BR O_CLOEXEC
169 flag in
170 .BR open (2)
171 for details and a remedy to the problem.
172 .SS File status flags
173 Each open file description has certain associated status flags,
174 initialized by
175 .BR open (2)
176 .\" or
177 .\" .BR creat (2),
178 and possibly modified by
179 .BR fcntl ().
180 Duplicated file descriptors
181 (made with
182 .BR dup (2),
183 .BR fcntl (F_DUPFD),
184 .BR fork (2),
185 etc.) refer to the same open file description, and thus
186 share the same file status flags.
187
188 The file status flags and their semantics are described in
189 .BR open (2).
190 .TP
191 .BR F_GETFL " (\fIvoid\fP)"
192 Get the file access mode and the file status flags;
193 .I arg
194 is ignored.
195 .TP
196 .BR F_SETFL " (\fIint\fP)"
197 Set the file status flags to the value specified by
198 .IR arg .
199 File access mode
200 .RB ( O_RDONLY ", " O_WRONLY ", " O_RDWR )
201 and file creation flags
202 (i.e.,
203 .BR O_CREAT ", " O_EXCL ", " O_NOCTTY ", " O_TRUNC )
204 in
205 .I arg
206 are ignored.
207 On Linux this command can change only the
208 .BR O_APPEND ,
209 .BR O_ASYNC ,
210 .BR O_DIRECT ,
211 .BR O_NOATIME ,
212 and
213 .B O_NONBLOCK
214 flags.
215 It is not possible to change the
216 .BR O_DSYNC
217 and
218 .BR O_SYNC
219 flags; see BUGS, below.
220 .SS Advisory record locking
221 Linux implements traditional ("process-associated") UNIX record locks,
222 as standardized by POSIX.
223 For a Linux-specific alternative with better semantics,
224 see the discussion of open file description locks below.
225
226 .BR F_SETLK ,
227 .BR F_SETLKW ,
228 and
229 .BR F_GETLK
230 are used to acquire, release, and test for the existence of record
231 locks (also known as byte-range, file-segment, or file-region locks).
232 The third argument,
233 .IR lock ,
234 is a pointer to a structure that has at least the following fields
235 (in unspecified order).
236 .in +4n
237 .nf
238 .sp
239 struct flock {
240 ...
241 short l_type; /* Type of lock: F_RDLCK,
242 F_WRLCK, F_UNLCK */
243 short l_whence; /* How to interpret l_start:
244 SEEK_SET, SEEK_CUR, SEEK_END */
245 off_t l_start; /* Starting offset for lock */
246 off_t l_len; /* Number of bytes to lock */
247 pid_t l_pid; /* PID of process blocking our lock
248 (set by F_GETLK and F_OFD_GETLK) */
249 ...
250 };
251 .fi
252 .in
253 .P
254 The
255 .IR l_whence ", " l_start ", and " l_len
256 fields of this structure specify the range of bytes we wish to lock.
257 Bytes past the end of the file may be locked,
258 but not bytes before the start of the file.
259
260 .I l_start
261 is the starting offset for the lock, and is interpreted
262 relative to either:
263 the start of the file (if
264 .I l_whence
265 is
266 .BR SEEK_SET );
267 the current file offset (if
268 .I l_whence
269 is
270 .BR SEEK_CUR );
271 or the end of the file (if
272 .I l_whence
273 is
274 .BR SEEK_END ).
275 In the final two cases,
276 .I l_start
277 can be a negative number provided the
278 offset does not lie before the start of the file.
279
280 .I l_len
281 specifies the number of bytes to be locked.
282 If
283 .I l_len
284 is positive, then the range to be locked covers bytes
285 .I l_start
286 up to and including
287 .IR l_start + l_len \-1.
288 Specifying 0 for
289 .I l_len
290 has the special meaning: lock all bytes starting at the
291 location specified by
292 .IR l_whence " and " l_start
293 through to the end of file, no matter how large the file grows.
294
295 POSIX.1-2001 allows (but does not require)
296 an implementation to support a negative
297 .I l_len
298 value; if
299 .I l_len
300 is negative, the interval described by
301 .I lock
302 covers bytes
303 .IR l_start + l_len
304 up to and including
305 .IR l_start \-1.
306 This is supported by Linux since kernel versions 2.4.21 and 2.5.49.
307
308 The
309 .I l_type
310 field can be used to place a read
311 .RB ( F_RDLCK )
312 or a write
313 .RB ( F_WRLCK )
314 lock on a file.
315 Any number of processes may hold a read lock (shared lock)
316 on a file region, but only one process may hold a write lock
317 (exclusive lock).
318 An exclusive lock excludes all other locks,
319 both shared and exclusive.
320 A single process can hold only one type of lock on a file region;
321 if a new lock is applied to an already-locked region,
322 then the existing lock is converted to the new lock type.
323 (Such conversions may involve splitting, shrinking, or coalescing with
324 an existing lock if the byte range specified by the new lock does not
325 precisely coincide with the range of the existing lock.)
326 .TP
327 .BR F_SETLK " (\fIstruct flock *\fP)"
328 Acquire a lock (when
329 .I l_type
330 is
331 .B F_RDLCK
332 or
333 .BR F_WRLCK )
334 or release a lock (when
335 .I l_type
336 is
337 .BR F_UNLCK )
338 on the bytes specified by the
339 .IR l_whence ", " l_start ", and " l_len
340 fields of
341 .IR lock .
342 If a conflicting lock is held by another process,
343 this call returns \-1 and sets
344 .I errno
345 to
346 .B EACCES
347 or
348 .BR EAGAIN .
349 .TP
350 .BR F_SETLKW " (\fIstruct flock *\fP)"
351 As for
352 .BR F_SETLK ,
353 but if a conflicting lock is held on the file, then wait for that
354 lock to be released.
355 If a signal is caught while waiting, then the call is interrupted
356 and (after the signal handler has returned)
357 returns immediately (with return value \-1 and
358 .I errno
359 set to
360 .BR EINTR ;
361 see
362 .BR signal (7)).
363 .TP
364 .BR F_GETLK " (\fIstruct flock *\fP)"
365 On input to this call,
366 .I lock
367 describes a lock we would like to place on the file.
368 If the lock could be placed,
369 .BR fcntl ()
370 does not actually place it, but returns
371 .B F_UNLCK
372 in the
373 .I l_type
374 field of
375 .I lock
376 and leaves the other fields of the structure unchanged.
377
378 If one or more incompatible locks would prevent
379 this lock being placed, then
380 .BR fcntl ()
381 returns details about one of those locks in the
382 .IR l_type ", " l_whence ", " l_start ", and " l_len
383 fields of
384 .IR lock .
385 If the conflicting lock is a traditional (process-associated) record lock,
386 then the
387 .I l_pid
388 field is set to the PID of the process holding that lock.
389 If the conflicting lock is an open file description lock, then
390 .I l_pid
391 is set to \-1.
392 Note that the returned information
393 may already be out of date by the time the caller inspects it.
394 .P
395 In order to place a read lock,
396 .I fd
397 must be open for reading.
398 In order to place a write lock,
399 .I fd
400 must be open for writing.
401 To place both types of lock, open a file read-write.
402 .P
403 As well as being removed by an explicit
404 .BR F_UNLCK ,
405 record locks are automatically released when the process terminates.
406 .P
407 Record locks are not inherited by a child created via
408 .BR fork (2),
409 but are preserved across an
410 .BR execve (2).
411 .P
412 Because of the buffering performed by the
413 .BR stdio (3)
414 library, the use of record locking with routines in that package
415 should be avoided; use
416 .BR read (2)
417 and
418 .BR write (2)
419 instead.
420
421 The record locks described above are associated with the process
422 (unlike the open file description locks described below).
423 This has some unfortunate consequences:
424 .IP * 3
425 If a process closes
426 .I any
427 file descriptor referring to a file,
428 then all of the process's locks on that file are released,
429 regardless of the file descriptor(s) on which the locks were obtained.
430 .\" (Additional file descriptors referring to the same file
431 .\" may have been obtained by calls to
432 .\" .BR open "(2), " dup "(2), " dup2 "(2), or " fcntl ().)
433 This is bad: it means that a process can lose its locks on
434 a file such as
435 .I /etc/passwd
436 or
437 .I /etc/mtab
438 when for some reason a library function decides to open, read,
439 and close the same file.
440 .IP *
441 The threads in a process share locks.
442 In other words,
443 a multithreaded program can't use record locking to ensure
444 that threads don't simultaneously access the same region of a file.
445 .PP
446 Open file description locks solve both of these problems.
447 .SS Open file description locks (non-POSIX)
448 Open file description locks are advisory byte-range locks whose operation is
449 in most respects identical to the traditional record locks described above.
450 This lock type is Linux-specific,
451 and available since Linux 3.15.
452
453 The principal difference between the two lock types
454 is that whereas traditional record locks
455 are associated with a process,
456 open file description locks are associated with the
457 open file description on which they are acquired,
458 much like locks acquired with
459 .BR flock (2).
460 Consequently (and unlike traditional advisory record locks),
461 open file description locks are inherited across
462 .BR fork (2)
463 (and
464 .BR clone (2)
465 with
466 .BR CLONE_FILES ),
467 and are only automatically released on the last close
468 of the open file description,
469 instead of being released on any close of the file.
470 .PP
471 Open file description locks always conflict with traditional record locks,
472 even when they are acquired by the same process on the same file descriptor.
473
474 Open file description locks placed via the same open file description
475 (i.e., via the same file descriptor,
476 or via a duplicate of the file descriptor created by
477 .BR fork (2),
478 .BR dup (2),
479 .BR fcntl (2)
480 .BR F_DUPFD ,
481 and so on) are always compatible:
482 if a new lock is placed on an already locked region,
483 then the existing lock is converted to the new lock type.
484 (Such conversions may result in splitting, shrinking, or coalescing with
485 an existing lock as discussed above.)
486
487 On the other hand, open file description locks may conflict with
488 each other when they are acquired via different open file descriptions.
489 Thus, the threads in a multithreaded program can use
490 open file description locks to synchronize access to a file region
491 by having each thread perform its own
492 .BR open (2)
493 on the file and applying locks via the resulting file descriptor.
494 .PP
495 As with traditional advisory locks, the third argument to
496 .BR fcntl (),
497 .IR lock ,
498 is a pointer to an
499 .IR flock
500 structure.
501 By contrast with traditional record locks, the
502 .I l_pid
503 field of that structure must be set to zero
504 when using the commands described below.
505
506 The commands for working with open file description locks are analogous
507 to those used with traditional locks:
508 .TP
509 .BR F_OFD_SETLK " (\fIstruct flock *\fP)"
510 Acquire an open file description lock (when
511 .I l_type
512 is
513 .B F_RDLCK
514 or
515 .BR F_WRLCK )
516 or release an open file description lock (when
517 .I l_type
518 is
519 .BR F_UNLCK )
520 on the bytes specified by the
521 .IR l_whence ", " l_start ", and " l_len
522 fields of
523 .IR lock .
524 If a conflicting lock is held by another process,
525 this call returns \-1 and sets
526 .I errno
527 to
528 .BR EAGAIN .
529 .TP
530 .BR F_OFD_SETLKW " (\fIstruct flock *\fP)"
531 As for
532 .BR F_OFD_SETLK ,
533 but if a conflicting lock is held on the file, then wait for that lock to be
534 released.
535 If a signal is caught while waiting, then the call is interrupted
536 and (after the signal handler has returned) returns immediately
537 (with return value \-1 and
538 .I errno
539 set to
540 .BR EINTR ;
541 see
542 .BR signal (7)).
543 .TP
544 .BR F_OFD_GETLK " (\fIstruct flock *\fP)"
545 On input to this call,
546 .I lock
547 describes an open file description lock we would like to place on the file.
548 If the lock could be placed,
549 .BR fcntl ()
550 does not actually place it, but returns
551 .B F_UNLCK
552 in the
553 .I l_type
554 field of
555 .I lock
556 and leaves the other fields of the structure unchanged.
557 If one or more incompatible locks would prevent this lock being placed,
558 then details about one of these locks are returned via
559 .IR lock ,
560 as described above for
561 .BR F_GETLK .
562 .SS Mandatory locking
563 .IR Warning :
564 the Linux implementation of mandatory locking is unreliable.
565 See BUGS below.
566
567 By default, both traditional (process-associated) and open file description
568 record locks are advisory.
569 Advisory locks are not enforced and are useful only between
570 cooperating processes.
571
572 Both lock types can also be mandatory.
573 Mandatory locks are enforced for all processes.
574 If a process tries to perform an incompatible access (e.g.,
575 .BR read (2)
576 or
577 .BR write (2))
578 on a file region that has an incompatible mandatory lock,
579 then the result depends upon whether the
580 .B O_NONBLOCK
581 flag is enabled for its open file description.
582 If the
583 .B O_NONBLOCK
584 flag is not enabled, then
585 the system call is blocked until the lock is removed
586 or converted to a mode that is compatible with the access.
587 If the
588 .B O_NONBLOCK
589 flag is enabled, then the system call fails with the error
590 .BR EAGAIN .
591
592 To make use of mandatory locks, mandatory locking must be enabled
593 both on the filesystem that contains the file to be locked,
594 and on the file itself.
595 Mandatory locking is enabled on a filesystem
596 using the "\-o mand" option to
597 .BR mount (8),
598 or the
599 .B MS_MANDLOCK
600 flag for
601 .BR mount (2).
602 Mandatory locking is enabled on a file by disabling
603 group execute permission on the file and enabling the set-group-ID
604 permission bit (see
605 .BR chmod (1)
606 and
607 .BR chmod (2)).
608
609 Mandatory locking is not specified by POSIX.
610 Some other systems also support mandatory locking,
611 although the details of how to enable it vary across systems.
612 .SS Managing signals
613 .BR F_GETOWN ,
614 .BR F_SETOWN ,
615 .BR F_GETOWN_EX ,
616 .BR F_SETOWN_EX ,
617 .BR F_GETSIG
618 and
619 .B F_SETSIG
620 are used to manage I/O availability signals:
621 .TP
622 .BR F_GETOWN " (\fIvoid\fP)"
623 Return (as the function result)
624 the process ID or process group currently receiving
625 .B SIGIO
626 and
627 .B SIGURG
628 signals for events on file descriptor
629 .IR fd .
630 Process IDs are returned as positive values;
631 process group IDs are returned as negative values (but see BUGS below).
632 .I arg
633 is ignored.
634 .TP
635 .BR F_SETOWN " (\fIint\fP)"
636 Set the process ID or process group ID that will receive
637 .B SIGIO
638 and
639 .B SIGURG
640 signals for events on file descriptor
641 .IR fd
642 to the ID given in
643 .IR arg .
644 A process ID is specified as a positive value;
645 a process group ID is specified as a negative value.
646 Most commonly, the calling process specifies itself as the owner
647 (that is,
648 .I arg
649 is specified as
650 .BR getpid (2)).
651
652 .\" From glibc.info:
653 If you set the
654 .B O_ASYNC
655 status flag on a file descriptor by using the
656 .B F_SETFL
657 command of
658 .BR fcntl (),
659 a
660 .B SIGIO
661 signal is sent whenever input or output becomes possible
662 on that file descriptor.
663 .B F_SETSIG
664 can be used to obtain delivery of a signal other than
665 .BR SIGIO .
666 If this permission check fails, then the signal is
667 silently discarded.
668
669 Sending a signal to the owner process (group) specified by
670 .B F_SETOWN
671 is subject to the same permissions checks as are described for
672 .BR kill (2),
673 where the sending process is the one that employs
674 .B F_SETOWN
675 (but see BUGS below).
676
677 If the file descriptor
678 .I fd
679 refers to a socket,
680 .B F_SETOWN
681 also selects
682 the recipient of
683 .B SIGURG
684 signals that are delivered when out-of-band
685 data arrives on that socket.
686 .RB ( SIGURG
687 is sent in any situation where
688 .BR select (2)
689 would report the socket as having an "exceptional condition".)
690 .\" The following appears to be rubbish. It doesn't seem to
691 .\" be true according to the kernel source, and I can write
692 .\" a program that gets a terminal-generated SIGIO even though
693 .\" it is not the foreground process group of the terminal.
694 .\" -- MTK, 8 Apr 05
695 .\"
696 .\" If the file descriptor
697 .\" .I fd
698 .\" refers to a terminal device, then SIGIO
699 .\" signals are sent to the foreground process group of the terminal.
700
701 The following was true in 2.6.x kernels up to and including
702 kernel 2.6.11:
703 .RS
704 .IP
705 If a nonzero value is given to
706 .B F_SETSIG
707 in a multithreaded process running with a threading library
708 that supports thread groups (e.g., NPTL),
709 then a positive value given to
710 .B F_SETOWN
711 has a different meaning:
712 .\" The relevant place in the (2.6) kernel source is the
713 .\" 'switch' in fs/fcntl.c::send_sigio_to_task() -- MTK, Apr 2005
714 instead of being a process ID identifying a whole process,
715 it is a thread ID identifying a specific thread within a process.
716 Consequently, it may be necessary to pass
717 .B F_SETOWN
718 the result of
719 .BR gettid (2)
720 instead of
721 .BR getpid (2)
722 to get sensible results when
723 .B F_SETSIG
724 is used.
725 (In current Linux threading implementations,
726 a main thread's thread ID is the same as its process ID.
727 This means that a single-threaded program can equally use
728 .BR gettid (2)
729 or
730 .BR getpid (2)
731 in this scenario.)
732 Note, however, that the statements in this paragraph do not apply
733 to the
734 .B SIGURG
735 signal generated for out-of-band data on a socket:
736 this signal is always sent to either a process or a process group,
737 depending on the value given to
738 .BR F_SETOWN .
739 .\" send_sigurg()/send_sigurg_to_task() bypasses
740 .\" kill_fasync()/send_sigio()/send_sigio_to_task()
741 .\" to directly call send_group_sig_info()
742 .\" -- MTK, Apr 2005 (kernel 2.6.11)
743 .RE
744 .IP
745 The above behavior was accidentally dropped in Linux 2.6.12,
746 and won't be restored.
747 From Linux 2.6.32 onward, use
748 .BR F_SETOWN_EX
749 to target
750 .B SIGIO
751 and
752 .B SIGURG
753 signals at a particular thread.
754 .TP
755 .BR F_GETOWN_EX " (struct f_owner_ex *) (since Linux 2.6.32)"
756 Return the current file descriptor owner settings
757 as defined by a previous
758 .BR F_SETOWN_EX
759 operation.
760 The information is returned in the structure pointed to by
761 .IR arg ,
762 which has the following form:
763 .nf
764 .in +4n
765
766 struct f_owner_ex {
767 int type;
768 pid_t pid;
769 };
770
771 .in
772 .fi
773 The
774 .I type
775 field will have one of the values
776 .BR F_OWNER_TID ,
777 .BR F_OWNER_PID ,
778 or
779 .BR F_OWNER_PGRP .
780 The
781 .I pid
782 field is a positive integer representing a thread ID, process ID,
783 or process group ID.
784 See
785 .B F_SETOWN_EX
786 for more details.
787 .TP
788 .BR F_SETOWN_EX " (struct f_owner_ex *) (since Linux 2.6.32)"
789 This operation performs a similar task to
790 .BR F_SETOWN .
791 It allows the caller to direct I/O availability signals
792 to a specific thread, process, or process group.
793 The caller specifies the target of signals via
794 .IR arg ,
795 which is a pointer to a
796 .IR f_owner_ex
797 structure.
798 The
799 .I type
800 field has one of the following values, which define how
801 .I pid
802 is interpreted:
803 .RS
804 .TP
805 .BR F_OWNER_TID
806 Send the signal to the thread whose thread ID
807 (the value returned by a call to
808 .BR clone (2)
809 or
810 .BR gettid (2))
811 is specified in
812 .IR pid .
813 .TP
814 .BR F_OWNER_PID
815 Send the signal to the process whose ID
816 is specified in
817 .IR pid .
818 .TP
819 .BR F_OWNER_PGRP
820 Send the signal to the process group whose ID
821 is specified in
822 .IR pid .
823 (Note that, unlike with
824 .BR F_SETOWN ,
825 a process group ID is specified as a positive value here.)
826 .RE
827 .TP
828 .BR F_GETSIG " (\fIvoid\fP)"
829 Return (as the function result)
830 the signal sent when input or output becomes possible.
831 A value of zero means
832 .B SIGIO
833 is sent.
834 Any other value (including
835 .BR SIGIO )
836 is the
837 signal sent instead, and in this case additional info is available to
838 the signal handler if installed with
839 .BR SA_SIGINFO .
840 .I arg
841 is ignored.
842 .TP
843 .BR F_SETSIG " (\fIint\fP)"
844 Set the signal sent when input or output becomes possible
845 to the value given in
846 .IR arg .
847 A value of zero means to send the default
848 .B SIGIO
849 signal.
850 Any other value (including
851 .BR SIGIO )
852 is the signal to send instead, and in this case additional info
853 is available to the signal handler if installed with
854 .BR SA_SIGINFO .
855 .\"
856 .\" The following was true only up until 2.6.11:
857 .\"
858 .\" Additionally, passing a nonzero value to
859 .\" .B F_SETSIG
860 .\" changes the signal recipient from a whole process to a specific thread
861 .\" within a process.
862 .\" See the description of
863 .\" .B F_SETOWN
864 .\" for more details.
865
866 By using
867 .B F_SETSIG
868 with a nonzero value, and setting
869 .B SA_SIGINFO
870 for the
871 signal handler (see
872 .BR sigaction (2)),
873 extra information about I/O events is passed to
874 the handler in a
875 .I siginfo_t
876 structure.
877 If the
878 .I si_code
879 field indicates the source is
880 .BR SI_SIGIO ,
881 the
882 .I si_fd
883 field gives the file descriptor associated with the event.
884 Otherwise,
885 there is no indication which file descriptors are pending, and you
886 should use the usual mechanisms
887 .RB ( select (2),
888 .BR poll (2),
889 .BR read (2)
890 with
891 .B O_NONBLOCK
892 set etc.) to determine which file descriptors are available for I/O.
893
894 By selecting a real time signal (value >=
895 .BR SIGRTMIN ),
896 multiple I/O events may be queued using the same signal numbers.
897 (Queuing is dependent on available memory).
898 Extra information is available
899 if
900 .B SA_SIGINFO
901 is set for the signal handler, as above.
902
903 Note that Linux imposes a limit on the
904 number of real-time signals that may be queued to a
905 process (see
906 .BR getrlimit (2)
907 and
908 .BR signal (7))
909 and if this limit is reached, then the kernel reverts to
910 delivering
911 .BR SIGIO ,
912 and this signal is delivered to the entire
913 process rather than to a specific thread.
914 .\" See fs/fcntl.c::send_sigio_to_task() (2.4/2.6) sources -- MTK, Apr 05
915 .PP
916 Using these mechanisms, a program can implement fully asynchronous I/O
917 without using
918 .BR select (2)
919 or
920 .BR poll (2)
921 most of the time.
922 .PP
923 The use of
924 .BR O_ASYNC
925 is specific to BSD and Linux.
926 The only use of
927 .BR F_GETOWN
928 and
929 .B F_SETOWN
930 specified in POSIX.1 is in conjunction with the use of the
931 .B SIGURG
932 signal on sockets.
933 (POSIX does not specify the
934 .BR SIGIO
935 signal.)
936 .BR F_GETOWN_EX ,
937 .BR F_SETOWN_EX ,
938 .BR F_GETSIG ,
939 and
940 .B F_SETSIG
941 are Linux-specific.
942 POSIX has asynchronous I/O and the
943 .I aio_sigevent
944 structure to achieve similar things; these are also available
945 in Linux as part of the GNU C Library (Glibc).
946 .SS Leases
947 .B F_SETLEASE
948 and
949 .B F_GETLEASE
950 (Linux 2.4 onward) are used (respectively) to establish a new lease,
951 and retrieve the current lease, on the open file description
952 referred to by the file descriptor
953 .IR fd .
954 A file lease provides a mechanism whereby the process holding
955 the lease (the "lease holder") is notified (via delivery of a signal)
956 when a process (the "lease breaker") tries to
957 .BR open (2)
958 or
959 .BR truncate (2)
960 the file referred to by that file descriptor.
961 .TP
962 .BR F_SETLEASE " (\fIint\fP)"
963 Set or remove a file lease according to which of the following
964 values is specified in the integer
965 .IR arg :
966 .RS
967 .TP
968 .B F_RDLCK
969 Take out a read lease.
970 This will cause the calling process to be notified when
971 the file is opened for writing or is truncated.
972 .\" The following became true in kernel 2.6.10:
973 .\" See the man-pages-2.09 Changelog for further info.
974 A read lease can be placed only on a file descriptor that
975 is opened read-only.
976 .TP
977 .B F_WRLCK
978 Take out a write lease.
979 This will cause the caller to be notified when
980 the file is opened for reading or writing or is truncated.
981 A write lease may be placed on a file only if there are no
982 other open file descriptors for the file.
983 .TP
984 .B F_UNLCK
985 Remove our lease from the file.
986 .RE
987 .P
988 Leases are associated with an open file description (see
989 .BR open (2)).
990 This means that duplicate file descriptors (created by, for example,
991 .BR fork (2)
992 or
993 .BR dup (2))
994 refer to the same lease, and this lease may be modified
995 or released using any of these descriptors.
996 Furthermore, the lease is released by either an explicit
997 .B F_UNLCK
998 operation on any of these duplicate descriptors, or when all
999 such descriptors have been closed.
1000 .P
1001 Leases may be taken out only on regular files.
1002 An unprivileged process may take out a lease only on a file whose
1003 UID (owner) matches the filesystem UID of the process.
1004 A process with the
1005 .B CAP_LEASE
1006 capability may take out leases on arbitrary files.
1007 .TP
1008 .BR F_GETLEASE " (\fIvoid\fP)"
1009 Indicates what type of lease is associated with the file descriptor
1010 .I fd
1011 by returning either
1012 .BR F_RDLCK ", " F_WRLCK ", or " F_UNLCK ,
1013 indicating, respectively, a read lease , a write lease, or no lease.
1014 .I arg
1015 is ignored.
1016 .PP
1017 When a process (the "lease breaker") performs an
1018 .BR open (2)
1019 or
1020 .BR truncate (2)
1021 that conflicts with a lease established via
1022 .BR F_SETLEASE ,
1023 the system call is blocked by the kernel and
1024 the kernel notifies the lease holder by sending it a signal
1025 .RB ( SIGIO
1026 by default).
1027 The lease holder should respond to receipt of this signal by doing
1028 whatever cleanup is required in preparation for the file to be
1029 accessed by another process (e.g., flushing cached buffers) and
1030 then either remove or downgrade its lease.
1031 A lease is removed by performing an
1032 .B F_SETLEASE
1033 command specifying
1034 .I arg
1035 as
1036 .BR F_UNLCK .
1037 If the lease holder currently holds a write lease on the file,
1038 and the lease breaker is opening the file for reading,
1039 then it is sufficient for the lease holder to downgrade
1040 the lease to a read lease.
1041 This is done by performing an
1042 .B F_SETLEASE
1043 command specifying
1044 .I arg
1045 as
1046 .BR F_RDLCK .
1047
1048 If the lease holder fails to downgrade or remove the lease within
1049 the number of seconds specified in
1050 .IR /proc/sys/fs/lease-break-time ,
1051 then the kernel forcibly removes or downgrades the lease holder's lease.
1052
1053 Once a lease break has been initiated,
1054 .B F_GETLEASE
1055 returns the target lease type (either
1056 .B F_RDLCK
1057 or
1058 .BR F_UNLCK ,
1059 depending on what would be compatible with the lease breaker)
1060 until the lease holder voluntarily downgrades or removes the lease or
1061 the kernel forcibly does so after the lease break timer expires.
1062
1063 Once the lease has been voluntarily or forcibly removed or downgraded,
1064 and assuming the lease breaker has not unblocked its system call,
1065 the kernel permits the lease breaker's system call to proceed.
1066
1067 If the lease breaker's blocked
1068 .BR open (2)
1069 or
1070 .BR truncate (2)
1071 is interrupted by a signal handler,
1072 then the system call fails with the error
1073 .BR EINTR ,
1074 but the other steps still occur as described above.
1075 If the lease breaker is killed by a signal while blocked in
1076 .BR open (2)
1077 or
1078 .BR truncate (2),
1079 then the other steps still occur as described above.
1080 If the lease breaker specifies the
1081 .B O_NONBLOCK
1082 flag when calling
1083 .BR open (2),
1084 then the call immediately fails with the error
1085 .BR EWOULDBLOCK ,
1086 but the other steps still occur as described above.
1087
1088 The default signal used to notify the lease holder is
1089 .BR SIGIO ,
1090 but this can be changed using the
1091 .B F_SETSIG
1092 command to
1093 .BR fcntl ().
1094 If a
1095 .B F_SETSIG
1096 command is performed (even one specifying
1097 .BR SIGIO ),
1098 and the signal
1099 handler is established using
1100 .BR SA_SIGINFO ,
1101 then the handler will receive a
1102 .I siginfo_t
1103 structure as its second argument, and the
1104 .I si_fd
1105 field of this argument will hold the descriptor of the leased file
1106 that has been accessed by another process.
1107 (This is useful if the caller holds leases against multiple files).
1108 .SS File and directory change notification (dnotify)
1109 .TP
1110 .BR F_NOTIFY " (\fIint\fP)"
1111 (Linux 2.4 onward)
1112 Provide notification when the directory referred to by
1113 .I fd
1114 or any of the files that it contains is changed.
1115 The events to be notified are specified in
1116 .IR arg ,
1117 which is a bit mask specified by ORing together zero or more of
1118 the following bits:
1119 .RS
1120 .sp
1121 .PD 0
1122 .TP 12
1123 .B DN_ACCESS
1124 A file was accessed (read, pread, readv)
1125 .TP
1126 .B DN_MODIFY
1127 A file was modified (write, pwrite, writev, truncate, ftruncate).
1128 .TP
1129 .B DN_CREATE
1130 A file was created (open, creat, mknod, mkdir, link, symlink, rename).
1131 .TP
1132 .B DN_DELETE
1133 A file was unlinked (unlink, rename to another directory, rmdir).
1134 .TP
1135 .B DN_RENAME
1136 A file was renamed within this directory (rename).
1137 .TP
1138 .B DN_ATTRIB
1139 The attributes of a file were changed (chown, chmod, utime[s]).
1140 .PD
1141 .RE
1142 .IP
1143 (In order to obtain these definitions, the
1144 .B _GNU_SOURCE
1145 feature test macro must be defined before including
1146 .I any
1147 header files.)
1148
1149 Directory notifications are normally "one-shot", and the application
1150 must reregister to receive further notifications.
1151 Alternatively, if
1152 .B DN_MULTISHOT
1153 is included in
1154 .IR arg ,
1155 then notification will remain in effect until explicitly removed.
1156
1157 .\" The following does seem a poor API-design choice...
1158 A series of
1159 .B F_NOTIFY
1160 requests is cumulative, with the events in
1161 .I arg
1162 being added to the set already monitored.
1163 To disable notification of all events, make an
1164 .B F_NOTIFY
1165 call specifying
1166 .I arg
1167 as 0.
1168
1169 Notification occurs via delivery of a signal.
1170 The default signal is
1171 .BR SIGIO ,
1172 but this can be changed using the
1173 .B F_SETSIG
1174 command to
1175 .BR fcntl ().
1176 In the latter case, the signal handler receives a
1177 .I siginfo_t
1178 structure as its second argument (if the handler was
1179 established using
1180 .BR SA_SIGINFO )
1181 and the
1182 .I si_fd
1183 field of this structure contains the file descriptor which
1184 generated the notification (useful when establishing notification
1185 on multiple directories).
1186
1187 Especially when using
1188 .BR DN_MULTISHOT ,
1189 a real time signal should be used for notification,
1190 so that multiple notifications can be queued.
1191
1192 .B NOTE:
1193 New applications should use the
1194 .I inotify
1195 interface (available since kernel 2.6.13),
1196 which provides a much superior interface for obtaining notifications of
1197 filesystem events.
1198 See
1199 .BR inotify (7).
1200 .SS Changing the capacity of a pipe
1201 .TP
1202 .BR F_SETPIPE_SZ " (\fIint\fP; since Linux 2.6.35)"
1203 Change the capacity of the pipe referred to by
1204 .I fd
1205 to be at least
1206 .I arg
1207 bytes.
1208 An unprivileged process can adjust the pipe capacity to any value
1209 between the system page size and the limit defined in
1210 .IR /proc/sys/fs/pipe-max-size
1211 (see
1212 .BR proc (5)).
1213 Attempts to set the pipe capacity below the page size are silently
1214 rounded up to the page size.
1215 Attempts by an unprivileged process to set the pipe capacity above the limit in
1216 .IR /proc/sys/fs/pipe-max-size
1217 yield the error
1218 .BR EPERM ;
1219 a privileged process
1220 .RB ( CAP_SYS_RESOURCE )
1221 can override the limit.
1222 When allocating the buffer for the pipe,
1223 the kernel may use a capacity larger than
1224 .IR arg ,
1225 if that is convenient for the implementation.
1226 The
1227 .B F_GETPIPE_SZ
1228 operation returns the actual size used.
1229 Attempting to set the pipe capacity smaller than the amount
1230 of buffer space currently used to store data produces the error
1231 .BR EBUSY .
1232 .TP
1233 .BR F_GETPIPE_SZ " (\fIvoid\fP; since Linux 2.6.35)"
1234 Return (as the function result) the capacity of the pipe referred to by
1235 .IR fd .
1236 .SH RETURN VALUE
1237 For a successful call, the return value depends on the operation:
1238 .TP 0.9i
1239 .B F_DUPFD
1240 The new descriptor.
1241 .TP
1242 .B F_GETFD
1243 Value of file descriptor flags.
1244 .TP
1245 .B F_GETFL
1246 Value of file status flags.
1247 .TP
1248 .B F_GETLEASE
1249 Type of lease held on file descriptor.
1250 .TP
1251 .B F_GETOWN
1252 Value of descriptor owner.
1253 .TP
1254 .B F_GETSIG
1255 Value of signal sent when read or write becomes possible, or zero
1256 for traditional
1257 .B SIGIO
1258 behavior.
1259 .TP
1260 .B F_GETPIPE_SZ
1261 The pipe capacity.
1262 .TP
1263 All other commands
1264 Zero.
1265 .PP
1266 On error, \-1 is returned, and
1267 .I errno
1268 is set appropriately.
1269 .SH ERRORS
1270 .TP
1271 .BR EACCES " or " EAGAIN
1272 Operation is prohibited by locks held by other processes.
1273 .TP
1274 .B EAGAIN
1275 The operation is prohibited because the file has been memory-mapped by
1276 another process.
1277 .TP
1278 .B EBADF
1279 .I fd
1280 is not an open file descriptor, or the command was
1281 .B F_SETLK
1282 or
1283 .B F_SETLKW
1284 and the file descriptor open mode doesn't match with the
1285 type of lock requested.
1286 .TP
1287 .B EDEADLK
1288 It was detected that the specified
1289 .B F_SETLKW
1290 command would cause a deadlock.
1291 .TP
1292 .B EFAULT
1293 .I lock
1294 is outside your accessible address space.
1295 .TP
1296 .B EINTR
1297 For
1298 .BR F_SETLKW ,
1299 the command was interrupted by a signal; see
1300 .BR signal (7).
1301 For
1302 .BR F_GETLK " and " F_SETLK ,
1303 the command was interrupted by a signal before the lock was checked or
1304 acquired.
1305 Most likely when locking a remote file (e.g., locking over
1306 NFS), but can sometimes happen locally.
1307 .TP
1308 .B EINVAL
1309 The value specified in
1310 .I cmd
1311 is not recognized by this kernel.
1312 .TP
1313 .B EINVAL
1314 For
1315 .BR F_DUPFD ,
1316 .I arg
1317 is negative or is greater than the maximum allowable value.
1318 For
1319 .BR F_SETSIG ,
1320 .I arg
1321 is not an allowable signal number.
1322 .TP
1323 .B EINVAL
1324 .I cmd
1325 is
1326 .BR F_OFD_SETLK ,
1327 .BR F_OFD_SETLKW ,
1328 or
1329 .BR F_OFD_GETLK ,
1330 and
1331 .I l_pid
1332 was not specified as zero.
1333 .TP
1334 .B EMFILE
1335 For
1336 .BR F_DUPFD ,
1337 the process already has the maximum number of file descriptors open.
1338 .TP
1339 .B ENOLCK
1340 Too many segment locks open, lock table is full, or a remote locking
1341 protocol failed (e.g., locking over NFS).
1342 .TP
1343 .B EPERM
1344 Attempted to clear the
1345 .B O_APPEND
1346 flag on a file that has the append-only attribute set.
1347 .SH CONFORMING TO
1348 SVr4, 4.3BSD, POSIX.1-2001.
1349 Only the operations
1350 .BR F_DUPFD ,
1351 .BR F_GETFD ,
1352 .BR F_SETFD ,
1353 .BR F_GETFL ,
1354 .BR F_SETFL ,
1355 .BR F_GETLK ,
1356 .BR F_SETLK ,
1357 and
1358 .BR F_SETLKW
1359 are specified in POSIX.1-2001.
1360
1361 .BR F_GETOWN
1362 and
1363 .B F_SETOWN
1364 are specified in POSIX.1-2001.
1365 (To get their definitions, define
1366 .BR _BSD_SOURCE ,
1367 or
1368 .BR _XOPEN_SOURCE
1369 with the value 500 or greater, or define
1370 .BR _POSIX_C_SOURCE
1371 with the value 200809L or greater.)
1372
1373 .B F_DUPFD_CLOEXEC
1374 is specified in POSIX.1-2008.
1375 (To get this definition, define
1376 .B _POSIX_C_SOURCE
1377 with the value 200809L or greater, or
1378 .B _XOPEN_SOURCE
1379 with the value 700 or greater.)
1380
1381 .BR F_GETOWN_EX ,
1382 .BR F_SETOWN_EX ,
1383 .BR F_SETPIPE_SZ ,
1384 .BR F_GETPIPE_SZ ,
1385 .BR F_GETSIG ,
1386 .BR F_SETSIG ,
1387 .BR F_NOTIFY ,
1388 .BR F_GETLEASE ,
1389 and
1390 .B F_SETLEASE
1391 are Linux-specific.
1392 (Define the
1393 .B _GNU_SOURCE
1394 macro to obtain these definitions.)
1395 .\" .PP
1396 .\" SVr4 documents additional EIO, ENOLINK and EOVERFLOW error conditions.
1397
1398 .BR F_OFD_SETLK ,
1399 .BR F_OFD_SETLKW ,
1400 and
1401 .BR F_OFD_GETLK
1402 are Linux-specific,
1403 but work is being done to have them included in the next version of POSIX.1.
1404 .\" FIXME Presumably, glibc will require _GNU_SOURCE
1405 .SH NOTES
1406 The errors returned by
1407 .BR dup2 (2)
1408 are different from those returned by
1409 .BR F_DUPFD .
1410 .\"
1411 .SS File locking
1412 The original Linux
1413 .BR fcntl ()
1414 system call was not designed to handle large file offsets
1415 (in the
1416 .I flock
1417 structure).
1418 Consequently, an
1419 .BR fcntl64 ()
1420 system call was added in Linux 2.4.
1421 The newer system call employs a different structure for file locking,
1422 .IR flock64 ,
1423 and corresponding commands,
1424 .BR F_GETLK64 ,
1425 .BR F_SETLK64 ,
1426 and
1427 .BR F_SETLKW64 .
1428 However, these details can be ignored by applications using glibc, whose
1429 .BR fcntl ()
1430 wrapper function transparently employs the more recent system call
1431 where it is available.
1432
1433 The errors returned by
1434 .BR dup2 (2)
1435 are different from those returned by
1436 .BR F_DUPFD .
1437 .SS Record locks
1438 Since kernel 2.0, there is no interaction between the types of lock
1439 placed by
1440 .BR flock (2)
1441 and
1442 .BR fcntl ().
1443
1444 Several systems have more fields in
1445 .I "struct flock"
1446 such as, for example,
1447 .IR l_sysid .
1448 .\" e.g., Solaris 8 documents this field in fcntl(2), and Irix 6.5
1449 .\" documents it in fcntl(5). mtk, May 2007
1450 .\" Also, FreeBSD documents it (Apr 2014).
1451 Clearly,
1452 .I l_pid
1453 alone is not going to be very useful if the process holding the lock
1454 may live on a different machine.
1455
1456 The original Linux
1457 .BR fcntl ()
1458 system call was not designed to handle large file offsets
1459 (in the
1460 .I flock
1461 structure).
1462 Consequently, an
1463 .BR fcntl64 ()
1464 system call was added in Linux 2.4.
1465 The newer system call employs a different structure for file locking,
1466 .IR flock64 ,
1467 and corresponding commands,
1468 .BR F_GETLK64 ,
1469 .BR F_SETLK64 ,
1470 and
1471 .BR F_SETLKW64 .
1472 However, these details can be ignored by applications using glibc, whose
1473 .BR fcntl ()
1474 wrapper function transparently employs the more recent system call
1475 where it is available.
1476 .SS Record locking and NFS
1477 Before Linux 3.12, if an NFSv4 client
1478 loses contact with the server for a period of time
1479 (defined as more than 90 seconds with no communication),
1480 .\"
1481 .\" Neil Brown: With NFSv3 the failure mode is the reverse. If
1482 .\" the server loses contact with a client then any lock stays in place
1483 .\" indefinitely ("why can't I read my mail"... I remember it well).
1484 .\"
1485 it might lose and regain a lock without ever being aware of the fact.
1486 (The period of time after which contact is assumed lost is known as
1487 the NFSv4 leasetime.
1488 On a Linux NFS server, this can be determined by looking at
1489 .IR /proc/fs/nfsd/nfsv4leasetime ,
1490 which expresses the period in seconds.
1491 The default value for this file is 90.)
1492 .\"
1493 .\" Jeff Layton:
1494 .\" Note that this is not a firm timeout. The server runs a job
1495 .\" periodically to clean out expired stateful objects, and it's likely
1496 .\" that there is some time (maybe even up to another whole lease period)
1497 .\" between when the timeout expires and the job actually runs. If the
1498 .\" client gets a RENEW in there within that window, its lease will be
1499 .\" renewed and its state preserved.
1500 .\"
1501 This scenario potentially risks data corruption,
1502 since another process might acquire a lock in the intervening period
1503 and perform file I/O.
1504
1505 Since Linux 3.12,
1506 .\" commit ef1820f9be27b6ad158f433ab38002ab8131db4d
1507 if an NFSv4 client loses contact with the server,
1508 any I/O to the file by a process which "thinks" it holds
1509 a lock will fail until that process closes and reopens the file.
1510 A kernel parameter,
1511 .IR nfs.recover_lost_locks ,
1512 can be set to 1 to obtain the pre-3.12 behavior,
1513 whereby the client will attempt to recover lost locks
1514 when contact is reestablished with the server.
1515 Because of the attendant risk of data corruption,
1516 .\" commit f6de7a39c181dfb8a2c534661a53c73afb3081cd
1517 this parameter defaults to 0 (disabled).
1518 .SH BUGS
1519 .SS F_SETFL
1520 It is not possible to use
1521 .BR F_SETFL
1522 to change the state of the
1523 .BR O_DSYNC
1524 and
1525 .BR O_SYNC
1526 flags.
1527 .\" FIXME . According to POSIX.1-2001, O_SYNC should also be modifiable
1528 .\" via fcntl(2), but currently Linux does not permit this
1529 .\" See http://bugzilla.kernel.org/show_bug.cgi?id=5994
1530 Attempts to change the state of these flags are silently ignored.
1531 .SS F_GETOWN
1532 A limitation of the Linux system call conventions on some
1533 architectures (notably i386) means that if a (negative)
1534 process group ID to be returned by
1535 .B F_GETOWN
1536 falls in the range \-1 to \-4095, then the return value is wrongly
1537 interpreted by glibc as an error in the system call;
1538 .\" glibc source: sysdeps/unix/sysv/linux/i386/sysdep.h
1539 that is, the return value of
1540 .BR fcntl ()
1541 will be \-1, and
1542 .I errno
1543 will contain the (positive) process group ID.
1544 The Linux-specific
1545 .BR F_GETOWN_EX
1546 operation avoids this problem.
1547 .\" mtk, Dec 04: some limited testing on alpha and ia64 seems to
1548 .\" indicate that ANY negative PGID value will cause F_GETOWN
1549 .\" to misinterpret the return as an error. Some other architectures
1550 .\" seem to have the same range check as i386.
1551 Since glibc version 2.11, glibc makes the kernel
1552 .B F_GETOWN
1553 problem invisible by implementing
1554 .B F_GETOWN
1555 using
1556 .BR F_GETOWN_EX .
1557 .SS F_SETOWN
1558 In Linux 2.4 and earlier, there is bug that can occur
1559 when an unprivileged process uses
1560 .B F_SETOWN
1561 to specify the owner
1562 of a socket file descriptor
1563 as a process (group) other than the caller.
1564 In this case,
1565 .BR fcntl ()
1566 can return \-1 with
1567 .I errno
1568 set to
1569 .BR EPERM ,
1570 even when the owner process (group) is one that the caller
1571 has permission to send signals to.
1572 Despite this error return, the file descriptor owner is set,
1573 and signals will be sent to the owner.
1574 .SS Mandatory locking
1575 The Linux implementation of mandatory locking
1576 is subject to race conditions which render it unreliable:
1577 .\" http://marc.info/?l=linux-kernel&m=119013491707153&w=2
1578 .\"
1579 .\" Reconfirmed by Jeff Layton
1580 .\" From: Jeff Layton <jlayton <at> redhat.com>
1581 .\" Subject: Re: Status of fcntl() mandatory locking
1582 .\" Newsgroups: gmane.linux.file-systems
1583 .\" Date: 2014-04-28 10:07:57 GMT
1584 .\" http://thread.gmane.org/gmane.linux.file-systems/84481/focus=84518
1585 .BR write (2)
1586 call that overlaps with a lock may modify data after the mandatory lock is
1587 acquired;
1588 a
1589 .BR read (2)
1590 call that overlaps with a lock may detect changes to data that were made
1591 only after a write lock was acquired.
1592 Similar races exist between mandatory locks and
1593 .BR mmap (2).
1594 It is therefore inadvisable to rely on mandatory locking.
1595 .SH SEE ALSO
1596 .BR dup2 (2),
1597 .BR flock (2),
1598 .BR open (2),
1599 .BR socket (2),
1600 .BR lockf (3),
1601 .BR capabilities (7),
1602 .BR feature_test_macros (7)
1603
1604 .IR locks.txt ,
1605 .IR mandatory-locking.txt ,
1606 and
1607 .I dnotify.txt
1608 in the Linux kernel source directory
1609 .IR Documentation/filesystems/
1610 (on older kernels, these files are directly under the
1611 .I Documentation/
1612 directory, and
1613 .I mandatory-locking.txt
1614 is called
1615 .IR mandatory.txt )