1 .\" Copyright (C) 2007 Michael Kerrisk <mtk.manpages@gmail.com>
2 .\" with some input from Stepan Kasal <kasal@ucw.cz>
4 .\" Some content retained from an earlier version of this page:
5 .\" Copyright (C) 1998 Andries Brouwer (aeb@cwi.nl)
6 .\" Modifications for 2.2 and 2.4 Copyright (C) 2002 Ian Redfern
7 .\" <redferni@logica.com>
9 .\" %%%LICENSE_START(verbatim)
10 .\" Permission is granted to make and distribute verbatim copies of this
11 .\" manual provided the copyright notice and this permission notice are
12 .\" preserved on all copies.
14 .\" Permission is granted to copy and distribute modified versions of this
15 .\" manual under the conditions for verbatim copying, provided that the
16 .\" entire resulting derived work is distributed under the terms of a
17 .\" permission notice identical to this one.
19 .\" Since the Linux kernel and libraries are constantly changing, this
20 .\" manual page may be incorrect or out-of-date. The author(s) assume no
21 .\" responsibility for errors or omissions, or for damages resulting from
22 .\" the use of the information contained herein. The author(s) may not
23 .\" have taken the same level of care in the production of this manual,
24 .\" which is licensed free of charge, as they might when working
27 .\" Formatted or processed versions of this manual, if unaccompanied by
28 .\" the source, must acknowledge the copyright and authors of this work.
31 .TH SYSCALLS 2 2013-02-10 "Linux" "Linux Programmer's Manual"
33 syscalls \- Linux system calls
37 The system call is the fundamental interface between an application
39 .SS System calls and library wrapper functions
40 System calls are generally not invoked directly,
41 but rather via wrapper functions in glibc (or perhaps some other library).
42 For details of direct invocation of a system call, see
44 Often, but not always, the name of the wrapper function is the same
45 as the name of the system call that it invokes.
46 For example, glibc contains a function
48 which invokes the underlying "truncate" system call.
50 Often the glibc wrapper function is quite thin, doing little work
51 other than copying arguments to the right registers
52 before invoking the system call,
55 appropriately after the system call has returned.
56 (These are the same steps that are performed by
58 which can be used to invoke system calls
59 for which no wrapper function is provided.)
60 Note: system calls indicate a failure by returning a negative error
63 the wrapper function negates the returned error number
64 (to make it positive), copies it to
66 and returns \-1 to the caller of the wrapper.
68 Sometimes, however, the wrapper function does some extra work
69 before invoking the system call.
70 For example, nowadays there are (for reasons described below) two
77 wrapper function checks which of those system calls
78 are provided by the kernel and determines which should be employed.
80 Below is a list of the Linux system calls.
83 column indicates the kernel version
84 for those system calls that were new in Linux 2.2,
85 or have appeared since that kernel version.
86 Note the following points:
88 Where no kernel version is indicated,
89 the system call appeared in kernel 1.0 or earlier.
91 Where a system call is marked "1.2"
92 this means the system call probably appeared in a 1.1.x kernel version,
93 and first appeared in a stable kernel with 1.2.
94 (Development of the 1.2 kernel was initiated from a branch of kernel
95 1.0.6 via the 1.1.x unstable kernel series.)
97 Where a system call is marked "2.0"
98 this means the system call probably appeared in a 1.3.x kernel version,
99 and first appeared in a stable kernel with 2.0.
100 (Development of the 2.0 kernel was initiated from a branch of kernel
101 1.2.x, somewhere around 1.2.10,
102 via the 1.3.x unstable kernel series.)
103 .\" Was kernel 2.0 started from a branch of 1.2.10?
104 .\" At least from the timestamps of the tarballs of
105 .\" of 1.2.10 and 1.3.0, that's how it looks, but in
106 .\" fact the diff doesn't seem very clear, the
107 .\" 1.3.0 .tar.bz is much bigger (2.0 MB) than the
108 .\" 1.2.10 .tar.bz2 (1.8 MB), and AEB points out the
109 .\" timestamps of some files in 1.3.0 seem to be older
110 .\" than those in 1.2.10. All of this suggests
111 .\" that there might not have been a clean branch point.
113 Where a system call is marked "2.2"
114 this means the system call probably appeared in a 2.1.x kernel version,
115 and first appeared in a stable kernel with 2.2.0.
116 (Development of the 2.2 kernel was initiated from a branch of kernel
117 2.0.21 via the 2.1.x unstable kernel series.)
119 Where a system call is marked "2.4"
120 this means the system call probably appeared in a 2.3.x kernel version,
121 and first appeared in a stable kernel with 2.4.0.
122 (Development of the 2.4 kernel was initiated from a branch of
123 kernel 2.2.8 via the 2.3.x unstable kernel series.)
125 Where a system call is marked "2.6"
126 this means the system call probably appeared in a 2.5.x kernel version,
127 and first appeared in a stable kernel with 2.6.0.
128 (Development of kernel 2.6 was initiated from a branch
129 of kernel 2.4.15 via the 2.5.x unstable kernel series.)
131 Starting with kernel 2.6.0, the development model changed,
132 and new system calls may appear in each 2.6.x release.
133 In this case, the exact version number where the system call appeared
135 This convention continues with the 3.x kernel series,
136 which followed on from kernel 2.6.39.
138 In some cases, a system call was added to a stable kernel
139 series after it branched from the previous stable kernel
140 series, and then backported into the earlier stable kernel series.
141 For example some system calls that appeared in 2.6.x were also backported
142 into a 2.4.x release after 2.4.15.
143 When this is so, the version where the system call appeared
144 in both of the major kernel series is listed.
146 The list of system calls that are available as at kernel 3.5
147 (or in a few cases only on older kernels) is as follows:
149 .\" Looking at scripts/checksyscalls.sh in the kernel source is
150 .\" instructive about x86 specifics.
157 \fBSystem call\fP \fBKernel\fP \fBNotes\fP
163 \fBaccept4\fP(2) 2.6.28
166 \fBadd_key\fP(2) 2.6.11
169 \fBalloc_hugepages\fP(2) 2.5.36 Removed in 2.5.44
171 Deprecated (does nothing) since 2.6
175 \fBcacheflush\fP(2) 1.2 Not on x86
183 \fBclock_adjtime\fP(2) 2.6.39
184 \fBclock_getres\fP(2) 2.6
185 \fBclock_gettime\fP(2) 2.6
186 \fBclock_nanosleep\fP(2) 2.6
187 \fBclock_settime\fP(2) 2.6
192 \fBcreate_module\fP(2) Removed in 2.6
193 \fBdelete_module\fP(2)
197 \fBepoll_create\fP(2) 2.6
198 \fBepoll_create1\fP(2) 2.6.27
199 \fBepoll_ctl\fP(2) 2.6
200 \fBepoll_pwait\fP(2) 2.6.19
201 \fBepoll_wait\fP(2) 2.6
202 \fBeventfd\fP(2) 2.6.22
203 \fBeventfd2\fP(2) 2.6.27
206 \fBexit_group\fP(2) 2.6
207 \fBfaccessat\fP(2) 2.6.16
208 \fBfadvise64\fP(2) 2.6
209 .\" Implements \fBposix_fadvise\fP(2)
210 \fBfadvise64_64\fP(2) 2.6
211 \fBfallocate\fP(2) 2.6.23
212 \fBfanotify_init\fP(2) 2.6.37
213 \fBfanotify_mark\fP(2) 2.6.37
214 .\" The fanotify calls were added in Linux 2.6.36,
215 .\" but disabled while the API was finalized.
218 \fBfchmodat\fP(2) 2.6.16
220 \fBfchown32\fP(2) 2.4
221 \fBfchownat\fP(2) 2.6.16
225 \fBfgetxattr\fP(2) 2.6; 2.4.18
226 \fBfinit_module\fP(2) 3.8
227 \fBflistxattr\fP(2) 2.6; 2.4.18
230 \fBfree_hugepages\fP(2) 2.5.36 Removed in 2.5.44
231 \fBfremovexattr\fP(2) 2.6; 2.4.18
232 \fBfsetxattr\fP(2) 2.6; 2.4.18
235 \fBfstatat64\fP(2) 2.6.16
237 \fBfstatfs64\fP(2) 2.6
240 \fBftruncate64\fP(2) 2.4
242 \fBfutimesat\fP(2) 2.6.16
243 \fBget_kernel_syms\fP(2) Removed in 2.6
244 \fBget_mempolicy\fP(2) 2.6.6
245 \fBget_robust_list\fP(2) 2.6.17
246 \fBget_thread_area\fP(2) 2.6
247 \fBgetcpu\fP(2) 2.6.19
249 \fBgetdents\fP(2) 2.0
250 \fBgetdents64\fP(2) 2.4
252 \fBgetegid32\fP(2) 2.4
254 \fBgeteuid32\fP(2) 2.4
256 \fBgetgid32\fP(2) 2.4
258 \fBgetgroups32\fP(2) 2.4
261 \fBgetpagesize\fP(2) 2.0 Not on x86
267 \fBgetresgid\fP(2) 2.2
268 \fBgetresgid32\fP(2) 2.4
269 \fBgetresuid\fP(2) 2.2
270 \fBgetresuid32\fP(2) 2.4
276 \fBgettid\fP(2) 2.4.11
277 \fBgettimeofday\fP(2)
279 \fBgetuid32\fP(2) 2.4
280 .\" \fBgetunwind\fP(2) 2.4.8 ia64; DEPRECATED
281 \fBgetxattr\fP(2) 2.6; 2.4.18
283 \fBinotify_add_watch\fP(2) 2.6.13
284 \fBinotify_init\fP(2) 2.6.13
285 \fBinotify_init1\fP(2) 2.6.27
286 \fBinotify_rm_watch\fP(2) 2.6.13
287 \fBio_cancel\fP(2) 2.6
288 \fBio_destroy\fP(2) 2.6
289 \fBio_getevents\fP(2) 2.6
290 \fBio_setup\fP(2) 2.6
291 \fBio_submit\fP(2) 2.6
295 \fBioprio_get\fP(2) 2.6.13
296 \fBioprio_set\fP(2) 2.6.13
298 .\" Implements System V IPC calls
300 \fBkern_features\fP(2) 3.7 Sparc64
301 \fBkexec_load\fP(2) 2.6.13
302 .\" The entry in the syscall table was reserved starting in 2.6.7
303 .\" Was named sys_kexec_load() from 2.6.7 to 2.6.16
304 \fBkeyctl\fP(2) 2.6.11
307 \fBlchown32\fP(2) 2.4
308 \fBlgetxattr\fP(2) 2.6; 2.4.18
310 \fBlinkat\fP(2) 2.6.16
312 \fBlistxattr\fP(2) 2.6; 2.4.18
313 \fBllistxattr\fP(2) 2.6; 2.4.18
314 \fBlookup_dcookie\fP(2) 2.6
315 \fBlremovexattr\fP(2) 2.6; 2.4.18
317 \fBlsetxattr\fP(2) 2.6; 2.4.18
321 \fBmadvise1\fP(2) 2.4
323 .\" \fBmemory_ordering\fP(2) ??? Sparc64
324 \fBmigrate_pages\fP(2) 2.6.16
327 \fBmkdirat\fP(2) 2.6.16
329 \fBmknodat\fP(2) 2.6.16
336 \fBmove_pages\fP(2) 2.6.18
338 \fBmq_getsetattr\fP(2) 2.6.6
339 .\" Implements \fBmq_getattr\fP(3) and \fBmq_setattr\fP(3)
340 \fBmq_notify\fP(2) 2.6.6
341 \fBmq_open\fP(2) 2.6.6
342 \fBmq_timedreceive\fP(2) 2.6.6
343 \fBmq_timedsend\fP(2) 2.6.6
351 .\" \fBmultiplexer\fP(2) ?? __NR_multiplexer reserved on
352 .\" PowerPC, but unimplemented?
356 \fBname_to_handle_at\fP(2) 2.6.39
357 \fBnanosleep\fP(2) 2.0
358 \fBnfsservctl\fP(2) 2.2 Removed in 3.1
366 \fBopen_by_handle_at\fP(2) 2.6.39
367 \fBopenat\fP(2) 2.6.16
369 \fBpciconfig_iobase\fP(2) 2.2.15; 2.4 Not on x86
370 .\" Alpha, PowerPC, ARM; not x86
371 \fBpciconfig_read\fP(2) 2.0.26; 2.2 Not on x86
372 .\" , PowerPC, ARM; not x86
373 \fBpciconfig_write\fP(2) 2.0.26; 2.2 Not on x86
374 .\" , PowerPC, ARM; not x86
375 \fBperf_event_open\fP(2) 2.6.31 T{
376 Was called perf_counter_open() in 2.6.31; renamed in 2.6.32
378 \fBpersonality\fP(2) 1.2
379 \fBperfctr\fP(2) 2.2 Sparc; removed in 2.6.34
380 .\" commit c7d5a0050773e98d1094eaa9f2a1a793fafac300 removed perfctr()
381 .\" \fBperfmonctl\fP(2) ??? ia64
383 \fBpipe2\fP(2) 2.6.27
384 \fBpivot_root\fP(2) 2.4
386 \fBppc_rtas\fP(2) PowerPC only
387 \fBppoll\fP(2) 2.6.16
390 Added as "pread" in 2.2; renamed "pread64" in 2.6
392 \fBpreadv\fP(2) 2.6.30
393 \fBprlimit\fP(2) 2.6.36
394 \fBprocess_vm_readv\fP(2) 3.2
395 \fBprocess_vm_writev\fP(2) 3.2
396 \fBpselect6\fP(2) 2.6.16
397 .\" Implements \fBpselect\fP(2)
400 Added as "pwrite" in 2.2; renamed "pwrite64" in 2.6
402 \fBpwritev\fP(2) 2.6.30
403 \fBquery_module\fP(2) 2.2 Removed in 2.6
406 \fBreadahead\fP(2) 2.4.13
408 .\" Supersedes \fBgetdents\fP(2)
410 \fBreadlinkat\fP(2) 2.6.16
416 \fBrecvmmsg\fP(2) 2.6.33
417 \fBremap_file_pages\fP(2) 2.6
418 \fBremovexattr\fP(2) 2.6; 2.4.18
420 \fBrenameat\fP(2) 2.6.16
421 \fBrequest_key\fP(2) 2.6.11
422 \fBrestart_syscall\fP(2) 2.6
424 \fBrt_sigaction\fP(2) 2.2
425 \fBrt_sigpending\fP(2) 2.2
426 \fBrt_sigprocmask\fP(2) 2.2
427 \fBrt_sigqueueinfo\fP(2) 2.2
428 \fBrt_sigreturn\fP(2) 2.2
429 \fBrt_sigsuspend\fP(2) 2.2
430 \fBrt_sigtimedwait\fP(2) 2.2
431 \fBrt_tgsigqueueinfo\fP(2) 2.6.31
432 \fBs390_runtime_instr\fP(2) 3.7 s390 only
433 \fBsched_get_priority_max\fP(2) 2.0
434 \fBsched_get_priority_min\fP(2) 2.0
435 \fBsched_getaffinity\fP(2) 2.6
436 \fBsched_getparam\fP(2) 2.0
437 \fBsched_getscheduler\fP(2) 2.0
438 \fBsched_rr_get_interval\fP(2) 2.0
439 \fBsched_setaffinity\fP(2) 2.6
440 \fBsched_setparam\fP(2) 2.0
441 \fBsched_setscheduler\fP(2) 2.0
442 \fBsched_yield\fP(2) 2.0
447 \fBsemtimedop\fP(2) 2.6; 2.4.22
449 \fBsendfile\fP(2) 2.2
450 \fBsendfile64\fP(2) 2.6; 2.4.19
451 \fBsendmmsg\fP(2) 3.0
454 \fBset_mempolicy\fP(2) 2.6.6
455 \fBset_robust_list\fP(2) 2.6.17
456 \fBset_thread_area\fP(2) 2.6
457 \fBset_tid_address\fP(2) 2.6
458 .\" See http://lkml.org/lkml/2005/8/1/83
459 .\" "[PATCH] remove sys_set_zone_reclaim()"
460 \fBsetdomainname\fP(2)
461 \fBsetfsgid\fP(2) 1.2
462 \fBsetfsgid32\fP(2) 2.4
463 \fBsetfsuid\fP(2) 1.2
464 \fBsetfsuid32\fP(2) 2.4
466 \fBsetgid32\fP(2) 2.4
468 \fBsetgroups32\fP(2) 2.4
475 \fBsetregid32\fP(2) 2.4
476 \fBsetresgid\fP(2) 2.2
477 \fBsetresgid32\fP(2) 2.4
478 \fBsetresuid\fP(2) 2.2
479 \fBsetresuid32\fP(2) 2.4
481 \fBsetreuid32\fP(2) 2.4
485 \fBsettimeofday\fP(2)
487 \fBsetuid32\fP(2) 2.4
488 \fBsetup\fP(2) Removed in 2.2
489 \fBsetxattr\fP(2) 2.6; 2.4.18
497 \fBsigaltstack\fP(2) 2.2
499 \fBsignalfd\fP(2) 2.6.22
500 \fBsignalfd4\fP(2) 2.6.27
507 .\" Implements BSD socket calls
509 \fBsplice\fP(2) 2.6.17
510 \fBspu_create\fP(2) 2.6.16 PowerPC only
511 \fBspu_run\fP(2) 2.6.16 PowerPC only
516 \fBstatfs64\fP(2) 2.6
518 \fBsubpage_prot\fP(2) 2.6.25 PowerPC if
523 \fBsymlinkat\fP(2) 2.6.16
525 \fBsync_file_range\fP(2) 2.6.17
526 \fBsync_file_range2\fP(2) 2.6.22 T{
527 Architecture-specific variant of \fBsync_file_range\fP(2)
529 .\" PowerPC, ARM, tile
530 .\" First appeared on ARM, as arm_sync_file_range(), but later renamed
531 .\" \fBsys_debug_setcontext\fP(2) ??? PowerPC if CONFIG_PPC32
532 \fBsyncfs\fP(2) 2.6.39
536 .\" glibc interface is \fBklogctl\fP(3)
540 \fBtimer_create\fP(2) 2.6
541 \fBtimer_delete\fP(2) 2.6
542 \fBtimer_getoverrun\fP(2) 2.6
543 \fBtimer_gettime\fP(2) 2.6
544 \fBtimer_settime\fP(2) 2.6
545 \fBtimerfd_create\fP(2) 2.6.25
546 \fBtimerfd_gettime\fP(2) 2.6.25
547 \fBtimerfd_settime\fP(2) 2.6.25
549 \fBtkill\fP(2) 2.6; 2.4.22
551 \fBtruncate64\fP(2) 2.4
552 \fBugetrlimit\fP(2) 2.4
555 .\" sys_oldumount() -- __NR_umount
557 .\" sys_umount() -- __NR_umount2
560 \fBunlinkat\fP(2) 2.6.16
561 \fBunshare\fP(2) 2.6.16
565 \fButimensat\fP(2) 2.6.22
567 \fButrap_install\fP(2) 2.2 Sparc
571 .\" Superseded by \fBvm86\fP(2)
572 \fBvmsplice\fP(2) 2.6.17
574 \fBwaitid\fP(2) 2.6.10
581 On many platforms, including x86-32, socket calls are all multiplexed
582 (via glibc wrapper functions) through
584 and similarly System V IPC calls are multiplexed through
587 Although slots are reserved for them in the system call table,
588 the following system calls are not implemented in the standard kernel:
589 .BR afs_syscall (2), \" __NR_afs_syscall is 53 on Linux 2.6.22/i386
590 .BR break (2), \" __NR_break is 17 on Linux 2.6.22/i386
591 .BR ftime (2), \" __NR_ftime is 35 on Linux 2.6.22/i386
592 .BR getpmsg (2), \" __NR_getpmsg is 188 on Linux 2.6.22/i386
593 .BR gtty (2), \" __NR_gtty is 32 on Linux 2.6.22/i386
594 .BR idle (2), \" __NR_idle is 112 on Linux 2.6.22/i386
595 .BR lock (2), \" __NR_lock is 53 on Linux 2.6.22/i386
596 .BR madvise1 (2), \" __NR_madvise1 is 219 on Linux 2.6.22/i386
597 .BR mpx (2), \" __NR_mpx is 66 on Linux 2.6.22/i386
598 .BR phys (2), \" Slot has been reused
599 .BR prof (2), \" __NR_prof is 44 on Linux 2.6.22/i386
600 .BR profil (2), \" __NR_profil is 98 on Linux 2.6.22/i386
601 .BR putpmsg (2), \" __NR_putpmsg is 189 on Linux 2.6.22/i386
602 .\" __NR_security is 223 on Linux 2.4/i386; absent on 2.6/i386, present
603 .\" on a couple of 2.6 architectures
604 .BR security (2), \" __NR_security is 223 on Linux 2.4/i386
605 .\" The security call is for future use.
606 .BR stty (2), \" __NR_stty is 31 on Linux 2.6.22/i386
607 .BR tuxcall (2), \" __NR_tuxcall is 184 on x86_64, also on PPC and alpha
608 .BR ulimit (2), \" __NR_ulimit is 58 on Linux 2.6.22/i386
610 .BR vserver (2) \" __NR_vserver is 273 on Linux 2.6.22/i386
612 .BR unimplemented (2)).
618 exist as library routines.
621 is in use since kernel 2.1.116 for
624 will never be implemented.
629 calls are for kernels patched to support STREAMS,
630 and may never be in the standard kernel.
633 .BR set_zone_reclaim (2),
634 added in Linux 2.6.13, and removed in 2.6.16;
635 this system call was never available to user space.
638 Roughly speaking, the code belonging to the system call
639 with number __NR_xxx defined in
640 .I /usr/include/asm/unistd.h
641 can be found in the Linux kernel source in the routine
643 (The dispatch table for i386 can be found in
644 .IR /usr/src/linux/arch/i386/kernel/entry.S .)
645 There are many exceptions, however, mostly because
646 older system calls were superseded by newer ones,
647 and this has been treated somewhat unsystematically.
649 proprietary operating-system emulation,
650 such as parisc, sparc, sparc64 and alpha,
651 there are many additional system calls; mips64 also contains a full
652 set of 32-bit system calls.
654 Over time, changes to the interfaces of some system calls have been
656 One reason for such changes was the need to increase the size of
657 structures or scalar values passed to the system call.
658 Because of these changes, there are now various groups
659 of related system calls
664 which perform similar tasks, but which vary in
665 details such as the size of their arguments.
666 (As noted earlier, applications are generally unaware of this:
667 the glibc wrapper functions do some work to ensure that the right
668 system call is invoked, and that ABI compatibility is
669 preserved for old binaries.)
670 Examples of systems calls that exist in multiple versions are
673 By now there are three different versions of
685 with the last being the most current.
686 .\" e.g., on 2.6.22/i386: __NR_oldstat 18, __NR_stat 106, __NR_stat64 195
687 .\" The stat system calls deal with three different data structures,
688 .\" defined in include/asm-i386/stat.h: __old_kernel_stat, stat, stat64
689 A similar story applies for
694 Similarly, the defines
695 .IR __NR_oldolduname ,
699 refer to the routines
705 In Linux 2.0, a new version of
707 appeared, with the old and the new kernel routines being named
712 In Linux 2.4, a new version of
714 appeared, with the old and the new kernel routines being named
715 .IR sys_old_getrlimit ()
721 .IR __NR_ugetrlimit ).
723 Linux 2.4 increased the size of user and group IDs from 16 to 32 bits.
724 .\" 64-bit off_t changes: ftruncate64, *stat64,
725 .\" fcntl64 (because of the flock structure), getdents64, *statfs64
726 To support this change, a range of system calls were added
731 .BR setresuid32 (2)),
732 superseding earlier calls of the same name without the
735 Linux 2.4 added support for applications on 32-bit architectures
736 to access large files (i.e., files for which the sizes and
737 file offsets can't be represented in 32 bits.)
738 To support this change, replacements were required for system calls
739 that deal with file offsets and sizes.
740 Thus the following system calls were added:
746 and their analogs that work with file descriptors or
748 These system calls supersede the older system calls
749 which, except in the case of the "stat" calls,
750 have the same name without the "64" suffix.
752 On newer platforms that only have 64-bit file access and 32-bit uids
753 (e.g., alpha, ia64, s390x) there are no *64 or *32 calls.
754 Where the *64 and *32 calls exist, the other versions are obsolete.
758 calls were added in kernel 2.2 to support the addition
759 of real-time signals (see
761 These system calls supersede the older system calls of the same
762 name without the "rt_" prefix.
768 system calls use five or more arguments,
769 which caused problems in the way
770 argument passing on the i386 used to be set up.
771 Thus, while other architectures have
783 (routines that use a pointer to a
784 argument block) instead.
785 These days passing five arguments
786 is not a problem any more, and there is a
789 that corresponds directly to
794 .\" Two system call numbers,
798 .\" have an additional underscore absent in
799 .\" .IR sys_llseek ()
801 .\" .IR sys_sysctl ().
803 .\" In kernel 2.1.81,
807 .\" were swapped; that is,
809 .\" was added with the semantics that were then current for
811 .\" and the semantics of the latter call were changed to what
815 .BR unimplemented (2),