]> git.ipfire.org Git - thirdparty/man-pages.git/blame - man2/clock_getres.2
gettimeofday.2: gettimeofday() is affected by discontinuous jumps in the system time
[thirdparty/man-pages.git] / man2 / clock_getres.2
CommitLineData
fea681da
MK
1.\" Hey Emacs! This file is -*- nroff -*- source.
2.\"
3.\" Copyright (c) 2003 Nick Clifford (zaf@nrc.co.nz), Jan 25, 2003
4.\" Copyright (c) 2003 Andries Brouwer (aeb@cwi.nl), Aug 24, 2003
5.\"
6.\" Permission is granted to make and distribute verbatim copies of this
7.\" manual provided the copyright notice and this permission notice are
8.\" preserved on all copies.
9.\"
10.\" Permission is granted to copy and distribute modified versions of this
11.\" manual under the conditions for verbatim copying, provided that the
12.\" entire resulting derived work is distributed under the terms of a
13.\" permission notice identical to this one.
c13182ef 14.\"
fea681da
MK
15.\" Since the Linux kernel and libraries are constantly changing, this
16.\" manual page may be incorrect or out-of-date. The author(s) assume no
17.\" responsibility for errors or omissions, or for damages resulting from
18.\" the use of the information contained herein. The author(s) may not
19.\" have taken the same level of care in the production of this manual,
20.\" which is licensed free of charge, as they might when working
21.\" professionally.
c13182ef 22.\"
fea681da
MK
23.\" Formatted or processed versions of this manual, if unaccompanied by
24.\" the source, must acknowledge the copyright and authors of this work.
25.\"
26.\" 2003-08-23 Martin Schulze <joey@infodrom.org> improvements
27.\" 2003-08-24 aeb, large parts rewritten
28.\" 2004-08-06 Christoph Lameter <clameter@sgi.com>, SMP note
29.\"
0245dc09
MK
30.\" FIXME: Linux 2.6.39 adds CLOCK_BOOTTIME
31.\"
a113440c 32.TH CLOCK_GETRES 2 2010-02-03 "" "Linux Programmer's Manual"
fea681da
MK
33.SH NAME
34clock_getres, clock_gettime, clock_settime \- clock and time functions
35.SH SYNOPSIS
36.B #include <time.h>
37.sp
38.BI "int clock_getres(clockid_t " clk_id ", struct timespec *" res );
5895e7eb 39
fea681da 40.BI "int clock_gettime(clockid_t " clk_id ", struct timespec *" tp );
5895e7eb 41
fea681da 42.BI "int clock_settime(clockid_t " clk_id ", const struct timespec *" tp );
cc4615cc 43.sp
b905d225
MK
44Link with \fI\-lrt\fP.
45.sp
cc4615cc
MK
46.in -4n
47Feature Test Macro Requirements for glibc (see
48.BR feature_test_macros (7)):
49.in
50.sp
51.ad l
52.BR clock_getres (),
53.BR clock_gettime (),
54.BR clock_settime ():
8260c01f 55.RS
cc4615cc 56_POSIX_C_SOURCE\ >=\ 199309L
8260c01f 57.RE
cc4615cc 58.ad b
fea681da
MK
59.SH DESCRIPTION
60The function
61.BR clock_getres ()
62finds the resolution (precision) of the specified clock
63.IR clk_id ,
64and, if
65.I res
0c2ec4f1 66is non-NULL, stores it in the \fIstruct timespec\fP pointed to by
fea681da
MK
67.IR res .
68The resolution of clocks depends on the implementation and cannot be
69configured by a particular process.
70If the time value pointed to by the argument
71.I tp
72of
73.BR clock_settime ()
74is not a multiple of
75.IR res ,
76then it is truncated to a multiple of
77.IR res .
78.PP
79The functions
80.BR clock_gettime ()
81and
82.BR clock_settime ()
83retrieve and set the time of the specified clock
84.IR clk_id .
85.PP
86The
87.I res
88and
89.I tp
90arguments are
f19a0f03 91.I timespec
4e836144 92structures, as specified in
c13182ef 93.IR <time.h> :
fea681da 94.sp
3ad4ddcd 95.in +4n
fea681da
MK
96.nf
97struct timespec {
cc4615cc
MK
98 time_t tv_sec; /* seconds */
99 long tv_nsec; /* nanoseconds */
fea681da
MK
100};
101.fi
3ad4ddcd 102.in
fea681da 103.PP
c13182ef 104The
fea681da
MK
105.I clk_id
106argument is the identifier of the particular clock on which to act.
107A clock may be system-wide and hence visible for all processes, or
108per-process if it measures time only within a single process.
109.LP
6f36deb4 110All implementations support the system-wide real-time clock,
fea681da
MK
111which is identified by
112.BR CLOCK_REALTIME .
113Its time represents seconds and nanoseconds since the Epoch.
114When its time is changed, timers for a relative interval are
115unaffected, but timers for an absolute point in time are affected.
116.LP
c13182ef
MK
117More clocks may be implemented.
118The interpretation of the
fea681da
MK
119corresponding time values and the effect on timers is unspecified.
120.LP
5260fe08 121Sufficiently recent versions of glibc and the Linux kernel
fea681da
MK
122support the following clocks:
123.TP
124.B CLOCK_REALTIME
6f36deb4 125System-wide real-time clock.
fea681da
MK
126Setting this clock requires appropriate privileges.
127.TP
128.B CLOCK_MONOTONIC
129Clock that cannot be set and represents monotonic time since
130some unspecified starting point.
131.TP
1ba3fba8
MK
132.BR CLOCK_MONOTONIC_RAW " (since Linux 2.6.28; Linux-specific)"
133.\" Added in commit 2d42244ae71d6c7b0884b5664cf2eda30fb2ae68, John Stultz
134Similar to
135.BR CLOCK_MONOTONIC ,
136but provides access to a raw hardware-based time
137that is not subject to NTP adjustments.
138.TP
fea681da
MK
139.B CLOCK_PROCESS_CPUTIME_ID
140High-resolution per-process timer from the CPU.
141.TP
142.B CLOCK_THREAD_CPUTIME_ID
143Thread-specific CPU-time clock.
144.SH "RETURN VALUE"
e1d6264d
MK
145.BR clock_gettime (),
146.BR clock_settime ()
fea681da
MK
147and
148.BR clock_getres ()
149return 0 for success, or \-1 for failure (in which case
150.I errno
151is set appropriately).
152.SH ERRORS
153.TP
154.B EFAULT
155.I tp
156points outside the accessible address space.
157.TP
158.B EINVAL
c13182ef 159The
fea681da
MK
160.I clk_id
161specified is not supported on this system.
a113440c
MK
162.\" Linux also gives this error on attempts to set CLOCK_PROCESS_CPUTIME_ID
163.\" and CLOCK_THREAD_CPUTIME_ID, when probably the proper error should be
164.\" EPERM.
fea681da
MK
165.TP
166.B EPERM
167.BR clock_settime ()
168does not have permission to set the clock indicated.
2b2581ee
MK
169.SH "CONFORMING TO"
170SUSv2, POSIX.1-2001.
171.SH AVAILABILITY
172On POSIX systems on which these functions are available, the symbol
173.B _POSIX_TIMERS
c84371c6 174is defined in \fI<unistd.h>\fP to a value greater than 0.
2b2581ee
MK
175The symbols
176.BR _POSIX_MONOTONIC_CLOCK ,
177.BR _POSIX_CPUTIME ,
178.B _POSIX_THREAD_CPUTIME
179indicate that
180.BR CLOCK_MONOTONIC ,
181.BR CLOCK_PROCESS_CPUTIME_ID ,
182.B CLOCK_THREAD_CPUTIME_ID
183are available.
184(See also
185.BR sysconf (3).)
19c98696 186.SH NOTES
19c98696 187.SS Note for SMP systems
fea681da
MK
188The
189.B CLOCK_PROCESS_CPUTIME_ID
190and
191.B CLOCK_THREAD_CPUTIME_ID
192clocks are realized on many platforms using timers from the CPUs
193(TSC on i386, AR.ITC on Itanium).
194These registers may differ between CPUs and as a consequence
195these clocks may return
196.B bogus results
197if a process is migrated to another CPU.
198.PP
199If the CPUs in an SMP system have different clock sources then
200there is no way to maintain a correlation between the timer registers since
c13182ef
MK
201each CPU will run at a slightly different frequency.
202If that is the case then
fea681da
MK
203.I clock_getcpuclockid(0)
204will return
205.B ENOENT
c13182ef
MK
206to signify this condition.
207The two clocks will then only be useful if it
fea681da
MK
208can be ensured that a process stays on a certain CPU.
209.PP
210The processors in an SMP system do not start all at exactly the same
211time and therefore the timer registers are typically running at an offset.
212Some architectures include code that attempts to limit these offsets on bootup.
213However, the code cannot guarantee to accurately tune the offsets.
214Glibc contains no provisions to deal with these offsets (unlike the Linux
c13182ef
MK
215Kernel).
216Typically these offsets are small and therefore the effects may be
fea681da 217negligible in most cases.
3847be78 218.SH BUGS
a113440c 219According to POSIX.1-2001, a process with "appropriate privileges" may set the
3847be78
MK
220.B CLOCK_PROCESS_CPUTIME_ID
221and
222.B CLOCK_THREAD_CPUTIME_ID
a113440c 223clocks using
3847be78 224.BR clock_settime ().
a113440c
MK
225On Linux, these clocks are not settable
226(i.e., no process has "appropriate privileges").
227.\" See http://bugzilla.kernel.org/show_bug.cgi?id=11972
fea681da
MK
228.SH "SEE ALSO"
229.BR date (1),
230.BR adjtimex (2),
231.BR gettimeofday (2),
232.BR settimeofday (2),
233.BR time (2),
edf60acb 234.BR clock_getcpuclockid (3),
fea681da
MK
235.BR ctime (3),
236.BR ftime (3),
24ad017d 237.BR pthread_getcpuclockid (3),
1d7c4d16
MK
238.BR sysconf (3),
239.BR time (7)