1 @node Date and Time, Non-Local Exits, Arithmetic, Top
4 This chapter describes functions for manipulating dates and times,
5 including functions for determining what the current time is and
6 conversion between different time representations.
8 The time functions fall into three main categories:
12 Functions for measuring elapsed CPU time are discussed in @ref{Processor
16 Functions for measuring absolute clock or calendar time are discussed in
20 Functions for setting alarms and timers are discussed in @ref{Setting
25 * Processor Time:: Measures processor time used by a program.
26 * Calendar Time:: Manipulation of ``real'' dates and times.
27 * Setting an Alarm:: Sending a signal after a specified time.
28 * Sleeping:: Waiting for a period of time.
29 * Resource Usage:: Measuring various resources used.
30 * Limits on Resources:: Specifying limits on resource usage.
31 * Priority:: Reading or setting process run priority.
35 @section Processor Time
37 If you're trying to optimize your program or measure its efficiency, it's
38 very useful to be able to know how much @dfn{processor time} or @dfn{CPU
39 time} it has used at any given point. Processor time is different from
40 actual wall clock time because it doesn't include any time spent waiting
41 for I/O or when some other process is running. Processor time is
42 represented by the data type @code{clock_t}, and is given as a number of
43 @dfn{clock ticks} relative to an arbitrary base time marking the beginning
44 of a single program invocation.
46 @cindex processor time
49 @cindex time, elapsed CPU
52 * Basic CPU Time:: The @code{clock} function.
53 * Detailed CPU Time:: The @code{times} function.
57 @subsection Basic CPU Time Inquiry
59 To get the elapsed CPU time used by a process, you can use the
60 @code{clock} function. This facility is declared in the header file
64 In typical usage, you call the @code{clock} function at the beginning and
65 end of the interval you want to time, subtract the values, and then divide
66 by @code{CLOCKS_PER_SEC} (the number of clock ticks per second), like this:
76 @dots{} /* @r{Do the work.} */
78 elapsed = ((double) (end - start)) / CLOCKS_PER_SEC;
82 Different computers and operating systems vary wildly in how they keep
83 track of processor time. It's common for the internal processor clock
84 to have a resolution somewhere between hundredths and millionths of a
87 In the GNU system, @code{clock_t} is equivalent to @code{long int} and
88 @code{CLOCKS_PER_SEC} is an integer value. But in other systems, both
89 @code{clock_t} and the type of the macro @code{CLOCKS_PER_SEC} can be
90 either integer or floating-point types. Casting processor time values
91 to @code{double}, as in the example above, makes sure that operations
92 such as arithmetic and printing work properly and consistently no matter
93 what the underlying representation is.
97 @deftypevr Macro int CLOCKS_PER_SEC
98 The value of this macro is the number of clock ticks per second measured
99 by the @code{clock} function.
104 @deftypevr Macro int CLK_TCK
105 This is an obsolete name for @code{CLOCKS_PER_SEC}.
110 @deftp {Data Type} clock_t
111 This is the type of the value returned by the @code{clock} function.
112 Values of type @code{clock_t} are in units of clock ticks.
117 @deftypefun clock_t clock (void)
118 This function returns the elapsed processor time. The base time is
119 arbitrary but doesn't change within a single process. If the processor
120 time is not available or cannot be represented, @code{clock} returns the
121 value @code{(clock_t)(-1)}.
125 @node Detailed CPU Time
126 @subsection Detailed Elapsed CPU Time Inquiry
128 The @code{times} function returns more detailed information about
129 elapsed processor time in a @w{@code{struct tms}} object. You should
130 include the header file @file{sys/times.h} to use this facility.
135 @deftp {Data Type} {struct tms}
136 The @code{tms} structure is used to return information about process
137 times. It contains at least the following members:
140 @item clock_t tms_utime
141 This is the CPU time used in executing the instructions of the calling
144 @item clock_t tms_stime
145 This is the CPU time used by the system on behalf of the calling process.
147 @item clock_t tms_cutime
148 This is the sum of the @code{tms_utime} values and the @code{tms_cutime}
149 values of all terminated child processes of the calling process, whose
150 status has been reported to the parent process by @code{wait} or
151 @code{waitpid}; see @ref{Process Completion}. In other words, it
152 represents the total CPU time used in executing the instructions of all
153 the terminated child processes of the calling process, excluding child
154 processes which have not yet been reported by @code{wait} or
157 @item clock_t tms_cstime
158 This is similar to @code{tms_cutime}, but represents the total CPU time
159 used by the system on behalf of all the terminated child processes of the
163 All of the times are given in clock ticks. These are absolute values; in a
164 newly created process, they are all zero. @xref{Creating a Process}.
169 @deftypefun clock_t times (struct tms *@var{buffer})
170 The @code{times} function stores the processor time information for
171 the calling process in @var{buffer}.
173 The return value is the same as the value of @code{clock()}: the elapsed
174 real time relative to an arbitrary base. The base is a constant within a
175 particular process, and typically represents the time since system
176 start-up. A value of @code{(clock_t)(-1)} is returned to indicate failure.
179 @strong{Portability Note:} The @code{clock} function described in
180 @ref{Basic CPU Time}, is specified by the ANSI C standard. The
181 @code{times} function is a feature of POSIX.1. In the GNU system, the
182 value returned by the @code{clock} function is equivalent to the sum of
183 the @code{tms_utime} and @code{tms_stime} fields returned by
187 @section Calendar Time
189 This section describes facilities for keeping track of dates and times
190 according to the Gregorian calendar.
191 @cindex Gregorian calendar
192 @cindex time, calendar
193 @cindex date and time
195 There are three representations for date and time information:
199 @dfn{Calendar time} (the @code{time_t} data type) is a compact
200 representation, typically giving the number of seconds elapsed since
201 some implementation-specific base time.
202 @cindex calendar time
205 There is also a @dfn{high-resolution time} representation (the @code{struct
206 timeval} data type) that includes fractions of a second. Use this time
207 representation instead of ordinary calendar time when you need greater
209 @cindex high-resolution time
212 @dfn{Local time} or @dfn{broken-down time} (the @code{struct
213 tm} data type) represents the date and time as a set of components
214 specifying the year, month, and so on, for a specific time zone.
215 This time representation is usually used in conjunction with formatting
216 date and time values.
218 @cindex broken-down time
222 * Simple Calendar Time:: Facilities for manipulating calendar time.
223 * High-Resolution Calendar:: A time representation with greater precision.
224 * Broken-down Time:: Facilities for manipulating local time.
225 * Formatting Date and Time:: Converting times to strings.
226 * TZ Variable:: How users specify the time zone.
227 * Time Zone Functions:: Functions to examine or specify the time zone.
228 * Time Functions Example:: An example program showing use of some of
232 @node Simple Calendar Time
233 @subsection Simple Calendar Time
235 This section describes the @code{time_t} data type for representing
236 calendar time, and the functions which operate on calendar time objects.
237 These facilities are declared in the header file @file{time.h}.
243 @deftp {Data Type} time_t
244 This is the data type used to represent calendar time. In the GNU C
245 library and other POSIX-compliant implementations, @code{time_t} is
246 equivalent to @code{long int}. When interpreted as an absolute time
247 value, it represents the number of seconds elapsed since 00:00:00 on
248 January 1, 1970, Coordinated Universal Time. (This date is sometimes
249 referred to as the @dfn{epoch}.)
251 In other systems, @code{time_t} might be either an integer or
257 @deftypefun double difftime (time_t @var{time1}, time_t @var{time0})
258 The @code{difftime} function returns the number of seconds elapsed
259 between time @var{time1} and time @var{time0}, as a value of type
262 In the GNU system, you can simply subtract @code{time_t} values. But on
263 other systems, the @code{time_t} data type might use some other encoding
264 where subtraction doesn't work directly.
269 @deftypefun time_t time (time_t *@var{result})
270 The @code{time} function returns the current time as a value of type
271 @code{time_t}. If the argument @var{result} is not a null pointer, the
272 time value is also stored in @code{*@var{result}}. If the calendar
273 time is not available, the value @w{@code{(time_t)(-1)}} is returned.
277 @node High-Resolution Calendar
278 @subsection High-Resolution Calendar
280 The @code{time_t} data type used to represent calendar times has a
281 resolution of only one second. Some applications need more precision.
283 So, the GNU C library also contains functions which are capable of
284 representing calendar times to a higher resolution than one second. The
285 functions and the associated data types described in this section are
286 declared in @file{sys/time.h}.
291 @deftp {Data Type} {struct timeval}
292 The @code{struct timeval} structure represents a calendar time. It
293 has the following members:
296 @item long int tv_sec
297 This represents the number of seconds since the epoch. It is equivalent
298 to a normal @code{time_t} value.
300 @item long int tv_usec
301 This is the fractional second value, represented as the number of
304 Some times struct timeval values are used for time intervals. Then the
305 @code{tv_sec} member is the number of seconds in the interval, and
306 @code{tv_usec} is the number of additional microseconds.
312 @deftp {Data Type} {struct timezone}
313 The @code{struct timezone} structure is used to hold minimal information
314 about the local time zone. It has the following members:
317 @item int tz_minuteswest
318 This is the number of minutes west of GMT.
321 If nonzero, daylight savings time applies during some part of the year.
324 The @code{struct timezone} type is obsolete and should never be used.
325 Instead, use the facilities described in @ref{Time Zone Functions}.
328 It is often necessary to subtract two values of type @w{@code{struct
329 timeval}}. Here is the best way to do this. It works even on some
330 peculiar operating systems where the @code{tv_sec} member has an
334 /* @r{Subtract the `struct timeval' values X and Y,}
335 @r{storing the result in RESULT.}
336 @r{Return 1 if the difference is negative, otherwise 0.} */
339 timeval_subtract (result, x, y)
340 struct timeval *result, *x, *y;
342 /* @r{Perform the carry for the later subtraction by updating @var{y}.} */
343 if (x->tv_usec < y->tv_usec) @{
344 int nsec = (y->tv_usec - x->tv_usec) / 1000000 + 1;
345 y->tv_usec -= 1000000 * nsec;
348 if (x->tv_usec - y->tv_usec > 1000000) @{
349 int nsec = (y->tv_usec - x->tv_usec) / 1000000;
350 y->tv_usec += 1000000 * nsec;
354 /* @r{Compute the time remaining to wait.}
355 @r{@code{tv_usec} is certainly positive.} */
356 result->tv_sec = x->tv_sec - y->tv_sec;
357 result->tv_usec = x->tv_usec - y->tv_usec;
359 /* @r{Return 1 if result is negative.} */
360 return x->tv_sec < y->tv_sec;
366 @deftypefun int gettimeofday (struct timeval *@var{tp}, struct timezone *@var{tzp})
367 The @code{gettimeofday} function returns the current date and time in the
368 @code{struct timeval} structure indicated by @var{tp}. Information about the
369 time zone is returned in the structure pointed at @var{tzp}. If the @var{tzp}
370 argument is a null pointer, time zone information is ignored.
372 The return value is @code{0} on success and @code{-1} on failure. The
373 following @code{errno} error condition is defined for this function:
377 The operating system does not support getting time zone information, and
378 @var{tzp} is not a null pointer. The GNU operating system does not
379 support using @w{@code{struct timezone}} to represent time zone
380 information; that is an obsolete feature of 4.3 BSD.
381 Instead, use the facilities described in @ref{Time Zone Functions}.
387 @deftypefun int settimeofday (const struct timeval *@var{tp}, const struct timezone *@var{tzp})
388 The @code{settimeofday} function sets the current date and time
389 according to the arguments. As for @code{gettimeofday}, time zone
390 information is ignored if @var{tzp} is a null pointer.
392 You must be a privileged user in order to use @code{settimeofday}.
394 The return value is @code{0} on success and @code{-1} on failure. The
395 following @code{errno} error conditions are defined for this function:
399 This process cannot set the time because it is not privileged.
402 The operating system does not support setting time zone information, and
403 @var{tzp} is not a null pointer.
409 @deftypefun int adjtime (const struct timeval *@var{delta}, struct timeval *@var{olddelta})
410 This function speeds up or slows down the system clock in order to make
411 gradual adjustments in the current time. This ensures that the time
412 reported by the system clock is always monotonically increasing, which
413 might not happen if you simply set the current time.
415 The @var{delta} argument specifies a relative adjustment to be made to
416 the current time. If negative, the system clock is slowed down for a
417 while until it has lost this much time. If positive, the system clock
418 is speeded up for a while.
420 If the @var{olddelta} argument is not a null pointer, the @code{adjtime}
421 function returns information about any previous time adjustment that
422 has not yet completed.
424 This function is typically used to synchronize the clocks of computers
425 in a local network. You must be a privileged user to use it.
426 The return value is @code{0} on success and @code{-1} on failure. The
427 following @code{errno} error condition is defined for this function:
431 You do not have privilege to set the time.
435 @strong{Portability Note:} The @code{gettimeofday}, @code{settimeofday},
436 and @code{adjtime} functions are derived from BSD.
439 @node Broken-down Time
440 @subsection Broken-down Time
441 @cindex broken-down time
442 @cindex calendar time and broken-down time
444 Calendar time is represented as a number of seconds. This is convenient
445 for calculation, but has no resemblance to the way people normally
446 represent dates and times. By contrast, @dfn{broken-down time} is a binary
447 representation separated into year, month, day, and so on. Broken down
448 time values are not useful for calculations, but they are useful for
449 printing human readable time.
451 A broken-down time value is always relative to a choice of local time
452 zone, and it also indicates which time zone was used.
454 The symbols in this section are declared in the header file @file{time.h}.
458 @deftp {Data Type} {struct tm}
459 This is the data type used to represent a broken-down time. The structure
460 contains at least the following members, which can appear in any order:
464 This is the number of seconds after the minute, normally in the range
465 @code{0} to @code{59}. (The actual upper limit is @code{61}, to allow
466 for ``leap seconds''.)
470 This is the number of minutes after the hour, in the range @code{0} to
474 This is the number of hours past midnight, in the range @code{0} to
478 This is the day of the month, in the range @code{1} to @code{31}.
481 This is the number of months since January, in the range @code{0} to
485 This is the number of years since @code{1900}.
488 This is the number of days since Sunday, in the range @code{0} to @code{6}.
491 This is the number of days since January 1, in the range @code{0} to
495 @cindex Daylight Saving Time
497 This is a flag that indicates whether Daylight Saving Time is (or was, or
498 will be) in effect at the time described. The value is positive if
499 Daylight Saving Time is in effect, zero if it is not, and negative if the
500 information is not available.
502 @item long int tm_gmtoff
503 This field describes the time zone that was used to compute this
504 broken-down time value; it is the amount you must add to the local time
505 in that zone to get GMT, in units of seconds. The value is like that of
506 the variable @code{timezone} (@pxref{Time Zone Functions}). You can
507 also think of this as the ``number of seconds west'' of GMT. The
508 @code{tm_gmtoff} field is a GNU library extension.
510 @item const char *tm_zone
511 This field is the three-letter name for the time zone that was used to
512 compute this broken-down time value. It is a GNU library extension.
518 @deftypefun {struct tm *} localtime (const time_t *@var{time})
519 The @code{localtime} function converts the calendar time pointed to by
520 @var{time} to broken-down time representation, expressed relative to the
521 user's specified time zone.
523 The return value is a pointer to a static broken-down time structure, which
524 might be overwritten by subsequent calls to any of the date and time
525 functions. (But no other library function overwrites the contents of this
528 Calling @code{localtime} has one other effect: it sets the variable
529 @code{tzname} with information about the current time zone. @xref{Time
535 @deftypefun {struct tm *} gmtime (const time_t *@var{time})
536 This function is similar to @code{localtime}, except that the broken-down
537 time is expressed as Coordinated Universal Time (UTC)---that is, as
538 Greenwich Mean Time (GMT) rather than relative to the local time zone.
540 Recall that calendar times are @emph{always} expressed in coordinated
546 @deftypefun time_t mktime (struct tm *@var{brokentime})
547 The @code{mktime} function is used to convert a broken-down time structure
548 to a calendar time representation. It also ``normalizes'' the contents of
549 the broken-down time structure, by filling in the day of week and day of
550 year based on the other date and time components.
552 The @code{mktime} function ignores the specified contents of the
553 @code{tm_wday} and @code{tm_yday} members of the broken-down time
554 structure. It uses the values of the other components to compute the
555 calendar time; it's permissible for these components to have
556 unnormalized values outside of their normal ranges. The last thing that
557 @code{mktime} does is adjust the components of the @var{brokentime}
558 structure (including the @code{tm_wday} and @code{tm_yday}).
560 If the specified broken-down time cannot be represented as a calendar time,
561 @code{mktime} returns a value of @code{(time_t)(-1)} and does not modify
562 the contents of @var{brokentime}.
564 Calling @code{mktime} also sets the variable @code{tzname} with
565 information about the current time zone. @xref{Time Zone Functions}.
568 @node Formatting Date and Time
569 @subsection Formatting Date and Time
571 The functions described in this section format time values as strings.
572 These functions are declared in the header file @file{time.h}.
577 @deftypefun {char *} asctime (const struct tm *@var{brokentime})
578 The @code{asctime} function converts the broken-down time value that
579 @var{brokentime} points to into a string in a standard format:
582 "Tue May 21 13:46:22 1991\n"
585 The abbreviations for the days of week are: @samp{Sun}, @samp{Mon},
586 @samp{Tue}, @samp{Wed}, @samp{Thu}, @samp{Fri}, and @samp{Sat}.
588 The abbreviations for the months are: @samp{Jan}, @samp{Feb},
589 @samp{Mar}, @samp{Apr}, @samp{May}, @samp{Jun}, @samp{Jul}, @samp{Aug},
590 @samp{Sep}, @samp{Oct}, @samp{Nov}, and @samp{Dec}.
592 The return value points to a statically allocated string, which might be
593 overwritten by subsequent calls to any of the date and time functions.
594 (But no other library function overwrites the contents of this
600 @deftypefun {char *} ctime (const time_t *@var{time})
601 The @code{ctime} function is similar to @code{asctime}, except that the
602 time value is specified as a @code{time_t} calendar time value rather
603 than in broken-down local time format. It is equivalent to
606 asctime (localtime (@var{time}))
609 @code{ctime} sets the variable @code{tzname}, because @code{localtime}
610 does so. @xref{Time Zone Functions}.
615 @deftypefun size_t strftime (char *@var{s}, size_t @var{size}, const char *@var{template}, const struct tm *@var{brokentime})
616 This function is similar to the @code{sprintf} function (@pxref{Formatted
617 Input}), but the conversion specifications that can appear in the format
618 template @var{template} are specialized for printing components of the date
619 and time @var{brokentime} according to the locale currently specified for
620 time conversion (@pxref{Locales}).
622 Ordinary characters appearing in the @var{template} are copied to the
623 output string @var{s}; this can include multibyte character sequences.
624 Conversion specifiers are introduced by a @samp{%} character, and are
625 replaced in the output string as follows:
629 The abbreviated weekday name according to the current locale.
632 The full weekday name according to the current locale.
635 The abbreviated month name according to the current locale.
638 The full month name according to the current locale.
641 The preferred date and time representation for the current locale.
644 The day of the month as a decimal number (range @code{01} to @code{31}).
647 The hour as a decimal number, using a 24-hour clock (range @code{00} to
651 The hour as a decimal number, using a 12-hour clock (range @code{01} to
655 The day of the year as a decimal number (range @code{001} to @code{366}).
658 The month as a decimal number (range @code{01} to @code{12}).
661 The minute as a decimal number.
664 Either @samp{am} or @samp{pm}, according to the given time value; or the
665 corresponding strings for the current locale.
668 The second as a decimal number.
671 The week number of the current year as a decimal number, starting with
672 the first Sunday as the first day of the first week.
675 The week number of the current year as a decimal number, starting with
676 the first Monday as the first day of the first week.
679 The day of the week as a decimal number, Sunday being @code{0}.
682 The preferred date representation for the current locale, but without the
686 The preferred time representation for the current locale, but with no date.
689 The year as a decimal number, but without a century (range @code{00} to
693 The year as a decimal number, including the century.
696 The time zone or name or abbreviation (empty if the time zone can't be
700 A literal @samp{%} character.
703 The @var{size} parameter can be used to specify the maximum number of
704 characters to be stored in the array @var{s}, including the terminating
705 null character. If the formatted time requires more than @var{size}
706 characters, the excess characters are discarded. The return value from
707 @code{strftime} is the number of characters placed in the array @var{s},
708 not including the terminating null character. If the value equals
709 @var{size}, it means that the array @var{s} was too small; you should
710 repeat the call, providing a bigger array.
712 If @var{s} is a null pointer, @code{strftime} does not actually write
713 anything, but instead returns the number of characters it would have written.
715 For an example of @code{strftime}, see @ref{Time Functions Example}.
719 @subsection Specifying the Time Zone with @code{TZ}
721 In POSIX systems, a user can specify the time zone by means of the
722 @code{TZ} environment variable. For information about how to set
723 environment variables, see @ref{Environment Variables}. The functions
724 for accessing the time zone are declared in @file{time.h}.
728 You should not normally need to set @code{TZ}. If the system is
729 configured properly, the default timezone will be correct. You might
730 set @code{TZ} if you are using a computer over the network from a
731 different timezone, and would like times reported to you in the timezone
732 that local for you, rather than what is local for the computer.
734 In POSIX.1 systems the value of the @code{TZ} variable can be of one of
735 three formats. With the GNU C library, the most common format is the
736 last one, which can specify a selection from a large database of time
737 zone information for many regions of the world. The first two formats
738 are used to describe the time zone information directly, which is both
739 more cumbersome and less precise. But the POSIX.1 standard only
740 specifies the details of the first two formats, so it is good to be
741 familiar with them in case you come across a POSIX.1 system that doesn't
742 support a time zone information database.
744 The first format is used when there is no Daylight Saving Time (or
745 summer time) in the local time zone:
748 @r{@var{std} @var{offset}}
751 The @var{std} string specifies the name of the time zone. It must be
752 three or more characters long and must not contain a leading colon or
753 embedded digits, commas, or plus or minus signs. There is no space
754 character separating the time zone name from the @var{offset}, so these
755 restrictions are necessary to parse the specification correctly.
757 The @var{offset} specifies the time value one must add to the local time
758 to get a Coordinated Universal Time value. It has syntax like
759 [@code{+}|@code{-}]@var{hh}[@code{:}@var{mm}[@code{:}@var{ss}]]. This
760 is positive if the local time zone is west of the Prime Meridian and
761 negative if it is east. The hour must be between @code{0} and
762 @code{24}, and the minute and seconds between @code{0} and @code{59}.
764 For example, here is how we would specify Eastern Standard Time, but
765 without any daylight savings time alternative:
771 The second format is used when there is Daylight Saving Time:
774 @r{@var{std} @var{offset} @var{dst} [@var{offset}]@code{,}@var{start}[@code{/}@var{time}]@code{,}@var{end}[@code{/}@var{time}]}
777 The initial @var{std} and @var{offset} specify the standard time zone, as
778 described above. The @var{dst} string and @var{offset} specify the name
779 and offset for the corresponding daylight savings time time zone; if the
780 @var{offset} is omitted, it defaults to one hour ahead of standard time.
782 The remainder of the specification describes when daylight savings time is
783 in effect. The @var{start} field is when daylight savings time goes into
784 effect and the @var{end} field is when the change is made back to standard
785 time. The following formats are recognized for these fields:
789 This specifies the Julian day, with @var{n} between @code{1} and @code{365}.
790 February 29 is never counted, even in leap years.
793 This specifies the Julian day, with @var{n} between @code{0} and @code{365}.
794 February 29 is counted in leap years.
796 @item M@var{m}.@var{w}.@var{d}
797 This specifies day @var{d} of week @var{w} of month @var{m}. The day
798 @var{d} must be between @code{0} (Sunday) and @code{6}. The week
799 @var{w} must be between @code{1} and @code{5}; week @code{1} is the
800 first week in which day @var{d} occurs, and week @code{5} specifies the
801 @emph{last} @var{d} day in the month. The month @var{m} should be
802 between @code{1} and @code{12}.
805 The @var{time} fields specify when, in the local time currently in
806 effect, the change to the other time occurs. If omitted, the default is
809 For example, here is how one would specify the Eastern time zone in the
810 United States, including the appropriate daylight saving time and its dates
811 of applicability. The normal offset from GMT is 5 hours; since this is
812 west of the prime meridian, the sign is positive. Summer time begins on
813 the first Sunday in April at 2:00am, and ends on the last Sunday in October
817 EST+5EDT,M4.1.0/M10.5.0
820 The schedule of daylight savings time in any particular jurisdiction has
821 changed over the years. To be strictly correct, the conversion of dates
822 and times in the past should be based on the schedule that was in effect
823 then. However, this format has no facilities to let you specify how the
824 schedule has changed from year to year. The most you can do is specify
825 one particular schedule---usually the present day schedule---and this is
826 used to convert any date, no matter when. For precise time zone
827 specifications, it is best to use the time zone information database
830 The third format looks like this:
836 Each operating system interprets this format differently; in the GNU C
837 library, @var{characters} is the name of a file which describes the time
840 @pindex /etc/localtime
842 If the @code{TZ} environment variable does not have a value, the
843 operation chooses a time zone by default. In the GNU C library, the
844 default time zone is like the specification @samp{TZ=:/etc/localtime}
845 (or @samp{TZ=:/usr/local/etc/localtime}, depending on how GNU C library
846 was configured; @pxref{Installation}). Other C libraries use their own
847 rule for choosing the default time zone, so there is little we can say
850 @cindex time zone database
851 @pindex /share/lib/zoneinfo
853 If @var{characters} begins with a slash, it is an absolute file name;
854 otherwise the library looks for the file
855 @w{@file{/share/lib/zoneinfo/@var{characters}}}. The @file{zoneinfo}
856 directory contains data files describing local time zones in many
857 different parts of the world. The names represent major cities, with
858 subdirectories for geographical areas; for example,
859 @file{America/New_York}, @file{Europe/London}, @file{Asia/Hong_Kong}.
860 These data files are installed by the system administrator, who also
861 sets @file{/etc/localtime} to point to the data file for the local time
862 zone. The GNU C library comes with a large database of time zone
863 information for most regions of the world, which is maintained by a
864 community of volunteers and put in the public domain.
866 @node Time Zone Functions
867 @subsection Functions and Variables for Time Zones
871 @deftypevar char * tzname [2]
872 The array @code{tzname} contains two strings, which are the standard
873 three-letter names of the pair of time zones (standard and daylight
874 savings) that the user has selected. @code{tzname[0]} is the name of
875 the standard time zone (for example, @code{"EST"}), and @code{tzname[1]}
876 is the name for the time zone when daylight savings time is in use (for
877 example, @code{"EDT"}). These correspond to the @var{std} and @var{dst}
878 strings (respectively) from the @code{TZ} environment variable.
880 The @code{tzname} array is initialized from the @code{TZ} environment
881 variable whenever @code{tzset}, @code{ctime}, @code{strftime},
882 @code{mktime}, or @code{localtime} is called.
887 @deftypefun void tzset (void)
888 The @code{tzset} function initializes the @code{tzname} variable from
889 the value of the @code{TZ} environment variable. It is not usually
890 necessary for your program to call this function, because it is called
891 automatically when you use the other time conversion functions that
892 depend on the time zone.
895 The following variables are defined for compatibility with System V
896 Unix. These variables are set by calling @code{localtime}.
900 @deftypevar {long int} timezone
901 This contains the difference between GMT and local standard time, in
902 seconds. For example, in the U.S. Eastern time zone, the value is
908 @deftypevar int daylight
909 This variable has a nonzero value if the standard U.S. daylight savings
913 @node Time Functions Example
914 @subsection Time Functions Example
916 Here is an example program showing the use of some of the local time and
917 calendar time functions.
920 @include strftim.c.texi
923 It produces output like this:
926 Wed Jul 31 13:02:36 1991
927 Today is Wednesday, July 31.
928 The time is 01:02 PM.
932 @node Setting an Alarm
933 @section Setting an Alarm
935 The @code{alarm} and @code{setitimer} functions provide a mechanism for a
936 process to interrupt itself at some future time. They do this by setting a
937 timer; when the timer expires, the process receives a signal.
939 @cindex setting an alarm
940 @cindex interval timer, setting
941 @cindex alarms, setting
942 @cindex timers, setting
943 Each process has three independent interval timers available:
947 A real-time timer that counts clock time. This timer sends a
948 @code{SIGALRM} signal to the process when it expires.
949 @cindex real-time timer
950 @cindex timer, real-time
953 A virtual timer that counts CPU time used by the process. This timer
954 sends a @code{SIGVTALRM} signal to the process when it expires.
955 @cindex virtual timer
956 @cindex timer, virtual
959 A profiling timer that counts both CPU time used by the process, and CPU
960 time spent in system calls on behalf of the process. This timer sends a
961 @code{SIGPROF} signal to the process when it expires.
962 @cindex profiling timer
963 @cindex timer, profiling
965 This timer is useful for profiling in interpreters. The interval timer
966 mechanism does not have the fine granularity necessary for profiling
971 You can only have one timer of each kind set at any given time. If you
972 set a timer that has not yet expired, that timer is simply reset to the
975 You should establish a handler for the appropriate alarm signal using
976 @code{signal} or @code{sigaction} before issuing a call to @code{setitimer}
977 or @code{alarm}. Otherwise, an unusual chain of events could cause the
978 timer to expire before your program establishes the handler, and in that
979 case it would be terminated, since that is the default action for the alarm
980 signals. @xref{Signal Handling}.
982 The @code{setitimer} function is the primary means for setting an alarm.
983 This facility is declared in the header file @file{sys/time.h}. The
984 @code{alarm} function, declared in @file{unistd.h}, provides a somewhat
985 simpler interface for setting the real-time timer.
991 @deftp {Data Type} {struct itimerval}
992 This structure is used to specify when a timer should expire. It contains
993 the following members:
995 @item struct timeval it_interval
996 This is the interval between successive timer interrupts. If zero, the
997 alarm will only be sent once.
999 @item struct timeval it_value
1000 This is the interval to the first timer interrupt. If zero, the alarm is
1004 The @code{struct timeval} data type is described in @ref{High-Resolution
1010 @deftypefun int setitimer (int @var{which}, struct itimerval *@var{new}, struct itimerval *@var{old})
1011 The @code{setitimer} function sets the timer specified by @var{which}
1012 according to @var{new}. The @var{which} argument can have a value of
1013 @code{ITIMER_REAL}, @code{ITIMER_VIRTUAL}, or @code{ITIMER_PROF}.
1015 If @var{old} is not a null pointer, @code{setitimer} returns information
1016 about any previous unexpired timer of the same kind in the structure it
1019 The return value is @code{0} on success and @code{-1} on failure. The
1020 following @code{errno} error conditions are defined for this function:
1024 The timer interval was too large.
1030 @deftypefun int getitimer (int @var{which}, struct itimerval *@var{old})
1031 The @code{getitimer} function stores information about the timer specified
1032 by @var{which} in the structure pointed at by @var{old}.
1034 The return value and error conditions are the same as for @code{setitimer}.
1042 This constant can be used as the @var{which} argument to the
1043 @code{setitimer} and @code{getitimer} functions to specify the real-time
1048 @item ITIMER_VIRTUAL
1049 @findex ITIMER_VIRTUAL
1050 This constant can be used as the @var{which} argument to the
1051 @code{setitimer} and @code{getitimer} functions to specify the virtual
1058 This constant can be used as the @var{which} argument to the
1059 @code{setitimer} and @code{getitimer} functions to specify the profiling
1065 @deftypefun {unsigned int} alarm (unsigned int @var{seconds})
1066 The @code{alarm} function sets the real-time timer to expire in
1067 @var{seconds} seconds. If you want to cancel any existing alarm, you
1068 can do this by calling @code{alarm} with a @var{seconds} argument of
1071 The return value indicates how many seconds remain before the previous
1072 alarm would have been sent. If there is no previous alarm, @code{alarm}
1076 The @code{alarm} function could be defined in terms of @code{setitimer}
1081 alarm (unsigned int seconds)
1083 struct itimerval old, new;
1084 new.it_interval.tv_usec = 0;
1085 new.it_interval.tv_sec = 0;
1086 new.it_value.tv_usec = 0;
1087 new.it_value.tv_sec = (long int) seconds;
1088 if (setitimer (ITIMER_REAL, &new, &old) < 0)
1091 return old.it_value.tv_sec;
1095 There is an example showing the use of the @code{alarm} function in
1096 @ref{Handler Returns}.
1098 If you simply want your process to wait for a given number of seconds,
1099 you should use the @code{sleep} function. @xref{Sleeping}.
1101 You shouldn't count on the signal arriving precisely when the timer
1102 expires. In a multiprocessing environment there is typically some
1103 amount of delay involved.
1105 @strong{Portability Note:} The @code{setitimer} and @code{getitimer}
1106 functions are derived from BSD Unix, while the @code{alarm} function is
1107 specified by the POSIX.1 standard. @code{setitimer} is more powerful than
1108 @code{alarm}, but @code{alarm} is more widely used.
1113 The function @code{sleep} gives a simple way to make the program wait
1114 for short periods of time. If your program doesn't use signals (except
1115 to terminate), then you can expect @code{sleep} to wait reliably for
1116 the specified amount of time. Otherwise, @code{sleep} can return sooner
1117 if a signal arrives; if you want to wait for a given period regardless
1118 of signals, use @code{select} (@pxref{Waiting for I/O}) and don't
1119 specify any descriptors to wait for.
1120 @c !!! select can get EINTR; using SA_RESTART makes sleep win too.
1124 @deftypefun {unsigned int} sleep (unsigned int @var{seconds})
1125 The @code{sleep} function waits for @var{seconds} or until a signal
1126 is delivered, whichever happens first.
1128 If @code{sleep} function returns because the requested time has
1129 elapsed, it returns a value of zero. If it returns because of delivery
1130 of a signal, its return value is the remaining time in the sleep period.
1132 The @code{sleep} function is declared in @file{unistd.h}.
1135 Resist the temptation to implement a sleep for a fixed amount of time by
1136 using the return value of @code{sleep}, when nonzero, to call
1137 @code{sleep} again. This will work with a certain amount of accuracy as
1138 long as signals arrive infrequently. But each signal can cause the
1139 eventual wakeup time to be off by an additional second or so. Suppose a
1140 few signals happen to arrive in rapid succession by bad luck---there is
1141 no limit on how much this could shorten or lengthen the wait.
1143 Instead, compute the time at which the program should stop waiting, and
1144 keep trying to wait until that time. This won't be off by more than a
1145 second. With just a little more work, you can use @code{select} and
1146 make the waiting period quite accurate. (Of course, heavy system load
1147 can cause unavoidable additional delays---unless the machine is
1148 dedicated to one application, there is no way you can avoid this.)
1150 On some systems, @code{sleep} can do strange things if your program uses
1151 @code{SIGALRM} explicitly. Even if @code{SIGALRM} signals are being
1152 ignored or blocked when @code{sleep} is called, @code{sleep} might
1153 return prematurely on delivery of a @code{SIGALRM} signal. If you have
1154 established a handler for @code{SIGALRM} signals and a @code{SIGALRM}
1155 signal is delivered while the process is sleeping, the action taken
1156 might be just to cause @code{sleep} to return instead of invoking your
1157 handler. And, if @code{sleep} is interrupted by delivery of a signal
1158 whose handler requests an alarm or alters the handling of @code{SIGALRM},
1159 this handler and @code{sleep} will interfere.
1161 On the GNU system, it is safe to use @code{sleep} and @code{SIGALRM} in
1162 the same program, because @code{sleep} does not work by means of
1165 @node Resource Usage
1166 @section Resource Usage
1168 @pindex sys/resource.h
1169 The function @code{getrusage} and the data type @code{struct rusage}
1170 are used for examining the usage figures of a process. They are declared
1171 in @file{sys/resource.h}.
1173 @comment sys/resource.h
1175 @deftypefun int getrusage (int @var{processes}, struct rusage *@var{rusage})
1176 This function reports the usage totals for processes specified by
1177 @var{processes}, storing the information in @code{*@var{rusage}}.
1179 In most systems, @var{processes} has only two valid values:
1182 @comment sys/resource.h
1185 Just the current process.
1187 @comment sys/resource.h
1189 @item RUSAGE_CHILDREN
1190 All child processes (direct and indirect) that have terminated already.
1193 In the GNU system, you can also inquire about a particular child process
1194 by specifying its process ID.
1196 The return value of @code{getrusage} is zero for success, and @code{-1}
1201 The argument @var{processes} is not valid.
1205 One way of getting usage figures for a particular child process is with
1206 the function @code{wait4}, which returns totals for a child when it
1207 terminates. @xref{BSD Wait Functions}.
1209 @comment sys/resource.h
1211 @deftp {Data Type} {struct rusage}
1212 This data type records a collection usage amounts for various sorts of
1213 resources. It has the following members, and possibly others:
1216 @item struct timeval ru_utime
1217 Time spent executing user instructions.
1219 @item struct timeval ru_stime
1220 Time spent in operating system code on behalf of @var{processes}.
1222 @item long int ru_maxrss
1223 The maximum resident set size used, in kilobytes. That is, the maximum
1224 number of kilobytes that @var{processes} used in real memory simultaneously.
1226 @item long int ru_ixrss
1227 An integral value expressed in kilobytes times ticks of execution, which
1228 indicates the amount of memory used by text that was shared with other
1231 @item long int ru_idrss
1232 An integral value expressed the same way, which is the amount of
1233 unshared memory used in data.
1235 @item long int ru_isrss
1236 An integral value expressed the same way, which is the amount of
1237 unshared memory used in stack space.
1239 @item long int ru_minflt
1240 The number of page faults which were serviced without requiring any I/O.
1242 @item long int ru_majflt
1243 The number of page faults which were serviced by doing I/O.
1245 @item long int ru_nswap
1246 The number of times @var{processes} was swapped entirely out of main memory.
1248 @item long int ru_inblock
1249 The number of times the file system had to read from the disk on behalf
1252 @item long int ru_oublock
1253 The number of times the file system had to write to the disk on behalf
1256 @item long int ru_msgsnd
1257 Number of IPC messages sent.
1259 @item long ru_msgrcv
1260 Number of IPC messages received.
1262 @item long int ru_nsignals
1263 Number of signals received.
1265 @item long int ru_nvcsw
1266 The number of times @var{processes} voluntarily invoked a context switch
1267 (usually to wait for some service).
1269 @item long int ru_nivcsw
1270 The number of times an involuntary context switch took place (because
1271 the time slice expired, or another process of higher priority became
1276 An additional historical function for examining usage figures,
1277 @code{vtimes}, is supported but not documented here. It is declared in
1278 @file{sys/vtimes.h}.
1280 @node Limits on Resources
1281 @section Limiting Resource Usage
1282 @cindex resource limits
1283 @cindex limits on resource usage
1284 @cindex usage limits
1286 You can specify limits for the resource usage of a process. When the
1287 process tries to exceed a limit, it may get a signal, or the system call
1288 by which it tried to do so may fail, depending on the limit. Each
1289 process initially inherits its limit values from its parent, but it can
1290 subsequently change them.
1292 @pindex sys/resource.h
1293 The symbols in this section are defined in @file{sys/resource.h}.
1295 @comment sys/resource.h
1297 @deftypefun int getrlimit (int @var{resource}, struct rlimit *@var{rlp})
1298 Read the current value and the maximum value of resource @var{resource}
1299 and store them in @code{*@var{rlp}}.
1301 The return value is @code{0} on success and @code{-1} on failure. The
1302 only possible @code{errno} error condition is @code{EFAULT}.
1305 @comment sys/resource.h
1307 @deftypefun int setrlimit (int @var{resource}, struct rlimit *@var{rlp})
1308 Store the current value and the maximum value of resource @var{resource}
1309 in @code{*@var{rlp}}.
1311 The return value is @code{0} on success and @code{-1} on failure. The
1312 following @code{errno} error condition is possible:
1316 You tried to change the maximum permissible limit value,
1317 but you don't have privileges to do so.
1321 @comment sys/resource.h
1323 @deftp {Data Type} {struct rlimit}
1324 This structure is used with @code{getrlimit} to receive limit values,
1325 and with @code{setrlimit} to specify limit values. It has two fields:
1329 The current value of the limit in question.
1330 This is also called the ``soft limit''.
1334 The maximum permissible value of the limit in question. You cannot set
1335 the current value of the limit to a larger number than this maximum.
1336 Only the super user can change the maximum permissible value.
1337 This is also called the ``hard limit''.
1341 In @code{getrlimit}, the structure is an output; it receives the current
1342 values. In @code{setrlimit}, it specifies the new values.
1345 Here is a list of resources that you can specify a limit for.
1346 Those that are sizes are measured in bytes.
1349 @comment sys/resource.h
1353 The maximum amount of cpu time the process can use. If it runs for
1354 longer than this, it gets a signal: @code{SIGXCPU}. The value is
1355 measured in seconds. @xref{Operation Error Signals}.
1357 @comment sys/resource.h
1360 @vindex RLIMIT_FSIZE
1361 The maximum size of file the process can create. Trying to write a
1362 larger file causes a signal: @code{SIGXFSZ}. @xref{Operation Error
1365 @comment sys/resource.h
1369 The maximum size of data memory for the process. If the process tries
1370 to allocate data memory beyond this amount, the allocation function
1373 @comment sys/resource.h
1376 @vindex RLIMIT_STACK
1377 The maximum stack size for the process. If the process tries to extend
1378 its stack past this size, it gets a @code{SIGSEGV} signal.
1379 @xref{Program Error Signals}.
1381 @comment sys/resource.h
1385 The maximum size core file that this process can create. If the process
1386 terminates and would dump a core file larger than this maximum size,
1387 then no core file is created. So setting this limit to zero prevents
1388 core files from ever being created.
1390 @comment sys/resource.h
1394 The maximum amount of physical memory that this process should get.
1395 This parameter is a guide for the system's scheduler and memory
1396 allocator; the system may give the process more memory when there is a
1399 @comment sys/resource.h
1401 @item RLIMIT_MEMLOCK
1402 The maximum amount of memory that can be locked into physical memory (so
1403 it will never be paged out).
1405 @comment sys/resource.h
1408 The maximum number of processes that can be created with the same user ID.
1409 If you have reached the limit for your user ID, @code{fork} will fail
1410 with @code{EAGAIN}. @xref{Creating a Process}.
1412 @comment sys/resource.h
1415 @vindex RLIMIT_NOFILE
1417 @vindex RLIMIT_OFILE
1418 The maximum number of files that the process can open. If it tries to
1419 open more files than this, it gets error code @code{EMFILE}.
1420 @xref{Error Codes}. Not all systems support this limit; GNU does, and
1423 @comment sys/resource.h
1426 @vindex RLIM_NLIMITS
1427 The number of different resource limits. Any valid @var{resource}
1428 operand must be less than @code{RLIM_NLIMITS}.
1431 @comment sys/resource.h
1433 @defvr Constant int RLIM_INFINITY
1434 This constant stands for a value of ``infinity'' when supplied as
1435 the limit value in @code{setrlimit}.
1438 @c ??? Someone want to finish these?
1439 Two historical functions for setting resource limits, @code{ulimit} and
1440 @code{vlimit}, are not documented here. The latter is declared in
1441 @file{sys/vlimit.h} and comes from BSD.
1444 @section Process Priority
1445 @cindex process priority
1446 @cindex priority of a process
1448 @pindex sys/resource.h
1449 When several processes try to run, their respective priorities determine
1450 what share of the CPU each process gets. This section describes how you
1451 can read and set the priority of a process. All these functions and
1452 macros are declared in @file{sys/resource.h}.
1454 The range of valid priority values depends on the operating system, but
1455 typically it runs from @code{-20} to @code{20}. A lower priority value
1456 means the process runs more often. These constants describe the range of
1460 @comment sys/resource.h
1464 The smallest valid priority value.
1466 @comment sys/resource.h
1470 The smallest valid priority value.
1473 @comment sys/resource.h
1475 @deftypefun int getpriority (int @var{class}, int @var{id})
1476 Read the priority of a class of processes; @var{class} and @var{id}
1477 specify which ones (see below). If the processes specified do not all
1478 have the same priority, this returns the smallest value that any of them
1481 The return value is the priority value on success, and @code{-1} on
1482 failure. The following @code{errno} error condition are possible for
1487 The combination of @var{class} and @var{id} does not match any existing
1491 The value of @var{class} is not valid.
1494 When the return value is @code{-1}, it could indicate failure, or it
1495 could be the priority value. The only way to make certain is to set
1496 @code{errno = 0} before calling @code{getpriority}, then use @code{errno
1497 != 0} afterward as the criterion for failure.
1500 @comment sys/resource.h
1502 @deftypefun int setpriority (int @var{class}, int @var{id}, int @var{priority})
1503 Set the priority of a class of processes to @var{priority}; @var{class}
1504 and @var{id} specify which ones (see below).
1506 The return value is @code{0} on success and @code{-1} on failure. The
1507 following @code{errno} error condition are defined for this function:
1511 The combination of @var{class} and @var{id} does not match any existing
1515 The value of @var{class} is not valid.
1518 You tried to set the priority of some other user's process, and you
1519 don't have privileges for that.
1522 You tried to lower the priority of a process, and you don't have
1523 privileges for that.
1527 The arguments @var{class} and @var{id} together specify a set of
1528 processes you are interested in. These are the possible values for
1532 @comment sys/resource.h
1535 @vindex PRIO_PROCESS
1536 Read or set the priority of one process. The argument @var{id} is a
1539 @comment sys/resource.h
1543 Read or set the priority of one process group. The argument @var{id} is
1546 @comment sys/resource.h
1550 Read or set the priority of one user's processes. The argument @var{id}
1554 If the argument @var{id} is 0, it stands for the current process,
1555 current process group, or the current user, according to @var{class}.
1557 @c ??? I don't know where we should say this comes from.
1560 @deftypefun int nice (int @var{increment})
1561 Increment the priority of the current process by @var{increment}.
1562 The return value is the same as for @code{setpriority}.
1564 Here is an equivalent definition for @code{nice}:
1568 nice (int increment)
1570 int old = getpriority (PRIO_PROCESS, 0);
1571 return setpriority (PRIO_PROCESS, 0, old + increment);