1 Revision history for the Perl extension Time::HiRes.
4 - [rt.cpan.org #29747]: Build failure with perl 5.005_05
5 Fixed by regenerating the ppport.h using Devel::PPPort 3.13.
8 - t/HiRes.t failed in Perl 5.6.2,
9 "action is not of type POSIX::SigAction at t/HiRes.t line 318",
10 reported and fixed by Anton Berezin, the reason was faulty
11 use of sigaction() when restoring the old action.
14 - with bleadperl in VMS the HiRes.t overrun the maximum number
15 of deferred signals because the libc SIGALRM was not strong
16 enough to interrupt select(), and select() got restarted every
17 time, solution is to use POSIX::SigAction if available.
18 A fix from Craig Berry (not 100% there, but helps).
19 - allow for more measuring noise for ualarm() tests 35..37
22 - nanosleep() and clock_nanosleep() detection and use were
23 quite broken; in Linux -lrt needed; fixes from Zefram
24 - [internal] slightly cleaner building of $DEFINE in Makefile.PL,
25 should avoid double/conflicting -D flags
28 - allow 10% of slop in test #14 (testing difference between
29 CORE::time() and Time::HiRes::time()), there seem to be often
30 transient failures from Perl smoke builds on this test
34 - use int main(int argc, char **argv) consistently in Makefile.PL,
36 [rt.cpan.org #23868] nanosleep not detected under Mac OS 10.3.9 starting with Time::HiRes 1.96
37 - if someone still has the locale-broken Perl 5.8.0,
38 suggest that they upgrade their Perl
41 - restore the -DATLEASTFIVEOHOHFIVE, Win32 needed it still
44 - upgrade to ppport.h 3.10_02
45 - remove the -DATLEASTFIVEOHOHFIVE
46 - use the ppport.h PL_ppaddr, PL_statcache, PL_laststatval
47 - use the ppport.h aTHXR for calling Perl stat()
48 - switch into four-digit version since 2.0 is coming up
49 awfully fast but not feeling like a major rewrite
52 - 1.95 broke building in Win32 (since pp_stat is not exported),
53 figured out how to call an op directly in 5.005 (use Perl_ppaddr
55 - backport to Perl 5.004_05 (requires using statcache
56 and laststatval instead of PL_statcache and PL_laststatval)
57 (also checked to work in 5.005_04, 5.6.1, and 5.8.8 with threads)
60 - 1.95 broke builds for threaded Perls, rt.cpan.org tickets:
61 [rt.cpan.org #23694] Time::HiRes fails tests on Solaris and Perl 5.6.1
62 [rt.cpan.org #23712] Time-HiRes 1.95 Fails make on AIX 5.2 with Perl 5.8.8
63 [rt.cpan.org #23730] Time::HiRes 1.95 fails make on MacOS X 10.3.9/perl 5.8.8
64 - use main() prototype consistently in Makefile.PL
67 - integrate core change #29180: Silence VC++ compiler warnings
69 - do not use PL_ppaddr in stat() because that is not available
71 - regenerate fallback/*.inc for older Perls without
72 ExtUtils::Constant because of d_hires_stat, resolves
73 [rt.cpan.org #23694] Time::HiRes fails tests on Solaris and Perl 5.6.1
74 - Make Makefile.PL more defensive against false PERL_CORE
77 - file timestamps oddities seen: the atime and mtime
78 can be out of sync (modify first and read second can leave
79 atime < mtime) and mtime can be subsecond while atime is not.
80 So make the test more forgiving.
83 - the ualarm() tests (34-37) assumed that ualarm(N)
84 could never alarm in less than N seconds, widened
85 the acceptable relative range to 0.9..1.5. Addresses
86 [rt.cpan.org #22090] and [rt.cpan.org #22091].
88 - skip the stat() tests in cygwin and win32, because
89 if run on FAT the timestamp granularity is only 2 seconds.
90 Any good way to detect (cygwin or win32) whether we are
91 being run on NTFS or anywhere with better timestamps?
92 Addresses [rt.cpan.org #22089] and [rt.cpan.org #22098].
95 - scan for subsecond resolution timestamps in struct stat,
96 some known possibilities:
98 (1) struct timespec st_atimespec;
104 (4) timestruc_t st_atim;
109 If something like this is found, one can do
112 my @stat = Time::HiRes::stat();
114 or even override the standard stat():
116 use Time::HiRes qw(stat);
118 to get the stat() timestamps
120 my ($atime, $mtime, $ctime) = @stat[8, 9, 10];
122 with subsecond resolution (assuming both the operating
123 system and the filesystem support that kind of thing).
125 Contributions for more systems (especially non-UNIX,
126 e.g. but not limited to: Win32, VMS, OS/2) gladly accepted.
127 (also more UNIX variants welcome: HP-UX? IRIX?)
129 Thanks to H.Merijn Brand, John Peacock, and Craig
130 Berry for brave beta testing.
133 - ualarm() in SuSE 10.1 was overflowing after ~4.2 seconds,
134 possibly due to a glibc bug/feature (suspected overflow at
135 2**32 microseconds?), workaround by using the setitimer()
136 implementation of ualarm() if either useconds or
137 interval > 999_999 (this case seems to vary between systems:
138 are useconds more than 999_999 for ualarm() defined or not)
139 Added more ualarm() tests to catch various overflow points,
140 hopefully no problems in various platforms.
141 (The problem report by Mark Seger and Jon Paul Sullivan of HP.)
144 - tweak still needed for Const64(), from Jerry Hedden
145 - get a freshly generated ppport.h
146 - update Copyright years
149 - Const64() already appends an 'LL' (or i64), so provide LL and i64
150 forms for the IV_1E[679] (effects Win32 and Cygwin), reported by
152 - the Changes entry for 1.88 talked about [IN]V_1[679],
156 - clean up the g++ warnings in HiRes.xs, all of them
157 about mixing integer and floating point, introduce
158 constants IV_1E[679] and NV_1E[679]
161 - [rt.cpan.org #17442] 'make test' frequently fails under
162 Cygwin Perl v5.8.8, reported and patched by J. R. Hedden
163 (two race condition bugs in the END block in the case the
164 main process dies before the timer process, unearthed
165 by a bug in Cygwin ualarm)
168 - HiRes.t:s/ok 32/ok 33/, from Dominic Dunlop
169 - tighten up the clock() test marginally by requiring non-negative
170 - clock_nanosleep() and clock() doc tweaks
173 - the interface to clock_nanosleep() is more natural
174 when it is like (hires) time() (instead of like nanosleep),
175 and the .xs implementation of clock_nanosleep() in 1.84
177 - the semantics of clock() are not quite so silly as I thought,
178 but still somewhat odd, documented as such
179 - additional enhancements to the clock() documentation
180 - add test for clock_nanosleep() (I cannot test this
181 since none of my systems have the function)
182 - add test for clock()
185 - add clock() which returns the processor time in
186 (floating point) seconds since an arbitrary era
187 - add clock_nanosleep() which suspends the current
188 thread until either absolute time or for relative time
189 - [rt.cpan.org #16486] printf missing value in HiRes.t
190 - add constants CLOCKS_PER_SEC, CLOCK_SOFTTIME, TIMER_ABSTIME
194 - has_symbol() was wrong since e.g. ITIMER_VIRTUAL is exported
195 via @EXPORT_OK even when it is not available. This is heinous.
196 @EXPORT_OK should be determined at Makefile.PL time.
197 - be more lenient is testing clock_gettime(): allow more slop,
198 and retry up to three times, sleeping a random nap between
200 - human months are one-based (noticed by Anton Berezin)
203 - CLOCK_REALTIME is an enum value (of the clockid_t enum)
204 in HP-UX (and might be so elsewhere, too), debugged by
206 - include const-c.inc as late as possible (from Randy Kobes,
207 [rt.cpan.org #15552] to avoid undefined usleep() on Win32
210 - try to be more robust and consistent in the detection of
211 CLOCK_REALTIME and ITIMER_VIRTUAL in HiRes.t: the proper
216 eval 'import Time::HiRes qw($symbol)';
217 return 0 unless $@ eq '';
218 return exists ${"Time::HiRes::$symbol"};
225 in the test. All these moves are needed because
227 1) one cannot directly do eval 'Time::HiRes::FOO_BAR'
228 because FOO_BAR might have a true value of zero
229 (or in the general case an empty string or even undef)
231 2) In case FOO_BAR is not available in this platform,
232 &FOO_BAR avoids the bareword warning
234 - wait more (1.5 seconds instead of 0.1) for the CLOCK_REALTIME test
235 but expect the 'customary' slop of 0.20 instead of 0.25
236 - fixed inside a comment HAS_POLL -> TIME_HIRES_NANOSLEEP
237 - at the end of HiRest.t tell how close we were to termination
240 - Gisle noticed a mistake (using HAS_NANOSLEEP) in 1.79
243 - try nanosleep for emulating usleep -- may help in some weird
244 embedded realtime places which have nanosleep but neither usleep
245 nor select nor poll (doesn't have to be weird embedded realtime
246 place, though -- in many places usleep is nanosleep anyway)
247 - try poll for emulating usleep -- this may help some obscure/old
248 SVR4 places that have neither usleep nor select
249 - a redundant test guard in HiRes.t
252 - ITIMER_VIRTUAL detection in HiRes.t had problems (that we cannot
253 in the general case fail already at 'use' phase is suboptimal)
254 - fixes to the documentation of clock_gettime() and clock_getres()
257 - add support for the POSIX clock_gettime() and clock_getres(),
258 if available, either as library calls or as syscalls
259 - be more defensive about missing functionality: break out
260 early (during 'use') if no e.g. clock_getres() is available,
261 and protect our back by trapping those cases also in HiRes.xs
262 - the test added in 1.76 could cause an endless loop e.g. in Solaris,
263 due to mixing of sleep() and alarm() (bad programmer, no cookie!)
266 - testing for nanosleep had wrong logic which caused nanosleep
267 to become undefined for e.g. Mac OS X
268 - added a test for a core dump that was introduced by Perl 5.8.0
269 safe signals and was fixed for the time of 5.8.1 (one report of
270 the core dump was [perl #20920]), the test skipped pre-5.8.1.
271 - *cough* s/unanosleep/nanosleep/g; *cough*
274 - installation patch from Gisle Aas: in Perls 5.8.x and later
275 use MakeMaker INSTALLDIRS value of 'perl' instead of 'site'.
278 - [cpan #14608] Solaris 8 perl 5.005_03 File::Spec module does not have method rel2abs
279 (the workaround is not to use rel2abs, should not be necessary)
280 - [cpan #14642] U2time wrongly exported on the C API
281 (patch supplied by the reporter, SALVA@cpan.org)
282 - add release dates to Changes
285 - Time::HiRes::nanosleep support for Solaris [PATCH]
286 (POSIX::uname() not available if building with core perl,
287 from Gisle Aas, via perl5-porters, perl change #25295)
290 - going back to the 1.68 loader setup (using DynaLoader)
291 since too many weird things starting breaking
292 - fix a typo in José Auguste-Etienne's name
295 - a thinko in the nanosleep() detection
296 - move more changes stuff from the README to Changes
297 - add -w to the Makefile.PL
300 - oops in 1.69 about @ISA (not affecting anything but silly)
301 - add copyright 2005 to HiRes.pm
302 - add copyright and license to HiRes.xs
303 - add copyrights 2003, 2004, 2005 to README
306 - actually run a test for nanosleep
307 (if there is no $Config{d_nanosleep}) since e.g. in AIX 4.2
308 it seems that one can link in nanosleep() but then calling
309 it fails instantly and sets errno to ENOSYS (Not implemented).
310 This may be fixable in the AIX case by figuring out the right
311 (realtime POSIX?) libs and whatnot, but in the general case
312 running a real test case is better. (Of course, this change
313 will no doubt run into portability problems because of the
314 execution step...) Note that because of hysterical raisins
315 most Perls do NOT have $Config{d_nanosleep} (scanning for
316 it by Configure would in many platforms require linking in
317 things like -lrt, which would in many platforms be a bad idea
319 (from José Auguste-Etienne)
320 - support XSLoader also since it's much faster
321 (from Alexey Tourbin)
322 - add SEE ALSO (BSD::Resource and Time::TAI64)
325 - somehow 1.67 had a lot of doubled lines (a major cut-and-paste
326 error suspected), but miraculously it still worked since the
327 doubling took place below the __END__ token
328 - undef Pause() before defining it to avoid redefinition warnings
329 during compilation in case perl.h had already defined Pause()
330 (part of perl change #24271)
334 - (internal) don't ignore the return value of gettimeofday()
335 - (external) return undef or an empty if the C gettimeofday() fails
336 (affects Time::HiRes gettimeofday() and the hires time())
340 - fix the 'hierachy' typo in Makefile.PL [rt.cpan.org #8492]
341 - should now build in Solaris [rt.cpan.org #7165] (since 1.64)
342 - should now build in Cygwin [rt.cpan.org #7535] (since 1.64)
343 - close also [rt.cpan.org #5933] "Time::HiRes::time does not
344 pick up time adjustments like ntp" since ever reproducing it
345 (and therefore verifying a possible fix) in the same environment
346 has become rather unlikely
349 - one should not mix u?alarm and sleep (the tests modified
350 by 1.65, #12 and #13, hung in Solaris), now we just busy
351 loop executing an empty block
352 - in the documentation underline the unspecificity of mixing
354 - small spelling fixes
357 - regenerate ppport.h with Devel::PPPort 3.03,
358 now the MY_CXT_CLONE is defined in ppport.h,
359 we no more need to do that.
361 - the test #12 would often hang in sigsuspend() (at least that's
362 where Mac OS X' ktrace shows it hanging). With the sleep()s
363 changed to sleep(1)s, the tests still pass but no hang after
364 a few hundred repeats.
367 - Win32 and any ithread build: ppport.h didn't define
368 MY_CXT_CLONE, which seems to be a Time-HiRes-ism.
371 - Skip testing if under PERL_CORE and Time::HiRes has not
372 been Configured (from Marcus Holland-Moritz, core change
374 - Use ppport.h generated by Devel::PPPort 3.01,
375 allowing cutting away our own portability code.
376 - Don't use $ENV{PERL_CORE} for < 5.6.0.
377 - Don't use "for my $i" for <= 5.003.
378 - Don't use Pause() for <= 5.003.
379 - Can't use newSVpvf for <= 5.003.
380 (most of the changes from Marcus)
383 - Win32: reset reading from the performance counters every
384 five minutes to better track wall clock time (thanks to
385 PC timers being often quite bad), should help long-running
389 - Win32: Patch from Steve Hay
390 [PATCH] Re: [perl #30755] [Win32] Different results from Time::HiRes::gettimeofdayunder the debugger
391 to [perl #30755] reported by Nigel Sandever
393 - Cygwin: Use the Win32 recalibration code also in Cygwin if the
394 <w32api/windows.h> APIs are available. Cygwin testing by
395 Yitzchak Scott-Thoennes.
397 - Solaris: use -lposix4 to get nanosleep for Solaris 2.6,
398 after that keep using -lrt, patch from Alan Burlison,
399 bug reported in [cpan #7165]
402 - Change the Win32 recalibration limit to 0.5 seconds and tweak
403 the documentation to blather less about the gory details of the
404 Win32 implementation and more about the complications in general
405 of meddling with the system clock.
408 - Document the 1.57 change better.
411 - Win32/Cygwin/MinGW: if the performance counter drifts by more
412 than two seconds from the system clock (due to ntp adjustments,
413 for example), recalibrate our internal counter: from Jan Dubois,
414 based on [cpan #5933] by Jerry D. Hedden.
417 - Give a clearer message if the tests timeout (perl change #22253)
418 - Don't use /tmp or its moral equivalents (perl bug #15036,
422 - Windows: mingw32 patch from Mike Pomraning (use Perl's Const64()
423 instead of VC-specific i64 suffix)
426 - Solaris: like Tru64 (dec_osf) also Solaris need -lrt for nanosleep
429 - Windows: higher resolution time() by using the Windows
430 performance counter API, from Jan Dubois and Anton Shcherbinin.
431 The exact new higher resolution depends on the hardware,
432 but it should be quite a bit better than using the basic
436 - In AIX (v?) with perl 5.6.1 the HiRes.t can hang after
437 the subtest 18. No known analysis nor fix, but added
438 an alarm (that requires fork() and alarm()) to the test.
441 - doc tweaks from mjd (perl change #20456)
442 - NCR MP-RAS hints file added (svr4.pl) (perl change #21249)
445 - add a message (for non-core builds) to Makefile.PL about
446 the LC_ALL=C workaround
447 - &Time::HiRes::d_nanosleep was broken (perl change #20131)
448 - the nanosleep() probe was broken (perl change #20061)
449 - use existence instead of definedness for feature probes
451 - MPE/iX tweak (perl change #20042)
452 - do not use HAS_NANOSLEEP (perl change #19898)
455 - UVuf for non-IVSIZE platforms (from Keiichiro Nagano)
456 - OS/2 can always mix subsecond sleeps with signals
457 (part of perl change #19789)
460 - workaround for buggy gcc 2.95.3 in openbsd/sparc64
464 - do not use -lrt in Linux (from March Lehmann, perl change #19449)
465 - unnecessary (nanosleep is in libc anyway)
466 - harmful (-lrt slows down execution)
467 - incompatible (with many distributions' pthreads)
470 - do not create files in blib directories under core
471 (perl change #19160, from rgs)
472 - detypo s/VTLARM/VTARLM/ (perl change #19328, from mjd)
475 - guarantee that $xdefine in HiRes.t is always defined
476 (perl change #19109, from IlyaZ)
477 - a cleaner way to detect PERL_CORE (perl change #19111,
481 - add hints/irix.pl to turn off overly POSIX flags that
482 cause hide struct timespec to be hidden (and compilation
483 to fail) (bleadperl change #19085)
484 - documentation tweaks
487 - add c:/temp to the list of temp directories to probe
488 so that cygwin (and win*?) builds are happy. This was
489 needed at least in my cygwin 1.3.20/w2k setup.
492 - modernize the constants code (from Nicholas Clark)
495 - At some point the ability to figure our the correct incdir
496 for EXTERN.h (either a core perl build, or an installed perl)
497 had broken (which lead into all test compiles failing with
498 a core perl build, but thanks to the robustness of Makefile.PL
499 nothing of this was visible). The brokenness seemed to be
500 caused by $ENV{PERL_CORE} not being on for core builds?
501 Now stole a trick from the Encode that sets $ENV{PERL_CORE}
502 right, and both styles of build should work again.
505 - Nicholas Clark noticed that the my_catdir() emulation function
506 was broken (which means that we didn't really work for Perls
508 - inspired by fixing the above made the whole Makefile.PL -w
510 - tightened up the Makefile.PL output, less whitespace
513 - fix from Craig Berry for better building in VMS with PERL_CORE
516 - no functional changes
517 - move lib/Time/HiRes.pm as Hires.pm
518 - libraries scanning was slightly broken (always scanned
519 for a library even when $Config{libs} already had it)
522 - Ray Zimmerman ran into a race condition in Mac OS X.
523 A 0.01-second alarm fired before the test expected.
524 The test first slept indefinitely (blocking for signals)
525 and only after that tested for the signal having been sent.
526 Since the signal had already been sent, the test #12 never
527 completed. The solution: test first, then block.
528 - default to being silent on all probing attempts, set the
529 environment variable VERBOSE to a true value to see the
530 details (the probing command and the possible errors)
533 - do not clear MAN3PODS in Makefile.PL (Radoslaw Zielinski)
534 - INSTALLDIRS => 'perl' missing which means that Time::HiRes
535 cannot be upgraded from CPAN to override the 5.8.0 version
537 - Time::HiRes 1.35 could not be dropped as-is to bleadperl
538 because the include directories did not adjust themselves
539 if $ENV{PERL_CORE} (Hugo van der Sanden)
540 - add documentation about the restart of select() under alarm()
543 - small documentation tweaks
547 - better VMS operation (Craig Berry)
550 - our time machine is accelerating: now works with Perl 5.004_01
551 (tried with 5.003_07 and 5.002 but I get segmentation faults
552 from running the Makefile.PL with those in Tru64 4.0D)
555 - backward compatibility (pre-5.6.0) tweaks:
556 - no XSLoader in 5.00503, use DynaLoader instead
557 - no SvPV_nolen, either
558 - no PerlProc_pause(), either
559 - now tested with 5.00404 and 5.00503
560 - Makefile.PL requires 5.00404 (no more 5.002)
561 - use nanosleep instead of usleep, if it is available (Wilson Snyder)
562 (this means that one can mix subsecond sleeps with alarms)
563 - because of nanosleep we probe for -lrt and -lposix4
564 - the existence of getitimer/nanosleep/setitimer/ualarm/usleep
565 is available by exportable constants Time::HiRes::d_func
566 (since older Perl do not have them in %Config, and even
567 5.8.0 does not probe for nanosleep)
570 - backward compatibility (pre-5.6.1) tweaks:
571 - define NV if no NVTYPE
572 - define IVdf if needed (note: the Devel::PPPort
573 in 5.8.0 does not try hard hard enough since
574 the IVSIZE might not be defined)
575 - define NVgf if needed
576 - grab the typemap from 5.8.0 for the NV stuff
578 1.31 and 1.32 add more backward compatibility (now all the way
579 back to Perl 5.00404), and using nanosleep() (if available) for
584 - release 1.29_02 as 1.30
586 1.30 adds all the changes made during the Perl 5.6->5.7->5.8
587 development cycle. Most notably portability across platforms has been
588 enhanced, and the interval timers (setitimer, getitimer) have been
589 added. Note that the version of Time::HiRes that is included in Perl
590 5.8.0 calls itself 1.20_00, but it is equivalent to this Time::HiRes
591 version. Note also that in 1.30 Wegscheid turns over the maintenance
592 to Jarkko Hietaniemi.
596 - fix a silly unclosed comment typo in HiRes.xs
597 - document and export REALTIME_REALPROF (Solaris)
601 - only getitimer(ITIMER_REAL) available in Cygwin and Win32
602 (need to patch this also in Perl 5.[89])
603 - remove CVS revision log from HiRes.xs
607 The following numbered patches refer to the Perl 5.7 changes,
608 you can browse them at http://public.activestate.com/cgi-bin/perlbrowse
610 - 17558: Add #!./perl to the .t
611 - 17201: linux + usemorebits fix, from Rafael Garcia-Suarez
612 - 16198: political correctness, from Simon Cozens
613 - 15857: doc tweaks, from Jarkko Hietaniemi
614 - 15593: optimization in .xs, from Paul Green
615 - 14892: pod fixes, from Robin Barker
616 - 14100: VOS fixes, from Paul Green
617 - 13422: XS segfault, from Marc Lehmann
618 - 13378: whether select() gets restarted on signals, depends
619 - 13354: timing constraints, again, from Andy Dougherty
620 - 13278: can't do subsecond alarms with ualarm;
621 break out early if alarms do not seem to be working
622 - 13266: test relaxation (cygwin gets lower hires
623 times than lores ones)
624 - 12846: protect against high load, from Jarkko Hietaniemi
625 - 12837: HiRes.t VMS tweak, from Craig A. Berry
626 - 12797: HiRes.t VMS tweak, from Charles Lane
627 - 12769: HiRes.t VMS tweak, from Craig A. Berry
628 - 12744: gcc vs MS 64-bit constant syntax, from Nick Ing-Simmons
629 - 12722: VMS ualarm for VMS without ualarm, from Charles Lane
630 - 12692: alarm() ain't gonna work if ualarm() ain't,
631 from Gurusamy Sarathy
632 - 12680: minor VMS tweak, from Charles Lane
633 - 12617: don't try to print ints as IVs, from Jarkko Hietaniemi
634 - 12609: croak on negative time, from Jarkko Hietaniemi
635 - 12595: Cygwin rounds up for time(), from Jarkko Hietaniemi
636 - 12594: MacOS Classic timeofday, from Chris Nandor
637 - 12473: allow for more than one second for sleep() and usleep()
638 - 12458: test tuning, relax timing constraints,
639 from Jarkko Hietaniemi
640 - 12449: make sleep() and usleep() to return the number
641 of seconds and microseconds actually slept (analogously
642 with the builtin sleep()), also make usleep() croak if
643 asked for more than 1_000_000 useconds, from Jarkko Hietaniemi
644 - 12366: Time::HiRes for VMS pre-7.0, from Charles Lane
645 - 12199: do not use ftime on Win32, from Gurusamy Sarathy
646 - 12196: use ftime() on Win32, from Artur Bergman
647 - 12184: fix Time::HiRes gettimeofday() on Win32, from Gurusamy Sarathy
648 - 12105: use GetSystemTime() on Win32, from Artur Bergman
649 - 12060: explain the 1e9 seconds problem, from Jarkko Hietaniemi
650 - 11901: UNICOS sloppy division, from Jarkko Hietaniemi
651 - 11797: problem in HiRes.t, from John P. Linderman
652 - 11414: prototype from Time::HiRes::sleep(), from Abhijit Menon-Sen
653 - 11409: Time::HiRes qw(sleep) failed, from Abhijit Menon-Sen
654 - 11270: dynix/ptx 4.5.2 hints fix, from Peter Prymmer
655 - 11032: VAX VMS s/div/lib\$ediv/ fix, from Peter Prymmer
656 - 11011: VAX VMS s/qdiv/div/ fix, from Peter Prymmer
657 - 10953: SCO OpenServer 5.0.5 requires an explicit -lc for usleep(),
659 - 10942: MPE/IX test tweaks, from Mark Bixby
660 - 10784: unnecessary pod2man calls, from Andy Dougherty
661 - 10354: ext/ + -Wall, from Doug MacEachern
662 - 10320: fix the BOOT section to call myU2time correctly
663 - 10317: correct casting for AIX< from H. Merijn Brand
664 - 10119: document that the core time() may be rounding, not truncating
665 - 10118: test fix, from John Peacock
666 - 9988: long =item, from Robin Barker
667 - 9714: correct test output
668 - 9708: test also the scalar aspect of getitimer()
669 - 9705: Add interval timers (setitimer, getitimer)
670 - 9692: do not require at least 5.005 using XS
672 The following changes were made on top of the changes
673 made for Time::HiRes during the Perl 5.7 development
674 cycle that culminated in the release of Perl 5.8.0.
676 - add "require 5.005" to the Makefile.PL
677 - remove the REVISION section (CVS log) from HiRes.pm
678 - add jhi's copyright alongside Douglas'
679 - move HiRes.pm to lib/Time/
681 - modify HiRes.t to use $ENV{PERL_CORE}
682 - modify the original Time::HiRes version 1.20 Makefile.PL
683 to work both with Perl 5.8.0 and the new code with pre-5.8.0
684 Perls (tried with 5.6.1)
685 - tiny tweaks and updates in README and TODO
686 - bump the VERSION to 1.29
688 1.20 Wed Feb 24 21:30 1999
689 - make our usleep and ualarm substitutes into hrt_usleep
690 and hrt_ualarm. This helps static links of Perl with other
691 packages that also have usleep, etc. From
692 Ilya Zakharevich <ilya@math.ohio-state.edu>
693 - add C API stuff. From Joshua Pritikin
694 <joshua.pritikin@db.com>
695 - VMS Makefile.PL fun. From pvhp@forte.com (Peter Prymmer)
696 - hopefully correct "-lc" fix for SCO.
699 1.20 adds a platform neutral set of C accessible routines if you are
700 running 5.005+. All other changes are packaging changes and build
701 fixes(?) for statically linked Perl, SCO, and VMS.
703 1.19 Tue Sep 29 22:30 1998
704 - put VMS gettimeofday() in. Patch is from Sebastian Bazley
705 <seb@stian.demon.co.uk>
706 - change GIMME_V to GIMME to help people with older versions of
708 - fix Win32 version of gettimeofday(). It didn't affect anything,
709 but it confuses people reading the code when the return value
710 is backwards (0 is success).
711 - fix Makefile.PL (more) so that detection of gettimeofday is
714 1.19 has better VMS support.
716 1.18 Mon Jul 6 22:40 1998
717 - add usleep() for Win32.
718 - fix Makefile.PL to fix reported HP/UX feature where unresolved
719 externals still cause an executable to be generated (though no
720 x bit set). Thanks to David Kozinn for report and explanation.
721 Problems with the fix are mine :)
723 1.18 has limited Win32 support (no ualarm). Added usleep for Win32.
724 Probably buggy. I'm sure I'll hear.
726 1.17 Wed Jul 1 20:10 1998
727 - fix setitimer calls so microseconds is not more than 1000000.
728 Hp/UX 9 doesn't like that. Provided by Roland B Robert, PhD.
729 - make Win32. We only get gettimeofday (the select hack doesn't
730 seem to work on my Win95 system).
731 - fix test 4 on 01test.t. add test to see if time() and
732 Time::HiRes::time() are close.
734 1.16 Wed Nov 12 21:05 1997
735 - add missing EXTEND in new gettimeofday scalar code.
737 1.16+ should be closer to building out of the box on Linux. Thanks
738 to Gisle Aas for patches, and the ualarm equivalent using setitimer.
740 If your underlying operating system doesn't implement ualarm(), then
741 a fake using setitimer() will be made. If the OS is missing usleep(),
742 a fake one using select() will be made. If a fake can't be made for
743 either ualarm() or usleep(), then the corresponding Perl function will
744 not be available. If the OS is missing gettimeofday(), you will get
745 unresolved externals, either at link- or run-time.
747 This is an improvement; the package used to not even build if
748 you were missing any of these bits. Roderick Schertler
750 <roderick@argon.org> did all the conditional compilation stuff,
751 look at HiRes.pm and the test suites; it's good educational reading.
753 1.15 Mon Nov 10 21:30 1997
754 - HiRes.pm: update pod. Provided by Gisle Aas.
755 - HiRes.xs: if gettimeofday() called in scalar context, do
756 something more useful than before. Provided by Gisle Aas.
757 - README: tell of xsubpp '-nolinenumber' woes. thanks to
758 Edward Henigin <ed@texas.net> for pointing out the problem.
760 1.14 Wed Nov 5 9:40 1997
761 - Makefile.PL: look for setitimer
762 - HiRes.xs: if missing ualarm, but we have setitimer, make up
763 our own setitimer. These were provided by Gisle Aas.
765 1.13 Tue Nov 4 23:30 1997
766 - Makefile.PL: fix autodetect mechanism to do try linking in addition
767 to just compiling; should fix Linux build problem. Fix was provided
770 1.12 Sun Oct 12 12:00:00 1997
771 - Makefile.PL: set XSOPT to '-nolinenumbers' to work around xsubpp bug;
772 you may need to comment this back out if you have an older xsubpp.
773 - HiRes.xs: set PROTOTYPES: DISABLE
775 1.11 Fri Sep 05 16:00:00 1997
777 Had some line commented out that shouldn't have been (testing
780 Previous version was corrupted.
782 1.10 Thu May 22 20:20:00 1997
783 - HiRes.xs, HiRes.pm, t/*:
784 - only compile what we have OS support for (or can
786 - only test what we compiled
787 - gross improvement to the test suite
789 This work was all done by Roderick Schertler
790 <roderick@argon.org>. If you run Linux or
791 one of the other ualarm-less platforms, and you like this
792 module, let Roderick know; without him, it still wouldn't
793 be working on those boxes...
794 - Makefile.PL: figure out what routines the OS has and
795 only build what we need. These bits were written by Jarkko
796 Hietaniemi <jhi@iki.fi>. Again, gratitude is due...
798 1.02 Mon Dec 30 08:00:00 1996
799 - HiRes.pm: update documentation to say what to do when missing
800 ualarm() and friends.
801 - README: update to warn that ualarm() and friends need to exist
803 1.01 Fri Oct 17 08:00:00 1996
804 - Makefile.PL: make XSPROTOARGS => '-noprototyopes'
805 - HiRes.pm: put blank line between __END__ and =head1 so that
808 1.00 Tue Sep 03 13:00:00 1996
809 - original version; created by h2xs 1.16