1 Revision history for Perl extension Time::HiRes.
4 - modernize the constants code (from Nicholas Clark)
7 - At some point the ability to figure our the correct incdir
8 for EXTERN.h (either a core perl build, or an installed perl)
9 had broken (which lead into all test compiles failing with
10 a core perl build, but thanks to the robustness of Makefile.PL
11 nothing of was visible). The brokenness seemed to be caused
12 by $ENV{PERL_CORE} not being on for core builds? Now stole
13 a trick from the Encode that sets $ENV{PERL_CORE} right, and
14 both styles of build should work again.
17 - Nicholas Clark noticed that the my_catdir() emulation function
18 was broken (which means that we didn't really work for Perls
20 - inspired by fixing the above made the whole Makefile.PL -w
22 - tightened up the Makefile.PL output, less whitespace
25 - fix from Craig Berry for better building in VMS with PERL_CORE
28 - no functional changes
29 - move lib/Time/HiRes.pm as Hires.pm
30 - libraries scanning was slightly broken (always scanned
31 for a library even when $Config{libs} already had it)
34 - Ray Zimmerman ran into a race condition in Mac OS X.
35 A 0.01-second alarm fired before the test expected.
36 The test first slept indefinitely (blocking for signals)
37 and only after that tested for the signal having been sent.
38 Since the signal had already been sent, the test #12 never
39 completed. The solution: test first, then block.
40 - default to being silent on all probing attempts, set the
41 environment variable VERBOSE to a true value to see the
42 details (the probing command and the possible errors)
45 - do not clear MAN3PODS in Makefile.PL (Radoslaw Zielinski)
46 - INSTALLDIRS => 'perl' missing which means that Time::HiRes
47 cannot be upgraded from CPAN to override the 5.8.0 version
49 - Time::HiRes 1.35 could not be dropped as-is to bleadperl
50 because the include directories did not adjust themselves
51 if $ENV{PERL_CORE} (Hugo van der Sanden)
52 - add documentation about the restart of select() under alarm()
55 - small documentation tweaks
59 - better VMS operation (Craig Berry)
62 - our time machine is accelerating: now works with Perl 5.004_01
63 (tried with 5.003_07 and 5.002 but I get segmentation faults
64 from running the Makefile.PL with those in Tru64 4.0D)
67 - backward compatibility (pre-5.6.0) tweaks:
68 - no XSLoader in 5.00503, use DynaLoader instead
69 - no SvPV_nolen, either
70 - no PerlProc_pause(), either
71 - now tested with 5.00404 and 5.00503
72 - Makefile.PL requires 5.00404 (no more 5.002)
73 - use nanosleep instead of usleep, if it is available (Wilson Snyder)
74 (this means that one can mix subsecond sleeps with alarms)
75 - because of nanosleep we probe for -lrt and -lposix4
76 - the existence of getitimer/nanosleep/setitimer/ualarm/usleep
77 is available by exportable constants Time::HiRes::d_func
78 (since older Perl do not have them in %Config, and even
79 5.8.0 does not probe for nanosleep)
82 - backward compatibility (pre-5.6.1) tweaks:
83 - define NV if no NVTYPE
84 - define IVdf if needed (note: the Devel::PPPort
85 in 5.8.0 does not try hard hard enough since
86 the IVSIZE might not be defined)
87 - define NVgf if needed
88 - grab the typemap from 5.8.0 for the NV stuff
91 - release 1.29_02 as 1.30
95 - fix a silly unclosed comment typo in HiRes.xs
96 - document and export REALTIME_REALPROF (Solaris)
100 - only getitimer(ITIMER_REAL) available in Cygwin and Win32
101 (need to patch this also in Perl 5.[89])
102 - remove CVS revision log from HiRes.xs
106 The following numbered patches refer to the Perl 5.7 changes,
107 you can browse them at http://public.activestate.com/cgi-bin/perlbrowse
109 - 17558: Add #!./perl to the .t
110 - 17201: linux + usemorebits fix, from Rafael Garcia-Suarez
111 - 16198: political correctness, from Simon Cozens
112 - 15857: doc tweaks, from Jarkko Hietaniemi
113 - 15593: optimization in .xs, from Paul Green
114 - 14892: pod fixes, from Robin Barker
115 - 14100: VOS fixes, from Paul Green
116 - 13422: XS segfault, from Marc Lehmann
117 - 13378: whether select() gets restarted on signals, depends
118 - 13354: timing constraints, again, from Andy Dougherty
119 - 13278: can't do subecond alarms with ualarm;
120 break out early if alarms do not seem to be working
121 - 13266: test relaxation (cygwin gets lower hires
122 times than lores ones)
123 - 12846: protect against high load, from Jarkko Hietaniemi
124 - 12837: HiRes.t VMS tweak, from Craig A. Berry
125 - 12797: HiRes.t VMS tweak, from Charles Lane
126 - 12769: HiRes.t VMS tweak, from Craig A. Berry
127 - 12744: gcc vs MS 64-bit constant syntax, from Nick Ing-Simmons
128 - 12722: VMS ualarm for VMS without ualarm, from Charles Lane
129 - 12692: alarm() ain't gonna work if ualarm() ain't,
130 from Gurusamy Sarathy
131 - 12680: minor VMS tweak, from Charles Lane
132 - 12617: don't try to print ints as IVs, from Jarkko Hietaniemi
133 - 12609: croak on negative time, from Jarkko Hietaniemi
134 - 12595: Cygwin rounds up for time(), from Jarkko Hietaniemi
135 - 12594: MacOS Classic timeofday, from Chris Nandor
136 - 12473: allow for more than one second for sleep() and usleep()
137 - 12458: test tuning, relax timing constraints,
138 from Jarkko Hietaniemi
139 - 12449: make sleep() and usleep() to return the number
140 of seconds and microseconds actually slept (analogously
141 with the builtin sleep()), also make usleep() croak if
142 asked for more than 1_000_000 useconds, from Jarkko Hietaniemi
143 - 12366: Time::HiRes for VMS pre-7.0, from Charles Lane
144 - 12199: do not use ftime on Win32, from Gurusamy Sarathy
145 - 12196: use ftime() on Win32, from Artur Bergman
146 - 12184: fix Time::HiRes gettimeofday() on Win32, from Gurusamy Sarathy
147 - 12105: use GetSystemTime() on Win32, from Artur Bergman
148 - 12060: explain the 1e9 seconds problem, from Jarkko Hietaniemi
149 - 11901: UNICOS sloppy division, from Jarkko Hietaniemi
150 - 11797: problem in HiRes.t, from John P. Linderman
151 - 11414: prototype from Time::HiRes::sleep(), from Abhijit Menon-Sen
152 - 11409: Time::HiRes qw(sleep) failed, from Abhijit Menon-Sen
153 - 11270: dynix/ptx 4.5.2 hints fix, from Peter Prymmer
154 - 11032: VAX VMS s/div/lib\$ediv/ fix, from Peter Prymmer
155 - 11011: VAX VMS s/qdiv/div/ fix, from Peter Prymmer
156 - 10953: SCO OpenServer 5.0.5 requires an explicit -lc for usleep(),
158 - 10942: MPE/IX test tweaks, from Mark Bixby
159 - 10784: unnecessary pod2man calls, from Andy Dougherty
160 - 10354: ext/ + -Wall, from Doug MacEachern
161 - 10320: fix the BOOT section to call myU2time correctly
162 - 10317: correct casting for AIX< from H. Merijn Brand
163 - 10119: document that the core time() may be rounding, not truncating
164 - 10118: test fix, from John Peacock
165 - 9988: long =item, from Robin Barker
166 - 9714: correct test output
167 - 9708: test also the scalar aspect of getitimer()
168 - 9705: Add interval timers (setitimer, getitimer)
169 - 9692: do not require at least 5.005 using XS
171 The following changes were made on top of the changes
172 made for Time::HiRes during the Perl 5.7 development
173 cycle that culminated in the release of Perl 5.8.0.
175 - add "require 5.005" to the Makefile.PL
176 - remove the REVISION section (CVS log) from HiRes.pm
177 - add jhi's copyright alongside Douglas'
178 - move HiRes.pm to lib/Time/
180 - modify HiRes.t to use $ENV{PERL_CORE}
181 - modify the original Time::HiRes version 1.20 Makefile.PL
182 to work both with Perl 5.8.0 and the new code with pre-5.8.0
183 Perls (tried with 5.6.1)
184 - tiny tweaks and updates in README and TODO
185 - bump the VERSION to 1.29
187 1.20 Wed Feb 24 21:30 1999
188 - make our usleep and ualarm substitutes into hrt_usleep
189 and hrt_ualarm. This helps static links of Perl with other
190 packages that also have usleep, etc. From
191 Ilya Zakharevich <ilya@math.ohio-state.edu>
192 - add C API stuff. From Joshua Pritikin
193 <joshua.pritikin@db.com>
194 - VMS Makefile.PL fun. From pvhp@forte.com (Peter Prymmer)
195 - hopefully correct "-lc" fix for SCO.
198 1.19 Tue Sep 29 22:30 1998
199 - put VMS gettimeofday() in. Patch is from Sebastian Bazley
200 <seb@stian.demon.co.uk>
201 - change GIMME_V to GIMME to help people with older versions of
203 - fix Win32 version of gettimeofday(). It didn't affect anything,
204 but it confuses people reading the code when the return value
205 is backwards (0 is success).
206 - fix Makefile.PL (more) so that detection of gettimeofday is
209 1.18 Mon Jul 6 22:40 1998
210 - add usleep() for Win32.
211 - fix Makefile.PL to fix reported HP/UX feature where unresolved
212 externals still cause an executable to be generated (though no
213 x bit set). Thanks to David Kozinn for report and explanation.
214 Problems with the fix are mine :)
216 1.17 Wed Jul 1 20:10 1998
217 - fix setitimer calls so microseconds is not more than 1000000.
218 Hp/UX 9 doesn't like that. Provided by Roland B Robert, PhD.
219 - make Win32. We only get gettimeofday (the select hack doesn't
220 seem to work on my Win95 system).
221 - fix test 4 on 01test.t. add test to see if time() and
222 Time::HiRes::time() are close.
224 1.16 Wed Nov 12 21:05 1997
225 - add missing EXTEND in new gettimeofday scalar code.
227 1.15 Mon Nov 10 21:30 1997
228 - HiRes.pm: update pod. Provided by Gisle Aas.
229 - HiRes.xs: if gettimeofday() called in scalar context, do
230 something more useful than before. Provided by Gisle Aas.
231 - README: tell of xsubpp '-nolinenumber' woes. thanks to
232 Edward Henigin <ed@texas.net> for pointing out the problem.
234 1.14 Wed Nov 5 9:40 1997
235 - Makefile.PL: look for setitimer
236 - HiRes.xs: if missing ualarm, but we have setitimer, make up
237 our own setitimer. These were provided by Gisle Aas.
239 1.13 Tue Nov 4 23:30 1997
240 - Makefile.PL: fix autodetect mechanism to do try linking in addition
241 to just compiling; should fix Linux build problem. Fix was provided
244 1.12 Sun Oct 12 12:00:00 1997
245 - Makefile.PL: set XSOPT to '-nolinenumbers' to work around xsubpp bug;
246 you may need to comment this back out if you have an older xsubpp.
247 - HiRes.xs: set PROTOTYPES: DISABLE
249 1.11 Fri Sep 05 16:00:00 1997
251 Had some line commented out that shouldn't have been (testing
254 Previous version was corrupted.
256 1.10 Thu May 22 20:20:00 1997
257 - HiRes.xs, HiRes.pm, t/*:
258 - only compile what we have OS support for (or can
260 - only test what we compiled
261 - gross improvement to the test suite
263 This work was all done by Roderick Schertler
264 <roderick@argon.org>. If you run Linux or
265 one of the other ualarm-less platoforms, and you like this
266 module, let Roderick know; without him, it still wouldn't
267 be working on those boxes...
268 - Makefile.PL: figure out what routines the OS has and
269 only build what we need. These bits were written by Jarkko
270 Hietaniemi <jhi@iki.fi>. Again, gratitude is due...
272 1.02 Mon Dec 30 08:00:00 1996
273 - HiRes.pm: update documentation to say what to do when missing
274 ualarm() and friends.
275 - README: update to warn that ualarm() and friends need to exist
277 1.01 Fri Oct 17 08:00:00 1996
278 - Makefile.PL: make XSPROTOARGS => '-noprototyopes'
279 - HiRes.pm: put blank line between __END__ and =head1 so that
282 1.00 Tue Sep 03 13:00:00 1996
283 - original version; created by h2xs 1.16