1 Revision history for Perl extension Time::HiRes.
4 - add c:/temp to the list of temp directories to probe
5 so that cygwin (and win*?) builds are happy. This was
6 needed at least in my cygwin 1.3.20/w2k setup.
9 - modernize the constants code (from Nicholas Clark)
12 - At some point the ability to figure our the correct incdir
13 for EXTERN.h (either a core perl build, or an installed perl)
14 had broken (which lead into all test compiles failing with
15 a core perl build, but thanks to the robustness of Makefile.PL
16 nothing of this was visible). The brokenness seemed to be
17 caused by $ENV{PERL_CORE} not being on for core builds?
18 Now stole a trick from the Encode that sets $ENV{PERL_CORE}
19 right, and both styles of build should work again.
22 - Nicholas Clark noticed that the my_catdir() emulation function
23 was broken (which means that we didn't really work for Perls
25 - inspired by fixing the above made the whole Makefile.PL -w
27 - tightened up the Makefile.PL output, less whitespace
30 - fix from Craig Berry for better building in VMS with PERL_CORE
33 - no functional changes
34 - move lib/Time/HiRes.pm as Hires.pm
35 - libraries scanning was slightly broken (always scanned
36 for a library even when $Config{libs} already had it)
39 - Ray Zimmerman ran into a race condition in Mac OS X.
40 A 0.01-second alarm fired before the test expected.
41 The test first slept indefinitely (blocking for signals)
42 and only after that tested for the signal having been sent.
43 Since the signal had already been sent, the test #12 never
44 completed. The solution: test first, then block.
45 - default to being silent on all probing attempts, set the
46 environment variable VERBOSE to a true value to see the
47 details (the probing command and the possible errors)
50 - do not clear MAN3PODS in Makefile.PL (Radoslaw Zielinski)
51 - INSTALLDIRS => 'perl' missing which means that Time::HiRes
52 cannot be upgraded from CPAN to override the 5.8.0 version
54 - Time::HiRes 1.35 could not be dropped as-is to bleadperl
55 because the include directories did not adjust themselves
56 if $ENV{PERL_CORE} (Hugo van der Sanden)
57 - add documentation about the restart of select() under alarm()
60 - small documentation tweaks
64 - better VMS operation (Craig Berry)
67 - our time machine is accelerating: now works with Perl 5.004_01
68 (tried with 5.003_07 and 5.002 but I get segmentation faults
69 from running the Makefile.PL with those in Tru64 4.0D)
72 - backward compatibility (pre-5.6.0) tweaks:
73 - no XSLoader in 5.00503, use DynaLoader instead
74 - no SvPV_nolen, either
75 - no PerlProc_pause(), either
76 - now tested with 5.00404 and 5.00503
77 - Makefile.PL requires 5.00404 (no more 5.002)
78 - use nanosleep instead of usleep, if it is available (Wilson Snyder)
79 (this means that one can mix subsecond sleeps with alarms)
80 - because of nanosleep we probe for -lrt and -lposix4
81 - the existence of getitimer/nanosleep/setitimer/ualarm/usleep
82 is available by exportable constants Time::HiRes::d_func
83 (since older Perl do not have them in %Config, and even
84 5.8.0 does not probe for nanosleep)
87 - backward compatibility (pre-5.6.1) tweaks:
88 - define NV if no NVTYPE
89 - define IVdf if needed (note: the Devel::PPPort
90 in 5.8.0 does not try hard hard enough since
91 the IVSIZE might not be defined)
92 - define NVgf if needed
93 - grab the typemap from 5.8.0 for the NV stuff
96 - release 1.29_02 as 1.30
100 - fix a silly unclosed comment typo in HiRes.xs
101 - document and export REALTIME_REALPROF (Solaris)
105 - only getitimer(ITIMER_REAL) available in Cygwin and Win32
106 (need to patch this also in Perl 5.[89])
107 - remove CVS revision log from HiRes.xs
111 The following numbered patches refer to the Perl 5.7 changes,
112 you can browse them at http://public.activestate.com/cgi-bin/perlbrowse
114 - 17558: Add #!./perl to the .t
115 - 17201: linux + usemorebits fix, from Rafael Garcia-Suarez
116 - 16198: political correctness, from Simon Cozens
117 - 15857: doc tweaks, from Jarkko Hietaniemi
118 - 15593: optimization in .xs, from Paul Green
119 - 14892: pod fixes, from Robin Barker
120 - 14100: VOS fixes, from Paul Green
121 - 13422: XS segfault, from Marc Lehmann
122 - 13378: whether select() gets restarted on signals, depends
123 - 13354: timing constraints, again, from Andy Dougherty
124 - 13278: can't do subecond alarms with ualarm;
125 break out early if alarms do not seem to be working
126 - 13266: test relaxation (cygwin gets lower hires
127 times than lores ones)
128 - 12846: protect against high load, from Jarkko Hietaniemi
129 - 12837: HiRes.t VMS tweak, from Craig A. Berry
130 - 12797: HiRes.t VMS tweak, from Charles Lane
131 - 12769: HiRes.t VMS tweak, from Craig A. Berry
132 - 12744: gcc vs MS 64-bit constant syntax, from Nick Ing-Simmons
133 - 12722: VMS ualarm for VMS without ualarm, from Charles Lane
134 - 12692: alarm() ain't gonna work if ualarm() ain't,
135 from Gurusamy Sarathy
136 - 12680: minor VMS tweak, from Charles Lane
137 - 12617: don't try to print ints as IVs, from Jarkko Hietaniemi
138 - 12609: croak on negative time, from Jarkko Hietaniemi
139 - 12595: Cygwin rounds up for time(), from Jarkko Hietaniemi
140 - 12594: MacOS Classic timeofday, from Chris Nandor
141 - 12473: allow for more than one second for sleep() and usleep()
142 - 12458: test tuning, relax timing constraints,
143 from Jarkko Hietaniemi
144 - 12449: make sleep() and usleep() to return the number
145 of seconds and microseconds actually slept (analogously
146 with the builtin sleep()), also make usleep() croak if
147 asked for more than 1_000_000 useconds, from Jarkko Hietaniemi
148 - 12366: Time::HiRes for VMS pre-7.0, from Charles Lane
149 - 12199: do not use ftime on Win32, from Gurusamy Sarathy
150 - 12196: use ftime() on Win32, from Artur Bergman
151 - 12184: fix Time::HiRes gettimeofday() on Win32, from Gurusamy Sarathy
152 - 12105: use GetSystemTime() on Win32, from Artur Bergman
153 - 12060: explain the 1e9 seconds problem, from Jarkko Hietaniemi
154 - 11901: UNICOS sloppy division, from Jarkko Hietaniemi
155 - 11797: problem in HiRes.t, from John P. Linderman
156 - 11414: prototype from Time::HiRes::sleep(), from Abhijit Menon-Sen
157 - 11409: Time::HiRes qw(sleep) failed, from Abhijit Menon-Sen
158 - 11270: dynix/ptx 4.5.2 hints fix, from Peter Prymmer
159 - 11032: VAX VMS s/div/lib\$ediv/ fix, from Peter Prymmer
160 - 11011: VAX VMS s/qdiv/div/ fix, from Peter Prymmer
161 - 10953: SCO OpenServer 5.0.5 requires an explicit -lc for usleep(),
163 - 10942: MPE/IX test tweaks, from Mark Bixby
164 - 10784: unnecessary pod2man calls, from Andy Dougherty
165 - 10354: ext/ + -Wall, from Doug MacEachern
166 - 10320: fix the BOOT section to call myU2time correctly
167 - 10317: correct casting for AIX< from H. Merijn Brand
168 - 10119: document that the core time() may be rounding, not truncating
169 - 10118: test fix, from John Peacock
170 - 9988: long =item, from Robin Barker
171 - 9714: correct test output
172 - 9708: test also the scalar aspect of getitimer()
173 - 9705: Add interval timers (setitimer, getitimer)
174 - 9692: do not require at least 5.005 using XS
176 The following changes were made on top of the changes
177 made for Time::HiRes during the Perl 5.7 development
178 cycle that culminated in the release of Perl 5.8.0.
180 - add "require 5.005" to the Makefile.PL
181 - remove the REVISION section (CVS log) from HiRes.pm
182 - add jhi's copyright alongside Douglas'
183 - move HiRes.pm to lib/Time/
185 - modify HiRes.t to use $ENV{PERL_CORE}
186 - modify the original Time::HiRes version 1.20 Makefile.PL
187 to work both with Perl 5.8.0 and the new code with pre-5.8.0
188 Perls (tried with 5.6.1)
189 - tiny tweaks and updates in README and TODO
190 - bump the VERSION to 1.29
192 1.20 Wed Feb 24 21:30 1999
193 - make our usleep and ualarm substitutes into hrt_usleep
194 and hrt_ualarm. This helps static links of Perl with other
195 packages that also have usleep, etc. From
196 Ilya Zakharevich <ilya@math.ohio-state.edu>
197 - add C API stuff. From Joshua Pritikin
198 <joshua.pritikin@db.com>
199 - VMS Makefile.PL fun. From pvhp@forte.com (Peter Prymmer)
200 - hopefully correct "-lc" fix for SCO.
203 1.19 Tue Sep 29 22:30 1998
204 - put VMS gettimeofday() in. Patch is from Sebastian Bazley
205 <seb@stian.demon.co.uk>
206 - change GIMME_V to GIMME to help people with older versions of
208 - fix Win32 version of gettimeofday(). It didn't affect anything,
209 but it confuses people reading the code when the return value
210 is backwards (0 is success).
211 - fix Makefile.PL (more) so that detection of gettimeofday is
214 1.18 Mon Jul 6 22:40 1998
215 - add usleep() for Win32.
216 - fix Makefile.PL to fix reported HP/UX feature where unresolved
217 externals still cause an executable to be generated (though no
218 x bit set). Thanks to David Kozinn for report and explanation.
219 Problems with the fix are mine :)
221 1.17 Wed Jul 1 20:10 1998
222 - fix setitimer calls so microseconds is not more than 1000000.
223 Hp/UX 9 doesn't like that. Provided by Roland B Robert, PhD.
224 - make Win32. We only get gettimeofday (the select hack doesn't
225 seem to work on my Win95 system).
226 - fix test 4 on 01test.t. add test to see if time() and
227 Time::HiRes::time() are close.
229 1.16 Wed Nov 12 21:05 1997
230 - add missing EXTEND in new gettimeofday scalar code.
232 1.15 Mon Nov 10 21:30 1997
233 - HiRes.pm: update pod. Provided by Gisle Aas.
234 - HiRes.xs: if gettimeofday() called in scalar context, do
235 something more useful than before. Provided by Gisle Aas.
236 - README: tell of xsubpp '-nolinenumber' woes. thanks to
237 Edward Henigin <ed@texas.net> for pointing out the problem.
239 1.14 Wed Nov 5 9:40 1997
240 - Makefile.PL: look for setitimer
241 - HiRes.xs: if missing ualarm, but we have setitimer, make up
242 our own setitimer. These were provided by Gisle Aas.
244 1.13 Tue Nov 4 23:30 1997
245 - Makefile.PL: fix autodetect mechanism to do try linking in addition
246 to just compiling; should fix Linux build problem. Fix was provided
249 1.12 Sun Oct 12 12:00:00 1997
250 - Makefile.PL: set XSOPT to '-nolinenumbers' to work around xsubpp bug;
251 you may need to comment this back out if you have an older xsubpp.
252 - HiRes.xs: set PROTOTYPES: DISABLE
254 1.11 Fri Sep 05 16:00:00 1997
256 Had some line commented out that shouldn't have been (testing
259 Previous version was corrupted.
261 1.10 Thu May 22 20:20:00 1997
262 - HiRes.xs, HiRes.pm, t/*:
263 - only compile what we have OS support for (or can
265 - only test what we compiled
266 - gross improvement to the test suite
268 This work was all done by Roderick Schertler
269 <roderick@argon.org>. If you run Linux or
270 one of the other ualarm-less platoforms, and you like this
271 module, let Roderick know; without him, it still wouldn't
272 be working on those boxes...
273 - Makefile.PL: figure out what routines the OS has and
274 only build what we need. These bits were written by Jarkko
275 Hietaniemi <jhi@iki.fi>. Again, gratitude is due...
277 1.02 Mon Dec 30 08:00:00 1996
278 - HiRes.pm: update documentation to say what to do when missing
279 ualarm() and friends.
280 - README: update to warn that ualarm() and friends need to exist
282 1.01 Fri Oct 17 08:00:00 1996
283 - Makefile.PL: make XSPROTOARGS => '-noprototyopes'
284 - HiRes.pm: put blank line between __END__ and =head1 so that
287 1.00 Tue Sep 03 13:00:00 1996
288 - original version; created by h2xs 1.16