ViewVC Help
View File | Revision Log | Show Annotations | Download File
/cvs/Coro/Changes
(Generate patch)

Comparing Coro/Changes (file contents):
Revision 1.45 by root, Wed Sep 12 16:45:59 2001 UTC vs.
Revision 1.249 by root, Tue Oct 2 23:16:23 2007 UTC

1Revision history for Perl extension Coro. 1Revision history for Perl extension Coro.
2 2
30.49 3TODO: destroy gets only called on cancel: fix(?)
4TODO: should explore PerlIO::via::CoroCede and PerlIO::via::CoroAIO.
5TODO: maybe implement a default message channel, very much like Erlang
6 (which is cool in a lot of important aspects (failures!),
7 but very lacking in others (higher level ipc)).
8
93.8
10 - incompatibly changed Coro::Storable::freeze.
11 - major new feature: added Coro::Debug, for interactive coroutine
12 debugging, tracing and much more.
13 - added Coro::Semaphore::adjust.
14 - added Coro::Util::fork_eval.
15 - added Coro::Storable::{nfreeze,blocking_{freeze,nfreeze}}.
16 - added Coro::killall.
17 - reduce initial stack sizes to allow for "micro-coroutines".
18 - better async_pool resource management, moved parts of async_pool
19 handling to XS (major speed improvement).
20 - actually croak before modifying important data structures.
21 - refuse to transfer while compiling.
22 - possibly support eval EXPR better now.
23 - enable assembly per default on linux+bsd x86+amd64.
24 - all internal members are named _something for easier subclassing.
25 - many minor tweaks.
26
273.63 Wed May 16 14:10:06 CEST 2007
28 - implement handcoded assembly for x86/amd64 SVR ABI.
29
303.62 Fri Apr 27 21:36:06 CEST 2007
31 - upgrade libcoro (which might set unwind info correctly).
32 - change default on linux to setjmp/longjmp.
33
343.61 Thu Apr 19 12:36:18 CEST 2007
35 - Coro::Storable caused an endless loop when thawing invalid
36 pst's sometimes.
37 - use a Semaphore in Coro::Storable, as Storable doesn't
38 seem to be reentrant (although it is documented to
39 be threadsafe...).
40 - fix Coro::Signal to bring back the original unreliable
41 but stateful semantics.
42 - fixed a lot of typos in Coro.pm (patch submitted by David
43 Steinbrunner, which applied flawlessly).
44
453.6 Sat Apr 14 17:13:31 CEST 2007
46 - added some bugfixes to get eg/myhttpd working again.
47 - added Coro::Storable for often-cede'ing freeze/thaw.
48 - try to do a clean exit when a coroutine calls exit
49 (EXPERIMENTAL).
50 - got rid of indirect call through _coro_init.
51 - updated the partly antique examples in eg/ to
52 work again and be a bit less magic, too.
53 - fixed Coro::Signal semantics to work as documented again.
54
553.55 Sun Mar 25 01:20:47 CET 2007
56 - add SAVE_DEFFH to save the default filehandle and enable
57 it by default.
58 - finally move socket-operations from Coro::Socket to Coro::Handle
59 to be able to unblock foreign sockets.
60 - add Coro::State::save_also and guarded_save.
61 - add count accessor to Coro::Semaphore.
62 - add Coro::State::cctx_stacksize.
63 - just for the fun of it, do not rely on implicit context,
64 which can dramatically improve performance, but people
65 using windows-process-emulation perls do not care much about
66 performance.
67
683.51 Sun Mar 4 14:18:04 CET 2007
69 - fixed a problem when you weakened references to Coro::State's
70 (patch by Yuval Kogman).
71
723.501 Wed Feb 28 12:44:07 CET 2007
73 - rename some global symbols as macosx from hell redefines
74 them without asking.
75
763.5 Tue Feb 13 20:22:53 CET 2007
77 - do AnyEvent model detection earlier, avoiding problems
78 caused by first using AnyEvent and later Coro::Event.
79 - implement and document Coro::Event event objects.
80 - fix a potential problem in Coro::Event causing crashes.
81 - initialise PL_comppad when creating a new coroutine,
82 avoids crashes on early coro destruction.
83
843.41 Mon Jan 22 19:19:49 CET 2007
85 - readline on Coro::Handle did not support undefined $/,
86 nor did it deliver partial lines on EOF or error.
87 - implement malloc fallback for stack allocation because
88 stupid broken idiotic OSX has a stupid broken
89 idiotic fits-the-whole-os mmap "implementation" and
90 my dick feels longer if Coro is portable even to
91 obsolete platforms.
92
933.4 Fri Jan 19 21:52:54 CET 2007
94 - remove t/09_timer.t, as it isn't really testing much
95 but was rather flaky in practise.
96 - async_pool coro would keep arguments and callback alive until
97 it was reused.
98 - cancellation of a coroutine could cause spurious idle calls
99 in cede_notself.
100
1013.3 Sat Jan 6 03:45:00 CET 2007
102 - implement $coro->on_destroy.
103 - Coro::Event blocking semantics have been changed,
104 documented and - hopefully - improved.
105 - fix nice adding, not subtracting, from priority.
106 - fix ->prio and api_is_ready (patch by Mark Hinds).
107 - fixed an assert ("... == prev__cctx->idle_te")
108 that could errronously trigger.
109 - fix various large and small memleaks.
110 - use a (hopefully) more stable cancel implementation
111 that immediately frees the coroutine data.
112 - cede/cede_notself return a status now.
113 - added Coro::guard function.
114 - added a global coroutine pool for jobs (on my machine,
115 I can create and execute 48k simple coros/s with async,
116 and 128k coros with async_pool).
117 - Coro::AIO now uses the coroutine priority as io priority.
118
1193.2 Fri Dec 22 05:07:09 CET 2006
120 - improve portability to slightly older perls.
121 - use cleaner coroutine destruction.
122 - simplify configuration for users.
123 - optionally (unrecommended) prefer perl functions over
124 their coro replacements.
125
1263.11 Tue Dec 5 13:11:24 CET 2006
127 - fixed some bogus assert's, but as perl.h disables assert even
128 without NDEBUG (thank you very much), not too many people should
129 notice (that did include myself). Andreas König noticed, though :)
130 - do not save/restore PL_sortcxix on >= 5.9.x, it doesn't seem to have
131 it. Also noticed by Andreas König :)
132 - save/restore tainted status.
133 - verified to pass the testsuite on my 5.9.5.
134
1353.1 Mon Dec 4 23:03:40 CET 2006
136 - INCOMPATIBLE CHANGE: $/ is now per-coroutine (but slow).
137 - incompatible change: transfer flags are now per-state.
138 - give a better error message on deadlock.
139 - document Coro::nready.
140 - enhanced testsuite.
141
1423.01 Sun Dec 3 23:47:42 CET 2006
143 - forgot to include Coro::Timer.
144
1453.0 Sun Dec 3 22:57:25 CET 2006
146 - the "FINALLY COMPLETELY STABLE" release (coming soon:
147 the "FAMOUS LAST WORDS" release).
148 - implement a new stack sharing algorithm, which uses a stack
149 pool (size currently hardcoded).
150 - make stack sharing mandatory (it no longer uses a heuristic).
151 - eval/die no longer cause weird problems under heavy use.
152 - Coro::Event could cause livelocks if it was used but
153 no Coro::Event watchers were used.
154 - Coro::Event now uses asynccheck as prepare does not
155 check for changed watchers.
156 - Coro::Event allows multiple waiting coros and will wake up one
157 per event.
158 - Coro::Event should be cleaner and more efficient now.
159 - new utility function Coro::unblock_sub.
160 - document the sad fact that Event is no longer reentrant.
161 - putting a coroutine into the ready queue twice could under
162 some circumstances lead to stack corruption.
163 - minor incompatibility: subclassing Coro::State is supported
164 directly now without going through a _coro_state member.
165 - state/coro switching is much faster now.
166 - very minor optimisations and code/documentation cleanup.
167 - avoid problems due to compiler inlining.
168 - removed timers from Coro::Timer -> use AnyEvent instead.
169 - replaced Coro::idle coroutine by (cleaner) idle handler.
170 - updated to newest libcoro.
171 - implement enhanced support for valgrind.
172 - implement is_ready and return value for ready.
173 - removed Coro::Cont, it was a misnomer (it's generators),
174 and it was somewhat annoying to get it right. Will come back
175 if somebody asks for it :)
176 - many ->wait methods and Coro::Event could return
177 spuriously without the event having happened.
178
1792.5 Tue Nov 7 12:22:33 CET 2006
180 - made Coro::Util, Coro::Select, Coro::Handle and Coro::Socket
181 use AnyEvent, moved them to Coro/.
182 - added Coro::LWP which contains all the uglyness required to
183 make LWP non-blocking.
184 - should work with perl 5.9.x now (Andreas König made me do it).
185 - fixed another bug in Coro::Select when the timeout was undef.
186 - reuse PL_start_env for all coros, saves some memory per coroutine.
187 - manage PL_top_env differently, hopefully to avoid panic: top_env.
188 - timeout argument was not properly used in Coro::Socket.
189 - allow limited forms of subclassing in Coro::Handle/Coro::Socket.
190 - emulate undocumented(!) functionality of IO::Socket required
191 by LWP(!!).
192 - updated eg/lwp to work with newer lwp's.
193 - remove "FATAL: uncaught exception" prefix. Coroutines that die
194 kill the whole process, just as exceptions in the main "coroutine"
195 did already.
196
1972.1 Wed Nov 1 23:01:13 CET 2006
198 - fix a long-standing bug in Coro::Select where select with
199 zero timeout would instead change the current default filehandle.
200 - use a simpler and hopefully more robust way to clone padlists
201 (uses less memory and a perl function instead of our own).
202 - coro can now create a stack guard area on many architectures.
203 - Coro::AIO properly reexports additional functions from IO::AIO.
204 - updated libcoro with a workaround for OS X,
205 pach and testing by Michael Schwern.
206
2072.0 Tue Oct 24 05:47:17 CEST 2006
208 - support additional aio requests in Coro::AIO.
209
2101.9 (never properly released due to a glitch)
211
2121.8 Thu Feb 2 00:59:06 CET 2006
213 - applied suggested patch by SAMV to avoid problems during stupid
214 mark & sweep gc run.
215 - applied patch by Scott Walters for 5.9.3 compatibility.
216
2171.7 Tue Dec 27 01:41:58 CET 2005
218 - added Coro::AIO, a thin wrapper around IO::AIO.
219 - improved Makefile.PL explanations.
220
2211.6 Mon Dec 12 21:30:05 CET 2005
222 - additionally save PM_curpm might fix as-of-yet
223 unseen problems with regex matches being attributed to the
224 wrong package.
225 - add t/10_bugs.t, which currently checks against imho broken
226 perls that use 0.26MB of stack space per Perl_magic_get
227 invocation (newer linuxes) as opposed to the 0.0002MB perl
228 normally uses.
229 - make stacksize configurable for the ultimate debian experience.
230
2311.51 Mon Dec 12 18:48:36 CET 2005
232 - remove debugging warn() accidentally left in Coro::Select.
233
2341.5 Tue Nov 29 13:32:44 CET 2005
235 - use Coro::Event inside Coro::Select to avoid spurious deadlocks.
236 - fix Coro::Select 'select'.
237 - strict'ify some modules.
238 - libcoro errornously restored the SIGUSR2 handler to SIGUSR1.
239 - use XSLoader in selected modules.
240 - remove some 5.6 compatibility cruft.
241
2421.4 Tue Sep 6 00:11:05 CEST 2005
243 - libcoro did not take into account the trampoline on amd64,
244 when 'l'inux method was used.
245
2461.31 Tue Aug 30 23:31:33 CEST 2005
247 - some portability fixes/workarounds.
248
2491.3 Sat Aug 20 03:08:56 CEST 2005
250 - no code changes. module seems to work fine.
251 - improve Coro::State docs, remove reference to nonexisting
252 Coro::State::flush method.
253 - no longer autodetect windows, present it as an option instead.
254
2551.2 Mon May 16 02:00:55 CEST 2005
256 - included libcoro.c earlier in State.xs, might improve portability.
257 - use a faster and possibly more stable Coro::Cont implementation.
258 - accept x86_64 in addition to amd64 for optimized linux-amd64 support.
259 - fix bugs Coro::SemaphoreSet that could cause locks never to be freed.
260 - fix bugs in CoroAPI.h and document it in Coro::MakeMaker.
261
2621.11 Thu Mar 3 18:00:52 CET 2005
263 - change of contact address.
264
2651.1 Tue Feb 22 20:51:16 CET 2005
266 - support [l]inux method on amd64.
267 - allow some context switches while compiling/eval'ing.
268
2691.0 Tue Aug 10 03:47:27 CEST 2004
270 - Event 0.89 fixes the bug again.
271 - fix the remaining(??) memleak. I didn't free the cache-AV for
272 padlists, so one AV per created anon-sub leaked.
273
2740.97 Fri May 14 15:23:32 CEST 2004
275 - remove stupid gcc extension that was an experiment only anyway.
276
2770.96 Thu May 13 18:09:29 CEST 2004
278 - tell people if they have yet another broken Event variant.
279 - made Coro to work _better_ with threads (it seems that, as long
280 as you use Coros in one thread only everything will work fine.
281 Looking at the perl source the problem might be missing locking
282 between Coro and some perl internal routines. That's not easily
283 fixable).
284 - (hopefully) got rid of the static xnext inside transfer.
285
2860.95 Sun Feb 15 03:20:28 CET 2004
287 - removed Coro::State::flush, implemented a new and much
288 faster caching mechanism by attaching magic to the CV.
289 - a memleak with (real) closures remains.
290 - dramatically reduced size of stacks to 65536 (linux/x86), after
291 observing that even myhttpd never had more than a kilobyte
292 of stack in use (stack is still only physically allocated in
293 pagesize increments for systems suporting mmap).
294 - fix for a "cygwin-64" architecture added, proposed by Stefan Traby.
295 - fix for threaded perl.
296
2970.9 Sun Nov 23 04:32:39 CET 2003
298 - fix a few very hard to track down but (of course) stupid
299 memory leaks.
300 - stringify version number to avoid locale problems :(.
301 - perl 5.9.0 does not have SvPADBUSY, reported by Scott Walters,
302 hopefully ignoring it "just works".
303
3040.8 Wed Nov 5 19:38:40 CET 2003
305 - port to perl5.005_03, but only 5.8.x+ is supported!
306 - honor LocalAddr even without LocalPort in Coro::Socket.
307
3080.7 Tue May 27 03:12:38 CEST 2003
309 - the version jump indicates some level of testing, not gobs
310 of new features.
311 - uh... I found the prompt function in ExtUtils::MakeMaker.
312 Highly correct stuff, that is...
313 - fixed(?) a bug with die's in coroutines causing "panic: top_env".
314 the fix is not well understood by the author, so beware :(.
315
3160.652 Thu May 8 02:54:46 CEST 2003
317 - Applied patch by Slaven Rezic to set default to "s" on FreeBSD,
318 cause version 4 doesn't have ucontext.
319 - Benjamin Reed reported that setjmp works fine on darwin,
320 so preselect it.
321
3220.651 Sat Mar 29 15:00:23 CET 2003
323 - fix a bug in Coro::Handle where some bytes could get lost
324 on reading, reported by jason@nichego.net.
325
3260.65 Sun Mar 23 00:08:26 CET 2003
327 - added a README.linux-glibc.
328 - new module Coro::Select.
329 - also save/restore PL_comppad, fixes segfaults with 5.8.1.
330 (I never claimed that I know what I am doing ;).
331 - set default to setjmp/longjmp on non-x86-linux.
332
3330.6 Thu Nov 21 11:09:06 CET 2002
334 - made lazy context switching the default.
335 - fixed the bug where SemaphoreSet::guard
336 was the same as timed_guard and timed_guard was missing.
337 - fix a memleak in Coro::Timer.
338 - reclassified context sharing as not experimental.
339
3400.534 Sun Apr 14 03:05:12 CEST 2002
341 - fixed backspace => tab in header parsing (myhttpd).
342 - added eg/readline, for elmex.
343 - Coro::Event now calls ->start on first call instead of
344 ->again, which created wrong timeouts for the first ->next.
345 - fixed a bug where Coro::Socket returned a closed socket instead
346 of nothing on ECONNREFUSED and similar error conditions.
347
3480.533 Mon Feb 18 18:49:40 CET 2002
349 - adapt to finally working Event-0.85.
350 - fixed buggy event initialization (trapped under windows, but how
351 could it possibly have worked anywhere?)
352
3530.532 Wed Jan 16 02:45:32 CET 2002
354 - removed the reference to pp_entersub, might work on
355 (some) win32 perls now (testsuite works under cygwin,
356 without Event, which freezes).
357
3580.531 Mon Dec 10 22:18:44 CET 2001
359 - Fixed a bug in SemaphoreSet::timed_down.
360
3610.53 Tue Nov 27 21:11:13 CET 2001
362 - some tests for Event.
363 - slightly different internal architecture to get refcnt into the object
364 and not the reference. might fix a few bugs, certainly did introduce
365 new ones.
366 - Coro::Timer (independent of Event).
367 - new timed_wait functions for Coro::Signal, Semaphore, SemaphoreSet.
368
3690.52 Tue Nov 6 21:36:18 CET 2001
370 - ported to cygwin (trivial).
371 patch by Gerrit P. Haase <gp@familiehaase.de>.
372 - small setjmp code fix by Sullivan.DanielJ@epamail.epa.gov.
373
3740.51 Thu Nov 1 20:39:01 CET 2001
375 - terminate/cancel now work properly (otherwise termination
376 could cause a "next coroutine is not and contains not..."
377 error.
378 - added Coro::Socket::shutdown.
379 - Coro::Event::loop no is the same as Event::loop.
380 - implemented terminate with args + join.
381
3820.5 Fri Sep 28 16:15:35 CEST 2001
383 - fixed "print" on a Coro::Handle. The print method worked.
384 - small tweaks (seem to reduce memory consumption a lot)
385 in various modules.
386 - splendid use of "no warnings" scattered throughout.
387 - added Coro::Handle::rbuf, fixed Coro::Handle::fh.
388
3890.49 Sun Sep 16 02:42:45 CEST 2001
4 - changed some method calls to function calls for speed 390 - changed some method calls to function calls for speed
5 inside Coro::Handle. 391 inside Coro::Handle.
6 - make Coro::Handle use an array instead of a hash for 392 - make Coro::Handle use an array instead of a hash for
7 speed reasons. 393 speed reasons.
8 - IRIX mystery solved: it's SGI's NT, after all: "standard, huh?". 394 - IRIX mystery solved: it's SGI's NT, after all: "standard, huh?".
9 sjlj and ucontext should now work. 395 sjlj and ucontext should now work.
10 - IRIX-specific port for libcoro. 396 - IRIX-specific port for libcoro.
11 - swapped order of accept results to match IO::Socket. 397 - swapped order of accept results to match IO::Socket.
398 - changed getsock/peername to sock/peername to match IO::Socket.
399 - fixed a bug that caused segfault when returning to main task
400 under some circumstances.
401 - other bugfixes.
12 402
130.45 Sun Sep 2 02:54:01 CEST 2001 4030.45 Sun Sep 2 02:54:01 CEST 2001
14 - new method Coro::Handle::timeout. 404 - new method Coro::Handle::timeout.
15 - corrected speling of set...name to get...name in Coro::Socket. 405 - corrected speling of set...name to get...name in Coro::Socket.
16 - Coro::Socket::accept now returns a Coro::Socket, not a Coro::Handle. 406 - Coro::Socket::accept now returns a Coro::Socket, not a Coro::Handle.

Diff Legend

Removed lines
+ Added lines
< Changed lines
> Changed lines