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

Comparing Coro/Changes (file contents):
Revision 1.332 by root, Sun Apr 6 18:39:02 2008 UTC vs.
Revision 1.570 by root, Tue Jun 14 06:17:42 2011 UTC

1Revision history for Perl extension Coro. 1Revision history for Perl extension Coro.
2 2
3TODO: should explore PerlIO::via::CoroCede and PerlIO::via::CoroAIO. 3TODO: should explore PerlIO::coroaio (perl leaks like hell).
4TODO: maybe implement a default message channel, very much like Erlang's 4TODO: channel->maxsize(newsize)?
5 actor model (which is cool in a lot of important aspects (failures!), 5TODO: http://www.microsoft.com/msj/archive/s2ce.aspx
6 but very lacking in others (higher level ipc)). 6
7TODO: tutorial
8TODO: in_eval, PL_tained
9 - INCOMPATIBLE CHANGE: unreferenced coro objects will now be
10 destroyed and cleaned up automatically (e.g. async { schedule }).
11 - implement a JIT compiler for part of the thread switch code,
12 which gives a 50% speed improvement on threaded perls, and
13 about 4% on non-thraeded perls (so threaded perls now finally
14 reach about half the speed of non-threaded perls).
15 - avoid DEFSV and ERRSV, giving another 10% improvement
16 in thread switching.
17 - Coro::State->is_destroyed is now called is_zombie.
18 - implement a Coro->safe_cancel method that might fail, but
19 cancels in a "safer" way if it succeeds.
20 - get rid of two hash-accesses when initialising a new Coro - this
21 speeds up coro creation by almost a factor of two.
22 - croak when a coro that is being cancelled tries to block
23 (e.g. while executing a guard block), instead of crashing or
24 deadlocking.
25 - use a more robust and also faster method to identify Coro::State
26 objects - speeds up everything a bit.
27 - implement Coro->cancel in XS for a 20% speed improvement, and to
28 be able to implement mutual cancellation.
29 - speed up context switches by a percent or two by more efficiently
30 allocating context stack entries.
31 - implement Coro->join and Coro->on_destroy in XS for a speedup and
32 a reduction in memory use.
33 - cancelling a coro while it itself is cancelling another coro is
34 now supported and working, instead of triggering an assertion.
35 - be a bit more crash-resistant when calling (buggy) on_destroy
36 callbacks (best effort).
37 - move on_destroy into the slf_frame, to allow extension slf
38 functions to have destructors.
39 - get rid if coro refcounting - simply crash in other interpreter
40 threads by nulling the pointers on clone.
41 - simplify warn/die hook handling when loading Coro - the convoluted
42 logic seems to be no longer neccessary.
43 - use libecb instead of our own home-grown gcc hacks.
44 - document alternatives to Coro::LWP. Please use them :)
45
465.372 Wed Feb 23 06:14:30 CET 2011
47 - apparently mingw doesn't provide a working gettimeofday, try to
48 work around that by relying on Time::HiRes (indirectly brought to
49 my attention by Max Maischein).
50 - fix some portability issues when Time::HiRes was used.
51
525.371 Mon Feb 21 14:36:08 CET 2011
53 - backport to windows process emulation code again.
54
555.37 Sat Feb 19 07:49:44 CET 2011
56 - add a big "Coro thread life cycle" section to "man Coro".
57 - try a tentative workaround against the breakage that 5.13 has
58 introduced without depreciation period. sigh.
59 - no longer use Time::HiRes if gettimeofday is available, which
60 saves quite a lot of memory.
61
625.36 Sun Feb 13 05:33:41 CET 2011
63 - automatically load Coro::Channel, Coro::RWLock, Coro::Semaphore,
64 Coro::SemaphoreSet, Coro::Signal and Coro::Specific on first "new"
65 method call.
66 - undocument Coro::Timer::sleep and obsolete whole module.
67 - optimise Coro::Timer::timeout memory and cpu usage.
68 - slightly updated Coro::Intro for recent changes.
69 - do not initialise PL_dirty anymore.
70
715.25 Thu Nov 11 01:08:39 CET 2010
72 - try a different approach on netbsd - netbsd 5 finally has marginally
73 working pthreads, but still broken ucontext/sigaltstack.
74 - openbsd 4.8 finally got their act together, Coro works out of the box
75 with asm, setjmp and pthreads (no change, just informational).
76
775.24 Sat Oct 23 11:27:12 CEST 2010
78 - port to the EV 4.0 API.
79 - work around bugs in mingw32, making strawberry perl work
80 out of the box.
81 - correctly modify Coro::AIO function prototypes
82 so that they reflect the "no optional parameters" rule.
83 - "ported" libcoro to C++.
84
855.23 Mon May 17 18:50:42 CEST 2010
86 - be more resistant to ordering changes when initialising
87 Coro::AnyEvent, Coro::EV and Coro::Event (reported by Matthias
88 Waldorf).
89 - document that perl 5.12 deliberately removed support for cloning.
90
915.22 Wed Apr 14 03:55:35 CEST 2010
92 - correctly return udnef on errors in Coro::Handle::read/write
93 (testcase by Marc Mims).
94 - convert Coro::Util into a "perl compatibility wrapper" - the functions
95 are less useful now, but are drop-in replacements for existing
96 functions, listing better alternatives in the documentation. This also
97 fixes a bug in Coro::LWP which naively substituted Socket::inet_aton
98 with Coro::Util::inet_aton.
99 - do not override $Coro::idle unconditionally in Coro.pm, as other
100 modules could have provided their own idle coro already
101 (for exmaple, Coro::AnyEvent).
102 - fix Coro::Util::gethost* functions.
103 - Coro::Timer corretcly exports it's symbols (reported by Hideki Yamamura).
104
1055.21 Wed Dec 16 07:19:51 CET 2009
106 - automatically load Coro::AnyEvent when AnyEvent and Coro are used
107 together.
108 - add some examples on how to combine other event loops with Coro in
109 Coro::AnyEvent, and how to run it (and not to block). Seems to be
110 the most common source of confusion.
111 - try to catch people naively blocking in an event callback.
112 - work around the perl filehandle bug issue in conjunction with
113 older common::sense (as indirectly pointed out by ZSystem).
114 - clarify the "not from signal handlers" section.
115
1165.2 Sun Oct 4 14:54:24 CEST 2009
117 - Coro::Storable destroyed the prototypes of the functions it wrapped.
118 - export rouse_cb and rouse_wait by default now.
119 - fix various prototype mismatches in Coro::AnyEvent and Coro::Handle.
120 - new method $state->swap_sv.
121 - added section on "windows process emulation" to the manpage,
122 after a not-so-fruitful (nor-friendly) "discussion" with chip
123 salzenberg (discussion implies arguments, but his only arguments
124 were ad-hominems, one wonders why he started it in the first
125 place). I hope this explains it well enough for him to understand,
126 and maybe well enough for others to understand.
127 - use common::sense everywhere now.
128 - idle callbacks are no longer supported, use idle coros instead.
129 - print a thread listing when a deadlock is detected.
130
1315.17 Sat Aug 22 23:09:31 CEST 2009
132 - work around a bug in the perl debugger causing crashes
133 when running under the debugger by marking _pool_handler as nodebug.
134 - speed up Coro::async considerably.
135 - try some hacks to get netbsd to work "more often" - their broken
136 setjmp/longjmp, ucontext *and* phtreads are really hard on Coro.
137 - convert Coro to AE 5.0 API.
138
1395.162 Tue Jul 28 04:04:03 CEST 2009
140 - perl 5.8.2 is now minimum requirement.
141 - skip t/19_handle.t on broken windows perls.
142
1435.161 Wed Jul 22 04:47:38 CEST 2009
144 - Coro::AnyEvent::poll could have a different prototype when EV was
145 used as backend (analyzed by Tatsuhiko Miyagawa).
146 - Coro::AnyEvent errornously initialised the event loop when loaded,
147 not on demand.
148 - try to workaround rare */t/01_unblock.t failures.
149
1505.16 Tue Jul 21 01:44:37 CEST 2009
151 - Coro::AnyEvent failed to hook into the event loop
152 when no threads had been readied between detecting
153 the event loop and actually running it.
154 - considerably speed up Coro::Select by taking avdantage
155 of AnyEvent > 4.8 and some other optimisations.
156 - implement paragraph readline mode in Coro::Handle
157 (based on patches by Zsbán Ambrus).
158 - replace WSAEINPROGRESS by WSAEWOULDBLOCK (reported
159 and analyzed by Yasuhiro MATSUMOTO).
160 - clarified libcoro license and copyright.
161 - someone stole my EXTRA_META!!!
162 - implement Coro::Select::patch_pp_sselect and it's brother,
163 for hardcode select overriding.
164
1655.151 Mon Jul 6 05:41:57 CEST 2009
166 - backport to windows process emulation code again (patch by
167 Yasuhiro MATSUMOTO).
168 - slightly update Coro::MakeMaker.
169
1705.15 Tue Jun 30 10:28:06 CEST 2009
171 - deprecate Coro::Socket, document how to get ipv6 support via
172 AnyEvent::Socket instead.
173 - implement signal->wait ($cb) interface, similar to semaphores.
174 - work around SvOK not supporting getmagic, so we have to getmagic
175 to test for undef :/ (reported by Matthias Waldorf).
176 - load Coro::AnyEvent in all modules using AnyEvent.
177 - work around perl corrupting our internal data structures,
178 reported by Tokuhiro Matsuno.
179 - enable per-coroutine real and cpu time gathering
180 (Coro::State::enable_times).
181
1825.14 Wed Jun 24 01:37:48 CEST 2009
183 - provide explicit functions to "cede" to the event loop in Coro::AnyEvent,
184 as this seems to have been a difficult concept (poll, sleep, idle,
185 idle_upto).
186 - add Coro::AnyEvent::readable/writable functions.
187 - clarify Coro::EV/Event/AnyEvent manpages.
188 - free per-thread global scalars in the thread calling ->cancel, to
189 avoid crashes when $_, $@ etc., are magical but some of those
190 had already been freed.
191 - "unexperimentalise" the callback interface for Coro::Semaphore.
192 - speed up ready queue management/context switching by using a linked
193 list instead of an array (~5-10%).
194 - implement "watch" command in Coro::Debug shells.
195 - for fun, implement time-slicing as an example in the manpage.
196 - if AnyEvent detects EV or Event, but we don't have Coro::EV or
197 Coro::Event, use the normal AnyEvent handling instead of dieing
198 (the same is true for Coro::Handle).
199 - properly document Coro::EV::timed_io_once.
200 - avoid unneccessary ->cancel calls in Coro::Handle.
201 - maybe make it work on mingw32 with win32 backend
202 (based on patch by Yasuhiro Matsumoto).
203
2045.132 Fri May 29 09:00:39 CEST 2009
205 - do not keep a reference to the argument itself in
206 Coro::Semaphore::guard, as it could change later.
207 - support SO_RCVBUF/SO_SNDBUF nonstandard Coro::Socket options,
208 should support a prepare callback.
209
2105.131 Mon Mar 16 23:20:37 CET 2009
211 - implement and document Coro->suspend, Coro->resume.
212 - fix Coro::Select implementation to not (often) close
213 the passed file descriptors (testcase provided by pippijn).
214
2155.13 Mon Dec 15 21:51:42 CET 2008
216 - EXPERIMENTAL: implement dynamic winds (on_enter/on_leave).
217 - don't set diehook to C<undef> but instead to NULL, to avoid
218 spurious warnings.
219 - fix a lot of bugs in Coro::SemaphoreSet.
220 - Coro::SemaphoreSet will less often create a semaphore needlessly.
221 - add Coro::SemaphoreSet::count and wait methods.
222 - take advantage of the new Guard module.
223 - deprecate Coro::guard.
224 - try to fix the dreaded 01_unblock tests once more. I hate it when
225 testsuites need more fixing than the code they are supposed to test.
226 - croak in more cases when a required callback isn't resolvable.
227 - fix some minor issues in Coro::State->call/eval.
228 - use current coroutine context instead of a temporary one
229 when temporarily switching to another coroutine.
230 - do not call C-level on_destroy handlers during global destruction,
231 to avoid needless segfaults.
232
2335.12 Sun Dec 7 13:30:38 CET 2008
234 - add default config for MirOS, which seems to be a bug-to-bug
235 compatible fork of openbsd ("world domination by releasing
236 openbsd cvs before the openbsd folks do it" or so :).
237 - free_padlist did destroy the names pad, not good, but didn't
238 seem to bother perl - this could fix issues such as eval ""
239 inside a function called from multiple coroutines.
240 - use a different method to detect destruction time.
241 - be more careful when freeing padlists just before global
242 destruction.
243 - fixed and expanded the call/cc example.
244 - renamed _terminate to _coro_run.
245 - new method Coro::Channel->shutdown.
246 - try pthreads on openbsd <4.4 (broken sigaltstack, will
247 pthreads fare better?).
248 - be less picky about destroying "a" running coroutine.
249
2505.11 Tue Nov 25 21:49:05 CET 2008
251 - DEBUGGING in 5.10.0 is a mess: it constantly flags perfectly
252 working code with failed assertions, introducing more bugs than
253 it fixes, requiring elaborate workarounds :(
254
2555.1 Mon Nov 24 08:54:59 CET 2008
256 - wrote a small introductory tutorial - Coro::Intro.
257 - convert Coro::Timer, Coro::Select and Coro::Util to rouse API.
258 - Coro::Select did errornously dup the file descriptors
259 and didn't work with all AnyEvent backends.
260 - Coro::Select wasn't imported correctly form Coro::LWP, causing blocking
261 LWP data transfers.
262 - disassociate c contexts from coro objects - this is agruably more
263 correct, but mostly allows sharing of cctxs between coro and state
264 objects, for added memory savings and speed increases.
265 - bumped $Coro::POOL_RSS up to 32kb by default.
266 - no longer set the optype to OP_CUSTOM, as B::* understandably
267 doesn't like this very much (and we *are* a type of entersub).
268 - implement state cloning, just to prove that call/cc can be done.
269 - automatically load Coro::AnyEvent in Coro::Handle.
270 - wrap ->cancel calls in eval inside Coro::Handle as EV watchers
271 do not have this method (and don't need it either).
272 - speed up generic anyevent methods in Coro::Handle by using rouse
273 callbacks.
274 - allow coroutines in $Coro::IDLE, speeding up Coro::AnyEvent and
275 others. It also makes the debugger happier, as you can trace
276 through the idle threads now.
277 - add comppad_name* and hints ($^H, %^H) to per-thread variables.
278 - eg/event was pretty broken.
279 - better 5.8.6 compatibility.
280
2815.0 Thu Nov 20 10:35:05 CET 2008
282 - NEW ARCHITECTURE: use the latest 4.x version if you experience
283 stability issues.
284 - bump API version to 7 - all dependents must be recompiled.
285 - removed timed_* functions - they were not being used anyways
286 and should be replaced by a more generic mechanism -
287 and were annoying to support anyways :)
288 - removed SemaphoreSet's waiter method - use sem method instead.
289 - Coro::Semaphore->adjust didn't correctly wake up enough waiters.
290 - async_pool did free a scalar value twice
291 ("Attempt to unreference...").
292 - fix a longstanding bug where calling terminate on a coro that
293 was waiting for a semaphore that was just becoming available
294 would cause a deadlock (semaphore would get into a state where
295 it was available but waiters were still blocked).
296 - calling throw on a coroutine that is waiting for a semaphore will
297 no longer make it acquire the semaphore (and thus leak a count).
298 - perl's process emulation is now not even theoretically supported
299 anymore.
300 - new functions Coro::rouse_cb and Coro::rouse_wait for easier
301 conversion of callback-style to blocking-style.
302 - new methods $coro->schedule_to and ->cede_to, to specifically
303 schedule or cede to a specific coroutine.
304 - new function Coro::Semaphore::wait.
305 - use named constants in Coro::Channel (Richard Hundt).
306 - directly patch the entersub opcode calling SLF functions (cede,
307 transfer and so on). this does speed up context switching, but
308 more importanly, it frees us from the hardcoded behaviour of
309 entersub, so we might actually be able to return something from
310 those functions and atcually create new ones.
311 - take advantage of __builtin_frame_address on gcc.
312 - expose THX in coroapi (not sure whether this was a wise decision,
313 as "threaded" perls are running at half speed anyways).
314 - implement execute_slf (schedule-like-function) interface that makes
315 it possible to implement schedule-like-functions in XS.
316 - use new SLF interface to massively speed up Coro::EV by roughly a
317 factor of two.
318 - used new SLF interface to massively speed up Coro::Semaphore by a
319 factor of three.
320 - used new SLF interface to speed up Coro::AIO by roughly a factor of
321 four and reduce its memory usage considerably.
322 - implement Coro::SemaphoreSet purely in terms of Coro::Semaphore,
323 for a nice speedup and vastly more correct behaviour. Also implement
324 a new method "sem" to get at the underlying semaphore object.
325 - implement Coro::Channel in terms of Coro::Semaphore, for a moderate
326 (in comparison) 20-40% speedup.
327 - used new SLF interface to reimplement Coro::Signal gaining
328 some unknown (because I was too lazy), but certain, speedup, and also
329 making signals reliable for the first time.
330 - used new SLF interface and other optimisations to speed up async_pool
331 by a factor of two. It also doesn't rely on perl's exception mechanism
332 to exit anymore. The overhead for terminating an async_pool, coro over
333 a normal async is now very small.
334 - sped up coroutine creation/destruction by 40%.
335 - forgot to include Coro/libcoro/README in the dist for all these years.
336 - work around a freebsd pthreads bug (manual testcancel is required as
337 pthread_cond_wait isn't a cancellation point on freebsd).
338 - use new rouse functions to speed up and simplify Coro::BDB.
339 - make "prefer perl native functions" work with threaded perls.
340 - condense Coro::Debug ps output, hint at v and w flags.
341 - (libcoro) lots of minor cleanups and portability improvements.
342
3434.914 Wed Nov 19 12:54:18 CET 2008
344 - fix a disastrous bug in the readline optimisation
345 introduced in 4.801.
346
3474.913 Sat Nov 15 07:58:28 CET 2008
348 - async_pool did free a scalar value twice
349 ("Attempt to unreference...").
350
3514.912 Thu Nov 13 18:31:23 CET 2008
352 - minor cleanups.
353 - use much larger stacks on linux and perl < 5.8.8.
354 - Coro::Debug::new_unix_server did not unlink the socket
355 when destroyed.
356
3574.911 Tue Nov 11 04:26:17 CET 2008
358 - "port" to threaded perls.
359
3604.91 Mon Nov 10 05:36:38 CET 2008
361 - the ->throw exception object no longer leaks.
362 - creating a new cctx leaked a scopestack entry (memleak).
363 - new coroutines didn't get created with a zero flags field
364 (unknown impact).
365 - calling ->throw on a not-yet-started coroutine should now work
366 instead of being ignored.
367 - ->throw is now supported on Coro::State objects.
368 - clean up cctx creation code a bit.
369 - entersub is actually a UNOP, not a LOGOP (not a bugfix).
370
3714.9 Sat Nov 8 17:45:27 CET 2008
372 - (libcoro) did not preserve rbp with CORO_ASM (we are getting there).
373 - (libcoro) no longer leak threads in the experimental pthread backend,
374 also speed it up considerably.
375 - (libcoro) do not rely on makecontext passing void *'s unscathed.
376 - fix compiletime dependencies on libcoro in the Makefile.
377 - cctx_count wasn't always updated properly.
378 - Coro::State::cctx_stacksize wasn't applied correctly.
379 - new function Coro::State::cctx_max_idle.
380 - the default max number of idle C contexts is now 4.
381 - (libcoro) try harder to get _setjmp/_longjmp.
382 - (libcoro) cleanup and extend the libcoro API to officially
383 allow the creation of empty source contexts.
384 - very experimental workaround for the totally broken netbsd platform.
385 - tried to hack around openbsd bugs.
386
3874.804 Wed Nov 5 16:36:00 CET 2008
388 - Coro::Debug::new_unix_server would not create a non-blocking listening
389 socket, sometimes causing freezes.
390 - (libcoro) fix misaligned stack points for setjmp and assembly
391 methods, which can cause crashes on x86/x86_64 with a sufficiently
392 aggressive compiler.
393 - new function: Coro::Debug::new_tcp_server.
394 - move ->throw into the Coro class because it only works on coro objects.
395
3964.803 Mon Nov 3 17:16:12 CET 2008
397 - (libcoro) use a global asm statement to become independent of gcc
398 otpimisations for CORO_ASM (thanks to pippijn for the idea).
399 - try to workaround yet another broken bsd, this time dragonfly.
400
4014.802 Thu Oct 30 10:56:12 CET 2008
402 - support -fno-omit-frame-pointer on x86 with the assembly method.
403 - tune 01_unblock.t tests a bit.
404
4054.801 Wed Oct 22 18:33:37 CEST 2008
406 - improve readline speed for very long "lines".
407 - backport to 5.8.8.
408
4094.8 Thu Oct 2 13:34:40 CEST 2008
410 - new function Coro::AIO::aio_wait.
411 - Coro.:AIO and Coro::BDB now "use Coro::AnyEvent".
412 - greatly speed up and reduce memory usage of Coro::AIO requests.
413 - implement some other µ-optimisations.
414
4154.749 Mon Sep 29 14:40:12 CEST 2008
416 - port to slow and broken pseudo-threaded perls. (courtesy pippijn).
417
4184.748 Sat Sep 27 14:03:19 CEST 2008
419 - implement, but do not document, PerlIO::cede(granularity).
420 - Coro::Storable forgot to wrap Storable::pstore.
421 - work around the multitude of leaks and memory corruption
422 bugs in PerlIO::via by using our own C-level perliol. As a side
423 effect, Coro::Storable is now much, much, much faster.
424
4254.747 Tue Sep 23 01:59:41 CEST 2008
426 - fix a per-cv memleak (one empty array was leaked per
427 code reference).
428 - avoid a crash in coro->call|rss when the coroutine was already
429 destroyed (most noticably when using Coro::Debug::ps :)
430 - also protect *Storable::FILE.
431 - push up default storable granularity to 20ms.
432
4334.746 Sun Sep 21 03:22:20 CEST 2008
434 - be more insistent on locking Storable against reentrancy
435 in Coro::Storable.
436 - move swap_def?v and throw to Coro::State, as documented.
437
4384.745 Thu Jul 24 00:14:38 CEST 2008
439 - remove debugging code related to MgPV_nolen_const, also try to
440 make it compile with perl 5.8.6 (yes, apple apparently loves
441 outdated software). Reported by John S.
442
4434.744 Tue Jul 8 22:06:35 CEST 2008
444 - correctly provide default DIE/WARN handlers as documented.
445 - also overwrite PL_vtbl_sigelem.svt_clear, even though current
446 implementations inside perl work fine for us.
447
4484.743 Mon Jun 16 00:21:57 CEST 2008
449 - when using Coro::EV without running EV::loop it could
450 result in busy-waiting for events, this has been fixed.
451 - reduce codesize and improve performance by using EV_DEFAULT_UC.
452
4534.742 Sat May 31 14:10:21 CEST 2008
454 - implement a workaround for (some) perl <5.8.8 versions.
455 - require EV 3.3+.
456
4574.741 Fri May 30 23:33:09 CEST 2008
458 - tell netbsd how utterly broken their imitation of an OS is
459 and refuse to build by default if pthreads are in use.
460 - switch to "s" method on all bsd's by default, as their ucontext
461 stuff seems just too broken.
462 - fix a bug in Coro::Select.
463
4644.74 Thu May 29 20:05:31 CEST 2008
465 - do not test Coro::LWP for lack of dependencies.
466 - do not test Coro::Select for lack of working perls.
467
4684.73 Thu May 29 2008
469 - fix a bug in Coro::EV which would cause it to block despite
470 there being runnable coroutines.
471 - sprinkle "no warnings" freely over everything, also suppress
472 warnings for some other modules.
473 - fix typo in WSAEWOULDBLOCK.
474
4754.72 Sun May 25 05:14:36 CEST 2008
476 - tweak META.yaml a bit, unfortunately, there is no documented way
477 to have optional dependencies with CPAN. doh :(
478 - avoid running some tests on windows because they would fail due to
479 perl bug (broken fork, broken pipes...).
480 - work around perl on windows bugs where perl returns undocumented
481 error codes for sysread, syswrite etc. by taking advantage
482 of AnyEvent's workaround for that problem.
483 - use AnyEvent::Util::fh_nonblocking in Coro::Handle to work around
484 a common perl implementation bug on windows.
485 - use unix domain sockets in testsuite to work around
486 common perl implementation bugs on widows (they are emulated by
487 tcp sockets on windows. ugh.)
488
4894.71 Sat May 24 20:01:27 CEST 2008
490 - fix a bug in Coro::AnyEvent ("Usage: Coro::AnyEvent::_schedule()").
491 - take advantage of async name resolution of AnyEvent::Util.
492 - work around brutal inet_aton override in Coro::LWP.
493 - take advantage of the readyhook in Coro::EV, for smoother
494 scheduling.
495
4964.7 Sun May 11 00:32:19 CEST 2008
497 - completely reworked the Coro manpage.
498 - added Coro::AnyEvent, generic event loop integration.
499 - implement cancel, ready and kill commands in Coro::Debug.
500 - document find_coro in Coro::Debug.
501 - incompatible change: rename has_stack to has_cctx.
502 - Coro::AIO and Coro::BDB no longer force event model detection,
503 use AnyEvent::AIO and AnyEvent::BDB.
504
5054.6 Sat Apr 26 10:05:14 CEST 2008
506 - INCOMPATIBLE CHANGE: sub/code attributes are no longer supported
507 by the Coro module. It was a mistake to have it in the first place.
508 - (experimental) support for activestate perl 5.10 (method "w").
509 - (experimental) support for strawberry perl (method "a").
510 - coro_sigelem_set did not return a value although it had to,
511 actual impact unknown.
512
5134.51 Mon Apr 14 13:28:27 CEST 2008
514 - make it compile again on 5.8.
515
5164.50 Thu Apr 10 09:43:17 CEST 2008
517 - I did it twice! (see 4.49).
518
5194.49 Sun Apr 6 21:23:31 CEST 2008
520 - grr, instead of compiling the recent changes on 5.10 only they
521 were compiled on 5.8 only.
7 522
84.48 Sun Apr 6 20:36:46 CEST 2008 5234.48 Sun Apr 6 20:36:46 CEST 2008
9 - allow coroutine switches during eval's under 5.10.x, as it seems 524 - allow coroutine switches during eval's under 5.10.x, as it seems
10 the parser is a per-interpreter option now, so this is safe 525 the parser is a per-interpreter option now, so this is safe
11 (this might fix the odd crashing bug). 526 (this might fix the odd crashing bug).

Diff Legend

– Removed lines
+ Added lines
< Changed lines
> Changed lines