ViewVC Help
View File | Revision Log | Show Annotations | Download File
/cvs/Coro/Changes
Revision: 1.417
Committed: Sat Nov 15 07:07:05 2008 UTC (15 years, 6 months ago) by root
Branch: MAIN
Changes since 1.416: +1 -0 lines
Log Message:
*** empty log message ***

File Contents

# Content
1 Revision history for Perl extension Coro.
2
3 TODO: should explore PerlIO::coroaio (perl leaks like hell).
4 TODO: maybe implement a default message channel, very much like Erlang's
5 actor model (which is cool in a lot of important aspects (failures!),
6 but very lacking in others (higher level ipc)).
7
8 5.0
9 - NEW ARCHITECTURE: use 4.912 for the stable version.
10 - bump API version to 7 - all dependents must be recompiled.
11 - Coro::Semaphore->adjust didn't correctly wake up enough waiters.
12 - async_pool did free a scalar value twice
13 ("Attempt to unreference...").
14 - perl's process emulation is now not even theoretically supported
15 anymore.
16 - directly patch the entersub opcode calling SLF functions (cede,
17 transfer and so on). this does speed up context switching, but
18 more importanly, it frees us from the hardcoded behaviour of
19 entersub, so we might actually be able to return something from
20 those functions and atcually create new ones.
21 - take advantage of __builtin_frame_address on gcc.
22 - expose THX in coroapi (not sure whether this was a wise decision,
23 as "threaded" perls are running at half speed anyways).
24 - implement execute_slf (schedule-like-function) interface that makes
25 it possible to implement schedule-like-functions in XS.
26 - use new SLF interface to massively speed up Coro::EV by roughly a factor
27 of two.
28 - removed timed_down and timed_guard functions - they were not being used
29 anyways and should be replaced by a more generic mechanism - and were
30 annoying to support anyways :
31 - used new SLF interface to massively speed up Coro::Semaphore by a factor
32 of three.
33 - forgot to include Coro/libcoro/README in the dist for all these years.
34
35 4.913 Sat Nov 15 07:58:28 CET 2008
36 - async_pool did free a scalar value twice
37 ("Attempt to unreference...").
38
39 4.912 Thu Nov 13 18:31:23 CET 2008
40 - minor cleanups.
41 - use much larger stacks on linux and perl < 5.8.8.
42 - Coro::Debug::new_unix_server did not unlink the socket
43 when destroyed.
44
45 4.911 Tue Nov 11 04:26:17 CET 2008
46 - "port" to threaded perls.
47
48 4.91 Mon Nov 10 05:36:38 CET 2008
49 - the ->throw exception object no longer leaks.
50 - creating a new cctx leaked a scopestack entry (memleak).
51 - new coroutines didn't get created with a zero flags field
52 (unknown impact).
53 - calling ->throw on a not-yet-started coroutine should now work
54 instead of being ignored.
55 - ->throw is now supported on Coro::State objects.
56 - clean up cctx creation code a bit.
57 - entersub is actually a UNOP, not a LOGOP (not a bugfix).
58
59 4.9 Sat Nov 8 17:45:27 CET 2008
60 - (libcoro) did not preserve rbp with CORO_ASM (we are getting there).
61 - (libcoro) no longer leak threads in the experimental pthread backend,
62 also speed it up considerably.
63 - (libcoro) do not rely on makecontext passing void *'s unscathed.
64 - fix compiletime dependencies on libcoro in the Makefile.
65 - cctx_count wasn't always updated properly.
66 - Coro::State::cctx_stacksize wasn't applied correctly.
67 - new function Coro::State::cctx_max_idle.
68 - the default max number of idle C contexts is now 4.
69 - (libcoro) try harder to get _setjmp/_longjmp.
70 - (libcoro) cleanup and extend the libcoro API to officially
71 allow the creation of empty source contexts.
72 - very experimental workaround for the totally broken netbsd platform.
73 - tried to hack around openbsd bugs.
74
75 4.804 Wed Nov 5 16:36:00 CET 2008
76 - Coro::Debug::new_unix_server would not create a non-blocking listening
77 socket, sometimes causing freezes.
78 - (libcoro) fix misaligned stack points for setjmp and assembly
79 methods, which can cause crashes on x86/x86_64 with a sufficiently
80 aggressive compiler.
81 - new function: Coro::Debug::new_tcp_server.
82 - move ->throw into the Coro class because it only works on coro objects.
83
84 4.803 Mon Nov 3 17:16:12 CET 2008
85 - (libcoro) use a global asm statement to become independent of gcc
86 otpimisations for CORO_ASM (thanks to pippijn for the idea).
87 - try to workaround yet another broken bsd, this time dragonfly.
88
89 4.802 Thu Oct 30 10:56:12 CET 2008
90 - support -fno-omit-frame-pointer on x86 with the assembly method.
91 - tune 01_unblock.t tests a bit.
92
93 4.801 Wed Oct 22 18:33:37 CEST 2008
94 - improve readline speed for very long "lines".
95 - backport to 5.8.8.
96
97 4.8 Thu Oct 2 13:34:40 CEST 2008
98 - new function Coro::AIO::aio_wait.
99 - Coro.:AIO and Coro::BDB now "use Coro::AnyEvent".
100 - greatly speed up and reduce memory usage of Coro::AIO requests.
101 - implement some other µ-optimisations.
102
103 4.749 Mon Sep 29 14:40:12 CEST 2008
104 - port to slow and broken pseudo-threaded perls. (courtesy pippijn).
105
106 4.748 Sat Sep 27 14:03:19 CEST 2008
107 - implement, but do not document, PerlIO::cede(granularity).
108 - Coro::Storable forgot to wrap Storable::pstore.
109 - work around the multitude of leaks and memory corruption
110 bugs in PerlIO::via by using our own C-level perliol. As a side
111 effect, Coro::Storable is now much, much, much faster.
112
113 4.747 Tue Sep 23 01:59:41 CEST 2008
114 - fix a per-cv memleak (one empty array was leaked per
115 code reference).
116 - avoid a crash in coro->call|rss when the coroutine was already
117 destroyed (most noticably when using Coro::Debug::ps :)
118 - also protect *Storable::FILE.
119 - push up default storable granularity to 20ms.
120
121 4.746 Sun Sep 21 03:22:20 CEST 2008
122 - be more insistent on locking Storable against reentrancy
123 in Coro::Storable.
124 - move swap_def?v and throw to Coro::State, as documented.
125
126 4.745 Thu Jul 24 00:14:38 CEST 2008
127 - remove debugging code related to MgPV_nolen_const, also try to
128 make it compile with perl 5.8.6 (yes, apple apparently loves
129 outdated software). Reported by John S.
130
131 4.744 Tue Jul 8 22:06:35 CEST 2008
132 - correctly provide default DIE/WARN handlers as documented.
133 - also overwrite PL_vtbl_sigelem.svt_clear, even though current
134 implementations inside perl work fine for us.
135
136 4.743 Mon Jun 16 00:21:57 CEST 2008
137 - when using Coro::EV without running EV::loop it could
138 result in busy-waiting for events, this has been fixed.
139 - reduce codesize and improve performance by using EV_DEFAULT_UC.
140
141 4.742 Sat May 31 14:10:21 CEST 2008
142 - implement a workaround for (some) perl <5.8.8 versions.
143 - require EV 3.3+.
144
145 4.741 Fri May 30 23:33:09 CEST 2008
146 - tell netbsd how utterly broken their imitation of an OS is
147 and refuse to build by default if pthreads are in use.
148 - switch to "s" method on all bsd's by default, as their ucontext
149 stuff seems just too broken.
150 - fix a bug in Coro::Select.
151
152 4.74 Thu May 29 20:05:31 CEST 2008
153 - do not test Coro::LWP for lack of dependencies.
154 - do not test Coro::Select for lack of working perls.
155
156 4.73 Thu May 29 2008
157 - fix a bug in Coro::EV which would cause it to block despite
158 there being runnable coroutines.
159 - sprinkle "no warnings" freely over everything, also suppress
160 warnings for some other modules.
161 - fix typo in WSAEWOULDBLOCK.
162
163 4.72 Sun May 25 05:14:36 CEST 2008
164 - tweak META.yaml a bit, unfortunately, there is no documented way
165 to have optional dependencies with CPAN. doh :(
166 - avoid running some tests on windows because they would fail due to
167 perl bug (broken fork, broken pipes...).
168 - work around perl on windows bugs where perl returns undocumented
169 error codes for sysread, syswrite etc. by taking advantage
170 of AnyEvent's workaround for that problem.
171 - use AnyEvent::Util::fh_nonblocking in Coro::Handle to work around
172 a common perl implementation bug on windows.
173 - use unix domain sockets in testsuite to work around
174 common perl implementation bugs on widows (they are emulated by
175 tcp sockets on windows. ugh.)
176
177 4.71 Sat May 24 20:01:27 CEST 2008
178 - fix a bug in Coro::AnyEvent ("Usage: Coro::AnyEvent::_schedule()").
179 - take advantage of async name resolution of AnyEvent::Util.
180 - work around brutal inet_aton override in Coro::LWP.
181 - take advantage of the readyhook in Coro::EV, for smoother
182 scheduling.
183
184 4.7 Sun May 11 00:32:19 CEST 2008
185 - completely reworked the Coro manpage.
186 - added Coro::AnyEvent, generic event loop integration.
187 - implement cancel, ready and kill commands in Coro::Debug.
188 - document find_coro in Coro::Debug.
189 - incompatible change: rename has_stack to has_cctx.
190 - Coro::AIO and Coro::BDB no longer force event model detection,
191 use AnyEvent::AIO and AnyEvent::BDB.
192
193 4.6 Sat Apr 26 10:05:14 CEST 2008
194 - INCOMPATIBLE CHANGE: sub/code attributes are no longer supported
195 by the Coro module. It was a mistake to have it in the first place.
196 - (experimental) support for activestate perl 5.10 (method "w").
197 - (experimental) support for strawberry perl (method "a").
198 - coro_sigelem_set did not return a value although it had to,
199 actual impact unknown.
200
201 4.51 Mon Apr 14 13:28:27 CEST 2008
202 - make it compile again on 5.8.
203
204 4.50 Thu Apr 10 09:43:17 CEST 2008
205 - I did it twice! (see 4.49).
206
207 4.49 Sun Apr 6 21:23:31 CEST 2008
208 - grr, instead of compiling the recent changes on 5.10 only they
209 were compiled on 5.8 only.
210
211 4.48 Sun Apr 6 20:36:46 CEST 2008
212 - allow coroutine switches during eval's under 5.10.x, as it seems
213 the parser is a per-interpreter option now, so this is safe
214 (this might fix the odd crashing bug).
215 - drop support for 5.9.x versions: they are dead, jim.
216
217 4.47 Sun Apr 6 00:37:52 CEST 2008
218 - force cctx allocation on API calls: we know we need to force one
219 and gcc actually manages to confuse our heuristic nowadays,
220 leading to crashes and worse.
221 - document force_cctx.
222
223 4.46 Fri Apr 4 22:05:43 CEST 2008
224 - upgrade libcoro, resulting in pthread-backend (which was only created
225 to fulfill the rules of the programming languages shootout).
226
227 4.45 Thu Mar 13 11:55:36 CET 2008
228 - fix a file leaking bug in eg/mhyttpd that would allow
229 downloading of any file (reported by oesi).
230 - fix deadlock bug in Coro::Channel (reported by Richard Hundt)
231 (also add testcase).
232 - support Broadcast option in Coro::Socket (patch by Richard Hundt,
233 apparently having loads of fun with that).
234 - implement $state->swap_defsv, ->swap_defav and document ->throw.
235
236 4.4 Wed Feb 13 16:44:29 CET 2008
237 - only assume .cfi pseudo isns exist on GNU/Linux.
238 - add get_timed method to Coro::Channels.
239 - fixed Coro::Debug manpage.
240 - perl 5.11 compatbility improvement based on patch by
241 Andreas König.
242
243 4.37 Sun Jan 20 11:25:23 CET 2008
244 - soften the check for an active parser for perl 5.10 (avoiding
245 "Coro::State::transfer called while parsing" in many cases).
246
247 4.36 Sun Jan 13 10:53:56 CET 2008
248 - reset diehook when terminating from an async_pool as to not
249 trigger any __DIE__ handlers.
250
251 4.35 Sun Jan 13 04:14:13 CET 2008
252 - "bt" debug command now displays any exceptions
253 from longmess and also skips the innermost
254 stackframes, giving more useufl output.
255 - allow backtraces in coroutines blocked in Coro::EV,
256 at a <1% speed hit.
257 - handle localising of $SIG{__WARN__} and __DIE__
258 properly (with a proper amount of dirty hacking).
259
260 4.34 Sat Dec 22 17:49:53 CET 2007
261 - upgrade to EV version 2.0 API.
262
263 4.33 Mon Dec 17 08:36:12 CET 2007
264 - make Coro::AIO etc. loadable in the absence of EV.
265
266 4.32 Mon Dec 17 07:46:02 CET 2007
267 - majorly improved Coro::Handle's performance with Coro::EV.
268 - implemented the until now mythical Coro::BDB module.
269 - specialcase EV in Coro::AIO and Coro::BDB for extra speed.
270
271 4.31 Wed Dec 5 12:32:39 CET 2007
272 - remove warn statement form Coro::Util.
273
274 4.3 Tue Dec 4 20:33:14 CET 2007
275 - calls to the idle function could cause stack corruption
276 when the stack changed.
277 - do no longer rely on the presence of EV::DNS if EV is used
278 (because it is gone), but instead take avdantage of EV::ADNS
279 if available.
280 - add ($) prototypes to all functions in Coro::Storable.
281 - use a conventional (safer) idle callback in Coro::EV.
282 - do accept quickly in Coro::Debug to avoid endless loops.
283
284 4.22 Fri Nov 30 16:04:04 CET 2007
285 - really use optimised versions for Event and EV in Coro::Util
286 and Coro::Handle.
287
288 4.21 Sun Nov 25 10:48:59 CET 2007
289 - fix a spurious memory read.
290 - Coro::EV no longer keeps the eventloop "alive".
291
292 4.2 Fri Nov 9 20:47:05 CET 2007
293 - enable/disable tracing from a new coroutine, not a pooled one.
294 - fix a memleak in Coro::Event.
295 - removed killall call from fork_eval.
296 - made sure store_fd is already loaded so that fork_eval
297 does not have to parse autoload in each subprocess.
298 - only use assembly method if -O switch is in $Config{optimize}.
299 - add (optional) Coro::EV module, so far the best event loop module
300 directly supported by Coro.
301 - if the event model is EV, use EV::DNS to resolve
302 stuff in Coro::Util.
303 - don't get confused by multiple event notifications in Coro::Handle.
304 - initial support for EV (libevent interface).
305 - require Event and EV using configure_requires, to force their existance.
306
307 4.13 Wed Oct 24 07:26:45 CEST 2007
308 - add Coro::Storable::blocking_thaw.
309 - use a vastly more complicated technique to localise
310 $SIG{__WARN/DIE__} that also works on perls <= 5.8.8.
311 - use a coroutine for the idle callback Coro::Event,
312 instead of running Event in the current coroutine context.
313 This also catches recursive invocations.
314 - actually report fork errors in gethostbyname and inet_aton.
315
316 4.11 Thu Oct 11 02:40:24 CEST 2007
317 - port to threaded perls.
318
319 4.1 Thu Oct 11 02:38:16 CEST 2007
320 - incompatible change: $SIG{__DIE__} and $SIG{__WARN__} will now
321 be local to each coro (see Coro::State).
322 - incompatible change: for very deep reasons, cede and cede_notself
323 cannot return anything, so nothing will be returned.
324 - possibly bring back 5.10 compatibility (untested).
325 - work around stupid (and wrong) warning on 5.10 :(.
326 - overlay the saved state over the context stack. This saves
327 a few hundred bytes per coroutine on average and also
328 speeds up context switching a bit.
329 - further tune default stack sizes.
330 - (more) correctly calculate stack usage in coro_rss.
331 - Coro::Storable::blocking_* did not properly lock
332 resulting in races between coroutines.
333 - added Coro::Storable::guard.
334 - stopping to trace a coroutine could destroy the cctx of
335 an unrelated coroutine.
336 - explain the relationship between Perl and C coroutines in
337 more detail in Coro::State.
338 - Coro::Util::inet_aton did not short-circuit dotted quad forms,
339 causing a fork per resolve. This also affected Coro::Socket.
340 - switch to a separate stack in $coro->call/eval to avoid
341 invalidating pointers.
342
343 4.03 Sat Oct 6 21:24:00 CEST 2007
344 - added Coro::throw method.
345 - minor code cleanups.
346
347 4.02 Sat Oct 6 02:36:47 CEST 2007
348 - fix a very minor per-coroutine memleak (a single codereference).
349 - fixed a bug where the currently in-use c context would be freed
350 prematurely (can happen only when programs change the stacksize
351 or use tracing).
352 - tracing can no longer keep a coro alive after it terminated.
353 - do static branch prediction in the common path for gcc. gives
354 about 2-5% speed improvement here.
355
356 4.01 Fri Oct 5 22:10:49 CEST 2007
357 - instead of recreating *a* standard output handle we simply
358 use STDOUT, which is faster and hopefully more robust.
359
360 4.0 Fri Oct 5 12:56:00 CEST 2007
361 - incompatibly changed Coro::Storable::freeze.
362 - major new feature: added Coro::Debug, for interactive coroutine
363 debugging, tracing and much more.
364 - major bug fix: unbelievable, but true: $_, $/ and many other
365 "saved" variables actually weren't being saved. This has been fixed,
366 of course, while increasing performance while losing all the save
367 flags.
368 - save flags are gone, and all the api functions dealing with them.
369 - added Coro::Semaphore::adjust.
370 - added Coro::Util::fork_eval.
371 - added Coro::Storable::{nfreeze,blocking_{freeze,nfreeze}}.
372 - added Coro::killall.
373 - reduce initial stack sizes to allow for "micro-coroutines".
374 - better async_pool resource management, moved parts of async_pool
375 handling to XS (major speed improvement).
376 - actually croak before modifying important data structures.
377 - refuse to transfer while compiling.
378 - possibly support eval EXPR better now.
379 - enable assembly per default on linux+bsd x86+amd64.
380 - all internal members were renamed _something for easier subclassing.
381 - many minor tweaks.
382
383 3.63 Wed May 16 14:10:06 CEST 2007
384 - implement handcoded assembly for x86/amd64 SVR ABI.
385
386 3.62 Fri Apr 27 21:36:06 CEST 2007
387 - upgrade libcoro (which might set unwind info correctly).
388 - change default on linux to setjmp/longjmp.
389
390 3.61 Thu Apr 19 12:36:18 CEST 2007
391 - Coro::Storable caused an endless loop when thawing invalid
392 pst's sometimes.
393 - use a Semaphore in Coro::Storable, as Storable doesn't
394 seem to be reentrant (although it is documented to
395 be threadsafe...).
396 - fix Coro::Signal to bring back the original unreliable
397 but stateful semantics.
398 - fixed a lot of typos in Coro.pm (patch submitted by David
399 Steinbrunner, which applied flawlessly).
400
401 3.6 Sat Apr 14 17:13:31 CEST 2007
402 - added some bugfixes to get eg/myhttpd working again.
403 - added Coro::Storable for often-cede'ing freeze/thaw.
404 - try to do a clean exit when a coroutine calls exit
405 (EXPERIMENTAL).
406 - got rid of indirect call through _coro_init.
407 - updated the partly antique examples in eg/ to
408 work again and be a bit less magic, too.
409 - fixed Coro::Signal semantics to work as documented again.
410
411 3.55 Sun Mar 25 01:20:47 CET 2007
412 - add SAVE_DEFFH to save the default filehandle and enable
413 it by default.
414 - finally move socket-operations from Coro::Socket to Coro::Handle
415 to be able to unblock foreign sockets.
416 - add Coro::State::save_also and guarded_save.
417 - add count accessor to Coro::Semaphore.
418 - add Coro::State::cctx_stacksize.
419 - just for the fun of it, do not rely on implicit context,
420 which can dramatically improve performance, but people
421 using windows-process-emulation perls do not care much about
422 performance.
423
424 3.51 Sun Mar 4 14:18:04 CET 2007
425 - fixed a problem when you weakened references to Coro::State's
426 (patch by Yuval Kogman).
427
428 3.501 Wed Feb 28 12:44:07 CET 2007
429 - rename some global symbols as macosx from hell redefines
430 them without asking.
431
432 3.5 Tue Feb 13 20:22:53 CET 2007
433 - do AnyEvent model detection earlier, avoiding problems
434 caused by first using AnyEvent and later Coro::Event.
435 - implement and document Coro::Event event objects.
436 - fix a potential problem in Coro::Event causing crashes.
437 - initialise PL_comppad when creating a new coroutine,
438 avoids crashes on early coro destruction.
439
440 3.41 Mon Jan 22 19:19:49 CET 2007
441 - readline on Coro::Handle did not support undefined $/,
442 nor did it deliver partial lines on EOF or error.
443 - implement malloc fallback for stack allocation because
444 stupid broken idiotic OSX has a stupid broken
445 idiotic fits-the-whole-os mmap "implementation" and
446 my dick feels longer if Coro is portable even to
447 obsolete platforms.
448
449 3.4 Fri Jan 19 21:52:54 CET 2007
450 - remove t/09_timer.t, as it isn't really testing much
451 but was rather flaky in practise.
452 - async_pool coro would keep arguments and callback alive until
453 it was reused.
454 - cancellation of a coroutine could cause spurious idle calls
455 in cede_notself.
456
457 3.3 Sat Jan 6 03:45:00 CET 2007
458 - implement $coro->on_destroy.
459 - Coro::Event blocking semantics have been changed,
460 documented and - hopefully - improved.
461 - fix nice adding, not subtracting, from priority.
462 - fix ->prio and api_is_ready (patch by Mark Hinds).
463 - fixed an assert ("... == prev__cctx->idle_te")
464 that could errronously trigger.
465 - fix various large and small memleaks.
466 - use a (hopefully) more stable cancel implementation
467 that immediately frees the coroutine data.
468 - cede/cede_notself return a status now.
469 - added Coro::guard function.
470 - added a global coroutine pool for jobs (on my machine,
471 I can create and execute 48k simple coros/s with async,
472 and 128k coros with async_pool).
473 - Coro::AIO now uses the coroutine priority as io priority.
474
475 3.2 Fri Dec 22 05:07:09 CET 2006
476 - improve portability to slightly older perls.
477 - use cleaner coroutine destruction.
478 - simplify configuration for users.
479 - optionally (unrecommended) prefer perl functions over
480 their coro replacements.
481
482 3.11 Tue Dec 5 13:11:24 CET 2006
483 - fixed some bogus assert's, but as perl.h disables assert even
484 without NDEBUG (thank you very much), not too many people should
485 notice (that did include myself). Andreas König noticed, though :)
486 - do not save/restore PL_sortcxix on >= 5.9.x, it doesn't seem to have
487 it. Also noticed by Andreas König :)
488 - save/restore tainted status.
489 - verified to pass the testsuite on my 5.9.5.
490
491 3.1 Mon Dec 4 23:03:40 CET 2006
492 - INCOMPATIBLE CHANGE: $/ is now per-coroutine (but slow).
493 - incompatible change: transfer flags are now per-state.
494 - give a better error message on deadlock.
495 - document Coro::nready.
496 - enhanced testsuite.
497
498 3.01 Sun Dec 3 23:47:42 CET 2006
499 - forgot to include Coro::Timer.
500
501 3.0 Sun Dec 3 22:57:25 CET 2006
502 - the "FINALLY COMPLETELY STABLE" release (coming soon:
503 the "FAMOUS LAST WORDS" release).
504 - implement a new stack sharing algorithm, which uses a stack
505 pool (size currently hardcoded).
506 - make stack sharing mandatory (it no longer uses a heuristic).
507 - eval/die no longer cause weird problems under heavy use.
508 - Coro::Event could cause livelocks if it was used but
509 no Coro::Event watchers were used.
510 - Coro::Event now uses asynccheck as prepare does not
511 check for changed watchers.
512 - Coro::Event allows multiple waiting coros and will wake up one
513 per event.
514 - Coro::Event should be cleaner and more efficient now.
515 - new utility function Coro::unblock_sub.
516 - document the sad fact that Event is no longer reentrant.
517 - putting a coroutine into the ready queue twice could under
518 some circumstances lead to stack corruption.
519 - minor incompatibility: subclassing Coro::State is supported
520 directly now without going through a _coro_state member.
521 - state/coro switching is much faster now.
522 - very minor optimisations and code/documentation cleanup.
523 - avoid problems due to compiler inlining.
524 - removed timers from Coro::Timer -> use AnyEvent instead.
525 - replaced Coro::idle coroutine by (cleaner) idle handler.
526 - updated to newest libcoro.
527 - implement enhanced support for valgrind.
528 - implement is_ready and return value for ready.
529 - removed Coro::Cont, it was a misnomer (it's generators),
530 and it was somewhat annoying to get it right. Will come back
531 if somebody asks for it :)
532 - many ->wait methods and Coro::Event could return
533 spuriously without the event having happened.
534
535 2.5 Tue Nov 7 12:22:33 CET 2006
536 - made Coro::Util, Coro::Select, Coro::Handle and Coro::Socket
537 use AnyEvent, moved them to Coro/.
538 - added Coro::LWP which contains all the uglyness required to
539 make LWP non-blocking.
540 - should work with perl 5.9.x now (Andreas König made me do it).
541 - fixed another bug in Coro::Select when the timeout was undef.
542 - reuse PL_start_env for all coros, saves some memory per coroutine.
543 - manage PL_top_env differently, hopefully to avoid panic: top_env.
544 - timeout argument was not properly used in Coro::Socket.
545 - allow limited forms of subclassing in Coro::Handle/Coro::Socket.
546 - emulate undocumented(!) functionality of IO::Socket required
547 by LWP(!!).
548 - updated eg/lwp to work with newer lwp's.
549 - remove "FATAL: uncaught exception" prefix. Coroutines that die
550 kill the whole process, just as exceptions in the main "coroutine"
551 did already.
552
553 2.1 Wed Nov 1 23:01:13 CET 2006
554 - fix a long-standing bug in Coro::Select where select with
555 zero timeout would instead change the current default filehandle.
556 - use a simpler and hopefully more robust way to clone padlists
557 (uses less memory and a perl function instead of our own).
558 - coro can now create a stack guard area on many architectures.
559 - Coro::AIO properly reexports additional functions from IO::AIO.
560 - updated libcoro with a workaround for OS X,
561 pach and testing by Michael Schwern.
562
563 2.0 Tue Oct 24 05:47:17 CEST 2006
564 - support additional aio requests in Coro::AIO.
565
566 1.9 (never properly released due to a glitch)
567
568 1.8 Thu Feb 2 00:59:06 CET 2006
569 - applied suggested patch by SAMV to avoid problems during stupid
570 mark & sweep gc run.
571 - applied patch by Scott Walters for 5.9.3 compatibility.
572
573 1.7 Tue Dec 27 01:41:58 CET 2005
574 - added Coro::AIO, a thin wrapper around IO::AIO.
575 - improved Makefile.PL explanations.
576
577 1.6 Mon Dec 12 21:30:05 CET 2005
578 - additionally save PM_curpm might fix as-of-yet
579 unseen problems with regex matches being attributed to the
580 wrong package.
581 - add t/10_bugs.t, which currently checks against imho broken
582 perls that use 0.26MB of stack space per Perl_magic_get
583 invocation (newer linuxes) as opposed to the 0.0002MB perl
584 normally uses.
585 - make stacksize configurable for the ultimate debian experience.
586
587 1.51 Mon Dec 12 18:48:36 CET 2005
588 - remove debugging warn() accidentally left in Coro::Select.
589
590 1.5 Tue Nov 29 13:32:44 CET 2005
591 - use Coro::Event inside Coro::Select to avoid spurious deadlocks.
592 - fix Coro::Select 'select'.
593 - strict'ify some modules.
594 - libcoro errornously restored the SIGUSR2 handler to SIGUSR1.
595 - use XSLoader in selected modules.
596 - remove some 5.6 compatibility cruft.
597
598 1.4 Tue Sep 6 00:11:05 CEST 2005
599 - libcoro did not take into account the trampoline on amd64,
600 when 'l'inux method was used.
601
602 1.31 Tue Aug 30 23:31:33 CEST 2005
603 - some portability fixes/workarounds.
604
605 1.3 Sat Aug 20 03:08:56 CEST 2005
606 - no code changes. module seems to work fine.
607 - improve Coro::State docs, remove reference to nonexisting
608 Coro::State::flush method.
609 - no longer autodetect windows, present it as an option instead.
610
611 1.2 Mon May 16 02:00:55 CEST 2005
612 - included libcoro.c earlier in State.xs, might improve portability.
613 - use a faster and possibly more stable Coro::Cont implementation.
614 - accept x86_64 in addition to amd64 for optimized linux-amd64 support.
615 - fix bugs Coro::SemaphoreSet that could cause locks never to be freed.
616 - fix bugs in CoroAPI.h and document it in Coro::MakeMaker.
617
618 1.11 Thu Mar 3 18:00:52 CET 2005
619 - change of contact address.
620
621 1.1 Tue Feb 22 20:51:16 CET 2005
622 - support [l]inux method on amd64.
623 - allow some context switches while compiling/eval'ing.
624
625 1.0 Tue Aug 10 03:47:27 CEST 2004
626 - Event 0.89 fixes the bug again.
627 - fix the remaining(??) memleak. I didn't free the cache-AV for
628 padlists, so one AV per created anon-sub leaked.
629
630 0.97 Fri May 14 15:23:32 CEST 2004
631 - remove stupid gcc extension that was an experiment only anyway.
632
633 0.96 Thu May 13 18:09:29 CEST 2004
634 - tell people if they have yet another broken Event variant.
635 - made Coro to work _better_ with threads (it seems that, as long
636 as you use Coros in one thread only everything will work fine.
637 Looking at the perl source the problem might be missing locking
638 between Coro and some perl internal routines. That's not easily
639 fixable).
640 - (hopefully) got rid of the static xnext inside transfer.
641
642 0.95 Sun Feb 15 03:20:28 CET 2004
643 - removed Coro::State::flush, implemented a new and much
644 faster caching mechanism by attaching magic to the CV.
645 - a memleak with (real) closures remains.
646 - dramatically reduced size of stacks to 65536 (linux/x86), after
647 observing that even myhttpd never had more than a kilobyte
648 of stack in use (stack is still only physically allocated in
649 pagesize increments for systems suporting mmap).
650 - fix for a "cygwin-64" architecture added, proposed by Stefan Traby.
651 - fix for threaded perl.
652
653 0.9 Sun Nov 23 04:32:39 CET 2003
654 - fix a few very hard to track down but (of course) stupid
655 memory leaks.
656 - stringify version number to avoid locale problems :(.
657 - perl 5.9.0 does not have SvPADBUSY, reported by Scott Walters,
658 hopefully ignoring it "just works".
659
660 0.8 Wed Nov 5 19:38:40 CET 2003
661 - port to perl5.005_03, but only 5.8.x+ is supported!
662 - honor LocalAddr even without LocalPort in Coro::Socket.
663
664 0.7 Tue May 27 03:12:38 CEST 2003
665 - the version jump indicates some level of testing, not gobs
666 of new features.
667 - uh... I found the prompt function in ExtUtils::MakeMaker.
668 Highly correct stuff, that is...
669 - fixed(?) a bug with die's in coroutines causing "panic: top_env".
670 the fix is not well understood by the author, so beware :(.
671
672 0.652 Thu May 8 02:54:46 CEST 2003
673 - Applied patch by Slaven Rezic to set default to "s" on FreeBSD,
674 cause version 4 doesn't have ucontext.
675 - Benjamin Reed reported that setjmp works fine on darwin,
676 so preselect it.
677
678 0.651 Sat Mar 29 15:00:23 CET 2003
679 - fix a bug in Coro::Handle where some bytes could get lost
680 on reading, reported by jason@nichego.net.
681
682 0.65 Sun Mar 23 00:08:26 CET 2003
683 - added a README.linux-glibc.
684 - new module Coro::Select.
685 - also save/restore PL_comppad, fixes segfaults with 5.8.1.
686 (I never claimed that I know what I am doing ;).
687 - set default to setjmp/longjmp on non-x86-linux.
688
689 0.6 Thu Nov 21 11:09:06 CET 2002
690 - made lazy context switching the default.
691 - fixed the bug where SemaphoreSet::guard
692 was the same as timed_guard and timed_guard was missing.
693 - fix a memleak in Coro::Timer.
694 - reclassified context sharing as not experimental.
695
696 0.534 Sun Apr 14 03:05:12 CEST 2002
697 - fixed backspace => tab in header parsing (myhttpd).
698 - added eg/readline, for elmex.
699 - Coro::Event now calls ->start on first call instead of
700 ->again, which created wrong timeouts for the first ->next.
701 - fixed a bug where Coro::Socket returned a closed socket instead
702 of nothing on ECONNREFUSED and similar error conditions.
703
704 0.533 Mon Feb 18 18:49:40 CET 2002
705 - adapt to finally working Event-0.85.
706 - fixed buggy event initialization (trapped under windows, but how
707 could it possibly have worked anywhere?)
708
709 0.532 Wed Jan 16 02:45:32 CET 2002
710 - removed the reference to pp_entersub, might work on
711 (some) win32 perls now (testsuite works under cygwin,
712 without Event, which freezes).
713
714 0.531 Mon Dec 10 22:18:44 CET 2001
715 - Fixed a bug in SemaphoreSet::timed_down.
716
717 0.53 Tue Nov 27 21:11:13 CET 2001
718 - some tests for Event.
719 - slightly different internal architecture to get refcnt into the object
720 and not the reference. might fix a few bugs, certainly did introduce
721 new ones.
722 - Coro::Timer (independent of Event).
723 - new timed_wait functions for Coro::Signal, Semaphore, SemaphoreSet.
724
725 0.52 Tue Nov 6 21:36:18 CET 2001
726 - ported to cygwin (trivial).
727 patch by Gerrit P. Haase <gp@familiehaase.de>.
728 - small setjmp code fix by Sullivan.DanielJ@epamail.epa.gov.
729
730 0.51 Thu Nov 1 20:39:01 CET 2001
731 - terminate/cancel now work properly (otherwise termination
732 could cause a "next coroutine is not and contains not..."
733 error.
734 - added Coro::Socket::shutdown.
735 - Coro::Event::loop no is the same as Event::loop.
736 - implemented terminate with args + join.
737
738 0.5 Fri Sep 28 16:15:35 CEST 2001
739 - fixed "print" on a Coro::Handle. The print method worked.
740 - small tweaks (seem to reduce memory consumption a lot)
741 in various modules.
742 - splendid use of "no warnings" scattered throughout.
743 - added Coro::Handle::rbuf, fixed Coro::Handle::fh.
744
745 0.49 Sun Sep 16 02:42:45 CEST 2001
746 - changed some method calls to function calls for speed
747 inside Coro::Handle.
748 - make Coro::Handle use an array instead of a hash for
749 speed reasons.
750 - IRIX mystery solved: it's SGI's NT, after all: "standard, huh?".
751 sjlj and ucontext should now work.
752 - IRIX-specific port for libcoro.
753 - swapped order of accept results to match IO::Socket.
754 - changed getsock/peername to sock/peername to match IO::Socket.
755 - fixed a bug that caused segfault when returning to main task
756 under some circumstances.
757 - other bugfixes.
758
759 0.45 Sun Sep 2 02:54:01 CEST 2001
760 - new method Coro::Handle::timeout.
761 - corrected speling of set...name to get...name in Coro::Socket.
762 - Coro::Socket::accept now returns a Coro::Socket, not a Coro::Handle.
763 - Coro::Handle now supports fileno and FILENO.
764 - added eg/myhttpd, a web server.
765 - Coro::Socket now accepts numeric ports as well as "port(number)"
766 syntax.
767 - moved some scheduling primitives into xs code. more to come.
768 - new simple priority system.
769 - implemented Coro::Event using XS, almost four times faster!
770 - small memory corruption problem fixed (boy that was difficult).
771
772 0.13 Wed Aug 8 16:53:07 CEST 2001
773 - new method Coro::Handle::readline.
774 - added eg/lwp and Coro::Handle::autoflush, to make LWP work.
775 (see eg/lwp on how to make LWP non-blocking).
776 - renamed Coro::Socket::new_inet to new.
777 - added Coro::Util, some utility functions.
778 - die/eval now works better (still get annoying (but true ;)
779 "Callback called exit" messages).
780 - Coro::Handle now supports timeout, and Coro::Socket Timeout.
781 - much better Coro::Socket support.
782
783 0.12 Fri Jul 27 04:19:01 CEST 2001
784 - do not use mmap if neither MAP_ANON nor MAP_ANONYMOUS is defined.
785 - very experimental stack sharing algorithm. not 100% safe but
786 should work well in practise ;)
787 - added Coro::RWLock.
788 - Coro::Specific now works.
789
790 0.11 Tue Jul 24 22:49:21 CEST 2001
791 - added specialized hack for newer and older linux versions (fast).
792 - renamed Coro::Event::IO to Coro::Handle.
793 - new module Coro::Socket.
794
795 0.1 Tue Jul 24 01:47:53 CEST 2001
796 - release candidate 3. A rather new internal structure :(
797 - the great renaming: Damian Conway gave me a suitable replacement
798 for yield.
799 - added Coro/libcoro, a portable coroutine implementation for C
800 (not even perl-dependent!!!), which can be used by Coro::State.
801 - renamed SAVE_DEFSV (on xs-level only) to avoid symbol clash in
802 perl-5.6.
803 - new function Coro::Event::idle.
804 - the idle process is now overriden
805 by default in Coro::Event.
806 - Coro::Channel now enforces the size.
807 - canceling events now works in all cases.
808 - Coro state now include $_ and $@.
809 - yet another bug workaround that I do not really understand :(
810 - new module Coro::Event::IO (very undocumented) to do
811 non-blocking i/o.
812 - performance tuning.
813
814 0.08 Thu Jul 19 06:13:25 CEST 2001
815 - release candidate 2 (new functionality).
816 - add Coro::State::flush function.
817 - transfer now supports a flags value (mostly for speed ATM).
818 - might compile and work in the presence of threads now.
819 - continuations have a saner syntax.
820 - no more memleaks.
821
822 0.07 Tue Jul 17 17:40:18 CEST 2001
823 - release candidate 1 ;)
824 - slightly nicer code.
825 - fixed a scheduling bug in Coro::Event.
826
827 0.06 Tue Jul 17 04:23:24 CEST 2001
828 - ok, I found the showstopper - the same sub must not be
829 re-entered in two different coroutines, otherwise => crash. I
830 see no easy solution to this problem, except by walking the call
831 chaing and saving/restoring the cv's, which is what I do now.
832 - memory leaks still latent, especially at thread termination.
833 - Coro::Event now works (haha).
834
835 0.05 Sun Jul 15 17:32:20 CEST 2001
836 - fixed a few issues in Coro::Event.
837 - I forgot to include Coro::Event in 0.04 :(:(:(
838
839 0.04 Sun Jul 15 05:24:59 CEST 2001
840 - @_ is now properly localized.
841 - Coro::State is now easier subclassable.
842 - Coro::Cont now coroutine-aware.
843 - Coro::Specific is a low-overhead module to create
844 coroutine-specific vars.
845 - Coro::Event provides a simple interface to Event.
846
847 0.03 Fri Jul 13 14:51:52 CEST 2001
848 - transfer() now implemented in XS (beware).
849 - new module Coro::Cont for really faked continuations.
850 - big internal architecture changes: Coro::State is now
851 really low-level and can thus be used to implement other
852 interesting things, While "Coro::" implements a process-like
853 model. Still crude and subject to change.
854 - $_ and $@ are no longer being localized.
855
856 0.02 Tue Jul 10 01:38:17 CEST 2001
857 - implemented "async" attribute.
858 - $_ and $@ are now localized.
859 - added Coro::Channel.
860 - more testcases, still no docs.
861
862 0.01 Tue Jul 3 02:18:41 CEST 2001
863 - original version; copied from Convert::Scalar.
864