ViewVC Help
View File | Revision Log | Show Annotations | Download File
/cvs/Coro/Changes
Revision: 1.467
Committed: Thu Nov 27 12:00:58 2008 UTC (15 years, 5 months ago) by root
Branch: MAIN
Changes since 1.466: +6 -1 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 - add default config for MirOS, which seems to be a bug-to-bug
9 compatible fork of openbsd ("world domination by releasing
10 openbsd cvs before the openbsd folks do it" or so :).
11 - use a different method to detect destruction time.
12 - be more careful when freeing padlists just before global
13 destruction.
14
15 5.11 Tue Nov 25 21:49:05 CET 2008
16 - DEBUGGING in 5.10.0 is a mess: it constantly flags perfectly
17 working code with failed assertions, introducing more bugs than
18 it fixes, requiring elaborate workarounds :(
19
20 5.1 Mon Nov 24 08:54:59 CET 2008
21 - wrote a small introductory tutorial - Coro::Intro.
22 - convert Coro::Timer, Coro::Select and Coro::Util to rouse API.
23 - Coro::Select did errornously dup the file descriptors
24 and didn't work with all AnyEvent backends.
25 - Coro::Select wasn't imported correctly form Coro::LWP, causing blocking
26 LWP data transfers.
27 - disassociate c contexts from coro objects - this is agruably more
28 correct, but mostly allows sharing of cctxs between coro and state
29 objects, for added memory savings and speed increases.
30 - bumped $Coro::POOL_RSS up to 32kb by default.
31 - no longer set the optype to OP_CUSTOM, as B::* understandably
32 doesn't like this very much (and we *are* a type of entersub).
33 - implement state cloning, just to prove that call/cc can be done.
34 - automatically load Coro::AnyEvent in Coro::Handle.
35 - wrap ->cancel calls in eval inside Coro::Handle as EV watchers
36 do not have this method (and don't need it either).
37 - speed up generic anyevent methods in Coro::Handle by using rouse
38 callbacks.
39 - allow coroutines in $Coro::IDLE, speeding up Coro::AnyEvent and
40 others. It also makes the debugger happier, as you can trace
41 through the idle threads now.
42 - add comppad_name* and hints ($^H, %^H) to per-thread variables.
43 - eg/event was pretty broken.
44 - better 5.8.6 compatibility.
45
46 5.0 Thu Nov 20 10:35:05 CET 2008
47 - NEW ARCHITECTURE: use the latest 4.x version if you experience
48 stability issues.
49 - bump API version to 7 - all dependents must be recompiled.
50 - removed timed_* functions - they were not being used anyways
51 and should be replaced by a more generic mechanism -
52 and were annoying to support anyways :)
53 - removed SemaphoreSet's waiter method - use sem method instead.
54 - Coro::Semaphore->adjust didn't correctly wake up enough waiters.
55 - async_pool did free a scalar value twice
56 ("Attempt to unreference...").
57 - fix a longstanding bug where calling terminate on a coro that
58 was waiting for a semaphore that was just becoming available
59 would cause a deadlock (semaphore would get into a state where
60 it was available but waiters were still blocked).
61 - calling throw on a coroutine that is waiting for a semaphore will
62 no longer make it acquire the semaphore (and thus leak a count).
63 - perl's process emulation is now not even theoretically supported
64 anymore.
65 - new functions Coro::rouse_cb and Coro::rouse_wait for easier
66 conversion of callback-style to blocking-style.
67 - new methods $coro->schedule_to and ->cede_to, to specifically
68 schedule or cede to a specific coroutine.
69 - new function Coro::Semaphore::wait.
70 - use named constants in Coro::Channel (Richard Hundt).
71 - directly patch the entersub opcode calling SLF functions (cede,
72 transfer and so on). this does speed up context switching, but
73 more importanly, it frees us from the hardcoded behaviour of
74 entersub, so we might actually be able to return something from
75 those functions and atcually create new ones.
76 - take advantage of __builtin_frame_address on gcc.
77 - expose THX in coroapi (not sure whether this was a wise decision,
78 as "threaded" perls are running at half speed anyways).
79 - implement execute_slf (schedule-like-function) interface that makes
80 it possible to implement schedule-like-functions in XS.
81 - use new SLF interface to massively speed up Coro::EV by roughly a
82 factor of two.
83 - used new SLF interface to massively speed up Coro::Semaphore by a
84 factor of three.
85 - used new SLF interface to speed up Coro::AIO by roughly a factor of
86 four and reduce its memory usage considerably.
87 - implement Coro::SemaphoreSet purely in terms of Coro::Semaphore,
88 for a nice speedup and vastly more correct behaviour. Also implement
89 a new method "sem" to get at the underlying semaphore object.
90 - implement Coro::Channel in terms of Coro::Semaphore, for a moderate
91 (in comparison) 20-40% speedup.
92 - used new SLF interface to reimplement Coro::Signal gaining
93 some unknown (because I was too lazy), but certain, speedup, and also
94 making signals reliable for the first time.
95 - used new SLF interface and other optimisations to speed up async_pool
96 by a factor of two. It also doesn't rely on perl's exception mechanism
97 to exit anymore. The overhead for terminating an async_pool, coro over
98 a normal async is now very small.
99 - sped up coroutine creation/destruction by 40%.
100 - forgot to include Coro/libcoro/README in the dist for all these years.
101 - work around a freebsd pthreads bug (manual testcancel is required as
102 pthread_cond_wait isn't a cancellation point on freebsd).
103 - use new rouse functions to speed up and simplify Coro::BDB.
104 - make "prefer perl native functions" work with threaded perls.
105 - condense Coro::Debug ps output, hint at v and w flags.
106 - (libcoro) lots of minor cleanups and portability improvements.
107
108 4.914 Wed Nov 19 12:54:18 CET 2008
109 - fix a disastrous bug in the readline optimisation
110 introduced in 4.801.
111
112 4.913 Sat Nov 15 07:58:28 CET 2008
113 - async_pool did free a scalar value twice
114 ("Attempt to unreference...").
115
116 4.912 Thu Nov 13 18:31:23 CET 2008
117 - minor cleanups.
118 - use much larger stacks on linux and perl < 5.8.8.
119 - Coro::Debug::new_unix_server did not unlink the socket
120 when destroyed.
121
122 4.911 Tue Nov 11 04:26:17 CET 2008
123 - "port" to threaded perls.
124
125 4.91 Mon Nov 10 05:36:38 CET 2008
126 - the ->throw exception object no longer leaks.
127 - creating a new cctx leaked a scopestack entry (memleak).
128 - new coroutines didn't get created with a zero flags field
129 (unknown impact).
130 - calling ->throw on a not-yet-started coroutine should now work
131 instead of being ignored.
132 - ->throw is now supported on Coro::State objects.
133 - clean up cctx creation code a bit.
134 - entersub is actually a UNOP, not a LOGOP (not a bugfix).
135
136 4.9 Sat Nov 8 17:45:27 CET 2008
137 - (libcoro) did not preserve rbp with CORO_ASM (we are getting there).
138 - (libcoro) no longer leak threads in the experimental pthread backend,
139 also speed it up considerably.
140 - (libcoro) do not rely on makecontext passing void *'s unscathed.
141 - fix compiletime dependencies on libcoro in the Makefile.
142 - cctx_count wasn't always updated properly.
143 - Coro::State::cctx_stacksize wasn't applied correctly.
144 - new function Coro::State::cctx_max_idle.
145 - the default max number of idle C contexts is now 4.
146 - (libcoro) try harder to get _setjmp/_longjmp.
147 - (libcoro) cleanup and extend the libcoro API to officially
148 allow the creation of empty source contexts.
149 - very experimental workaround for the totally broken netbsd platform.
150 - tried to hack around openbsd bugs.
151
152 4.804 Wed Nov 5 16:36:00 CET 2008
153 - Coro::Debug::new_unix_server would not create a non-blocking listening
154 socket, sometimes causing freezes.
155 - (libcoro) fix misaligned stack points for setjmp and assembly
156 methods, which can cause crashes on x86/x86_64 with a sufficiently
157 aggressive compiler.
158 - new function: Coro::Debug::new_tcp_server.
159 - move ->throw into the Coro class because it only works on coro objects.
160
161 4.803 Mon Nov 3 17:16:12 CET 2008
162 - (libcoro) use a global asm statement to become independent of gcc
163 otpimisations for CORO_ASM (thanks to pippijn for the idea).
164 - try to workaround yet another broken bsd, this time dragonfly.
165
166 4.802 Thu Oct 30 10:56:12 CET 2008
167 - support -fno-omit-frame-pointer on x86 with the assembly method.
168 - tune 01_unblock.t tests a bit.
169
170 4.801 Wed Oct 22 18:33:37 CEST 2008
171 - improve readline speed for very long "lines".
172 - backport to 5.8.8.
173
174 4.8 Thu Oct 2 13:34:40 CEST 2008
175 - new function Coro::AIO::aio_wait.
176 - Coro.:AIO and Coro::BDB now "use Coro::AnyEvent".
177 - greatly speed up and reduce memory usage of Coro::AIO requests.
178 - implement some other ยต-optimisations.
179
180 4.749 Mon Sep 29 14:40:12 CEST 2008
181 - port to slow and broken pseudo-threaded perls. (courtesy pippijn).
182
183 4.748 Sat Sep 27 14:03:19 CEST 2008
184 - implement, but do not document, PerlIO::cede(granularity).
185 - Coro::Storable forgot to wrap Storable::pstore.
186 - work around the multitude of leaks and memory corruption
187 bugs in PerlIO::via by using our own C-level perliol. As a side
188 effect, Coro::Storable is now much, much, much faster.
189
190 4.747 Tue Sep 23 01:59:41 CEST 2008
191 - fix a per-cv memleak (one empty array was leaked per
192 code reference).
193 - avoid a crash in coro->call|rss when the coroutine was already
194 destroyed (most noticably when using Coro::Debug::ps :)
195 - also protect *Storable::FILE.
196 - push up default storable granularity to 20ms.
197
198 4.746 Sun Sep 21 03:22:20 CEST 2008
199 - be more insistent on locking Storable against reentrancy
200 in Coro::Storable.
201 - move swap_def?v and throw to Coro::State, as documented.
202
203 4.745 Thu Jul 24 00:14:38 CEST 2008
204 - remove debugging code related to MgPV_nolen_const, also try to
205 make it compile with perl 5.8.6 (yes, apple apparently loves
206 outdated software). Reported by John S.
207
208 4.744 Tue Jul 8 22:06:35 CEST 2008
209 - correctly provide default DIE/WARN handlers as documented.
210 - also overwrite PL_vtbl_sigelem.svt_clear, even though current
211 implementations inside perl work fine for us.
212
213 4.743 Mon Jun 16 00:21:57 CEST 2008
214 - when using Coro::EV without running EV::loop it could
215 result in busy-waiting for events, this has been fixed.
216 - reduce codesize and improve performance by using EV_DEFAULT_UC.
217
218 4.742 Sat May 31 14:10:21 CEST 2008
219 - implement a workaround for (some) perl <5.8.8 versions.
220 - require EV 3.3+.
221
222 4.741 Fri May 30 23:33:09 CEST 2008
223 - tell netbsd how utterly broken their imitation of an OS is
224 and refuse to build by default if pthreads are in use.
225 - switch to "s" method on all bsd's by default, as their ucontext
226 stuff seems just too broken.
227 - fix a bug in Coro::Select.
228
229 4.74 Thu May 29 20:05:31 CEST 2008
230 - do not test Coro::LWP for lack of dependencies.
231 - do not test Coro::Select for lack of working perls.
232
233 4.73 Thu May 29 2008
234 - fix a bug in Coro::EV which would cause it to block despite
235 there being runnable coroutines.
236 - sprinkle "no warnings" freely over everything, also suppress
237 warnings for some other modules.
238 - fix typo in WSAEWOULDBLOCK.
239
240 4.72 Sun May 25 05:14:36 CEST 2008
241 - tweak META.yaml a bit, unfortunately, there is no documented way
242 to have optional dependencies with CPAN. doh :(
243 - avoid running some tests on windows because they would fail due to
244 perl bug (broken fork, broken pipes...).
245 - work around perl on windows bugs where perl returns undocumented
246 error codes for sysread, syswrite etc. by taking advantage
247 of AnyEvent's workaround for that problem.
248 - use AnyEvent::Util::fh_nonblocking in Coro::Handle to work around
249 a common perl implementation bug on windows.
250 - use unix domain sockets in testsuite to work around
251 common perl implementation bugs on widows (they are emulated by
252 tcp sockets on windows. ugh.)
253
254 4.71 Sat May 24 20:01:27 CEST 2008
255 - fix a bug in Coro::AnyEvent ("Usage: Coro::AnyEvent::_schedule()").
256 - take advantage of async name resolution of AnyEvent::Util.
257 - work around brutal inet_aton override in Coro::LWP.
258 - take advantage of the readyhook in Coro::EV, for smoother
259 scheduling.
260
261 4.7 Sun May 11 00:32:19 CEST 2008
262 - completely reworked the Coro manpage.
263 - added Coro::AnyEvent, generic event loop integration.
264 - implement cancel, ready and kill commands in Coro::Debug.
265 - document find_coro in Coro::Debug.
266 - incompatible change: rename has_stack to has_cctx.
267 - Coro::AIO and Coro::BDB no longer force event model detection,
268 use AnyEvent::AIO and AnyEvent::BDB.
269
270 4.6 Sat Apr 26 10:05:14 CEST 2008
271 - INCOMPATIBLE CHANGE: sub/code attributes are no longer supported
272 by the Coro module. It was a mistake to have it in the first place.
273 - (experimental) support for activestate perl 5.10 (method "w").
274 - (experimental) support for strawberry perl (method "a").
275 - coro_sigelem_set did not return a value although it had to,
276 actual impact unknown.
277
278 4.51 Mon Apr 14 13:28:27 CEST 2008
279 - make it compile again on 5.8.
280
281 4.50 Thu Apr 10 09:43:17 CEST 2008
282 - I did it twice! (see 4.49).
283
284 4.49 Sun Apr 6 21:23:31 CEST 2008
285 - grr, instead of compiling the recent changes on 5.10 only they
286 were compiled on 5.8 only.
287
288 4.48 Sun Apr 6 20:36:46 CEST 2008
289 - allow coroutine switches during eval's under 5.10.x, as it seems
290 the parser is a per-interpreter option now, so this is safe
291 (this might fix the odd crashing bug).
292 - drop support for 5.9.x versions: they are dead, jim.
293
294 4.47 Sun Apr 6 00:37:52 CEST 2008
295 - force cctx allocation on API calls: we know we need to force one
296 and gcc actually manages to confuse our heuristic nowadays,
297 leading to crashes and worse.
298 - document force_cctx.
299
300 4.46 Fri Apr 4 22:05:43 CEST 2008
301 - upgrade libcoro, resulting in pthread-backend (which was only created
302 to fulfill the rules of the programming languages shootout).
303
304 4.45 Thu Mar 13 11:55:36 CET 2008
305 - fix a file leaking bug in eg/mhyttpd that would allow
306 downloading of any file (reported by oesi).
307 - fix deadlock bug in Coro::Channel (reported by Richard Hundt)
308 (also add testcase).
309 - support Broadcast option in Coro::Socket (patch by Richard Hundt,
310 apparently having loads of fun with that).
311 - implement $state->swap_defsv, ->swap_defav and document ->throw.
312
313 4.4 Wed Feb 13 16:44:29 CET 2008
314 - only assume .cfi pseudo isns exist on GNU/Linux.
315 - add get_timed method to Coro::Channels.
316 - fixed Coro::Debug manpage.
317 - perl 5.11 compatbility improvement based on patch by
318 Andreas Kรถnig.
319
320 4.37 Sun Jan 20 11:25:23 CET 2008
321 - soften the check for an active parser for perl 5.10 (avoiding
322 "Coro::State::transfer called while parsing" in many cases).
323
324 4.36 Sun Jan 13 10:53:56 CET 2008
325 - reset diehook when terminating from an async_pool as to not
326 trigger any __DIE__ handlers.
327
328 4.35 Sun Jan 13 04:14:13 CET 2008
329 - "bt" debug command now displays any exceptions
330 from longmess and also skips the innermost
331 stackframes, giving more useufl output.
332 - allow backtraces in coroutines blocked in Coro::EV,
333 at a <1% speed hit.
334 - handle localising of $SIG{__WARN__} and __DIE__
335 properly (with a proper amount of dirty hacking).
336
337 4.34 Sat Dec 22 17:49:53 CET 2007
338 - upgrade to EV version 2.0 API.
339
340 4.33 Mon Dec 17 08:36:12 CET 2007
341 - make Coro::AIO etc. loadable in the absence of EV.
342
343 4.32 Mon Dec 17 07:46:02 CET 2007
344 - majorly improved Coro::Handle's performance with Coro::EV.
345 - implemented the until now mythical Coro::BDB module.
346 - specialcase EV in Coro::AIO and Coro::BDB for extra speed.
347
348 4.31 Wed Dec 5 12:32:39 CET 2007
349 - remove warn statement form Coro::Util.
350
351 4.3 Tue Dec 4 20:33:14 CET 2007
352 - calls to the idle function could cause stack corruption
353 when the stack changed.
354 - do no longer rely on the presence of EV::DNS if EV is used
355 (because it is gone), but instead take avdantage of EV::ADNS
356 if available.
357 - add ($) prototypes to all functions in Coro::Storable.
358 - use a conventional (safer) idle callback in Coro::EV.
359 - do accept quickly in Coro::Debug to avoid endless loops.
360
361 4.22 Fri Nov 30 16:04:04 CET 2007
362 - really use optimised versions for Event and EV in Coro::Util
363 and Coro::Handle.
364
365 4.21 Sun Nov 25 10:48:59 CET 2007
366 - fix a spurious memory read.
367 - Coro::EV no longer keeps the eventloop "alive".
368
369 4.2 Fri Nov 9 20:47:05 CET 2007
370 - enable/disable tracing from a new coroutine, not a pooled one.
371 - fix a memleak in Coro::Event.
372 - removed killall call from fork_eval.
373 - made sure store_fd is already loaded so that fork_eval
374 does not have to parse autoload in each subprocess.
375 - only use assembly method if -O switch is in $Config{optimize}.
376 - add (optional) Coro::EV module, so far the best event loop module
377 directly supported by Coro.
378 - if the event model is EV, use EV::DNS to resolve
379 stuff in Coro::Util.
380 - don't get confused by multiple event notifications in Coro::Handle.
381 - initial support for EV (libevent interface).
382 - require Event and EV using configure_requires, to force their existance.
383
384 4.13 Wed Oct 24 07:26:45 CEST 2007
385 - add Coro::Storable::blocking_thaw.
386 - use a vastly more complicated technique to localise
387 $SIG{__WARN/DIE__} that also works on perls <= 5.8.8.
388 - use a coroutine for the idle callback Coro::Event,
389 instead of running Event in the current coroutine context.
390 This also catches recursive invocations.
391 - actually report fork errors in gethostbyname and inet_aton.
392
393 4.11 Thu Oct 11 02:40:24 CEST 2007
394 - port to threaded perls.
395
396 4.1 Thu Oct 11 02:38:16 CEST 2007
397 - incompatible change: $SIG{__DIE__} and $SIG{__WARN__} will now
398 be local to each coro (see Coro::State).
399 - incompatible change: for very deep reasons, cede and cede_notself
400 cannot return anything, so nothing will be returned.
401 - possibly bring back 5.10 compatibility (untested).
402 - work around stupid (and wrong) warning on 5.10 :(.
403 - overlay the saved state over the context stack. This saves
404 a few hundred bytes per coroutine on average and also
405 speeds up context switching a bit.
406 - further tune default stack sizes.
407 - (more) correctly calculate stack usage in coro_rss.
408 - Coro::Storable::blocking_* did not properly lock
409 resulting in races between coroutines.
410 - added Coro::Storable::guard.
411 - stopping to trace a coroutine could destroy the cctx of
412 an unrelated coroutine.
413 - explain the relationship between Perl and C coroutines in
414 more detail in Coro::State.
415 - Coro::Util::inet_aton did not short-circuit dotted quad forms,
416 causing a fork per resolve. This also affected Coro::Socket.
417 - switch to a separate stack in $coro->call/eval to avoid
418 invalidating pointers.
419
420 4.03 Sat Oct 6 21:24:00 CEST 2007
421 - added Coro::throw method.
422 - minor code cleanups.
423
424 4.02 Sat Oct 6 02:36:47 CEST 2007
425 - fix a very minor per-coroutine memleak (a single codereference).
426 - fixed a bug where the currently in-use c context would be freed
427 prematurely (can happen only when programs change the stacksize
428 or use tracing).
429 - tracing can no longer keep a coro alive after it terminated.
430 - do static branch prediction in the common path for gcc. gives
431 about 2-5% speed improvement here.
432
433 4.01 Fri Oct 5 22:10:49 CEST 2007
434 - instead of recreating *a* standard output handle we simply
435 use STDOUT, which is faster and hopefully more robust.
436
437 4.0 Fri Oct 5 12:56:00 CEST 2007
438 - incompatibly changed Coro::Storable::freeze.
439 - major new feature: added Coro::Debug, for interactive coroutine
440 debugging, tracing and much more.
441 - major bug fix: unbelievable, but true: $_, $/ and many other
442 "saved" variables actually weren't being saved. This has been fixed,
443 of course, while increasing performance while losing all the save
444 flags.
445 - save flags are gone, and all the api functions dealing with them.
446 - added Coro::Semaphore::adjust.
447 - added Coro::Util::fork_eval.
448 - added Coro::Storable::{nfreeze,blocking_{freeze,nfreeze}}.
449 - added Coro::killall.
450 - reduce initial stack sizes to allow for "micro-coroutines".
451 - better async_pool resource management, moved parts of async_pool
452 handling to XS (major speed improvement).
453 - actually croak before modifying important data structures.
454 - refuse to transfer while compiling.
455 - possibly support eval EXPR better now.
456 - enable assembly per default on linux+bsd x86+amd64.
457 - all internal members were renamed _something for easier subclassing.
458 - many minor tweaks.
459
460 3.63 Wed May 16 14:10:06 CEST 2007
461 - implement handcoded assembly for x86/amd64 SVR ABI.
462
463 3.62 Fri Apr 27 21:36:06 CEST 2007
464 - upgrade libcoro (which might set unwind info correctly).
465 - change default on linux to setjmp/longjmp.
466
467 3.61 Thu Apr 19 12:36:18 CEST 2007
468 - Coro::Storable caused an endless loop when thawing invalid
469 pst's sometimes.
470 - use a Semaphore in Coro::Storable, as Storable doesn't
471 seem to be reentrant (although it is documented to
472 be threadsafe...).
473 - fix Coro::Signal to bring back the original unreliable
474 but stateful semantics.
475 - fixed a lot of typos in Coro.pm (patch submitted by David
476 Steinbrunner, which applied flawlessly).
477
478 3.6 Sat Apr 14 17:13:31 CEST 2007
479 - added some bugfixes to get eg/myhttpd working again.
480 - added Coro::Storable for often-cede'ing freeze/thaw.
481 - try to do a clean exit when a coroutine calls exit
482 (EXPERIMENTAL).
483 - got rid of indirect call through _coro_init.
484 - updated the partly antique examples in eg/ to
485 work again and be a bit less magic, too.
486 - fixed Coro::Signal semantics to work as documented again.
487
488 3.55 Sun Mar 25 01:20:47 CET 2007
489 - add SAVE_DEFFH to save the default filehandle and enable
490 it by default.
491 - finally move socket-operations from Coro::Socket to Coro::Handle
492 to be able to unblock foreign sockets.
493 - add Coro::State::save_also and guarded_save.
494 - add count accessor to Coro::Semaphore.
495 - add Coro::State::cctx_stacksize.
496 - just for the fun of it, do not rely on implicit context,
497 which can dramatically improve performance, but people
498 using windows-process-emulation perls do not care much about
499 performance.
500
501 3.51 Sun Mar 4 14:18:04 CET 2007
502 - fixed a problem when you weakened references to Coro::State's
503 (patch by Yuval Kogman).
504
505 3.501 Wed Feb 28 12:44:07 CET 2007
506 - rename some global symbols as macosx from hell redefines
507 them without asking.
508
509 3.5 Tue Feb 13 20:22:53 CET 2007
510 - do AnyEvent model detection earlier, avoiding problems
511 caused by first using AnyEvent and later Coro::Event.
512 - implement and document Coro::Event event objects.
513 - fix a potential problem in Coro::Event causing crashes.
514 - initialise PL_comppad when creating a new coroutine,
515 avoids crashes on early coro destruction.
516
517 3.41 Mon Jan 22 19:19:49 CET 2007
518 - readline on Coro::Handle did not support undefined $/,
519 nor did it deliver partial lines on EOF or error.
520 - implement malloc fallback for stack allocation because
521 stupid broken idiotic OSX has a stupid broken
522 idiotic fits-the-whole-os mmap "implementation" and
523 my dick feels longer if Coro is portable even to
524 obsolete platforms.
525
526 3.4 Fri Jan 19 21:52:54 CET 2007
527 - remove t/09_timer.t, as it isn't really testing much
528 but was rather flaky in practise.
529 - async_pool coro would keep arguments and callback alive until
530 it was reused.
531 - cancellation of a coroutine could cause spurious idle calls
532 in cede_notself.
533
534 3.3 Sat Jan 6 03:45:00 CET 2007
535 - implement $coro->on_destroy.
536 - Coro::Event blocking semantics have been changed,
537 documented and - hopefully - improved.
538 - fix nice adding, not subtracting, from priority.
539 - fix ->prio and api_is_ready (patch by Mark Hinds).
540 - fixed an assert ("... == prev__cctx->idle_te")
541 that could errronously trigger.
542 - fix various large and small memleaks.
543 - use a (hopefully) more stable cancel implementation
544 that immediately frees the coroutine data.
545 - cede/cede_notself return a status now.
546 - added Coro::guard function.
547 - added a global coroutine pool for jobs (on my machine,
548 I can create and execute 48k simple coros/s with async,
549 and 128k coros with async_pool).
550 - Coro::AIO now uses the coroutine priority as io priority.
551
552 3.2 Fri Dec 22 05:07:09 CET 2006
553 - improve portability to slightly older perls.
554 - use cleaner coroutine destruction.
555 - simplify configuration for users.
556 - optionally (unrecommended) prefer perl functions over
557 their coro replacements.
558
559 3.11 Tue Dec 5 13:11:24 CET 2006
560 - fixed some bogus assert's, but as perl.h disables assert even
561 without NDEBUG (thank you very much), not too many people should
562 notice (that did include myself). Andreas Kรถnig noticed, though :)
563 - do not save/restore PL_sortcxix on >= 5.9.x, it doesn't seem to have
564 it. Also noticed by Andreas Kรถnig :)
565 - save/restore tainted status.
566 - verified to pass the testsuite on my 5.9.5.
567
568 3.1 Mon Dec 4 23:03:40 CET 2006
569 - INCOMPATIBLE CHANGE: $/ is now per-coroutine (but slow).
570 - incompatible change: transfer flags are now per-state.
571 - give a better error message on deadlock.
572 - document Coro::nready.
573 - enhanced testsuite.
574
575 3.01 Sun Dec 3 23:47:42 CET 2006
576 - forgot to include Coro::Timer.
577
578 3.0 Sun Dec 3 22:57:25 CET 2006
579 - the "FINALLY COMPLETELY STABLE" release (coming soon:
580 the "FAMOUS LAST WORDS" release).
581 - implement a new stack sharing algorithm, which uses a stack
582 pool (size currently hardcoded).
583 - make stack sharing mandatory (it no longer uses a heuristic).
584 - eval/die no longer cause weird problems under heavy use.
585 - Coro::Event could cause livelocks if it was used but
586 no Coro::Event watchers were used.
587 - Coro::Event now uses asynccheck as prepare does not
588 check for changed watchers.
589 - Coro::Event allows multiple waiting coros and will wake up one
590 per event.
591 - Coro::Event should be cleaner and more efficient now.
592 - new utility function Coro::unblock_sub.
593 - document the sad fact that Event is no longer reentrant.
594 - putting a coroutine into the ready queue twice could under
595 some circumstances lead to stack corruption.
596 - minor incompatibility: subclassing Coro::State is supported
597 directly now without going through a _coro_state member.
598 - state/coro switching is much faster now.
599 - very minor optimisations and code/documentation cleanup.
600 - avoid problems due to compiler inlining.
601 - removed timers from Coro::Timer -> use AnyEvent instead.
602 - replaced Coro::idle coroutine by (cleaner) idle handler.
603 - updated to newest libcoro.
604 - implement enhanced support for valgrind.
605 - implement is_ready and return value for ready.
606 - removed Coro::Cont, it was a misnomer (it's generators),
607 and it was somewhat annoying to get it right. Will come back
608 if somebody asks for it :)
609 - many ->wait methods and Coro::Event could return
610 spuriously without the event having happened.
611
612 2.5 Tue Nov 7 12:22:33 CET 2006
613 - made Coro::Util, Coro::Select, Coro::Handle and Coro::Socket
614 use AnyEvent, moved them to Coro/.
615 - added Coro::LWP which contains all the uglyness required to
616 make LWP non-blocking.
617 - should work with perl 5.9.x now (Andreas Kรถnig made me do it).
618 - fixed another bug in Coro::Select when the timeout was undef.
619 - reuse PL_start_env for all coros, saves some memory per coroutine.
620 - manage PL_top_env differently, hopefully to avoid panic: top_env.
621 - timeout argument was not properly used in Coro::Socket.
622 - allow limited forms of subclassing in Coro::Handle/Coro::Socket.
623 - emulate undocumented(!) functionality of IO::Socket required
624 by LWP(!!).
625 - updated eg/lwp to work with newer lwp's.
626 - remove "FATAL: uncaught exception" prefix. Coroutines that die
627 kill the whole process, just as exceptions in the main "coroutine"
628 did already.
629
630 2.1 Wed Nov 1 23:01:13 CET 2006
631 - fix a long-standing bug in Coro::Select where select with
632 zero timeout would instead change the current default filehandle.
633 - use a simpler and hopefully more robust way to clone padlists
634 (uses less memory and a perl function instead of our own).
635 - coro can now create a stack guard area on many architectures.
636 - Coro::AIO properly reexports additional functions from IO::AIO.
637 - updated libcoro with a workaround for OS X,
638 pach and testing by Michael Schwern.
639
640 2.0 Tue Oct 24 05:47:17 CEST 2006
641 - support additional aio requests in Coro::AIO.
642
643 1.9 (never properly released due to a glitch)
644
645 1.8 Thu Feb 2 00:59:06 CET 2006
646 - applied suggested patch by SAMV to avoid problems during stupid
647 mark & sweep gc run.
648 - applied patch by Scott Walters for 5.9.3 compatibility.
649
650 1.7 Tue Dec 27 01:41:58 CET 2005
651 - added Coro::AIO, a thin wrapper around IO::AIO.
652 - improved Makefile.PL explanations.
653
654 1.6 Mon Dec 12 21:30:05 CET 2005
655 - additionally save PM_curpm might fix as-of-yet
656 unseen problems with regex matches being attributed to the
657 wrong package.
658 - add t/10_bugs.t, which currently checks against imho broken
659 perls that use 0.26MB of stack space per Perl_magic_get
660 invocation (newer linuxes) as opposed to the 0.0002MB perl
661 normally uses.
662 - make stacksize configurable for the ultimate debian experience.
663
664 1.51 Mon Dec 12 18:48:36 CET 2005
665 - remove debugging warn() accidentally left in Coro::Select.
666
667 1.5 Tue Nov 29 13:32:44 CET 2005
668 - use Coro::Event inside Coro::Select to avoid spurious deadlocks.
669 - fix Coro::Select 'select'.
670 - strict'ify some modules.
671 - libcoro errornously restored the SIGUSR2 handler to SIGUSR1.
672 - use XSLoader in selected modules.
673 - remove some 5.6 compatibility cruft.
674
675 1.4 Tue Sep 6 00:11:05 CEST 2005
676 - libcoro did not take into account the trampoline on amd64,
677 when 'l'inux method was used.
678
679 1.31 Tue Aug 30 23:31:33 CEST 2005
680 - some portability fixes/workarounds.
681
682 1.3 Sat Aug 20 03:08:56 CEST 2005
683 - no code changes. module seems to work fine.
684 - improve Coro::State docs, remove reference to nonexisting
685 Coro::State::flush method.
686 - no longer autodetect windows, present it as an option instead.
687
688 1.2 Mon May 16 02:00:55 CEST 2005
689 - included libcoro.c earlier in State.xs, might improve portability.
690 - use a faster and possibly more stable Coro::Cont implementation.
691 - accept x86_64 in addition to amd64 for optimized linux-amd64 support.
692 - fix bugs Coro::SemaphoreSet that could cause locks never to be freed.
693 - fix bugs in CoroAPI.h and document it in Coro::MakeMaker.
694
695 1.11 Thu Mar 3 18:00:52 CET 2005
696 - change of contact address.
697
698 1.1 Tue Feb 22 20:51:16 CET 2005
699 - support [l]inux method on amd64.
700 - allow some context switches while compiling/eval'ing.
701
702 1.0 Tue Aug 10 03:47:27 CEST 2004
703 - Event 0.89 fixes the bug again.
704 - fix the remaining(??) memleak. I didn't free the cache-AV for
705 padlists, so one AV per created anon-sub leaked.
706
707 0.97 Fri May 14 15:23:32 CEST 2004
708 - remove stupid gcc extension that was an experiment only anyway.
709
710 0.96 Thu May 13 18:09:29 CEST 2004
711 - tell people if they have yet another broken Event variant.
712 - made Coro to work _better_ with threads (it seems that, as long
713 as you use Coros in one thread only everything will work fine.
714 Looking at the perl source the problem might be missing locking
715 between Coro and some perl internal routines. That's not easily
716 fixable).
717 - (hopefully) got rid of the static xnext inside transfer.
718
719 0.95 Sun Feb 15 03:20:28 CET 2004
720 - removed Coro::State::flush, implemented a new and much
721 faster caching mechanism by attaching magic to the CV.
722 - a memleak with (real) closures remains.
723 - dramatically reduced size of stacks to 65536 (linux/x86), after
724 observing that even myhttpd never had more than a kilobyte
725 of stack in use (stack is still only physically allocated in
726 pagesize increments for systems suporting mmap).
727 - fix for a "cygwin-64" architecture added, proposed by Stefan Traby.
728 - fix for threaded perl.
729
730 0.9 Sun Nov 23 04:32:39 CET 2003
731 - fix a few very hard to track down but (of course) stupid
732 memory leaks.
733 - stringify version number to avoid locale problems :(.
734 - perl 5.9.0 does not have SvPADBUSY, reported by Scott Walters,
735 hopefully ignoring it "just works".
736
737 0.8 Wed Nov 5 19:38:40 CET 2003
738 - port to perl5.005_03, but only 5.8.x+ is supported!
739 - honor LocalAddr even without LocalPort in Coro::Socket.
740
741 0.7 Tue May 27 03:12:38 CEST 2003
742 - the version jump indicates some level of testing, not gobs
743 of new features.
744 - uh... I found the prompt function in ExtUtils::MakeMaker.
745 Highly correct stuff, that is...
746 - fixed(?) a bug with die's in coroutines causing "panic: top_env".
747 the fix is not well understood by the author, so beware :(.
748
749 0.652 Thu May 8 02:54:46 CEST 2003
750 - Applied patch by Slaven Rezic to set default to "s" on FreeBSD,
751 cause version 4 doesn't have ucontext.
752 - Benjamin Reed reported that setjmp works fine on darwin,
753 so preselect it.
754
755 0.651 Sat Mar 29 15:00:23 CET 2003
756 - fix a bug in Coro::Handle where some bytes could get lost
757 on reading, reported by jason@nichego.net.
758
759 0.65 Sun Mar 23 00:08:26 CET 2003
760 - added a README.linux-glibc.
761 - new module Coro::Select.
762 - also save/restore PL_comppad, fixes segfaults with 5.8.1.
763 (I never claimed that I know what I am doing ;).
764 - set default to setjmp/longjmp on non-x86-linux.
765
766 0.6 Thu Nov 21 11:09:06 CET 2002
767 - made lazy context switching the default.
768 - fixed the bug where SemaphoreSet::guard
769 was the same as timed_guard and timed_guard was missing.
770 - fix a memleak in Coro::Timer.
771 - reclassified context sharing as not experimental.
772
773 0.534 Sun Apr 14 03:05:12 CEST 2002
774 - fixed backspace => tab in header parsing (myhttpd).
775 - added eg/readline, for elmex.
776 - Coro::Event now calls ->start on first call instead of
777 ->again, which created wrong timeouts for the first ->next.
778 - fixed a bug where Coro::Socket returned a closed socket instead
779 of nothing on ECONNREFUSED and similar error conditions.
780
781 0.533 Mon Feb 18 18:49:40 CET 2002
782 - adapt to finally working Event-0.85.
783 - fixed buggy event initialization (trapped under windows, but how
784 could it possibly have worked anywhere?)
785
786 0.532 Wed Jan 16 02:45:32 CET 2002
787 - removed the reference to pp_entersub, might work on
788 (some) win32 perls now (testsuite works under cygwin,
789 without Event, which freezes).
790
791 0.531 Mon Dec 10 22:18:44 CET 2001
792 - Fixed a bug in SemaphoreSet::timed_down.
793
794 0.53 Tue Nov 27 21:11:13 CET 2001
795 - some tests for Event.
796 - slightly different internal architecture to get refcnt into the object
797 and not the reference. might fix a few bugs, certainly did introduce
798 new ones.
799 - Coro::Timer (independent of Event).
800 - new timed_wait functions for Coro::Signal, Semaphore, SemaphoreSet.
801
802 0.52 Tue Nov 6 21:36:18 CET 2001
803 - ported to cygwin (trivial).
804 patch by Gerrit P. Haase <gp@familiehaase.de>.
805 - small setjmp code fix by Sullivan.DanielJ@epamail.epa.gov.
806
807 0.51 Thu Nov 1 20:39:01 CET 2001
808 - terminate/cancel now work properly (otherwise termination
809 could cause a "next coroutine is not and contains not..."
810 error.
811 - added Coro::Socket::shutdown.
812 - Coro::Event::loop no is the same as Event::loop.
813 - implemented terminate with args + join.
814
815 0.5 Fri Sep 28 16:15:35 CEST 2001
816 - fixed "print" on a Coro::Handle. The print method worked.
817 - small tweaks (seem to reduce memory consumption a lot)
818 in various modules.
819 - splendid use of "no warnings" scattered throughout.
820 - added Coro::Handle::rbuf, fixed Coro::Handle::fh.
821
822 0.49 Sun Sep 16 02:42:45 CEST 2001
823 - changed some method calls to function calls for speed
824 inside Coro::Handle.
825 - make Coro::Handle use an array instead of a hash for
826 speed reasons.
827 - IRIX mystery solved: it's SGI's NT, after all: "standard, huh?".
828 sjlj and ucontext should now work.
829 - IRIX-specific port for libcoro.
830 - swapped order of accept results to match IO::Socket.
831 - changed getsock/peername to sock/peername to match IO::Socket.
832 - fixed a bug that caused segfault when returning to main task
833 under some circumstances.
834 - other bugfixes.
835
836 0.45 Sun Sep 2 02:54:01 CEST 2001
837 - new method Coro::Handle::timeout.
838 - corrected speling of set...name to get...name in Coro::Socket.
839 - Coro::Socket::accept now returns a Coro::Socket, not a Coro::Handle.
840 - Coro::Handle now supports fileno and FILENO.
841 - added eg/myhttpd, a web server.
842 - Coro::Socket now accepts numeric ports as well as "port(number)"
843 syntax.
844 - moved some scheduling primitives into xs code. more to come.
845 - new simple priority system.
846 - implemented Coro::Event using XS, almost four times faster!
847 - small memory corruption problem fixed (boy that was difficult).
848
849 0.13 Wed Aug 8 16:53:07 CEST 2001
850 - new method Coro::Handle::readline.
851 - added eg/lwp and Coro::Handle::autoflush, to make LWP work.
852 (see eg/lwp on how to make LWP non-blocking).
853 - renamed Coro::Socket::new_inet to new.
854 - added Coro::Util, some utility functions.
855 - die/eval now works better (still get annoying (but true ;)
856 "Callback called exit" messages).
857 - Coro::Handle now supports timeout, and Coro::Socket Timeout.
858 - much better Coro::Socket support.
859
860 0.12 Fri Jul 27 04:19:01 CEST 2001
861 - do not use mmap if neither MAP_ANON nor MAP_ANONYMOUS is defined.
862 - very experimental stack sharing algorithm. not 100% safe but
863 should work well in practise ;)
864 - added Coro::RWLock.
865 - Coro::Specific now works.
866
867 0.11 Tue Jul 24 22:49:21 CEST 2001
868 - added specialized hack for newer and older linux versions (fast).
869 - renamed Coro::Event::IO to Coro::Handle.
870 - new module Coro::Socket.
871
872 0.1 Tue Jul 24 01:47:53 CEST 2001
873 - release candidate 3. A rather new internal structure :(
874 - the great renaming: Damian Conway gave me a suitable replacement
875 for yield.
876 - added Coro/libcoro, a portable coroutine implementation for C
877 (not even perl-dependent!!!), which can be used by Coro::State.
878 - renamed SAVE_DEFSV (on xs-level only) to avoid symbol clash in
879 perl-5.6.
880 - new function Coro::Event::idle.
881 - the idle process is now overriden
882 by default in Coro::Event.
883 - Coro::Channel now enforces the size.
884 - canceling events now works in all cases.
885 - Coro state now include $_ and $@.
886 - yet another bug workaround that I do not really understand :(
887 - new module Coro::Event::IO (very undocumented) to do
888 non-blocking i/o.
889 - performance tuning.
890
891 0.08 Thu Jul 19 06:13:25 CEST 2001
892 - release candidate 2 (new functionality).
893 - add Coro::State::flush function.
894 - transfer now supports a flags value (mostly for speed ATM).
895 - might compile and work in the presence of threads now.
896 - continuations have a saner syntax.
897 - no more memleaks.
898
899 0.07 Tue Jul 17 17:40:18 CEST 2001
900 - release candidate 1 ;)
901 - slightly nicer code.
902 - fixed a scheduling bug in Coro::Event.
903
904 0.06 Tue Jul 17 04:23:24 CEST 2001
905 - ok, I found the showstopper - the same sub must not be
906 re-entered in two different coroutines, otherwise => crash. I
907 see no easy solution to this problem, except by walking the call
908 chaing and saving/restoring the cv's, which is what I do now.
909 - memory leaks still latent, especially at thread termination.
910 - Coro::Event now works (haha).
911
912 0.05 Sun Jul 15 17:32:20 CEST 2001
913 - fixed a few issues in Coro::Event.
914 - I forgot to include Coro::Event in 0.04 :(:(:(
915
916 0.04 Sun Jul 15 05:24:59 CEST 2001
917 - @_ is now properly localized.
918 - Coro::State is now easier subclassable.
919 - Coro::Cont now coroutine-aware.
920 - Coro::Specific is a low-overhead module to create
921 coroutine-specific vars.
922 - Coro::Event provides a simple interface to Event.
923
924 0.03 Fri Jul 13 14:51:52 CEST 2001
925 - transfer() now implemented in XS (beware).
926 - new module Coro::Cont for really faked continuations.
927 - big internal architecture changes: Coro::State is now
928 really low-level and can thus be used to implement other
929 interesting things, While "Coro::" implements a process-like
930 model. Still crude and subject to change.
931 - $_ and $@ are no longer being localized.
932
933 0.02 Tue Jul 10 01:38:17 CEST 2001
934 - implemented "async" attribute.
935 - $_ and $@ are now localized.
936 - added Coro::Channel.
937 - more testcases, still no docs.
938
939 0.01 Tue Jul 3 02:18:41 CEST 2001
940 - original version; copied from Convert::Scalar.
941