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

Comparing Coro/Changes (file contents):
Revision 1.39 by root, Sat Aug 11 23:10:56 2001 UTC vs.
Revision 1.461 by root, Mon Nov 24 04:56:38 2008 UTC

1Revision history for Perl extension Coro. 1Revision history for Perl extension Coro.
2 2
30.45 3TODO: should explore PerlIO::coroaio (perl leaks like hell).
4TODO: 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 TODO: Coro/Intro
9 - wrote a small introduction - Coro::Intro.
10 - convert Coro::Timer, Coro::Select and Coro::Util to rouse API.
11 - Coro::Select did errornously dup the file descriptors
12 and didn't work with all AnyEvent backends.
13 - disassociate c contexts from coro objects - this is agruably more
14 correct, but mostly allows sharing of cctxs between coro and state
15 objects, for added memory savings and speed increases.
16 - bumped $Coro::POOL_RSS up to 32kb by default.
17 - no longer set the optype to OP_CUSTOM, as B::* understandably
18 doesn't like this very much (and we *are* a type of entersub).
19 - implement state cloning, just to prove that call/cc can be done.
20 - automatically load Coro::AnyEvent in Coro::Handle.
21 - wrap ->cancel calls in eval inside Coro::Handle as EV watchers
22 do not have this method (and don't need it either).
23 - speed up generic anyevent methods in Coro::Handle by using rouse
24 callbacks.
25 - allow coroutines in $Coro::IDLE, speeding up Coro::AnyEvent and
26 others. It also makes the debugger happier, as you can trace
27 through the idle threads now.
28 - eg/event was pretty broken.
29 - better 5.8.6 compatibility.
30
315.0 Thu Nov 20 10:35:05 CET 2008
32 - NEW ARCHITECTURE: use the latest 4.x version if you experience
33 stability issues.
34 - bump API version to 7 - all dependents must be recompiled.
35 - removed timed_* functions - they were not being used anyways
36 and should be replaced by a more generic mechanism -
37 and were annoying to support anyways :)
38 - removed SemaphoreSet's waiter method - use sem method instead.
39 - Coro::Semaphore->adjust didn't correctly wake up enough waiters.
40 - async_pool did free a scalar value twice
41 ("Attempt to unreference...").
42 - fix a longstanding bug where calling terminate on a coro that
43 was waiting for a semaphore that was just becoming available
44 would cause a deadlock (semaphore would get into a state where
45 it was available but waiters were still blocked).
46 - calling throw on a coroutine that is waiting for a semaphore will
47 no longer make it acquire the semaphore (and thus leak a count).
48 - perl's process emulation is now not even theoretically supported
49 anymore.
50 - new functions Coro::rouse_cb and Coro::rouse_wait for easier
51 conversion of callback-style to blocking-style.
52 - new methods $coro->schedule_to and ->cede_to, to specifically
53 schedule or cede to a specific coroutine.
54 - new function Coro::Semaphore::wait.
55 - use named constants in Coro::Channel (Richard Hundt).
56 - directly patch the entersub opcode calling SLF functions (cede,
57 transfer and so on). this does speed up context switching, but
58 more importanly, it frees us from the hardcoded behaviour of
59 entersub, so we might actually be able to return something from
60 those functions and atcually create new ones.
61 - take advantage of __builtin_frame_address on gcc.
62 - expose THX in coroapi (not sure whether this was a wise decision,
63 as "threaded" perls are running at half speed anyways).
64 - implement execute_slf (schedule-like-function) interface that makes
65 it possible to implement schedule-like-functions in XS.
66 - use new SLF interface to massively speed up Coro::EV by roughly a
67 factor of two.
68 - used new SLF interface to massively speed up Coro::Semaphore by a
69 factor of three.
70 - used new SLF interface to speed up Coro::AIO by roughly a factor of
71 four and reduce its memory usage considerably.
72 - implement Coro::SemaphoreSet purely in terms of Coro::Semaphore,
73 for a nice speedup and vastly more correct behaviour. Also implement
74 a new method "sem" to get at the underlying semaphore object.
75 - implement Coro::Channel in terms of Coro::Semaphore, for a moderate
76 (in comparison) 20-40% speedup.
77 - used new SLF interface to reimplement Coro::Signal gaining
78 some unknown (because I was too lazy), but certain, speedup, and also
79 making signals reliable for the first time.
80 - used new SLF interface and other optimisations to speed up async_pool
81 by a factor of two. It also doesn't rely on perl's exception mechanism
82 to exit anymore. The overhead for terminating an async_pool, coro over
83 a normal async is now very small.
84 - sped up coroutine creation/destruction by 40%.
85 - forgot to include Coro/libcoro/README in the dist for all these years.
86 - work around a freebsd pthreads bug (manual testcancel is required as
87 pthread_cond_wait isn't a cancellation point on freebsd).
88 - use new rouse functions to speed up and simplify Coro::BDB.
89 - make "prefer perl native functions" work with threaded perls.
90 - condense Coro::Debug ps output, hint at v and w flags.
91 - (libcoro) lots of minor cleanups and portability improvements.
92
934.914 Wed Nov 19 12:54:18 CET 2008
94 - fix a disastrous bug in the readline optimisation
95 introduced in 4.801.
96
974.913 Sat Nov 15 07:58:28 CET 2008
98 - async_pool did free a scalar value twice
99 ("Attempt to unreference...").
100
1014.912 Thu Nov 13 18:31:23 CET 2008
102 - minor cleanups.
103 - use much larger stacks on linux and perl < 5.8.8.
104 - Coro::Debug::new_unix_server did not unlink the socket
105 when destroyed.
106
1074.911 Tue Nov 11 04:26:17 CET 2008
108 - "port" to threaded perls.
109
1104.91 Mon Nov 10 05:36:38 CET 2008
111 - the ->throw exception object no longer leaks.
112 - creating a new cctx leaked a scopestack entry (memleak).
113 - new coroutines didn't get created with a zero flags field
114 (unknown impact).
115 - calling ->throw on a not-yet-started coroutine should now work
116 instead of being ignored.
117 - ->throw is now supported on Coro::State objects.
118 - clean up cctx creation code a bit.
119 - entersub is actually a UNOP, not a LOGOP (not a bugfix).
120
1214.9 Sat Nov 8 17:45:27 CET 2008
122 - (libcoro) did not preserve rbp with CORO_ASM (we are getting there).
123 - (libcoro) no longer leak threads in the experimental pthread backend,
124 also speed it up considerably.
125 - (libcoro) do not rely on makecontext passing void *'s unscathed.
126 - fix compiletime dependencies on libcoro in the Makefile.
127 - cctx_count wasn't always updated properly.
128 - Coro::State::cctx_stacksize wasn't applied correctly.
129 - new function Coro::State::cctx_max_idle.
130 - the default max number of idle C contexts is now 4.
131 - (libcoro) try harder to get _setjmp/_longjmp.
132 - (libcoro) cleanup and extend the libcoro API to officially
133 allow the creation of empty source contexts.
134 - very experimental workaround for the totally broken netbsd platform.
135 - tried to hack around openbsd bugs.
136
1374.804 Wed Nov 5 16:36:00 CET 2008
138 - Coro::Debug::new_unix_server would not create a non-blocking listening
139 socket, sometimes causing freezes.
140 - (libcoro) fix misaligned stack points for setjmp and assembly
141 methods, which can cause crashes on x86/x86_64 with a sufficiently
142 aggressive compiler.
143 - new function: Coro::Debug::new_tcp_server.
144 - move ->throw into the Coro class because it only works on coro objects.
145
1464.803 Mon Nov 3 17:16:12 CET 2008
147 - (libcoro) use a global asm statement to become independent of gcc
148 otpimisations for CORO_ASM (thanks to pippijn for the idea).
149 - try to workaround yet another broken bsd, this time dragonfly.
150
1514.802 Thu Oct 30 10:56:12 CET 2008
152 - support -fno-omit-frame-pointer on x86 with the assembly method.
153 - tune 01_unblock.t tests a bit.
154
1554.801 Wed Oct 22 18:33:37 CEST 2008
156 - improve readline speed for very long "lines".
157 - backport to 5.8.8.
158
1594.8 Thu Oct 2 13:34:40 CEST 2008
160 - new function Coro::AIO::aio_wait.
161 - Coro.:AIO and Coro::BDB now "use Coro::AnyEvent".
162 - greatly speed up and reduce memory usage of Coro::AIO requests.
163 - implement some other µ-optimisations.
164
1654.749 Mon Sep 29 14:40:12 CEST 2008
166 - port to slow and broken pseudo-threaded perls. (courtesy pippijn).
167
1684.748 Sat Sep 27 14:03:19 CEST 2008
169 - implement, but do not document, PerlIO::cede(granularity).
170 - Coro::Storable forgot to wrap Storable::pstore.
171 - work around the multitude of leaks and memory corruption
172 bugs in PerlIO::via by using our own C-level perliol. As a side
173 effect, Coro::Storable is now much, much, much faster.
174
1754.747 Tue Sep 23 01:59:41 CEST 2008
176 - fix a per-cv memleak (one empty array was leaked per
177 code reference).
178 - avoid a crash in coro->call|rss when the coroutine was already
179 destroyed (most noticably when using Coro::Debug::ps :)
180 - also protect *Storable::FILE.
181 - push up default storable granularity to 20ms.
182
1834.746 Sun Sep 21 03:22:20 CEST 2008
184 - be more insistent on locking Storable against reentrancy
185 in Coro::Storable.
186 - move swap_def?v and throw to Coro::State, as documented.
187
1884.745 Thu Jul 24 00:14:38 CEST 2008
189 - remove debugging code related to MgPV_nolen_const, also try to
190 make it compile with perl 5.8.6 (yes, apple apparently loves
191 outdated software). Reported by John S.
192
1934.744 Tue Jul 8 22:06:35 CEST 2008
194 - correctly provide default DIE/WARN handlers as documented.
195 - also overwrite PL_vtbl_sigelem.svt_clear, even though current
196 implementations inside perl work fine for us.
197
1984.743 Mon Jun 16 00:21:57 CEST 2008
199 - when using Coro::EV without running EV::loop it could
200 result in busy-waiting for events, this has been fixed.
201 - reduce codesize and improve performance by using EV_DEFAULT_UC.
202
2034.742 Sat May 31 14:10:21 CEST 2008
204 - implement a workaround for (some) perl <5.8.8 versions.
205 - require EV 3.3+.
206
2074.741 Fri May 30 23:33:09 CEST 2008
208 - tell netbsd how utterly broken their imitation of an OS is
209 and refuse to build by default if pthreads are in use.
210 - switch to "s" method on all bsd's by default, as their ucontext
211 stuff seems just too broken.
212 - fix a bug in Coro::Select.
213
2144.74 Thu May 29 20:05:31 CEST 2008
215 - do not test Coro::LWP for lack of dependencies.
216 - do not test Coro::Select for lack of working perls.
217
2184.73 Thu May 29 2008
219 - fix a bug in Coro::EV which would cause it to block despite
220 there being runnable coroutines.
221 - sprinkle "no warnings" freely over everything, also suppress
222 warnings for some other modules.
223 - fix typo in WSAEWOULDBLOCK.
224
2254.72 Sun May 25 05:14:36 CEST 2008
226 - tweak META.yaml a bit, unfortunately, there is no documented way
227 to have optional dependencies with CPAN. doh :(
228 - avoid running some tests on windows because they would fail due to
229 perl bug (broken fork, broken pipes...).
230 - work around perl on windows bugs where perl returns undocumented
231 error codes for sysread, syswrite etc. by taking advantage
232 of AnyEvent's workaround for that problem.
233 - use AnyEvent::Util::fh_nonblocking in Coro::Handle to work around
234 a common perl implementation bug on windows.
235 - use unix domain sockets in testsuite to work around
236 common perl implementation bugs on widows (they are emulated by
237 tcp sockets on windows. ugh.)
238
2394.71 Sat May 24 20:01:27 CEST 2008
240 - fix a bug in Coro::AnyEvent ("Usage: Coro::AnyEvent::_schedule()").
241 - take advantage of async name resolution of AnyEvent::Util.
242 - work around brutal inet_aton override in Coro::LWP.
243 - take advantage of the readyhook in Coro::EV, for smoother
244 scheduling.
245
2464.7 Sun May 11 00:32:19 CEST 2008
247 - completely reworked the Coro manpage.
248 - added Coro::AnyEvent, generic event loop integration.
249 - implement cancel, ready and kill commands in Coro::Debug.
250 - document find_coro in Coro::Debug.
251 - incompatible change: rename has_stack to has_cctx.
252 - Coro::AIO and Coro::BDB no longer force event model detection,
253 use AnyEvent::AIO and AnyEvent::BDB.
254
2554.6 Sat Apr 26 10:05:14 CEST 2008
256 - INCOMPATIBLE CHANGE: sub/code attributes are no longer supported
257 by the Coro module. It was a mistake to have it in the first place.
258 - (experimental) support for activestate perl 5.10 (method "w").
259 - (experimental) support for strawberry perl (method "a").
260 - coro_sigelem_set did not return a value although it had to,
261 actual impact unknown.
262
2634.51 Mon Apr 14 13:28:27 CEST 2008
264 - make it compile again on 5.8.
265
2664.50 Thu Apr 10 09:43:17 CEST 2008
267 - I did it twice! (see 4.49).
268
2694.49 Sun Apr 6 21:23:31 CEST 2008
270 - grr, instead of compiling the recent changes on 5.10 only they
271 were compiled on 5.8 only.
272
2734.48 Sun Apr 6 20:36:46 CEST 2008
274 - allow coroutine switches during eval's under 5.10.x, as it seems
275 the parser is a per-interpreter option now, so this is safe
276 (this might fix the odd crashing bug).
277 - drop support for 5.9.x versions: they are dead, jim.
278
2794.47 Sun Apr 6 00:37:52 CEST 2008
280 - force cctx allocation on API calls: we know we need to force one
281 and gcc actually manages to confuse our heuristic nowadays,
282 leading to crashes and worse.
283 - document force_cctx.
284
2854.46 Fri Apr 4 22:05:43 CEST 2008
286 - upgrade libcoro, resulting in pthread-backend (which was only created
287 to fulfill the rules of the programming languages shootout).
288
2894.45 Thu Mar 13 11:55:36 CET 2008
290 - fix a file leaking bug in eg/mhyttpd that would allow
291 downloading of any file (reported by oesi).
292 - fix deadlock bug in Coro::Channel (reported by Richard Hundt)
293 (also add testcase).
294 - support Broadcast option in Coro::Socket (patch by Richard Hundt,
295 apparently having loads of fun with that).
296 - implement $state->swap_defsv, ->swap_defav and document ->throw.
297
2984.4 Wed Feb 13 16:44:29 CET 2008
299 - only assume .cfi pseudo isns exist on GNU/Linux.
300 - add get_timed method to Coro::Channels.
301 - fixed Coro::Debug manpage.
302 - perl 5.11 compatbility improvement based on patch by
303 Andreas König.
304
3054.37 Sun Jan 20 11:25:23 CET 2008
306 - soften the check for an active parser for perl 5.10 (avoiding
307 "Coro::State::transfer called while parsing" in many cases).
308
3094.36 Sun Jan 13 10:53:56 CET 2008
310 - reset diehook when terminating from an async_pool as to not
311 trigger any __DIE__ handlers.
312
3134.35 Sun Jan 13 04:14:13 CET 2008
314 - "bt" debug command now displays any exceptions
315 from longmess and also skips the innermost
316 stackframes, giving more useufl output.
317 - allow backtraces in coroutines blocked in Coro::EV,
318 at a <1% speed hit.
319 - handle localising of $SIG{__WARN__} and __DIE__
320 properly (with a proper amount of dirty hacking).
321
3224.34 Sat Dec 22 17:49:53 CET 2007
323 - upgrade to EV version 2.0 API.
324
3254.33 Mon Dec 17 08:36:12 CET 2007
326 - make Coro::AIO etc. loadable in the absence of EV.
327
3284.32 Mon Dec 17 07:46:02 CET 2007
329 - majorly improved Coro::Handle's performance with Coro::EV.
330 - implemented the until now mythical Coro::BDB module.
331 - specialcase EV in Coro::AIO and Coro::BDB for extra speed.
332
3334.31 Wed Dec 5 12:32:39 CET 2007
334 - remove warn statement form Coro::Util.
335
3364.3 Tue Dec 4 20:33:14 CET 2007
337 - calls to the idle function could cause stack corruption
338 when the stack changed.
339 - do no longer rely on the presence of EV::DNS if EV is used
340 (because it is gone), but instead take avdantage of EV::ADNS
341 if available.
342 - add ($) prototypes to all functions in Coro::Storable.
343 - use a conventional (safer) idle callback in Coro::EV.
344 - do accept quickly in Coro::Debug to avoid endless loops.
345
3464.22 Fri Nov 30 16:04:04 CET 2007
347 - really use optimised versions for Event and EV in Coro::Util
348 and Coro::Handle.
349
3504.21 Sun Nov 25 10:48:59 CET 2007
351 - fix a spurious memory read.
352 - Coro::EV no longer keeps the eventloop "alive".
353
3544.2 Fri Nov 9 20:47:05 CET 2007
355 - enable/disable tracing from a new coroutine, not a pooled one.
356 - fix a memleak in Coro::Event.
357 - removed killall call from fork_eval.
358 - made sure store_fd is already loaded so that fork_eval
359 does not have to parse autoload in each subprocess.
360 - only use assembly method if -O switch is in $Config{optimize}.
361 - add (optional) Coro::EV module, so far the best event loop module
362 directly supported by Coro.
363 - if the event model is EV, use EV::DNS to resolve
364 stuff in Coro::Util.
365 - don't get confused by multiple event notifications in Coro::Handle.
366 - initial support for EV (libevent interface).
367 - require Event and EV using configure_requires, to force their existance.
368
3694.13 Wed Oct 24 07:26:45 CEST 2007
370 - add Coro::Storable::blocking_thaw.
371 - use a vastly more complicated technique to localise
372 $SIG{__WARN/DIE__} that also works on perls <= 5.8.8.
373 - use a coroutine for the idle callback Coro::Event,
374 instead of running Event in the current coroutine context.
375 This also catches recursive invocations.
376 - actually report fork errors in gethostbyname and inet_aton.
377
3784.11 Thu Oct 11 02:40:24 CEST 2007
379 - port to threaded perls.
380
3814.1 Thu Oct 11 02:38:16 CEST 2007
382 - incompatible change: $SIG{__DIE__} and $SIG{__WARN__} will now
383 be local to each coro (see Coro::State).
384 - incompatible change: for very deep reasons, cede and cede_notself
385 cannot return anything, so nothing will be returned.
386 - possibly bring back 5.10 compatibility (untested).
387 - work around stupid (and wrong) warning on 5.10 :(.
388 - overlay the saved state over the context stack. This saves
389 a few hundred bytes per coroutine on average and also
390 speeds up context switching a bit.
391 - further tune default stack sizes.
392 - (more) correctly calculate stack usage in coro_rss.
393 - Coro::Storable::blocking_* did not properly lock
394 resulting in races between coroutines.
395 - added Coro::Storable::guard.
396 - stopping to trace a coroutine could destroy the cctx of
397 an unrelated coroutine.
398 - explain the relationship between Perl and C coroutines in
399 more detail in Coro::State.
400 - Coro::Util::inet_aton did not short-circuit dotted quad forms,
401 causing a fork per resolve. This also affected Coro::Socket.
402 - switch to a separate stack in $coro->call/eval to avoid
403 invalidating pointers.
404
4054.03 Sat Oct 6 21:24:00 CEST 2007
406 - added Coro::throw method.
407 - minor code cleanups.
408
4094.02 Sat Oct 6 02:36:47 CEST 2007
410 - fix a very minor per-coroutine memleak (a single codereference).
411 - fixed a bug where the currently in-use c context would be freed
412 prematurely (can happen only when programs change the stacksize
413 or use tracing).
414 - tracing can no longer keep a coro alive after it terminated.
415 - do static branch prediction in the common path for gcc. gives
416 about 2-5% speed improvement here.
417
4184.01 Fri Oct 5 22:10:49 CEST 2007
419 - instead of recreating *a* standard output handle we simply
420 use STDOUT, which is faster and hopefully more robust.
421
4224.0 Fri Oct 5 12:56:00 CEST 2007
423 - incompatibly changed Coro::Storable::freeze.
424 - major new feature: added Coro::Debug, for interactive coroutine
425 debugging, tracing and much more.
426 - major bug fix: unbelievable, but true: $_, $/ and many other
427 "saved" variables actually weren't being saved. This has been fixed,
428 of course, while increasing performance while losing all the save
429 flags.
430 - save flags are gone, and all the api functions dealing with them.
431 - added Coro::Semaphore::adjust.
432 - added Coro::Util::fork_eval.
433 - added Coro::Storable::{nfreeze,blocking_{freeze,nfreeze}}.
434 - added Coro::killall.
435 - reduce initial stack sizes to allow for "micro-coroutines".
436 - better async_pool resource management, moved parts of async_pool
437 handling to XS (major speed improvement).
438 - actually croak before modifying important data structures.
439 - refuse to transfer while compiling.
440 - possibly support eval EXPR better now.
441 - enable assembly per default on linux+bsd x86+amd64.
442 - all internal members were renamed _something for easier subclassing.
443 - many minor tweaks.
444
4453.63 Wed May 16 14:10:06 CEST 2007
446 - implement handcoded assembly for x86/amd64 SVR ABI.
447
4483.62 Fri Apr 27 21:36:06 CEST 2007
449 - upgrade libcoro (which might set unwind info correctly).
450 - change default on linux to setjmp/longjmp.
451
4523.61 Thu Apr 19 12:36:18 CEST 2007
453 - Coro::Storable caused an endless loop when thawing invalid
454 pst's sometimes.
455 - use a Semaphore in Coro::Storable, as Storable doesn't
456 seem to be reentrant (although it is documented to
457 be threadsafe...).
458 - fix Coro::Signal to bring back the original unreliable
459 but stateful semantics.
460 - fixed a lot of typos in Coro.pm (patch submitted by David
461 Steinbrunner, which applied flawlessly).
462
4633.6 Sat Apr 14 17:13:31 CEST 2007
464 - added some bugfixes to get eg/myhttpd working again.
465 - added Coro::Storable for often-cede'ing freeze/thaw.
466 - try to do a clean exit when a coroutine calls exit
467 (EXPERIMENTAL).
468 - got rid of indirect call through _coro_init.
469 - updated the partly antique examples in eg/ to
470 work again and be a bit less magic, too.
471 - fixed Coro::Signal semantics to work as documented again.
472
4733.55 Sun Mar 25 01:20:47 CET 2007
474 - add SAVE_DEFFH to save the default filehandle and enable
475 it by default.
476 - finally move socket-operations from Coro::Socket to Coro::Handle
477 to be able to unblock foreign sockets.
478 - add Coro::State::save_also and guarded_save.
479 - add count accessor to Coro::Semaphore.
480 - add Coro::State::cctx_stacksize.
481 - just for the fun of it, do not rely on implicit context,
482 which can dramatically improve performance, but people
483 using windows-process-emulation perls do not care much about
484 performance.
485
4863.51 Sun Mar 4 14:18:04 CET 2007
487 - fixed a problem when you weakened references to Coro::State's
488 (patch by Yuval Kogman).
489
4903.501 Wed Feb 28 12:44:07 CET 2007
491 - rename some global symbols as macosx from hell redefines
492 them without asking.
493
4943.5 Tue Feb 13 20:22:53 CET 2007
495 - do AnyEvent model detection earlier, avoiding problems
496 caused by first using AnyEvent and later Coro::Event.
497 - implement and document Coro::Event event objects.
498 - fix a potential problem in Coro::Event causing crashes.
499 - initialise PL_comppad when creating a new coroutine,
500 avoids crashes on early coro destruction.
501
5023.41 Mon Jan 22 19:19:49 CET 2007
503 - readline on Coro::Handle did not support undefined $/,
504 nor did it deliver partial lines on EOF or error.
505 - implement malloc fallback for stack allocation because
506 stupid broken idiotic OSX has a stupid broken
507 idiotic fits-the-whole-os mmap "implementation" and
508 my dick feels longer if Coro is portable even to
509 obsolete platforms.
510
5113.4 Fri Jan 19 21:52:54 CET 2007
512 - remove t/09_timer.t, as it isn't really testing much
513 but was rather flaky in practise.
514 - async_pool coro would keep arguments and callback alive until
515 it was reused.
516 - cancellation of a coroutine could cause spurious idle calls
517 in cede_notself.
518
5193.3 Sat Jan 6 03:45:00 CET 2007
520 - implement $coro->on_destroy.
521 - Coro::Event blocking semantics have been changed,
522 documented and - hopefully - improved.
523 - fix nice adding, not subtracting, from priority.
524 - fix ->prio and api_is_ready (patch by Mark Hinds).
525 - fixed an assert ("... == prev__cctx->idle_te")
526 that could errronously trigger.
527 - fix various large and small memleaks.
528 - use a (hopefully) more stable cancel implementation
529 that immediately frees the coroutine data.
530 - cede/cede_notself return a status now.
531 - added Coro::guard function.
532 - added a global coroutine pool for jobs (on my machine,
533 I can create and execute 48k simple coros/s with async,
534 and 128k coros with async_pool).
535 - Coro::AIO now uses the coroutine priority as io priority.
536
5373.2 Fri Dec 22 05:07:09 CET 2006
538 - improve portability to slightly older perls.
539 - use cleaner coroutine destruction.
540 - simplify configuration for users.
541 - optionally (unrecommended) prefer perl functions over
542 their coro replacements.
543
5443.11 Tue Dec 5 13:11:24 CET 2006
545 - fixed some bogus assert's, but as perl.h disables assert even
546 without NDEBUG (thank you very much), not too many people should
547 notice (that did include myself). Andreas König noticed, though :)
548 - do not save/restore PL_sortcxix on >= 5.9.x, it doesn't seem to have
549 it. Also noticed by Andreas König :)
550 - save/restore tainted status.
551 - verified to pass the testsuite on my 5.9.5.
552
5533.1 Mon Dec 4 23:03:40 CET 2006
554 - INCOMPATIBLE CHANGE: $/ is now per-coroutine (but slow).
555 - incompatible change: transfer flags are now per-state.
556 - give a better error message on deadlock.
557 - document Coro::nready.
558 - enhanced testsuite.
559
5603.01 Sun Dec 3 23:47:42 CET 2006
561 - forgot to include Coro::Timer.
562
5633.0 Sun Dec 3 22:57:25 CET 2006
564 - the "FINALLY COMPLETELY STABLE" release (coming soon:
565 the "FAMOUS LAST WORDS" release).
566 - implement a new stack sharing algorithm, which uses a stack
567 pool (size currently hardcoded).
568 - make stack sharing mandatory (it no longer uses a heuristic).
569 - eval/die no longer cause weird problems under heavy use.
570 - Coro::Event could cause livelocks if it was used but
571 no Coro::Event watchers were used.
572 - Coro::Event now uses asynccheck as prepare does not
573 check for changed watchers.
574 - Coro::Event allows multiple waiting coros and will wake up one
575 per event.
576 - Coro::Event should be cleaner and more efficient now.
577 - new utility function Coro::unblock_sub.
578 - document the sad fact that Event is no longer reentrant.
579 - putting a coroutine into the ready queue twice could under
580 some circumstances lead to stack corruption.
581 - minor incompatibility: subclassing Coro::State is supported
582 directly now without going through a _coro_state member.
583 - state/coro switching is much faster now.
584 - very minor optimisations and code/documentation cleanup.
585 - avoid problems due to compiler inlining.
586 - removed timers from Coro::Timer -> use AnyEvent instead.
587 - replaced Coro::idle coroutine by (cleaner) idle handler.
588 - updated to newest libcoro.
589 - implement enhanced support for valgrind.
590 - implement is_ready and return value for ready.
591 - removed Coro::Cont, it was a misnomer (it's generators),
592 and it was somewhat annoying to get it right. Will come back
593 if somebody asks for it :)
594 - many ->wait methods and Coro::Event could return
595 spuriously without the event having happened.
596
5972.5 Tue Nov 7 12:22:33 CET 2006
598 - made Coro::Util, Coro::Select, Coro::Handle and Coro::Socket
599 use AnyEvent, moved them to Coro/.
600 - added Coro::LWP which contains all the uglyness required to
601 make LWP non-blocking.
602 - should work with perl 5.9.x now (Andreas König made me do it).
603 - fixed another bug in Coro::Select when the timeout was undef.
604 - reuse PL_start_env for all coros, saves some memory per coroutine.
605 - manage PL_top_env differently, hopefully to avoid panic: top_env.
606 - timeout argument was not properly used in Coro::Socket.
607 - allow limited forms of subclassing in Coro::Handle/Coro::Socket.
608 - emulate undocumented(!) functionality of IO::Socket required
609 by LWP(!!).
610 - updated eg/lwp to work with newer lwp's.
611 - remove "FATAL: uncaught exception" prefix. Coroutines that die
612 kill the whole process, just as exceptions in the main "coroutine"
613 did already.
614
6152.1 Wed Nov 1 23:01:13 CET 2006
616 - fix a long-standing bug in Coro::Select where select with
617 zero timeout would instead change the current default filehandle.
618 - use a simpler and hopefully more robust way to clone padlists
619 (uses less memory and a perl function instead of our own).
620 - coro can now create a stack guard area on many architectures.
621 - Coro::AIO properly reexports additional functions from IO::AIO.
622 - updated libcoro with a workaround for OS X,
623 pach and testing by Michael Schwern.
624
6252.0 Tue Oct 24 05:47:17 CEST 2006
626 - support additional aio requests in Coro::AIO.
627
6281.9 (never properly released due to a glitch)
629
6301.8 Thu Feb 2 00:59:06 CET 2006
631 - applied suggested patch by SAMV to avoid problems during stupid
632 mark & sweep gc run.
633 - applied patch by Scott Walters for 5.9.3 compatibility.
634
6351.7 Tue Dec 27 01:41:58 CET 2005
636 - added Coro::AIO, a thin wrapper around IO::AIO.
637 - improved Makefile.PL explanations.
638
6391.6 Mon Dec 12 21:30:05 CET 2005
640 - additionally save PM_curpm might fix as-of-yet
641 unseen problems with regex matches being attributed to the
642 wrong package.
643 - add t/10_bugs.t, which currently checks against imho broken
644 perls that use 0.26MB of stack space per Perl_magic_get
645 invocation (newer linuxes) as opposed to the 0.0002MB perl
646 normally uses.
647 - make stacksize configurable for the ultimate debian experience.
648
6491.51 Mon Dec 12 18:48:36 CET 2005
650 - remove debugging warn() accidentally left in Coro::Select.
651
6521.5 Tue Nov 29 13:32:44 CET 2005
653 - use Coro::Event inside Coro::Select to avoid spurious deadlocks.
654 - fix Coro::Select 'select'.
655 - strict'ify some modules.
656 - libcoro errornously restored the SIGUSR2 handler to SIGUSR1.
657 - use XSLoader in selected modules.
658 - remove some 5.6 compatibility cruft.
659
6601.4 Tue Sep 6 00:11:05 CEST 2005
661 - libcoro did not take into account the trampoline on amd64,
662 when 'l'inux method was used.
663
6641.31 Tue Aug 30 23:31:33 CEST 2005
665 - some portability fixes/workarounds.
666
6671.3 Sat Aug 20 03:08:56 CEST 2005
668 - no code changes. module seems to work fine.
669 - improve Coro::State docs, remove reference to nonexisting
670 Coro::State::flush method.
671 - no longer autodetect windows, present it as an option instead.
672
6731.2 Mon May 16 02:00:55 CEST 2005
674 - included libcoro.c earlier in State.xs, might improve portability.
675 - use a faster and possibly more stable Coro::Cont implementation.
676 - accept x86_64 in addition to amd64 for optimized linux-amd64 support.
677 - fix bugs Coro::SemaphoreSet that could cause locks never to be freed.
678 - fix bugs in CoroAPI.h and document it in Coro::MakeMaker.
679
6801.11 Thu Mar 3 18:00:52 CET 2005
681 - change of contact address.
682
6831.1 Tue Feb 22 20:51:16 CET 2005
684 - support [l]inux method on amd64.
685 - allow some context switches while compiling/eval'ing.
686
6871.0 Tue Aug 10 03:47:27 CEST 2004
688 - Event 0.89 fixes the bug again.
689 - fix the remaining(??) memleak. I didn't free the cache-AV for
690 padlists, so one AV per created anon-sub leaked.
691
6920.97 Fri May 14 15:23:32 CEST 2004
693 - remove stupid gcc extension that was an experiment only anyway.
694
6950.96 Thu May 13 18:09:29 CEST 2004
696 - tell people if they have yet another broken Event variant.
697 - made Coro to work _better_ with threads (it seems that, as long
698 as you use Coros in one thread only everything will work fine.
699 Looking at the perl source the problem might be missing locking
700 between Coro and some perl internal routines. That's not easily
701 fixable).
702 - (hopefully) got rid of the static xnext inside transfer.
703
7040.95 Sun Feb 15 03:20:28 CET 2004
705 - removed Coro::State::flush, implemented a new and much
706 faster caching mechanism by attaching magic to the CV.
707 - a memleak with (real) closures remains.
708 - dramatically reduced size of stacks to 65536 (linux/x86), after
709 observing that even myhttpd never had more than a kilobyte
710 of stack in use (stack is still only physically allocated in
711 pagesize increments for systems suporting mmap).
712 - fix for a "cygwin-64" architecture added, proposed by Stefan Traby.
713 - fix for threaded perl.
714
7150.9 Sun Nov 23 04:32:39 CET 2003
716 - fix a few very hard to track down but (of course) stupid
717 memory leaks.
718 - stringify version number to avoid locale problems :(.
719 - perl 5.9.0 does not have SvPADBUSY, reported by Scott Walters,
720 hopefully ignoring it "just works".
721
7220.8 Wed Nov 5 19:38:40 CET 2003
723 - port to perl5.005_03, but only 5.8.x+ is supported!
724 - honor LocalAddr even without LocalPort in Coro::Socket.
725
7260.7 Tue May 27 03:12:38 CEST 2003
727 - the version jump indicates some level of testing, not gobs
728 of new features.
729 - uh... I found the prompt function in ExtUtils::MakeMaker.
730 Highly correct stuff, that is...
731 - fixed(?) a bug with die's in coroutines causing "panic: top_env".
732 the fix is not well understood by the author, so beware :(.
733
7340.652 Thu May 8 02:54:46 CEST 2003
735 - Applied patch by Slaven Rezic to set default to "s" on FreeBSD,
736 cause version 4 doesn't have ucontext.
737 - Benjamin Reed reported that setjmp works fine on darwin,
738 so preselect it.
739
7400.651 Sat Mar 29 15:00:23 CET 2003
741 - fix a bug in Coro::Handle where some bytes could get lost
742 on reading, reported by jason@nichego.net.
743
7440.65 Sun Mar 23 00:08:26 CET 2003
745 - added a README.linux-glibc.
746 - new module Coro::Select.
747 - also save/restore PL_comppad, fixes segfaults with 5.8.1.
748 (I never claimed that I know what I am doing ;).
749 - set default to setjmp/longjmp on non-x86-linux.
750
7510.6 Thu Nov 21 11:09:06 CET 2002
752 - made lazy context switching the default.
753 - fixed the bug where SemaphoreSet::guard
754 was the same as timed_guard and timed_guard was missing.
755 - fix a memleak in Coro::Timer.
756 - reclassified context sharing as not experimental.
757
7580.534 Sun Apr 14 03:05:12 CEST 2002
759 - fixed backspace => tab in header parsing (myhttpd).
760 - added eg/readline, for elmex.
761 - Coro::Event now calls ->start on first call instead of
762 ->again, which created wrong timeouts for the first ->next.
763 - fixed a bug where Coro::Socket returned a closed socket instead
764 of nothing on ECONNREFUSED and similar error conditions.
765
7660.533 Mon Feb 18 18:49:40 CET 2002
767 - adapt to finally working Event-0.85.
768 - fixed buggy event initialization (trapped under windows, but how
769 could it possibly have worked anywhere?)
770
7710.532 Wed Jan 16 02:45:32 CET 2002
772 - removed the reference to pp_entersub, might work on
773 (some) win32 perls now (testsuite works under cygwin,
774 without Event, which freezes).
775
7760.531 Mon Dec 10 22:18:44 CET 2001
777 - Fixed a bug in SemaphoreSet::timed_down.
778
7790.53 Tue Nov 27 21:11:13 CET 2001
780 - some tests for Event.
781 - slightly different internal architecture to get refcnt into the object
782 and not the reference. might fix a few bugs, certainly did introduce
783 new ones.
784 - Coro::Timer (independent of Event).
785 - new timed_wait functions for Coro::Signal, Semaphore, SemaphoreSet.
786
7870.52 Tue Nov 6 21:36:18 CET 2001
788 - ported to cygwin (trivial).
789 patch by Gerrit P. Haase <gp@familiehaase.de>.
790 - small setjmp code fix by Sullivan.DanielJ@epamail.epa.gov.
791
7920.51 Thu Nov 1 20:39:01 CET 2001
793 - terminate/cancel now work properly (otherwise termination
794 could cause a "next coroutine is not and contains not..."
795 error.
796 - added Coro::Socket::shutdown.
797 - Coro::Event::loop no is the same as Event::loop.
798 - implemented terminate with args + join.
799
8000.5 Fri Sep 28 16:15:35 CEST 2001
801 - fixed "print" on a Coro::Handle. The print method worked.
802 - small tweaks (seem to reduce memory consumption a lot)
803 in various modules.
804 - splendid use of "no warnings" scattered throughout.
805 - added Coro::Handle::rbuf, fixed Coro::Handle::fh.
806
8070.49 Sun Sep 16 02:42:45 CEST 2001
808 - changed some method calls to function calls for speed
809 inside Coro::Handle.
810 - make Coro::Handle use an array instead of a hash for
811 speed reasons.
812 - IRIX mystery solved: it's SGI's NT, after all: "standard, huh?".
813 sjlj and ucontext should now work.
814 - IRIX-specific port for libcoro.
815 - swapped order of accept results to match IO::Socket.
816 - changed getsock/peername to sock/peername to match IO::Socket.
817 - fixed a bug that caused segfault when returning to main task
818 under some circumstances.
819 - other bugfixes.
820
8210.45 Sun Sep 2 02:54:01 CEST 2001
4 - new method Coro::Handle::timeout. 822 - new method Coro::Handle::timeout.
5 - corrected speling of set...name to get...name in Coro::Socket. 823 - corrected speling of set...name to get...name in Coro::Socket.
6 - Coro::Socket::accept now returns a Coro::Socket, not a Coro::Handle. 824 - Coro::Socket::accept now returns a Coro::Socket, not a Coro::Handle.
7 - Coro::Handle now supports fileno and FILENO. 825 - Coro::Handle now supports fileno and FILENO.
8 - added eg/myhttpd, a web server. 826 - added eg/myhttpd, a web server.
9 - Coro::Socket now accepts numeric ports as well as "port(number)" 827 - Coro::Socket now accepts numeric ports as well as "port(number)"
10 syntax. 828 syntax.
11 - moved some scheduling primitives into xs code. more to come. 829 - moved some scheduling primitives into xs code. more to come.
12 - new simple priority system. 830 - new simple priority system.
831 - implemented Coro::Event using XS, almost four times faster!
832 - small memory corruption problem fixed (boy that was difficult).
13 833
140.13 Wed Aug 8 16:53:07 CEST 2001 8340.13 Wed Aug 8 16:53:07 CEST 2001
15 - new method Coro::Handle::readline. 835 - new method Coro::Handle::readline.
16 - added eg/lwp and Coro::Handle::autoflush, to make LWP work. 836 - added eg/lwp and Coro::Handle::autoflush, to make LWP work.
17 (see eg/lwp on how to make LWP non-blocking). 837 (see eg/lwp on how to make LWP non-blocking).

Diff Legend

Removed lines
+ Added lines
< Changed lines
> Changed lines