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

Comparing Coro/Changes (file contents):
Revision 1.246 by root, Sun Sep 30 13:57:34 2007 UTC vs.
Revision 1.603 by root, Wed Dec 5 01:09:12 2012 UTC

1Revision history for Perl extension Coro. 1Revision history for Perl extension Coro.
2 2
3TODO: destroy gets only called on cancel: fix(?) 3TODO: should explore PerlIO::coroaio (perl leaks like hell).
4TODO: should explore PerlIO::via::CoroCede and PerlIO::via::CoroAIO. 4TODO: channel->maxsize(newsize)?
5TODO: maybe implement a default message channel, very much like Erlang 5TODO: http://www.microsoft.com/msj/archive/s2ce.aspx
6 (which is cool in a lot of important aspects (failures!), 6TODO: __GCC_HAVE_DWARF2_CFI_ASM
7 but very lacking in others).
8 7
93.8 8TODO: add peerhost/peerport from IO::Socket::INET etc.
9
10 - implement 5.17 compatibility by almost blindly applying a
11 good-looking patch by Father Chrysostomos.
12 - use experimental fiber implementation on native windows
13 perls.
14 - support magic values as timed_io_once args.
15 - recommend AnyEvent 7+ or EV 4+, also require EV
16 version 4 or newer for Coro::EV.
17
186.10 Tue Oct 9 01:14:27 CEST 2012
19 - updated ecb.h, it had a typo that caused it to not compile on many
20 big endian systems (reported by many people).
21 - disable memory fences in ecb.h to improve portability.
22
236.09 Sat Oct 6 23:25:02 CEST 2012
24 - Coro::EV I/O watchers were not interruptible by exceptions
25 (Coro::State::throw) (testcase by sten).
26 - ->throw now puts threads into the ready queue, as this seems to
27 be expected by existing code, and code that doesn't cope with spurious
28 wakeups needs fixing anyway.
29 - use fd -1 in mmap.
30 - cast I32 to int in error message printf.
31 - warn about broken so-called "hardened" kernels.
32
336.08 Fri Apr 13 12:05:47 CEST 2012
34 - be more aggressive about exiting like perl does - formerly,
35 exiting from the non-main thread would not execute END blocks.
36
376.07 Fri Nov 11 21:21:48 CET 2011
38 - work around a bug in PerlIO (setting $SIG{__WARN__} to a PVCV).
39 - update ecb.h.
40
416.06 Mon Aug 8 23:59:48 CEST 2011
42 - cygwin unfortunately patches the stack at runtime, so we use the pthreads
43 backend, which is an order of magnitude slower. unfortunately, cygwins
44 pthread implementation isn't very complete either, so allocate the stack
45 twice just to be sure.
46 (note: cygwin also enables mymalloc, which is NOT THREADSAFE ON WINDOWS,
47 in its ithreaded perl - best recompile cygwin and use the 'w'indows
48 backend for much better performance. also disable ithreads for
49 even better performance...).
50
516.05 Thu Aug 4 21:36:36 CEST 2011
52 - blush, condvar values would not be propagated from send to recv anymore
53 (reported by Chip Salzenberg).
54 - use exponential increase for the readline buffer length in
55 Coro::Handle. also reduce initial allocation to 1020 from 4096 bytes.
56
576.04 Wed Aug 3 17:15:45 CEST 2011
58 - use even more efficient and more compatible condvars for
59 compatibility to AnyEvent 6.x :)
60 - more inconsequential ecb.h updates.
61
626.03 Wed Aug 3 11:41:30 CEST 2011
63 - change how Coro patches AnyEvent condvars for compatibility to
64 AnyEvent 6.x.
65 - update ecb.h, to no longer include <pthread.h> in case WinNT.h
66 hasn't been included.
67
686.02 Wed Jul 13 04:35:19 CEST 2011
69 - "improve portability to Gentoo" - gentoo manages to put perl variables
70 in memory areas that are farther than 2gb apart, which the jit couldn't
71 handle and barfed. now it's just a bit slower on gentoo and similar
72 systems.
73
746.01 Sun Jul 3 12:31:14 CEST 2011
75 - workarounds are good, but the test for whether pthreads are used
76 was not good. thisone should be better.
77 - check differently whether gcc generates cfi instructions itself.
78
796.0 Wed Jun 29 19:43:35 CEST 2011
80 - INCOMPATIBLE CHANGE: unreferenced coro objects will now be
81 destroyed and cleaned up automatically (e.g. async { schedule }).
82 - implement a JIT compiler for part of the thread switch code,
83 which gives a 50% speed improvement on threaded perls, and
84 about 4% on non-thraeded perls (so threaded perls now finally
85 reach about half the speed of non-threaded perls).
86 - slightly modernise Coro::Intro, add section about rouse functions.
87 - avoid DEFSV and ERRSV, giving another 10% improvement
88 in thread switching.
89 - Coro::State->is_destroyed is now called is_zombie.
90 - implement a Coro->safe_cancel method that might fail, but
91 cancels in a "safer" way if it succeeds.
92 - add preliminary support for DEBUGGING perls.
93 - get rid of two hash-accesses when initialising a new Coro - this
94 speeds up coro creation by almost a factor of two.
95 - croak when a coro that is being cancelled tries to block
96 (e.g. while executing a guard block), instead of crashing or
97 deadlocking.
98 - use a more robust and also faster method to identify Coro::State
99 objects - speeds up everything a bit.
100 - implement Coro->cancel in XS for a 20% speed improvement, and to
101 be able to implement mutual cancellation.
102 - speed up context switches by a percent or two by more efficiently
103 allocating context stack entries.
104 - implement Coro->join and Coro->on_destroy in XS for a speedup and
105 a reduction in memory use.
106 - cancelling a coro while it itself is cancelling another coro is
107 now supported and working, instead of triggering an assertion.
108 - be a bit more crash-resistant when calling (buggy) on_destroy
109 callbacks (best effort).
110 - move on_destroy into the slf_frame, to allow extension slf
111 functions to have destructors.
112 - get rid if coro refcounting - simply crash in other interpreter
113 threads by nulling the pointers on clone.
114 - simplify warn/die hook handling when loading Coro - the convoluted
115 logic seems to be no longer neccessary.
116 - use libecb instead of our own home-grown gcc hacks.
117 - document alternatives to Coro::LWP. Please use them :)
118 - work around another mindless idiotic NEEDLESS bug in openbsd/mirbsds
119 sigaltstack. Really. wine suffers from it, erlang suffers from it,
120 and it's known since at least 2006.
121
1225.372 Wed Feb 23 06:14:30 CET 2011
123 - apparently mingw doesn't provide a working gettimeofday, try to
124 work around that by relying on Time::HiRes (indirectly brought to
125 my attention by Max Maischein).
126 - fix some portability issues when Time::HiRes was used.
127
1285.371 Mon Feb 21 14:36:08 CET 2011
129 - backport to windows process emulation code again.
130
1315.37 Sat Feb 19 07:49:44 CET 2011
132 - add a big "Coro thread life cycle" section to "man Coro".
133 - try a tentative workaround against the breakage that 5.13 has
134 introduced without depreciation period. sigh.
135 - no longer use Time::HiRes if gettimeofday is available, which
136 saves quite a lot of memory.
137
1385.36 Sun Feb 13 05:33:41 CET 2011
139 - automatically load Coro::Channel, Coro::RWLock, Coro::Semaphore,
140 Coro::SemaphoreSet, Coro::Signal and Coro::Specific on first "new"
141 method call.
142 - undocument Coro::Timer::sleep and obsolete whole module.
143 - optimise Coro::Timer::timeout memory and cpu usage.
144 - slightly updated Coro::Intro for recent changes.
145 - do not initialise PL_dirty anymore.
146
1475.25 Thu Nov 11 01:08:39 CET 2010
148 - try a different approach on netbsd - netbsd 5 finally has marginally
149 working pthreads, but still broken ucontext/sigaltstack.
150 - openbsd 4.8 finally got their act together, Coro works out of the box
151 with asm, setjmp and pthreads (no change, just informational).
152
1535.24 Sat Oct 23 11:27:12 CEST 2010
154 - port to the EV 4.0 API.
155 - work around bugs in mingw32, making strawberry perl work
156 out of the box.
157 - correctly modify Coro::AIO function prototypes
158 so that they reflect the "no optional parameters" rule.
159 - "ported" libcoro to C++.
160
1615.23 Mon May 17 18:50:42 CEST 2010
162 - be more resistant to ordering changes when initialising
163 Coro::AnyEvent, Coro::EV and Coro::Event (reported by Matthias
164 Waldorf).
165 - document that perl 5.12 deliberately removed support for cloning.
166
1675.22 Wed Apr 14 03:55:35 CEST 2010
168 - correctly return udnef on errors in Coro::Handle::read/write
169 (testcase by Marc Mims).
170 - convert Coro::Util into a "perl compatibility wrapper" - the functions
171 are less useful now, but are drop-in replacements for existing
172 functions, listing better alternatives in the documentation. This also
173 fixes a bug in Coro::LWP which naively substituted Socket::inet_aton
174 with Coro::Util::inet_aton.
175 - do not override $Coro::idle unconditionally in Coro.pm, as other
176 modules could have provided their own idle coro already
177 (for exmaple, Coro::AnyEvent).
178 - fix Coro::Util::gethost* functions.
179 - Coro::Timer corretcly exports it's symbols (reported by Hideki Yamamura).
180
1815.21 Wed Dec 16 07:19:51 CET 2009
182 - automatically load Coro::AnyEvent when AnyEvent and Coro are used
183 together.
184 - add some examples on how to combine other event loops with Coro in
185 Coro::AnyEvent, and how to run it (and not to block). Seems to be
186 the most common source of confusion.
187 - try to catch people naively blocking in an event callback.
188 - work around the perl filehandle bug issue in conjunction with
189 older common::sense (as indirectly pointed out by ZSystem).
190 - clarify the "not from signal handlers" section.
191
1925.2 Sun Oct 4 14:54:24 CEST 2009
193 - Coro::Storable destroyed the prototypes of the functions it wrapped.
194 - export rouse_cb and rouse_wait by default now.
195 - fix various prototype mismatches in Coro::AnyEvent and Coro::Handle.
196 - new method $state->swap_sv.
197 - added section on "windows process emulation" to the manpage,
198 after a not-so-fruitful (nor-friendly) "discussion" with chip
199 salzenberg (discussion implies arguments, but his only arguments
200 were ad-hominems, one wonders why he started it in the first
201 place). I hope this explains it well enough for him to understand,
202 and maybe well enough for others to understand.
203 - use common::sense everywhere now.
204 - idle callbacks are no longer supported, use idle coros instead.
205 - print a thread listing when a deadlock is detected.
206
2075.17 Sat Aug 22 23:09:31 CEST 2009
208 - work around a bug in the perl debugger causing crashes
209 when running under the debugger by marking _pool_handler as nodebug.
210 - speed up Coro::async considerably.
211 - try some hacks to get netbsd to work "more often" - their broken
212 setjmp/longjmp, ucontext *and* phtreads are really hard on Coro.
213 - convert Coro to AE 5.0 API.
214
2155.162 Tue Jul 28 04:04:03 CEST 2009
216 - perl 5.8.2 is now minimum requirement.
217 - skip t/19_handle.t on broken windows perls.
218
2195.161 Wed Jul 22 04:47:38 CEST 2009
220 - Coro::AnyEvent::poll could have a different prototype when EV was
221 used as backend (analyzed by Tatsuhiko Miyagawa).
222 - Coro::AnyEvent errornously initialised the event loop when loaded,
223 not on demand.
224 - try to workaround rare */t/01_unblock.t failures.
225
2265.16 Tue Jul 21 01:44:37 CEST 2009
227 - Coro::AnyEvent failed to hook into the event loop
228 when no threads had been readied between detecting
229 the event loop and actually running it.
230 - considerably speed up Coro::Select by taking avdantage
231 of AnyEvent > 4.8 and some other optimisations.
232 - implement paragraph readline mode in Coro::Handle
233 (based on patches by Zsbán Ambrus).
234 - replace WSAEINPROGRESS by WSAEWOULDBLOCK (reported
235 and analyzed by Yasuhiro MATSUMOTO).
236 - clarified libcoro license and copyright.
237 - someone stole my EXTRA_META!!!
238 - implement Coro::Select::patch_pp_sselect and it's brother,
239 for hardcode select overriding.
240
2415.151 Mon Jul 6 05:41:57 CEST 2009
242 - backport to windows process emulation code again (patch by
243 Yasuhiro MATSUMOTO).
244 - slightly update Coro::MakeMaker.
245
2465.15 Tue Jun 30 10:28:06 CEST 2009
247 - deprecate Coro::Socket, document how to get ipv6 support via
248 AnyEvent::Socket instead.
249 - implement signal->wait ($cb) interface, similar to semaphores.
250 - work around SvOK not supporting getmagic, so we have to getmagic
251 to test for undef :/ (reported by Matthias Waldorf).
252 - load Coro::AnyEvent in all modules using AnyEvent.
253 - work around perl corrupting our internal data structures,
254 reported by Tokuhiro Matsuno.
255 - enable per-coroutine real and cpu time gathering
256 (Coro::State::enable_times).
257
2585.14 Wed Jun 24 01:37:48 CEST 2009
259 - provide explicit functions to "cede" to the event loop in Coro::AnyEvent,
260 as this seems to have been a difficult concept (poll, sleep, idle,
261 idle_upto).
262 - add Coro::AnyEvent::readable/writable functions.
263 - clarify Coro::EV/Event/AnyEvent manpages.
264 - free per-thread global scalars in the thread calling ->cancel, to
265 avoid crashes when $_, $@ etc., are magical but some of those
266 had already been freed.
267 - "unexperimentalise" the callback interface for Coro::Semaphore.
268 - speed up ready queue management/context switching by using a linked
269 list instead of an array (~5-10%).
270 - implement "watch" command in Coro::Debug shells.
271 - for fun, implement time-slicing as an example in the manpage.
272 - if AnyEvent detects EV or Event, but we don't have Coro::EV or
273 Coro::Event, use the normal AnyEvent handling instead of dieing
274 (the same is true for Coro::Handle).
275 - properly document Coro::EV::timed_io_once.
276 - avoid unneccessary ->cancel calls in Coro::Handle.
277 - maybe make it work on mingw32 with win32 backend
278 (based on patch by Yasuhiro Matsumoto).
279
2805.132 Fri May 29 09:00:39 CEST 2009
281 - do not keep a reference to the argument itself in
282 Coro::Semaphore::guard, as it could change later.
283 - support SO_RCVBUF/SO_SNDBUF nonstandard Coro::Socket options,
284 should support a prepare callback.
285
2865.131 Mon Mar 16 23:20:37 CET 2009
287 - implement and document Coro->suspend, Coro->resume.
288 - fix Coro::Select implementation to not (often) close
289 the passed file descriptors (testcase provided by pippijn).
290
2915.13 Mon Dec 15 21:51:42 CET 2008
292 - EXPERIMENTAL: implement dynamic winds (on_enter/on_leave).
293 - don't set diehook to C<undef> but instead to NULL, to avoid
294 spurious warnings.
295 - fix a lot of bugs in Coro::SemaphoreSet.
296 - Coro::SemaphoreSet will less often create a semaphore needlessly.
297 - add Coro::SemaphoreSet::count and wait methods.
298 - take advantage of the new Guard module.
299 - deprecate Coro::guard.
300 - try to fix the dreaded 01_unblock tests once more. I hate it when
301 testsuites need more fixing than the code they are supposed to test.
302 - croak in more cases when a required callback isn't resolvable.
303 - fix some minor issues in Coro::State->call/eval.
304 - use current coroutine context instead of a temporary one
305 when temporarily switching to another coroutine.
306 - do not call C-level on_destroy handlers during global destruction,
307 to avoid needless segfaults.
308
3095.12 Sun Dec 7 13:30:38 CET 2008
310 - add default config for MirOS, which seems to be a bug-to-bug
311 compatible fork of openbsd ("world domination by releasing
312 openbsd cvs before the openbsd folks do it" or so :).
313 - free_padlist did destroy the names pad, not good, but didn't
314 seem to bother perl - this could fix issues such as eval ""
315 inside a function called from multiple coroutines.
316 - use a different method to detect destruction time.
317 - be more careful when freeing padlists just before global
318 destruction.
319 - fixed and expanded the call/cc example.
320 - renamed _terminate to _coro_run.
321 - new method Coro::Channel->shutdown.
322 - try pthreads on openbsd <4.4 (broken sigaltstack, will
323 pthreads fare better?).
324 - be less picky about destroying "a" running coroutine.
325
3265.11 Tue Nov 25 21:49:05 CET 2008
327 - DEBUGGING in 5.10.0 is a mess: it constantly flags perfectly
328 working code with failed assertions, introducing more bugs than
329 it fixes, requiring elaborate workarounds :(
330
3315.1 Mon Nov 24 08:54:59 CET 2008
332 - wrote a small introductory tutorial - Coro::Intro.
333 - convert Coro::Timer, Coro::Select and Coro::Util to rouse API.
334 - Coro::Select did errornously dup the file descriptors
335 and didn't work with all AnyEvent backends.
336 - Coro::Select wasn't imported correctly form Coro::LWP, causing blocking
337 LWP data transfers.
338 - disassociate c contexts from coro objects - this is agruably more
339 correct, but mostly allows sharing of cctxs between coro and state
340 objects, for added memory savings and speed increases.
341 - bumped $Coro::POOL_RSS up to 32kb by default.
342 - no longer set the optype to OP_CUSTOM, as B::* understandably
343 doesn't like this very much (and we *are* a type of entersub).
344 - implement state cloning, just to prove that call/cc can be done.
345 - automatically load Coro::AnyEvent in Coro::Handle.
346 - wrap ->cancel calls in eval inside Coro::Handle as EV watchers
347 do not have this method (and don't need it either).
348 - speed up generic anyevent methods in Coro::Handle by using rouse
349 callbacks.
350 - allow coroutines in $Coro::IDLE, speeding up Coro::AnyEvent and
351 others. It also makes the debugger happier, as you can trace
352 through the idle threads now.
353 - add comppad_name* and hints ($^H, %^H) to per-thread variables.
354 - eg/event was pretty broken.
355 - better 5.8.6 compatibility.
356
3575.0 Thu Nov 20 10:35:05 CET 2008
358 - NEW ARCHITECTURE: use the latest 4.x version if you experience
359 stability issues.
360 - bump API version to 7 - all dependents must be recompiled.
361 - removed timed_* functions - they were not being used anyways
362 and should be replaced by a more generic mechanism -
363 and were annoying to support anyways :)
364 - removed SemaphoreSet's waiter method - use sem method instead.
365 - Coro::Semaphore->adjust didn't correctly wake up enough waiters.
366 - async_pool did free a scalar value twice
367 ("Attempt to unreference...").
368 - fix a longstanding bug where calling terminate on a coro that
369 was waiting for a semaphore that was just becoming available
370 would cause a deadlock (semaphore would get into a state where
371 it was available but waiters were still blocked).
372 - calling throw on a coroutine that is waiting for a semaphore will
373 no longer make it acquire the semaphore (and thus leak a count).
374 - perl's process emulation is now not even theoretically supported
375 anymore.
376 - new functions Coro::rouse_cb and Coro::rouse_wait for easier
377 conversion of callback-style to blocking-style.
378 - new methods $coro->schedule_to and ->cede_to, to specifically
379 schedule or cede to a specific coroutine.
380 - new function Coro::Semaphore::wait.
381 - use named constants in Coro::Channel (Richard Hundt).
382 - directly patch the entersub opcode calling SLF functions (cede,
383 transfer and so on). this does speed up context switching, but
384 more importanly, it frees us from the hardcoded behaviour of
385 entersub, so we might actually be able to return something from
386 those functions and atcually create new ones.
387 - take advantage of __builtin_frame_address on gcc.
388 - expose THX in coroapi (not sure whether this was a wise decision,
389 as "threaded" perls are running at half speed anyways).
390 - implement execute_slf (schedule-like-function) interface that makes
391 it possible to implement schedule-like-functions in XS.
392 - use new SLF interface to massively speed up Coro::EV by roughly a
393 factor of two.
394 - used new SLF interface to massively speed up Coro::Semaphore by a
395 factor of three.
396 - used new SLF interface to speed up Coro::AIO by roughly a factor of
397 four and reduce its memory usage considerably.
398 - implement Coro::SemaphoreSet purely in terms of Coro::Semaphore,
399 for a nice speedup and vastly more correct behaviour. Also implement
400 a new method "sem" to get at the underlying semaphore object.
401 - implement Coro::Channel in terms of Coro::Semaphore, for a moderate
402 (in comparison) 20-40% speedup.
403 - used new SLF interface to reimplement Coro::Signal gaining
404 some unknown (because I was too lazy), but certain, speedup, and also
405 making signals reliable for the first time.
406 - used new SLF interface and other optimisations to speed up async_pool
407 by a factor of two. It also doesn't rely on perl's exception mechanism
408 to exit anymore. The overhead for terminating an async_pool, coro over
409 a normal async is now very small.
410 - sped up coroutine creation/destruction by 40%.
411 - forgot to include Coro/libcoro/README in the dist for all these years.
412 - work around a freebsd pthreads bug (manual testcancel is required as
413 pthread_cond_wait isn't a cancellation point on freebsd).
414 - use new rouse functions to speed up and simplify Coro::BDB.
415 - make "prefer perl native functions" work with threaded perls.
416 - condense Coro::Debug ps output, hint at v and w flags.
417 - (libcoro) lots of minor cleanups and portability improvements.
418
4194.914 Wed Nov 19 12:54:18 CET 2008
420 - fix a disastrous bug in the readline optimisation
421 introduced in 4.801.
422
4234.913 Sat Nov 15 07:58:28 CET 2008
424 - async_pool did free a scalar value twice
425 ("Attempt to unreference...").
426
4274.912 Thu Nov 13 18:31:23 CET 2008
428 - minor cleanups.
429 - use much larger stacks on linux and perl < 5.8.8.
430 - Coro::Debug::new_unix_server did not unlink the socket
431 when destroyed.
432
4334.911 Tue Nov 11 04:26:17 CET 2008
434 - "port" to threaded perls.
435
4364.91 Mon Nov 10 05:36:38 CET 2008
437 - the ->throw exception object no longer leaks.
438 - creating a new cctx leaked a scopestack entry (memleak).
439 - new coroutines didn't get created with a zero flags field
440 (unknown impact).
441 - calling ->throw on a not-yet-started coroutine should now work
442 instead of being ignored.
443 - ->throw is now supported on Coro::State objects.
444 - clean up cctx creation code a bit.
445 - entersub is actually a UNOP, not a LOGOP (not a bugfix).
446
4474.9 Sat Nov 8 17:45:27 CET 2008
448 - (libcoro) did not preserve rbp with CORO_ASM (we are getting there).
449 - (libcoro) no longer leak threads in the experimental pthread backend,
450 also speed it up considerably.
451 - (libcoro) do not rely on makecontext passing void *'s unscathed.
452 - fix compiletime dependencies on libcoro in the Makefile.
453 - cctx_count wasn't always updated properly.
454 - Coro::State::cctx_stacksize wasn't applied correctly.
455 - new function Coro::State::cctx_max_idle.
456 - the default max number of idle C contexts is now 4.
457 - (libcoro) try harder to get _setjmp/_longjmp.
458 - (libcoro) cleanup and extend the libcoro API to officially
459 allow the creation of empty source contexts.
460 - very experimental workaround for the totally broken netbsd platform.
461 - tried to hack around openbsd bugs.
462
4634.804 Wed Nov 5 16:36:00 CET 2008
464 - Coro::Debug::new_unix_server would not create a non-blocking listening
465 socket, sometimes causing freezes.
466 - (libcoro) fix misaligned stack points for setjmp and assembly
467 methods, which can cause crashes on x86/x86_64 with a sufficiently
468 aggressive compiler.
469 - new function: Coro::Debug::new_tcp_server.
470 - move ->throw into the Coro class because it only works on coro objects.
471
4724.803 Mon Nov 3 17:16:12 CET 2008
473 - (libcoro) use a global asm statement to become independent of gcc
474 otpimisations for CORO_ASM (thanks to pippijn for the idea).
475 - try to workaround yet another broken bsd, this time dragonfly.
476
4774.802 Thu Oct 30 10:56:12 CET 2008
478 - support -fno-omit-frame-pointer on x86 with the assembly method.
479 - tune 01_unblock.t tests a bit.
480
4814.801 Wed Oct 22 18:33:37 CEST 2008
482 - improve readline speed for very long "lines".
483 - backport to 5.8.8.
484
4854.8 Thu Oct 2 13:34:40 CEST 2008
486 - new function Coro::AIO::aio_wait.
487 - Coro.:AIO and Coro::BDB now "use Coro::AnyEvent".
488 - greatly speed up and reduce memory usage of Coro::AIO requests.
489 - implement some other µ-optimisations.
490
4914.749 Mon Sep 29 14:40:12 CEST 2008
492 - port to slow and broken pseudo-threaded perls. (courtesy pippijn).
493
4944.748 Sat Sep 27 14:03:19 CEST 2008
495 - implement, but do not document, PerlIO::cede(granularity).
496 - Coro::Storable forgot to wrap Storable::pstore.
497 - work around the multitude of leaks and memory corruption
498 bugs in PerlIO::via by using our own C-level perliol. As a side
499 effect, Coro::Storable is now much, much, much faster.
500
5014.747 Tue Sep 23 01:59:41 CEST 2008
502 - fix a per-cv memleak (one empty array was leaked per
503 code reference).
504 - avoid a crash in coro->call|rss when the coroutine was already
505 destroyed (most noticably when using Coro::Debug::ps :)
506 - also protect *Storable::FILE.
507 - push up default storable granularity to 20ms.
508
5094.746 Sun Sep 21 03:22:20 CEST 2008
510 - be more insistent on locking Storable against reentrancy
511 in Coro::Storable.
512 - move swap_def?v and throw to Coro::State, as documented.
513
5144.745 Thu Jul 24 00:14:38 CEST 2008
515 - remove debugging code related to MgPV_nolen_const, also try to
516 make it compile with perl 5.8.6 (yes, apple apparently loves
517 outdated software). Reported by John S.
518
5194.744 Tue Jul 8 22:06:35 CEST 2008
520 - correctly provide default DIE/WARN handlers as documented.
521 - also overwrite PL_vtbl_sigelem.svt_clear, even though current
522 implementations inside perl work fine for us.
523
5244.743 Mon Jun 16 00:21:57 CEST 2008
525 - when using Coro::EV without running EV::loop it could
526 result in busy-waiting for events, this has been fixed.
527 - reduce codesize and improve performance by using EV_DEFAULT_UC.
528
5294.742 Sat May 31 14:10:21 CEST 2008
530 - implement a workaround for (some) perl <5.8.8 versions.
531 - require EV 3.3+.
532
5334.741 Fri May 30 23:33:09 CEST 2008
534 - tell netbsd how utterly broken their imitation of an OS is
535 and refuse to build by default if pthreads are in use.
536 - switch to "s" method on all bsd's by default, as their ucontext
537 stuff seems just too broken.
538 - fix a bug in Coro::Select.
539
5404.74 Thu May 29 20:05:31 CEST 2008
541 - do not test Coro::LWP for lack of dependencies.
542 - do not test Coro::Select for lack of working perls.
543
5444.73 Thu May 29 2008
545 - fix a bug in Coro::EV which would cause it to block despite
546 there being runnable coroutines.
547 - sprinkle "no warnings" freely over everything, also suppress
548 warnings for some other modules.
549 - fix typo in WSAEWOULDBLOCK.
550
5514.72 Sun May 25 05:14:36 CEST 2008
552 - tweak META.yaml a bit, unfortunately, there is no documented way
553 to have optional dependencies with CPAN. doh :(
554 - avoid running some tests on windows because they would fail due to
555 perl bug (broken fork, broken pipes...).
556 - work around perl on windows bugs where perl returns undocumented
557 error codes for sysread, syswrite etc. by taking advantage
558 of AnyEvent's workaround for that problem.
559 - use AnyEvent::Util::fh_nonblocking in Coro::Handle to work around
560 a common perl implementation bug on windows.
561 - use unix domain sockets in testsuite to work around
562 common perl implementation bugs on widows (they are emulated by
563 tcp sockets on windows. ugh.)
564
5654.71 Sat May 24 20:01:27 CEST 2008
566 - fix a bug in Coro::AnyEvent ("Usage: Coro::AnyEvent::_schedule()").
567 - take advantage of async name resolution of AnyEvent::Util.
568 - work around brutal inet_aton override in Coro::LWP.
569 - take advantage of the readyhook in Coro::EV, for smoother
570 scheduling.
571
5724.7 Sun May 11 00:32:19 CEST 2008
573 - completely reworked the Coro manpage.
574 - added Coro::AnyEvent, generic event loop integration.
575 - implement cancel, ready and kill commands in Coro::Debug.
576 - document find_coro in Coro::Debug.
577 - incompatible change: rename has_stack to has_cctx.
578 - Coro::AIO and Coro::BDB no longer force event model detection,
579 use AnyEvent::AIO and AnyEvent::BDB.
580
5814.6 Sat Apr 26 10:05:14 CEST 2008
582 - INCOMPATIBLE CHANGE: sub/code attributes are no longer supported
583 by the Coro module. It was a mistake to have it in the first place.
584 - (experimental) support for activestate perl 5.10 (method "w").
585 - (experimental) support for strawberry perl (method "a").
586 - coro_sigelem_set did not return a value although it had to,
587 actual impact unknown.
588
5894.51 Mon Apr 14 13:28:27 CEST 2008
590 - make it compile again on 5.8.
591
5924.50 Thu Apr 10 09:43:17 CEST 2008
593 - I did it twice! (see 4.49).
594
5954.49 Sun Apr 6 21:23:31 CEST 2008
596 - grr, instead of compiling the recent changes on 5.10 only they
597 were compiled on 5.8 only.
598
5994.48 Sun Apr 6 20:36:46 CEST 2008
600 - allow coroutine switches during eval's under 5.10.x, as it seems
601 the parser is a per-interpreter option now, so this is safe
602 (this might fix the odd crashing bug).
603 - drop support for 5.9.x versions: they are dead, jim.
604
6054.47 Sun Apr 6 00:37:52 CEST 2008
606 - force cctx allocation on API calls: we know we need to force one
607 and gcc actually manages to confuse our heuristic nowadays,
608 leading to crashes and worse.
609 - document force_cctx.
610
6114.46 Fri Apr 4 22:05:43 CEST 2008
612 - upgrade libcoro, resulting in pthread-backend (which was only created
613 to fulfill the rules of the programming languages shootout).
614
6154.45 Thu Mar 13 11:55:36 CET 2008
616 - fix a file leaking bug in eg/mhyttpd that would allow
617 downloading of any file (reported by oesi).
618 - fix deadlock bug in Coro::Channel (reported by Richard Hundt)
619 (also add testcase).
620 - support Broadcast option in Coro::Socket (patch by Richard Hundt,
621 apparently having loads of fun with that).
622 - implement $state->swap_defsv, ->swap_defav and document ->throw.
623
6244.4 Wed Feb 13 16:44:29 CET 2008
625 - only assume .cfi pseudo isns exist on GNU/Linux.
626 - add get_timed method to Coro::Channels.
627 - fixed Coro::Debug manpage.
628 - perl 5.11 compatbility improvement based on patch by
629 Andreas König.
630
6314.37 Sun Jan 20 11:25:23 CET 2008
632 - soften the check for an active parser for perl 5.10 (avoiding
633 "Coro::State::transfer called while parsing" in many cases).
634
6354.36 Sun Jan 13 10:53:56 CET 2008
636 - reset diehook when terminating from an async_pool as to not
637 trigger any __DIE__ handlers.
638
6394.35 Sun Jan 13 04:14:13 CET 2008
640 - "bt" debug command now displays any exceptions
641 from longmess and also skips the innermost
642 stackframes, giving more useufl output.
643 - allow backtraces in coroutines blocked in Coro::EV,
644 at a <1% speed hit.
645 - handle localising of $SIG{__WARN__} and __DIE__
646 properly (with a proper amount of dirty hacking).
647
6484.34 Sat Dec 22 17:49:53 CET 2007
649 - upgrade to EV version 2.0 API.
650
6514.33 Mon Dec 17 08:36:12 CET 2007
652 - make Coro::AIO etc. loadable in the absence of EV.
653
6544.32 Mon Dec 17 07:46:02 CET 2007
655 - majorly improved Coro::Handle's performance with Coro::EV.
656 - implemented the until now mythical Coro::BDB module.
657 - specialcase EV in Coro::AIO and Coro::BDB for extra speed.
658
6594.31 Wed Dec 5 12:32:39 CET 2007
660 - remove warn statement form Coro::Util.
661
6624.3 Tue Dec 4 20:33:14 CET 2007
663 - calls to the idle function could cause stack corruption
664 when the stack changed.
665 - do no longer rely on the presence of EV::DNS if EV is used
666 (because it is gone), but instead take avdantage of EV::ADNS
667 if available.
668 - add ($) prototypes to all functions in Coro::Storable.
669 - use a conventional (safer) idle callback in Coro::EV.
670 - do accept quickly in Coro::Debug to avoid endless loops.
671
6724.22 Fri Nov 30 16:04:04 CET 2007
673 - really use optimised versions for Event and EV in Coro::Util
674 and Coro::Handle.
675
6764.21 Sun Nov 25 10:48:59 CET 2007
677 - fix a spurious memory read.
678 - Coro::EV no longer keeps the eventloop "alive".
679
6804.2 Fri Nov 9 20:47:05 CET 2007
681 - enable/disable tracing from a new coroutine, not a pooled one.
682 - fix a memleak in Coro::Event.
683 - removed killall call from fork_eval.
684 - made sure store_fd is already loaded so that fork_eval
685 does not have to parse autoload in each subprocess.
686 - only use assembly method if -O switch is in $Config{optimize}.
687 - add (optional) Coro::EV module, so far the best event loop module
688 directly supported by Coro.
689 - if the event model is EV, use EV::DNS to resolve
690 stuff in Coro::Util.
691 - don't get confused by multiple event notifications in Coro::Handle.
692 - initial support for EV (libevent interface).
693 - require Event and EV using configure_requires, to force their existance.
694
6954.13 Wed Oct 24 07:26:45 CEST 2007
696 - add Coro::Storable::blocking_thaw.
697 - use a vastly more complicated technique to localise
698 $SIG{__WARN/DIE__} that also works on perls <= 5.8.8.
699 - use a coroutine for the idle callback Coro::Event,
700 instead of running Event in the current coroutine context.
701 This also catches recursive invocations.
702 - actually report fork errors in gethostbyname and inet_aton.
703
7044.11 Thu Oct 11 02:40:24 CEST 2007
705 - port to threaded perls.
706
7074.1 Thu Oct 11 02:38:16 CEST 2007
708 - incompatible change: $SIG{__DIE__} and $SIG{__WARN__} will now
709 be local to each coro (see Coro::State).
710 - incompatible change: for very deep reasons, cede and cede_notself
711 cannot return anything, so nothing will be returned.
712 - possibly bring back 5.10 compatibility (untested).
713 - work around stupid (and wrong) warning on 5.10 :(.
714 - overlay the saved state over the context stack. This saves
715 a few hundred bytes per coroutine on average and also
716 speeds up context switching a bit.
717 - further tune default stack sizes.
718 - (more) correctly calculate stack usage in coro_rss.
719 - Coro::Storable::blocking_* did not properly lock
720 resulting in races between coroutines.
721 - added Coro::Storable::guard.
722 - stopping to trace a coroutine could destroy the cctx of
723 an unrelated coroutine.
724 - explain the relationship between Perl and C coroutines in
725 more detail in Coro::State.
726 - Coro::Util::inet_aton did not short-circuit dotted quad forms,
727 causing a fork per resolve. This also affected Coro::Socket.
728 - switch to a separate stack in $coro->call/eval to avoid
729 invalidating pointers.
730
7314.03 Sat Oct 6 21:24:00 CEST 2007
732 - added Coro::throw method.
733 - minor code cleanups.
734
7354.02 Sat Oct 6 02:36:47 CEST 2007
736 - fix a very minor per-coroutine memleak (a single codereference).
737 - fixed a bug where the currently in-use c context would be freed
738 prematurely (can happen only when programs change the stacksize
739 or use tracing).
740 - tracing can no longer keep a coro alive after it terminated.
741 - do static branch prediction in the common path for gcc. gives
742 about 2-5% speed improvement here.
743
7444.01 Fri Oct 5 22:10:49 CEST 2007
745 - instead of recreating *a* standard output handle we simply
746 use STDOUT, which is faster and hopefully more robust.
747
7484.0 Fri Oct 5 12:56:00 CEST 2007
10 - incompatibly changed Coro::Storable::freeze. 749 - incompatibly changed Coro::Storable::freeze.
11 - major new feature: added Coro::Debug, for interactive coroutine 750 - major new feature: added Coro::Debug, for interactive coroutine
12 debugging, tracing and much more. 751 debugging, tracing and much more.
752 - major bug fix: unbelievable, but true: $_, $/ and many other
753 "saved" variables actually weren't being saved. This has been fixed,
754 of course, while increasing performance while losing all the save
755 flags.
756 - save flags are gone, and all the api functions dealing with them.
13 - added Coro::Semaphore::adjust. 757 - added Coro::Semaphore::adjust.
14 - added Coro::Util::fork_eval. 758 - added Coro::Util::fork_eval.
15 - added Coro::Storable::{nfreeze,blocking_{freeze,nfreeze}}. 759 - added Coro::Storable::{nfreeze,blocking_{freeze,nfreeze}}.
760 - added Coro::killall.
16 - reduce initial stack sizes to allow for "micro-coroutines". 761 - reduce initial stack sizes to allow for "micro-coroutines".
17 - better async_pool resource management, moved parts of async_pool 762 - better async_pool resource management, moved parts of async_pool
18 handling to XS (major speed improvement). 763 handling to XS (major speed improvement).
19 - actually croak before modifying important data structures. 764 - actually croak before modifying important data structures.
20 - refuse to transfer while compiling. 765 - refuse to transfer while compiling.
21 - possibly support eval EXPR better now. 766 - possibly support eval EXPR better now.
22 - enable assembly method per default on linux+bsd x86+amd64. 767 - enable assembly per default on linux+bsd x86+amd64.
768 - all internal members were renamed _something for easier subclassing.
23 - many minor tweaks. 769 - many minor tweaks.
24 770
253.63 Wed May 16 14:10:06 CEST 2007 7713.63 Wed May 16 14:10:06 CEST 2007
26 - implement handcoded assembly for x86/amd64 SVR ABI. 772 - implement handcoded assembly for x86/amd64 SVR ABI.
27 773

Diff Legend

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