ViewVC Help
View File | Revision Log | Show Annotations | Download File
/cvs/Coro/Changes
Revision: 1.664
Committed: Fri Jul 14 02:43:50 2017 UTC (6 years, 10 months ago) by root
Branch: MAIN
Changes since 1.663: +1 -0 lines
Log Message:
*** empty log message ***

File Contents

# Content
1 Revision history for Perl extension Coro.
2
3 TODO: should explore PerlIO::coroaio (perl leaks like hell).
4 TODO: channel->maxsize(newsize)?
5 TODO: __GCC_HAVE_DWARF2_CFI_ASM
6 TODO: swap_sv, maybe add scope_swap_sv?
7 TODO: croak when async_pool tries to run canceled thread?
8
9 - experimental 5.24 compatibility changes.
10 - re-try JIT allocation without PROT_EXEC, to hopefully improve
11 portability.
12
13 6.511 Sun Jun 26 23:44:50 CEST 2016
14 - make it compile with threaded perls again.
15 - simplify and speed up __DIE__ and __WARN__ handling.
16
17 6.51 Sat Jun 25 20:57:20 CEST 2016
18 - 6.5 release didn't compile on most older perls due to
19 a logic error enabling experimental 5.24 code for all versions.
20 - since various broken or deficient patches float around
21 to work around the vtbl API breakage in perl, let's provide
22 our own workaround, which seems to be much less invasive
23 and more compatible than the approaches seen so far. YMMV.
24 - work around assertion failure in perl_destruct on 5.24
25 (analyzed by Dave Mitchell).
26
27 6.5 Wed Jun 22 22:23:50 CEST 2016
28 - swap_sv swapping order was not symmetrical, causing
29 wrong swaps when swap_sv was used multiple times on the
30 same sv in the same thread.
31 - swap_sv calls can now be undone by calling it again
32 with the same variables.
33 - swap_sv calls will now be undone in async_pool threads.
34 - split Coro::Semaphore::up/adjust into separate xs functions
35 for better error reporting, at slight codesize increase.
36 - (libcoro) arm assembly support, please test and report.
37 - adjust to PL_savestack changes in perl 5.24 (adapted from
38 the debian patch, which unfortunately gets it wrong).
39
40 6.49 Sat Oct 17 01:40:12 CEST 2015
41 - throwing an exception to a thread waiting in
42 Coro::Handle using Coro::EV did not stop the watchers,
43 causing the next call to fail (testcase by Martin Pritchard).
44 - bump minimum perl version to 5.10.
45
46 6.48 Sun Oct 4 19:03:51 CEST 2015
47 - fix memory leak when Coro::Handle uses Coro::EV internally
48 (testcase by Sten Sten).
49 - update code to libev 4 API (internally, Coro still used the
50 version 3 API calls).
51 - remove unused hv_sig.
52
53 6.47 Sat Jul 11 03:58:20 CEST 2015
54 - swap_svs - NVs can be in the SV head beginning with 5.20.
55
56 6.46 Tue Jun 30 14:36:30 CEST 2015
57 - restore portability to perls with windows fork emulation
58 (patch by Petr Písař).
59
60 6.45 Tue Jun 30 01:40:08 CEST 2015
61 - restore portability to perl 5.16 and below.
62
63 6.44 Tue Jun 30 00:41:54 CEST 2015
64 - avoid segfaulting (or worse) when tracing a canceled thread.
65 - the Event module silently broke it's public hook API, causing
66 failures on perls compiled with -Duselongdouble. Requiring
67 the latest version of Event because I don't know which release
68 changed this (it's not mentioned in the ChangeLog).
69 - allow xs level enterleave hooks via CoroAPI.
70
71 6.43 Thu Jun 4 15:38:14 CEST 2015
72 - use stability canary.
73 - port to stableperl-5.22.0-1.001.
74 - update libecb.
75
76 6.42 Wed Feb 11 20:29:52 CET 2015
77 - Coro::SemaphoreSet->try did not actually work (analyzed by
78 SATO Kentaro).
79 - upgrade libecb to be C11 compliant.
80
81 6.41 Sat Sep 6 22:08:46 CEST 2014
82 - restore portability to perl 5.8.x.
83 - give new Coro's a valid GvHV(PL_hintgv) - this is slow and takes
84 up some memory, but fixes "use feature" and similar modules
85 when used inside a Coro.
86 - allow zero as argument to Coro::Channel to mean the same thing
87 as no argument. this works with older versions as well,
88 but wasn't legal until now.
89 - slightly better c header file detection.
90
91 6.39 Mon Jun 2 00:00:08 CEST 2014
92 - work around more incompatible changes in 5.20.
93
94 6.38 Sun Jun 1 21:54:23 CEST 2014
95 - check that perl slots actually have enough space to hold
96 interpreter variables.
97 - untested port to perl 5.19 (64 bit tmps indices) (reported
98 by Andreas König).
99 - croak when cancel is called without a thread context.
100
101 6.37 Tue Mar 4 13:27:33 CET 2014
102 - *sigh*, removed leftover debugging code from debugging a
103 perl bug, of all things.
104
105 6.36 Tue Mar 4 07:11:59 CET 2014
106 - semaphores would not clear the destroy hook when interrupted
107 by ->throw, causing segfaults or worse.
108 - ->throw on a thread waiting for a semaphore did not acquire
109 the semaphore, but also didn't wake up other waiters,
110 possibly causing a deadlock.
111 - "FATAL: $Coro::IDLE blocked itself" will now use Carp::confess
112 to report a full stacktrace, which should help find out
113 where the actual call is.
114 - "atomically" destroy data in slf_destroy, because it is
115 the right thing to do, just in case.
116 - disable ecb.h libm dependency, to compile on ancient systems
117 or under adverse conditions.
118
119 6.33 Mon Nov 18 11:26:27 CET 2013
120 - do not crash when freeing padlists with holes (in 5.18).
121 - tentative SVt_BIND 5.19 port/fix.
122
123 6.32 Tue Nov 5 15:35:35 CET 2013
124 - use a new algorithm to derive padlists for perl 5.18. The old
125 one could lead to 0-pointer accesses inside perl (reported
126 by Darin McBride).
127
128 6.31 Thu May 9 07:39:48 CEST 2013
129 - Coro::AIO requests would crash if the thread was ready'd
130 while the request was ongoing.
131
132 6.29 Wed May 8 02:55:18 CEST 2013
133 - when an on_destroy handler destructs the coro currently being
134 destructed a perl scalar could be accessed after being freed,
135 likely causing a crash.
136
137 6.28 Wed Mar 6 06:58:02 CET 2013
138 - clean remnants of existing __DIE__ and __WARN__ handlers so
139 they lose their magic and will not cause segfaults later
140 (testcase by Andrey Sagulin).
141 - improved Coro::State documentation a bit.
142 - Coro::Debug::command now flushes the output.
143 - add hack detection code for x32 abi, because the braindead slugs
144 who designed that made it look exactly like x86_64 without
145 providing proper compile time symbols to test for it. as a result,
146 this detection cannot work reliably.
147 - valgrind stack registering was broken.
148 - do not rely on Time::HiRes anymore in Coro::Debug.
149
150 6.23 Fri Dec 7 23:36:37 CET 2012
151 - use experimental fiber implementation on native windows
152 perls.
153 - use sizeof (void *) as multiplication factor for stack sizes,
154 to accomodate the totally braindamaged microsoft 64 bit "os".
155 - changed verifier host from win2k-ap510-32 to win7-sp516-32/64.
156 activeperl 5.16 crashes when PerlIO_define_layer is called due
157 to some bug in the perl dll, strawberry perl at least passes
158 the testsuite.
159 - implement Coro::Handle->peeraddr/host/port, for slightly
160 improved compatibility with LWP.
161 - implement 5.17 compatibility by almost blindly applying a
162 good-looking patch by Father Chrysostomos.
163 - move stack management functions into libcoro 3.
164 - libcoro version 3 "released".
165 - support magic values as timed_io_once args.
166 - recommend AnyEvent 7+ or EV 4+, also require EV
167 version 4 or newer for Coro::EV.
168
169 6.10 Tue Oct 9 01:14:27 CEST 2012
170 - updated ecb.h, it had a typo that caused it to not compile on many
171 big endian systems (reported by many people).
172 - disable memory fences in ecb.h to improve portability.
173
174 6.09 Sat Oct 6 23:25:02 CEST 2012
175 - Coro::EV I/O watchers were not interruptible by exceptions
176 (Coro::State::throw) (testcase by sten).
177 - ->throw now puts threads into the ready queue, as this seems to
178 be expected by existing code, and code that doesn't cope with spurious
179 wakeups needs fixing anyway.
180 - use fd -1 in mmap.
181 - cast I32 to int in error message printf.
182 - warn about broken so-called "hardened" kernels.
183
184 6.08 Fri Apr 13 12:05:47 CEST 2012
185 - be more aggressive about exiting like perl does - formerly,
186 exiting from the non-main thread would not execute END blocks.
187
188 6.07 Fri Nov 11 21:21:48 CET 2011
189 - work around a bug in PerlIO (setting $SIG{__WARN__} to a PVCV).
190 - update ecb.h.
191
192 6.06 Mon Aug 8 23:59:48 CEST 2011
193 - cygwin unfortunately patches the stack at runtime, so we use the pthreads
194 backend, which is an order of magnitude slower. unfortunately, cygwins
195 pthread implementation isn't very complete either, so allocate the stack
196 twice just to be sure.
197 (note: cygwin also enables mymalloc, which is NOT THREADSAFE ON WINDOWS,
198 in its ithreaded perl - best recompile cygwin and use the 'w'indows
199 backend for much better performance. also disable ithreads for
200 even better performance...).
201
202 6.05 Thu Aug 4 21:36:36 CEST 2011
203 - blush, condvar values would not be propagated from send to recv anymore
204 (reported by Chip Salzenberg).
205 - use exponential increase for the readline buffer length in
206 Coro::Handle. also reduce initial allocation to 1020 from 4096 bytes.
207
208 6.04 Wed Aug 3 17:15:45 CEST 2011
209 - use even more efficient and more compatible condvars for
210 compatibility to AnyEvent 6.x :)
211 - more inconsequential ecb.h updates.
212
213 6.03 Wed Aug 3 11:41:30 CEST 2011
214 - change how Coro patches AnyEvent condvars for compatibility to
215 AnyEvent 6.x.
216 - update ecb.h, to no longer include <pthread.h> in case WinNT.h
217 hasn't been included.
218
219 6.02 Wed Jul 13 04:35:19 CEST 2011
220 - "improve portability to Gentoo" - gentoo manages to put perl variables
221 in memory areas that are farther than 2gb apart, which the jit couldn't
222 handle and barfed. now it's just a bit slower on gentoo and similar
223 systems.
224
225 6.01 Sun Jul 3 12:31:14 CEST 2011
226 - workarounds are good, but the test for whether pthreads are used
227 was not good. this one should be better.
228 - check differently whether gcc generates cfi instructions itself.
229
230 6.0 Wed Jun 29 19:43:35 CEST 2011
231 - INCOMPATIBLE CHANGE: unreferenced coro objects will now be
232 destroyed and cleaned up automatically (e.g. async { schedule }).
233 - implement a JIT compiler for part of the thread switch code,
234 which gives a 50% speed improvement on threaded perls, and
235 about 4% on non-threaded perls (so threaded perls now finally
236 reach about half the speed of non-threaded perls).
237 - slightly modernise Coro::Intro, add section about rouse functions.
238 - avoid DEFSV and ERRSV, giving another 10% improvement
239 in thread switching.
240 - Coro::State->is_destroyed is now called is_zombie.
241 - implement a Coro->safe_cancel method that might fail, but
242 cancels in a "safer" way if it succeeds.
243 - add preliminary support for DEBUGGING perls.
244 - get rid of two hash-accesses when initialising a new Coro - this
245 speeds up coro creation by almost a factor of two.
246 - croak when a coro that is being cancelled tries to block
247 (e.g. while executing a guard block), instead of crashing or
248 deadlocking.
249 - use a more robust and also faster method to identify Coro::State
250 objects - speeds up everything a bit.
251 - implement Coro->cancel in XS for a 20% speed improvement, and to
252 be able to implement mutual cancellation.
253 - speed up context switches by a percent or two by more efficiently
254 allocating context stack entries.
255 - implement Coro->join and Coro->on_destroy in XS for a speedup and
256 a reduction in memory use.
257 - cancelling a coro while it itself is cancelling another coro is
258 now supported and working, instead of triggering an assertion.
259 - be a bit more crash-resistant when calling (buggy) on_destroy
260 callbacks (best effort).
261 - move on_destroy into the slf_frame, to allow extension slf
262 functions to have destructors.
263 - get rid if coro refcounting - simply crash in other interpreter
264 threads by nulling the pointers on clone.
265 - simplify warn/die hook handling when loading Coro - the convoluted
266 logic seems to be no longer neccessary.
267 - use libecb instead of our own home-grown gcc hacks.
268 - document alternatives to Coro::LWP. Please use them :)
269 - work around another mindless idiotic NEEDLESS bug in openbsd/mirbsds
270 sigaltstack. Really. wine suffers from it, erlang suffers from it,
271 and it's known since at least 2006.
272
273 5.372 Wed Feb 23 06:14:30 CET 2011
274 - apparently mingw doesn't provide a working gettimeofday, try to
275 work around that by relying on Time::HiRes (indirectly brought to
276 my attention by Max Maischein).
277 - fix some portability issues when Time::HiRes was used.
278
279 5.371 Mon Feb 21 14:36:08 CET 2011
280 - backport to windows process emulation code again.
281
282 5.37 Sat Feb 19 07:49:44 CET 2011
283 - add a big "Coro thread life cycle" section to "man Coro".
284 - try a tentative workaround against the breakage that 5.13 has
285 introduced without depreciation period. sigh.
286 - no longer use Time::HiRes if gettimeofday is available, which
287 saves quite a lot of memory.
288
289 5.36 Sun Feb 13 05:33:41 CET 2011
290 - automatically load Coro::Channel, Coro::RWLock, Coro::Semaphore,
291 Coro::SemaphoreSet, Coro::Signal and Coro::Specific on first "new"
292 method call.
293 - undocument Coro::Timer::sleep and obsolete whole module.
294 - optimise Coro::Timer::timeout memory and cpu usage.
295 - slightly updated Coro::Intro for recent changes.
296 - do not initialise PL_dirty anymore.
297
298 5.25 Thu Nov 11 01:08:39 CET 2010
299 - try a different approach on netbsd - netbsd 5 finally has marginally
300 working pthreads, but still broken ucontext/sigaltstack.
301 - openbsd 4.8 finally got their act together, Coro works out of the box
302 with asm, setjmp and pthreads (no change, just informational).
303
304 5.24 Sat Oct 23 11:27:12 CEST 2010
305 - port to the EV 4.0 API.
306 - work around bugs in mingw32, making strawberry perl work
307 out of the box.
308 - correctly modify Coro::AIO function prototypes
309 so that they reflect the "no optional parameters" rule.
310 - "ported" libcoro to C++.
311
312 5.23 Mon May 17 18:50:42 CEST 2010
313 - be more resistant to ordering changes when initialising
314 Coro::AnyEvent, Coro::EV and Coro::Event (reported by Matthias
315 Waldorf).
316 - document that perl 5.12 deliberately removed support for cloning.
317
318 5.22 Wed Apr 14 03:55:35 CEST 2010
319 - correctly return udnef on errors in Coro::Handle::read/write
320 (testcase by Marc Mims).
321 - convert Coro::Util into a "perl compatibility wrapper" - the functions
322 are less useful now, but are drop-in replacements for existing
323 functions, listing better alternatives in the documentation. This also
324 fixes a bug in Coro::LWP which naively substituted Socket::inet_aton
325 with Coro::Util::inet_aton.
326 - do not override $Coro::idle unconditionally in Coro.pm, as other
327 modules could have provided their own idle coro already
328 (for exmaple, Coro::AnyEvent).
329 - fix Coro::Util::gethost* functions.
330 - Coro::Timer corretcly exports it's symbols (reported by Hideki Yamamura).
331
332 5.21 Wed Dec 16 07:19:51 CET 2009
333 - automatically load Coro::AnyEvent when AnyEvent and Coro are used
334 together.
335 - add some examples on how to combine other event loops with Coro in
336 Coro::AnyEvent, and how to run it (and not to block). Seems to be
337 the most common source of confusion.
338 - try to catch people naively blocking in an event callback.
339 - work around the perl filehandle bug issue in conjunction with
340 older common::sense (as indirectly pointed out by ZSystem).
341 - clarify the "not from signal handlers" section.
342
343 5.2 Sun Oct 4 14:54:24 CEST 2009
344 - Coro::Storable destroyed the prototypes of the functions it wrapped.
345 - export rouse_cb and rouse_wait by default now.
346 - fix various prototype mismatches in Coro::AnyEvent and Coro::Handle.
347 - new method $state->swap_sv.
348 - added section on "windows process emulation" to the manpage,
349 after a not-so-fruitful (nor-friendly) "discussion" with chip
350 salzenberg (discussion implies arguments, but his only arguments
351 were ad-hominems, one wonders why he started it in the first
352 place). I hope this explains it well enough for him to understand,
353 and maybe well enough for others to understand.
354 - use common::sense everywhere now.
355 - idle callbacks are no longer supported, use idle coros instead.
356 - print a thread listing when a deadlock is detected.
357
358 5.17 Sat Aug 22 23:09:31 CEST 2009
359 - work around a bug in the perl debugger causing crashes
360 when running under the debugger by marking _pool_handler as nodebug.
361 - speed up Coro::async considerably.
362 - try some hacks to get netbsd to work "more often" - their broken
363 setjmp/longjmp, ucontext *and* phtreads are really hard on Coro.
364 - convert Coro to AE 5.0 API.
365
366 5.162 Tue Jul 28 04:04:03 CEST 2009
367 - perl 5.8.2 is now minimum requirement.
368 - skip t/19_handle.t on broken windows perls.
369
370 5.161 Wed Jul 22 04:47:38 CEST 2009
371 - Coro::AnyEvent::poll could have a different prototype when EV was
372 used as backend (analyzed by Tatsuhiko Miyagawa).
373 - Coro::AnyEvent errornously initialised the event loop when loaded,
374 not on demand.
375 - try to workaround rare */t/01_unblock.t failures.
376
377 5.16 Tue Jul 21 01:44:37 CEST 2009
378 - Coro::AnyEvent failed to hook into the event loop
379 when no threads had been readied between detecting
380 the event loop and actually running it.
381 - considerably speed up Coro::Select by taking avdantage
382 of AnyEvent > 4.8 and some other optimisations.
383 - implement paragraph readline mode in Coro::Handle
384 (based on patches by Zsbán Ambrus).
385 - replace WSAEINPROGRESS by WSAEWOULDBLOCK (reported
386 and analyzed by Yasuhiro MATSUMOTO).
387 - clarified libcoro license and copyright.
388 - someone stole my EXTRA_META!!!
389 - implement Coro::Select::patch_pp_sselect and it's brother,
390 for hardcode select overriding.
391
392 5.151 Mon Jul 6 05:41:57 CEST 2009
393 - backport to windows process emulation code again (patch by
394 Yasuhiro MATSUMOTO).
395 - slightly update Coro::MakeMaker.
396
397 5.15 Tue Jun 30 10:28:06 CEST 2009
398 - deprecate Coro::Socket, document how to get ipv6 support via
399 AnyEvent::Socket instead.
400 - implement signal->wait ($cb) interface, similar to semaphores.
401 - work around SvOK not supporting getmagic, so we have to getmagic
402 to test for undef :/ (reported by Matthias Waldorf).
403 - load Coro::AnyEvent in all modules using AnyEvent.
404 - work around perl corrupting our internal data structures,
405 reported by Tokuhiro Matsuno.
406 - enable per-coroutine real and cpu time gathering
407 (Coro::State::enable_times).
408
409 5.14 Wed Jun 24 01:37:48 CEST 2009
410 - provide explicit functions to "cede" to the event loop in Coro::AnyEvent,
411 as this seems to have been a difficult concept (poll, sleep, idle,
412 idle_upto).
413 - add Coro::AnyEvent::readable/writable functions.
414 - clarify Coro::EV/Event/AnyEvent manpages.
415 - free per-thread global scalars in the thread calling ->cancel, to
416 avoid crashes when $_, $@ etc., are magical but some of those
417 had already been freed.
418 - "unexperimentalise" the callback interface for Coro::Semaphore.
419 - speed up ready queue management/context switching by using a linked
420 list instead of an array (~5-10%).
421 - implement "watch" command in Coro::Debug shells.
422 - for fun, implement time-slicing as an example in the manpage.
423 - if AnyEvent detects EV or Event, but we don't have Coro::EV or
424 Coro::Event, use the normal AnyEvent handling instead of dieing
425 (the same is true for Coro::Handle).
426 - properly document Coro::EV::timed_io_once.
427 - avoid unneccessary ->cancel calls in Coro::Handle.
428 - maybe make it work on mingw32 with win32 backend
429 (based on patch by Yasuhiro Matsumoto).
430
431 5.132 Fri May 29 09:00:39 CEST 2009
432 - do not keep a reference to the argument itself in
433 Coro::Semaphore::guard, as it could change later.
434 - support SO_RCVBUF/SO_SNDBUF nonstandard Coro::Socket options,
435 should support a prepare callback.
436
437 5.131 Mon Mar 16 23:20:37 CET 2009
438 - implement and document Coro->suspend, Coro->resume.
439 - fix Coro::Select implementation to not (often) close
440 the passed file descriptors (testcase provided by pippijn).
441
442 5.13 Mon Dec 15 21:51:42 CET 2008
443 - EXPERIMENTAL: implement dynamic winds (on_enter/on_leave).
444 - don't set diehook to C<undef> but instead to NULL, to avoid
445 spurious warnings.
446 - fix a lot of bugs in Coro::SemaphoreSet.
447 - Coro::SemaphoreSet will less often create a semaphore needlessly.
448 - add Coro::SemaphoreSet::count and wait methods.
449 - take advantage of the new Guard module.
450 - deprecate Coro::guard.
451 - try to fix the dreaded 01_unblock tests once more. I hate it when
452 testsuites need more fixing than the code they are supposed to test.
453 - croak in more cases when a required callback isn't resolvable.
454 - fix some minor issues in Coro::State->call/eval.
455 - use current coroutine context instead of a temporary one
456 when temporarily switching to another coroutine.
457 - do not call C-level on_destroy handlers during global destruction,
458 to avoid needless segfaults.
459
460 5.12 Sun Dec 7 13:30:38 CET 2008
461 - add default config for MirOS, which seems to be a bug-to-bug
462 compatible fork of openbsd ("world domination by releasing
463 openbsd cvs before the openbsd folks do it" or so :).
464 - free_padlist did destroy the names pad, not good, but didn't
465 seem to bother perl - this could fix issues such as eval ""
466 inside a function called from multiple coroutines.
467 - use a different method to detect destruction time.
468 - be more careful when freeing padlists just before global
469 destruction.
470 - fixed and expanded the call/cc example.
471 - renamed _terminate to _coro_run.
472 - new method Coro::Channel->shutdown.
473 - try pthreads on openbsd <4.4 (broken sigaltstack, will
474 pthreads fare better?).
475 - be less picky about destroying "a" running coroutine.
476
477 5.11 Tue Nov 25 21:49:05 CET 2008
478 - DEBUGGING in 5.10.0 is a mess: it constantly flags perfectly
479 working code with failed assertions, introducing more bugs than
480 it fixes, requiring elaborate workarounds :(
481
482 5.1 Mon Nov 24 08:54:59 CET 2008
483 - wrote a small introductory tutorial - Coro::Intro.
484 - convert Coro::Timer, Coro::Select and Coro::Util to rouse API.
485 - Coro::Select did errornously dup the file descriptors
486 and didn't work with all AnyEvent backends.
487 - Coro::Select wasn't imported correctly form Coro::LWP, causing blocking
488 LWP data transfers.
489 - disassociate c contexts from coro objects - this is agruably more
490 correct, but mostly allows sharing of cctxs between coro and state
491 objects, for added memory savings and speed increases.
492 - bumped $Coro::POOL_RSS up to 32kb by default.
493 - no longer set the optype to OP_CUSTOM, as B::* understandably
494 doesn't like this very much (and we *are* a type of entersub).
495 - implement state cloning, just to prove that call/cc can be done.
496 - automatically load Coro::AnyEvent in Coro::Handle.
497 - wrap ->cancel calls in eval inside Coro::Handle as EV watchers
498 do not have this method (and don't need it either).
499 - speed up generic anyevent methods in Coro::Handle by using rouse
500 callbacks.
501 - allow coroutines in $Coro::IDLE, speeding up Coro::AnyEvent and
502 others. It also makes the debugger happier, as you can trace
503 through the idle threads now.
504 - add comppad_name* and hints ($^H, %^H) to per-thread variables.
505 - eg/event was pretty broken.
506 - better 5.8.6 compatibility.
507
508 5.0 Thu Nov 20 10:35:05 CET 2008
509 - NEW ARCHITECTURE: use the latest 4.x version if you experience
510 stability issues.
511 - bump API version to 7 - all dependents must be recompiled.
512 - removed timed_* functions - they were not being used anyways
513 and should be replaced by a more generic mechanism -
514 and were annoying to support anyways :)
515 - removed SemaphoreSet's waiter method - use sem method instead.
516 - Coro::Semaphore->adjust didn't correctly wake up enough waiters.
517 - async_pool did free a scalar value twice
518 ("Attempt to unreference...").
519 - fix a longstanding bug where calling terminate on a coro that
520 was waiting for a semaphore that was just becoming available
521 would cause a deadlock (semaphore would get into a state where
522 it was available but waiters were still blocked).
523 - calling throw on a coroutine that is waiting for a semaphore will
524 no longer make it acquire the semaphore (and thus leak a count).
525 - perl's process emulation is now not even theoretically supported
526 anymore.
527 - new functions Coro::rouse_cb and Coro::rouse_wait for easier
528 conversion of callback-style to blocking-style.
529 - new methods $coro->schedule_to and ->cede_to, to specifically
530 schedule or cede to a specific coroutine.
531 - new function Coro::Semaphore::wait.
532 - use named constants in Coro::Channel (Richard Hundt).
533 - directly patch the entersub opcode calling SLF functions (cede,
534 transfer and so on). this does speed up context switching, but
535 more importanly, it frees us from the hardcoded behaviour of
536 entersub, so we might actually be able to return something from
537 those functions and atcually create new ones.
538 - take advantage of __builtin_frame_address on gcc.
539 - expose THX in coroapi (not sure whether this was a wise decision,
540 as "threaded" perls are running at half speed anyways).
541 - implement execute_slf (schedule-like-function) interface that makes
542 it possible to implement schedule-like-functions in XS.
543 - use new SLF interface to massively speed up Coro::EV by roughly a
544 factor of two.
545 - used new SLF interface to massively speed up Coro::Semaphore by a
546 factor of three.
547 - used new SLF interface to speed up Coro::AIO by roughly a factor of
548 four and reduce its memory usage considerably.
549 - implement Coro::SemaphoreSet purely in terms of Coro::Semaphore,
550 for a nice speedup and vastly more correct behaviour. Also implement
551 a new method "sem" to get at the underlying semaphore object.
552 - implement Coro::Channel in terms of Coro::Semaphore, for a moderate
553 (in comparison) 20-40% speedup.
554 - used new SLF interface to reimplement Coro::Signal gaining
555 some unknown (because I was too lazy), but certain, speedup, and also
556 making signals reliable for the first time.
557 - used new SLF interface and other optimisations to speed up async_pool
558 by a factor of two. It also doesn't rely on perl's exception mechanism
559 to exit anymore. The overhead for terminating an async_pool, coro over
560 a normal async is now very small.
561 - sped up coroutine creation/destruction by 40%.
562 - forgot to include Coro/libcoro/README in the dist for all these years.
563 - work around a freebsd pthreads bug (manual testcancel is required as
564 pthread_cond_wait isn't a cancellation point on freebsd).
565 - use new rouse functions to speed up and simplify Coro::BDB.
566 - make "prefer perl native functions" work with threaded perls.
567 - condense Coro::Debug ps output, hint at v and w flags.
568 - (libcoro) lots of minor cleanups and portability improvements.
569
570 4.914 Wed Nov 19 12:54:18 CET 2008
571 - fix a disastrous bug in the readline optimisation
572 introduced in 4.801.
573
574 4.913 Sat Nov 15 07:58:28 CET 2008
575 - async_pool did free a scalar value twice
576 ("Attempt to unreference...").
577
578 4.912 Thu Nov 13 18:31:23 CET 2008
579 - minor cleanups.
580 - use much larger stacks on linux and perl < 5.8.8.
581 - Coro::Debug::new_unix_server did not unlink the socket
582 when destroyed.
583
584 4.911 Tue Nov 11 04:26:17 CET 2008
585 - "port" to threaded perls.
586
587 4.91 Mon Nov 10 05:36:38 CET 2008
588 - the ->throw exception object no longer leaks.
589 - creating a new cctx leaked a scopestack entry (memleak).
590 - new coroutines didn't get created with a zero flags field
591 (unknown impact).
592 - calling ->throw on a not-yet-started coroutine should now work
593 instead of being ignored.
594 - ->throw is now supported on Coro::State objects.
595 - clean up cctx creation code a bit.
596 - entersub is actually a UNOP, not a LOGOP (not a bugfix).
597
598 4.9 Sat Nov 8 17:45:27 CET 2008
599 - (libcoro) did not preserve rbp with CORO_ASM (we are getting there).
600 - (libcoro) no longer leak threads in the experimental pthread backend,
601 also speed it up considerably.
602 - (libcoro) do not rely on makecontext passing void *'s unscathed.
603 - fix compiletime dependencies on libcoro in the Makefile.
604 - cctx_count wasn't always updated properly.
605 - Coro::State::cctx_stacksize wasn't applied correctly.
606 - new function Coro::State::cctx_max_idle.
607 - the default max number of idle C contexts is now 4.
608 - (libcoro) try harder to get _setjmp/_longjmp.
609 - (libcoro) cleanup and extend the libcoro API to officially
610 allow the creation of empty source contexts.
611 - very experimental workaround for the totally broken netbsd platform.
612 - tried to hack around openbsd bugs.
613
614 4.804 Wed Nov 5 16:36:00 CET 2008
615 - Coro::Debug::new_unix_server would not create a non-blocking listening
616 socket, sometimes causing freezes.
617 - (libcoro) fix misaligned stack points for setjmp and assembly
618 methods, which can cause crashes on x86/x86_64 with a sufficiently
619 aggressive compiler.
620 - new function: Coro::Debug::new_tcp_server.
621 - move ->throw into the Coro class because it only works on coro objects.
622
623 4.803 Mon Nov 3 17:16:12 CET 2008
624 - (libcoro) use a global asm statement to become independent of gcc
625 otpimisations for CORO_ASM (thanks to pippijn for the idea).
626 - try to workaround yet another broken bsd, this time dragonfly.
627
628 4.802 Thu Oct 30 10:56:12 CET 2008
629 - support -fno-omit-frame-pointer on x86 with the assembly method.
630 - tune 01_unblock.t tests a bit.
631
632 4.801 Wed Oct 22 18:33:37 CEST 2008
633 - improve readline speed for very long "lines".
634 - backport to 5.8.8.
635
636 4.8 Thu Oct 2 13:34:40 CEST 2008
637 - new function Coro::AIO::aio_wait.
638 - Coro.:AIO and Coro::BDB now "use Coro::AnyEvent".
639 - greatly speed up and reduce memory usage of Coro::AIO requests.
640 - implement some other µ-optimisations.
641
642 4.749 Mon Sep 29 14:40:12 CEST 2008
643 - port to slow and broken pseudo-threaded perls. (courtesy pippijn).
644
645 4.748 Sat Sep 27 14:03:19 CEST 2008
646 - implement, but do not document, PerlIO::cede(granularity).
647 - Coro::Storable forgot to wrap Storable::pstore.
648 - work around the multitude of leaks and memory corruption
649 bugs in PerlIO::via by using our own C-level perliol. As a side
650 effect, Coro::Storable is now much, much, much faster.
651
652 4.747 Tue Sep 23 01:59:41 CEST 2008
653 - fix a per-cv memleak (one empty array was leaked per
654 code reference).
655 - avoid a crash in coro->call|rss when the coroutine was already
656 destroyed (most noticably when using Coro::Debug::ps :)
657 - also protect *Storable::FILE.
658 - push up default storable granularity to 20ms.
659
660 4.746 Sun Sep 21 03:22:20 CEST 2008
661 - be more insistent on locking Storable against reentrancy
662 in Coro::Storable.
663 - move swap_def?v and throw to Coro::State, as documented.
664
665 4.745 Thu Jul 24 00:14:38 CEST 2008
666 - remove debugging code related to MgPV_nolen_const, also try to
667 make it compile with perl 5.8.6 (yes, apple apparently loves
668 outdated software). Reported by John S.
669
670 4.744 Tue Jul 8 22:06:35 CEST 2008
671 - correctly provide default DIE/WARN handlers as documented.
672 - also overwrite PL_vtbl_sigelem.svt_clear, even though current
673 implementations inside perl work fine for us.
674
675 4.743 Mon Jun 16 00:21:57 CEST 2008
676 - when using Coro::EV without running EV::loop it could
677 result in busy-waiting for events, this has been fixed.
678 - reduce codesize and improve performance by using EV_DEFAULT_UC.
679
680 4.742 Sat May 31 14:10:21 CEST 2008
681 - implement a workaround for (some) perl <5.8.8 versions.
682 - require EV 3.3+.
683
684 4.741 Fri May 30 23:33:09 CEST 2008
685 - tell netbsd how utterly broken their imitation of an OS is
686 and refuse to build by default if pthreads are in use.
687 - switch to "s" method on all bsd's by default, as their ucontext
688 stuff seems just too broken.
689 - fix a bug in Coro::Select.
690
691 4.74 Thu May 29 20:05:31 CEST 2008
692 - do not test Coro::LWP for lack of dependencies.
693 - do not test Coro::Select for lack of working perls.
694
695 4.73 Thu May 29 2008
696 - fix a bug in Coro::EV which would cause it to block despite
697 there being runnable coroutines.
698 - sprinkle "no warnings" freely over everything, also suppress
699 warnings for some other modules.
700 - fix typo in WSAEWOULDBLOCK.
701
702 4.72 Sun May 25 05:14:36 CEST 2008
703 - tweak META.yaml a bit, unfortunately, there is no documented way
704 to have optional dependencies with CPAN. doh :(
705 - avoid running some tests on windows because they would fail due to
706 perl bug (broken fork, broken pipes...).
707 - work around perl on windows bugs where perl returns undocumented
708 error codes for sysread, syswrite etc. by taking advantage
709 of AnyEvent's workaround for that problem.
710 - use AnyEvent::Util::fh_nonblocking in Coro::Handle to work around
711 a common perl implementation bug on windows.
712 - use unix domain sockets in testsuite to work around
713 common perl implementation bugs on widows (they are emulated by
714 tcp sockets on windows. ugh.)
715
716 4.71 Sat May 24 20:01:27 CEST 2008
717 - fix a bug in Coro::AnyEvent ("Usage: Coro::AnyEvent::_schedule()").
718 - take advantage of async name resolution of AnyEvent::Util.
719 - work around brutal inet_aton override in Coro::LWP.
720 - take advantage of the readyhook in Coro::EV, for smoother
721 scheduling.
722
723 4.7 Sun May 11 00:32:19 CEST 2008
724 - completely reworked the Coro manpage.
725 - added Coro::AnyEvent, generic event loop integration.
726 - implement cancel, ready and kill commands in Coro::Debug.
727 - document find_coro in Coro::Debug.
728 - incompatible change: rename has_stack to has_cctx.
729 - Coro::AIO and Coro::BDB no longer force event model detection,
730 use AnyEvent::AIO and AnyEvent::BDB.
731
732 4.6 Sat Apr 26 10:05:14 CEST 2008
733 - INCOMPATIBLE CHANGE: sub/code attributes are no longer supported
734 by the Coro module. It was a mistake to have it in the first place.
735 - (experimental) support for activestate perl 5.10 (method "w").
736 - (experimental) support for strawberry perl (method "a").
737 - coro_sigelem_set did not return a value although it had to,
738 actual impact unknown.
739
740 4.51 Mon Apr 14 13:28:27 CEST 2008
741 - make it compile again on 5.8.
742
743 4.50 Thu Apr 10 09:43:17 CEST 2008
744 - I did it twice! (see 4.49).
745
746 4.49 Sun Apr 6 21:23:31 CEST 2008
747 - grr, instead of compiling the recent changes on 5.10 only they
748 were compiled on 5.8 only.
749
750 4.48 Sun Apr 6 20:36:46 CEST 2008
751 - allow coroutine switches during eval's under 5.10.x, as it seems
752 the parser is a per-interpreter option now, so this is safe
753 (this might fix the odd crashing bug).
754 - drop support for 5.9.x versions: they are dead, jim.
755
756 4.47 Sun Apr 6 00:37:52 CEST 2008
757 - force cctx allocation on API calls: we know we need to force one
758 and gcc actually manages to confuse our heuristic nowadays,
759 leading to crashes and worse.
760 - document force_cctx.
761
762 4.46 Fri Apr 4 22:05:43 CEST 2008
763 - upgrade libcoro, resulting in pthread-backend (which was only created
764 to fulfill the rules of the programming languages shootout).
765
766 4.45 Thu Mar 13 11:55:36 CET 2008
767 - fix a file leaking bug in eg/mhyttpd that would allow
768 downloading of any file (reported by oesi).
769 - fix deadlock bug in Coro::Channel (reported by Richard Hundt)
770 (also add testcase).
771 - support Broadcast option in Coro::Socket (patch by Richard Hundt,
772 apparently having loads of fun with that).
773 - implement $state->swap_defsv, ->swap_defav and document ->throw.
774
775 4.4 Wed Feb 13 16:44:29 CET 2008
776 - only assume .cfi pseudo isns exist on GNU/Linux.
777 - add get_timed method to Coro::Channels.
778 - fixed Coro::Debug manpage.
779 - perl 5.11 compatbility improvement based on patch by
780 Andreas König.
781
782 4.37 Sun Jan 20 11:25:23 CET 2008
783 - soften the check for an active parser for perl 5.10 (avoiding
784 "Coro::State::transfer called while parsing" in many cases).
785
786 4.36 Sun Jan 13 10:53:56 CET 2008
787 - reset diehook when terminating from an async_pool as to not
788 trigger any __DIE__ handlers.
789
790 4.35 Sun Jan 13 04:14:13 CET 2008
791 - "bt" debug command now displays any exceptions
792 from longmess and also skips the innermost
793 stackframes, giving more useufl output.
794 - allow backtraces in coroutines blocked in Coro::EV,
795 at a <1% speed hit.
796 - handle localising of $SIG{__WARN__} and __DIE__
797 properly (with a proper amount of dirty hacking).
798
799 4.34 Sat Dec 22 17:49:53 CET 2007
800 - upgrade to EV version 2.0 API.
801
802 4.33 Mon Dec 17 08:36:12 CET 2007
803 - make Coro::AIO etc. loadable in the absence of EV.
804
805 4.32 Mon Dec 17 07:46:02 CET 2007
806 - majorly improved Coro::Handle's performance with Coro::EV.
807 - implemented the until now mythical Coro::BDB module.
808 - specialcase EV in Coro::AIO and Coro::BDB for extra speed.
809
810 4.31 Wed Dec 5 12:32:39 CET 2007
811 - remove warn statement form Coro::Util.
812
813 4.3 Tue Dec 4 20:33:14 CET 2007
814 - calls to the idle function could cause stack corruption
815 when the stack changed.
816 - do no longer rely on the presence of EV::DNS if EV is used
817 (because it is gone), but instead take avdantage of EV::ADNS
818 if available.
819 - add ($) prototypes to all functions in Coro::Storable.
820 - use a conventional (safer) idle callback in Coro::EV.
821 - do accept quickly in Coro::Debug to avoid endless loops.
822
823 4.22 Fri Nov 30 16:04:04 CET 2007
824 - really use optimised versions for Event and EV in Coro::Util
825 and Coro::Handle.
826
827 4.21 Sun Nov 25 10:48:59 CET 2007
828 - fix a spurious memory read.
829 - Coro::EV no longer keeps the eventloop "alive".
830
831 4.2 Fri Nov 9 20:47:05 CET 2007
832 - enable/disable tracing from a new coroutine, not a pooled one.
833 - fix a memleak in Coro::Event.
834 - removed killall call from fork_eval.
835 - made sure store_fd is already loaded so that fork_eval
836 does not have to parse autoload in each subprocess.
837 - only use assembly method if -O switch is in $Config{optimize}.
838 - add (optional) Coro::EV module, so far the best event loop module
839 directly supported by Coro.
840 - if the event model is EV, use EV::DNS to resolve
841 stuff in Coro::Util.
842 - don't get confused by multiple event notifications in Coro::Handle.
843 - initial support for EV (libevent interface).
844 - require Event and EV using configure_requires, to force their existance.
845
846 4.13 Wed Oct 24 07:26:45 CEST 2007
847 - add Coro::Storable::blocking_thaw.
848 - use a vastly more complicated technique to localise
849 $SIG{__WARN/DIE__} that also works on perls <= 5.8.8.
850 - use a coroutine for the idle callback Coro::Event,
851 instead of running Event in the current coroutine context.
852 This also catches recursive invocations.
853 - actually report fork errors in gethostbyname and inet_aton.
854
855 4.11 Thu Oct 11 02:40:24 CEST 2007
856 - port to threaded perls.
857
858 4.1 Thu Oct 11 02:38:16 CEST 2007
859 - incompatible change: $SIG{__DIE__} and $SIG{__WARN__} will now
860 be local to each coro (see Coro::State).
861 - incompatible change: for very deep reasons, cede and cede_notself
862 cannot return anything, so nothing will be returned.
863 - possibly bring back 5.10 compatibility (untested).
864 - work around stupid (and wrong) warning on 5.10 :(.
865 - overlay the saved state over the context stack. This saves
866 a few hundred bytes per coroutine on average and also
867 speeds up context switching a bit.
868 - further tune default stack sizes.
869 - (more) correctly calculate stack usage in coro_rss.
870 - Coro::Storable::blocking_* did not properly lock
871 resulting in races between coroutines.
872 - added Coro::Storable::guard.
873 - stopping to trace a coroutine could destroy the cctx of
874 an unrelated coroutine.
875 - explain the relationship between Perl and C coroutines in
876 more detail in Coro::State.
877 - Coro::Util::inet_aton did not short-circuit dotted quad forms,
878 causing a fork per resolve. This also affected Coro::Socket.
879 - switch to a separate stack in $coro->call/eval to avoid
880 invalidating pointers.
881
882 4.03 Sat Oct 6 21:24:00 CEST 2007
883 - added Coro::throw method.
884 - minor code cleanups.
885
886 4.02 Sat Oct 6 02:36:47 CEST 2007
887 - fix a very minor per-coroutine memleak (a single codereference).
888 - fixed a bug where the currently in-use c context would be freed
889 prematurely (can happen only when programs change the stacksize
890 or use tracing).
891 - tracing can no longer keep a coro alive after it terminated.
892 - do static branch prediction in the common path for gcc. gives
893 about 2-5% speed improvement here.
894
895 4.01 Fri Oct 5 22:10:49 CEST 2007
896 - instead of recreating *a* standard output handle we simply
897 use STDOUT, which is faster and hopefully more robust.
898
899 4.0 Fri Oct 5 12:56:00 CEST 2007
900 - incompatibly changed Coro::Storable::freeze.
901 - major new feature: added Coro::Debug, for interactive coroutine
902 debugging, tracing and much more.
903 - major bug fix: unbelievable, but true: $_, $/ and many other
904 "saved" variables actually weren't being saved. This has been fixed,
905 of course, while increasing performance while losing all the save
906 flags.
907 - save flags are gone, and all the api functions dealing with them.
908 - added Coro::Semaphore::adjust.
909 - added Coro::Util::fork_eval.
910 - added Coro::Storable::{nfreeze,blocking_{freeze,nfreeze}}.
911 - added Coro::killall.
912 - reduce initial stack sizes to allow for "micro-coroutines".
913 - better async_pool resource management, moved parts of async_pool
914 handling to XS (major speed improvement).
915 - actually croak before modifying important data structures.
916 - refuse to transfer while compiling.
917 - possibly support eval EXPR better now.
918 - enable assembly per default on linux+bsd x86+amd64.
919 - all internal members were renamed _something for easier subclassing.
920 - many minor tweaks.
921
922 3.63 Wed May 16 14:10:06 CEST 2007
923 - implement handcoded assembly for x86/amd64 SVR ABI.
924
925 3.62 Fri Apr 27 21:36:06 CEST 2007
926 - upgrade libcoro (which might set unwind info correctly).
927 - change default on linux to setjmp/longjmp.
928
929 3.61 Thu Apr 19 12:36:18 CEST 2007
930 - Coro::Storable caused an endless loop when thawing invalid
931 pst's sometimes.
932 - use a Semaphore in Coro::Storable, as Storable doesn't
933 seem to be reentrant (although it is documented to
934 be threadsafe...).
935 - fix Coro::Signal to bring back the original unreliable
936 but stateful semantics.
937 - fixed a lot of typos in Coro.pm (patch submitted by David
938 Steinbrunner, which applied flawlessly).
939
940 3.6 Sat Apr 14 17:13:31 CEST 2007
941 - added some bugfixes to get eg/myhttpd working again.
942 - added Coro::Storable for often-cede'ing freeze/thaw.
943 - try to do a clean exit when a coroutine calls exit
944 (EXPERIMENTAL).
945 - got rid of indirect call through _coro_init.
946 - updated the partly antique examples in eg/ to
947 work again and be a bit less magic, too.
948 - fixed Coro::Signal semantics to work as documented again.
949
950 3.55 Sun Mar 25 01:20:47 CET 2007
951 - add SAVE_DEFFH to save the default filehandle and enable
952 it by default.
953 - finally move socket-operations from Coro::Socket to Coro::Handle
954 to be able to unblock foreign sockets.
955 - add Coro::State::save_also and guarded_save.
956 - add count accessor to Coro::Semaphore.
957 - add Coro::State::cctx_stacksize.
958 - just for the fun of it, do not rely on implicit context,
959 which can dramatically improve performance, but people
960 using windows-process-emulation perls do not care much about
961 performance.
962
963 3.51 Sun Mar 4 14:18:04 CET 2007
964 - fixed a problem when you weakened references to Coro::State's
965 (patch by Yuval Kogman).
966
967 3.501 Wed Feb 28 12:44:07 CET 2007
968 - rename some global symbols as macosx from hell redefines
969 them without asking.
970
971 3.5 Tue Feb 13 20:22:53 CET 2007
972 - do AnyEvent model detection earlier, avoiding problems
973 caused by first using AnyEvent and later Coro::Event.
974 - implement and document Coro::Event event objects.
975 - fix a potential problem in Coro::Event causing crashes.
976 - initialise PL_comppad when creating a new coroutine,
977 avoids crashes on early coro destruction.
978
979 3.41 Mon Jan 22 19:19:49 CET 2007
980 - readline on Coro::Handle did not support undefined $/,
981 nor did it deliver partial lines on EOF or error.
982 - implement malloc fallback for stack allocation because
983 stupid broken idiotic OSX has a stupid broken
984 idiotic fits-the-whole-os mmap "implementation" and
985 my dick feels longer if Coro is portable even to
986 obsolete platforms.
987
988 3.4 Fri Jan 19 21:52:54 CET 2007
989 - remove t/09_timer.t, as it isn't really testing much
990 but was rather flaky in practise.
991 - async_pool coro would keep arguments and callback alive until
992 it was reused.
993 - cancellation of a coroutine could cause spurious idle calls
994 in cede_notself.
995
996 3.3 Sat Jan 6 03:45:00 CET 2007
997 - implement $coro->on_destroy.
998 - Coro::Event blocking semantics have been changed,
999 documented and - hopefully - improved.
1000 - fix nice adding, not subtracting, from priority.
1001 - fix ->prio and api_is_ready (patch by Mark Hinds).
1002 - fixed an assert ("... == prev__cctx->idle_te")
1003 that could errronously trigger.
1004 - fix various large and small memleaks.
1005 - use a (hopefully) more stable cancel implementation
1006 that immediately frees the coroutine data.
1007 - cede/cede_notself return a status now.
1008 - added Coro::guard function.
1009 - added a global coroutine pool for jobs (on my machine,
1010 I can create and execute 48k simple coros/s with async,
1011 and 128k coros with async_pool).
1012 - Coro::AIO now uses the coroutine priority as io priority.
1013
1014 3.2 Fri Dec 22 05:07:09 CET 2006
1015 - improve portability to slightly older perls.
1016 - use cleaner coroutine destruction.
1017 - simplify configuration for users.
1018 - optionally (unrecommended) prefer perl functions over
1019 their coro replacements.
1020
1021 3.11 Tue Dec 5 13:11:24 CET 2006
1022 - fixed some bogus assert's, but as perl.h disables assert even
1023 without NDEBUG (thank you very much), not too many people should
1024 notice (that did include myself). Andreas König noticed, though :)
1025 - do not save/restore PL_sortcxix on >= 5.9.x, it doesn't seem to have
1026 it. Also noticed by Andreas König :)
1027 - save/restore tainted status.
1028 - verified to pass the testsuite on my 5.9.5.
1029
1030 3.1 Mon Dec 4 23:03:40 CET 2006
1031 - INCOMPATIBLE CHANGE: $/ is now per-coroutine (but slow).
1032 - incompatible change: transfer flags are now per-state.
1033 - give a better error message on deadlock.
1034 - document Coro::nready.
1035 - enhanced testsuite.
1036
1037 3.01 Sun Dec 3 23:47:42 CET 2006
1038 - forgot to include Coro::Timer.
1039
1040 3.0 Sun Dec 3 22:57:25 CET 2006
1041 - the "FINALLY COMPLETELY STABLE" release (coming soon:
1042 the "FAMOUS LAST WORDS" release).
1043 - implement a new stack sharing algorithm, which uses a stack
1044 pool (size currently hardcoded).
1045 - make stack sharing mandatory (it no longer uses a heuristic).
1046 - eval/die no longer cause weird problems under heavy use.
1047 - Coro::Event could cause livelocks if it was used but
1048 no Coro::Event watchers were used.
1049 - Coro::Event now uses asynccheck as prepare does not
1050 check for changed watchers.
1051 - Coro::Event allows multiple waiting coros and will wake up one
1052 per event.
1053 - Coro::Event should be cleaner and more efficient now.
1054 - new utility function Coro::unblock_sub.
1055 - document the sad fact that Event is no longer reentrant.
1056 - putting a coroutine into the ready queue twice could under
1057 some circumstances lead to stack corruption.
1058 - minor incompatibility: subclassing Coro::State is supported
1059 directly now without going through a _coro_state member.
1060 - state/coro switching is much faster now.
1061 - very minor optimisations and code/documentation cleanup.
1062 - avoid problems due to compiler inlining.
1063 - removed timers from Coro::Timer -> use AnyEvent instead.
1064 - replaced Coro::idle coroutine by (cleaner) idle handler.
1065 - updated to newest libcoro.
1066 - implement enhanced support for valgrind.
1067 - implement is_ready and return value for ready.
1068 - removed Coro::Cont, it was a misnomer (it's generators),
1069 and it was somewhat annoying to get it right. Will come back
1070 if somebody asks for it :)
1071 - many ->wait methods and Coro::Event could return
1072 spuriously without the event having happened.
1073
1074 2.5 Tue Nov 7 12:22:33 CET 2006
1075 - made Coro::Util, Coro::Select, Coro::Handle and Coro::Socket
1076 use AnyEvent, moved them to Coro/.
1077 - added Coro::LWP which contains all the uglyness required to
1078 make LWP non-blocking.
1079 - should work with perl 5.9.x now (Andreas König made me do it).
1080 - fixed another bug in Coro::Select when the timeout was undef.
1081 - reuse PL_start_env for all coros, saves some memory per coroutine.
1082 - manage PL_top_env differently, hopefully to avoid panic: top_env.
1083 - timeout argument was not properly used in Coro::Socket.
1084 - allow limited forms of subclassing in Coro::Handle/Coro::Socket.
1085 - emulate undocumented(!) functionality of IO::Socket required
1086 by LWP(!!).
1087 - updated eg/lwp to work with newer lwp's.
1088 - remove "FATAL: uncaught exception" prefix. Coroutines that die
1089 kill the whole process, just as exceptions in the main "coroutine"
1090 did already.
1091
1092 2.1 Wed Nov 1 23:01:13 CET 2006
1093 - fix a long-standing bug in Coro::Select where select with
1094 zero timeout would instead change the current default filehandle.
1095 - use a simpler and hopefully more robust way to clone padlists
1096 (uses less memory and a perl function instead of our own).
1097 - coro can now create a stack guard area on many architectures.
1098 - Coro::AIO properly reexports additional functions from IO::AIO.
1099 - updated libcoro with a workaround for OS X,
1100 pach and testing by Michael Schwern.
1101
1102 2.0 Tue Oct 24 05:47:17 CEST 2006
1103 - support additional aio requests in Coro::AIO.
1104
1105 1.9 (never properly released due to a glitch)
1106
1107 1.8 Thu Feb 2 00:59:06 CET 2006
1108 - applied suggested patch by SAMV to avoid problems during stupid
1109 mark & sweep gc run.
1110 - applied patch by Scott Walters for 5.9.3 compatibility.
1111
1112 1.7 Tue Dec 27 01:41:58 CET 2005
1113 - added Coro::AIO, a thin wrapper around IO::AIO.
1114 - improved Makefile.PL explanations.
1115
1116 1.6 Mon Dec 12 21:30:05 CET 2005
1117 - additionally save PM_curpm might fix as-of-yet
1118 unseen problems with regex matches being attributed to the
1119 wrong package.
1120 - add t/10_bugs.t, which currently checks against imho broken
1121 perls that use 0.26MB of stack space per Perl_magic_get
1122 invocation (newer linuxes) as opposed to the 0.0002MB perl
1123 normally uses.
1124 - make stacksize configurable for the ultimate debian experience.
1125
1126 1.51 Mon Dec 12 18:48:36 CET 2005
1127 - remove debugging warn() accidentally left in Coro::Select.
1128
1129 1.5 Tue Nov 29 13:32:44 CET 2005
1130 - use Coro::Event inside Coro::Select to avoid spurious deadlocks.
1131 - fix Coro::Select 'select'.
1132 - strict'ify some modules.
1133 - libcoro errornously restored the SIGUSR2 handler to SIGUSR1.
1134 - use XSLoader in selected modules.
1135 - remove some 5.6 compatibility cruft.
1136
1137 1.4 Tue Sep 6 00:11:05 CEST 2005
1138 - libcoro did not take into account the trampoline on amd64,
1139 when 'l'inux method was used.
1140
1141 1.31 Tue Aug 30 23:31:33 CEST 2005
1142 - some portability fixes/workarounds.
1143
1144 1.3 Sat Aug 20 03:08:56 CEST 2005
1145 - no code changes. module seems to work fine.
1146 - improve Coro::State docs, remove reference to nonexisting
1147 Coro::State::flush method.
1148 - no longer autodetect windows, present it as an option instead.
1149
1150 1.2 Mon May 16 02:00:55 CEST 2005
1151 - included libcoro.c earlier in State.xs, might improve portability.
1152 - use a faster and possibly more stable Coro::Cont implementation.
1153 - accept x86_64 in addition to amd64 for optimized linux-amd64 support.
1154 - fix bugs Coro::SemaphoreSet that could cause locks never to be freed.
1155 - fix bugs in CoroAPI.h and document it in Coro::MakeMaker.
1156
1157 1.11 Thu Mar 3 18:00:52 CET 2005
1158 - change of contact address.
1159
1160 1.1 Tue Feb 22 20:51:16 CET 2005
1161 - support [l]inux method on amd64.
1162 - allow some context switches while compiling/eval'ing.
1163
1164 1.0 Tue Aug 10 03:47:27 CEST 2004
1165 - Event 0.89 fixes the bug again.
1166 - fix the remaining(??) memleak. I didn't free the cache-AV for
1167 padlists, so one AV per created anon-sub leaked.
1168
1169 0.97 Fri May 14 15:23:32 CEST 2004
1170 - remove stupid gcc extension that was an experiment only anyway.
1171
1172 0.96 Thu May 13 18:09:29 CEST 2004
1173 - tell people if they have yet another broken Event variant.
1174 - made Coro to work _better_ with threads (it seems that, as long
1175 as you use Coros in one thread only everything will work fine.
1176 Looking at the perl source the problem might be missing locking
1177 between Coro and some perl internal routines. That's not easily
1178 fixable).
1179 - (hopefully) got rid of the static xnext inside transfer.
1180
1181 0.95 Sun Feb 15 03:20:28 CET 2004
1182 - removed Coro::State::flush, implemented a new and much
1183 faster caching mechanism by attaching magic to the CV.
1184 - a memleak with (real) closures remains.
1185 - dramatically reduced size of stacks to 65536 (linux/x86), after
1186 observing that even myhttpd never had more than a kilobyte
1187 of stack in use (stack is still only physically allocated in
1188 pagesize increments for systems suporting mmap).
1189 - fix for a "cygwin-64" architecture added, proposed by Stefan Traby.
1190 - fix for threaded perl.
1191
1192 0.9 Sun Nov 23 04:32:39 CET 2003
1193 - fix a few very hard to track down but (of course) stupid
1194 memory leaks.
1195 - stringify version number to avoid locale problems :(.
1196 - perl 5.9.0 does not have SvPADBUSY, reported by Scott Walters,
1197 hopefully ignoring it "just works".
1198
1199 0.8 Wed Nov 5 19:38:40 CET 2003
1200 - port to perl5.005_03, but only 5.8.x+ is supported!
1201 - honor LocalAddr even without LocalPort in Coro::Socket.
1202
1203 0.7 Tue May 27 03:12:38 CEST 2003
1204 - the version jump indicates some level of testing, not gobs
1205 of new features.
1206 - uh... I found the prompt function in ExtUtils::MakeMaker.
1207 Highly correct stuff, that is...
1208 - fixed(?) a bug with die's in coroutines causing "panic: top_env".
1209 the fix is not well understood by the author, so beware :(.
1210
1211 0.652 Thu May 8 02:54:46 CEST 2003
1212 - Applied patch by Slaven Rezic to set default to "s" on FreeBSD,
1213 cause version 4 doesn't have ucontext.
1214 - Benjamin Reed reported that setjmp works fine on darwin,
1215 so preselect it.
1216
1217 0.651 Sat Mar 29 15:00:23 CET 2003
1218 - fix a bug in Coro::Handle where some bytes could get lost
1219 on reading, reported by jason@nichego.net.
1220
1221 0.65 Sun Mar 23 00:08:26 CET 2003
1222 - added a README.linux-glibc.
1223 - new module Coro::Select.
1224 - also save/restore PL_comppad, fixes segfaults with 5.8.1.
1225 (I never claimed that I know what I am doing ;).
1226 - set default to setjmp/longjmp on non-x86-linux.
1227
1228 0.6 Thu Nov 21 11:09:06 CET 2002
1229 - made lazy context switching the default.
1230 - fixed the bug where SemaphoreSet::guard
1231 was the same as timed_guard and timed_guard was missing.
1232 - fix a memleak in Coro::Timer.
1233 - reclassified context sharing as not experimental.
1234
1235 0.534 Sun Apr 14 03:05:12 CEST 2002
1236 - fixed backspace => tab in header parsing (myhttpd).
1237 - added eg/readline, for elmex.
1238 - Coro::Event now calls ->start on first call instead of
1239 ->again, which created wrong timeouts for the first ->next.
1240 - fixed a bug where Coro::Socket returned a closed socket instead
1241 of nothing on ECONNREFUSED and similar error conditions.
1242
1243 0.533 Mon Feb 18 18:49:40 CET 2002
1244 - adapt to finally working Event-0.85.
1245 - fixed buggy event initialization (trapped under windows, but how
1246 could it possibly have worked anywhere?)
1247
1248 0.532 Wed Jan 16 02:45:32 CET 2002
1249 - removed the reference to pp_entersub, might work on
1250 (some) win32 perls now (testsuite works under cygwin,
1251 without Event, which freezes).
1252
1253 0.531 Mon Dec 10 22:18:44 CET 2001
1254 - Fixed a bug in SemaphoreSet::timed_down.
1255
1256 0.53 Tue Nov 27 21:11:13 CET 2001
1257 - some tests for Event.
1258 - slightly different internal architecture to get refcnt into the object
1259 and not the reference. might fix a few bugs, certainly did introduce
1260 new ones.
1261 - Coro::Timer (independent of Event).
1262 - new timed_wait functions for Coro::Signal, Semaphore, SemaphoreSet.
1263
1264 0.52 Tue Nov 6 21:36:18 CET 2001
1265 - ported to cygwin (trivial).
1266 patch by Gerrit P. Haase <gp@familiehaase.de>.
1267 - small setjmp code fix by Sullivan.DanielJ@epamail.epa.gov.
1268
1269 0.51 Thu Nov 1 20:39:01 CET 2001
1270 - terminate/cancel now work properly (otherwise termination
1271 could cause a "next coroutine is not and contains not..."
1272 error.
1273 - added Coro::Socket::shutdown.
1274 - Coro::Event::loop no is the same as Event::loop.
1275 - implemented terminate with args + join.
1276
1277 0.5 Fri Sep 28 16:15:35 CEST 2001
1278 - fixed "print" on a Coro::Handle. The print method worked.
1279 - small tweaks (seem to reduce memory consumption a lot)
1280 in various modules.
1281 - splendid use of "no warnings" scattered throughout.
1282 - added Coro::Handle::rbuf, fixed Coro::Handle::fh.
1283
1284 0.49 Sun Sep 16 02:42:45 CEST 2001
1285 - changed some method calls to function calls for speed
1286 inside Coro::Handle.
1287 - make Coro::Handle use an array instead of a hash for
1288 speed reasons.
1289 - IRIX mystery solved: it's SGI's NT, after all: "standard, huh?".
1290 sjlj and ucontext should now work.
1291 - IRIX-specific port for libcoro.
1292 - swapped order of accept results to match IO::Socket.
1293 - changed getsock/peername to sock/peername to match IO::Socket.
1294 - fixed a bug that caused segfault when returning to main task
1295 under some circumstances.
1296 - other bugfixes.
1297
1298 0.45 Sun Sep 2 02:54:01 CEST 2001
1299 - new method Coro::Handle::timeout.
1300 - corrected speling of set...name to get...name in Coro::Socket.
1301 - Coro::Socket::accept now returns a Coro::Socket, not a Coro::Handle.
1302 - Coro::Handle now supports fileno and FILENO.
1303 - added eg/myhttpd, a web server.
1304 - Coro::Socket now accepts numeric ports as well as "port(number)"
1305 syntax.
1306 - moved some scheduling primitives into xs code. more to come.
1307 - new simple priority system.
1308 - implemented Coro::Event using XS, almost four times faster!
1309 - small memory corruption problem fixed (boy that was difficult).
1310
1311 0.13 Wed Aug 8 16:53:07 CEST 2001
1312 - new method Coro::Handle::readline.
1313 - added eg/lwp and Coro::Handle::autoflush, to make LWP work.
1314 (see eg/lwp on how to make LWP non-blocking).
1315 - renamed Coro::Socket::new_inet to new.
1316 - added Coro::Util, some utility functions.
1317 - die/eval now works better (still get annoying (but true ;)
1318 "Callback called exit" messages).
1319 - Coro::Handle now supports timeout, and Coro::Socket Timeout.
1320 - much better Coro::Socket support.
1321
1322 0.12 Fri Jul 27 04:19:01 CEST 2001
1323 - do not use mmap if neither MAP_ANON nor MAP_ANONYMOUS is defined.
1324 - very experimental stack sharing algorithm. not 100% safe but
1325 should work well in practise ;)
1326 - added Coro::RWLock.
1327 - Coro::Specific now works.
1328
1329 0.11 Tue Jul 24 22:49:21 CEST 2001
1330 - added specialized hack for newer and older linux versions (fast).
1331 - renamed Coro::Event::IO to Coro::Handle.
1332 - new module Coro::Socket.
1333
1334 0.1 Tue Jul 24 01:47:53 CEST 2001
1335 - release candidate 3. A rather new internal structure :(
1336 - the great renaming: Damian Conway gave me a suitable replacement
1337 for yield.
1338 - added Coro/libcoro, a portable coroutine implementation for C
1339 (not even perl-dependent!!!), which can be used by Coro::State.
1340 - renamed SAVE_DEFSV (on xs-level only) to avoid symbol clash in
1341 perl-5.6.
1342 - new function Coro::Event::idle.
1343 - the idle process is now overriden
1344 by default in Coro::Event.
1345 - Coro::Channel now enforces the size.
1346 - canceling events now works in all cases.
1347 - Coro state now include $_ and $@.
1348 - yet another bug workaround that I do not really understand :(
1349 - new module Coro::Event::IO (very undocumented) to do
1350 non-blocking i/o.
1351 - performance tuning.
1352
1353 0.08 Thu Jul 19 06:13:25 CEST 2001
1354 - release candidate 2 (new functionality).
1355 - add Coro::State::flush function.
1356 - transfer now supports a flags value (mostly for speed ATM).
1357 - might compile and work in the presence of threads now.
1358 - continuations have a saner syntax.
1359 - no more memleaks.
1360
1361 0.07 Tue Jul 17 17:40:18 CEST 2001
1362 - release candidate 1 ;)
1363 - slightly nicer code.
1364 - fixed a scheduling bug in Coro::Event.
1365
1366 0.06 Tue Jul 17 04:23:24 CEST 2001
1367 - ok, I found the showstopper - the same sub must not be
1368 re-entered in two different coroutines, otherwise => crash. I
1369 see no easy solution to this problem, except by walking the call
1370 chaing and saving/restoring the cv's, which is what I do now.
1371 - memory leaks still latent, especially at thread termination.
1372 - Coro::Event now works (haha).
1373
1374 0.05 Sun Jul 15 17:32:20 CEST 2001
1375 - fixed a few issues in Coro::Event.
1376 - I forgot to include Coro::Event in 0.04 :(:(:(
1377
1378 0.04 Sun Jul 15 05:24:59 CEST 2001
1379 - @_ is now properly localized.
1380 - Coro::State is now easier subclassable.
1381 - Coro::Cont now coroutine-aware.
1382 - Coro::Specific is a low-overhead module to create
1383 coroutine-specific vars.
1384 - Coro::Event provides a simple interface to Event.
1385
1386 0.03 Fri Jul 13 14:51:52 CEST 2001
1387 - transfer() now implemented in XS (beware).
1388 - new module Coro::Cont for really faked continuations.
1389 - big internal architecture changes: Coro::State is now
1390 really low-level and can thus be used to implement other
1391 interesting things, While "Coro::" implements a process-like
1392 model. Still crude and subject to change.
1393 - $_ and $@ are no longer being localized.
1394
1395 0.02 Tue Jul 10 01:38:17 CEST 2001
1396 - implemented "async" attribute.
1397 - $_ and $@ are now localized.
1398 - added Coro::Channel.
1399 - more testcases, still no docs.
1400
1401 0.01 Tue Jul 3 02:18:41 CEST 2001
1402 - original version; copied from Convert::Scalar.
1403