ViewVC Help
View File | Revision Log | Show Annotations | Download File
/cvs/Coro/Changes
Revision: 1.531
Committed: Sat Dec 12 01:30:26 2009 UTC (14 years, 5 months ago) by root
Branch: MAIN
Changes since 1.530: +5 -2 lines
Log Message:
*** empty log message ***

File Contents

# User Rev Content
1 root 1.1 Revision history for Perl extension Coro.
2 root 1.113
3 root 1.382 TODO: should explore PerlIO::coroaio (perl leaks like hell).
4 root 1.328 TODO: maybe implement a default message channel, very much like Erlang's
5     actor model (which is cool in a lot of important aspects (failures!),
6 root 1.247 but very lacking in others (higher level ipc)).
7 root 1.487 TODO: unready_all
8 root 1.490 TODO: myhttpd header parsing
9 root 1.494 TODO: channel->maxsize(newsize)?
10 root 1.173
11 root 1.530 5.21
12     - automatically load Coro::AnyEvent when AnyEvent and Coro are used
13     together.
14     - add some examples on how to combine other event loops with Coro in
15 root 1.531 Coro::AnyEvent, and how to run it (and not to block). Seems to be
16     the most common source of confusion.
17     - try to catch people naively blocking in an event callback.
18     - work around the perl filehandle bug issue in conjunction with
19     older common::sense (as indirectly pointed out by ZSystem).
20 root 1.529 - clarify the "not from signal handlers" section.
21    
22 root 1.528 5.2 Sun Oct 4 14:54:24 CEST 2009
23 root 1.524 - Coro::Storable destroyed the prototypes of the functions it wrapped.
24 root 1.527 - export rouse_cb and rouse_wait by default now.
25 root 1.524 - fix various prototype mismatches in Coro::AnyEvent and Coro::Handle.
26 root 1.527 - new method $state->swap_sv.
27 root 1.522 - added section on "windows process emulation" to the manpage,
28     after a not-so-fruitful (nor-friendly) "discussion" with chip
29     salzenberg (discussion implies arguments, but his only arguments
30     were ad-hominems, one wonders why he started it in the first
31 root 1.523 place). I hope this explains it well enough for him to understand,
32     and maybe well enough for others to understand.
33 root 1.524 - use common::sense everywhere now.
34 root 1.525 - idle callbacks are no longer supported, use idle coros instead.
35     - print a thread listing when a deadlock is detected.
36 root 1.522
37 root 1.521 5.17 Sat Aug 22 23:09:31 CEST 2009
38 root 1.520 - work around a bug in the perl debugger causing crashes
39     when running under the debugger by marking _pool_handler as nodebug.
40 root 1.519 - speed up Coro::async considerably.
41 root 1.521 - try some hacks to get netbsd to work "more often" - their broken
42     setjmp/longjmp, ucontext *and* phtreads are really hard on Coro.
43     - convert Coro to AE 5.0 API.
44 root 1.519
45 root 1.518 5.162 Tue Jul 28 04:04:03 CEST 2009
46 root 1.517 - perl 5.8.2 is now minimum requirement.
47 root 1.518 - skip t/19_handle.t on broken windows perls.
48 root 1.517
49 root 1.516 5.161 Wed Jul 22 04:47:38 CEST 2009
50     - Coro::AnyEvent::poll could have a different prototype when EV was
51     used as backend (analyzed by Tatsuhiko Miyagawa).
52 root 1.515 - Coro::AnyEvent errornously initialised the event loop when loaded,
53     not on demand.
54 root 1.516 - try to workaround rare */t/01_unblock.t failures.
55 root 1.515
56 root 1.514 5.16 Tue Jul 21 01:44:37 CEST 2009
57 root 1.512 - Coro::AnyEvent failed to hook into the event loop
58     when no threads had been readied between detecting
59     the event loop and actually running it.
60 root 1.509 - considerably speed up Coro::Select by taking avdantage
61     of AnyEvent > 4.8 and some other optimisations.
62 root 1.514 - implement paragraph readline mode in Coro::Handle
63     (based on patches by Zsbán Ambrus).
64 root 1.513 - replace WSAEINPROGRESS by WSAEWOULDBLOCK (reported
65     and analyzed by Yasuhiro MATSUMOTO).
66 root 1.511 - clarified libcoro license and copyright.
67 root 1.510 - someone stole my EXTRA_META!!!
68 root 1.512 - implement Coro::Select::patch_pp_sselect and it's brother,
69     for hardcode select overriding.
70 root 1.509
71 root 1.508 5.151 Mon Jul 6 05:41:57 CEST 2009
72 root 1.506 - backport to windows process emulation code again (patch by
73     Yasuhiro MATSUMOTO).
74 root 1.507 - slightly update Coro::MakeMaker.
75 root 1.506
76 root 1.505 5.15 Tue Jun 30 10:28:06 CEST 2009
77 root 1.502 - deprecate Coro::Socket, document how to get ipv6 support via
78     AnyEvent::Socket instead.
79 root 1.500 - implement signal->wait ($cb) interface, similar to semaphores.
80 root 1.501 - work around SvOK not supporting getmagic, so we have to getmagic
81     to test for undef :/ (reported by Matthias Waldorf).
82 root 1.502 - load Coro::AnyEvent in all modules using AnyEvent.
83 root 1.503 - work around perl corrupting our internal data structures,
84     reported by Tokuhiro Matsuno.
85 root 1.504 - enable per-coroutine real and cpu time gathering
86     (Coro::State::enable_times).
87 root 1.500
88 root 1.499 5.14 Wed Jun 24 01:37:48 CEST 2009
89 root 1.497 - provide explicit functions to "cede" to the event loop in Coro::AnyEvent,
90     as this seems to have been a difficult concept (poll, sleep, idle,
91     idle_upto).
92     - add Coro::AnyEvent::readable/writable functions.
93     - clarify Coro::EV/Event/AnyEvent manpages.
94 root 1.496 - free per-thread global scalars in the thread calling ->cancel, to
95     avoid crashes when $_, $@ etc., are magical but some of those
96     had already been freed.
97 root 1.499 - "unexperimentalise" the callback interface for Coro::Semaphore.
98 root 1.493 - speed up ready queue management/context switching by using a linked
99 root 1.492 list instead of an array (~5-10%).
100 root 1.493 - implement "watch" command in Coro::Debug shells.
101 root 1.494 - for fun, implement time-slicing as an example in the manpage.
102 root 1.497 - if AnyEvent detects EV or Event, but we don't have Coro::EV or
103     Coro::Event, use the normal AnyEvent handling instead of dieing
104     (the same is true for Coro::Handle).
105     - properly document Coro::EV::timed_io_once.
106 root 1.498 - avoid unneccessary ->cancel calls in Coro::Handle.
107     - maybe make it work on mingw32 with win32 backend
108     (based on patch by Yasuhiro Matsumoto).
109 root 1.492
110 root 1.490 5.132 Fri May 29 09:00:39 CEST 2009
111 root 1.488 - do not keep a reference to the argument itself in
112 root 1.491 Coro::Semaphore::guard, as it could change later.
113 root 1.489 - support SO_RCVBUF/SO_SNDBUF nonstandard Coro::Socket options,
114     should support a prepare callback.
115 root 1.488
116 root 1.487 5.131 Mon Mar 16 23:20:37 CET 2009
117 root 1.485 - implement and document Coro->suspend, Coro->resume.
118 root 1.486 - fix Coro::Select implementation to not (often) close
119     the passed file descriptors (testcase provided by pippijn).
120 root 1.485
121 root 1.484 5.13 Mon Dec 15 21:51:42 CET 2008
122 root 1.481 - EXPERIMENTAL: implement dynamic winds (on_enter/on_leave).
123 root 1.476 - don't set diehook to C<undef> but instead to NULL, to avoid
124     spurious warnings.
125 root 1.478 - fix a lot of bugs in Coro::SemaphoreSet.
126     - Coro::SemaphoreSet will less often create a semaphore needlessly.
127     - add Coro::SemaphoreSet::count and wait methods.
128 root 1.477 - take advantage of the new Guard module.
129     - deprecate Coro::guard.
130 root 1.479 - try to fix the dreaded 01_unblock tests once more. I hate it when
131     testsuites need more fixing than the code they are supposed to test.
132 root 1.480 - croak in more cases when a required callback isn't resolvable.
133     - fix some minor issues in Coro::State->call/eval.
134     - use current coroutine context instead of a temporary one
135     when temporarily switching to another coroutine.
136 root 1.484 - do not call C-level on_destroy handlers during global destruction,
137     to avoid needless segfaults.
138 root 1.476
139 root 1.475 5.12 Sun Dec 7 13:30:38 CET 2008
140 root 1.467 - add default config for MirOS, which seems to be a bug-to-bug
141     compatible fork of openbsd ("world domination by releasing
142     openbsd cvs before the openbsd folks do it" or so :).
143 root 1.468 - free_padlist did destroy the names pad, not good, but didn't
144 root 1.470 seem to bother perl - this could fix issues such as eval ""
145 root 1.477 inside a function called from multiple coroutines.
146 root 1.467 - use a different method to detect destruction time.
147     - be more careful when freeing padlists just before global
148     destruction.
149 root 1.469 - fixed and expanded the call/cc example.
150 root 1.471 - renamed _terminate to _coro_run.
151 root 1.473 - new method Coro::Channel->shutdown.
152     - try pthreads on openbsd <4.4 (broken sigaltstack, will
153     pthreads fare better?).
154     - be less picky about destroying "a" running coroutine.
155 root 1.466
156 root 1.465 5.11 Tue Nov 25 21:49:05 CET 2008
157 root 1.464 - DEBUGGING in 5.10.0 is a mess: it constantly flags perfectly
158     working code with failed assertions, introducing more bugs than
159     it fixes, requiring elaborate workarounds :(
160    
161 root 1.463 5.1 Mon Nov 24 08:54:59 CET 2008
162     - wrote a small introductory tutorial - Coro::Intro.
163 root 1.454 - convert Coro::Timer, Coro::Select and Coro::Util to rouse API.
164 root 1.453 - Coro::Select did errornously dup the file descriptors
165     and didn't work with all AnyEvent backends.
166 root 1.463 - Coro::Select wasn't imported correctly form Coro::LWP, causing blocking
167     LWP data transfers.
168 root 1.458 - disassociate c contexts from coro objects - this is agruably more
169     correct, but mostly allows sharing of cctxs between coro and state
170     objects, for added memory savings and speed increases.
171 root 1.455 - bumped $Coro::POOL_RSS up to 32kb by default.
172 root 1.456 - no longer set the optype to OP_CUSTOM, as B::* understandably
173     doesn't like this very much (and we *are* a type of entersub).
174 root 1.458 - implement state cloning, just to prove that call/cc can be done.
175 root 1.460 - automatically load Coro::AnyEvent in Coro::Handle.
176 root 1.461 - wrap ->cancel calls in eval inside Coro::Handle as EV watchers
177     do not have this method (and don't need it either).
178     - speed up generic anyevent methods in Coro::Handle by using rouse
179     callbacks.
180     - allow coroutines in $Coro::IDLE, speeding up Coro::AnyEvent and
181     others. It also makes the debugger happier, as you can trace
182     through the idle threads now.
183 root 1.462 - add comppad_name* and hints ($^H, %^H) to per-thread variables.
184 root 1.461 - eg/event was pretty broken.
185 root 1.457 - better 5.8.6 compatibility.
186 root 1.453
187 root 1.452 5.0 Thu Nov 20 10:35:05 CET 2008
188 root 1.424 - NEW ARCHITECTURE: use the latest 4.x version if you experience
189     stability issues.
190 root 1.411 - bump API version to 7 - all dependents must be recompiled.
191 root 1.430 - removed timed_* functions - they were not being used anyways
192     and should be replaced by a more generic mechanism -
193 root 1.429 and were annoying to support anyways :)
194 root 1.451 - removed SemaphoreSet's waiter method - use sem method instead.
195 root 1.415 - Coro::Semaphore->adjust didn't correctly wake up enough waiters.
196 root 1.416 - async_pool did free a scalar value twice
197     ("Attempt to unreference...").
198 root 1.420 - fix a longstanding bug where calling terminate on a coro that
199     was waiting for a semaphore that was just becoming available
200     would cause a deadlock (semaphore would get into a state where
201 root 1.433 it was available but waiters were still blocked).
202 root 1.432 - calling throw on a coroutine that is waiting for a semaphore will
203     no longer make it acquire the semaphore (and thus leak a count).
204 root 1.412 - perl's process emulation is now not even theoretically supported
205     anymore.
206 root 1.447 - new functions Coro::rouse_cb and Coro::rouse_wait for easier
207     conversion of callback-style to blocking-style.
208     - new methods $coro->schedule_to and ->cede_to, to specifically
209     schedule or cede to a specific coroutine.
210 root 1.433 - new function Coro::Semaphore::wait.
211 root 1.428 - use named constants in Coro::Channel (Richard Hundt).
212 root 1.412 - directly patch the entersub opcode calling SLF functions (cede,
213     transfer and so on). this does speed up context switching, but
214     more importanly, it frees us from the hardcoded behaviour of
215     entersub, so we might actually be able to return something from
216     those functions and atcually create new ones.
217     - take advantage of __builtin_frame_address on gcc.
218     - expose THX in coroapi (not sure whether this was a wise decision,
219     as "threaded" perls are running at half speed anyways).
220     - implement execute_slf (schedule-like-function) interface that makes
221     it possible to implement schedule-like-functions in XS.
222 root 1.424 - use new SLF interface to massively speed up Coro::EV by roughly a
223     factor of two.
224     - used new SLF interface to massively speed up Coro::Semaphore by a
225     factor of three.
226 root 1.426 - used new SLF interface to speed up Coro::AIO by roughly a factor of
227     four and reduce its memory usage considerably.
228 root 1.429 - implement Coro::SemaphoreSet purely in terms of Coro::Semaphore,
229     for a nice speedup and vastly more correct behaviour. Also implement
230     a new method "sem" to get at the underlying semaphore object.
231 root 1.430 - implement Coro::Channel in terms of Coro::Semaphore, for a moderate
232     (in comparison) 20-40% speedup.
233 root 1.435 - used new SLF interface to reimplement Coro::Signal gaining
234     some unknown (because I was too lazy), but certain, speedup, and also
235     making signals reliable for the first time.
236 root 1.445 - used new SLF interface and other optimisations to speed up async_pool
237     by a factor of two. It also doesn't rely on perl's exception mechanism
238 root 1.446 to exit anymore. The overhead for terminating an async_pool, coro over
239     a normal async is now very small.
240 root 1.444 - sped up coroutine creation/destruction by 40%.
241 root 1.417 - forgot to include Coro/libcoro/README in the dist for all these years.
242 root 1.418 - work around a freebsd pthreads bug (manual testcancel is required as
243     pthread_cond_wait isn't a cancellation point on freebsd).
244 root 1.438 - use new rouse functions to speed up and simplify Coro::BDB.
245 root 1.439 - make "prefer perl native functions" work with threaded perls.
246 root 1.440 - condense Coro::Debug ps output, hint at v and w flags.
247 root 1.436 - (libcoro) lots of minor cleanups and portability improvements.
248 root 1.416
249 root 1.449 4.914 Wed Nov 19 12:54:18 CET 2008
250     - fix a disastrous bug in the readline optimisation
251     introduced in 4.801.
252    
253 root 1.416 4.913 Sat Nov 15 07:58:28 CET 2008
254     - async_pool did free a scalar value twice
255     ("Attempt to unreference...").
256 root 1.408
257 root 1.407 4.912 Thu Nov 13 18:31:23 CET 2008
258 root 1.404 - minor cleanups.
259 root 1.405 - use much larger stacks on linux and perl < 5.8.8.
260 root 1.406 - Coro::Debug::new_unix_server did not unlink the socket
261     when destroyed.
262 root 1.404
263 root 1.403 4.911 Tue Nov 11 04:26:17 CET 2008
264 root 1.401 - "port" to threaded perls.
265    
266 root 1.400 4.91 Mon Nov 10 05:36:38 CET 2008
267     - the ->throw exception object no longer leaks.
268     - creating a new cctx leaked a scopestack entry (memleak).
269     - new coroutines didn't get created with a zero flags field
270     (unknown impact).
271 root 1.399 - calling ->throw on a not-yet-started coroutine should now work
272     instead of being ignored.
273 root 1.400 - ->throw is now supported on Coro::State objects.
274 root 1.398 - clean up cctx creation code a bit.
275 root 1.400 - entersub is actually a UNOP, not a LOGOP (not a bugfix).
276 root 1.398
277 root 1.397 4.9 Sat Nov 8 17:45:27 CET 2008
278 root 1.392 - (libcoro) did not preserve rbp with CORO_ASM (we are getting there).
279     - (libcoro) no longer leak threads in the experimental pthread backend,
280     also speed it up considerably.
281 root 1.394 - (libcoro) do not rely on makecontext passing void *'s unscathed.
282 root 1.392 - fix compiletime dependencies on libcoro in the Makefile.
283     - cctx_count wasn't always updated properly.
284 root 1.393 - Coro::State::cctx_stacksize wasn't applied correctly.
285     - new function Coro::State::cctx_max_idle.
286     - the default max number of idle C contexts is now 4.
287 root 1.394 - (libcoro) try harder to get _setjmp/_longjmp.
288     - (libcoro) cleanup and extend the libcoro API to officially
289     allow the creation of empty source contexts.
290 root 1.395 - very experimental workaround for the totally broken netbsd platform.
291 root 1.396 - tried to hack around openbsd bugs.
292 root 1.391
293 root 1.390 4.804 Wed Nov 5 16:36:00 CET 2008
294 root 1.389 - Coro::Debug::new_unix_server would not create a non-blocking listening
295 root 1.390 socket, sometimes causing freezes.
296 root 1.392 - (libcoro) fix misaligned stack points for setjmp and assembly
297     methods, which can cause crashes on x86/x86_64 with a sufficiently
298     aggressive compiler.
299 root 1.389 - new function: Coro::Debug::new_tcp_server.
300 root 1.388 - move ->throw into the Coro class because it only works on coro objects.
301    
302 root 1.387 4.803 Mon Nov 3 17:16:12 CET 2008
303 root 1.392 - (libcoro) use a global asm statement to become independent of gcc
304     otpimisations for CORO_ASM (thanks to pippijn for the idea).
305 root 1.387 - try to workaround yet another broken bsd, this time dragonfly.
306 root 1.386
307 root 1.385 4.802 Thu Oct 30 10:56:12 CET 2008
308     - support -fno-omit-frame-pointer on x86 with the assembly method.
309 root 1.383 - tune 01_unblock.t tests a bit.
310    
311 root 1.382 4.801 Wed Oct 22 18:33:37 CEST 2008
312 root 1.381 - improve readline speed for very long "lines".
313 root 1.382 - backport to 5.8.8.
314 root 1.381
315 root 1.380 4.8 Thu Oct 2 13:34:40 CEST 2008
316 root 1.378 - new function Coro::AIO::aio_wait.
317     - Coro.:AIO and Coro::BDB now "use Coro::AnyEvent".
318     - greatly speed up and reduce memory usage of Coro::AIO requests.
319 root 1.379 - implement some other µ-optimisations.
320 root 1.378
321 root 1.377 4.749 Mon Sep 29 14:40:12 CEST 2008
322     - port to slow and broken pseudo-threaded perls. (courtesy pippijn).
323    
324 root 1.376 4.748 Sat Sep 27 14:03:19 CEST 2008
325 root 1.374 - implement, but do not document, PerlIO::cede(granularity).
326 root 1.372 - Coro::Storable forgot to wrap Storable::pstore.
327 root 1.374 - work around the multitude of leaks and memory corruption
328 root 1.375 bugs in PerlIO::via by using our own C-level perliol. As a side
329     effect, Coro::Storable is now much, much, much faster.
330 root 1.372
331 root 1.371 4.747 Tue Sep 23 01:59:41 CEST 2008
332 root 1.369 - fix a per-cv memleak (one empty array was leaked per
333     code reference).
334 root 1.370 - avoid a crash in coro->call|rss when the coroutine was already
335     destroyed (most noticably when using Coro::Debug::ps :)
336     - also protect *Storable::FILE.
337 root 1.371 - push up default storable granularity to 20ms.
338 root 1.369
339 root 1.368 4.746 Sun Sep 21 03:22:20 CEST 2008
340     - be more insistent on locking Storable against reentrancy
341     in Coro::Storable.
342 root 1.366 - move swap_def?v and throw to Coro::State, as documented.
343    
344 root 1.365 4.745 Thu Jul 24 00:14:38 CEST 2008
345 root 1.364 - remove debugging code related to MgPV_nolen_const, also try to
346     make it compile with perl 5.8.6 (yes, apple apparently loves
347     outdated software). Reported by John S.
348 root 1.363
349 root 1.362 4.744 Tue Jul 8 22:06:35 CEST 2008
350 root 1.361 - correctly provide default DIE/WARN handlers as documented.
351     - also overwrite PL_vtbl_sigelem.svt_clear, even though current
352     implementations inside perl work fine for us.
353    
354 root 1.360 4.743 Mon Jun 16 00:21:57 CEST 2008
355     - when using Coro::EV without running EV::loop it could
356     result in busy-waiting for events, this has been fixed.
357     - reduce codesize and improve performance by using EV_DEFAULT_UC.
358    
359 root 1.359 4.742 Sat May 31 14:10:21 CEST 2008
360     - implement a workaround for (some) perl <5.8.8 versions.
361     - require EV 3.3+.
362    
363 root 1.358 4.741 Fri May 30 23:33:09 CEST 2008
364     - tell netbsd how utterly broken their imitation of an OS is
365     and refuse to build by default if pthreads are in use.
366     - switch to "s" method on all bsd's by default, as their ucontext
367     stuff seems just too broken.
368     - fix a bug in Coro::Select.
369    
370 root 1.357 4.74 Thu May 29 20:05:31 CEST 2008
371     - do not test Coro::LWP for lack of dependencies.
372     - do not test Coro::Select for lack of working perls.
373    
374     4.73 Thu May 29 2008
375 root 1.356 - fix a bug in Coro::EV which would cause it to block despite
376     there being runnable coroutines.
377 root 1.354 - sprinkle "no warnings" freely over everything, also suppress
378     warnings for some other modules.
379 root 1.355 - fix typo in WSAEWOULDBLOCK.
380 root 1.354
381 root 1.353 4.72 Sun May 25 05:14:36 CEST 2008
382 root 1.351 - tweak META.yaml a bit, unfortunately, there is no documented way
383     to have optional dependencies with CPAN. doh :(
384 root 1.350 - avoid running some tests on windows because they would fail due to
385 root 1.351 perl bug (broken fork, broken pipes...).
386     - work around perl on windows bugs where perl returns undocumented
387     error codes for sysread, syswrite etc. by taking advantage
388     of AnyEvent's workaround for that problem.
389 root 1.350 - use AnyEvent::Util::fh_nonblocking in Coro::Handle to work around
390     a common perl implementation bug on windows.
391     - use unix domain sockets in testsuite to work around
392 root 1.351 common perl implementation bugs on widows (they are emulated by
393     tcp sockets on windows. ugh.)
394 root 1.350
395 root 1.349 4.71 Sat May 24 20:01:27 CEST 2008
396 root 1.348 - fix a bug in Coro::AnyEvent ("Usage: Coro::AnyEvent::_schedule()").
397 root 1.347 - take advantage of async name resolution of AnyEvent::Util.
398     - work around brutal inet_aton override in Coro::LWP.
399 root 1.348 - take advantage of the readyhook in Coro::EV, for smoother
400     scheduling.
401 root 1.347
402 root 1.346 4.7 Sun May 11 00:32:19 CEST 2008
403 root 1.342 - completely reworked the Coro manpage.
404     - added Coro::AnyEvent, generic event loop integration.
405 root 1.341 - implement cancel, ready and kill commands in Coro::Debug.
406     - document find_coro in Coro::Debug.
407 root 1.343 - incompatible change: rename has_stack to has_cctx.
408 root 1.346 - Coro::AIO and Coro::BDB no longer force event model detection,
409     use AnyEvent::AIO and AnyEvent::BDB.
410 root 1.341
411 root 1.340 4.6 Sat Apr 26 10:05:14 CEST 2008
412 root 1.336 - INCOMPATIBLE CHANGE: sub/code attributes are no longer supported
413 root 1.339 by the Coro module. It was a mistake to have it in the first place.
414     - (experimental) support for activestate perl 5.10 (method "w").
415     - (experimental) support for strawberry perl (method "a").
416 root 1.337 - coro_sigelem_set did not return a value although it had to,
417     actual impact unknown.
418 root 1.336
419 root 1.335 4.51 Mon Apr 14 13:28:27 CEST 2008
420     - make it compile again on 5.8.
421    
422 root 1.334 4.50 Thu Apr 10 09:43:17 CEST 2008
423     - I did it twice! (see 4.49).
424    
425 root 1.333 4.49 Sun Apr 6 21:23:31 CEST 2008
426     - grr, instead of compiling the recent changes on 5.10 only they
427     were compiled on 5.8 only.
428    
429 root 1.332 4.48 Sun Apr 6 20:36:46 CEST 2008
430 root 1.331 - allow coroutine switches during eval's under 5.10.x, as it seems
431     the parser is a per-interpreter option now, so this is safe
432     (this might fix the odd crashing bug).
433 root 1.332 - drop support for 5.9.x versions: they are dead, jim.
434 root 1.331
435 root 1.330 4.47 Sun Apr 6 00:37:52 CEST 2008
436 root 1.329 - force cctx allocation on API calls: we know we need to force one
437     and gcc actually manages to confuse our heuristic nowadays,
438     leading to crashes and worse.
439 root 1.330 - document force_cctx.
440 root 1.329
441 root 1.328 4.46 Fri Apr 4 22:05:43 CEST 2008
442     - upgrade libcoro, resulting in pthread-backend (which was only created
443     to fulfill the rules of the programming languages shootout).
444    
445 root 1.327 4.45 Thu Mar 13 11:55:36 CET 2008
446 root 1.322 - fix a file leaking bug in eg/mhyttpd that would allow
447     downloading of any file (reported by oesi).
448 root 1.325 - fix deadlock bug in Coro::Channel (reported by Richard Hundt)
449     (also add testcase).
450 root 1.323 - support Broadcast option in Coro::Socket (patch by Richard Hundt,
451     apparently having loads of fun with that).
452 root 1.324 - implement $state->swap_defsv, ->swap_defav and document ->throw.
453 root 1.322
454 root 1.321 4.4 Wed Feb 13 16:44:29 CET 2008
455 root 1.319 - only assume .cfi pseudo isns exist on GNU/Linux.
456 root 1.320 - add get_timed method to Coro::Channels.
457 root 1.321 - fixed Coro::Debug manpage.
458     - perl 5.11 compatbility improvement based on patch by
459     Andreas König.
460 root 1.319
461 root 1.318 4.37 Sun Jan 20 11:25:23 CET 2008
462 root 1.317 - soften the check for an active parser for perl 5.10 (avoiding
463     "Coro::State::transfer called while parsing" in many cases).
464 root 1.316
465 root 1.315 4.36 Sun Jan 13 10:53:56 CET 2008
466 root 1.314 - reset diehook when terminating from an async_pool as to not
467     trigger any __DIE__ handlers.
468    
469 root 1.313 4.35 Sun Jan 13 04:14:13 CET 2008
470 root 1.308 - "bt" debug command now displays any exceptions
471 root 1.311 from longmess and also skips the innermost
472     stackframes, giving more useufl output.
473 root 1.309 - allow backtraces in coroutines blocked in Coro::EV,
474 root 1.310 at a <1% speed hit.
475 root 1.312 - handle localising of $SIG{__WARN__} and __DIE__
476 root 1.314 properly (with a proper amount of dirty hacking).
477 root 1.308
478 root 1.307 4.34 Sat Dec 22 17:49:53 CET 2007
479 root 1.306 - upgrade to EV version 2.0 API.
480    
481 root 1.305 4.33 Mon Dec 17 08:36:12 CET 2007
482     - make Coro::AIO etc. loadable in the absence of EV.
483    
484 root 1.304 4.32 Mon Dec 17 07:46:02 CET 2007
485 root 1.301 - majorly improved Coro::Handle's performance with Coro::EV.
486 root 1.302 - implemented the until now mythical Coro::BDB module.
487 root 1.304 - specialcase EV in Coro::AIO and Coro::BDB for extra speed.
488 root 1.301
489 root 1.300 4.31 Wed Dec 5 12:32:39 CET 2007
490     - remove warn statement form Coro::Util.
491    
492 root 1.299 4.3 Tue Dec 4 20:33:14 CET 2007
493     - calls to the idle function could cause stack corruption
494     when the stack changed.
495 root 1.295 - do no longer rely on the presence of EV::DNS if EV is used
496     (because it is gone), but instead take avdantage of EV::ADNS
497     if available.
498 root 1.296 - add ($) prototypes to all functions in Coro::Storable.
499 root 1.297 - use a conventional (safer) idle callback in Coro::EV.
500 root 1.298 - do accept quickly in Coro::Debug to avoid endless loops.
501 root 1.295
502 root 1.294 4.22 Fri Nov 30 16:04:04 CET 2007
503 root 1.293 - really use optimised versions for Event and EV in Coro::Util
504     and Coro::Handle.
505    
506 root 1.292 4.21 Sun Nov 25 10:48:59 CET 2007
507 root 1.290 - fix a spurious memory read.
508 root 1.291 - Coro::EV no longer keeps the eventloop "alive".
509 root 1.290
510 root 1.289 4.2 Fri Nov 9 20:47:05 CET 2007
511 root 1.282 - enable/disable tracing from a new coroutine, not a pooled one.
512 root 1.283 - fix a memleak in Coro::Event.
513 root 1.284 - removed killall call from fork_eval.
514     - made sure store_fd is already loaded so that fork_eval
515     does not have to parse autoload in each subprocess.
516 root 1.288 - only use assembly method if -O switch is in $Config{optimize}.
517 root 1.289 - add (optional) Coro::EV module, so far the best event loop module
518     directly supported by Coro.
519     - if the event model is EV, use EV::DNS to resolve
520     stuff in Coro::Util.
521 root 1.286 - don't get confused by multiple event notifications in Coro::Handle.
522 root 1.287 - initial support for EV (libevent interface).
523 root 1.289 - require Event and EV using configure_requires, to force their existance.
524 root 1.282
525 root 1.281 4.13 Wed Oct 24 07:26:45 CEST 2007
526 root 1.277 - add Coro::Storable::blocking_thaw.
527 root 1.278 - use a vastly more complicated technique to localise
528     $SIG{__WARN/DIE__} that also works on perls <= 5.8.8.
529 root 1.279 - use a coroutine for the idle callback Coro::Event,
530     instead of running Event in the current coroutine context.
531     This also catches recursive invocations.
532 root 1.280 - actually report fork errors in gethostbyname and inet_aton.
533 root 1.277
534 root 1.276 4.11 Thu Oct 11 02:40:24 CEST 2007
535     - port to threaded perls.
536    
537 root 1.275 4.1 Thu Oct 11 02:38:16 CEST 2007
538 root 1.264 - incompatible change: $SIG{__DIE__} and $SIG{__WARN__} will now
539     be local to each coro (see Coro::State).
540 root 1.274 - incompatible change: for very deep reasons, cede and cede_notself
541     cannot return anything, so nothing will be returned.
542 root 1.264 - possibly bring back 5.10 compatibility (untested).
543     - work around stupid (and wrong) warning on 5.10 :(.
544     - overlay the saved state over the context stack. This saves
545     a few hundred bytes per coroutine on average and also
546     speeds up context switching a bit.
547 root 1.265 - further tune default stack sizes.
548 root 1.264 - (more) correctly calculate stack usage in coro_rss.
549 root 1.267 - Coro::Storable::blocking_* did not properly lock
550     resulting in races between coroutines.
551     - added Coro::Storable::guard.
552 root 1.268 - stopping to trace a coroutine could destroy the cctx of
553     an unrelated coroutine.
554 root 1.269 - explain the relationship between Perl and C coroutines in
555     more detail in Coro::State.
556 root 1.270 - Coro::Util::inet_aton did not short-circuit dotted quad forms,
557     causing a fork per resolve. This also affected Coro::Socket.
558 root 1.272 - switch to a separate stack in $coro->call/eval to avoid
559     invalidating pointers.
560 root 1.264
561 root 1.263 4.03 Sat Oct 6 21:24:00 CEST 2007
562 root 1.262 - added Coro::throw method.
563     - minor code cleanups.
564    
565 root 1.261 4.02 Sat Oct 6 02:36:47 CEST 2007
566 root 1.258 - fix a very minor per-coroutine memleak (a single codereference).
567 root 1.259 - fixed a bug where the currently in-use c context would be freed
568     prematurely (can happen only when programs change the stacksize
569     or use tracing).
570     - tracing can no longer keep a coro alive after it terminated.
571 root 1.260 - do static branch prediction in the common path for gcc. gives
572     about 2-5% speed improvement here.
573 root 1.258
574 root 1.256 4.01 Fri Oct 5 22:10:49 CEST 2007
575 root 1.257 - instead of recreating *a* standard output handle we simply
576     use STDOUT, which is faster and hopefully more robust.
577 root 1.255
578 root 1.254 4.0 Fri Oct 5 12:56:00 CEST 2007
579 root 1.238 - incompatibly changed Coro::Storable::freeze.
580 root 1.244 - major new feature: added Coro::Debug, for interactive coroutine
581     debugging, tracing and much more.
582 root 1.252 - major bug fix: unbelievable, but true: $_, $/ and many other
583     "saved" variables actually weren't being saved. This has been fixed,
584     of course, while increasing performance while losing all the save
585     flags.
586 root 1.251 - save flags are gone, and all the api functions dealing with them.
587 root 1.232 - added Coro::Semaphore::adjust.
588 root 1.233 - added Coro::Util::fork_eval.
589     - added Coro::Storable::{nfreeze,blocking_{freeze,nfreeze}}.
590 root 1.248 - added Coro::killall.
591 root 1.236 - reduce initial stack sizes to allow for "micro-coroutines".
592 root 1.239 - better async_pool resource management, moved parts of async_pool
593 root 1.240 handling to XS (major speed improvement).
594 root 1.245 - actually croak before modifying important data structures.
595     - refuse to transfer while compiling.
596     - possibly support eval EXPR better now.
597 root 1.249 - enable assembly per default on linux+bsd x86+amd64.
598 root 1.253 - all internal members were renamed _something for easier subclassing.
599 root 1.237 - many minor tweaks.
600 root 1.232
601 root 1.231 3.63 Wed May 16 14:10:06 CEST 2007
602 root 1.230 - implement handcoded assembly for x86/amd64 SVR ABI.
603    
604 root 1.229 3.62 Fri Apr 27 21:36:06 CEST 2007
605     - upgrade libcoro (which might set unwind info correctly).
606 root 1.228 - change default on linux to setjmp/longjmp.
607 root 1.227
608 root 1.226 3.61 Thu Apr 19 12:36:18 CEST 2007
609 root 1.225 - Coro::Storable caused an endless loop when thawing invalid
610     pst's sometimes.
611 root 1.223 - use a Semaphore in Coro::Storable, as Storable doesn't
612 root 1.224 seem to be reentrant (although it is documented to
613     be threadsafe...).
614     - fix Coro::Signal to bring back the original unreliable
615     but stateful semantics.
616 root 1.226 - fixed a lot of typos in Coro.pm (patch submitted by David
617     Steinbrunner, which applied flawlessly).
618 root 1.223
619     3.6 Sat Apr 14 17:13:31 CEST 2007
620 root 1.217 - added some bugfixes to get eg/myhttpd working again.
621 root 1.218 - added Coro::Storable for often-cede'ing freeze/thaw.
622 root 1.219 - try to do a clean exit when a coroutine calls exit
623     (EXPERIMENTAL).
624 root 1.220 - got rid of indirect call through _coro_init.
625     - updated the partly antique examples in eg/ to
626     work again and be a bit less magic, too.
627     - fixed Coro::Signal semantics to work as documented again.
628 root 1.217
629 root 1.216 3.55 Sun Mar 25 01:20:47 CET 2007
630 root 1.215 - add SAVE_DEFFH to save the default filehandle and enable
631     it by default.
632     - finally move socket-operations from Coro::Socket to Coro::Handle
633     to be able to unblock foreign sockets.
634     - add Coro::State::save_also and guarded_save.
635 root 1.211 - add count accessor to Coro::Semaphore.
636 root 1.212 - add Coro::State::cctx_stacksize.
637 root 1.213 - just for the fun of it, do not rely on implicit context,
638     which can dramatically improve performance, but people
639     using windows-process-emulation perls do not care much about
640     performance.
641 root 1.211
642 root 1.210 3.51 Sun Mar 4 14:18:04 CET 2007
643 root 1.209 - fixed a problem when you weakened references to Coro::State's
644     (patch by Yuval Kogman).
645    
646 root 1.208 3.501 Wed Feb 28 12:44:07 CET 2007
647     - rename some global symbols as macosx from hell redefines
648     them without asking.
649    
650 root 1.207 3.5 Tue Feb 13 20:22:53 CET 2007
651 root 1.204 - do AnyEvent model detection earlier, avoiding problems
652     caused by first using AnyEvent and later Coro::Event.
653 root 1.205 - implement and document Coro::Event event objects.
654     - fix a potential problem in Coro::Event causing crashes.
655 root 1.206 - initialise PL_comppad when creating a new coroutine,
656     avoids crashes on early coro destruction.
657 root 1.204
658 root 1.203 3.41 Mon Jan 22 19:19:49 CET 2007
659 root 1.202 - readline on Coro::Handle did not support undefined $/,
660     nor did it deliver partial lines on EOF or error.
661 root 1.203 - implement malloc fallback for stack allocation because
662     stupid broken idiotic OSX has a stupid broken
663     idiotic fits-the-whole-os mmap "implementation" and
664     my dick feels longer if Coro is portable even to
665     obsolete platforms.
666 root 1.202
667 root 1.201 3.4 Fri Jan 19 21:52:54 CET 2007
668 root 1.197 - remove t/09_timer.t, as it isn't really testing much
669     but was rather flaky in practise.
670 root 1.198 - async_pool coro would keep arguments and callback alive until
671     it was reused.
672 root 1.199 - cancellation of a coroutine could cause spurious idle calls
673     in cede_notself.
674 root 1.197
675 root 1.196 3.3 Sat Jan 6 03:45:00 CET 2007
676 root 1.187 - implement $coro->on_destroy.
677 root 1.188 - Coro::Event blocking semantics have been changed,
678     documented and - hopefully - improved.
679 root 1.184 - fix nice adding, not subtracting, from priority.
680 root 1.185 - fix ->prio and api_is_ready (patch by Mark Hinds).
681 root 1.190 - fixed an assert ("... == prev__cctx->idle_te")
682     that could errronously trigger.
683 root 1.192 - fix various large and small memleaks.
684 root 1.191 - use a (hopefully) more stable cancel implementation
685     that immediately frees the coroutine data.
686 root 1.194 - cede/cede_notself return a status now.
687 root 1.191 - added Coro::guard function.
688 root 1.193 - added a global coroutine pool for jobs (on my machine,
689     I can create and execute 48k simple coros/s with async,
690     and 128k coros with async_pool).
691 root 1.195 - Coro::AIO now uses the coroutine priority as io priority.
692 root 1.184
693 root 1.183 3.2 Fri Dec 22 05:07:09 CET 2006
694 root 1.179 - improve portability to slightly older perls.
695 root 1.180 - use cleaner coroutine destruction.
696 root 1.182 - simplify configuration for users.
697     - optionally (unrecommended) prefer perl functions over
698     their coro replacements.
699 root 1.179
700 root 1.178 3.11 Tue Dec 5 13:11:24 CET 2006
701     - fixed some bogus assert's, but as perl.h disables assert even
702     without NDEBUG (thank you very much), not too many people should
703     notice (that did include myself). Andreas König noticed, though :)
704     - do not save/restore PL_sortcxix on >= 5.9.x, it doesn't seem to have
705     it. Also noticed by Andreas König :)
706     - save/restore tainted status.
707     - verified to pass the testsuite on my 5.9.5.
708    
709 root 1.177 3.1 Mon Dec 4 23:03:40 CET 2006
710     - INCOMPATIBLE CHANGE: $/ is now per-coroutine (but slow).
711 root 1.175 - incompatible change: transfer flags are now per-state.
712 root 1.173 - give a better error message on deadlock.
713 root 1.175 - document Coro::nready.
714 root 1.174 - enhanced testsuite.
715 root 1.169
716 root 1.172 3.01 Sun Dec 3 23:47:42 CET 2006
717     - forgot to include Coro::Timer.
718    
719 root 1.171 3.0 Sun Dec 3 22:57:25 CET 2006
720 root 1.163 - the "FINALLY COMPLETELY STABLE" release (coming soon:
721     the "FAMOUS LAST WORDS" release).
722 root 1.166 - implement a new stack sharing algorithm, which uses a stack
723     pool (size currently hardcoded).
724 root 1.155 - make stack sharing mandatory (it no longer uses a heuristic).
725 root 1.157 - eval/die no longer cause weird problems under heavy use.
726 root 1.152 - Coro::Event could cause livelocks if it was used but
727 root 1.161 no Coro::Event watchers were used.
728 root 1.167 - Coro::Event now uses asynccheck as prepare does not
729     check for changed watchers.
730 root 1.170 - Coro::Event allows multiple waiting coros and will wake up one
731     per event.
732 root 1.152 - Coro::Event should be cleaner and more efficient now.
733 root 1.165 - new utility function Coro::unblock_sub.
734 root 1.164 - document the sad fact that Event is no longer reentrant.
735 root 1.163 - putting a coroutine into the ready queue twice could under
736     some circumstances lead to stack corruption.
737 root 1.161 - minor incompatibility: subclassing Coro::State is supported
738     directly now without going through a _coro_state member.
739 root 1.156 - state/coro switching is much faster now.
740 root 1.154 - very minor optimisations and code/documentation cleanup.
741 root 1.153 - avoid problems due to compiler inlining.
742 root 1.154 - removed timers from Coro::Timer -> use AnyEvent instead.
743     - replaced Coro::idle coroutine by (cleaner) idle handler.
744 root 1.156 - updated to newest libcoro.
745 root 1.159 - implement enhanced support for valgrind.
746 root 1.162 - implement is_ready and return value for ready.
747 root 1.168 - removed Coro::Cont, it was a misnomer (it's generators),
748     and it was somewhat annoying to get it right. Will come back
749     if somebody asks for it :)
750 root 1.163 - many ->wait methods and Coro::Event could return
751     spuriously without the event having happened.
752 root 1.146
753 root 1.145 2.5 Tue Nov 7 12:22:33 CET 2006
754 root 1.143 - made Coro::Util, Coro::Select, Coro::Handle and Coro::Socket
755     use AnyEvent, moved them to Coro/.
756 root 1.140 - added Coro::LWP which contains all the uglyness required to
757     make LWP non-blocking.
758     - should work with perl 5.9.x now (Andreas König made me do it).
759 root 1.141 - fixed another bug in Coro::Select when the timeout was undef.
760 root 1.136 - reuse PL_start_env for all coros, saves some memory per coroutine.
761     - manage PL_top_env differently, hopefully to avoid panic: top_env.
762 root 1.137 - timeout argument was not properly used in Coro::Socket.
763     - allow limited forms of subclassing in Coro::Handle/Coro::Socket.
764 root 1.138 - emulate undocumented(!) functionality of IO::Socket required
765     by LWP(!!).
766     - updated eg/lwp to work with newer lwp's.
767 root 1.140 - remove "FATAL: uncaught exception" prefix. Coroutines that die
768     kill the whole process, just as exceptions in the main "coroutine"
769     did already.
770 root 1.135
771 root 1.134 2.1 Wed Nov 1 23:01:13 CET 2006
772     - fix a long-standing bug in Coro::Select where select with
773     zero timeout would instead change the current default filehandle.
774 root 1.132 - use a simpler and hopefully more robust way to clone padlists
775     (uses less memory and a perl function instead of our own).
776     - coro can now create a stack guard area on many architectures.
777 root 1.134 - Coro::AIO properly reexports additional functions from IO::AIO.
778 root 1.132 - updated libcoro with a workaround for OS X,
779 root 1.134 pach and testing by Michael Schwern.
780 root 1.130
781 root 1.129 2.0 Tue Oct 24 05:47:17 CEST 2006
782 root 1.128 - support additional aio requests in Coro::AIO.
783    
784 root 1.129 1.9 (never properly released due to a glitch)
785    
786 root 1.127 1.8 Thu Feb 2 00:59:06 CET 2006
787 root 1.125 - applied suggested patch by SAMV to avoid problems during stupid
788     mark & sweep gc run.
789 root 1.126 - applied patch by Scott Walters for 5.9.3 compatibility.
790 root 1.125
791 root 1.124 1.7 Tue Dec 27 01:41:58 CET 2005
792 root 1.123 - added Coro::AIO, a thin wrapper around IO::AIO.
793 root 1.122 - improved Makefile.PL explanations.
794    
795 root 1.119 1.6 Mon Dec 12 21:30:05 CET 2005
796 root 1.117 - additionally save PM_curpm might fix as-of-yet
797 root 1.121 unseen problems with regex matches being attributed to the
798 root 1.117 wrong package.
799 root 1.118 - add t/10_bugs.t, which currently checks against imho broken
800     perls that use 0.26MB of stack space per Perl_magic_get
801 root 1.121 invocation (newer linuxes) as opposed to the 0.0002MB perl
802     normally uses.
803 root 1.120 - make stacksize configurable for the ultimate debian experience.
804 root 1.117
805 root 1.116 1.51 Mon Dec 12 18:48:36 CET 2005
806     - remove debugging warn() accidentally left in Coro::Select.
807    
808 root 1.115 1.5 Tue Nov 29 13:32:44 CET 2005
809 root 1.114 - use Coro::Event inside Coro::Select to avoid spurious deadlocks.
810 root 1.115 - fix Coro::Select 'select'.
811     - strict'ify some modules.
812 root 1.112 - libcoro errornously restored the SIGUSR2 handler to SIGUSR1.
813 root 1.115 - use XSLoader in selected modules.
814     - remove some 5.6 compatibility cruft.
815 root 1.112
816 root 1.111 1.4 Tue Sep 6 00:11:05 CEST 2005
817     - libcoro did not take into account the trampoline on amd64,
818     when 'l'inux method was used.
819    
820 root 1.110 1.31 Tue Aug 30 23:31:33 CEST 2005
821     - some portability fixes/workarounds.
822    
823 root 1.107 1.3 Sat Aug 20 03:08:56 CEST 2005
824 root 1.108 - no code changes. module seems to work fine.
825 root 1.106 - improve Coro::State docs, remove reference to nonexisting
826     Coro::State::flush method.
827 root 1.107 - no longer autodetect windows, present it as an option instead.
828 root 1.106
829 root 1.105 1.2 Mon May 16 02:00:55 CEST 2005
830 root 1.100 - included libcoro.c earlier in State.xs, might improve portability.
831 root 1.101 - use a faster and possibly more stable Coro::Cont implementation.
832 root 1.102 - accept x86_64 in addition to amd64 for optimized linux-amd64 support.
833 root 1.104 - fix bugs Coro::SemaphoreSet that could cause locks never to be freed.
834 root 1.105 - fix bugs in CoroAPI.h and document it in Coro::MakeMaker.
835 root 1.100
836 root 1.99 1.11 Thu Mar 3 18:00:52 CET 2005
837     - change of contact address.
838    
839 root 1.98 1.1 Tue Feb 22 20:51:16 CET 2005
840 root 1.97 - support [l]inux method on amd64.
841 root 1.109 - allow some context switches while compiling/eval'ing.
842 root 1.97
843 root 1.96 1.0 Tue Aug 10 03:47:27 CEST 2004
844 root 1.95 - Event 0.89 fixes the bug again.
845 root 1.96 - fix the remaining(??) memleak. I didn't free the cache-AV for
846     padlists, so one AV per created anon-sub leaked.
847 root 1.95
848 pcg 1.94 0.97 Fri May 14 15:23:32 CEST 2004
849     - remove stupid gcc extension that was an experiment only anyway.
850    
851 pcg 1.93 0.96 Thu May 13 18:09:29 CEST 2004
852 root 1.103 - tell people if they have yet another broken Event variant.
853 pcg 1.91 - made Coro to work _better_ with threads (it seems that, as long
854     as you use Coros in one thread only everything will work fine.
855     Looking at the perl source the problem might be missing locking
856     between Coro and some perl internal routines. That's not easily
857     fixable).
858 pcg 1.92 - (hopefully) got rid of the static xnext inside transfer.
859 pcg 1.91
860 pcg 1.90 0.95 Sun Feb 15 03:20:28 CET 2004
861 pcg 1.86 - removed Coro::State::flush, implemented a new and much
862     faster caching mechanism by attaching magic to the CV.
863 pcg 1.87 - a memleak with (real) closures remains.
864 pcg 1.88 - dramatically reduced size of stacks to 65536 (linux/x86), after
865     observing that even myhttpd never had more than a kilobyte
866     of stack in use (stack is still only physically allocated in
867     pagesize increments for systems suporting mmap).
868 pcg 1.89 - fix for a "cygwin-64" architecture added, proposed by Stefan Traby.
869 pcg 1.90 - fix for threaded perl.
870 pcg 1.86
871 pcg 1.85 0.9 Sun Nov 23 04:32:39 CET 2003
872 pcg 1.84 - fix a few very hard to track down but (of course) stupid
873     memory leaks.
874     - stringify version number to avoid locale problems :(.
875 pcg 1.85 - perl 5.9.0 does not have SvPADBUSY, reported by Scott Walters,
876     hopefully ignoring it "just works".
877 pcg 1.84
878 pcg 1.83 0.8 Wed Nov 5 19:38:40 CET 2003
879     - port to perl5.005_03, but only 5.8.x+ is supported!
880 pcg 1.82 - honor LocalAddr even without LocalPort in Coro::Socket.
881    
882 root 1.81 0.7 Tue May 27 03:12:38 CEST 2003
883     - the version jump indicates some level of testing, not gobs
884     of new features.
885 root 1.80 - uh... I found the prompt function in ExtUtils::MakeMaker.
886     Highly correct stuff, that is...
887 root 1.81 - fixed(?) a bug with die's in coroutines causing "panic: top_env".
888     the fix is not well understood by the author, so beware :(.
889 root 1.80
890 root 1.79 0.652 Thu May 8 02:54:46 CEST 2003
891 root 1.76 - Applied patch by Slaven Rezic to set default to "s" on FreeBSD,
892 root 1.78 cause version 4 doesn't have ucontext.
893 root 1.77 - Benjamin Reed reported that setjmp works fine on darwin,
894     so preselect it.
895 root 1.76
896 root 1.75 0.651 Sat Mar 29 15:00:23 CET 2003
897     - fix a bug in Coro::Handle where some bytes could get lost
898     on reading, reported by jason@nichego.net.
899 root 1.74
900 root 1.73 0.65 Sun Mar 23 00:08:26 CET 2003
901 root 1.72 - added a README.linux-glibc.
902     - new module Coro::Select.
903     - also save/restore PL_comppad, fixes segfaults with 5.8.1.
904     (I never claimed that I know what I am doing ;).
905 root 1.73 - set default to setjmp/longjmp on non-x86-linux.
906 root 1.72
907 root 1.71 0.6 Thu Nov 21 11:09:06 CET 2002
908     - made lazy context switching the default.
909     - fixed the bug where SemaphoreSet::guard
910     was the same as timed_guard and timed_guard was missing.
911     - fix a memleak in Coro::Timer.
912     - reclassified context sharing as not experimental.
913    
914 root 1.70 0.534 Sun Apr 14 03:05:12 CEST 2002
915 root 1.69 - fixed backspace => tab in header parsing (myhttpd).
916 root 1.70 - added eg/readline, for elmex.
917     - Coro::Event now calls ->start on first call instead of
918     ->again, which created wrong timeouts for the first ->next.
919     - fixed a bug where Coro::Socket returned a closed socket instead
920     of nothing on ECONNREFUSED and similar error conditions.
921 root 1.69
922 root 1.68 0.533 Mon Feb 18 18:49:40 CET 2002
923     - adapt to finally working Event-0.85.
924     - fixed buggy event initialization (trapped under windows, but how
925     could it possibly have worked anywhere?)
926    
927 root 1.67 0.532 Wed Jan 16 02:45:32 CET 2002
928 root 1.66 - removed the reference to pp_entersub, might work on
929     (some) win32 perls now (testsuite works under cygwin,
930     without Event, which freezes).
931 root 1.65
932 root 1.64 0.531 Mon Dec 10 22:18:44 CET 2001
933 root 1.63 - Fixed a bug in SemaphoreSet::timed_down.
934    
935 root 1.64 0.53 Tue Nov 27 21:11:13 CET 2001
936 root 1.60 - some tests for Event.
937 root 1.61 - slightly different internal architecture to get refcnt into the object
938     and not the reference. might fix a few bugs, certainly did introduce
939     new ones.
940     - Coro::Timer (independent of Event).
941 root 1.62 - new timed_wait functions for Coro::Signal, Semaphore, SemaphoreSet.
942 root 1.60
943 root 1.64 0.52 Tue Nov 6 21:36:18 CET 2001
944 root 1.58 - ported to cygwin (trivial).
945     patch by Gerrit P. Haase <gp@familiehaase.de>.
946     - small setjmp code fix by Sullivan.DanielJ@epamail.epa.gov.
947    
948 root 1.64 0.51 Thu Nov 1 20:39:01 CET 2001
949 root 1.54 - terminate/cancel now work properly (otherwise termination
950     could cause a "next coroutine is not and contains not..."
951     error.
952 root 1.55 - added Coro::Socket::shutdown.
953 root 1.56 - Coro::Event::loop no is the same as Event::loop.
954     - implemented terminate with args + join.
955 root 1.54
956 root 1.64 0.5 Fri Sep 28 16:15:35 CEST 2001
957 root 1.53 - fixed "print" on a Coro::Handle. The print method worked.
958 root 1.50 - small tweaks (seem to reduce memory consumption a lot)
959     in various modules.
960     - splendid use of "no warnings" scattered throughout.
961 root 1.52 - added Coro::Handle::rbuf, fixed Coro::Handle::fh.
962 root 1.48
963 root 1.47 0.49 Sun Sep 16 02:42:45 CEST 2001
964 root 1.43 - changed some method calls to function calls for speed
965     inside Coro::Handle.
966 root 1.44 - make Coro::Handle use an array instead of a hash for
967     speed reasons.
968     - IRIX mystery solved: it's SGI's NT, after all: "standard, huh?".
969     sjlj and ucontext should now work.
970     - IRIX-specific port for libcoro.
971 root 1.45 - swapped order of accept results to match IO::Socket.
972 root 1.46 - changed getsock/peername to sock/peername to match IO::Socket.
973     - fixed a bug that caused segfault when returning to main task
974     under some circumstances.
975     - other bugfixes.
976 root 1.43
977 root 1.42 0.45 Sun Sep 2 02:54:01 CEST 2001
978 root 1.36 - new method Coro::Handle::timeout.
979     - corrected speling of set...name to get...name in Coro::Socket.
980     - Coro::Socket::accept now returns a Coro::Socket, not a Coro::Handle.
981     - Coro::Handle now supports fileno and FILENO.
982     - added eg/myhttpd, a web server.
983 root 1.37 - Coro::Socket now accepts numeric ports as well as "port(number)"
984     syntax.
985 root 1.38 - moved some scheduling primitives into xs code. more to come.
986 root 1.39 - new simple priority system.
987 root 1.40 - implemented Coro::Event using XS, almost four times faster!
988 root 1.41 - small memory corruption problem fixed (boy that was difficult).
989 root 1.36
990 root 1.35 0.13 Wed Aug 8 16:53:07 CEST 2001
991 root 1.31 - new method Coro::Handle::readline.
992 root 1.34 - added eg/lwp and Coro::Handle::autoflush, to make LWP work.
993     (see eg/lwp on how to make LWP non-blocking).
994 root 1.31 - renamed Coro::Socket::new_inet to new.
995 root 1.32 - added Coro::Util, some utility functions.
996 root 1.33 - die/eval now works better (still get annoying (but true ;)
997     "Callback called exit" messages).
998 root 1.34 - Coro::Handle now supports timeout, and Coro::Socket Timeout.
999     - much better Coro::Socket support.
1000 root 1.31
1001     0.12 Fri Jul 27 04:19:01 CEST 2001
1002 root 1.29 - do not use mmap if neither MAP_ANON nor MAP_ANONYMOUS is defined.
1003 root 1.34 - very experimental stack sharing algorithm. not 100% safe but
1004 root 1.28 should work well in practise ;)
1005 root 1.30 - added Coro::RWLock.
1006 root 1.31 - Coro::Specific now works.
1007 root 1.28
1008     0.11 Tue Jul 24 22:49:21 CEST 2001
1009 root 1.27 - added specialized hack for newer and older linux versions (fast).
1010     - renamed Coro::Event::IO to Coro::Handle.
1011     - new module Coro::Socket.
1012    
1013 root 1.26 0.1 Tue Jul 24 01:47:53 CEST 2001
1014 root 1.23 - release candidate 3. A rather new internal structure :(
1015 root 1.24 - the great renaming: Damian Conway gave me a suitable replacement
1016     for yield.
1017 root 1.23 - added Coro/libcoro, a portable coroutine implementation for C
1018 root 1.22 (not even perl-dependent!!!), which can be used by Coro::State.
1019 root 1.20 - renamed SAVE_DEFSV (on xs-level only) to avoid symbol clash in
1020     perl-5.6.
1021 root 1.16 - new function Coro::Event::idle.
1022     - the idle process is now overriden
1023     by default in Coro::Event.
1024 root 1.17 - Coro::Channel now enforces the size.
1025 root 1.18 - canceling events now works in all cases.
1026 root 1.19 - Coro state now include $_ and $@.
1027     - yet another bug workaround that I do not really understand :(
1028 root 1.21 - new module Coro::Event::IO (very undocumented) to do
1029     non-blocking i/o.
1030 root 1.25 - performance tuning.
1031 root 1.16
1032 root 1.15 0.08 Thu Jul 19 06:13:25 CEST 2001
1033 root 1.14 - release candidate 2 (new functionality).
1034 root 1.13 - add Coro::State::flush function.
1035     - transfer now supports a flags value (mostly for speed ATM).
1036     - might compile and work in the presence of threads now.
1037 root 1.14 - continuations have a saner syntax.
1038 root 1.19 - no more memleaks.
1039 root 1.13
1040 root 1.12 0.07 Tue Jul 17 17:40:18 CEST 2001
1041     - release candidate 1 ;)
1042     - slightly nicer code.
1043     - fixed a scheduling bug in Coro::Event.
1044    
1045 root 1.11 0.06 Tue Jul 17 04:23:24 CEST 2001
1046     - ok, I found the showstopper - the same sub must not be
1047 root 1.9 re-entered in two different coroutines, otherwise => crash. I
1048     see no easy solution to this problem, except by walking the call
1049     chaing and saving/restoring the cv's, which is what I do now.
1050 root 1.11 - memory leaks still latent, especially at thread termination.
1051     - Coro::Event now works (haha).
1052 root 1.9
1053 root 1.8 0.05 Sun Jul 15 17:32:20 CEST 2001
1054     - fixed a few issues in Coro::Event.
1055     - I forgot to include Coro::Event in 0.04 :(:(:(
1056    
1057 root 1.7 0.04 Sun Jul 15 05:24:59 CEST 2001
1058 root 1.5 - @_ is now properly localized.
1059 root 1.4 - Coro::State is now easier subclassable.
1060     - Coro::Cont now coroutine-aware.
1061     - Coro::Specific is a low-overhead module to create
1062     coroutine-specific vars.
1063 root 1.6 - Coro::Event provides a simple interface to Event.
1064 root 1.4
1065 root 1.3 0.03 Fri Jul 13 14:51:52 CEST 2001
1066 root 1.4 - transfer() now implemented in XS (beware).
1067     - new module Coro::Cont for really faked continuations.
1068     - big internal architecture changes: Coro::State is now
1069 root 1.3 really low-level and can thus be used to implement other
1070 root 1.4 interesting things, While "Coro::" implements a process-like
1071     model. Still crude and subject to change.
1072 root 1.3 - $_ and $@ are no longer being localized.
1073    
1074 root 1.2 0.02 Tue Jul 10 01:38:17 CEST 2001
1075     - implemented "async" attribute.
1076     - $_ and $@ are now localized.
1077     - added Coro::Channel.
1078     - more testcases, still no docs.
1079    
1080 root 1.1 0.01 Tue Jul 3 02:18:41 CEST 2001
1081     - original version; copied from Convert::Scalar.
1082