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