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

Comparing Coro/Changes (file contents):
Revision 1.444 by root, Wed Nov 19 16:01:32 2008 UTC vs.
Revision 1.537 by root, Tue May 4 21:51:27 2010 UTC

1Revision history for Perl extension Coro. 1Revision history for Perl extension Coro.
2 2
3TODO: should explore PerlIO::coroaio (perl leaks like hell). 3TODO: should explore PerlIO::coroaio (perl leaks like hell).
4TODO: maybe implement a default message channel, very much like Erlang's 4TODO: unready_all
5 actor model (which is cool in a lot of important aspects (failures!), 5TODO: myhttpd header parsing
6 but very lacking in others (higher level ipc)). 6TODO: channel->maxsize(newsize)?
7TODO: http://www.microsoft.com/msj/archive/s2ce.aspx
7 8
8TODO: async_pool to SLF 9 - document that perl 5.12 deliberately removed support for cloning.
9TODO: _cancel to xs 10
105.0 115.22 Wed Apr 14 03:55:35 CEST 2010
12 - correctly return udnef on errors in Coro::Handle::read/write
13 (testcase by Marc Mims).
14 - convert Coro::Util into a "perl compatibility wrapper" - the functions
15 are less useful now, but are drop-in replacements for existing
16 functions, listing better alternatives in the documentation. This also
17 fixes a bug in Coro::LWP which naively substituted Socket::inet_aton
18 with Coro::Util::inet_aton.
19 - do not override $Coro::idle unconditionally in Coro.pm, as other
20 modules could have provided their own idle coro already
21 (for exmaple, Coro::AnyEvent).
22 - fix Coro::Util::gethost* functions.
23 - Coro::Timer corretcly exports it's symbols (reported by Hideki Yamamura).
24
255.21 Wed Dec 16 07:19:51 CET 2009
26 - automatically load Coro::AnyEvent when AnyEvent and Coro are used
27 together.
28 - add some examples on how to combine other event loops with Coro in
29 Coro::AnyEvent, and how to run it (and not to block). Seems to be
30 the most common source of confusion.
31 - try to catch people naively blocking in an event callback.
32 - work around the perl filehandle bug issue in conjunction with
33 older common::sense (as indirectly pointed out by ZSystem).
34 - clarify the "not from signal handlers" section.
35
365.2 Sun Oct 4 14:54:24 CEST 2009
37 - Coro::Storable destroyed the prototypes of the functions it wrapped.
38 - export rouse_cb and rouse_wait by default now.
39 - fix various prototype mismatches in Coro::AnyEvent and Coro::Handle.
40 - new method $state->swap_sv.
41 - added section on "windows process emulation" to the manpage,
42 after a not-so-fruitful (nor-friendly) "discussion" with chip
43 salzenberg (discussion implies arguments, but his only arguments
44 were ad-hominems, one wonders why he started it in the first
45 place). I hope this explains it well enough for him to understand,
46 and maybe well enough for others to understand.
47 - use common::sense everywhere now.
48 - idle callbacks are no longer supported, use idle coros instead.
49 - print a thread listing when a deadlock is detected.
50
515.17 Sat Aug 22 23:09:31 CEST 2009
52 - work around a bug in the perl debugger causing crashes
53 when running under the debugger by marking _pool_handler as nodebug.
54 - speed up Coro::async considerably.
55 - try some hacks to get netbsd to work "more often" - their broken
56 setjmp/longjmp, ucontext *and* phtreads are really hard on Coro.
57 - convert Coro to AE 5.0 API.
58
595.162 Tue Jul 28 04:04:03 CEST 2009
60 - perl 5.8.2 is now minimum requirement.
61 - skip t/19_handle.t on broken windows perls.
62
635.161 Wed Jul 22 04:47:38 CEST 2009
64 - Coro::AnyEvent::poll could have a different prototype when EV was
65 used as backend (analyzed by Tatsuhiko Miyagawa).
66 - Coro::AnyEvent errornously initialised the event loop when loaded,
67 not on demand.
68 - try to workaround rare */t/01_unblock.t failures.
69
705.16 Tue Jul 21 01:44:37 CEST 2009
71 - Coro::AnyEvent failed to hook into the event loop
72 when no threads had been readied between detecting
73 the event loop and actually running it.
74 - considerably speed up Coro::Select by taking avdantage
75 of AnyEvent > 4.8 and some other optimisations.
76 - implement paragraph readline mode in Coro::Handle
77 (based on patches by Zsbán Ambrus).
78 - replace WSAEINPROGRESS by WSAEWOULDBLOCK (reported
79 and analyzed by Yasuhiro MATSUMOTO).
80 - clarified libcoro license and copyright.
81 - someone stole my EXTRA_META!!!
82 - implement Coro::Select::patch_pp_sselect and it's brother,
83 for hardcode select overriding.
84
855.151 Mon Jul 6 05:41:57 CEST 2009
86 - backport to windows process emulation code again (patch by
87 Yasuhiro MATSUMOTO).
88 - slightly update Coro::MakeMaker.
89
905.15 Tue Jun 30 10:28:06 CEST 2009
91 - deprecate Coro::Socket, document how to get ipv6 support via
92 AnyEvent::Socket instead.
93 - implement signal->wait ($cb) interface, similar to semaphores.
94 - work around SvOK not supporting getmagic, so we have to getmagic
95 to test for undef :/ (reported by Matthias Waldorf).
96 - load Coro::AnyEvent in all modules using AnyEvent.
97 - work around perl corrupting our internal data structures,
98 reported by Tokuhiro Matsuno.
99 - enable per-coroutine real and cpu time gathering
100 (Coro::State::enable_times).
101
1025.14 Wed Jun 24 01:37:48 CEST 2009
103 - provide explicit functions to "cede" to the event loop in Coro::AnyEvent,
104 as this seems to have been a difficult concept (poll, sleep, idle,
105 idle_upto).
106 - add Coro::AnyEvent::readable/writable functions.
107 - clarify Coro::EV/Event/AnyEvent manpages.
108 - free per-thread global scalars in the thread calling ->cancel, to
109 avoid crashes when $_, $@ etc., are magical but some of those
110 had already been freed.
111 - "unexperimentalise" the callback interface for Coro::Semaphore.
112 - speed up ready queue management/context switching by using a linked
113 list instead of an array (~5-10%).
114 - implement "watch" command in Coro::Debug shells.
115 - for fun, implement time-slicing as an example in the manpage.
116 - if AnyEvent detects EV or Event, but we don't have Coro::EV or
117 Coro::Event, use the normal AnyEvent handling instead of dieing
118 (the same is true for Coro::Handle).
119 - properly document Coro::EV::timed_io_once.
120 - avoid unneccessary ->cancel calls in Coro::Handle.
121 - maybe make it work on mingw32 with win32 backend
122 (based on patch by Yasuhiro Matsumoto).
123
1245.132 Fri May 29 09:00:39 CEST 2009
125 - do not keep a reference to the argument itself in
126 Coro::Semaphore::guard, as it could change later.
127 - support SO_RCVBUF/SO_SNDBUF nonstandard Coro::Socket options,
128 should support a prepare callback.
129
1305.131 Mon Mar 16 23:20:37 CET 2009
131 - implement and document Coro->suspend, Coro->resume.
132 - fix Coro::Select implementation to not (often) close
133 the passed file descriptors (testcase provided by pippijn).
134
1355.13 Mon Dec 15 21:51:42 CET 2008
136 - EXPERIMENTAL: implement dynamic winds (on_enter/on_leave).
137 - don't set diehook to C<undef> but instead to NULL, to avoid
138 spurious warnings.
139 - fix a lot of bugs in Coro::SemaphoreSet.
140 - Coro::SemaphoreSet will less often create a semaphore needlessly.
141 - add Coro::SemaphoreSet::count and wait methods.
142 - take advantage of the new Guard module.
143 - deprecate Coro::guard.
144 - try to fix the dreaded 01_unblock tests once more. I hate it when
145 testsuites need more fixing than the code they are supposed to test.
146 - croak in more cases when a required callback isn't resolvable.
147 - fix some minor issues in Coro::State->call/eval.
148 - use current coroutine context instead of a temporary one
149 when temporarily switching to another coroutine.
150 - do not call C-level on_destroy handlers during global destruction,
151 to avoid needless segfaults.
152
1535.12 Sun Dec 7 13:30:38 CET 2008
154 - add default config for MirOS, which seems to be a bug-to-bug
155 compatible fork of openbsd ("world domination by releasing
156 openbsd cvs before the openbsd folks do it" or so :).
157 - free_padlist did destroy the names pad, not good, but didn't
158 seem to bother perl - this could fix issues such as eval ""
159 inside a function called from multiple coroutines.
160 - use a different method to detect destruction time.
161 - be more careful when freeing padlists just before global
162 destruction.
163 - fixed and expanded the call/cc example.
164 - renamed _terminate to _coro_run.
165 - new method Coro::Channel->shutdown.
166 - try pthreads on openbsd <4.4 (broken sigaltstack, will
167 pthreads fare better?).
168 - be less picky about destroying "a" running coroutine.
169
1705.11 Tue Nov 25 21:49:05 CET 2008
171 - DEBUGGING in 5.10.0 is a mess: it constantly flags perfectly
172 working code with failed assertions, introducing more bugs than
173 it fixes, requiring elaborate workarounds :(
174
1755.1 Mon Nov 24 08:54:59 CET 2008
176 - wrote a small introductory tutorial - Coro::Intro.
177 - convert Coro::Timer, Coro::Select and Coro::Util to rouse API.
178 - Coro::Select did errornously dup the file descriptors
179 and didn't work with all AnyEvent backends.
180 - Coro::Select wasn't imported correctly form Coro::LWP, causing blocking
181 LWP data transfers.
182 - disassociate c contexts from coro objects - this is agruably more
183 correct, but mostly allows sharing of cctxs between coro and state
184 objects, for added memory savings and speed increases.
185 - bumped $Coro::POOL_RSS up to 32kb by default.
186 - no longer set the optype to OP_CUSTOM, as B::* understandably
187 doesn't like this very much (and we *are* a type of entersub).
188 - implement state cloning, just to prove that call/cc can be done.
189 - automatically load Coro::AnyEvent in Coro::Handle.
190 - wrap ->cancel calls in eval inside Coro::Handle as EV watchers
191 do not have this method (and don't need it either).
192 - speed up generic anyevent methods in Coro::Handle by using rouse
193 callbacks.
194 - allow coroutines in $Coro::IDLE, speeding up Coro::AnyEvent and
195 others. It also makes the debugger happier, as you can trace
196 through the idle threads now.
197 - add comppad_name* and hints ($^H, %^H) to per-thread variables.
198 - eg/event was pretty broken.
199 - better 5.8.6 compatibility.
200
2015.0 Thu Nov 20 10:35:05 CET 2008
11 - NEW ARCHITECTURE: use the latest 4.x version if you experience 202 - NEW ARCHITECTURE: use the latest 4.x version if you experience
12 stability issues. 203 stability issues.
13 - bump API version to 7 - all dependents must be recompiled. 204 - bump API version to 7 - all dependents must be recompiled.
14 - removed timed_* functions - they were not being used anyways 205 - removed timed_* functions - they were not being used anyways
15 and should be replaced by a more generic mechanism - 206 and should be replaced by a more generic mechanism -
16 and were annoying to support anyways :) 207 and were annoying to support anyways :)
17 - removed SemaphoreSet's waiter method - use sem instead. 208 - removed SemaphoreSet's waiter method - use sem method instead.
18 - Coro::Semaphore->adjust didn't correctly wake up enough waiters. 209 - Coro::Semaphore->adjust didn't correctly wake up enough waiters.
19 - async_pool did free a scalar value twice 210 - async_pool did free a scalar value twice
20 ("Attempt to unreference..."). 211 ("Attempt to unreference...").
21 - fix a disastrous bug in the readline optimisation
22 introduced in 4.801.
23 - fix a longstanding bug where calling terminate on a coro that 212 - fix a longstanding bug where calling terminate on a coro that
24 was waiting for a semaphore that was just becoming available 213 was waiting for a semaphore that was just becoming available
25 would cause a deadlock (semaphore would get into a state where 214 would cause a deadlock (semaphore would get into a state where
26 it was available but waiters were still blocked). 215 it was available but waiters were still blocked).
27 - calling throw on a coroutine that is waiting for a semaphore will 216 - calling throw on a coroutine that is waiting for a semaphore will
28 no longer make it acquire the semaphore (and thus leak a count). 217 no longer make it acquire the semaphore (and thus leak a count).
29 - perl's process emulation is now not even theoretically supported 218 - perl's process emulation is now not even theoretically supported
30 anymore. 219 anymore.
31 - new functions Coro::rouse_cb and Coro::rouse_wait. 220 - new functions Coro::rouse_cb and Coro::rouse_wait for easier
221 conversion of callback-style to blocking-style.
222 - new methods $coro->schedule_to and ->cede_to, to specifically
223 schedule or cede to a specific coroutine.
32 - new function Coro::Semaphore::wait. 224 - new function Coro::Semaphore::wait.
33 - use named constants in Coro::Channel (Richard Hundt). 225 - use named constants in Coro::Channel (Richard Hundt).
34 - directly patch the entersub opcode calling SLF functions (cede, 226 - directly patch the entersub opcode calling SLF functions (cede,
35 transfer and so on). this does speed up context switching, but 227 transfer and so on). this does speed up context switching, but
36 more importanly, it frees us from the hardcoded behaviour of 228 more importanly, it frees us from the hardcoded behaviour of
… …
53 - implement Coro::Channel in terms of Coro::Semaphore, for a moderate 245 - implement Coro::Channel in terms of Coro::Semaphore, for a moderate
54 (in comparison) 20-40% speedup. 246 (in comparison) 20-40% speedup.
55 - used new SLF interface to reimplement Coro::Signal gaining 247 - used new SLF interface to reimplement Coro::Signal gaining
56 some unknown (because I was too lazy), but certain, speedup, and also 248 some unknown (because I was too lazy), but certain, speedup, and also
57 making signals reliable for the first time. 249 making signals reliable for the first time.
250 - used new SLF interface and other optimisations to speed up async_pool
251 by a factor of two. It also doesn't rely on perl's exception mechanism
252 to exit anymore. The overhead for terminating an async_pool, coro over
253 a normal async is now very small.
58 - sped up coroutine creation/destruction by 40%. 254 - sped up coroutine creation/destruction by 40%.
59 - forgot to include Coro/libcoro/README in the dist for all these years. 255 - forgot to include Coro/libcoro/README in the dist for all these years.
60 - work around a freebsd pthreads bug (manual testcancel is required as 256 - work around a freebsd pthreads bug (manual testcancel is required as
61 pthread_cond_wait isn't a cancellation point on freebsd). 257 pthread_cond_wait isn't a cancellation point on freebsd).
62 - use new rouse functions to speed up and simplify Coro::BDB. 258 - use new rouse functions to speed up and simplify Coro::BDB.
63 - make "prefer perl native functions" work with threaded perls. 259 - make "prefer perl native functions" work with threaded perls.
64 - condense Coro::Debug ps output, hint at v and w flags. 260 - condense Coro::Debug ps output, hint at v and w flags.
65 - (libcoro) lots of minor cleanups and portability improvements. 261 - (libcoro) lots of minor cleanups and portability improvements.
262
2634.914 Wed Nov 19 12:54:18 CET 2008
264 - fix a disastrous bug in the readline optimisation
265 introduced in 4.801.
66 266
674.913 Sat Nov 15 07:58:28 CET 2008 2674.913 Sat Nov 15 07:58:28 CET 2008
68 - async_pool did free a scalar value twice 268 - async_pool did free a scalar value twice
69 ("Attempt to unreference..."). 269 ("Attempt to unreference...").
70 270

Diff Legend

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