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

Comparing Coro/Changes (file contents):
Revision 1.446 by root, Thu Nov 20 03:11:29 2008 UTC vs.
Revision 1.522 by root, Wed Aug 26 07:42:25 2009 UTC

… …
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: maybe implement a default message channel, very much like Erlang's
5 actor model (which is cool in a lot of important aspects (failures!), 5 actor model (which is cool in a lot of important aspects (failures!),
6 but very lacking in others (higher level ipc)). 6 but very lacking in others (higher level ipc)).
7TODO: unready_all
8TODO: myhttpd header parsing
9TODO: channel->maxsize(newsize)?
7 10
8TODO: async_pool to SLF 11 - added section on "windows process emulation" to the manpage,
9TODO: _cancel to xs 12 after a not-so-fruitful (nor-friendly) "discussion" with chip
105.0 13 salzenberg (discussion implies arguments, but his only arguments
14 were ad-hominems, one wonders why he started it in the first
15 place).
16
175.17 Sat Aug 22 23:09:31 CEST 2009
18 - work around a bug in the perl debugger causing crashes
19 when running under the debugger by marking _pool_handler as nodebug.
20 - speed up Coro::async considerably.
21 - try some hacks to get netbsd to work "more often" - their broken
22 setjmp/longjmp, ucontext *and* phtreads are really hard on Coro.
23 - convert Coro to AE 5.0 API.
24
255.162 Tue Jul 28 04:04:03 CEST 2009
26 - perl 5.8.2 is now minimum requirement.
27 - skip t/19_handle.t on broken windows perls.
28
295.161 Wed Jul 22 04:47:38 CEST 2009
30 - Coro::AnyEvent::poll could have a different prototype when EV was
31 used as backend (analyzed by Tatsuhiko Miyagawa).
32 - Coro::AnyEvent errornously initialised the event loop when loaded,
33 not on demand.
34 - try to workaround rare */t/01_unblock.t failures.
35
365.16 Tue Jul 21 01:44:37 CEST 2009
37 - Coro::AnyEvent failed to hook into the event loop
38 when no threads had been readied between detecting
39 the event loop and actually running it.
40 - considerably speed up Coro::Select by taking avdantage
41 of AnyEvent > 4.8 and some other optimisations.
42 - implement paragraph readline mode in Coro::Handle
43 (based on patches by Zsbán Ambrus).
44 - replace WSAEINPROGRESS by WSAEWOULDBLOCK (reported
45 and analyzed by Yasuhiro MATSUMOTO).
46 - clarified libcoro license and copyright.
47 - someone stole my EXTRA_META!!!
48 - implement Coro::Select::patch_pp_sselect and it's brother,
49 for hardcode select overriding.
50
515.151 Mon Jul 6 05:41:57 CEST 2009
52 - backport to windows process emulation code again (patch by
53 Yasuhiro MATSUMOTO).
54 - slightly update Coro::MakeMaker.
55
565.15 Tue Jun 30 10:28:06 CEST 2009
57 - deprecate Coro::Socket, document how to get ipv6 support via
58 AnyEvent::Socket instead.
59 - implement signal->wait ($cb) interface, similar to semaphores.
60 - work around SvOK not supporting getmagic, so we have to getmagic
61 to test for undef :/ (reported by Matthias Waldorf).
62 - load Coro::AnyEvent in all modules using AnyEvent.
63 - work around perl corrupting our internal data structures,
64 reported by Tokuhiro Matsuno.
65 - enable per-coroutine real and cpu time gathering
66 (Coro::State::enable_times).
67
685.14 Wed Jun 24 01:37:48 CEST 2009
69 - provide explicit functions to "cede" to the event loop in Coro::AnyEvent,
70 as this seems to have been a difficult concept (poll, sleep, idle,
71 idle_upto).
72 - add Coro::AnyEvent::readable/writable functions.
73 - clarify Coro::EV/Event/AnyEvent manpages.
74 - free per-thread global scalars in the thread calling ->cancel, to
75 avoid crashes when $_, $@ etc., are magical but some of those
76 had already been freed.
77 - "unexperimentalise" the callback interface for Coro::Semaphore.
78 - speed up ready queue management/context switching by using a linked
79 list instead of an array (~5-10%).
80 - implement "watch" command in Coro::Debug shells.
81 - for fun, implement time-slicing as an example in the manpage.
82 - if AnyEvent detects EV or Event, but we don't have Coro::EV or
83 Coro::Event, use the normal AnyEvent handling instead of dieing
84 (the same is true for Coro::Handle).
85 - properly document Coro::EV::timed_io_once.
86 - avoid unneccessary ->cancel calls in Coro::Handle.
87 - maybe make it work on mingw32 with win32 backend
88 (based on patch by Yasuhiro Matsumoto).
89
905.132 Fri May 29 09:00:39 CEST 2009
91 - do not keep a reference to the argument itself in
92 Coro::Semaphore::guard, as it could change later.
93 - support SO_RCVBUF/SO_SNDBUF nonstandard Coro::Socket options,
94 should support a prepare callback.
95
965.131 Mon Mar 16 23:20:37 CET 2009
97 - implement and document Coro->suspend, Coro->resume.
98 - fix Coro::Select implementation to not (often) close
99 the passed file descriptors (testcase provided by pippijn).
100
1015.13 Mon Dec 15 21:51:42 CET 2008
102 - EXPERIMENTAL: implement dynamic winds (on_enter/on_leave).
103 - don't set diehook to C<undef> but instead to NULL, to avoid
104 spurious warnings.
105 - fix a lot of bugs in Coro::SemaphoreSet.
106 - Coro::SemaphoreSet will less often create a semaphore needlessly.
107 - add Coro::SemaphoreSet::count and wait methods.
108 - take advantage of the new Guard module.
109 - deprecate Coro::guard.
110 - try to fix the dreaded 01_unblock tests once more. I hate it when
111 testsuites need more fixing than the code they are supposed to test.
112 - croak in more cases when a required callback isn't resolvable.
113 - fix some minor issues in Coro::State->call/eval.
114 - use current coroutine context instead of a temporary one
115 when temporarily switching to another coroutine.
116 - do not call C-level on_destroy handlers during global destruction,
117 to avoid needless segfaults.
118
1195.12 Sun Dec 7 13:30:38 CET 2008
120 - add default config for MirOS, which seems to be a bug-to-bug
121 compatible fork of openbsd ("world domination by releasing
122 openbsd cvs before the openbsd folks do it" or so :).
123 - free_padlist did destroy the names pad, not good, but didn't
124 seem to bother perl - this could fix issues such as eval ""
125 inside a function called from multiple coroutines.
126 - use a different method to detect destruction time.
127 - be more careful when freeing padlists just before global
128 destruction.
129 - fixed and expanded the call/cc example.
130 - renamed _terminate to _coro_run.
131 - new method Coro::Channel->shutdown.
132 - try pthreads on openbsd <4.4 (broken sigaltstack, will
133 pthreads fare better?).
134 - be less picky about destroying "a" running coroutine.
135
1365.11 Tue Nov 25 21:49:05 CET 2008
137 - DEBUGGING in 5.10.0 is a mess: it constantly flags perfectly
138 working code with failed assertions, introducing more bugs than
139 it fixes, requiring elaborate workarounds :(
140
1415.1 Mon Nov 24 08:54:59 CET 2008
142 - wrote a small introductory tutorial - Coro::Intro.
143 - convert Coro::Timer, Coro::Select and Coro::Util to rouse API.
144 - Coro::Select did errornously dup the file descriptors
145 and didn't work with all AnyEvent backends.
146 - Coro::Select wasn't imported correctly form Coro::LWP, causing blocking
147 LWP data transfers.
148 - disassociate c contexts from coro objects - this is agruably more
149 correct, but mostly allows sharing of cctxs between coro and state
150 objects, for added memory savings and speed increases.
151 - bumped $Coro::POOL_RSS up to 32kb by default.
152 - no longer set the optype to OP_CUSTOM, as B::* understandably
153 doesn't like this very much (and we *are* a type of entersub).
154 - implement state cloning, just to prove that call/cc can be done.
155 - automatically load Coro::AnyEvent in Coro::Handle.
156 - wrap ->cancel calls in eval inside Coro::Handle as EV watchers
157 do not have this method (and don't need it either).
158 - speed up generic anyevent methods in Coro::Handle by using rouse
159 callbacks.
160 - allow coroutines in $Coro::IDLE, speeding up Coro::AnyEvent and
161 others. It also makes the debugger happier, as you can trace
162 through the idle threads now.
163 - add comppad_name* and hints ($^H, %^H) to per-thread variables.
164 - eg/event was pretty broken.
165 - better 5.8.6 compatibility.
166
1675.0 Thu Nov 20 10:35:05 CET 2008
11 - NEW ARCHITECTURE: use the latest 4.x version if you experience 168 - NEW ARCHITECTURE: use the latest 4.x version if you experience
12 stability issues. 169 stability issues.
13 - bump API version to 7 - all dependents must be recompiled. 170 - bump API version to 7 - all dependents must be recompiled.
14 - removed timed_* functions - they were not being used anyways 171 - removed timed_* functions - they were not being used anyways
15 and should be replaced by a more generic mechanism - 172 and should be replaced by a more generic mechanism -
16 and were annoying to support anyways :) 173 and were annoying to support anyways :)
17 - removed SemaphoreSet's waiter method - use sem instead. 174 - removed SemaphoreSet's waiter method - use sem method instead.
18 - Coro::Semaphore->adjust didn't correctly wake up enough waiters. 175 - Coro::Semaphore->adjust didn't correctly wake up enough waiters.
19 - async_pool did free a scalar value twice 176 - async_pool did free a scalar value twice
20 ("Attempt to unreference..."). 177 ("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 178 - fix a longstanding bug where calling terminate on a coro that
24 was waiting for a semaphore that was just becoming available 179 was waiting for a semaphore that was just becoming available
25 would cause a deadlock (semaphore would get into a state where 180 would cause a deadlock (semaphore would get into a state where
26 it was available but waiters were still blocked). 181 it was available but waiters were still blocked).
27 - calling throw on a coroutine that is waiting for a semaphore will 182 - 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). 183 no longer make it acquire the semaphore (and thus leak a count).
29 - perl's process emulation is now not even theoretically supported 184 - perl's process emulation is now not even theoretically supported
30 anymore. 185 anymore.
31 - new functions Coro::rouse_cb and Coro::rouse_wait. 186 - new functions Coro::rouse_cb and Coro::rouse_wait for easier
187 conversion of callback-style to blocking-style.
188 - new methods $coro->schedule_to and ->cede_to, to specifically
189 schedule or cede to a specific coroutine.
32 - new function Coro::Semaphore::wait. 190 - new function Coro::Semaphore::wait.
33 - use named constants in Coro::Channel (Richard Hundt). 191 - use named constants in Coro::Channel (Richard Hundt).
34 - directly patch the entersub opcode calling SLF functions (cede, 192 - directly patch the entersub opcode calling SLF functions (cede,
35 transfer and so on). this does speed up context switching, but 193 transfer and so on). this does speed up context switching, but
36 more importanly, it frees us from the hardcoded behaviour of 194 more importanly, it frees us from the hardcoded behaviour of
… …
66 - use new rouse functions to speed up and simplify Coro::BDB. 224 - use new rouse functions to speed up and simplify Coro::BDB.
67 - make "prefer perl native functions" work with threaded perls. 225 - make "prefer perl native functions" work with threaded perls.
68 - condense Coro::Debug ps output, hint at v and w flags. 226 - condense Coro::Debug ps output, hint at v and w flags.
69 - (libcoro) lots of minor cleanups and portability improvements. 227 - (libcoro) lots of minor cleanups and portability improvements.
70 228
2294.914 Wed Nov 19 12:54:18 CET 2008
230 - fix a disastrous bug in the readline optimisation
231 introduced in 4.801.
232
714.913 Sat Nov 15 07:58:28 CET 2008 2334.913 Sat Nov 15 07:58:28 CET 2008
72 - async_pool did free a scalar value twice 234 - async_pool did free a scalar value twice
73 ("Attempt to unreference..."). 235 ("Attempt to unreference...").
74 236
754.912 Thu Nov 13 18:31:23 CET 2008 2374.912 Thu Nov 13 18:31:23 CET 2008

Diff Legend

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