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

Comparing Coro/Changes (file contents):
Revision 1.434 by root, Wed Nov 19 00:06:54 2008 UTC vs.
Revision 1.502 by root, Sat Jun 27 14:26:09 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: testcancel 11 - deprecate Coro::Socket, document how to get ipv6 support via
9TODO: guard as SLF - to avoid deadlocks. 12 AnyEvent::Socket instead.
10TODO: Coro::Signal as SLF. 13 - implement signal->wait ($cb) interface, similar to semaphores.
11 - avoid deadlock in Coro::Channel if maxsize == 1 (Richard Hundt). 14 - work around SvOK not supporting getmagic, so we have to getmagic
15 to test for undef :/ (reported by Matthias Waldorf).
16 - load Coro::AnyEvent in all modules using AnyEvent.
12 17
135.0 185.14 Wed Jun 24 01:37:48 CEST 2009
19 - provide explicit functions to "cede" to the event loop in Coro::AnyEvent,
20 as this seems to have been a difficult concept (poll, sleep, idle,
21 idle_upto).
22 - add Coro::AnyEvent::readable/writable functions.
23 - clarify Coro::EV/Event/AnyEvent manpages.
24 - free per-thread global scalars in the thread calling ->cancel, to
25 avoid crashes when $_, $@ etc., are magical but some of those
26 had already been freed.
27 - "unexperimentalise" the callback interface for Coro::Semaphore.
28 - speed up ready queue management/context switching by using a linked
29 list instead of an array (~5-10%).
30 - implement "watch" command in Coro::Debug shells.
31 - for fun, implement time-slicing as an example in the manpage.
32 - if AnyEvent detects EV or Event, but we don't have Coro::EV or
33 Coro::Event, use the normal AnyEvent handling instead of dieing
34 (the same is true for Coro::Handle).
35 - properly document Coro::EV::timed_io_once.
36 - avoid unneccessary ->cancel calls in Coro::Handle.
37 - maybe make it work on mingw32 with win32 backend
38 (based on patch by Yasuhiro Matsumoto).
39
405.132 Fri May 29 09:00:39 CEST 2009
41 - do not keep a reference to the argument itself in
42 Coro::Semaphore::guard, as it could change later.
43 - support SO_RCVBUF/SO_SNDBUF nonstandard Coro::Socket options,
44 should support a prepare callback.
45
465.131 Mon Mar 16 23:20:37 CET 2009
47 - implement and document Coro->suspend, Coro->resume.
48 - fix Coro::Select implementation to not (often) close
49 the passed file descriptors (testcase provided by pippijn).
50
515.13 Mon Dec 15 21:51:42 CET 2008
52 - EXPERIMENTAL: implement dynamic winds (on_enter/on_leave).
53 - don't set diehook to C<undef> but instead to NULL, to avoid
54 spurious warnings.
55 - fix a lot of bugs in Coro::SemaphoreSet.
56 - Coro::SemaphoreSet will less often create a semaphore needlessly.
57 - add Coro::SemaphoreSet::count and wait methods.
58 - take advantage of the new Guard module.
59 - deprecate Coro::guard.
60 - try to fix the dreaded 01_unblock tests once more. I hate it when
61 testsuites need more fixing than the code they are supposed to test.
62 - croak in more cases when a required callback isn't resolvable.
63 - fix some minor issues in Coro::State->call/eval.
64 - use current coroutine context instead of a temporary one
65 when temporarily switching to another coroutine.
66 - do not call C-level on_destroy handlers during global destruction,
67 to avoid needless segfaults.
68
695.12 Sun Dec 7 13:30:38 CET 2008
70 - add default config for MirOS, which seems to be a bug-to-bug
71 compatible fork of openbsd ("world domination by releasing
72 openbsd cvs before the openbsd folks do it" or so :).
73 - free_padlist did destroy the names pad, not good, but didn't
74 seem to bother perl - this could fix issues such as eval ""
75 inside a function called from multiple coroutines.
76 - use a different method to detect destruction time.
77 - be more careful when freeing padlists just before global
78 destruction.
79 - fixed and expanded the call/cc example.
80 - renamed _terminate to _coro_run.
81 - new method Coro::Channel->shutdown.
82 - try pthreads on openbsd <4.4 (broken sigaltstack, will
83 pthreads fare better?).
84 - be less picky about destroying "a" running coroutine.
85
865.11 Tue Nov 25 21:49:05 CET 2008
87 - DEBUGGING in 5.10.0 is a mess: it constantly flags perfectly
88 working code with failed assertions, introducing more bugs than
89 it fixes, requiring elaborate workarounds :(
90
915.1 Mon Nov 24 08:54:59 CET 2008
92 - wrote a small introductory tutorial - Coro::Intro.
93 - convert Coro::Timer, Coro::Select and Coro::Util to rouse API.
94 - Coro::Select did errornously dup the file descriptors
95 and didn't work with all AnyEvent backends.
96 - Coro::Select wasn't imported correctly form Coro::LWP, causing blocking
97 LWP data transfers.
98 - disassociate c contexts from coro objects - this is agruably more
99 correct, but mostly allows sharing of cctxs between coro and state
100 objects, for added memory savings and speed increases.
101 - bumped $Coro::POOL_RSS up to 32kb by default.
102 - no longer set the optype to OP_CUSTOM, as B::* understandably
103 doesn't like this very much (and we *are* a type of entersub).
104 - implement state cloning, just to prove that call/cc can be done.
105 - automatically load Coro::AnyEvent in Coro::Handle.
106 - wrap ->cancel calls in eval inside Coro::Handle as EV watchers
107 do not have this method (and don't need it either).
108 - speed up generic anyevent methods in Coro::Handle by using rouse
109 callbacks.
110 - allow coroutines in $Coro::IDLE, speeding up Coro::AnyEvent and
111 others. It also makes the debugger happier, as you can trace
112 through the idle threads now.
113 - add comppad_name* and hints ($^H, %^H) to per-thread variables.
114 - eg/event was pretty broken.
115 - better 5.8.6 compatibility.
116
1175.0 Thu Nov 20 10:35:05 CET 2008
14 - NEW ARCHITECTURE: use the latest 4.x version if you experience 118 - NEW ARCHITECTURE: use the latest 4.x version if you experience
15 stability issues. 119 stability issues.
16 - bump API version to 7 - all dependents must be recompiled. 120 - bump API version to 7 - all dependents must be recompiled.
17 - removed timed_* functions - they were not being used anyways 121 - removed timed_* functions - they were not being used anyways
18 and should be replaced by a more generic mechanism - 122 and should be replaced by a more generic mechanism -
19 and were annoying to support anyways :) 123 and were annoying to support anyways :)
20 - removed SemaphoreSet's waiter method - use sem instead. 124 - removed SemaphoreSet's waiter method - use sem method instead.
21 - Coro::Semaphore->adjust didn't correctly wake up enough waiters. 125 - Coro::Semaphore->adjust didn't correctly wake up enough waiters.
22 - async_pool did free a scalar value twice 126 - async_pool did free a scalar value twice
23 ("Attempt to unreference..."). 127 ("Attempt to unreference...").
24 - fix a longstanding bug where calling terminate on a coro that 128 - fix a longstanding bug where calling terminate on a coro that
25 was waiting for a semaphore that was just becoming available 129 was waiting for a semaphore that was just becoming available
… …
27 it was available but waiters were still blocked). 131 it was available but waiters were still blocked).
28 - calling throw on a coroutine that is waiting for a semaphore will 132 - calling throw on a coroutine that is waiting for a semaphore will
29 no longer make it acquire the semaphore (and thus leak a count). 133 no longer make it acquire the semaphore (and thus leak a count).
30 - perl's process emulation is now not even theoretically supported 134 - perl's process emulation is now not even theoretically supported
31 anymore. 135 anymore.
136 - new functions Coro::rouse_cb and Coro::rouse_wait for easier
137 conversion of callback-style to blocking-style.
138 - new methods $coro->schedule_to and ->cede_to, to specifically
139 schedule or cede to a specific coroutine.
32 - new function Coro::Semaphore::wait. 140 - new function Coro::Semaphore::wait.
33 - use named constants in Coro::Channel (Richard Hundt). 141 - use named constants in Coro::Channel (Richard Hundt).
34 - directly patch the entersub opcode calling SLF functions (cede, 142 - directly patch the entersub opcode calling SLF functions (cede,
35 transfer and so on). this does speed up context switching, but 143 transfer and so on). this does speed up context switching, but
36 more importanly, it frees us from the hardcoded behaviour of 144 more importanly, it frees us from the hardcoded behaviour of
… …
50 - implement Coro::SemaphoreSet purely in terms of Coro::Semaphore, 158 - implement Coro::SemaphoreSet purely in terms of Coro::Semaphore,
51 for a nice speedup and vastly more correct behaviour. Also implement 159 for a nice speedup and vastly more correct behaviour. Also implement
52 a new method "sem" to get at the underlying semaphore object. 160 a new method "sem" to get at the underlying semaphore object.
53 - implement Coro::Channel in terms of Coro::Semaphore, for a moderate 161 - implement Coro::Channel in terms of Coro::Semaphore, for a moderate
54 (in comparison) 20-40% speedup. 162 (in comparison) 20-40% speedup.
55 - implement Coro::Signal (mostly) in terms of Coro::Semaphore, 163 - used new SLF interface to reimplement Coro::Signal gaining
56 for some unknown (because I was too lazy), but certain, speedup. 164 some unknown (because I was too lazy), but certain, speedup, and also
165 making signals reliable for the first time.
166 - used new SLF interface and other optimisations to speed up async_pool
167 by a factor of two. It also doesn't rely on perl's exception mechanism
168 to exit anymore. The overhead for terminating an async_pool, coro over
169 a normal async is now very small.
170 - sped up coroutine creation/destruction by 40%.
57 - forgot to include Coro/libcoro/README in the dist for all these years. 171 - forgot to include Coro/libcoro/README in the dist for all these years.
58 - work around a freebsd pthreads bug (manual testcancel is required as 172 - work around a freebsd pthreads bug (manual testcancel is required as
59 pthread_cond_wait isn't a cancellation point on freebsd). 173 pthread_cond_wait isn't a cancellation point on freebsd).
174 - use new rouse functions to speed up and simplify Coro::BDB.
175 - make "prefer perl native functions" work with threaded perls.
176 - condense Coro::Debug ps output, hint at v and w flags.
177 - (libcoro) lots of minor cleanups and portability improvements.
178
1794.914 Wed Nov 19 12:54:18 CET 2008
180 - fix a disastrous bug in the readline optimisation
181 introduced in 4.801.
60 182
614.913 Sat Nov 15 07:58:28 CET 2008 1834.913 Sat Nov 15 07:58:28 CET 2008
62 - async_pool did free a scalar value twice 184 - async_pool did free a scalar value twice
63 ("Attempt to unreference..."). 185 ("Attempt to unreference...").
64 186

Diff Legend

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