| 1 | Revision history for Perl extension Coro. |
1 | Revision history for Perl extension Coro. |
| 2 | |
2 | |
| 3 | TODO: better (earlier) error messages on broken transfer's |
3 | TODO: better (earlier) error messages on broken transfer's |
| 4 | TODO: "internals mode" |
4 | TODO: destroy gets only called on cancel: fix(?) |
|
|
5 | TODO: should explore PerlIO::via::CoroCede and PerlIO::via::CoroAIO. |
|
|
6 | TODO: maybe implement a default message channel, very much like Erlang |
|
|
7 | (which is cool in a lot of important aspects (failures!), |
|
|
8 | but very lacking in others). |
| 5 | |
9 | |
|
|
10 | - added Coro::Semaphore::adjust. |
|
|
11 | |
|
|
12 | 3.63 Wed May 16 14:10:06 CEST 2007 |
|
|
13 | - implement handcoded assembly for x86/amd64 SVR ABI. |
|
|
14 | |
|
|
15 | 3.62 Fri Apr 27 21:36:06 CEST 2007 |
|
|
16 | - upgrade libcoro (which might set unwind info correctly). |
|
|
17 | - change default on linux to setjmp/longjmp. |
|
|
18 | |
|
|
19 | 3.61 Thu Apr 19 12:36:18 CEST 2007 |
|
|
20 | - Coro::Storable caused an endless loop when thawing invalid |
|
|
21 | pst's sometimes. |
|
|
22 | - use a Semaphore in Coro::Storable, as Storable doesn't |
|
|
23 | seem to be reentrant (although it is documented to |
|
|
24 | be threadsafe...). |
|
|
25 | - fix Coro::Signal to bring back the original unreliable |
|
|
26 | but stateful semantics. |
|
|
27 | - fixed a lot of typos in Coro.pm (patch submitted by David |
|
|
28 | Steinbrunner, which applied flawlessly). |
|
|
29 | |
|
|
30 | 3.6 Sat Apr 14 17:13:31 CEST 2007 |
|
|
31 | - added some bugfixes to get eg/myhttpd working again. |
|
|
32 | - added Coro::Storable for often-cede'ing freeze/thaw. |
|
|
33 | - try to do a clean exit when a coroutine calls exit |
|
|
34 | (EXPERIMENTAL). |
|
|
35 | - got rid of indirect call through _coro_init. |
|
|
36 | - updated the partly antique examples in eg/ to |
|
|
37 | work again and be a bit less magic, too. |
|
|
38 | - fixed Coro::Signal semantics to work as documented again. |
|
|
39 | |
|
|
40 | 3.55 Sun Mar 25 01:20:47 CET 2007 |
|
|
41 | - add SAVE_DEFFH to save the default filehandle and enable |
|
|
42 | it by default. |
|
|
43 | - finally move socket-operations from Coro::Socket to Coro::Handle |
|
|
44 | to be able to unblock foreign sockets. |
|
|
45 | - add Coro::State::save_also and guarded_save. |
|
|
46 | - add count accessor to Coro::Semaphore. |
|
|
47 | - add Coro::State::cctx_stacksize. |
|
|
48 | - just for the fun of it, do not rely on implicit context, |
|
|
49 | which can dramatically improve performance, but people |
|
|
50 | using windows-process-emulation perls do not care much about |
|
|
51 | performance. |
|
|
52 | |
|
|
53 | 3.51 Sun Mar 4 14:18:04 CET 2007 |
|
|
54 | - fixed a problem when you weakened references to Coro::State's |
|
|
55 | (patch by Yuval Kogman). |
|
|
56 | |
|
|
57 | 3.501 Wed Feb 28 12:44:07 CET 2007 |
|
|
58 | - rename some global symbols as macosx from hell redefines |
|
|
59 | them without asking. |
|
|
60 | |
|
|
61 | 3.5 Tue Feb 13 20:22:53 CET 2007 |
|
|
62 | - do AnyEvent model detection earlier, avoiding problems |
|
|
63 | caused by first using AnyEvent and later Coro::Event. |
|
|
64 | - implement and document Coro::Event event objects. |
|
|
65 | - fix a potential problem in Coro::Event causing crashes. |
|
|
66 | - initialise PL_comppad when creating a new coroutine, |
|
|
67 | avoids crashes on early coro destruction. |
|
|
68 | |
|
|
69 | 3.41 Mon Jan 22 19:19:49 CET 2007 |
|
|
70 | - readline on Coro::Handle did not support undefined $/, |
|
|
71 | nor did it deliver partial lines on EOF or error. |
|
|
72 | - implement malloc fallback for stack allocation because |
|
|
73 | stupid broken idiotic OSX has a stupid broken |
|
|
74 | idiotic fits-the-whole-os mmap "implementation" and |
|
|
75 | my dick feels longer if Coro is portable even to |
|
|
76 | obsolete platforms. |
|
|
77 | |
|
|
78 | 3.4 Fri Jan 19 21:52:54 CET 2007 |
|
|
79 | - remove t/09_timer.t, as it isn't really testing much |
|
|
80 | but was rather flaky in practise. |
|
|
81 | - async_pool coro would keep arguments and callback alive until |
|
|
82 | it was reused. |
|
|
83 | - cancellation of a coroutine could cause spurious idle calls |
|
|
84 | in cede_notself. |
|
|
85 | |
|
|
86 | 3.3 Sat Jan 6 03:45:00 CET 2007 |
|
|
87 | - implement $coro->on_destroy. |
|
|
88 | - Coro::Event blocking semantics have been changed, |
|
|
89 | documented and - hopefully - improved. |
|
|
90 | - fix nice adding, not subtracting, from priority. |
|
|
91 | - fix ->prio and api_is_ready (patch by Mark Hinds). |
|
|
92 | - fixed an assert ("... == prev__cctx->idle_te") |
|
|
93 | that could errronously trigger. |
|
|
94 | - fix various large and small memleaks. |
|
|
95 | - use a (hopefully) more stable cancel implementation |
|
|
96 | that immediately frees the coroutine data. |
|
|
97 | - cede/cede_notself return a status now. |
|
|
98 | - added Coro::guard function. |
|
|
99 | - added a global coroutine pool for jobs (on my machine, |
|
|
100 | I can create and execute 48k simple coros/s with async, |
|
|
101 | and 128k coros with async_pool). |
|
|
102 | - Coro::AIO now uses the coroutine priority as io priority. |
|
|
103 | |
|
|
104 | 3.2 Fri Dec 22 05:07:09 CET 2006 |
| 6 | - improve portability to slightly older perls. |
105 | - improve portability to slightly older perls. |
| 7 | - use cleaner coroutine destruction. |
106 | - use cleaner coroutine destruction. |
|
|
107 | - simplify configuration for users. |
|
|
108 | - optionally (unrecommended) prefer perl functions over |
|
|
109 | their coro replacements. |
| 8 | |
110 | |
| 9 | 3.11 Tue Dec 5 13:11:24 CET 2006 |
111 | 3.11 Tue Dec 5 13:11:24 CET 2006 |
| 10 | - fixed some bogus assert's, but as perl.h disables assert even |
112 | - fixed some bogus assert's, but as perl.h disables assert even |
| 11 | without NDEBUG (thank you very much), not too many people should |
113 | without NDEBUG (thank you very much), not too many people should |
| 12 | notice (that did include myself). Andreas König noticed, though :) |
114 | notice (that did include myself). Andreas König noticed, though :) |
