| … | |
… | |
| 3 | TODO: should explore PerlIO::coroaio (perl leaks like hell). |
3 | TODO: should explore PerlIO::coroaio (perl leaks like hell). |
| 4 | TODO: maybe implement a default message channel, very much like Erlang's |
4 | TODO: 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)). |
| 7 | |
7 | |
|
|
8 | TODO: testcancel |
|
|
9 | TODO: aio |
|
|
10 | TODO: Coro::Semaphore |
| 8 | 5.0 |
11 | 5.0 |
| 9 | - NEW ARCHITECTURE: use 4.912 for the stable version. |
12 | - NEW ARCHITECTURE: use 4.912 for the stable version. |
|
|
13 | - bump API version to 7 - all dependents must be recompiled. |
|
|
14 | - Coro::Semaphore->adjust didn't correctly wake up enough waiters. |
|
|
15 | - async_pool did free a scalar value twice |
|
|
16 | ("Attempt to unreference..."). |
|
|
17 | - fix a longstanding bug where calling terminate on a coro that |
|
|
18 | was waiting for a semaphore that was just becoming available |
|
|
19 | would cause a deadlock (semaphore would get into a state where |
|
|
20 | it was available but waiters were waiting(. |
|
|
21 | - perl's process emulation is now not even theoretically supported |
|
|
22 | anymore. |
| 10 | - directly patch the entersub opcode calling ssl functions (cede, |
23 | - directly patch the entersub opcode calling SLF functions (cede, |
| 11 | transfer and so on). this does speed up context switching by |
24 | transfer and so on). this does speed up context switching, but |
| 12 | up to 10%. More importanly, it frees us from the hardcoded |
25 | more importanly, it frees us from the hardcoded behaviour of |
| 13 | behaviour of entersub, so we might actually be able to return |
26 | entersub, so we might actually be able to return something from |
| 14 | something from those functions. |
27 | those functions and atcually create new ones. |
|
|
28 | - take advantage of __builtin_frame_address on gcc. |
|
|
29 | - expose THX in coroapi (not sure whether this was a wise decision, |
|
|
30 | as "threaded" perls are running at half speed anyways). |
|
|
31 | - implement execute_slf (schedule-like-function) interface that makes |
|
|
32 | it possible to implement schedule-like-functions in XS. |
|
|
33 | - use new SLF interface to massively speed up Coro::EV by roughly a factor |
|
|
34 | of two. |
|
|
35 | - removed timed_down and timed_guard functions - they were not being used |
|
|
36 | anyways and should be replaced by a more generic mechanism - and were |
|
|
37 | annoying to support anyways : |
|
|
38 | - used new SLF interface to massively speed up Coro::Semaphore by a factor |
|
|
39 | of three. |
|
|
40 | - forgot to include Coro/libcoro/README in the dist for all these years. |
|
|
41 | - work around a freebsd pthreads bug (manual testcancel is required as |
|
|
42 | pthread_cond_wait isn't a cancellation point on freebsd). |
|
|
43 | |
|
|
44 | 4.913 Sat Nov 15 07:58:28 CET 2008 |
|
|
45 | - async_pool did free a scalar value twice |
|
|
46 | ("Attempt to unreference..."). |
| 15 | |
47 | |
| 16 | 4.912 Thu Nov 13 18:31:23 CET 2008 |
48 | 4.912 Thu Nov 13 18:31:23 CET 2008 |
| 17 | - minor cleanups. |
49 | - minor cleanups. |
| 18 | - use much larger stacks on linux and perl < 5.8.8. |
50 | - use much larger stacks on linux and perl < 5.8.8. |
| 19 | - Coro::Debug::new_unix_server did not unlink the socket |
51 | - Coro::Debug::new_unix_server did not unlink the socket |
