| 1 | Revision history for libev, a high-performance and full-featured event loop. |
1 | Revision history for libev, a high-performance and full-featured event loop. |
| 2 | |
2 | |
| 3 | TODO: ev_loop_wakeup |
3 | - move the darwin select workaround highe rin ev.c, as newre versions of |
| 4 | TODO: EV_STANDALONE == NO_HASSEL (do not use clock_gettime in ev_standalone) |
4 | darwin managed to break their broken select even more. |
| 5 | TODO: faq, process a thing in each iteration |
5 | - ANDROID => __ANDROID__ (reported by enh@google.com). |
| 6 | TODO: dbeugging tips, ev_verify, ev_init twice |
6 | - disable epoll_create1 on android because it has broken header files |
| 7 | TODO: ev_break for immediate exit (EVBREAK_NOW?) |
7 | and google is unwilling to fix them (reported by enh@google.com). |
| 8 | TODO: ev_feed_child_event |
8 | - avoid a minor compilation warning on win32. |
| 9 | TODO: document the special problem of signals around fork. |
9 | |
| 10 | TODO: store pid for each signal |
10 | 4.24 Wed Dec 28 05:19:55 CET 2016 |
| 11 | TODO: document file descriptor usage per loop |
11 | - bump version to 4.24, as the release tarball inexplicably |
| 12 | TODO: store loop pid_t and compare isndie signal handler,store 1 for same, 2 for differign pid, clean up in loop_fork |
12 | didn't have the right version in ev.h, even though the cvs-tagged |
| 13 | TODO: embed watchers need updating when fd changes |
13 | version did have the right one (reported by Ales Teska). |
| 14 | TODO: document portability requirements for atomic pointer access |
14 | |
| 15 | TODO: possible cb aliasing? |
15 | 4.23 Wed Nov 16 18:23:41 CET 2016 |
| 16 | TODO: document requirements for function pointers and calling conventions. |
16 | - move some declarations at the beginning to help certain retarded |
|
|
17 | microsoft compilers, even though their documentation claims |
|
|
18 | otherwise (reported by Ruslan Osmanov). |
|
|
19 | |
|
|
20 | 4.22 Sun Dec 20 22:11:50 CET 2015 |
|
|
21 | - when epoll detects unremovable fds in the fd set, rebuild |
|
|
22 | only the epoll descriptor, not the signal pipe, to avoid |
|
|
23 | SIGPIPE in ev_async_send. This doesn't solve it on fork, |
|
|
24 | so document what needs to be done in ev_loop_fork |
|
|
25 | (analyzed by Benjamin Mahler). |
|
|
26 | - remove superfluous sys/timeb.h include on win32 |
|
|
27 | (analyzed by Jason Madden). |
|
|
28 | - updated libecb. |
|
|
29 | |
|
|
30 | 4.20 Sat Jun 20 13:01:43 CEST 2015 |
|
|
31 | - prefer noexcept over throw () with C++ 11. |
|
|
32 | - update ecb.h due to incompatibilities with c11. |
|
|
33 | - fix a potential aliasing issue when reading and writing |
|
|
34 | watcher callbacks. |
|
|
35 | |
|
|
36 | 4.19 Thu Sep 25 08:18:25 CEST 2014 |
|
|
37 | - ev.h wasn't valid C++ anymore, which tripped compilers other than |
|
|
38 | clang, msvc or gcc (analyzed by Raphael 'kena' Poss). Unfortunately, |
|
|
39 | C++ doesn't support typedefs for function pointers fully, so the affected |
|
|
40 | declarations have to spell out the types each time. |
|
|
41 | - when not using autoconf, tighten the check for clock_gettime and related |
|
|
42 | functionality. |
| 17 | |
43 | |
| 18 | 4.18 Fri Sep 5 17:55:26 CEST 2014 |
44 | 4.18 Fri Sep 5 17:55:26 CEST 2014 |
| 19 | - events on files were not always generated properly with the |
45 | - events on files were not always generated properly with the |
| 20 | epoll backend (testcase by Assaf Inbal). |
46 | epoll backend (testcase by Assaf Inbal). |
| 21 | - mark event pipe fd as cloexec after a fork (analyzed by Sami Farin). |
47 | - mark event pipe fd as cloexec after a fork (analyzed by Sami Farin). |
