| … | |
… | |
| 9 | TODO: document the special problem of signals around fork. |
9 | TODO: document the special problem of signals around fork. |
| 10 | TODO: store pid for each signal |
10 | TODO: store pid for each signal |
| 11 | TODO: document file descriptor usage per loop |
11 | TODO: document file descriptor usage per loop |
| 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 | TODO: store loop pid_t and compare isndie signal handler,store 1 for same, 2 for differign pid, clean up in loop_fork |
| 13 | TODO: embed watchers need updating when fd changes |
13 | TODO: embed watchers need updating when fd changes |
| 14 | TODO: document portbaility requirements for atomic pointer access |
14 | TODO: document portability requirements for atomic pointer access |
|
|
15 | TODO: possible cb aliasing? |
|
|
16 | TODO: document requirements for function pointers and calling conventions. |
|
|
17 | |
|
|
18 | - prefer noexcept over throw () with C++ 11. |
|
|
19 | - update ecb.h due to incompatibilities with c11. |
|
|
20 | - fix a potentgial aliasing issue when reading and writing |
|
|
21 | watcher callbacks. |
|
|
22 | |
|
|
23 | 4.19 Thu Sep 25 08:18:25 CEST 2014 |
|
|
24 | - ev.h wasn't valid C++ anymore, which tripped compilers other than |
|
|
25 | clang, msvc or gcc (analyzed by Raphael 'kena' Poss). Unfortunately, |
|
|
26 | C++ doesn't support typedefs for function pointers fully, so the affected |
|
|
27 | declarations have to spell out the types each time. |
|
|
28 | - when not using autoconf, tighten the check for clock_gettime and related |
|
|
29 | functionality. |
|
|
30 | |
|
|
31 | 4.18 Fri Sep 5 17:55:26 CEST 2014 |
|
|
32 | - events on files were not always generated properly with the |
|
|
33 | epoll backend (testcase by Assaf Inbal). |
|
|
34 | - mark event pipe fd as cloexec after a fork (analyzed by Sami Farin). |
|
|
35 | - (ecb) support m68k, m88k and sh (patch by Miod Vallat). |
|
|
36 | - use a reasonable fallback for EV_NSIG instead of erroring out |
|
|
37 | when we can't detect the signal set size. |
|
|
38 | - in the absence of autoconf, do not use the clock syscall |
|
|
39 | on glibc >= 2.17 (avoids the syscall AND -lrt on systems |
|
|
40 | doing clock_gettime in userspace). |
|
|
41 | - ensure extern "C" function pointers are used for externally-visible |
|
|
42 | loop callbacks (not watcher callbacks yet). |
|
|
43 | - (ecb) work around memory barriers and volatile apparently both being |
|
|
44 | broken in visual studio 2008 and later (analysed and patch by Nicolas Noble). |
| 15 | |
45 | |
| 16 | 4.15 Fri Mar 1 12:04:50 CET 2013 |
46 | 4.15 Fri Mar 1 12:04:50 CET 2013 |
| 17 | - destroying a non-default loop would stop the global waitpid |
47 | - destroying a non-default loop would stop the global waitpid |
| 18 | watcher (Denis Bilenko). |
48 | watcher (Denis Bilenko). |
| 19 | - queueing pending watchers of higher priority from a watcher now invokes |
49 | - queueing pending watchers of higher priority from a watcher now invokes |
