| 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 | 3.45 Tue Oct 21 21:59:26 CEST 2008 |
|
|
4 | - disable inotify usage on linux <2.6.25, as it is broken |
|
|
5 | (reported by Yoann Vandoorselaere). |
|
|
6 | - ev_stat errornously would try to add inotify watchers |
|
|
7 | even when inotify wasn't available (this should only |
|
|
8 | have a performance impact). |
|
|
9 | - ev_once now passes both timeout and io to the callback if both |
|
|
10 | occur concurrently, instead of giving timeouts precedence. |
|
|
11 | - disable EV_USE_INOTIFY when sys/inotify.h is too old. |
|
|
12 | |
|
|
13 | 3.44 Mon Sep 29 05:18:39 CEST 2008 |
|
|
14 | - embed watchers now automatically invoke ev_loop_fork on the |
|
|
15 | embedded loop when the parent loop forks. |
|
|
16 | - new function: ev_now_update (loop). |
|
|
17 | - verify_watcher was not marked static. |
|
|
18 | - improve the "associating..." manpage section. |
|
|
19 | - documentation tweaks here and there. |
|
|
20 | |
|
|
21 | 3.43 Sun Jul 6 05:34:41 CEST 2008 |
|
|
22 | - include more include files on windows to get struct _stati64 |
|
|
23 | (reported by Chris Hulbert, but doesn't quite fix his issue). |
|
|
24 | - add missing #include <io.h> in ev.c on windows (reported by |
|
|
25 | Matt Tolton). |
|
|
26 | |
|
|
27 | 3.42 Tue Jun 17 12:12:07 CEST 2008 |
|
|
28 | - work around yet another windows bug: FD_SET actually adds fd's |
|
|
29 | multiple times to the fd_*SET*, despite official MSN docs claiming |
|
|
30 | otherwise. Reported and well-analysed by Matt Tolton. |
|
|
31 | - define NFDBITS to 0 when EV_SELECT_IS_WINSOCKET to make it compile |
|
|
32 | (reported any analysed by Chris Hulbert). |
|
|
33 | - fix a bug in ev_ebadf (this function is only used to catch |
|
|
34 | programming errors in the libev user). reported by Matt Tolton. |
|
|
35 | - fix a bug in fd_intern on win32 (could lead to compile errors |
|
|
36 | under some circumstances, but would work correctly if it compiles). |
|
|
37 | reported by Matt Tolton. |
|
|
38 | - (try to) work around missing lstat on windows. |
| 3 | - pass in the write fd set as except fd set under windows. windows |
39 | - pass in the write fd set as except fd set under windows. windows |
| 4 | is so uncontrollably lame that it requires this. |
40 | is so uncontrollably lame that it requires this. this means that |
|
|
41 | switching off oobinline is not supported (but tcp/ip doesn't |
|
|
42 | have oob, so that would be stupid anyways. |
|
|
43 | - use posix module symbol to auto-detect monotonic clock presence |
|
|
44 | and some other default values. |
| 5 | |
45 | |
| 6 | 3.41 Fri May 23 18:42:54 CEST 2008 |
46 | 3.41 Fri May 23 18:42:54 CEST 2008 |
| 7 | - work around an (undocumented) bug in winsocket select: if you |
47 | - work around an obscure bug in winsocket select: if you |
| 8 | provide only empty fd sets then select returns WSAEINVAL. how sucky. |
48 | provide only empty fd sets then select returns WSAEINVAL. how sucky. |
| 9 | - improve timer scheduling stability and reduce use of time_epsilon. |
49 | - improve timer scheduling stability and reduce use of time_epsilon. |
| 10 | - use 1-based 2-heap for EV_MINIMAL, simplifies code, reduces |
50 | - use 1-based 2-heap for EV_MINIMAL, simplifies code, reduces |
| 11 | codesize and makes for better cache-efficiency. |
51 | codesize and makes for better cache-efficiency. |
| 12 | - use 3-based 4-heap for !EV_MINIMAL. this makes better use |
52 | - use 3-based 4-heap for !EV_MINIMAL. this makes better use |
