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 | WISH? monotonic clocks times/GetTickCount for coarse corrections? |
3 | 3.49 Wed Nov 19 11:26:53 CET 2008 |
|
|
4 | - ev_stat watchers will now use inotify as a mere hint on |
|
|
5 | kernels <2.6.25, or if the filesystem is not in the |
|
|
6 | "known to be good" list. |
|
|
7 | - better mingw32 compatibility (it's not as borked as native win32) |
|
|
8 | (analysed by Roger Pack). |
|
|
9 | - include stdio.h in the example program, as too many people are |
|
|
10 | confused by the weird C language otherwise. I guess the next thing |
|
|
11 | I get told is that the "..." ellipses in the examples don't compile |
|
|
12 | with their C compiler. |
4 | |
13 | |
|
|
14 | 3.48 Thu Oct 30 09:02:37 CET 2008 |
5 | - further optimise away the EPOLL_CTL_ADD/MOD combo in the epoll |
15 | - further optimise away the EPOLL_CTL_ADD/MOD combo in the epoll |
6 | backend by assuming the kernel event mask hasn't changed if |
16 | backend by assuming the kernel event mask hasn't changed if |
7 | ADD fails with EEXIST. |
17 | ADD fails with EEXIST. |
|
|
18 | - work around spurious event notification bugs in epoll by using |
|
|
19 | a 32-bit generation counter. recreate kernel state if we receive |
|
|
20 | spurious notifications or unwanted events. this is very costly, |
|
|
21 | but I didn't come up with this horrible design. |
8 | - use memset to initialise most arrays now and do away with the |
22 | - use memset to initialise most arrays now and do away with the |
9 | init functions. |
23 | init functions. |
10 | - expand time-out strategies into a "Be smart about timeouts" section. |
24 | - expand time-out strategies into a "Be smart about timeouts" section. |
11 | - drop the "struct" from all ev_watcher declarations in the |
25 | - drop the "struct" from all ev_watcher declarations in the |
12 | documentation and did other clarifications (yeah, it was a mistake |
26 | documentation and did other clarifications (yeah, it was a mistake |
13 | to have a struct AND a function called ev_loop). |
27 | to have a struct AND a function called ev_loop). |
|
|
28 | - fix a bug where ev_default would not initialise the default |
|
|
29 | loop again after it was destroyed with ev_default_destroy. |
|
|
30 | - rename syserr to ev_syserr to avoid name clashes when embedding, |
|
|
31 | do similar changes for event.c. |
14 | |
32 | |
15 | 3.45 Tue Oct 21 21:59:26 CEST 2008 |
33 | 3.45 Tue Oct 21 21:59:26 CEST 2008 |
16 | - disable inotify usage on linux <2.6.25, as it is broken |
34 | - disable inotify usage on linux <2.6.25, as it is broken |
17 | (reported by Yoann Vandoorselaere). |
35 | (reported by Yoann Vandoorselaere). |
18 | - ev_stat errornously would try to add inotify watchers |
36 | - ev_stat errornously would try to add inotify watchers |