ViewVC Help
View File | Revision Log | Show Annotations | Download File
/cvs/libev/Changes
(Generate patch)

Comparing libev/Changes (file contents):
Revision 1.88 by root, Fri Nov 21 02:10:34 2008 UTC vs.
Revision 1.128 by root, Wed Jul 15 16:58:53 2009 UTC

1Revision history for libev, a high-performance and full-featured event loop. 1Revision history for libev, a high-performance and full-featured event loop.
2 2
3TODO: ev_walk
4TODO: remain
5TODO: EV_MINIMAL
6
7 - ev_unloop and ev_loop wrongly used a global variable to exit loops,
8 instead of using a per-loop variable (bug caught by accident...).
9 - the ev_set_io_collect_interval interpretation has changed.
10 - add new functionality: ev_set_userdata, ev_userdata,
11 ev_set_invoke_pending_cb, ev_set_loop_release_cb,
12 ev_invoke_pending, together with a long example about thread
13 locking.
14 - add ev_timer_remaining (as requested by Denis F. Latypoff).
15 - add ev_loop_depth.
16 - calling ev_unloop in fork/prepare watchers will no longer poll
17 for new events.
18 - Denis F. Latypoff corrected many typos in example code snippets.
19 - honor autoconf detection of EV_USE_CLOCK_SYSCALL, also double-
20 check that the syscall number is available before trying to
21 use it (reported by ry@tinyclouds).
22 - use GetSystemTimeAsFileTime instead of _timeb on windows, for
23 slightly higher accuracy.
24 - properly declare ev_loop_verify and ev_now_update even when
25 !EV_MULTIPLICITY.
26 - do not compile in any priority code when EV_MAXPRI == EV_MINPRI.
27 - support EV_MINIMAL==2 for a reduced API.
28 - actually 0-initialise struct sigaction when installing signals.
29 - add section on hibernate and stopped processes to ev_timer docs.
30
313.6 Tue Apr 28 02:49:30 CEST 2009
32 - multiple timers becoming ready within an event loop iteration
33 will be invoked in the "correct" order now.
34 - do not leave the event loop early just because we have no active
35 watchers, fixing a problem when embedding a kqueue loop
36 that has active kernel events but no registered watchers
37 (reported by blacksand blacksand).
38 - correctly zero the idx values for arrays, so destroying and
39 reinitialising the default loop actually works (patch by
40 Malek Hadj-Ali).
41 - implement ev_suspend and ev_resume.
42 - new EV_CUSTOM revents flag for use by applications.
43 - add documentation section about priorites.
44 - add a glossary to the dcoumentation.
45 - extend the ev_fork description slightly.
46 - optimize a jump out of call_pending.
47
483.53 Sun Feb 15 02:38:20 CET 2009
49 - fix a bug in event pipe creation on win32 that would cause a
50 failed assertion on event loop creation (patch by Malek Hadj-Ali).
51 - probe for CLOCK_REALTIME support at runtime as well and fall
52 back to gettimeofday if there is an error, to support older
53 operating systems with newer header files/libraries.
54 - prefer gettimeofday over clock_gettime with USE_CLOCK_SYSCALL
55 (default most everywhere), otherwise not.
56
573.52 Wed Jan 7 21:43:02 CET 2009
58 - fix compilation of select backend in fd_set mode when NFDBITS is
59 missing (to get it to compile on QNX, reported by Rodrigo Campos).
60 - better select-nfds handling when select backend is in fd_set mode.
61 - diagnose fd_set overruns when select backend is in fd_set mode.
62 - due to a thinko, instead of disabling everything but
63 select on the borked OS X platform, everything but select was
64 allowed (reported by Emanuele Giaquinta).
65 - actually verify that local and remote port are matching in
66 libev's socketpair emulation, which makes denial-of-service
67 attacks harder (but not impossible - it's windows). Make sure
68 it even works under vista, which thinks that getpeer/sockname
69 should return fantasy port numbers.
70 - include "libev" in all assertion messages for potentially
71 clearer diagnostics.
72 - event_get_version (libevent compatibility) returned
73 a useless string instead of the expected version string
74 (patch by W.C.A. Wijngaards).
75
763.51 Wed Dec 24 23:00:11 CET 2008
77 - fix a bug where an inotify watcher was added twice, causing
78 freezes on hash collisions (reported and analysed by Graham Leggett).
3 - new config symbol, EV_USE_CLOCK_SYSCALL, to make libev use 79 - new config symbol, EV_USE_CLOCK_SYSCALL, to make libev use
4 a direct syscall - slower, but no dependency on librt et al. 80 a direct syscall - slower, but no dependency on librt et al.
5 - assume negative return values != -1 signals success of port_getn 81 - assume negative return values != -1 signals success of port_getn
6 (http://cvs.epicsol.org/cgi/viewcvs.cgi/epic5/source/newio.c?rev=1.52) 82 (http://cvs.epicsol.org/cgi/viewcvs.cgi/epic5/source/newio.c?rev=1.52)
7 (no known failure reports, but it doesn't hurt). 83 (no known failure reports, but it doesn't hurt).
84 - fork detection in ev_embed now stops and restarts the watcher
85 automatically.
86 - EXPERIMENTAL: default the method to operator () in ev++.h,
87 to make it nicer to use functors (requested by Benedek László).
88 - fixed const object callbacks in ev++.h.
89 - replaced loop_ref argument of watcher.set (loop) by a direct
90 ev_loop * in ev++.h, to avoid clashes with functor patch.
91 - do not try to watch the empty string via inotify.
92 - inotify watchers could be leaked under certain circumstances.
93 - OS X 10.5 is actually even more broken than earlier versions,
94 so fall back to select on that piece of garbage.
95 - fixed some weirdness in the ev_embed documentation.
8 96
93.49 Wed Nov 19 11:26:53 CET 2008 973.49 Wed Nov 19 11:26:53 CET 2008
10 - ev_stat watchers will now use inotify as a mere hint on 98 - ev_stat watchers will now use inotify as a mere hint on
11 kernels <2.6.25, or if the filesystem is not in the 99 kernels <2.6.25, or if the filesystem is not in the
12 "known to be good" list. 100 "known to be good" list.

Diff Legend

– Removed lines
+ Added lines
< Changed lines
> Changed lines