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 |
|
|
4 | |
|
|
5 | TODO: EV_NO_THREADS/NO_SMP and/or EV_USE_PTHREADS or so. also document. |
|
|
6 | TODO: #define EV_API_DECL static? ev_default_loop_ptr? |
|
|
7 | TODO: EV_STANDALONE == NO_HASSEL (do not use clock_gettime in ev_standalone) |
|
|
8 | TODO: memory fences for clang |
|
|
9 | 4.10 |
|
|
10 | - fix a race where the workaround against the epoll fork bugs |
|
|
11 | caused signals to not be handled anymore. |
|
|
12 | - correct backend_fudge for most backends, and implement a windows |
|
|
13 | specific workaround to avoid looping because we call both |
|
|
14 | select and Sleep, both with different time resolutions. |
|
|
15 | - document range and guarantees of ev_sleep. |
|
|
16 | - document reasonable ranges for periodics interval and offset. |
|
|
17 | - rename backend_fudge to backend_mintime to avoid future confusion :) |
|
|
18 | - change the default periodic reschedule function to hopefully be more |
|
|
19 | exact and correct even in corner cases or in the far future. |
|
|
20 | - do not rely on -lm anymore: use it when available but use our |
|
|
21 | own floor () if it is missing. This should make it easier to embed, |
|
|
22 | as no external libraries are required. |
|
|
23 | - strategically import macros from libecb and mark rarely-used functions |
|
|
24 | as cache-cold (saving almost 2k code size on typical amd64 setups). |
|
|
25 | - add Symbols.ev and Symbols.event files, that were missing. |
|
|
26 | |
|
|
27 | 4.04 Wed Feb 16 09:01:51 CET 2011 |
|
|
28 | - fix two problems in the native win32 backend, where reuse of fd's |
|
|
29 | with different underlying handles caused handles not to be removed |
|
|
30 | or added to the select set (analyzed and tested by Bert Belder). |
|
|
31 | - do no rely on ceil() in ev_e?poll.c. |
|
|
32 | - backport libev to HP-UX versions before 11 v3. |
|
|
33 | - configure did not detect nanosleep and clock_gettime properly when |
|
|
34 | they are available in the libc (as opposed to -lrt). |
|
|
35 | |
|
|
36 | 4.03 Tue Jan 11 14:37:25 CET 2011 |
|
|
37 | - officially support polling files with all backends. |
|
|
38 | - support files, /dev/zero etc. the same way as select in the epoll |
|
|
39 | backend, by generating events on our own. |
|
|
40 | - ports backend: work around solaris bug 6874410 and many related ones |
|
|
41 | (EINTR, maybe more), with no performance loss (note that the solaris |
|
|
42 | bug report is actually wrong, reality is far more bizarre and broken |
|
|
43 | than that). |
|
|
44 | - define EV_READ/EV_WRITE as macros in event.h, as some programs use |
|
|
45 | #ifdef to test for them. |
|
|
46 | - new (experimental) function: ev_feed_signal. |
|
|
47 | - new (to become default) EVFLAG_NOSIGMASK flag. |
|
|
48 | - new EVBACKEND_MASK symbol. |
|
|
49 | - updated COMMON IDIOMS SECTION. |
|
|
50 | |
|
|
51 | 4.01 Fri Nov 5 21:51:29 CET 2010 |
|
|
52 | - automake fucked it up, apparently, --add-missing -f is not quite enough |
|
|
53 | to make it update its files, so 4.00 didn't install ev++.h and |
|
|
54 | event.h on make install. grrr. |
|
|
55 | - ev_loop(count|depth) didn't return anything (Robin Haberkorn). |
|
|
56 | - change EV_UNDEF to 0xffffffff to silence some overzealous compilers. |
|
|
57 | - use "(libev) " prefix for all libev error messages now. |
|
|
58 | |
|
|
59 | 4.00 Mon Oct 25 12:32:12 CEST 2010 |
|
|
60 | - "PORTING FROM LIBEV 3.X TO 4.X" (in ev.pod) is recommended reading. |
|
|
61 | - ev_embed_stop did not correctly stop the watcher (very good |
|
|
62 | testcase by Vladimir Timofeev). |
|
|
63 | - ev_run will now always update the current loop time - it erroneously |
|
|
64 | didn't when idle watchers were active, causing timers not to fire. |
|
|
65 | - fix a bug where a timeout of zero caused the timer not to fire |
|
|
66 | in the libevent emulation (testcase by Péter Szabó). |
|
|
67 | - applied win32 fixes by Michael Lenaghan (also James Mansion). |
|
|
68 | - replace EV_MINIMAL by EV_FEATURES. |
|
|
69 | - prefer EPOLL_CTL_ADD over EPOLL_CTL_MOD in some more cases, as it |
|
|
70 | seems the former is *much* faster than the latter. |
|
|
71 | - linux kernel version detection (for inotify bug workarounds) |
|
|
72 | did not work properly. |
|
|
73 | - reduce the number of spurious wake-ups with the ports backend. |
|
|
74 | - remove dependency on sys/queue.h on freebsd (patch by Vanilla Hsu). |
|
|
75 | - do async init within ev_async_start, not ev_async_set, which avoids |
|
|
76 | an API quirk where the set function must be called in the C++ API |
|
|
77 | even when there is nothing to set. |
|
|
78 | - add (undocumented) EV_ENABLE when adding events with kqueue, |
|
|
79 | this might help with OS X, which seems to need it despite documenting |
|
|
80 | not to need it (helpfully pointed out by Tilghman Lesher). |
|
|
81 | - do not use poll by default on freebsd, it's broken (what isn't |
|
|
82 | on freebsd...). |
|
|
83 | - allow to embed epoll on kernels >= 2.6.32. |
|
|
84 | - configure now prepends -O3, not appends it, so one can still |
|
|
85 | override it. |
|
|
86 | - ev.pod: greatly expanded the portability section, added a porting |
|
|
87 | section, a description of watcher states and made lots of minor fixes. |
|
|
88 | - disable poll backend on AIX, the poll header spams the namespace |
|
|
89 | and it's not worth working around dead platforms (reported |
|
|
90 | and analyzed by Aivars Kalvans). |
|
|
91 | - improve header file compatibility of the standalone eventfd code |
|
|
92 | in an obscure case. |
|
|
93 | - implement EV_AVOID_STDIO option. |
|
|
94 | - do not use sscanf to parse linux version number (smaller, faster, |
|
|
95 | no sscanf dependency). |
|
|
96 | - new EV_CHILD_ENABLE and EV_SIGNAL_ENABLE configurable settings. |
|
|
97 | - update libev.m4 HAVE_CLOCK_SYSCALL test for newer glibcs. |
|
|
98 | - add section on accept() problems to the manpage. |
|
|
99 | - rename EV_TIMEOUT to EV_TIMER. |
|
|
100 | - rename ev_loop_count/depth/verify/loop/unloop. |
|
|
101 | - remove ev_default_destroy and ev_default_fork. |
|
|
102 | - switch to two-digit minor version. |
|
|
103 | - work around an apparent gentoo compiler bug. |
|
|
104 | - define _DARWIN_UNLIMITED_SELECT. just so. |
|
|
105 | - use enum instead of #define for most constants. |
|
|
106 | - improve compatibility to older C++ compilers. |
|
|
107 | - (experimental) ev_run/ev_default_loop/ev_break/ev_loop_new have now |
|
|
108 | default arguments when compiled as C++. |
|
|
109 | - enable automake dependency tracking. |
|
|
110 | - ev_loop_new no longer leaks memory when loop creation failed. |
|
|
111 | - new ev_cleanup watcher type. |
|
|
112 | |
|
|
113 | 3.9 Thu Dec 31 07:59:59 CET 2009 |
|
|
114 | - signalfd is no longer used by default and has to be requested |
|
|
115 | explicitly - this means that easy to catch bugs become hard to |
|
|
116 | catch race conditions, but the users have spoken. |
|
|
117 | - point out the unspecified signal mask in the documentation, and |
|
|
118 | that this is a race condition regardless of EV_SIGNALFD. |
3 | - backport inotify code to C89. |
119 | - backport inotify code to C89. |
|
|
120 | - inotify file descriptors could leak into child processes. |
|
|
121 | - ev_stat watchers could keep an erroneous extra ref on the loop, |
|
|
122 | preventing exit when unregistering all watchers (testcases |
|
|
123 | provided by ry@tinyclouds.org). |
4 | - implement EV_WIN32_HANDLE_TO_FD and EV_WIN32_CLOSE_FD configuration |
124 | - implement EV_WIN32_HANDLE_TO_FD and EV_WIN32_CLOSE_FD configuration |
5 | symbols to make ti easier for apps to do their own fd management. |
125 | symbols to make it easier for apps to do their own fd management. |
|
|
126 | - support EV_IDLE_ENABLE being disabled in ev++.h |
|
|
127 | (patch by Didier Spezia). |
|
|
128 | - take advantage of inotify_init1, if available, to set cloexec/nonblock |
|
|
129 | on fd creation, to avoid races. |
|
|
130 | - the signal handling pipe wasn't always initialised under windows |
|
|
131 | (analysed by lekma). |
|
|
132 | - changed minimum glibc requirement from glibc 2.9 to 2.7, for |
|
|
133 | signalfd. |
|
|
134 | - add missing string.h include (Denis F. Latypoff). |
|
|
135 | - only replace ev_stat.prev when we detect an actual difference, |
|
|
136 | so prev is (almost) always different to attr. this might |
|
|
137 | have caused the problems with 04_stat.t. |
|
|
138 | - add ev::timer->remaining () method to C++ API. |
6 | |
139 | |
7 | 3.8 Sun Aug 9 14:30:45 CEST 2009 |
140 | 3.8 Sun Aug 9 14:30:45 CEST 2009 |
8 | - incompatible change: do not necessarily reset signal handler |
141 | - incompatible change: do not necessarily reset signal handler |
9 | to SIG_DFL when a sighandler is stopped. |
142 | to SIG_DFL when a sighandler is stopped. |
10 | - ev_default_destroy did not properly free or zero some members, |
143 | - ev_default_destroy did not properly free or zero some members, |
11 | potentially causing crashes and memory corruption on repated |
144 | potentially causing crashes and memory corruption on repeated |
12 | ev_default_destroy/ev_default_loop calls. |
145 | ev_default_destroy/ev_default_loop calls. |
13 | - take advantage of signalfd on GNU/Linux systems. |
146 | - take advantage of signalfd on GNU/Linux systems. |
14 | - document that the signal mask might be in an unspecified |
147 | - document that the signal mask might be in an unspecified |
15 | state when using libev's signal handling. |
148 | state when using libev's signal handling. |
16 | - take advantage of some GNU/Linux calls to set cloexec/nonblock |
149 | - take advantage of some GNU/Linux calls to set cloexec/nonblock |
… | |
… | |
51 | - correctly zero the idx values for arrays, so destroying and |
184 | - correctly zero the idx values for arrays, so destroying and |
52 | reinitialising the default loop actually works (patch by |
185 | reinitialising the default loop actually works (patch by |
53 | Malek Hadj-Ali). |
186 | Malek Hadj-Ali). |
54 | - implement ev_suspend and ev_resume. |
187 | - implement ev_suspend and ev_resume. |
55 | - new EV_CUSTOM revents flag for use by applications. |
188 | - new EV_CUSTOM revents flag for use by applications. |
56 | - add documentation section about priorites. |
189 | - add documentation section about priorities. |
57 | - add a glossary to the dcoumentation. |
190 | - add a glossary to the dcoumentation. |
58 | - extend the ev_fork description slightly. |
191 | - extend the ev_fork description slightly. |
59 | - optimize a jump out of call_pending. |
192 | - optimize a jump out of call_pending. |
60 | |
193 | |
61 | 3.53 Sun Feb 15 02:38:20 CET 2009 |
194 | 3.53 Sun Feb 15 02:38:20 CET 2009 |
… | |
… | |
138 | do similar changes for event.c. |
271 | do similar changes for event.c. |
139 | |
272 | |
140 | 3.45 Tue Oct 21 21:59:26 CEST 2008 |
273 | 3.45 Tue Oct 21 21:59:26 CEST 2008 |
141 | - disable inotify usage on linux <2.6.25, as it is broken |
274 | - disable inotify usage on linux <2.6.25, as it is broken |
142 | (reported by Yoann Vandoorselaere). |
275 | (reported by Yoann Vandoorselaere). |
143 | - ev_stat errornously would try to add inotify watchers |
276 | - ev_stat erroneously would try to add inotify watchers |
144 | even when inotify wasn't available (this should only |
277 | even when inotify wasn't available (this should only |
145 | have a performance impact). |
278 | have a performance impact). |
146 | - ev_once now passes both timeout and io to the callback if both |
279 | - ev_once now passes both timeout and io to the callback if both |
147 | occur concurrently, instead of giving timeouts precedence. |
280 | occur concurrently, instead of giving timeouts precedence. |
148 | - disable EV_USE_INOTIFY when sys/inotify.h is too old. |
281 | - disable EV_USE_INOTIFY when sys/inotify.h is too old. |
… | |
… | |
216 | - enable inlining with gcc versions 2 and 3. |
349 | - enable inlining with gcc versions 2 and 3. |
217 | - work around broken poll implementations potentially |
350 | - work around broken poll implementations potentially |
218 | not clearing revents field in ev_poll (Brandon Black) |
351 | not clearing revents field in ev_poll (Brandon Black) |
219 | (no such systems are known at this time). |
352 | (no such systems are known at this time). |
220 | - work around a bug in realloc on openbsd and darwin, |
353 | - work around a bug in realloc on openbsd and darwin, |
221 | also makes the errornous valgrind complaints |
354 | also makes the erroneous valgrind complaints |
222 | go away (noted by various people). |
355 | go away (noted by various people). |
223 | - fix ev_async_pending, add c++ wrapper for ev_async |
356 | - fix ev_async_pending, add c++ wrapper for ev_async |
224 | (based on patch sent by Johannes Deisenhofer. |
357 | (based on patch sent by Johannes Deisenhofer). |
225 | - add sensible set method to ev::embed. |
358 | - add sensible set method to ev::embed. |
226 | - made integer constants type int in ev.h. |
359 | - made integer constants type int in ev.h. |
227 | |
360 | |
228 | 3.2 Wed Apr 2 17:11:19 CEST 2008 |
361 | 3.2 Wed Apr 2 17:11:19 CEST 2008 |
229 | - fix a 64 bit overflow issue in the select backend, |
362 | - fix a 64 bit overflow issue in the select backend, |