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

Comparing libev/Changes (file contents):
Revision 1.245 by root, Tue Dec 20 04:08:35 2011 UTC vs.
Revision 1.321 by root, Fri Dec 21 06:57:09 2018 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_loop_wakeup 34.25 Fri Dec 21 07:49:20 CET 2018
4 - INCOMPATIBLE CHANGE: EV_THROW was renamed to EV_NOEXCEPT
5 (EV_THROW sitll provided) and now uses noexcept on C++11 or newer.
6 - move the darwin select workaround highe rin ev.c, as newer versions of
7 darwin managed to break their broken select even more.
8 - ANDROID => __ANDROID__ (reported by enh@google.com).
9 - disable epoll_create1 on android because it has broken header files
10 and google is unwilling to fix them (reported by enh@google.com).
11 - avoid a minor compilation warning on win32.
12 - c++: remove deprecated dynamic throw() specifications.
13 - c++: improve the (unsupported) bad_loop exception class.
14 - backport perl ev_periodic example to C, untested.
15 - update libecb, biggets change is to include a memory fence
16 in ECB_MEMORY_FENCE_RELEASE on x86/amd64.
4 17
5TODO: EV_NO_THREADS/NO_SMP and/or EV_USE_PTHREADS or so. also document. 184.24 Wed Dec 28 05:19:55 CET 2016
6TODO: EV_STANDALONE == NO_HASSEL (do not use clock_gettime in ev_standalone) 19 - bump version to 4.24, as the release tarball inexplicably
7TODO: memory fences for clang 20 didn't have the right version in ev.h, even though the cvs-tagged
8TODO: fix repeat=0 in http://pod.tst.eu/http://cvs.schmorp.de/libev/ev.pod#Be_smart_about_timeouts 21 version did have the right one (reported by Ales Teska).
94.10 22
234.23 Wed Nov 16 18:23:41 CET 2016
24 - move some declarations at the beginning to help certain retarded
25 microsoft compilers, even though their documentation claims
26 otherwise (reported by Ruslan Osmanov).
27
284.22 Sun Dec 20 22:11:50 CET 2015
29 - when epoll detects unremovable fds in the fd set, rebuild
30 only the epoll descriptor, not the signal pipe, to avoid
31 SIGPIPE in ev_async_send. This doesn't solve it on fork,
32 so document what needs to be done in ev_loop_fork
33 (analyzed by Benjamin Mahler).
34 - remove superfluous sys/timeb.h include on win32
35 (analyzed by Jason Madden).
36 - updated libecb.
37
384.20 Sat Jun 20 13:01:43 CEST 2015
39 - prefer noexcept over throw () with C++ 11.
40 - update ecb.h due to incompatibilities with c11.
41 - fix a potential aliasing issue when reading and writing
42 watcher callbacks.
43
444.19 Thu Sep 25 08:18:25 CEST 2014
45 - ev.h wasn't valid C++ anymore, which tripped compilers other than
46 clang, msvc or gcc (analyzed by Raphael 'kena' Poss). Unfortunately,
47 C++ doesn't support typedefs for function pointers fully, so the affected
48 declarations have to spell out the types each time.
49 - when not using autoconf, tighten the check for clock_gettime and related
50 functionality.
51
524.18 Fri Sep 5 17:55:26 CEST 2014
53 - events on files were not always generated properly with the
54 epoll backend (testcase by Assaf Inbal).
55 - mark event pipe fd as cloexec after a fork (analyzed by Sami Farin).
56 - (ecb) support m68k, m88k and sh (patch by Miod Vallat).
57 - use a reasonable fallback for EV_NSIG instead of erroring out
58 when we can't detect the signal set size.
59 - in the absence of autoconf, do not use the clock syscall
60 on glibc >= 2.17 (avoids the syscall AND -lrt on systems
61 doing clock_gettime in userspace).
62 - ensure extern "C" function pointers are used for externally-visible
63 loop callbacks (not watcher callbacks yet).
64 - (ecb) work around memory barriers and volatile apparently both being
65 broken in visual studio 2008 and later (analysed and patch by Nicolas Noble).
66
674.15 Fri Mar 1 12:04:50 CET 2013
68 - destroying a non-default loop would stop the global waitpid
69 watcher (Denis Bilenko).
70 - queueing pending watchers of higher priority from a watcher now invokes
71 them in a timely fashion (reported by Denis Bilenko).
72 - add throw() to all libev functions that cannot throw exceptions, for
73 further code size decrease when compiling for C++.
74 - add throw () to callbacks that must not throw exceptions (allocator,
75 syserr, loop acquire/release, periodic reschedule cbs).
76 - fix event_base_loop return code, add event_get_callback, event_base_new,
77 event_base_get_method calls to improve libevent 1.x emulation and add
78 some libevent 2.x functionality (based on a patch by Jeff Davey).
79 - add more memory fences to fix a bug reported by Jeff Davey. Better
80 be overfenced than underprotected.
81 - ev_run now returns a boolean status (true meaning watchers are
82 still active).
83 - ev_once: undef EV_ERROR in ev_kqueue.c, to avoid clashing with
84 libev's EV_ERROR (reported by 191919).
85 - (ecb) add memory fence support for xlC (Darin McBride).
86 - (ecb) add memory fence support for gcc-mips (Anton Kirilov).
87 - (ecb) add memory fence support for gcc-alpha (Christian Weisgerber).
88 - work around some kernels losing file descriptors by leaking
89 the kqueue descriptor in the child.
90 - work around linux inotify not reporting IN_ATTRIB changes for directories
91 in many cases.
92 - include sys/syscall.h instead of plain syscall.h.
93 - check for io watcher loops in ev_verify, check for the most
94 common reported usage bug in ev_io_start.
95 - choose socket vs. WSASocket at compiletime using EV_USE_WSASOCKET.
96 - always use WSASend/WSARecv directly on windows, hoping that this
97 works in all cases (unlike read/write/send/recv...).
98 - try to detect signals around a fork faster (test program by
99 Denis Bilenko).
100 - work around recent glibc versions that leak memory in realloc.
101 - rename ev::embed::set to ev::embed::set_embed to avoid clashing
102 the watcher base set (loop) method.
103 - rewrite the async/signal pipe logic to always keep a valid fd, which
104 simplifies (and hopefully correctifies :) the race checking
105 on fork, at the cost of one extra fd.
106 - add fat, msdos, jffs2, ramfs, ntfs and btrfs to the list of
107 inotify-supporting filesystems.
108 - move orig_CFLAGS assignment to after AC_INIT, as newer autoconf
109 versions ignore it before
110 (https://bugzilla.redhat.com/show_bug.cgi?id=908096).
111 - add some untested android support.
112 - enum expressions must be of type int (reported by Juan Pablo L).
113
1144.11 Sat Feb 4 19:52:39 CET 2012
115 - INCOMPATIBLE CHANGE: ev_timer_again now clears the pending status, as
116 was documented already, but not implemented in the repeating case.
117 - new compiletime symbols: EV_NO_SMP and EV_NO_THREADS.
10 - fix a race where the workaround against the epoll fork bugs 118 - fix a race where the workaround against the epoll fork bugs
11 caused signals to not be handled anymore. 119 caused signals to not be handled anymore.
12 - correct backend_fudge for most backends, and implement a windows 120 - correct backend_fudge for most backends, and implement a windows
13 specific workaround to avoid looping because we call both 121 specific workaround to avoid looping because we call both
14 select and Sleep, both with different time resolutions. 122 select and Sleep, both with different time resolutions.
… …
27 - fix #3 "be smart about timeouts" to not "deadlock" when 135 - fix #3 "be smart about timeouts" to not "deadlock" when
28 timeout == now, also improve the section overall. 136 timeout == now, also improve the section overall.
29 - avoid "AVOIDING FINISHING BEFORE RETURNING" idiom. 137 - avoid "AVOIDING FINISHING BEFORE RETURNING" idiom.
30 - support new EV_API_STATIC mode to make all libev symbols 138 - support new EV_API_STATIC mode to make all libev symbols
31 static. 139 static.
140 - supply default CFLAGS of -g -O3 with gcc when original CFLAGS
141 were empty.
32 142
334.04 Wed Feb 16 09:01:51 CET 2011 1434.04 Wed Feb 16 09:01:51 CET 2011
34 - fix two problems in the native win32 backend, where reuse of fd's 144 - fix two problems in the native win32 backend, where reuse of fd's
35 with different underlying handles caused handles not to be removed 145 with different underlying handles caused handles not to be removed
36 or added to the select set (analyzed and tested by Bert Belder). 146 or added to the select set (analyzed and tested by Bert Belder).
… …
191 reinitialising the default loop actually works (patch by 301 reinitialising the default loop actually works (patch by
192 Malek Hadj-Ali). 302 Malek Hadj-Ali).
193 - implement ev_suspend and ev_resume. 303 - implement ev_suspend and ev_resume.
194 - new EV_CUSTOM revents flag for use by applications. 304 - new EV_CUSTOM revents flag for use by applications.
195 - add documentation section about priorities. 305 - add documentation section about priorities.
196 - add a glossary to the dcoumentation. 306 - add a glossary to the documentation.
197 - extend the ev_fork description slightly. 307 - extend the ev_fork description slightly.
198 - optimize a jump out of call_pending. 308 - optimize a jump out of call_pending.
199 309
2003.53 Sun Feb 15 02:38:20 CET 2009 3103.53 Sun Feb 15 02:38:20 CET 2009
201 - fix a bug in event pipe creation on win32 that would cause a 311 - fix a bug in event pipe creation on win32 that would cause a

Diff Legend

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