[ViewVC] Diff of: cvs/libev/ev.html

Comparing libev/ev.html (file contents):
Revision 1.6 by root, Mon Nov 12 08:12:14 2007 UTC vs.
Revision 1.52 by root, Tue Nov 27 19:41:52 2007 UTC

…		…
4	<head>	4	<head>
5	<title>libev</title>	5	<title>libev</title>
6	<meta name="description" content="Pod documentation for libev" />	6	<meta name="description" content="Pod documentation for libev" />
7	<meta name="inputfile" content="<standard input>" />	7	<meta name="inputfile" content="<standard input>" />
8	<meta name="outputfile" content="<standard output>" />	8	<meta name="outputfile" content="<standard output>" />
9	<meta name="created" content="Mon Nov 12 09:12:14 2007" />	9	<meta name="created" content="Tue Nov 27 20:38:24 2007" />
10	<meta name="generator" content="Pod::Xhtml 1.57" />	10	<meta name="generator" content="Pod::Xhtml 1.57" />
11	<link rel="stylesheet" href="http://res.tst.eu/pod.css"/></head>	11	<link rel="stylesheet" href="http://res.tst.eu/pod.css"/></head>
12	<body>	12	<body>
13	<div class="pod">	13	<div class="pod">
14	<!-- INDEX START -->	14	<!-- INDEX START -->
…		…
17	<ul><li><a href="#NAME">NAME</a></li>	17	<ul><li><a href="#NAME">NAME</a></li>
18	<li><a href="#SYNOPSIS">SYNOPSIS</a></li>	18	<li><a href="#SYNOPSIS">SYNOPSIS</a></li>
19	<li><a href="#DESCRIPTION">DESCRIPTION</a></li>	19	<li><a href="#DESCRIPTION">DESCRIPTION</a></li>
20	<li><a href="#FEATURES">FEATURES</a></li>	20	<li><a href="#FEATURES">FEATURES</a></li>
21	<li><a href="#CONVENTIONS">CONVENTIONS</a></li>	21	<li><a href="#CONVENTIONS">CONVENTIONS</a></li>
		22	<li><a href="#TIME_REPRESENTATION">TIME REPRESENTATION</a></li>
22	<li><a href="#TIME_AND_OTHER_GLOBAL_FUNCTIONS">TIME AND OTHER GLOBAL FUNCTIONS</a></li>	23	<li><a href="#GLOBAL_FUNCTIONS">GLOBAL FUNCTIONS</a></li>
23	<li><a href="#FUNCTIONS_CONTROLLING_THE_EVENT_LOOP">FUNCTIONS CONTROLLING THE EVENT LOOP</a></li>	24	<li><a href="#FUNCTIONS_CONTROLLING_THE_EVENT_LOOP">FUNCTIONS CONTROLLING THE EVENT LOOP</a></li>
24	<li><a href="#ANATOMY_OF_A_WATCHER">ANATOMY OF A WATCHER</a>	25	<li><a href="#ANATOMY_OF_A_WATCHER">ANATOMY OF A WATCHER</a>
		26	<ul><li><a href="#GENERIC_WATCHER_FUNCTIONS">GENERIC WATCHER FUNCTIONS</a></li>
25	<ul><li><a href="#ASSOCIATING_CUSTOM_DATA_WITH_A_WATCH">ASSOCIATING CUSTOM DATA WITH A WATCHER</a></li>	27	<li><a href="#ASSOCIATING_CUSTOM_DATA_WITH_A_WATCH">ASSOCIATING CUSTOM DATA WITH A WATCHER</a></li>
26	</ul>	28	</ul>
27	</li>	29	</li>
28	<li><a href="#WATCHER_TYPES">WATCHER TYPES</a>	30	<li><a href="#WATCHER_TYPES">WATCHER TYPES</a>
29	<ul><li><a href="#struct_ev_io_is_my_file_descriptor_r">struct ev_io - is my file descriptor readable or writable</a></li>	31	<ul><li><a href="#code_ev_io_code_is_this_file_descrip"><code>ev_io</code> - is this file descriptor readable or writable?</a></li>
30	<li><a href="#struct_ev_timer_relative_and_optiona">struct ev_timer - relative and optionally recurring timeouts</a></li>	32	<li><a href="#code_ev_timer_code_relative_and_opti"><code>ev_timer</code> - relative and optionally repeating timeouts</a></li>
31	<li><a href="#ev_periodic_to_cron_or_not_to_cron_i">ev_periodic - to cron or not to cron it</a></li>	33	<li><a href="#code_ev_periodic_code_to_cron_or_not"><code>ev_periodic</code> - to cron or not to cron?</a></li>
32	<li><a href="#ev_signal_signal_me_when_a_signal_ge">ev_signal - signal me when a signal gets signalled</a></li>	34	<li><a href="#code_ev_signal_code_signal_me_when_a"><code>ev_signal</code> - signal me when a signal gets signalled!</a></li>
33	<li><a href="#ev_child_wait_for_pid_status_changes">ev_child - wait for pid status changes</a></li>	35	<li><a href="#code_ev_child_code_watch_out_for_pro"><code>ev_child</code> - watch out for process status changes</a></li>
		36	<li><a href="#code_ev_stat_code_did_the_file_attri"><code>ev_stat</code> - did the file attributes just change?</a></li>
34	<li><a href="#ev_idle_when_you_ve_got_nothing_bett">ev_idle - when you've got nothing better to do</a></li>	37	<li><a href="#code_ev_idle_code_when_you_ve_got_no"><code>ev_idle</code> - when you've got nothing better to do...</a></li>
35	<li><a href="#prepare_and_check_your_hooks_into_th">prepare and check - your hooks into the event loop</a></li>	38	<li><a href="#code_ev_prepare_code_and_code_ev_che"><code>ev_prepare</code> and <code>ev_check</code> - customise your event loop!</a></li>
		39	<li><a href="#code_ev_embed_code_when_one_backend_"><code>ev_embed</code> - when one backend isn't enough...</a></li>
		40	<li><a href="#code_ev_fork_code_the_audacity_to_re"><code>ev_fork</code> - the audacity to resume the event loop after a fork</a></li>
36	</ul>	41	</ul>
37	</li>	42	</li>
38	<li><a href="#OTHER_FUNCTIONS">OTHER FUNCTIONS</a></li>	43	<li><a href="#OTHER_FUNCTIONS">OTHER FUNCTIONS</a></li>
		44	<li><a href="#LIBEVENT_EMULATION">LIBEVENT EMULATION</a></li>
		45	<li><a href="#C_SUPPORT">C++ SUPPORT</a></li>
		46	<li><a href="#MACRO_MAGIC">MACRO MAGIC</a></li>
		47	<li><a href="#EMBEDDING">EMBEDDING</a>
		48	<ul><li><a href="#FILESETS">FILESETS</a>
		49	<ul><li><a href="#CORE_EVENT_LOOP">CORE EVENT LOOP</a></li>
		50	<li><a href="#LIBEVENT_COMPATIBILITY_API">LIBEVENT COMPATIBILITY API</a></li>
		51	<li><a href="#AUTOCONF_SUPPORT">AUTOCONF SUPPORT</a></li>
		52	</ul>
		53	</li>
		54	<li><a href="#PREPROCESSOR_SYMBOLS_MACROS">PREPROCESSOR SYMBOLS/MACROS</a></li>
		55	<li><a href="#EXAMPLES">EXAMPLES</a></li>
		56	</ul>
		57	</li>
		58	<li><a href="#COMPLEXITIES">COMPLEXITIES</a></li>
39	<li><a href="#AUTHOR">AUTHOR</a>	59	<li><a href="#AUTHOR">AUTHOR</a>
40	</li>	60	</li>
41	</ul><hr />	61	</ul><hr />
42	<!-- INDEX END -->	62	<!-- INDEX END -->
43		63
…		…
72	<p>Libev supports select, poll, the linux-specific epoll and the bsd-specific	92	<p>Libev supports select, poll, the linux-specific epoll and the bsd-specific
73	kqueue mechanisms for file descriptor events, relative timers, absolute	93	kqueue mechanisms for file descriptor events, relative timers, absolute
74	timers with customised rescheduling, signal events, process status change	94	timers with customised rescheduling, signal events, process status change
75	events (related to SIGCHLD), and event watchers dealing with the event	95	events (related to SIGCHLD), and event watchers dealing with the event
76	loop mechanism itself (idle, prepare and check watchers). It also is quite	96	loop mechanism itself (idle, prepare and check watchers). It also is quite
77	fast (see a <a href="http://libev.schmorp.de/bench.html">benchmark</a> comparing it	97	fast (see this <a href="http://libev.schmorp.de/bench.html">benchmark</a> comparing
78	to libevent).</p>	98	it to libevent for example).</p>
79		99
80	</div>	100	</div>
81	<h1 id="CONVENTIONS">CONVENTIONS</h1><p><a href="#TOP" class="toplink">Top</a></p>	101	<h1 id="CONVENTIONS">CONVENTIONS</h1><p><a href="#TOP" class="toplink">Top</a></p>
82	<div id="CONVENTIONS_CONTENT">	102	<div id="CONVENTIONS_CONTENT">
83	<p>Libev is very configurable. In this manual the default configuration	103	<p>Libev is very configurable. In this manual the default configuration
84	will be described, which supports multiple event loops. For more info	104	will be described, which supports multiple event loops. For more info
85	about various configuraiton options please have a look at the file	105	about various configuration options please have a look at the file
86	<cite>README.embed</cite> in the libev distribution. If libev was configured without	106	<cite>README.embed</cite> in the libev distribution. If libev was configured without
87	support for multiple event loops, then all functions taking an initial	107	support for multiple event loops, then all functions taking an initial
88	argument of name <code>loop</code> (which is always of type <code>struct ev_loop *</code>)	108	argument of name <code>loop</code> (which is always of type <code>struct ev_loop *</code>)
89	will not have this argument.</p>	109	will not have this argument.</p>
90		110
91	</div>	111	</div>
92	<h1 id="TIME_AND_OTHER_GLOBAL_FUNCTIONS">TIME AND OTHER GLOBAL FUNCTIONS</h1><p><a href="#TOP" class="toplink">Top</a></p>	112	<h1 id="TIME_REPRESENTATION">TIME REPRESENTATION</h1><p><a href="#TOP" class="toplink">Top</a></p>
93	<div id="TIME_AND_OTHER_GLOBAL_FUNCTIONS_CONT">	113	<div id="TIME_REPRESENTATION_CONTENT">
94	<p>Libev represents time as a single floating point number, representing the	114	<p>Libev represents time as a single floating point number, representing the
95	(fractional) number of seconds since the (POSIX) epoch (somewhere near	115	(fractional) number of seconds since the (POSIX) epoch (somewhere near
96	the beginning of 1970, details are complicated, don't ask). This type is	116	the beginning of 1970, details are complicated, don't ask). This type is
97	called <code>ev_tstamp</code>, which is what you should use too. It usually aliases	117	called <code>ev_tstamp</code>, which is what you should use too. It usually aliases
98	to the double type in C.</p>	118	to the <code>double</code> type in C, and when you need to do any calculations on
		119	it, you should treat it as such.</p>
		120
		121	</div>
		122	<h1 id="GLOBAL_FUNCTIONS">GLOBAL FUNCTIONS</h1><p><a href="#TOP" class="toplink">Top</a></p>
		123	<div id="GLOBAL_FUNCTIONS_CONTENT">
		124	<p>These functions can be called anytime, even before initialising the
		125	library in any way.</p>
99	<dl>	126	<dl>
100	<dt>ev_tstamp ev_time ()</dt>	127	<dt>ev_tstamp ev_time ()</dt>
101	<dd>	128	<dd>
102	<p>Returns the current time as libev would use it.</p>	129	<p>Returns the current time as libev would use it. Please note that the
		130	<code>ev_now</code> function is usually faster and also often returns the timestamp
		131	you actually want to know.</p>
103	</dd>	132	</dd>
104	<dt>int ev_version_major ()</dt>	133	<dt>int ev_version_major ()</dt>
105	<dt>int ev_version_minor ()</dt>	134	<dt>int ev_version_minor ()</dt>
106	<dd>	135	<dd>
107	<p>You can find out the major and minor version numbers of the library	136	<p>You can find out the major and minor version numbers of the library
108	you linked against by calling the functions <code>ev_version_major</code> and	137	you linked against by calling the functions <code>ev_version_major</code> and
109	<code>ev_version_minor</code>. If you want, you can compare against the global	138	<code>ev_version_minor</code>. If you want, you can compare against the global
110	symbols <code>EV_VERSION_MAJOR</code> and <code>EV_VERSION_MINOR</code>, which specify the	139	symbols <code>EV_VERSION_MAJOR</code> and <code>EV_VERSION_MINOR</code>, which specify the
111	version of the library your program was compiled against.</p>	140	version of the library your program was compiled against.</p>
112	<p>Usually, its a good idea to terminate if the major versions mismatch,	141	<p>Usually, it's a good idea to terminate if the major versions mismatch,
113	as this indicates an incompatible change. Minor versions are usually	142	as this indicates an incompatible change. Minor versions are usually
114	compatible to older versions, so a larger minor version alone is usually	143	compatible to older versions, so a larger minor version alone is usually
115	not a problem.</p>	144	not a problem.</p>
		145	<p>Example: make sure we haven't accidentally been linked against the wrong
		146	version:</p>
		147	<pre> assert (("libev version mismatch",
		148	ev_version_major () == EV_VERSION_MAJOR
		149	&& ev_version_minor () >= EV_VERSION_MINOR));
		150
		151	</pre>
		152	</dd>
		153	<dt>unsigned int ev_supported_backends ()</dt>
116	</dd>	154	<dd>
		155	<p>Return the set of all backends (i.e. their corresponding <code>EV_BACKEND_*</code>
		156	value) compiled into this binary of libev (independent of their
		157	availability on the system you are running on). See <code>ev_default_loop</code> for
		158	a description of the set values.</p>
		159	<p>Example: make sure we have the epoll method, because yeah this is cool and
		160	a must have and can we have a torrent of it please!!!11</p>
		161	<pre> assert (("sorry, no epoll, no sex",
		162	ev_supported_backends () & EVBACKEND_EPOLL));
		163
		164	</pre>
		165	</dd>
		166	<dt>unsigned int ev_recommended_backends ()</dt>
		167	<dd>
		168	<p>Return the set of all backends compiled into this binary of libev and also
		169	recommended for this platform. This set is often smaller than the one
		170	returned by <code>ev_supported_backends</code>, as for example kqueue is broken on
		171	most BSDs and will not be autodetected unless you explicitly request it
		172	(assuming you know what you are doing). This is the set of backends that
		173	libev will probe for if you specify no backends explicitly.</p>
		174	</dd>
		175	<dt>unsigned int ev_embeddable_backends ()</dt>
		176	<dd>
		177	<p>Returns the set of backends that are embeddable in other event loops. This
		178	is the theoretical, all-platform, value. To find which backends
		179	might be supported on the current system, you would need to look at
		180	<code>ev_embeddable_backends () & ev_supported_backends ()</code>, likewise for
		181	recommended ones.</p>
		182	<p>See the description of <code>ev_embed</code> watchers for more info.</p>
		183	</dd>
117	<dt>ev_set_allocator (void (cb)(void *ptr, long size))</dt>	184	<dt>ev_set_allocator (void (cb)(void *ptr, size_t size))</dt>
118	<dd>	185	<dd>
119	<p>Sets the allocation function to use (the prototype is similar to the	186	<p>Sets the allocation function to use (the prototype and semantics are
120	realloc function). It is used to allocate and free memory (no surprises	187	identical to the realloc C function). It is used to allocate and free
121	here). If it returns zero when memory needs to be allocated, the library	188	memory (no surprises here). If it returns zero when memory needs to be
122	might abort or take some potentially destructive action. The default is	189	allocated, the library might abort or take some potentially destructive
123	your system realloc function.</p>	190	action. The default is your system realloc function.</p>
124	<p>You could override this function in high-availability programs to, say,	191	<p>You could override this function in high-availability programs to, say,
125	free some memory if it cannot allocate memory, to use a special allocator,	192	free some memory if it cannot allocate memory, to use a special allocator,
126	or even to sleep a while and retry until some memory is available.</p>	193	or even to sleep a while and retry until some memory is available.</p>
		194	<p>Example: replace the libev allocator with one that waits a bit and then
		195	retries: better than mine).</p>
		196	<pre> static void *
		197	persistent_realloc (void *ptr, size_t size)
		198	{
		199	for (;;)
		200	{
		201	void *newptr = realloc (ptr, size);
		202
		203	if (newptr)
		204	return newptr;
		205
		206	sleep (60);
		207	}
		208	}
		209
		210	...
		211	ev_set_allocator (persistent_realloc);
		212
		213	</pre>
127	</dd>	214	</dd>
128	<dt>ev_set_syserr_cb (void (cb)(const char msg));</dt>	215	<dt>ev_set_syserr_cb (void (cb)(const char msg));</dt>
129	<dd>	216	<dd>
130	<p>Set the callback function to call on a retryable syscall error (such	217	<p>Set the callback function to call on a retryable syscall error (such
131	as failed select, poll, epoll_wait). The message is a printable string	218	as failed select, poll, epoll_wait). The message is a printable string
132	indicating the system call or subsystem causing the problem. If this	219	indicating the system call or subsystem causing the problem. If this
133	callback is set, then libev will expect it to remedy the sitution, no	220	callback is set, then libev will expect it to remedy the sitution, no
134	matter what, when it returns. That is, libev will geenrally retry the	221	matter what, when it returns. That is, libev will generally retry the
135	requested operation, or, if the condition doesn't go away, do bad stuff	222	requested operation, or, if the condition doesn't go away, do bad stuff
136	(such as abort).</p>	223	(such as abort).</p>
		224	<p>Example: do the same thing as libev does internally:</p>
		225	<pre> static void
		226	fatal_error (const char *msg)
		227	{
		228	perror (msg);
		229	abort ();
		230	}
		231
		232	...
		233	ev_set_syserr_cb (fatal_error);
		234
		235	</pre>
137	</dd>	236	</dd>
138	</dl>	237	</dl>
139		238
140	</div>	239	</div>
141	<h1 id="FUNCTIONS_CONTROLLING_THE_EVENT_LOOP">FUNCTIONS CONTROLLING THE EVENT LOOP</h1><p><a href="#TOP" class="toplink">Top</a></p>	240	<h1 id="FUNCTIONS_CONTROLLING_THE_EVENT_LOOP">FUNCTIONS CONTROLLING THE EVENT LOOP</h1><p><a href="#TOP" class="toplink">Top</a></p>
142	<div id="FUNCTIONS_CONTROLLING_THE_EVENT_LOOP-2">	241	<div id="FUNCTIONS_CONTROLLING_THE_EVENT_LOOP-2">
143	<p>An event loop is described by a <code>struct ev_loop *</code>. The library knows two	242	<p>An event loop is described by a <code>struct ev_loop *</code>. The library knows two
144	types of such loops, the <i>default</i> loop, which supports signals and child	243	types of such loops, the <i>default</i> loop, which supports signals and child
145	events, and dynamically created loops which do not.</p>	244	events, and dynamically created loops which do not.</p>
146	<p>If you use threads, a common model is to run the default event loop	245	<p>If you use threads, a common model is to run the default event loop
147	in your main thread (or in a separate thrad) and for each thread you	246	in your main thread (or in a separate thread) and for each thread you
148	create, you also create another event loop. Libev itself does no lockign	247	create, you also create another event loop. Libev itself does no locking
149	whatsoever, so if you mix calls to different event loops, make sure you	248	whatsoever, so if you mix calls to the same event loop in different
150	lock (this is usually a bad idea, though, even if done right).</p>	249	threads, make sure you lock (this is usually a bad idea, though, even if
		250	done correctly, because it's hideous and inefficient).</p>
151	<dl>	251	<dl>
152	<dt>struct ev_loop *ev_default_loop (unsigned int flags)</dt>	252	<dt>struct ev_loop *ev_default_loop (unsigned int flags)</dt>
153	<dd>	253	<dd>
154	<p>This will initialise the default event loop if it hasn't been initialised	254	<p>This will initialise the default event loop if it hasn't been initialised
155	yet and return it. If the default loop could not be initialised, returns	255	yet and return it. If the default loop could not be initialised, returns
156	false. If it already was initialised it simply returns it (and ignores the	256	false. If it already was initialised it simply returns it (and ignores the
157	flags).</p>	257	flags. If that is troubling you, check <code>ev_backend ()</code> afterwards).</p>
158	<p>If you don't know what event loop to use, use the one returned from this	258	<p>If you don't know what event loop to use, use the one returned from this
159	function.</p>	259	function.</p>
160	<p>The flags argument can be used to specify special behaviour or specific	260	<p>The flags argument can be used to specify special behaviour or specific
161	backends to use, and is usually specified as 0 (or EVFLAG_AUTO)</p>	261	backends to use, and is usually specified as <code>0</code> (or <code>EVFLAG_AUTO</code>).</p>
162	<p>It supports the following flags:</p>	262	<p>The following flags are supported:</p>
163	<p>	263	<p>
164	<dl>	264	<dl>
165	<dt>EVFLAG_AUTO</dt>	265	<dt><code>EVFLAG_AUTO</code></dt>
166	<dd>	266	<dd>
167	<p>The default flags value. Use this if you have no clue (its the right	267	<p>The default flags value. Use this if you have no clue (it's the right
168	thing, believe me).</p>	268	thing, believe me).</p>
169	</dd>	269	</dd>
170	<dt>EVFLAG_NOENV</dt>	270	<dt><code>EVFLAG_NOENV</code></dt>
171	<dd>	271	<dd>
172	<p>If this flag bit is ored into the flag value then libev will <i>not</i> look	272	<p>If this flag bit is ored into the flag value (or the program runs setuid
173	at the environment variable <code>LIBEV_FLAGS</code>. Otherwise (the default), this	273	or setgid) then libev will <i>not</i> look at the environment variable
174	environment variable will override the flags completely. This is useful	274	<code>LIBEV_FLAGS</code>. Otherwise (the default), this environment variable will
		275	override the flags completely if it is found in the environment. This is
175	to try out specific backends to tets their performance, or to work around	276	useful to try out specific backends to test their performance, or to work
176	bugs.</p>	277	around bugs.</p>
177	</dd>	278	</dd>
178	<dt>EVMETHOD_SELECT portable select backend</dt>	279	<dt><code>EVBACKEND_SELECT</code> (value 1, portable select backend)</dt>
179	<dt>EVMETHOD_POLL poll backend (everywhere except windows)</dt>
180	<dt>EVMETHOD_EPOLL linux only</dt>
181	<dt>EVMETHOD_KQUEUE some bsds only</dt>
182	<dt>EVMETHOD_DEVPOLL solaris 8 only</dt>
183	<dt>EVMETHOD_PORT solaris 10 only</dt>
184	<dd>	280	<dd>
185	<p>If one or more of these are ored into the flags value, then only these	281	<p>This is your standard select(2) backend. Not <i>completely</i> standard, as
186	backends will be tried (in the reverse order as given here). If one are	282	libev tries to roll its own fd_set with no limits on the number of fds,
187	specified, any backend will do.</p>	283	but if that fails, expect a fairly low limit on the number of fds when
		284	using this backend. It doesn't scale too well (O(highest_fd)), but its usually
		285	the fastest backend for a low number of fds.</p>
		286	</dd>
		287	<dt><code>EVBACKEND_POLL</code> (value 2, poll backend, available everywhere except on windows)</dt>
		288	<dd>
		289	<p>And this is your standard poll(2) backend. It's more complicated than
		290	select, but handles sparse fds better and has no artificial limit on the
		291	number of fds you can use (except it will slow down considerably with a
		292	lot of inactive fds). It scales similarly to select, i.e. O(total_fds).</p>
		293	</dd>
		294	<dt><code>EVBACKEND_EPOLL</code> (value 4, Linux)</dt>
		295	<dd>
		296	<p>For few fds, this backend is a bit little slower than poll and select,
		297	but it scales phenomenally better. While poll and select usually scale like
		298	O(total_fds) where n is the total number of fds (or the highest fd), epoll scales
		299	either O(1) or O(active_fds).</p>
		300	<p>While stopping and starting an I/O watcher in the same iteration will
		301	result in some caching, there is still a syscall per such incident
		302	(because the fd could point to a different file description now), so its
		303	best to avoid that. Also, dup()ed file descriptors might not work very
		304	well if you register events for both fds.</p>
		305	<p>Please note that epoll sometimes generates spurious notifications, so you
		306	need to use non-blocking I/O or other means to avoid blocking when no data
		307	(or space) is available.</p>
		308	</dd>
		309	<dt><code>EVBACKEND_KQUEUE</code> (value 8, most BSD clones)</dt>
		310	<dd>
		311	<p>Kqueue deserves special mention, as at the time of this writing, it
		312	was broken on all BSDs except NetBSD (usually it doesn't work with
		313	anything but sockets and pipes, except on Darwin, where of course its
		314	completely useless). For this reason its not being "autodetected"
		315	unless you explicitly specify it explicitly in the flags (i.e. using
		316	<code>EVBACKEND_KQUEUE</code>).</p>
		317	<p>It scales in the same way as the epoll backend, but the interface to the
		318	kernel is more efficient (which says nothing about its actual speed, of
		319	course). While starting and stopping an I/O watcher does not cause an
		320	extra syscall as with epoll, it still adds up to four event changes per
		321	incident, so its best to avoid that.</p>
		322	</dd>
		323	<dt><code>EVBACKEND_DEVPOLL</code> (value 16, Solaris 8)</dt>
		324	<dd>
		325	<p>This is not implemented yet (and might never be).</p>
		326	</dd>
		327	<dt><code>EVBACKEND_PORT</code> (value 32, Solaris 10)</dt>
		328	<dd>
		329	<p>This uses the Solaris 10 port mechanism. As with everything on Solaris,
		330	it's really slow, but it still scales very well (O(active_fds)).</p>
		331	<p>Please note that solaris ports can result in a lot of spurious
		332	notifications, so you need to use non-blocking I/O or other means to avoid
		333	blocking when no data (or space) is available.</p>
		334	</dd>
		335	<dt><code>EVBACKEND_ALL</code></dt>
		336	<dd>
		337	<p>Try all backends (even potentially broken ones that wouldn't be tried
		338	with <code>EVFLAG_AUTO</code>). Since this is a mask, you can do stuff such as
		339	<code>EVBACKEND_ALL & ~EVBACKEND_KQUEUE</code>.</p>
188	</dd>	340	</dd>
189	</dl>	341	</dl>
190	</p>	342	</p>
		343	<p>If one or more of these are ored into the flags value, then only these
		344	backends will be tried (in the reverse order as given here). If none are
		345	specified, most compiled-in backend will be tried, usually in reverse
		346	order of their flag values :)</p>
		347	<p>The most typical usage is like this:</p>
		348	<pre> if (!ev_default_loop (0))
		349	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
		350
		351	</pre>
		352	<p>Restrict libev to the select and poll backends, and do not allow
		353	environment settings to be taken into account:</p>
		354	<pre> ev_default_loop (EVBACKEND_POLL \| EVBACKEND_SELECT \| EVFLAG_NOENV);
		355
		356	</pre>
		357	<p>Use whatever libev has to offer, but make sure that kqueue is used if
		358	available (warning, breaks stuff, best use only with your own private
		359	event loop and only if you know the OS supports your types of fds):</p>
		360	<pre> ev_default_loop (ev_recommended_backends () \| EVBACKEND_KQUEUE);
		361
		362	</pre>
191	</dd>	363	</dd>
192	<dt>struct ev_loop *ev_loop_new (unsigned int flags)</dt>	364	<dt>struct ev_loop *ev_loop_new (unsigned int flags)</dt>
193	<dd>	365	<dd>
194	<p>Similar to <code>ev_default_loop</code>, but always creates a new event loop that is	366	<p>Similar to <code>ev_default_loop</code>, but always creates a new event loop that is
195	always distinct from the default loop. Unlike the default loop, it cannot	367	always distinct from the default loop. Unlike the default loop, it cannot
196	handle signal and child watchers, and attempts to do so will be greeted by	368	handle signal and child watchers, and attempts to do so will be greeted by
197	undefined behaviour (or a failed assertion if assertions are enabled).</p>	369	undefined behaviour (or a failed assertion if assertions are enabled).</p>
		370	<p>Example: try to create a event loop that uses epoll and nothing else.</p>
		371	<pre> struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL \| EVFLAG_NOENV);
		372	if (!epoller)
		373	fatal ("no epoll found here, maybe it hides under your chair");
		374
		375	</pre>
198	</dd>	376	</dd>
199	<dt>ev_default_destroy ()</dt>	377	<dt>ev_default_destroy ()</dt>
200	<dd>	378	<dd>
201	<p>Destroys the default loop again (frees all memory and kernel state	379	<p>Destroys the default loop again (frees all memory and kernel state
202	etc.). This stops all registered event watchers (by not touching them in	380	etc.). None of the active event watchers will be stopped in the normal
203	any way whatsoever, although you cnanot rely on this :).</p>	381	sense, so e.g. <code>ev_is_active</code> might still return true. It is your
		382	responsibility to either stop all watchers cleanly yoursef <i>before</i>
		383	calling this function, or cope with the fact afterwards (which is usually
		384	the easiest thing, youc na just ignore the watchers and/or <code>free ()</code> them
		385	for example).</p>
204	</dd>	386	</dd>
205	<dt>ev_loop_destroy (loop)</dt>	387	<dt>ev_loop_destroy (loop)</dt>
206	<dd>	388	<dd>
207	<p>Like <code>ev_default_destroy</code>, but destroys an event loop created by an	389	<p>Like <code>ev_default_destroy</code>, but destroys an event loop created by an
208	earlier call to <code>ev_loop_new</code>.</p>	390	earlier call to <code>ev_loop_new</code>.</p>
…		…
211	<dd>	393	<dd>
212	<p>This function reinitialises the kernel state for backends that have	394	<p>This function reinitialises the kernel state for backends that have
213	one. Despite the name, you can call it anytime, but it makes most sense	395	one. Despite the name, you can call it anytime, but it makes most sense
214	after forking, in either the parent or child process (or both, but that	396	after forking, in either the parent or child process (or both, but that
215	again makes little sense).</p>	397	again makes little sense).</p>
216	<p>You <i>must</i> call this function after forking if and only if you want to	398	<p>You <i>must</i> call this function in the child process after forking if and
217	use the event library in both processes. If you just fork+exec, you don't	399	only if you want to use the event library in both processes. If you just
218	have to call it.</p>	400	fork+exec, you don't have to call it.</p>
219	<p>The function itself is quite fast and its usually not a problem to call	401	<p>The function itself is quite fast and it's usually not a problem to call
220	it just in case after a fork. To make this easy, the function will fit in	402	it just in case after a fork. To make this easy, the function will fit in
221	quite nicely into a call to <code>pthread_atfork</code>:</p>	403	quite nicely into a call to <code>pthread_atfork</code>:</p>
222	<pre> pthread_atfork (0, 0, ev_default_fork);	404	<pre> pthread_atfork (0, 0, ev_default_fork);
223		405
224	</pre>	406	</pre>
		407	<p>At the moment, <code>EVBACKEND_SELECT</code> and <code>EVBACKEND_POLL</code> are safe to use
		408	without calling this function, so if you force one of those backends you
		409	do not need to care.</p>
225	</dd>	410	</dd>
226	<dt>ev_loop_fork (loop)</dt>	411	<dt>ev_loop_fork (loop)</dt>
227	<dd>	412	<dd>
228	<p>Like <code>ev_default_fork</code>, but acts on an event loop created by	413	<p>Like <code>ev_default_fork</code>, but acts on an event loop created by
229	<code>ev_loop_new</code>. Yes, you have to call this on every allocated event loop	414	<code>ev_loop_new</code>. Yes, you have to call this on every allocated event loop
230	after fork, and how you do this is entirely your own problem.</p>	415	after fork, and how you do this is entirely your own problem.</p>
231	</dd>	416	</dd>
232	<dt>unsigned int ev_method (loop)</dt>	417	<dt>unsigned int ev_backend (loop)</dt>
233	<dd>	418	<dd>
234	<p>Returns one of the <code>EVMETHOD_*</code> flags indicating the event backend in	419	<p>Returns one of the <code>EVBACKEND_*</code> flags indicating the event backend in
235	use.</p>	420	use.</p>
236	</dd>	421	</dd>
237	<dt>ev_tstamp = ev_now (loop)</dt>	422	<dt>ev_tstamp ev_now (loop)</dt>
238	<dd>	423	<dd>
239	<p>Returns the current "event loop time", which is the time the event loop	424	<p>Returns the current "event loop time", which is the time the event loop
240	got events and started processing them. This timestamp does not change	425	received events and started processing them. This timestamp does not
241	as long as callbacks are being processed, and this is also the base time	426	change as long as callbacks are being processed, and this is also the base
242	used for relative timers. You can treat it as the timestamp of the event	427	time used for relative timers. You can treat it as the timestamp of the
243	occuring (or more correctly, the mainloop finding out about it).</p>	428	event occuring (or more correctly, libev finding out about it).</p>
244	</dd>	429	</dd>
245	<dt>ev_loop (loop, int flags)</dt>	430	<dt>ev_loop (loop, int flags)</dt>
246	<dd>	431	<dd>
247	<p>Finally, this is it, the event handler. This function usually is called	432	<p>Finally, this is it, the event handler. This function usually is called
248	after you initialised all your watchers and you want to start handling	433	after you initialised all your watchers and you want to start handling
249	events.</p>	434	events.</p>
250	<p>If the flags argument is specified as 0, it will not return until either	435	<p>If the flags argument is specified as <code>0</code>, it will not return until
251	no event watchers are active anymore or <code>ev_unloop</code> was called.</p>	436	either no event watchers are active anymore or <code>ev_unloop</code> was called.</p>
		437	<p>Please note that an explicit <code>ev_unloop</code> is usually better than
		438	relying on all watchers to be stopped when deciding when a program has
		439	finished (especially in interactive programs), but having a program that
		440	automatically loops as long as it has to and no longer by virtue of
		441	relying on its watchers stopping correctly is a thing of beauty.</p>
252	<p>A flags value of <code>EVLOOP_NONBLOCK</code> will look for new events, will handle	442	<p>A flags value of <code>EVLOOP_NONBLOCK</code> will look for new events, will handle
253	those events and any outstanding ones, but will not block your process in	443	those events and any outstanding ones, but will not block your process in
254	case there are no events.</p>	444	case there are no events and will return after one iteration of the loop.</p>
255	<p>A flags value of <code>EVLOOP_ONESHOT</code> will look for new events (waiting if	445	<p>A flags value of <code>EVLOOP_ONESHOT</code> will look for new events (waiting if
256	neccessary) and will handle those and any outstanding ones. It will block	446	neccessary) and will handle those and any outstanding ones. It will block
257	your process until at least one new event arrives.</p>	447	your process until at least one new event arrives, and will return after
258	<p>This flags value could be used to implement alternative looping	448	one iteration of the loop. This is useful if you are waiting for some
259	constructs, but the <code>prepare</code> and <code>check</code> watchers provide a better and	449	external event in conjunction with something not expressible using other
260	more generic mechanism.</p>	450	libev watchers. However, a pair of <code>ev_prepare</code>/<code>ev_check</code> watchers is
		451	usually a better approach for this kind of thing.</p>
		452	<p>Here are the gory details of what <code>ev_loop</code> does:</p>
		453	<pre> * If there are no active watchers (reference count is zero), return.
		454	- Queue prepare watchers and then call all outstanding watchers.
		455	- If we have been forked, recreate the kernel state.
		456	- Update the kernel state with all outstanding changes.
		457	- Update the "event loop time".
		458	- Calculate for how long to block.
		459	- Block the process, waiting for any events.
		460	- Queue all outstanding I/O (fd) events.
		461	- Update the "event loop time" and do time jump handling.
		462	- Queue all outstanding timers.
		463	- Queue all outstanding periodics.
		464	- If no events are pending now, queue all idle watchers.
		465	- Queue all check watchers.
		466	- Call all queued watchers in reverse order (i.e. check watchers first).
		467	Signals and child watchers are implemented as I/O watchers, and will
		468	be handled here by queueing them when their watcher gets executed.
		469	- If ev_unloop has been called or EVLOOP_ONESHOT or EVLOOP_NONBLOCK
		470	were used, return, otherwise continue with step *.
		471
		472	</pre>
		473	<p>Example: queue some jobs and then loop until no events are outsanding
		474	anymore.</p>
		475	<pre> ... queue jobs here, make sure they register event watchers as long
		476	... as they still have work to do (even an idle watcher will do..)
		477	ev_loop (my_loop, 0);
		478	... jobs done. yeah!
		479
		480	</pre>
261	</dd>	481	</dd>
262	<dt>ev_unloop (loop, how)</dt>	482	<dt>ev_unloop (loop, how)</dt>
263	<dd>	483	<dd>
264	<p>Can be used to make a call to <code>ev_loop</code> return early. The <code>how</code> argument	484	<p>Can be used to make a call to <code>ev_loop</code> return early (but only after it
		485	has processed all outstanding events). The <code>how</code> argument must be either
265	must be either <code>EVUNLOOP_ONCE</code>, which will make the innermost <code>ev_loop</code>	486	<code>EVUNLOOP_ONE</code>, which will make the innermost <code>ev_loop</code> call return, or
266	call return, or <code>EVUNLOOP_ALL</code>, which will make all nested <code>ev_loop</code>	487	<code>EVUNLOOP_ALL</code>, which will make all nested <code>ev_loop</code> calls return.</p>
267	calls return.</p>
268	</dd>	488	</dd>
269	<dt>ev_ref (loop)</dt>	489	<dt>ev_ref (loop)</dt>
270	<dt>ev_unref (loop)</dt>	490	<dt>ev_unref (loop)</dt>
271	<dd>	491	<dd>
272	<p>Ref/unref can be used to add or remove a refcount on the event loop: Every	492	<p>Ref/unref can be used to add or remove a reference count on the event
273	watcher keeps one reference. If you have a long-runing watcher you never	493	loop: Every watcher keeps one reference, and as long as the reference
274	unregister that should not keep ev_loop from running, ev_unref() after	494	count is nonzero, <code>ev_loop</code> will not return on its own. If you have
275	starting, and ev_ref() before stopping it. Libev itself uses this for	495	a watcher you never unregister that should not keep <code>ev_loop</code> from
276	example for its internal signal pipe: It is not visible to you as a user	496	returning, ev_unref() after starting, and ev_ref() before stopping it. For
277	and should not keep <code>ev_loop</code> from exiting if the work is done. It is	497	example, libev itself uses this for its internal signal pipe: It is not
278	also an excellent way to do this for generic recurring timers or from	498	visible to the libev user and should not keep <code>ev_loop</code> from exiting if
279	within third-party libraries. Just remember to unref after start and ref	499	no event watchers registered by it are active. It is also an excellent
280	before stop.</p>	500	way to do this for generic recurring timers or from within third-party
		501	libraries. Just remember to <i>unref after start</i> and <i>ref before stop</i>.</p>
		502	<p>Example: create a signal watcher, but keep it from keeping <code>ev_loop</code>
		503	running when nothing else is active.</p>
		504	<pre> struct dv_signal exitsig;
		505	ev_signal_init (&exitsig, sig_cb, SIGINT);
		506	ev_signal_start (myloop, &exitsig);
		507	evf_unref (myloop);
		508
		509	</pre>
		510	<p>Example: for some weird reason, unregister the above signal handler again.</p>
		511	<pre> ev_ref (myloop);
		512	ev_signal_stop (myloop, &exitsig);
		513
		514	</pre>
281	</dd>	515	</dd>
282	</dl>	516	</dl>
		517
		518
		519
		520
283		521
284	</div>	522	</div>
285	<h1 id="ANATOMY_OF_A_WATCHER">ANATOMY OF A WATCHER</h1><p><a href="#TOP" class="toplink">Top</a></p>	523	<h1 id="ANATOMY_OF_A_WATCHER">ANATOMY OF A WATCHER</h1><p><a href="#TOP" class="toplink">Top</a></p>
286	<div id="ANATOMY_OF_A_WATCHER_CONTENT">	524	<div id="ANATOMY_OF_A_WATCHER_CONTENT">
287	<p>A watcher is a structure that you create and register to record your	525	<p>A watcher is a structure that you create and register to record your
288	interest in some event. For instance, if you want to wait for STDIN to	526	interest in some event. For instance, if you want to wait for STDIN to
289	become readable, you would create an ev_io watcher for that:</p>	527	become readable, you would create an <code>ev_io</code> watcher for that:</p>
290	<pre> static void my_cb (struct ev_loop loop, struct ev_io w, int revents)	528	<pre> static void my_cb (struct ev_loop loop, struct ev_io w, int revents)
291	{	529	{
292	ev_io_stop (w);	530	ev_io_stop (w);
293	ev_unloop (loop, EVUNLOOP_ALL);	531	ev_unloop (loop, EVUNLOOP_ALL);
294	}	532	}
…		…
317	with a watcher-specific start function (<code>ev_<type>_start (loop, watcher	555	with a watcher-specific start function (<code>ev_<type>_start (loop, watcher
318	*)</code>), and you can stop watching for events at any time by calling the	556	*)</code>), and you can stop watching for events at any time by calling the
319	corresponding stop function (<code>ev_<type>_stop (loop, watcher *)</code>.</p>	557	corresponding stop function (<code>ev_<type>_stop (loop, watcher *)</code>.</p>
320	<p>As long as your watcher is active (has been started but not stopped) you	558	<p>As long as your watcher is active (has been started but not stopped) you
321	must not touch the values stored in it. Most specifically you must never	559	must not touch the values stored in it. Most specifically you must never
322	reinitialise it or call its set method.</p>	560	reinitialise it or call its <code>set</code> macro.</p>
323	<p>You cna check whether an event is active by calling the <code>ev_is_active
324	(watcher *)</code> macro. To see whether an event is outstanding (but the
325	callback for it has not been called yet) you cna use the <code>ev_is_pending
326	(watcher *)</code> macro.</p>
327	<p>Each and every callback receives the event loop pointer as first, the	561	<p>Each and every callback receives the event loop pointer as first, the
328	registered watcher structure as second, and a bitset of received events as	562	registered watcher structure as second, and a bitset of received events as
329	third argument.</p>	563	third argument.</p>
330	<p>The rceeived events usually include a single bit per event type received	564	<p>The received events usually include a single bit per event type received
331	(you can receive multiple events at the same time). The possible bit masks	565	(you can receive multiple events at the same time). The possible bit masks
332	are:</p>	566	are:</p>
333	<dl>	567	<dl>
334	<dt>EV_READ</dt>	568	<dt><code>EV_READ</code></dt>
335	<dt>EV_WRITE</dt>	569	<dt><code>EV_WRITE</code></dt>
336	<dd>	570	<dd>
337	<p>The file descriptor in the ev_io watcher has become readable and/or	571	<p>The file descriptor in the <code>ev_io</code> watcher has become readable and/or
338	writable.</p>	572	writable.</p>
339	</dd>	573	</dd>
340	<dt>EV_TIMEOUT</dt>	574	<dt><code>EV_TIMEOUT</code></dt>
341	<dd>
342	<p>The ev_timer watcher has timed out.</p>
343	</dd>	575	<dd>
344	<dt>EV_PERIODIC</dt>	576	<p>The <code>ev_timer</code> watcher has timed out.</p>
345	<dd>	577	</dd>
346	<p>The ev_periodic watcher has timed out.</p>	578	<dt><code>EV_PERIODIC</code></dt>
347	</dd>	579	<dd>
348	<dt>EV_SIGNAL</dt>	580	<p>The <code>ev_periodic</code> watcher has timed out.</p>
349	<dd>	581	</dd>
		582	<dt><code>EV_SIGNAL</code></dt>
		583	<dd>
350	<p>The signal specified in the ev_signal watcher has been received by a thread.</p>	584	<p>The signal specified in the <code>ev_signal</code> watcher has been received by a thread.</p>
351	</dd>
352	<dt>EV_CHILD</dt>
353	<dd>	585	</dd>
		586	<dt><code>EV_CHILD</code></dt>
		587	<dd>
354	<p>The pid specified in the ev_child watcher has received a status change.</p>	588	<p>The pid specified in the <code>ev_child</code> watcher has received a status change.</p>
355	</dd>
356	<dt>EV_IDLE</dt>
357	<dd>	589	</dd>
		590	<dt><code>EV_STAT</code></dt>
		591	<dd>
		592	<p>The path specified in the <code>ev_stat</code> watcher changed its attributes somehow.</p>
		593	</dd>
		594	<dt><code>EV_IDLE</code></dt>
		595	<dd>
358	<p>The ev_idle watcher has determined that you have nothing better to do.</p>	596	<p>The <code>ev_idle</code> watcher has determined that you have nothing better to do.</p>
359	</dd>
360	<dt>EV_PREPARE</dt>
361	<dt>EV_CHECK</dt>
362	<dd>	597	</dd>
		598	<dt><code>EV_PREPARE</code></dt>
		599	<dt><code>EV_CHECK</code></dt>
		600	<dd>
363	<p>All ev_prepare watchers are invoked just <i>before</i> <code>ev_loop</code> starts	601	<p>All <code>ev_prepare</code> watchers are invoked just <i>before</i> <code>ev_loop</code> starts
364	to gather new events, and all ev_check watchers are invoked just after	602	to gather new events, and all <code>ev_check</code> watchers are invoked just after
365	<code>ev_loop</code> has gathered them, but before it invokes any callbacks for any	603	<code>ev_loop</code> has gathered them, but before it invokes any callbacks for any
366	received events. Callbacks of both watcher types can start and stop as	604	received events. Callbacks of both watcher types can start and stop as
367	many watchers as they want, and all of them will be taken into account	605	many watchers as they want, and all of them will be taken into account
368	(for example, a ev_prepare watcher might start an idle watcher to keep	606	(for example, a <code>ev_prepare</code> watcher might start an idle watcher to keep
369	<code>ev_loop</code> from blocking).</p>	607	<code>ev_loop</code> from blocking).</p>
370	</dd>	608	</dd>
371	<dt>EV_ERROR</dt>	609	<dt><code>EV_EMBED</code></dt>
		610	<dd>
		611	<p>The embedded event loop specified in the <code>ev_embed</code> watcher needs attention.</p>
		612	</dd>
		613	<dt><code>EV_FORK</code></dt>
		614	<dd>
		615	<p>The event loop has been resumed in the child process after fork (see
		616	<code>ev_fork</code>).</p>
		617	</dd>
		618	<dt><code>EV_ERROR</code></dt>
372	<dd>	619	<dd>
373	<p>An unspecified error has occured, the watcher has been stopped. This might	620	<p>An unspecified error has occured, the watcher has been stopped. This might
374	happen because the watcher could not be properly started because libev	621	happen because the watcher could not be properly started because libev
375	ran out of memory, a file descriptor was found to be closed or any other	622	ran out of memory, a file descriptor was found to be closed or any other
376	problem. You best act on it by reporting the problem and somehow coping	623	problem. You best act on it by reporting the problem and somehow coping
…		…
382	programs, though, so beware.</p>	629	programs, though, so beware.</p>
383	</dd>	630	</dd>
384	</dl>	631	</dl>
385		632
386	</div>	633	</div>
		634	<h2 id="GENERIC_WATCHER_FUNCTIONS">GENERIC WATCHER FUNCTIONS</h2>
		635	<div id="GENERIC_WATCHER_FUNCTIONS_CONTENT">
		636	<p>In the following description, <code>TYPE</code> stands for the watcher type,
		637	e.g. <code>timer</code> for <code>ev_timer</code> watchers and <code>io</code> for <code>ev_io</code> watchers.</p>
		638	<dl>
		639	<dt><code>ev_init</code> (ev_TYPE *watcher, callback)</dt>
		640	<dd>
		641	<p>This macro initialises the generic portion of a watcher. The contents
		642	of the watcher object can be arbitrary (so <code>malloc</code> will do). Only
		643	the generic parts of the watcher are initialised, you <i>need</i> to call
		644	the type-specific <code>ev_TYPE_set</code> macro afterwards to initialise the
		645	type-specific parts. For each type there is also a <code>ev_TYPE_init</code> macro
		646	which rolls both calls into one.</p>
		647	<p>You can reinitialise a watcher at any time as long as it has been stopped
		648	(or never started) and there are no pending events outstanding.</p>
		649	<p>The callback is always of type <code>void ()(ev_loop loop, ev_TYPE *watcher,
		650	int revents)</code>.</p>
		651	</dd>
		652	<dt><code>ev_TYPE_set</code> (ev_TYPE *, [args])</dt>
		653	<dd>
		654	<p>This macro initialises the type-specific parts of a watcher. You need to
		655	call <code>ev_init</code> at least once before you call this macro, but you can
		656	call <code>ev_TYPE_set</code> any number of times. You must not, however, call this
		657	macro on a watcher that is active (it can be pending, however, which is a
		658	difference to the <code>ev_init</code> macro).</p>
		659	<p>Although some watcher types do not have type-specific arguments
		660	(e.g. <code>ev_prepare</code>) you still need to call its <code>set</code> macro.</p>
		661	</dd>
		662	<dt><code>ev_TYPE_init</code> (ev_TYPE *watcher, callback, [args])</dt>
		663	<dd>
		664	<p>This convinience macro rolls both <code>ev_init</code> and <code>ev_TYPE_set</code> macro
		665	calls into a single call. This is the most convinient method to initialise
		666	a watcher. The same limitations apply, of course.</p>
		667	</dd>
		668	<dt><code>ev_TYPE_start</code> (loop , ev_TYPE watcher)</dt>
		669	<dd>
		670	<p>Starts (activates) the given watcher. Only active watchers will receive
		671	events. If the watcher is already active nothing will happen.</p>
		672	</dd>
		673	<dt><code>ev_TYPE_stop</code> (loop , ev_TYPE watcher)</dt>
		674	<dd>
		675	<p>Stops the given watcher again (if active) and clears the pending
		676	status. It is possible that stopped watchers are pending (for example,
		677	non-repeating timers are being stopped when they become pending), but
		678	<code>ev_TYPE_stop</code> ensures that the watcher is neither active nor pending. If
		679	you want to free or reuse the memory used by the watcher it is therefore a
		680	good idea to always call its <code>ev_TYPE_stop</code> function.</p>
		681	</dd>
		682	<dt>bool ev_is_active (ev_TYPE *watcher)</dt>
		683	<dd>
		684	<p>Returns a true value iff the watcher is active (i.e. it has been started
		685	and not yet been stopped). As long as a watcher is active you must not modify
		686	it.</p>
		687	</dd>
		688	<dt>bool ev_is_pending (ev_TYPE *watcher)</dt>
		689	<dd>
		690	<p>Returns a true value iff the watcher is pending, (i.e. it has outstanding
		691	events but its callback has not yet been invoked). As long as a watcher
		692	is pending (but not active) you must not call an init function on it (but
		693	<code>ev_TYPE_set</code> is safe) and you must make sure the watcher is available to
		694	libev (e.g. you cnanot <code>free ()</code> it).</p>
		695	</dd>
		696	<dt>callback = ev_cb (ev_TYPE *watcher)</dt>
		697	<dd>
		698	<p>Returns the callback currently set on the watcher.</p>
		699	</dd>
		700	<dt>ev_cb_set (ev_TYPE *watcher, callback)</dt>
		701	<dd>
		702	<p>Change the callback. You can change the callback at virtually any time
		703	(modulo threads).</p>
		704	</dd>
		705	</dl>
		706
		707
		708
		709
		710
		711	</div>
387	<h2 id="ASSOCIATING_CUSTOM_DATA_WITH_A_WATCH">ASSOCIATING CUSTOM DATA WITH A WATCHER</h2>	712	<h2 id="ASSOCIATING_CUSTOM_DATA_WITH_A_WATCH">ASSOCIATING CUSTOM DATA WITH A WATCHER</h2>
388	<div id="ASSOCIATING_CUSTOM_DATA_WITH_A_WATCH-2">	713	<div id="ASSOCIATING_CUSTOM_DATA_WITH_A_WATCH-2">
389	<p>Each watcher has, by default, a member <code>void *data</code> that you can change	714	<p>Each watcher has, by default, a member <code>void *data</code> that you can change
390	and read at any time, libev will completely ignore it. This cna be used	715	and read at any time, libev will completely ignore it. This can be used
391	to associate arbitrary data with your watcher. If you need more data and	716	to associate arbitrary data with your watcher. If you need more data and
392	don't want to allocate memory and store a pointer to it in that data	717	don't want to allocate memory and store a pointer to it in that data
393	member, you can also "subclass" the watcher type and provide your own	718	member, you can also "subclass" the watcher type and provide your own
394	data:</p>	719	data:</p>
395	<pre> struct my_io	720	<pre> struct my_io
…		…
419		744
420	</div>	745	</div>
421	<h1 id="WATCHER_TYPES">WATCHER TYPES</h1><p><a href="#TOP" class="toplink">Top</a></p>	746	<h1 id="WATCHER_TYPES">WATCHER TYPES</h1><p><a href="#TOP" class="toplink">Top</a></p>
422	<div id="WATCHER_TYPES_CONTENT">	747	<div id="WATCHER_TYPES_CONTENT">
423	<p>This section describes each watcher in detail, but will not repeat	748	<p>This section describes each watcher in detail, but will not repeat
424	information given in the last section.</p>	749	information given in the last section. Any initialisation/set macros,
		750	functions and members specific to the watcher type are explained.</p>
		751	<p>Members are additionally marked with either <i>[read-only]</i>, meaning that,
		752	while the watcher is active, you can look at the member and expect some
		753	sensible content, but you must not modify it (you can modify it while the
		754	watcher is stopped to your hearts content), or <i>[read-write]</i>, which
		755	means you can expect it to have some sensible content while the watcher
		756	is active, but you can also modify it. Modifying it may not do something
		757	sensible or take immediate effect (or do anything at all), but libev will
		758	not crash or malfunction in any way.</p>
425		759
		760
		761
		762
		763
426	</div>	764	</div>
427	<h2 id="struct_ev_io_is_my_file_descriptor_r">struct ev_io - is my file descriptor readable or writable</h2>	765	<h2 id="code_ev_io_code_is_this_file_descrip"><code>ev_io</code> - is this file descriptor readable or writable?</h2>
428	<div id="struct_ev_io_is_my_file_descriptor_r-2">	766	<div id="code_ev_io_code_is_this_file_descrip-2">
429	<p>I/O watchers check whether a file descriptor is readable or writable	767	<p>I/O watchers check whether a file descriptor is readable or writable
430	in each iteration of the event loop (This behaviour is called	768	in each iteration of the event loop, or, more precisely, when reading
431	level-triggering because you keep receiving events as long as the	769	would not block the process and writing would at least be able to write
432	condition persists. Remember you cna stop the watcher if you don't want to	770	some data. This behaviour is called level-triggering because you keep
433	act on the event and neither want to receive future events).</p>	771	receiving events as long as the condition persists. Remember you can stop
		772	the watcher if you don't want to act on the event and neither want to
		773	receive future events.</p>
		774	<p>In general you can register as many read and/or write event watchers per
		775	fd as you want (as long as you don't confuse yourself). Setting all file
		776	descriptors to non-blocking mode is also usually a good idea (but not
		777	required if you know what you are doing).</p>
		778	<p>You have to be careful with dup'ed file descriptors, though. Some backends
		779	(the linux epoll backend is a notable example) cannot handle dup'ed file
		780	descriptors correctly if you register interest in two or more fds pointing
		781	to the same underlying file/socket/etc. description (that is, they share
		782	the same underlying "file open").</p>
		783	<p>If you must do this, then force the use of a known-to-be-good backend
		784	(at the time of this writing, this includes only <code>EVBACKEND_SELECT</code> and
		785	<code>EVBACKEND_POLL</code>).</p>
		786	<p>Another thing you have to watch out for is that it is quite easy to
		787	receive "spurious" readyness notifications, that is your callback might
		788	be called with <code>EV_READ</code> but a subsequent <code>read</code>(2) will actually block
		789	because there is no data. Not only are some backends known to create a
		790	lot of those (for example solaris ports), it is very easy to get into
		791	this situation even with a relatively standard program structure. Thus
		792	it is best to always use non-blocking I/O: An extra <code>read</code>(2) returning
		793	<code>EAGAIN</code> is far preferable to a program hanging until some data arrives.</p>
		794	<p>If you cannot run the fd in non-blocking mode (for example you should not
		795	play around with an Xlib connection), then you have to seperately re-test
		796	wether a file descriptor is really ready with a known-to-be good interface
		797	such as poll (fortunately in our Xlib example, Xlib already does this on
		798	its own, so its quite safe to use).</p>
434	<dl>	799	<dl>
435	<dt>ev_io_init (ev_io *, callback, int fd, int events)</dt>	800	<dt>ev_io_init (ev_io *, callback, int fd, int events)</dt>
436	<dt>ev_io_set (ev_io *, int fd, int events)</dt>	801	<dt>ev_io_set (ev_io *, int fd, int events)</dt>
437	<dd>	802	<dd>
438	<p>Configures an ev_io watcher. The fd is the file descriptor to rceeive	803	<p>Configures an <code>ev_io</code> watcher. The <code>fd</code> is the file descriptor to
439	events for and events is either <code>EV_READ</code>, <code>EV_WRITE</code> or <code>EV_READ \|	804	rceeive events for and events is either <code>EV_READ</code>, <code>EV_WRITE</code> or
440	EV_WRITE</code> to receive the given events.</p>	805	<code>EV_READ \| EV_WRITE</code> to receive the given events.</p>
		806	</dd>
		807	<dt>int fd [read-only]</dt>
		808	<dd>
		809	<p>The file descriptor being watched.</p>
		810	</dd>
		811	<dt>int events [read-only]</dt>
		812	<dd>
		813	<p>The events being watched.</p>
441	</dd>	814	</dd>
442	</dl>	815	</dl>
		816	<p>Example: call <code>stdin_readable_cb</code> when STDIN_FILENO has become, well
		817	readable, but only once. Since it is likely line-buffered, you could
		818	attempt to read a whole line in the callback:</p>
		819	<pre> static void
		820	stdin_readable_cb (struct ev_loop loop, struct ev_io w, int revents)
		821	{
		822	ev_io_stop (loop, w);
		823	.. read from stdin here (or from w->fd) and haqndle any I/O errors
		824	}
443		825
		826	...
		827	struct ev_loop *loop = ev_default_init (0);
		828	struct ev_io stdin_readable;
		829	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);
		830	ev_io_start (loop, &stdin_readable);
		831	ev_loop (loop, 0);
		832
		833
		834
		835
		836	</pre>
		837
444	</div>	838	</div>
445	<h2 id="struct_ev_timer_relative_and_optiona">struct ev_timer - relative and optionally recurring timeouts</h2>	839	<h2 id="code_ev_timer_code_relative_and_opti"><code>ev_timer</code> - relative and optionally repeating timeouts</h2>
446	<div id="struct_ev_timer_relative_and_optiona-2">	840	<div id="code_ev_timer_code_relative_and_opti-2">
447	<p>Timer watchers are simple relative timers that generate an event after a	841	<p>Timer watchers are simple relative timers that generate an event after a
448	given time, and optionally repeating in regular intervals after that.</p>	842	given time, and optionally repeating in regular intervals after that.</p>
449	<p>The timers are based on real time, that is, if you register an event that	843	<p>The timers are based on real time, that is, if you register an event that
450	times out after an hour and youreset your system clock to last years	844	times out after an hour and you reset your system clock to last years
451	time, it will still time out after (roughly) and hour. "Roughly" because	845	time, it will still time out after (roughly) and hour. "Roughly" because
452	detecting time jumps is hard, and soem inaccuracies are unavoidable (the	846	detecting time jumps is hard, and some inaccuracies are unavoidable (the
453	monotonic clock option helps a lot here).</p>	847	monotonic clock option helps a lot here).</p>
		848	<p>The relative timeouts are calculated relative to the <code>ev_now ()</code>
		849	time. This is usually the right thing as this timestamp refers to the time
		850	of the event triggering whatever timeout you are modifying/starting. If
		851	you suspect event processing to be delayed and you <i>need</i> to base the timeout
		852	on the current time, use something like this to adjust for this:</p>
		853	<pre> ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);
		854
		855	</pre>
		856	<p>The callback is guarenteed to be invoked only when its timeout has passed,
		857	but if multiple timers become ready during the same loop iteration then
		858	order of execution is undefined.</p>
454	<dl>	859	<dl>
455	<dt>ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)</dt>	860	<dt>ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)</dt>
456	<dt>ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)</dt>	861	<dt>ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)</dt>
457	<dd>	862	<dd>
458	<p>Configure the timer to trigger after <code>after</code> seconds. If <code>repeat</code> is	863	<p>Configure the timer to trigger after <code>after</code> seconds. If <code>repeat</code> is
…		…
460	timer will automatically be configured to trigger again <code>repeat</code> seconds	865	timer will automatically be configured to trigger again <code>repeat</code> seconds
461	later, again, and again, until stopped manually.</p>	866	later, again, and again, until stopped manually.</p>
462	<p>The timer itself will do a best-effort at avoiding drift, that is, if you	867	<p>The timer itself will do a best-effort at avoiding drift, that is, if you
463	configure a timer to trigger every 10 seconds, then it will trigger at	868	configure a timer to trigger every 10 seconds, then it will trigger at
464	exactly 10 second intervals. If, however, your program cannot keep up with	869	exactly 10 second intervals. If, however, your program cannot keep up with
465	the timer (ecause it takes longer than those 10 seconds to do stuff) the	870	the timer (because it takes longer than those 10 seconds to do stuff) the
466	timer will not fire more than once per event loop iteration.</p>	871	timer will not fire more than once per event loop iteration.</p>
467	</dd>	872	</dd>
468	<dt>ev_timer_again (loop)</dt>	873	<dt>ev_timer_again (loop)</dt>
469	<dd>	874	<dd>
470	<p>This will act as if the timer timed out and restart it again if it is	875	<p>This will act as if the timer timed out and restart it again if it is
471	repeating. The exact semantics are:</p>	876	repeating. The exact semantics are:</p>
472	<p>If the timer is started but nonrepeating, stop it.</p>	877	<p>If the timer is started but nonrepeating, stop it.</p>
473	<p>If the timer is repeating, either start it if necessary (with the repeat	878	<p>If the timer is repeating, either start it if necessary (with the repeat
474	value), or reset the running timer to the repeat value.</p>	879	value), or reset the running timer to the repeat value.</p>
475	<p>This sounds a bit complicated, but here is a useful and typical	880	<p>This sounds a bit complicated, but here is a useful and typical
476	example: Imagine you have a tcp connection and you want a so-called idle	881	example: Imagine you have a tcp connection and you want a so-called
477	timeout, that is, you want to be called when there have been, say, 60	882	idle timeout, that is, you want to be called when there have been,
478	seconds of inactivity on the socket. The easiest way to do this is to	883	say, 60 seconds of inactivity on the socket. The easiest way to do
479	configure an ev_timer with after=repeat=60 and calling ev_timer_again each	884	this is to configure an <code>ev_timer</code> with <code>after</code>=<code>repeat</code>=<code>60</code> and calling
480	time you successfully read or write some data. If you go into an idle	885	<code>ev_timer_again</code> each time you successfully read or write some data. If
481	state where you do not expect data to travel on the socket, you can stop	886	you go into an idle state where you do not expect data to travel on the
482	the timer, and again will automatically restart it if need be.</p>	887	socket, you can stop the timer, and again will automatically restart it if
		888	need be.</p>
		889	<p>You can also ignore the <code>after</code> value and <code>ev_timer_start</code> altogether
		890	and only ever use the <code>repeat</code> value:</p>
		891	<pre> ev_timer_init (timer, callback, 0., 5.);
		892	ev_timer_again (loop, timer);
		893	...
		894	timer->again = 17.;
		895	ev_timer_again (loop, timer);
		896	...
		897	timer->again = 10.;
		898	ev_timer_again (loop, timer);
		899
		900	</pre>
		901	<p>This is more efficient then stopping/starting the timer eahc time you want
		902	to modify its timeout value.</p>
		903	</dd>
		904	<dt>ev_tstamp repeat [read-write]</dt>
		905	<dd>
		906	<p>The current <code>repeat</code> value. Will be used each time the watcher times out
		907	or <code>ev_timer_again</code> is called and determines the next timeout (if any),
		908	which is also when any modifications are taken into account.</p>
483	</dd>	909	</dd>
484	</dl>	910	</dl>
		911	<p>Example: create a timer that fires after 60 seconds.</p>
		912	<pre> static void
		913	one_minute_cb (struct ev_loop loop, struct ev_timer w, int revents)
		914	{
		915	.. one minute over, w is actually stopped right here
		916	}
485		917
		918	struct ev_timer mytimer;
		919	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);
		920	ev_timer_start (loop, &mytimer);
		921
		922	</pre>
		923	<p>Example: create a timeout timer that times out after 10 seconds of
		924	inactivity.</p>
		925	<pre> static void
		926	timeout_cb (struct ev_loop loop, struct ev_timer w, int revents)
		927	{
		928	.. ten seconds without any activity
		929	}
		930
		931	struct ev_timer mytimer;
		932	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */
		933	ev_timer_again (&mytimer); /* start timer */
		934	ev_loop (loop, 0);
		935
		936	// and in some piece of code that gets executed on any "activity":
		937	// reset the timeout to start ticking again at 10 seconds
		938	ev_timer_again (&mytimer);
		939
		940
		941
		942
		943	</pre>
		944
486	</div>	945	</div>
487	<h2 id="ev_periodic_to_cron_or_not_to_cron_i">ev_periodic - to cron or not to cron it</h2>	946	<h2 id="code_ev_periodic_code_to_cron_or_not"><code>ev_periodic</code> - to cron or not to cron?</h2>
488	<div id="ev_periodic_to_cron_or_not_to_cron_i-2">	947	<div id="code_ev_periodic_code_to_cron_or_not-2">
489	<p>Periodic watchers are also timers of a kind, but they are very versatile	948	<p>Periodic watchers are also timers of a kind, but they are very versatile
490	(and unfortunately a bit complex).</p>	949	(and unfortunately a bit complex).</p>
491	<p>Unlike ev_timer's, they are not based on real time (or relative time)	950	<p>Unlike <code>ev_timer</code>'s, they are not based on real time (or relative time)
492	but on wallclock time (absolute time). You can tell a periodic watcher	951	but on wallclock time (absolute time). You can tell a periodic watcher
493	to trigger "at" some specific point in time. For example, if you tell a	952	to trigger "at" some specific point in time. For example, if you tell a
494	periodic watcher to trigger in 10 seconds (by specifiying e.g. c<ev_now ()	953	periodic watcher to trigger in 10 seconds (by specifiying e.g. <code>ev_now ()
495	+ 10.>) and then reset your system clock to the last year, then it will	954	+ 10.</code>) and then reset your system clock to the last year, then it will
496	take a year to trigger the event (unlike an ev_timer, which would trigger	955	take a year to trigger the event (unlike an <code>ev_timer</code>, which would trigger
497	roughly 10 seconds later and of course not if you reset your system time	956	roughly 10 seconds later and of course not if you reset your system time
498	again).</p>	957	again).</p>
499	<p>They can also be used to implement vastly more complex timers, such as	958	<p>They can also be used to implement vastly more complex timers, such as
500	triggering an event on eahc midnight, local time.</p>	959	triggering an event on eahc midnight, local time.</p>
		960	<p>As with timers, the callback is guarenteed to be invoked only when the
		961	time (<code>at</code>) has been passed, but if multiple periodic timers become ready
		962	during the same loop iteration then order of execution is undefined.</p>
501	<dl>	963	<dl>
502	<dt>ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)</dt>	964	<dt>ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)</dt>
503	<dt>ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)</dt>	965	<dt>ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)</dt>
504	<dd>	966	<dd>
505	<p>Lots of arguments, lets sort it out... There are basically three modes of	967	<p>Lots of arguments, lets sort it out... There are basically three modes of
506	operation, and we will explain them from simplest to complex:</p>	968	operation, and we will explain them from simplest to complex:</p>
507
508
509
510
511	<p>	969	<p>
512	<dl>	970	<dl>
513	<dt>* absolute timer (interval = reschedule_cb = 0)</dt>	971	<dt>* absolute timer (interval = reschedule_cb = 0)</dt>
514	<dd>	972	<dd>
515	<p>In this configuration the watcher triggers an event at the wallclock time	973	<p>In this configuration the watcher triggers an event at the wallclock time
…		…
527	<pre> ev_periodic_set (&periodic, 0., 3600., 0);	985	<pre> ev_periodic_set (&periodic, 0., 3600., 0);
528		986
529	</pre>	987	</pre>
530	<p>This doesn't mean there will always be 3600 seconds in between triggers,	988	<p>This doesn't mean there will always be 3600 seconds in between triggers,
531	but only that the the callback will be called when the system time shows a	989	but only that the the callback will be called when the system time shows a
532	full hour (UTC), or more correct, when the system time is evenly divisible	990	full hour (UTC), or more correctly, when the system time is evenly divisible
533	by 3600.</p>	991	by 3600.</p>
534	<p>Another way to think about it (for the mathematically inclined) is that	992	<p>Another way to think about it (for the mathematically inclined) is that
535	ev_periodic will try to run the callback in this mode at the next possible	993	<code>ev_periodic</code> will try to run the callback in this mode at the next possible
536	time where <code>time = at (mod interval)</code>, regardless of any time jumps.</p>	994	time where <code>time = at (mod interval)</code>, regardless of any time jumps.</p>
537	</dd>	995	</dd>
538	<dt>* manual reschedule mode (reschedule_cb = callback)</dt>	996	<dt>* manual reschedule mode (reschedule_cb = callback)</dt>
539	<dd>	997	<dd>
540	<p>In this mode the values for <code>interval</code> and <code>at</code> are both being	998	<p>In this mode the values for <code>interval</code> and <code>at</code> are both being
541	ignored. Instead, each time the periodic watcher gets scheduled, the	999	ignored. Instead, each time the periodic watcher gets scheduled, the
542	reschedule callback will be called with the watcher as first, and the	1000	reschedule callback will be called with the watcher as first, and the
543	current time as second argument.</p>	1001	current time as second argument.</p>
544	<p>NOTE: <i>This callback MUST NOT stop or destroy the periodic or any other	1002	<p>NOTE: <i>This callback MUST NOT stop or destroy any periodic watcher,
545	periodic watcher, ever, or make any event loop modificstions</i>. If you need	1003	ever, or make any event loop modifications</i>. If you need to stop it,
546	to stop it, return 1e30 (or so, fudge fudge) and stop it afterwards.</p>	1004	return <code>now + 1e30</code> (or so, fudge fudge) and stop it afterwards (e.g. by
		1005	starting a prepare watcher).</p>
547	<p>Its prototype is c<ev_tstamp (reschedule_cb)(struct ev_periodic w,	1006	<p>Its prototype is <code>ev_tstamp (reschedule_cb)(struct ev_periodic w,
548	ev_tstamp now)>, e.g.:</p>	1007	ev_tstamp now)</code>, e.g.:</p>
549	<pre> static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now)	1008	<pre> static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now)
550	{	1009	{
551	return now + 60.;	1010	return now + 60.;
552	}	1011	}
553		1012
554	</pre>	1013	</pre>
555	<p>It must return the next time to trigger, based on the passed time value	1014	<p>It must return the next time to trigger, based on the passed time value
556	(that is, the lowest time value larger than to the second argument). It	1015	(that is, the lowest time value larger than to the second argument). It
557	will usually be called just before the callback will be triggered, but	1016	will usually be called just before the callback will be triggered, but
558	might be called at other times, too.</p>	1017	might be called at other times, too.</p>
		1018	<p>NOTE: <i>This callback must always return a time that is later than the
		1019	passed <code>now</code> value</i>. Not even <code>now</code> itself will do, it <i>must</i> be larger.</p>
559	<p>This can be used to create very complex timers, such as a timer that	1020	<p>This can be used to create very complex timers, such as a timer that
560	triggers on each midnight, local time. To do this, you would calculate the	1021	triggers on each midnight, local time. To do this, you would calculate the
561	next midnight after <code>now</code> and return the timestamp value for this. How you do this	1022	next midnight after <code>now</code> and return the timestamp value for this. How
562	is, again, up to you (but it is not trivial).</p>	1023	you do this is, again, up to you (but it is not trivial, which is the main
		1024	reason I omitted it as an example).</p>
563	</dd>	1025	</dd>
564	</dl>	1026	</dl>
565	</p>	1027	</p>
566	</dd>	1028	</dd>
567	<dt>ev_periodic_again (loop, ev_periodic *)</dt>	1029	<dt>ev_periodic_again (loop, ev_periodic *)</dt>
…		…
569	<p>Simply stops and restarts the periodic watcher again. This is only useful	1031	<p>Simply stops and restarts the periodic watcher again. This is only useful
570	when you changed some parameters or the reschedule callback would return	1032	when you changed some parameters or the reschedule callback would return
571	a different time than the last time it was called (e.g. in a crond like	1033	a different time than the last time it was called (e.g. in a crond like
572	program when the crontabs have changed).</p>	1034	program when the crontabs have changed).</p>
573	</dd>	1035	</dd>
		1036	<dt>ev_tstamp interval [read-write]</dt>
		1037	<dd>
		1038	<p>The current interval value. Can be modified any time, but changes only
		1039	take effect when the periodic timer fires or <code>ev_periodic_again</code> is being
		1040	called.</p>
		1041	</dd>
		1042	<dt>ev_tstamp (reschedule_cb)(struct ev_periodic w, ev_tstamp now) [read-write]</dt>
		1043	<dd>
		1044	<p>The current reschedule callback, or <code>0</code>, if this functionality is
		1045	switched off. Can be changed any time, but changes only take effect when
		1046	the periodic timer fires or <code>ev_periodic_again</code> is being called.</p>
		1047	</dd>
574	</dl>	1048	</dl>
		1049	<p>Example: call a callback every hour, or, more precisely, whenever the
		1050	system clock is divisible by 3600. The callback invocation times have
		1051	potentially a lot of jittering, but good long-term stability.</p>
		1052	<pre> static void
		1053	clock_cb (struct ev_loop loop, struct ev_io w, int revents)
		1054	{
		1055	... its now a full hour (UTC, or TAI or whatever your clock follows)
		1056	}
575		1057
		1058	struct ev_periodic hourly_tick;
		1059	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);
		1060	ev_periodic_start (loop, &hourly_tick);
		1061
		1062	</pre>
		1063	<p>Example: the same as above, but use a reschedule callback to do it:</p>
		1064	<pre> #include <math.h>
		1065
		1066	static ev_tstamp
		1067	my_scheduler_cb (struct ev_periodic *w, ev_tstamp now)
		1068	{
		1069	return fmod (now, 3600.) + 3600.;
		1070	}
		1071
		1072	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);
		1073
		1074	</pre>
		1075	<p>Example: call a callback every hour, starting now:</p>
		1076	<pre> struct ev_periodic hourly_tick;
		1077	ev_periodic_init (&hourly_tick, clock_cb,
		1078	fmod (ev_now (loop), 3600.), 3600., 0);
		1079	ev_periodic_start (loop, &hourly_tick);
		1080
		1081
		1082
		1083
		1084	</pre>
		1085
576	</div>	1086	</div>
577	<h2 id="ev_signal_signal_me_when_a_signal_ge">ev_signal - signal me when a signal gets signalled</h2>	1087	<h2 id="code_ev_signal_code_signal_me_when_a"><code>ev_signal</code> - signal me when a signal gets signalled!</h2>
578	<div id="ev_signal_signal_me_when_a_signal_ge-2">	1088	<div id="code_ev_signal_code_signal_me_when_a-2">
579	<p>Signal watchers will trigger an event when the process receives a specific	1089	<p>Signal watchers will trigger an event when the process receives a specific
580	signal one or more times. Even though signals are very asynchronous, libev	1090	signal one or more times. Even though signals are very asynchronous, libev
581	will try its best to deliver signals synchronously, i.e. as part of the	1091	will try it's best to deliver signals synchronously, i.e. as part of the
582	normal event processing, like any other event.</p>	1092	normal event processing, like any other event.</p>
583	<p>You cna configure as many watchers as you like per signal. Only when the	1093	<p>You can configure as many watchers as you like per signal. Only when the
584	first watcher gets started will libev actually register a signal watcher	1094	first watcher gets started will libev actually register a signal watcher
585	with the kernel (thus it coexists with your own signal handlers as long	1095	with the kernel (thus it coexists with your own signal handlers as long
586	as you don't register any with libev). Similarly, when the last signal	1096	as you don't register any with libev). Similarly, when the last signal
587	watcher for a signal is stopped libev will reset the signal handler to	1097	watcher for a signal is stopped libev will reset the signal handler to
588	SIG_DFL (regardless of what it was set to before).</p>	1098	SIG_DFL (regardless of what it was set to before).</p>
…		…
591	<dt>ev_signal_set (ev_signal *, int signum)</dt>	1101	<dt>ev_signal_set (ev_signal *, int signum)</dt>
592	<dd>	1102	<dd>
593	<p>Configures the watcher to trigger on the given signal number (usually one	1103	<p>Configures the watcher to trigger on the given signal number (usually one
594	of the <code>SIGxxx</code> constants).</p>	1104	of the <code>SIGxxx</code> constants).</p>
595	</dd>	1105	</dd>
		1106	<dt>int signum [read-only]</dt>
		1107	<dd>
		1108	<p>The signal the watcher watches out for.</p>
		1109	</dd>
596	</dl>	1110	</dl>
597		1111
		1112
		1113
		1114
		1115
598	</div>	1116	</div>
599	<h2 id="ev_child_wait_for_pid_status_changes">ev_child - wait for pid status changes</h2>	1117	<h2 id="code_ev_child_code_watch_out_for_pro"><code>ev_child</code> - watch out for process status changes</h2>
600	<div id="ev_child_wait_for_pid_status_changes-2">	1118	<div id="code_ev_child_code_watch_out_for_pro-2">
601	<p>Child watchers trigger when your process receives a SIGCHLD in response to	1119	<p>Child watchers trigger when your process receives a SIGCHLD in response to
602	some child status changes (most typically when a child of yours dies).</p>	1120	some child status changes (most typically when a child of yours dies).</p>
603	<dl>	1121	<dl>
604	<dt>ev_child_init (ev_child *, callback, int pid)</dt>	1122	<dt>ev_child_init (ev_child *, callback, int pid)</dt>
605	<dt>ev_child_set (ev_child *, int pid)</dt>	1123	<dt>ev_child_set (ev_child *, int pid)</dt>
606	<dd>	1124	<dd>
607	<p>Configures the watcher to wait for status changes of process <code>pid</code> (or	1125	<p>Configures the watcher to wait for status changes of process <code>pid</code> (or
608	<i>any</i> process if <code>pid</code> is specified as <code>0</code>). The callback can look	1126	<i>any</i> process if <code>pid</code> is specified as <code>0</code>). The callback can look
609	at the <code>rstatus</code> member of the <code>ev_child</code> watcher structure to see	1127	at the <code>rstatus</code> member of the <code>ev_child</code> watcher structure to see
610	the status word (use the macros from <code>sys/wait.h</code>). The <code>rpid</code> member	1128	the status word (use the macros from <code>sys/wait.h</code> and see your systems
		1129	<code>waitpid</code> documentation). The <code>rpid</code> member contains the pid of the
611	contains the pid of the process causing the status change.</p>	1130	process causing the status change.</p>
		1131	</dd>
		1132	<dt>int pid [read-only]</dt>
		1133	<dd>
		1134	<p>The process id this watcher watches out for, or <code>0</code>, meaning any process id.</p>
		1135	</dd>
		1136	<dt>int rpid [read-write]</dt>
		1137	<dd>
		1138	<p>The process id that detected a status change.</p>
		1139	</dd>
		1140	<dt>int rstatus [read-write]</dt>
		1141	<dd>
		1142	<p>The process exit/trace status caused by <code>rpid</code> (see your systems
		1143	<code>waitpid</code> and <code>sys/wait.h</code> documentation for details).</p>
612	</dd>	1144	</dd>
613	</dl>	1145	</dl>
		1146	<p>Example: try to exit cleanly on SIGINT and SIGTERM.</p>
		1147	<pre> static void
		1148	sigint_cb (struct ev_loop loop, struct ev_signal w, int revents)
		1149	{
		1150	ev_unloop (loop, EVUNLOOP_ALL);
		1151	}
614		1152
		1153	struct ev_signal signal_watcher;
		1154	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);
		1155	ev_signal_start (loop, &sigint_cb);
		1156
		1157
		1158
		1159
		1160	</pre>
		1161
615	</div>	1162	</div>
		1163	<h2 id="code_ev_stat_code_did_the_file_attri"><code>ev_stat</code> - did the file attributes just change?</h2>
		1164	<div id="code_ev_stat_code_did_the_file_attri-2">
		1165	<p>This watches a filesystem path for attribute changes. That is, it calls
		1166	<code>stat</code> regularly (or when the OS says it changed) and sees if it changed
		1167	compared to the last time, invoking the callback if it did.</p>
		1168	<p>The path does not need to exist: changing from "path exists" to "path does
		1169	not exist" is a status change like any other. The condition "path does
		1170	not exist" is signified by the <code>st_nlink</code> field being zero (which is
		1171	otherwise always forced to be at least one) and all the other fields of
		1172	the stat buffer having unspecified contents.</p>
		1173	<p>Since there is no standard to do this, the portable implementation simply
		1174	calls <code>stat (2)</code> regulalry on the path to see if it changed somehow. You
		1175	can specify a recommended polling interval for this case. If you specify
		1176	a polling interval of <code>0</code> (highly recommended!) then a <i>suitable,
		1177	unspecified default</i> value will be used (which you can expect to be around
		1178	five seconds, although this might change dynamically). Libev will also
		1179	impose a minimum interval which is currently around <code>0.1</code>, but thats
		1180	usually overkill.</p>
		1181	<p>This watcher type is not meant for massive numbers of stat watchers,
		1182	as even with OS-supported change notifications, this can be
		1183	resource-intensive.</p>
		1184	<p>At the time of this writing, no specific OS backends are implemented, but
		1185	if demand increases, at least a kqueue and inotify backend will be added.</p>
		1186	<dl>
		1187	<dt>ev_stat_init (ev_stat , callback, const char path, ev_tstamp interval)</dt>
		1188	<dt>ev_stat_set (ev_stat , const char path, ev_tstamp interval)</dt>
		1189	<dd>
		1190	<p>Configures the watcher to wait for status changes of the given
		1191	<code>path</code>. The <code>interval</code> is a hint on how quickly a change is expected to
		1192	be detected and should normally be specified as <code>0</code> to let libev choose
		1193	a suitable value. The memory pointed to by <code>path</code> must point to the same
		1194	path for as long as the watcher is active.</p>
		1195	<p>The callback will be receive <code>EV_STAT</code> when a change was detected,
		1196	relative to the attributes at the time the watcher was started (or the
		1197	last change was detected).</p>
		1198	</dd>
		1199	<dt>ev_stat_stat (ev_stat *)</dt>
		1200	<dd>
		1201	<p>Updates the stat buffer immediately with new values. If you change the
		1202	watched path in your callback, you could call this fucntion to avoid
		1203	detecting this change (while introducing a race condition). Can also be
		1204	useful simply to find out the new values.</p>
		1205	</dd>
		1206	<dt>ev_statdata attr [read-only]</dt>
		1207	<dd>
		1208	<p>The most-recently detected attributes of the file. Although the type is of
		1209	<code>ev_statdata</code>, this is usually the (or one of the) <code>struct stat</code> types
		1210	suitable for your system. If the <code>st_nlink</code> member is <code>0</code>, then there
		1211	was some error while <code>stat</code>ing the file.</p>
		1212	</dd>
		1213	<dt>ev_statdata prev [read-only]</dt>
		1214	<dd>
		1215	<p>The previous attributes of the file. The callback gets invoked whenever
		1216	<code>prev</code> != <code>attr</code>.</p>
		1217	</dd>
		1218	<dt>ev_tstamp interval [read-only]</dt>
		1219	<dd>
		1220	<p>The specified interval.</p>
		1221	</dd>
		1222	<dt>const char *path [read-only]</dt>
		1223	<dd>
		1224	<p>The filesystem path that is being watched.</p>
		1225	</dd>
		1226	</dl>
		1227	<p>Example: Watch <code>/etc/passwd</code> for attribute changes.</p>
		1228	<pre> static void
		1229	passwd_cb (struct ev_loop loop, ev_stat w, int revents)
		1230	{
		1231	/* /etc/passwd changed in some way */
		1232	if (w->attr.st_nlink)
		1233	{
		1234	printf ("passwd current size %ld\n", (long)w->attr.st_size);
		1235	printf ("passwd current atime %ld\n", (long)w->attr.st_mtime);
		1236	printf ("passwd current mtime %ld\n", (long)w->attr.st_mtime);
		1237	}
		1238	else
		1239	/* you shalt not abuse printf for puts */
		1240	puts ("wow, /etc/passwd is not there, expect problems. "
		1241	"if this is windows, they already arrived\n");
		1242	}
		1243
		1244	...
		1245	ev_stat passwd;
		1246
		1247	ev_stat_init (&passwd, passwd_cb, "/etc/passwd");
		1248	ev_stat_start (loop, &passwd);
		1249
		1250
		1251
		1252
		1253	</pre>
		1254
		1255	</div>
616	<h2 id="ev_idle_when_you_ve_got_nothing_bett">ev_idle - when you've got nothing better to do</h2>	1256	<h2 id="code_ev_idle_code_when_you_ve_got_no"><code>ev_idle</code> - when you've got nothing better to do...</h2>
617	<div id="ev_idle_when_you_ve_got_nothing_bett-2">	1257	<div id="code_ev_idle_code_when_you_ve_got_no-2">
618	<p>Idle watchers trigger events when there are no other I/O or timer (or	1258	<p>Idle watchers trigger events when there are no other events are pending
619	periodic) events pending. That is, as long as your process is busy	1259	(prepare, check and other idle watchers do not count). That is, as long
620	handling sockets or timeouts it will not be called. But when your process	1260	as your process is busy handling sockets or timeouts (or even signals,
621	is idle all idle watchers are being called again and again - until	1261	imagine) it will not be triggered. But when your process is idle all idle
		1262	watchers are being called again and again, once per event loop iteration -
622	stopped, that is, or your process receives more events.</p>	1263	until stopped, that is, or your process receives more events and becomes
		1264	busy.</p>
623	<p>The most noteworthy effect is that as long as any idle watchers are	1265	<p>The most noteworthy effect is that as long as any idle watchers are
624	active, the process will not block when waiting for new events.</p>	1266	active, the process will not block when waiting for new events.</p>
625	<p>Apart from keeping your process non-blocking (which is a useful	1267	<p>Apart from keeping your process non-blocking (which is a useful
626	effect on its own sometimes), idle watchers are a good place to do	1268	effect on its own sometimes), idle watchers are a good place to do
627	"pseudo-background processing", or delay processing stuff to after the	1269	"pseudo-background processing", or delay processing stuff to after the
…		…
632	<p>Initialises and configures the idle watcher - it has no parameters of any	1274	<p>Initialises and configures the idle watcher - it has no parameters of any
633	kind. There is a <code>ev_idle_set</code> macro, but using it is utterly pointless,	1275	kind. There is a <code>ev_idle_set</code> macro, but using it is utterly pointless,
634	believe me.</p>	1276	believe me.</p>
635	</dd>	1277	</dd>
636	</dl>	1278	</dl>
		1279	<p>Example: dynamically allocate an <code>ev_idle</code>, start it, and in the
		1280	callback, free it. Alos, use no error checking, as usual.</p>
		1281	<pre> static void
		1282	idle_cb (struct ev_loop loop, struct ev_idle w, int revents)
		1283	{
		1284	free (w);
		1285	// now do something you wanted to do when the program has
		1286	// no longer asnything immediate to do.
		1287	}
637		1288
		1289	struct ev_idle *idle_watcher = malloc (sizeof (struct ev_idle));
		1290	ev_idle_init (idle_watcher, idle_cb);
		1291	ev_idle_start (loop, idle_cb);
		1292
		1293
		1294
		1295
		1296	</pre>
		1297
638	</div>	1298	</div>
639	<h2 id="prepare_and_check_your_hooks_into_th">prepare and check - your hooks into the event loop</h2>	1299	<h2 id="code_ev_prepare_code_and_code_ev_che"><code>ev_prepare</code> and <code>ev_check</code> - customise your event loop!</h2>
640	<div id="prepare_and_check_your_hooks_into_th-2">	1300	<div id="code_ev_prepare_code_and_code_ev_che-2">
641	<p>Prepare and check watchers usually (but not always) are used in	1301	<p>Prepare and check watchers are usually (but not always) used in tandem:
642	tandom. Prepare watchers get invoked before the process blocks and check	1302	prepare watchers get invoked before the process blocks and check watchers
643	watchers afterwards.</p>	1303	afterwards.</p>
		1304	<p>You <i>must not</i> call <code>ev_loop</code> or similar functions that enter
		1305	the current event loop from either <code>ev_prepare</code> or <code>ev_check</code>
		1306	watchers. Other loops than the current one are fine, however. The
		1307	rationale behind this is that you do not need to check for recursion in
		1308	those watchers, i.e. the sequence will always be <code>ev_prepare</code>, blocking,
		1309	<code>ev_check</code> so if you have one watcher of each kind they will always be
		1310	called in pairs bracketing the blocking call.</p>
644	<p>Their main purpose is to integrate other event mechanisms into libev. This	1311	<p>Their main purpose is to integrate other event mechanisms into libev and
645	could be used, for example, to track variable changes, implement your own	1312	their use is somewhat advanced. This could be used, for example, to track
646	watchers, integrate net-snmp or a coroutine library and lots more.</p>	1313	variable changes, implement your own watchers, integrate net-snmp or a
		1314	coroutine library and lots more. They are also occasionally useful if
		1315	you cache some data and want to flush it before blocking (for example,
		1316	in X programs you might want to do an <code>XFlush ()</code> in an <code>ev_prepare</code>
		1317	watcher).</p>
647	<p>This is done by examining in each prepare call which file descriptors need	1318	<p>This is done by examining in each prepare call which file descriptors need
648	to be watched by the other library, registering ev_io watchers for them	1319	to be watched by the other library, registering <code>ev_io</code> watchers for
649	and starting an ev_timer watcher for any timeouts (many libraries provide	1320	them and starting an <code>ev_timer</code> watcher for any timeouts (many libraries
650	just this functionality). Then, in the check watcher you check for any	1321	provide just this functionality). Then, in the check watcher you check for
651	events that occured (by making your callbacks set soem flags for example)	1322	any events that occured (by checking the pending status of all watchers
652	and call back into the library.</p>	1323	and stopping them) and call back into the library. The I/O and timer
		1324	callbacks will never actually be called (but must be valid nevertheless,
		1325	because you never know, you know?).</p>
653	<p>As another example, the perl Coro module uses these hooks to integrate	1326	<p>As another example, the Perl Coro module uses these hooks to integrate
654	coroutines into libev programs, by yielding to other active coroutines	1327	coroutines into libev programs, by yielding to other active coroutines
655	during each prepare and only letting the process block if no coroutines	1328	during each prepare and only letting the process block if no coroutines
656	are ready to run.</p>	1329	are ready to run (it's actually more complicated: it only runs coroutines
		1330	with priority higher than or equal to the event loop and one coroutine
		1331	of lower priority, but only once, using idle watchers to keep the event
		1332	loop from blocking if lower-priority coroutines are active, thus mapping
		1333	low-priority coroutines to idle/background tasks).</p>
657	<dl>	1334	<dl>
658	<dt>ev_prepare_init (ev_prepare *, callback)</dt>	1335	<dt>ev_prepare_init (ev_prepare *, callback)</dt>
659	<dt>ev_check_init (ev_check *, callback)</dt>	1336	<dt>ev_check_init (ev_check *, callback)</dt>
660	<dd>	1337	<dd>
661	<p>Initialises and configures the prepare or check watcher - they have no	1338	<p>Initialises and configures the prepare or check watcher - they have no
662	parameters of any kind. There are <code>ev_prepare_set</code> and <code>ev_check_set</code>	1339	parameters of any kind. There are <code>ev_prepare_set</code> and <code>ev_check_set</code>
663	macros, but using them is utterly, utterly pointless.</p>	1340	macros, but using them is utterly, utterly and completely pointless.</p>
664	</dd>	1341	</dd>
665	</dl>	1342	</dl>
		1343	<p>Example: To include a library such as adns, you would add IO watchers
		1344	and a timeout watcher in a prepare handler, as required by libadns, and
		1345	in a check watcher, destroy them and call into libadns. What follows is
		1346	pseudo-code only of course:</p>
		1347	<pre> static ev_io iow [nfd];
		1348	static ev_timer tw;
		1349
		1350	static void
		1351	io_cb (ev_loop loop, ev_io w, int revents)
		1352	{
		1353	// set the relevant poll flags
		1354	// could also call adns_processreadable etc. here
		1355	struct pollfd fd = (struct pollfd )w->data;
		1356	if (revents & EV_READ ) fd->revents \|= fd->events & POLLIN;
		1357	if (revents & EV_WRITE) fd->revents \|= fd->events & POLLOUT;
		1358	}
		1359
		1360	// create io watchers for each fd and a timer before blocking
		1361	static void
		1362	adns_prepare_cb (ev_loop loop, ev_prepare w, int revents)
		1363	{
		1364	int timeout = 3600000;truct pollfd fds [nfd];
		1365	// actual code will need to loop here and realloc etc.
		1366	adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));
		1367
		1368	/* the callback is illegal, but won't be called as we stop during check */
		1369	ev_timer_init (&tw, 0, timeout * 1e-3);
		1370	ev_timer_start (loop, &tw);
		1371
		1372	// create on ev_io per pollfd
		1373	for (int i = 0; i < nfd; ++i)
		1374	{
		1375	ev_io_init (iow + i, io_cb, fds [i].fd,
		1376	((fds [i].events & POLLIN ? EV_READ : 0)
		1377	\| (fds [i].events & POLLOUT ? EV_WRITE : 0)));
		1378
		1379	fds [i].revents = 0;
		1380	iow [i].data = fds + i;
		1381	ev_io_start (loop, iow + i);
		1382	}
		1383	}
		1384
		1385	// stop all watchers after blocking
		1386	static void
		1387	adns_check_cb (ev_loop loop, ev_check w, int revents)
		1388	{
		1389	ev_timer_stop (loop, &tw);
		1390
		1391	for (int i = 0; i < nfd; ++i)
		1392	ev_io_stop (loop, iow + i);
		1393
		1394	adns_afterpoll (adns, fds, nfd, timeval_from (ev_now (loop));
		1395	}
		1396
		1397
		1398
		1399
		1400	</pre>
		1401
		1402	</div>
		1403	<h2 id="code_ev_embed_code_when_one_backend_"><code>ev_embed</code> - when one backend isn't enough...</h2>
		1404	<div id="code_ev_embed_code_when_one_backend_-2">
		1405	<p>This is a rather advanced watcher type that lets you embed one event loop
		1406	into another (currently only <code>ev_io</code> events are supported in the embedded
		1407	loop, other types of watchers might be handled in a delayed or incorrect
		1408	fashion and must not be used).</p>
		1409	<p>There are primarily two reasons you would want that: work around bugs and
		1410	prioritise I/O.</p>
		1411	<p>As an example for a bug workaround, the kqueue backend might only support
		1412	sockets on some platform, so it is unusable as generic backend, but you
		1413	still want to make use of it because you have many sockets and it scales
		1414	so nicely. In this case, you would create a kqueue-based loop and embed it
		1415	into your default loop (which might use e.g. poll). Overall operation will
		1416	be a bit slower because first libev has to poll and then call kevent, but
		1417	at least you can use both at what they are best.</p>
		1418	<p>As for prioritising I/O: rarely you have the case where some fds have
		1419	to be watched and handled very quickly (with low latency), and even
		1420	priorities and idle watchers might have too much overhead. In this case
		1421	you would put all the high priority stuff in one loop and all the rest in
		1422	a second one, and embed the second one in the first.</p>
		1423	<p>As long as the watcher is active, the callback will be invoked every time
		1424	there might be events pending in the embedded loop. The callback must then
		1425	call <code>ev_embed_sweep (mainloop, watcher)</code> to make a single sweep and invoke
		1426	their callbacks (you could also start an idle watcher to give the embedded
		1427	loop strictly lower priority for example). You can also set the callback
		1428	to <code>0</code>, in which case the embed watcher will automatically execute the
		1429	embedded loop sweep.</p>
		1430	<p>As long as the watcher is started it will automatically handle events. The
		1431	callback will be invoked whenever some events have been handled. You can
		1432	set the callback to <code>0</code> to avoid having to specify one if you are not
		1433	interested in that.</p>
		1434	<p>Also, there have not currently been made special provisions for forking:
		1435	when you fork, you not only have to call <code>ev_loop_fork</code> on both loops,
		1436	but you will also have to stop and restart any <code>ev_embed</code> watchers
		1437	yourself.</p>
		1438	<p>Unfortunately, not all backends are embeddable, only the ones returned by
		1439	<code>ev_embeddable_backends</code> are, which, unfortunately, does not include any
		1440	portable one.</p>
		1441	<p>So when you want to use this feature you will always have to be prepared
		1442	that you cannot get an embeddable loop. The recommended way to get around
		1443	this is to have a separate variables for your embeddable loop, try to
		1444	create it, and if that fails, use the normal loop for everything:</p>
		1445	<pre> struct ev_loop *loop_hi = ev_default_init (0);
		1446	struct ev_loop *loop_lo = 0;
		1447	struct ev_embed embed;
		1448
		1449	// see if there is a chance of getting one that works
		1450	// (remember that a flags value of 0 means autodetection)
		1451	loop_lo = ev_embeddable_backends () & ev_recommended_backends ()
		1452	? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())
		1453	: 0;
		1454
		1455	// if we got one, then embed it, otherwise default to loop_hi
		1456	if (loop_lo)
		1457	{
		1458	ev_embed_init (&embed, 0, loop_lo);
		1459	ev_embed_start (loop_hi, &embed);
		1460	}
		1461	else
		1462	loop_lo = loop_hi;
		1463
		1464	</pre>
		1465	<dl>
		1466	<dt>ev_embed_init (ev_embed , callback, struct ev_loop embedded_loop)</dt>
		1467	<dt>ev_embed_set (ev_embed , callback, struct ev_loop embedded_loop)</dt>
		1468	<dd>
		1469	<p>Configures the watcher to embed the given loop, which must be
		1470	embeddable. If the callback is <code>0</code>, then <code>ev_embed_sweep</code> will be
		1471	invoked automatically, otherwise it is the responsibility of the callback
		1472	to invoke it (it will continue to be called until the sweep has been done,
		1473	if you do not want thta, you need to temporarily stop the embed watcher).</p>
		1474	</dd>
		1475	<dt>ev_embed_sweep (loop, ev_embed *)</dt>
		1476	<dd>
		1477	<p>Make a single, non-blocking sweep over the embedded loop. This works
		1478	similarly to <code>ev_loop (embedded_loop, EVLOOP_NONBLOCK)</code>, but in the most
		1479	apropriate way for embedded loops.</p>
		1480	</dd>
		1481	<dt>struct ev_loop *loop [read-only]</dt>
		1482	<dd>
		1483	<p>The embedded event loop.</p>
		1484	</dd>
		1485	</dl>
		1486
		1487
		1488
		1489
		1490
		1491	</div>
		1492	<h2 id="code_ev_fork_code_the_audacity_to_re"><code>ev_fork</code> - the audacity to resume the event loop after a fork</h2>
		1493	<div id="code_ev_fork_code_the_audacity_to_re-2">
		1494	<p>Fork watchers are called when a <code>fork ()</code> was detected (usually because
		1495	whoever is a good citizen cared to tell libev about it by calling
		1496	<code>ev_default_fork</code> or <code>ev_loop_fork</code>). The invocation is done before the
		1497	event loop blocks next and before <code>ev_check</code> watchers are being called,
		1498	and only in the child after the fork. If whoever good citizen calling
		1499	<code>ev_default_fork</code> cheats and calls it in the wrong process, the fork
		1500	handlers will be invoked, too, of course.</p>
		1501	<dl>
		1502	<dt>ev_fork_init (ev_signal *, callback)</dt>
		1503	<dd>
		1504	<p>Initialises and configures the fork watcher - it has no parameters of any
		1505	kind. There is a <code>ev_fork_set</code> macro, but using it is utterly pointless,
		1506	believe me.</p>
		1507	</dd>
		1508	</dl>
		1509
		1510
		1511
		1512
666		1513
667	</div>	1514	</div>
668	<h1 id="OTHER_FUNCTIONS">OTHER FUNCTIONS</h1><p><a href="#TOP" class="toplink">Top</a></p>	1515	<h1 id="OTHER_FUNCTIONS">OTHER FUNCTIONS</h1><p><a href="#TOP" class="toplink">Top</a></p>
669	<div id="OTHER_FUNCTIONS_CONTENT">	1516	<div id="OTHER_FUNCTIONS_CONTENT">
670	<p>There are some other fucntions of possible interest. Described. Here. Now.</p>	1517	<p>There are some other functions of possible interest. Described. Here. Now.</p>
671	<dl>	1518	<dl>
672	<dt>ev_once (loop, int fd, int events, ev_tstamp timeout, callback)</dt>	1519	<dt>ev_once (loop, int fd, int events, ev_tstamp timeout, callback)</dt>
673	<dd>	1520	<dd>
674	<p>This function combines a simple timer and an I/O watcher, calls your	1521	<p>This function combines a simple timer and an I/O watcher, calls your
675	callback on whichever event happens first and automatically stop both	1522	callback on whichever event happens first and automatically stop both
676	watchers. This is useful if you want to wait for a single event on an fd	1523	watchers. This is useful if you want to wait for a single event on an fd
677	or timeout without havign to allocate/configure/start/stop/free one or	1524	or timeout without having to allocate/configure/start/stop/free one or
678	more watchers yourself.</p>	1525	more watchers yourself.</p>
679	<p>If <code>fd</code> is less than 0, then no I/O watcher will be started and events is	1526	<p>If <code>fd</code> is less than 0, then no I/O watcher will be started and events
680	ignored. Otherwise, an ev_io watcher for the given <code>fd</code> and <code>events</code> set	1527	is being ignored. Otherwise, an <code>ev_io</code> watcher for the given <code>fd</code> and
681	will be craeted and started.</p>	1528	<code>events</code> set will be craeted and started.</p>
682	<p>If <code>timeout</code> is less than 0, then no timeout watcher will be	1529	<p>If <code>timeout</code> is less than 0, then no timeout watcher will be
683	started. Otherwise an ev_timer watcher with after = <code>timeout</code> (and repeat	1530	started. Otherwise an <code>ev_timer</code> watcher with after = <code>timeout</code> (and
684	= 0) will be started.</p>	1531	repeat = 0) will be started. While <code>0</code> is a valid timeout, it is of
		1532	dubious value.</p>
685	<p>The callback has the type <code>void (cb)(int revents, void arg)</code> and	1533	<p>The callback has the type <code>void (cb)(int revents, void arg)</code> and gets
686	gets passed an events set (normally a combination of EV_ERROR, EV_READ,	1534	passed an <code>revents</code> set like normal event callbacks (a combination of
687	EV_WRITE or EV_TIMEOUT) and the <code>arg</code> value passed to <code>ev_once</code>:</p>	1535	<code>EV_ERROR</code>, <code>EV_READ</code>, <code>EV_WRITE</code> or <code>EV_TIMEOUT</code>) and the <code>arg</code>
		1536	value passed to <code>ev_once</code>:</p>
688	<pre> static void stdin_ready (int revents, void *arg)	1537	<pre> static void stdin_ready (int revents, void *arg)
689	{	1538	{
690	if (revents & EV_TIMEOUT)	1539	if (revents & EV_TIMEOUT)
691	/* doh, nothing entered */	1540	/* doh, nothing entered */;
692	else if (revents & EV_READ)	1541	else if (revents & EV_READ)
693	/* stdin might have data for us, joy! */	1542	/* stdin might have data for us, joy! */;
694	}	1543	}
695		1544
696	ev_once (STDIN_FILENO, EV_READm 10., stdin_ready, 0);	1545	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);
697		1546
698	</pre>	1547	</pre>
699	</dd>	1548	</dd>
700	<dt>ev_feed_event (loop, watcher, int events)</dt>	1549	<dt>ev_feed_event (ev_loop , watcher , int revents)</dt>
701	<dd>	1550	<dd>
702	<p>Feeds the given event set into the event loop, as if the specified event	1551	<p>Feeds the given event set into the event loop, as if the specified event
703	has happened for the specified watcher (which must be a pointer to an	1552	had happened for the specified watcher (which must be a pointer to an
704	initialised but not necessarily active event watcher).</p>	1553	initialised but not necessarily started event watcher).</p>
705	</dd>	1554	</dd>
706	<dt>ev_feed_fd_event (loop, int fd, int revents)</dt>	1555	<dt>ev_feed_fd_event (ev_loop *, int fd, int revents)</dt>
707	<dd>	1556	<dd>
708	<p>Feed an event on the given fd, as if a file descriptor backend detected it.</p>	1557	<p>Feed an event on the given fd, as if a file descriptor backend detected
		1558	the given events it.</p>
709	</dd>	1559	</dd>
710	<dt>ev_feed_signal_event (loop, int signum)</dt>	1560	<dt>ev_feed_signal_event (ev_loop *loop, int signum)</dt>
711	<dd>	1561	<dd>
712	<p>Feed an event as if the given signal occured (loop must be the default loop!).</p>	1562	<p>Feed an event as if the given signal occured (<code>loop</code> must be the default
		1563	loop!).</p>
713	</dd>	1564	</dd>
714	</dl>	1565	</dl>
		1566
		1567
		1568
		1569
		1570
		1571	</div>
		1572	<h1 id="LIBEVENT_EMULATION">LIBEVENT EMULATION</h1><p><a href="#TOP" class="toplink">Top</a></p>
		1573	<div id="LIBEVENT_EMULATION_CONTENT">
		1574	<p>Libev offers a compatibility emulation layer for libevent. It cannot
		1575	emulate the internals of libevent, so here are some usage hints:</p>
		1576	<dl>
		1577	<dt>* Use it by including <event.h>, as usual.</dt>
		1578	<dt>* The following members are fully supported: ev_base, ev_callback,
		1579	ev_arg, ev_fd, ev_res, ev_events.</dt>
		1580	<dt>* Avoid using ev_flags and the EVLIST_*-macros, while it is
		1581	maintained by libev, it does not work exactly the same way as in libevent (consider
		1582	it a private API).</dt>
		1583	<dt>* Priorities are not currently supported. Initialising priorities
		1584	will fail and all watchers will have the same priority, even though there
		1585	is an ev_pri field.</dt>
		1586	<dt>* Other members are not supported.</dt>
		1587	<dt>* The libev emulation is <i>not</i> ABI compatible to libevent, you need
		1588	to use the libev header file and library.</dt>
		1589	</dl>
		1590
		1591	</div>
		1592	<h1 id="C_SUPPORT">C++ SUPPORT</h1><p><a href="#TOP" class="toplink">Top</a></p>
		1593	<div id="C_SUPPORT_CONTENT">
		1594	<p>Libev comes with some simplistic wrapper classes for C++ that mainly allow
		1595	you to use some convinience methods to start/stop watchers and also change
		1596	the callback model to a model using method callbacks on objects.</p>
		1597	<p>To use it,</p>
		1598	<pre> #include <ev++.h>
		1599
		1600	</pre>
		1601	<p>(it is not installed by default). This automatically includes <cite>ev.h</cite>
		1602	and puts all of its definitions (many of them macros) into the global
		1603	namespace. All C++ specific things are put into the <code>ev</code> namespace.</p>
		1604	<p>It should support all the same embedding options as <cite>ev.h</cite>, most notably
		1605	<code>EV_MULTIPLICITY</code>.</p>
		1606	<p>Here is a list of things available in the <code>ev</code> namespace:</p>
		1607	<dl>
		1608	<dt><code>ev::READ</code>, <code>ev::WRITE</code> etc.</dt>
		1609	<dd>
		1610	<p>These are just enum values with the same values as the <code>EV_READ</code> etc.
		1611	macros from <cite>ev.h</cite>.</p>
		1612	</dd>
		1613	<dt><code>ev::tstamp</code>, <code>ev::now</code></dt>
		1614	<dd>
		1615	<p>Aliases to the same types/functions as with the <code>ev_</code> prefix.</p>
		1616	</dd>
		1617	<dt><code>ev::io</code>, <code>ev::timer</code>, <code>ev::periodic</code>, <code>ev::idle</code>, <code>ev::sig</code> etc.</dt>
		1618	<dd>
		1619	<p>For each <code>ev_TYPE</code> watcher in <cite>ev.h</cite> there is a corresponding class of
		1620	the same name in the <code>ev</code> namespace, with the exception of <code>ev_signal</code>
		1621	which is called <code>ev::sig</code> to avoid clashes with the <code>signal</code> macro
		1622	defines by many implementations.</p>
		1623	<p>All of those classes have these methods:</p>
		1624	<p>
		1625	<dl>
		1626	<dt>ev::TYPE::TYPE (object , object::method )</dt>
		1627	<dt>ev::TYPE::TYPE (object , object::method , struct ev_loop *)</dt>
		1628	<dt>ev::TYPE::~TYPE</dt>
		1629	<dd>
		1630	<p>The constructor takes a pointer to an object and a method pointer to
		1631	the event handler callback to call in this class. The constructor calls
		1632	<code>ev_init</code> for you, which means you have to call the <code>set</code> method
		1633	before starting it. If you do not specify a loop then the constructor
		1634	automatically associates the default loop with this watcher.</p>
		1635	<p>The destructor automatically stops the watcher if it is active.</p>
		1636	</dd>
		1637	<dt>w->set (struct ev_loop *)</dt>
		1638	<dd>
		1639	<p>Associates a different <code>struct ev_loop</code> with this watcher. You can only
		1640	do this when the watcher is inactive (and not pending either).</p>
		1641	</dd>
		1642	<dt>w->set ([args])</dt>
		1643	<dd>
		1644	<p>Basically the same as <code>ev_TYPE_set</code>, with the same args. Must be
		1645	called at least once. Unlike the C counterpart, an active watcher gets
		1646	automatically stopped and restarted.</p>
		1647	</dd>
		1648	<dt>w->start ()</dt>
		1649	<dd>
		1650	<p>Starts the watcher. Note that there is no <code>loop</code> argument as the
		1651	constructor already takes the loop.</p>
		1652	</dd>
		1653	<dt>w->stop ()</dt>
		1654	<dd>
		1655	<p>Stops the watcher if it is active. Again, no <code>loop</code> argument.</p>
		1656	</dd>
		1657	<dt>w->again () <code>ev::timer</code>, <code>ev::periodic</code> only</dt>
		1658	<dd>
		1659	<p>For <code>ev::timer</code> and <code>ev::periodic</code>, this invokes the corresponding
		1660	<code>ev_TYPE_again</code> function.</p>
		1661	</dd>
		1662	<dt>w->sweep () <code>ev::embed</code> only</dt>
		1663	<dd>
		1664	<p>Invokes <code>ev_embed_sweep</code>.</p>
		1665	</dd>
		1666	<dt>w->update () <code>ev::stat</code> only</dt>
		1667	<dd>
		1668	<p>Invokes <code>ev_stat_stat</code>.</p>
		1669	</dd>
		1670	</dl>
		1671	</p>
		1672	</dd>
		1673	</dl>
		1674	<p>Example: Define a class with an IO and idle watcher, start one of them in
		1675	the constructor.</p>
		1676	<pre> class myclass
		1677	{
		1678	ev_io io; void io_cb (ev::io &w, int revents);
		1679	ev_idle idle void idle_cb (ev::idle &w, int revents);
		1680
		1681	myclass ();
		1682	}
		1683
		1684	myclass::myclass (int fd)
		1685	: io (this, &myclass::io_cb),
		1686	idle (this, &myclass::idle_cb)
		1687	{
		1688	io.start (fd, ev::READ);
		1689	}
		1690
		1691
		1692
		1693
		1694	</pre>
		1695
		1696	</div>
		1697	<h1 id="MACRO_MAGIC">MACRO MAGIC</h1><p><a href="#TOP" class="toplink">Top</a></p>
		1698	<div id="MACRO_MAGIC_CONTENT">
		1699	<p>Libev can be compiled with a variety of options, the most fundemantal is
		1700	<code>EV_MULTIPLICITY</code>. This option determines wether (most) functions and
		1701	callbacks have an initial <code>struct ev_loop *</code> argument.</p>
		1702	<p>To make it easier to write programs that cope with either variant, the
		1703	following macros are defined:</p>
		1704	<dl>
		1705	<dt><code>EV_A</code>, <code>EV_A_</code></dt>
		1706	<dd>
		1707	<p>This provides the loop <i>argument</i> for functions, if one is required ("ev
		1708	loop argument"). The <code>EV_A</code> form is used when this is the sole argument,
		1709	<code>EV_A_</code> is used when other arguments are following. Example:</p>
		1710	<pre> ev_unref (EV_A);
		1711	ev_timer_add (EV_A_ watcher);
		1712	ev_loop (EV_A_ 0);
		1713
		1714	</pre>
		1715	<p>It assumes the variable <code>loop</code> of type <code>struct ev_loop *</code> is in scope,
		1716	which is often provided by the following macro.</p>
		1717	</dd>
		1718	<dt><code>EV_P</code>, <code>EV_P_</code></dt>
		1719	<dd>
		1720	<p>This provides the loop <i>parameter</i> for functions, if one is required ("ev
		1721	loop parameter"). The <code>EV_P</code> form is used when this is the sole parameter,
		1722	<code>EV_P_</code> is used when other parameters are following. Example:</p>
		1723	<pre> // this is how ev_unref is being declared
		1724	static void ev_unref (EV_P);
		1725
		1726	// this is how you can declare your typical callback
		1727	static void cb (EV_P_ ev_timer *w, int revents)
		1728
		1729	</pre>
		1730	<p>It declares a parameter <code>loop</code> of type <code>struct ev_loop *</code>, quite
		1731	suitable for use with <code>EV_A</code>.</p>
		1732	</dd>
		1733	<dt><code>EV_DEFAULT</code>, <code>EV_DEFAULT_</code></dt>
		1734	<dd>
		1735	<p>Similar to the other two macros, this gives you the value of the default
		1736	loop, if multiple loops are supported ("ev loop default").</p>
		1737	</dd>
		1738	</dl>
		1739	<p>Example: Declare and initialise a check watcher, working regardless of
		1740	wether multiple loops are supported or not.</p>
		1741	<pre> static void
		1742	check_cb (EV_P_ ev_timer *w, int revents)
		1743	{
		1744	ev_check_stop (EV_A_ w);
		1745	}
		1746
		1747	ev_check check;
		1748	ev_check_init (&check, check_cb);
		1749	ev_check_start (EV_DEFAULT_ &check);
		1750	ev_loop (EV_DEFAULT_ 0);
		1751
		1752
		1753
		1754
		1755	</pre>
		1756
		1757	</div>
		1758	<h1 id="EMBEDDING">EMBEDDING</h1><p><a href="#TOP" class="toplink">Top</a></p>
		1759	<div id="EMBEDDING_CONTENT">
		1760	<p>Libev can (and often is) directly embedded into host
		1761	applications. Examples of applications that embed it include the Deliantra
		1762	Game Server, the EV perl module, the GNU Virtual Private Ethernet (gvpe)
		1763	and rxvt-unicode.</p>
		1764	<p>The goal is to enable you to just copy the neecssary files into your
		1765	source directory without having to change even a single line in them, so
		1766	you can easily upgrade by simply copying (or having a checked-out copy of
		1767	libev somewhere in your source tree).</p>
		1768
		1769	</div>
		1770	<h2 id="FILESETS">FILESETS</h2>
		1771	<div id="FILESETS_CONTENT">
		1772	<p>Depending on what features you need you need to include one or more sets of files
		1773	in your app.</p>
		1774
		1775	</div>
		1776	<h3 id="CORE_EVENT_LOOP">CORE EVENT LOOP</h3>
		1777	<div id="CORE_EVENT_LOOP_CONTENT">
		1778	<p>To include only the libev core (all the <code>ev_*</code> functions), with manual
		1779	configuration (no autoconf):</p>
		1780	<pre> #define EV_STANDALONE 1
		1781	#include "ev.c"
		1782
		1783	</pre>
		1784	<p>This will automatically include <cite>ev.h</cite>, too, and should be done in a
		1785	single C source file only to provide the function implementations. To use
		1786	it, do the same for <cite>ev.h</cite> in all files wishing to use this API (best
		1787	done by writing a wrapper around <cite>ev.h</cite> that you can include instead and
		1788	where you can put other configuration options):</p>
		1789	<pre> #define EV_STANDALONE 1
		1790	#include "ev.h"
		1791
		1792	</pre>
		1793	<p>Both header files and implementation files can be compiled with a C++
		1794	compiler (at least, thats a stated goal, and breakage will be treated
		1795	as a bug).</p>
		1796	<p>You need the following files in your source tree, or in a directory
		1797	in your include path (e.g. in libev/ when using -Ilibev):</p>
		1798	<pre> ev.h
		1799	ev.c
		1800	ev_vars.h
		1801	ev_wrap.h
		1802
		1803	ev_win32.c required on win32 platforms only
		1804
		1805	ev_select.c only when select backend is enabled (which is by default)
		1806	ev_poll.c only when poll backend is enabled (disabled by default)
		1807	ev_epoll.c only when the epoll backend is enabled (disabled by default)
		1808	ev_kqueue.c only when the kqueue backend is enabled (disabled by default)
		1809	ev_port.c only when the solaris port backend is enabled (disabled by default)
		1810
		1811	</pre>
		1812	<p><cite>ev.c</cite> includes the backend files directly when enabled, so you only need
		1813	to compile this single file.</p>
		1814
		1815	</div>
		1816	<h3 id="LIBEVENT_COMPATIBILITY_API">LIBEVENT COMPATIBILITY API</h3>
		1817	<div id="LIBEVENT_COMPATIBILITY_API_CONTENT">
		1818	<p>To include the libevent compatibility API, also include:</p>
		1819	<pre> #include "event.c"
		1820
		1821	</pre>
		1822	<p>in the file including <cite>ev.c</cite>, and:</p>
		1823	<pre> #include "event.h"
		1824
		1825	</pre>
		1826	<p>in the files that want to use the libevent API. This also includes <cite>ev.h</cite>.</p>
		1827	<p>You need the following additional files for this:</p>
		1828	<pre> event.h
		1829	event.c
		1830
		1831	</pre>
		1832
		1833	</div>
		1834	<h3 id="AUTOCONF_SUPPORT">AUTOCONF SUPPORT</h3>
		1835	<div id="AUTOCONF_SUPPORT_CONTENT">
		1836	<p>Instead of using <code>EV_STANDALONE=1</code> and providing your config in
		1837	whatever way you want, you can also <code>m4_include([libev.m4])</code> in your
		1838	<cite>configure.ac</cite> and leave <code>EV_STANDALONE</code> undefined. <cite>ev.c</cite> will then
		1839	include <cite>config.h</cite> and configure itself accordingly.</p>
		1840	<p>For this of course you need the m4 file:</p>
		1841	<pre> libev.m4
		1842
		1843	</pre>
		1844
		1845	</div>
		1846	<h2 id="PREPROCESSOR_SYMBOLS_MACROS">PREPROCESSOR SYMBOLS/MACROS</h2>
		1847	<div id="PREPROCESSOR_SYMBOLS_MACROS_CONTENT">
		1848	<p>Libev can be configured via a variety of preprocessor symbols you have to define
		1849	before including any of its files. The default is not to build for multiplicity
		1850	and only include the select backend.</p>
		1851	<dl>
		1852	<dt>EV_STANDALONE</dt>
		1853	<dd>
		1854	<p>Must always be <code>1</code> if you do not use autoconf configuration, which
		1855	keeps libev from including <cite>config.h</cite>, and it also defines dummy
		1856	implementations for some libevent functions (such as logging, which is not
		1857	supported). It will also not define any of the structs usually found in
		1858	<cite>event.h</cite> that are not directly supported by the libev core alone.</p>
		1859	</dd>
		1860	<dt>EV_USE_MONOTONIC</dt>
		1861	<dd>
		1862	<p>If defined to be <code>1</code>, libev will try to detect the availability of the
		1863	monotonic clock option at both compiletime and runtime. Otherwise no use
		1864	of the monotonic clock option will be attempted. If you enable this, you
		1865	usually have to link against librt or something similar. Enabling it when
		1866	the functionality isn't available is safe, though, althoguh you have
		1867	to make sure you link against any libraries where the <code>clock_gettime</code>
		1868	function is hiding in (often <cite>-lrt</cite>).</p>
		1869	</dd>
		1870	<dt>EV_USE_REALTIME</dt>
		1871	<dd>
		1872	<p>If defined to be <code>1</code>, libev will try to detect the availability of the
		1873	realtime clock option at compiletime (and assume its availability at
		1874	runtime if successful). Otherwise no use of the realtime clock option will
		1875	be attempted. This effectively replaces <code>gettimeofday</code> by <code>clock_get
		1876	(CLOCK_REALTIME, ...)</code> and will not normally affect correctness. See tzhe note about libraries
		1877	in the description of <code>EV_USE_MONOTONIC</code>, though.</p>
		1878	</dd>
		1879	<dt>EV_USE_SELECT</dt>
		1880	<dd>
		1881	<p>If undefined or defined to be <code>1</code>, libev will compile in support for the
		1882	<code>select</code>(2) backend. No attempt at autodetection will be done: if no
		1883	other method takes over, select will be it. Otherwise the select backend
		1884	will not be compiled in.</p>
		1885	</dd>
		1886	<dt>EV_SELECT_USE_FD_SET</dt>
		1887	<dd>
		1888	<p>If defined to <code>1</code>, then the select backend will use the system <code>fd_set</code>
		1889	structure. This is useful if libev doesn't compile due to a missing
		1890	<code>NFDBITS</code> or <code>fd_mask</code> definition or it misguesses the bitset layout on
		1891	exotic systems. This usually limits the range of file descriptors to some
		1892	low limit such as 1024 or might have other limitations (winsocket only
		1893	allows 64 sockets). The <code>FD_SETSIZE</code> macro, set before compilation, might
		1894	influence the size of the <code>fd_set</code> used.</p>
		1895	</dd>
		1896	<dt>EV_SELECT_IS_WINSOCKET</dt>
		1897	<dd>
		1898	<p>When defined to <code>1</code>, the select backend will assume that
		1899	select/socket/connect etc. don't understand file descriptors but
		1900	wants osf handles on win32 (this is the case when the select to
		1901	be used is the winsock select). This means that it will call
		1902	<code>_get_osfhandle</code> on the fd to convert it to an OS handle. Otherwise,
		1903	it is assumed that all these functions actually work on fds, even
		1904	on win32. Should not be defined on non-win32 platforms.</p>
		1905	</dd>
		1906	<dt>EV_USE_POLL</dt>
		1907	<dd>
		1908	<p>If defined to be <code>1</code>, libev will compile in support for the <code>poll</code>(2)
		1909	backend. Otherwise it will be enabled on non-win32 platforms. It
		1910	takes precedence over select.</p>
		1911	</dd>
		1912	<dt>EV_USE_EPOLL</dt>
		1913	<dd>
		1914	<p>If defined to be <code>1</code>, libev will compile in support for the Linux
		1915	<code>epoll</code>(7) backend. Its availability will be detected at runtime,
		1916	otherwise another method will be used as fallback. This is the
		1917	preferred backend for GNU/Linux systems.</p>
		1918	</dd>
		1919	<dt>EV_USE_KQUEUE</dt>
		1920	<dd>
		1921	<p>If defined to be <code>1</code>, libev will compile in support for the BSD style
		1922	<code>kqueue</code>(2) backend. Its actual availability will be detected at runtime,
		1923	otherwise another method will be used as fallback. This is the preferred
		1924	backend for BSD and BSD-like systems, although on most BSDs kqueue only
		1925	supports some types of fds correctly (the only platform we found that
		1926	supports ptys for example was NetBSD), so kqueue might be compiled in, but
		1927	not be used unless explicitly requested. The best way to use it is to find
		1928	out whether kqueue supports your type of fd properly and use an embedded
		1929	kqueue loop.</p>
		1930	</dd>
		1931	<dt>EV_USE_PORT</dt>
		1932	<dd>
		1933	<p>If defined to be <code>1</code>, libev will compile in support for the Solaris
		1934	10 port style backend. Its availability will be detected at runtime,
		1935	otherwise another method will be used as fallback. This is the preferred
		1936	backend for Solaris 10 systems.</p>
		1937	</dd>
		1938	<dt>EV_USE_DEVPOLL</dt>
		1939	<dd>
		1940	<p>reserved for future expansion, works like the USE symbols above.</p>
		1941	</dd>
		1942	<dt>EV_H</dt>
		1943	<dd>
		1944	<p>The name of the <cite>ev.h</cite> header file used to include it. The default if
		1945	undefined is <code><ev.h></code> in <cite>event.h</cite> and <code>"ev.h"</code> in <cite>ev.c</cite>. This
		1946	can be used to virtually rename the <cite>ev.h</cite> header file in case of conflicts.</p>
		1947	</dd>
		1948	<dt>EV_CONFIG_H</dt>
		1949	<dd>
		1950	<p>If <code>EV_STANDALONE</code> isn't <code>1</code>, this variable can be used to override
		1951	<cite>ev.c</cite>'s idea of where to find the <cite>config.h</cite> file, similarly to
		1952	<code>EV_H</code>, above.</p>
		1953	</dd>
		1954	<dt>EV_EVENT_H</dt>
		1955	<dd>
		1956	<p>Similarly to <code>EV_H</code>, this macro can be used to override <cite>event.c</cite>'s idea
		1957	of how the <cite>event.h</cite> header can be found.</p>
		1958	</dd>
		1959	<dt>EV_PROTOTYPES</dt>
		1960	<dd>
		1961	<p>If defined to be <code>0</code>, then <cite>ev.h</cite> will not define any function
		1962	prototypes, but still define all the structs and other symbols. This is
		1963	occasionally useful if you want to provide your own wrapper functions
		1964	around libev functions.</p>
		1965	</dd>
		1966	<dt>EV_MULTIPLICITY</dt>
		1967	<dd>
		1968	<p>If undefined or defined to <code>1</code>, then all event-loop-specific functions
		1969	will have the <code>struct ev_loop *</code> as first argument, and you can create
		1970	additional independent event loops. Otherwise there will be no support
		1971	for multiple event loops and there is no first event loop pointer
		1972	argument. Instead, all functions act on the single default loop.</p>
		1973	</dd>
		1974	<dt>EV_PERIODIC_ENABLE</dt>
		1975	<dd>
		1976	<p>If undefined or defined to be <code>1</code>, then periodic timers are supported. If
		1977	defined to be <code>0</code>, then they are not. Disabling them saves a few kB of
		1978	code.</p>
		1979	</dd>
		1980	<dt>EV_EMBED_ENABLE</dt>
		1981	<dd>
		1982	<p>If undefined or defined to be <code>1</code>, then embed watchers are supported. If
		1983	defined to be <code>0</code>, then they are not.</p>
		1984	</dd>
		1985	<dt>EV_STAT_ENABLE</dt>
		1986	<dd>
		1987	<p>If undefined or defined to be <code>1</code>, then stat watchers are supported. If
		1988	defined to be <code>0</code>, then they are not.</p>
		1989	</dd>
		1990	<dt>EV_FORK_ENABLE</dt>
		1991	<dd>
		1992	<p>If undefined or defined to be <code>1</code>, then fork watchers are supported. If
		1993	defined to be <code>0</code>, then they are not.</p>
		1994	</dd>
		1995	<dt>EV_MINIMAL</dt>
		1996	<dd>
		1997	<p>If you need to shave off some kilobytes of code at the expense of some
		1998	speed, define this symbol to <code>1</code>. Currently only used for gcc to override
		1999	some inlining decisions, saves roughly 30% codesize of amd64.</p>
		2000	</dd>
		2001	<dt>EV_PID_HASHSIZE</dt>
		2002	<dd>
		2003	<p><code>ev_child</code> watchers use a small hash table to distribute workload by
		2004	pid. The default size is <code>16</code> (or <code>1</code> with <code>EV_MINIMAL</code>), usually more
		2005	than enough. If you need to manage thousands of children you might want to
		2006	increase this value.</p>
		2007	</dd>
		2008	<dt>EV_COMMON</dt>
		2009	<dd>
		2010	<p>By default, all watchers have a <code>void *data</code> member. By redefining
		2011	this macro to a something else you can include more and other types of
		2012	members. You have to define it each time you include one of the files,
		2013	though, and it must be identical each time.</p>
		2014	<p>For example, the perl EV module uses something like this:</p>
		2015	<pre> #define EV_COMMON \
		2016	SV self; / contains this struct */ \
		2017	SV cb_sv, fh /* note no trailing ";" */
		2018
		2019	</pre>
		2020	</dd>
		2021	<dt>EV_CB_DECLARE (type)</dt>
		2022	<dt>EV_CB_INVOKE (watcher, revents)</dt>
		2023	<dt>ev_set_cb (ev, cb)</dt>
		2024	<dd>
		2025	<p>Can be used to change the callback member declaration in each watcher,
		2026	and the way callbacks are invoked and set. Must expand to a struct member
		2027	definition and a statement, respectively. See the <cite>ev.v</cite> header file for
		2028	their default definitions. One possible use for overriding these is to
		2029	avoid the <code>struct ev_loop *</code> as first argument in all cases, or to use
		2030	method calls instead of plain function calls in C++.</p>
		2031
		2032	</div>
		2033	<h2 id="EXAMPLES">EXAMPLES</h2>
		2034	<div id="EXAMPLES_CONTENT">
		2035	<p>For a real-world example of a program the includes libev
		2036	verbatim, you can have a look at the EV perl module
		2037	(<a href="http://software.schmorp.de/pkg/EV.html">http://software.schmorp.de/pkg/EV.html</a>). It has the libev files in
		2038	the <cite>libev/</cite> subdirectory and includes them in the <cite>EV/EVAPI.h</cite> (public
		2039	interface) and <cite>EV.xs</cite> (implementation) files. Only the <cite>EV.xs</cite> file
		2040	will be compiled. It is pretty complex because it provides its own header
		2041	file.</p>
		2042	<p>The usage in rxvt-unicode is simpler. It has a <cite>ev_cpp.h</cite> header file
		2043	that everybody includes and which overrides some autoconf choices:</p>
		2044	<pre> #define EV_USE_POLL 0
		2045	#define EV_MULTIPLICITY 0
		2046	#define EV_PERIODICS 0
		2047	#define EV_CONFIG_H <config.h>
		2048
		2049	#include "ev++.h"
		2050
		2051	</pre>
		2052	<p>And a <cite>ev_cpp.C</cite> implementation file that contains libev proper and is compiled:</p>
		2053	<pre> #include "ev_cpp.h"
		2054	#include "ev.c"
		2055
		2056
		2057
		2058
		2059	</pre>
		2060
		2061	</div>
		2062	<h1 id="COMPLEXITIES">COMPLEXITIES</h1><p><a href="#TOP" class="toplink">Top</a></p>
		2063	<div id="COMPLEXITIES_CONTENT">
		2064	<p>In this section the complexities of (many of) the algorithms used inside
		2065	libev will be explained. For complexity discussions about backends see the
		2066	documentation for <code>ev_default_init</code>.</p>
		2067	<p>
		2068	<dl>
		2069	<dt>Starting and stopping timer/periodic watchers: O(log skipped_other_timers)</dt>
		2070	<dt>Changing timer/periodic watchers (by autorepeat, again): O(log skipped_other_timers)</dt>
		2071	<dt>Starting io/check/prepare/idle/signal/child watchers: O(1)</dt>
		2072	<dt>Stopping check/prepare/idle watchers: O(1)</dt>
		2073	<dt>Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % 16))</dt>
		2074	<dt>Finding the next timer per loop iteration: O(1)</dt>
		2075	<dt>Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)</dt>
		2076	<dt>Activating one watcher: O(1)</dt>
		2077	</dl>
		2078	</p>
		2079
		2080
		2081
		2082
715		2083
716	</div>	2084	</div>
717	<h1 id="AUTHOR">AUTHOR</h1><p><a href="#TOP" class="toplink">Top</a></p>	2085	<h1 id="AUTHOR">AUTHOR</h1><p><a href="#TOP" class="toplink">Top</a></p>
718	<div id="AUTHOR_CONTENT">	2086	<div id="AUTHOR_CONTENT">
719	<p>Marc Lehmann <libev@schmorp.de>.</p>	2087	<p>Marc Lehmann <libev@schmorp.de>.</p>
720		2088
721	</div>	2089	</div>
722	</div></body>	2090	</div></body>
723	</html>	2091	</html>

Diff Legend

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

Comparing libev/ev.html (file contents): Revision 1.6 by root, Mon Nov 12 08:12:14 2007 UTC vs. Revision 1.52 by root, Tue Nov 27 19:41:52 2007 UTC

Diff Legend

Comparing libev/ev.html (file contents):
Revision 1.6 by root, Mon Nov 12 08:12:14 2007 UTC vs.
Revision 1.52 by root, Tue Nov 27 19:41:52 2007 UTC