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

Comparing libev/ev.html (file contents):
Revision 1.3 by root, Mon Nov 12 08:03:31 2007 UTC vs.
Revision 1.36 by root, Sat Nov 24 06:23:27 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="&lt;standard input&gt;" /> 7 <meta name="inputfile" content="&lt;standard input&gt;" />
8 <meta name="outputfile" content="&lt;standard output&gt;" /> 8 <meta name="outputfile" content="&lt;standard output&gt;" />
9 <meta name="created" content="Mon Nov 12 09:03:30 2007" /> 9 <meta name="created" content="Sat Nov 24 05:58:35 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>
25<ul><li><a href="#ASSOCIATING_CUSTOM_DATA_WITH_A_WATCH">ASSOCIATING CUSTOM DATA WITH A WATCHER</a></li> 26<ul><li><a href="#ASSOCIATING_CUSTOM_DATA_WITH_A_WATCH">ASSOCIATING CUSTOM DATA WITH A WATCHER</a></li>
26</ul> 27</ul>
27</li> 28</li>
28<li><a href="#WATCHER_TYPES">WATCHER TYPES</a> 29<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> 30<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> 31<li><a href="#code_ev_timer_code_relative_and_opti"><code>ev_timer</code> - relative and optionally recurring 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> 32<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> 33<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> 34<li><a href="#code_ev_child_code_wait_for_pid_stat"><code>ev_child</code> - wait for pid status changes</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> 35<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> 36<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>
37<li><a href="#code_ev_embed_code_when_one_backend_"><code>ev_embed</code> - when one backend isn't enough</a></li>
36</ul> 38</ul>
37</li> 39</li>
38<li><a href="#OTHER_FUNCTIONS">OTHER FUNCTIONS</a></li> 40<li><a href="#OTHER_FUNCTIONS">OTHER FUNCTIONS</a></li>
41<li><a href="#LIBEVENT_EMULATION">LIBEVENT EMULATION</a></li>
42<li><a href="#C_SUPPORT">C++ SUPPORT</a></li>
39<li><a href="#AUTHOR">AUTHOR</a> 43<li><a href="#AUTHOR">AUTHOR</a>
40</li> 44</li>
41</ul><hr /> 45</ul><hr />
42<!-- INDEX END --> 46<!-- INDEX END -->
43 47
… …
55</div> 59</div>
56<h1 id="DESCRIPTION">DESCRIPTION</h1><p><a href="#TOP" class="toplink">Top</a></p> 60<h1 id="DESCRIPTION">DESCRIPTION</h1><p><a href="#TOP" class="toplink">Top</a></p>
57<div id="DESCRIPTION_CONTENT"> 61<div id="DESCRIPTION_CONTENT">
58<p>Libev is an event loop: you register interest in certain events (such as a 62<p>Libev is an event loop: you register interest in certain events (such as a
59file descriptor being readable or a timeout occuring), and it will manage 63file descriptor being readable or a timeout occuring), and it will manage
60these event sources and provide your program events.</p> 64these event sources and provide your program with events.</p>
61<p>To do this, it must take more or less complete control over your process 65<p>To do this, it must take more or less complete control over your process
62(or thread) by executing the <i>event loop</i> handler, and will then 66(or thread) by executing the <i>event loop</i> handler, and will then
63communicate events via a callback mechanism.</p> 67communicate events via a callback mechanism.</p>
64<p>You register interest in certain events by registering so-called <i>event 68<p>You register interest in certain events by registering so-called <i>event
65watchers</i>, which are relatively small C structures you initialise with the 69watchers</i>, which are relatively small C structures you initialise with the
… …
71<div id="FEATURES_CONTENT"> 75<div id="FEATURES_CONTENT">
72<p>Libev supports select, poll, the linux-specific epoll and the bsd-specific 76<p>Libev supports select, poll, the linux-specific epoll and the bsd-specific
73kqueue mechanisms for file descriptor events, relative timers, absolute 77kqueue mechanisms for file descriptor events, relative timers, absolute
74timers with customised rescheduling, signal events, process status change 78timers with customised rescheduling, signal events, process status change
75events (related to SIGCHLD), and event watchers dealing with the event 79events (related to SIGCHLD), and event watchers dealing with the event
76loop mechanism itself (idle, prepare and check watchers).</p> 80loop mechanism itself (idle, prepare and check watchers). It also is quite
81fast (see this <a href="http://libev.schmorp.de/bench.html">benchmark</a> comparing
82it to libevent for example).</p>
77 83
78</div> 84</div>
79<h1 id="CONVENTIONS">CONVENTIONS</h1><p><a href="#TOP" class="toplink">Top</a></p> 85<h1 id="CONVENTIONS">CONVENTIONS</h1><p><a href="#TOP" class="toplink">Top</a></p>
80<div id="CONVENTIONS_CONTENT"> 86<div id="CONVENTIONS_CONTENT">
81<p>Libev is very configurable. In this manual the default configuration 87<p>Libev is very configurable. In this manual the default configuration
82will be described, which supports multiple event loops. For more info 88will be described, which supports multiple event loops. For more info
83about various configuraiton options please have a look at the file 89about various configuration options please have a look at the file
84<cite>README.embed</cite> in the libev distribution. If libev was configured without 90<cite>README.embed</cite> in the libev distribution. If libev was configured without
85support for multiple event loops, then all functions taking an initial 91support for multiple event loops, then all functions taking an initial
86argument of name <code>loop</code> (which is always of type <code>struct ev_loop *</code>) 92argument of name <code>loop</code> (which is always of type <code>struct ev_loop *</code>)
87will not have this argument.</p> 93will not have this argument.</p>
88 94
89</div> 95</div>
90<h1 id="TIME_AND_OTHER_GLOBAL_FUNCTIONS">TIME AND OTHER GLOBAL FUNCTIONS</h1><p><a href="#TOP" class="toplink">Top</a></p> 96<h1 id="TIME_REPRESENTATION">TIME REPRESENTATION</h1><p><a href="#TOP" class="toplink">Top</a></p>
91<div id="TIME_AND_OTHER_GLOBAL_FUNCTIONS_CONT"> 97<div id="TIME_REPRESENTATION_CONTENT">
92<p>Libev represents time as a single floating point number, representing the 98<p>Libev represents time as a single floating point number, representing the
93(fractional) number of seconds since the (POSIX) epoch (somewhere near 99(fractional) number of seconds since the (POSIX) epoch (somewhere near
94the beginning of 1970, details are complicated, don't ask). This type is 100the beginning of 1970, details are complicated, don't ask). This type is
95called <code>ev_tstamp</code>, which is what you should use too. It usually aliases 101called <code>ev_tstamp</code>, which is what you should use too. It usually aliases
96to the double type in C.</p> 102to the <code>double</code> type in C, and when you need to do any calculations on
103it, you should treat it as such.</p>
104
105
106
107
108
109</div>
110<h1 id="GLOBAL_FUNCTIONS">GLOBAL FUNCTIONS</h1><p><a href="#TOP" class="toplink">Top</a></p>
111<div id="GLOBAL_FUNCTIONS_CONTENT">
112<p>These functions can be called anytime, even before initialising the
113library in any way.</p>
97<dl> 114<dl>
98 <dt>ev_tstamp ev_time ()</dt> 115 <dt>ev_tstamp ev_time ()</dt>
99 <dd> 116 <dd>
100 <p>Returns the current time as libev would use it.</p> 117 <p>Returns the current time as libev would use it. Please note that the
118<code>ev_now</code> function is usually faster and also often returns the timestamp
119you actually want to know.</p>
101 </dd> 120 </dd>
102 <dt>int ev_version_major ()</dt> 121 <dt>int ev_version_major ()</dt>
103 <dt>int ev_version_minor ()</dt> 122 <dt>int ev_version_minor ()</dt>
104 <dd> 123 <dd>
105 <p>You can find out the major and minor version numbers of the library 124 <p>You can find out the major and minor version numbers of the library
106you linked against by calling the functions <code>ev_version_major</code> and 125you linked against by calling the functions <code>ev_version_major</code> and
107<code>ev_version_minor</code>. If you want, you can compare against the global 126<code>ev_version_minor</code>. If you want, you can compare against the global
108symbols <code>EV_VERSION_MAJOR</code> and <code>EV_VERSION_MINOR</code>, which specify the 127symbols <code>EV_VERSION_MAJOR</code> and <code>EV_VERSION_MINOR</code>, which specify the
109version of the library your program was compiled against.</p> 128version of the library your program was compiled against.</p>
110 <p>Usually, its a good idea to terminate if the major versions mismatch, 129 <p>Usually, it's a good idea to terminate if the major versions mismatch,
111as this indicates an incompatible change. Minor versions are usually 130as this indicates an incompatible change. Minor versions are usually
112compatible to older versions, so a larger minor version alone is usually 131compatible to older versions, so a larger minor version alone is usually
113not a problem.</p> 132not a problem.</p>
133 <p>Example: make sure we haven't accidentally been linked against the wrong
134version:</p>
135<pre> assert ((&quot;libev version mismatch&quot;,
136 ev_version_major () == EV_VERSION_MAJOR
137 &amp;&amp; ev_version_minor () &gt;= EV_VERSION_MINOR));
138
139</pre>
140 </dd>
141 <dt>unsigned int ev_supported_backends ()</dt>
142 <dd>
143 <p>Return the set of all backends (i.e. their corresponding <code>EV_BACKEND_*</code>
144value) compiled into this binary of libev (independent of their
145availability on the system you are running on). See <code>ev_default_loop</code> for
146a description of the set values.</p>
147 <p>Example: make sure we have the epoll method, because yeah this is cool and
148a must have and can we have a torrent of it please!!!11</p>
149<pre> assert ((&quot;sorry, no epoll, no sex&quot;,
150 ev_supported_backends () &amp; EVBACKEND_EPOLL));
151
152</pre>
153 </dd>
154 <dt>unsigned int ev_recommended_backends ()</dt>
155 <dd>
156 <p>Return the set of all backends compiled into this binary of libev and also
157recommended for this platform. This set is often smaller than the one
158returned by <code>ev_supported_backends</code>, as for example kqueue is broken on
159most BSDs and will not be autodetected unless you explicitly request it
160(assuming you know what you are doing). This is the set of backends that
161libev will probe for if you specify no backends explicitly.</p>
162 </dd>
163 <dt>unsigned int ev_embeddable_backends ()</dt>
164 <dd>
165 <p>Returns the set of backends that are embeddable in other event loops. This
166is the theoretical, all-platform, value. To find which backends
167might be supported on the current system, you would need to look at
168<code>ev_embeddable_backends () &amp; ev_supported_backends ()</code>, likewise for
169recommended ones.</p>
170 <p>See the description of <code>ev_embed</code> watchers for more info.</p>
114 </dd> 171 </dd>
115 <dt>ev_set_allocator (void *(*cb)(void *ptr, long size))</dt> 172 <dt>ev_set_allocator (void *(*cb)(void *ptr, long size))</dt>
116 <dd> 173 <dd>
117 <p>Sets the allocation function to use (the prototype is similar to the 174 <p>Sets the allocation function to use (the prototype is similar to the
118realloc function). It is used to allocate and free memory (no surprises 175realloc C function, the semantics are identical). It is used to allocate
119here). If it returns zero when memory needs to be allocated, the library 176and free memory (no surprises here). If it returns zero when memory
120might abort or take some potentially destructive action. The default is 177needs to be allocated, the library might abort or take some potentially
121your system realloc function.</p> 178destructive action. The default is your system realloc function.</p>
122 <p>You could override this function in high-availability programs to, say, 179 <p>You could override this function in high-availability programs to, say,
123free some memory if it cannot allocate memory, to use a special allocator, 180free some memory if it cannot allocate memory, to use a special allocator,
124or even to sleep a while and retry until some memory is available.</p> 181or even to sleep a while and retry until some memory is available.</p>
182 <p>Example: replace the libev allocator with one that waits a bit and then
183retries: better than mine).</p>
184<pre> static void *
185 persistent_realloc (void *ptr, long size)
186 {
187 for (;;)
188 {
189 void *newptr = realloc (ptr, size);
190
191 if (newptr)
192 return newptr;
193
194 sleep (60);
195 }
196 }
197
198 ...
199 ev_set_allocator (persistent_realloc);
200
201</pre>
125 </dd> 202 </dd>
126 <dt>ev_set_syserr_cb (void (*cb)(const char *msg));</dt> 203 <dt>ev_set_syserr_cb (void (*cb)(const char *msg));</dt>
127 <dd> 204 <dd>
128 <p>Set the callback function to call on a retryable syscall error (such 205 <p>Set the callback function to call on a retryable syscall error (such
129as failed select, poll, epoll_wait). The message is a printable string 206as failed select, poll, epoll_wait). The message is a printable string
130indicating the system call or subsystem causing the problem. If this 207indicating the system call or subsystem causing the problem. If this
131callback is set, then libev will expect it to remedy the sitution, no 208callback is set, then libev will expect it to remedy the sitution, no
132matter what, when it returns. That is, libev will geenrally retry the 209matter what, when it returns. That is, libev will generally retry the
133requested operation, or, if the condition doesn't go away, do bad stuff 210requested operation, or, if the condition doesn't go away, do bad stuff
134(such as abort).</p> 211(such as abort).</p>
212 <p>Example: do the same thing as libev does internally:</p>
213<pre> static void
214 fatal_error (const char *msg)
215 {
216 perror (msg);
217 abort ();
218 }
219
220 ...
221 ev_set_syserr_cb (fatal_error);
222
223</pre>
135 </dd> 224 </dd>
136</dl> 225</dl>
137 226
138</div> 227</div>
139<h1 id="FUNCTIONS_CONTROLLING_THE_EVENT_LOOP">FUNCTIONS CONTROLLING THE EVENT LOOP</h1><p><a href="#TOP" class="toplink">Top</a></p> 228<h1 id="FUNCTIONS_CONTROLLING_THE_EVENT_LOOP">FUNCTIONS CONTROLLING THE EVENT LOOP</h1><p><a href="#TOP" class="toplink">Top</a></p>
140<div id="FUNCTIONS_CONTROLLING_THE_EVENT_LOOP-2"> 229<div id="FUNCTIONS_CONTROLLING_THE_EVENT_LOOP-2">
141<p>An event loop is described by a <code>struct ev_loop *</code>. The library knows two 230<p>An event loop is described by a <code>struct ev_loop *</code>. The library knows two
142types of such loops, the <i>default</i> loop, which supports signals and child 231types of such loops, the <i>default</i> loop, which supports signals and child
143events, and dynamically created loops which do not.</p> 232events, and dynamically created loops which do not.</p>
144<p>If you use threads, a common model is to run the default event loop 233<p>If you use threads, a common model is to run the default event loop
145in your main thread (or in a separate thrad) and for each thread you 234in your main thread (or in a separate thread) and for each thread you
146create, you also create another event loop. Libev itself does no lockign 235create, you also create another event loop. Libev itself does no locking
147whatsoever, so if you mix calls to different event loops, make sure you 236whatsoever, so if you mix calls to the same event loop in different
148lock (this is usually a bad idea, though, even if done right).</p> 237threads, make sure you lock (this is usually a bad idea, though, even if
238done correctly, because it's hideous and inefficient).</p>
149<dl> 239<dl>
150 <dt>struct ev_loop *ev_default_loop (unsigned int flags)</dt> 240 <dt>struct ev_loop *ev_default_loop (unsigned int flags)</dt>
151 <dd> 241 <dd>
152 <p>This will initialise the default event loop if it hasn't been initialised 242 <p>This will initialise the default event loop if it hasn't been initialised
153yet and return it. If the default loop could not be initialised, returns 243yet and return it. If the default loop could not be initialised, returns
154false. If it already was initialised it simply returns it (and ignores the 244false. If it already was initialised it simply returns it (and ignores the
155flags).</p> 245flags. If that is troubling you, check <code>ev_backend ()</code> afterwards).</p>
156 <p>If you don't know what event loop to use, use the one returned from this 246 <p>If you don't know what event loop to use, use the one returned from this
157function.</p> 247function.</p>
158 <p>The flags argument can be used to specify special behaviour or specific 248 <p>The flags argument can be used to specify special behaviour or specific
159backends to use, and is usually specified as 0 (or EVFLAG_AUTO)</p> 249backends to use, and is usually specified as <code>0</code> (or <code>EVFLAG_AUTO</code>).</p>
160 <p>It supports the following flags:</p> 250 <p>The following flags are supported:</p>
161 <p> 251 <p>
162 <dl> 252 <dl>
163 <dt>EVFLAG_AUTO</dt> 253 <dt><code>EVFLAG_AUTO</code></dt>
164 <dd> 254 <dd>
165 <p>The default flags value. Use this if you have no clue (its the right 255 <p>The default flags value. Use this if you have no clue (it's the right
166thing, believe me).</p> 256thing, believe me).</p>
167 </dd> 257 </dd>
168 <dt>EVFLAG_NOENV</dt> 258 <dt><code>EVFLAG_NOENV</code></dt>
169 <dd> 259 <dd>
170 <p>If this flag bit is ored into the flag value then libev will <i>not</i> look 260 <p>If this flag bit is ored into the flag value (or the program runs setuid
171at the environment variable <code>LIBEV_FLAGS</code>. Otherwise (the default), this 261or setgid) then libev will <i>not</i> look at the environment variable
172environment variable will override the flags completely. This is useful 262<code>LIBEV_FLAGS</code>. Otherwise (the default), this environment variable will
263override the flags completely if it is found in the environment. This is
173to try out specific backends to tets their performance, or to work around 264useful to try out specific backends to test their performance, or to work
174bugs.</p> 265around bugs.</p>
175 </dd> 266 </dd>
176 <dt>EVMETHOD_SELECT portable select backend</dt> 267 <dt><code>EVBACKEND_SELECT</code> (value 1, portable select backend)</dt>
177 <dt>EVMETHOD_POLL poll backend (everywhere except windows)</dt>
178 <dt>EVMETHOD_EPOLL linux only</dt>
179 <dt>EVMETHOD_KQUEUE some bsds only</dt>
180 <dt>EVMETHOD_DEVPOLL solaris 8 only</dt>
181 <dt>EVMETHOD_PORT solaris 10 only</dt>
182 <dd> 268 <dd>
183 <p>If one or more of these are ored into the flags value, then only these 269 <p>This is your standard select(2) backend. Not <i>completely</i> standard, as
184backends will be tried (in the reverse order as given here). If one are 270libev tries to roll its own fd_set with no limits on the number of fds,
185specified, any backend will do.</p> 271but if that fails, expect a fairly low limit on the number of fds when
272using this backend. It doesn't scale too well (O(highest_fd)), but its usually
273the fastest backend for a low number of fds.</p>
274 </dd>
275 <dt><code>EVBACKEND_POLL</code> (value 2, poll backend, available everywhere except on windows)</dt>
276 <dd>
277 <p>And this is your standard poll(2) backend. It's more complicated than
278select, but handles sparse fds better and has no artificial limit on the
279number of fds you can use (except it will slow down considerably with a
280lot of inactive fds). It scales similarly to select, i.e. O(total_fds).</p>
281 </dd>
282 <dt><code>EVBACKEND_EPOLL</code> (value 4, Linux)</dt>
283 <dd>
284 <p>For few fds, this backend is a bit little slower than poll and select,
285but it scales phenomenally better. While poll and select usually scale like
286O(total_fds) where n is the total number of fds (or the highest fd), epoll scales
287either O(1) or O(active_fds).</p>
288 <p>While stopping and starting an I/O watcher in the same iteration will
289result in some caching, there is still a syscall per such incident
290(because the fd could point to a different file description now), so its
291best to avoid that. Also, dup()ed file descriptors might not work very
292well if you register events for both fds.</p>
293 <p>Please note that epoll sometimes generates spurious notifications, so you
294need to use non-blocking I/O or other means to avoid blocking when no data
295(or space) is available.</p>
296 </dd>
297 <dt><code>EVBACKEND_KQUEUE</code> (value 8, most BSD clones)</dt>
298 <dd>
299 <p>Kqueue deserves special mention, as at the time of this writing, it
300was broken on all BSDs except NetBSD (usually it doesn't work with
301anything but sockets and pipes, except on Darwin, where of course its
302completely useless). For this reason its not being &quot;autodetected&quot;
303unless you explicitly specify it explicitly in the flags (i.e. using
304<code>EVBACKEND_KQUEUE</code>).</p>
305 <p>It scales in the same way as the epoll backend, but the interface to the
306kernel is more efficient (which says nothing about its actual speed, of
307course). While starting and stopping an I/O watcher does not cause an
308extra syscall as with epoll, it still adds up to four event changes per
309incident, so its best to avoid that.</p>
310 </dd>
311 <dt><code>EVBACKEND_DEVPOLL</code> (value 16, Solaris 8)</dt>
312 <dd>
313 <p>This is not implemented yet (and might never be).</p>
314 </dd>
315 <dt><code>EVBACKEND_PORT</code> (value 32, Solaris 10)</dt>
316 <dd>
317 <p>This uses the Solaris 10 port mechanism. As with everything on Solaris,
318it's really slow, but it still scales very well (O(active_fds)).</p>
319 <p>Please note that solaris ports can result in a lot of spurious
320notifications, so you need to use non-blocking I/O or other means to avoid
321blocking when no data (or space) is available.</p>
322 </dd>
323 <dt><code>EVBACKEND_ALL</code></dt>
324 <dd>
325 <p>Try all backends (even potentially broken ones that wouldn't be tried
326with <code>EVFLAG_AUTO</code>). Since this is a mask, you can do stuff such as
327<code>EVBACKEND_ALL &amp; ~EVBACKEND_KQUEUE</code>.</p>
186 </dd> 328 </dd>
187 </dl> 329 </dl>
188 </p> 330 </p>
331 <p>If one or more of these are ored into the flags value, then only these
332backends will be tried (in the reverse order as given here). If none are
333specified, most compiled-in backend will be tried, usually in reverse
334order of their flag values :)</p>
335 <p>The most typical usage is like this:</p>
336<pre> if (!ev_default_loop (0))
337 fatal (&quot;could not initialise libev, bad $LIBEV_FLAGS in environment?&quot;);
338
339</pre>
340 <p>Restrict libev to the select and poll backends, and do not allow
341environment settings to be taken into account:</p>
342<pre> ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);
343
344</pre>
345 <p>Use whatever libev has to offer, but make sure that kqueue is used if
346available (warning, breaks stuff, best use only with your own private
347event loop and only if you know the OS supports your types of fds):</p>
348<pre> ev_default_loop (ev_recommended_backends () | EVBACKEND_KQUEUE);
349
350</pre>
189 </dd> 351 </dd>
190 <dt>struct ev_loop *ev_loop_new (unsigned int flags)</dt> 352 <dt>struct ev_loop *ev_loop_new (unsigned int flags)</dt>
191 <dd> 353 <dd>
192 <p>Similar to <code>ev_default_loop</code>, but always creates a new event loop that is 354 <p>Similar to <code>ev_default_loop</code>, but always creates a new event loop that is
193always distinct from the default loop. Unlike the default loop, it cannot 355always distinct from the default loop. Unlike the default loop, it cannot
194handle signal and child watchers, and attempts to do so will be greeted by 356handle signal and child watchers, and attempts to do so will be greeted by
195undefined behaviour (or a failed assertion if assertions are enabled).</p> 357undefined behaviour (or a failed assertion if assertions are enabled).</p>
358 <p>Example: try to create a event loop that uses epoll and nothing else.</p>
359<pre> struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV);
360 if (!epoller)
361 fatal (&quot;no epoll found here, maybe it hides under your chair&quot;);
362
363</pre>
196 </dd> 364 </dd>
197 <dt>ev_default_destroy ()</dt> 365 <dt>ev_default_destroy ()</dt>
198 <dd> 366 <dd>
199 <p>Destroys the default loop again (frees all memory and kernel state 367 <p>Destroys the default loop again (frees all memory and kernel state
200etc.). This stops all registered event watchers (by not touching them in 368etc.). This stops all registered event watchers (by not touching them in
201any way whatsoever, although you cnanot rely on this :).</p> 369any way whatsoever, although you cannot rely on this :).</p>
202 </dd> 370 </dd>
203 <dt>ev_loop_destroy (loop)</dt> 371 <dt>ev_loop_destroy (loop)</dt>
204 <dd> 372 <dd>
205 <p>Like <code>ev_default_destroy</code>, but destroys an event loop created by an 373 <p>Like <code>ev_default_destroy</code>, but destroys an event loop created by an
206earlier call to <code>ev_loop_new</code>.</p> 374earlier call to <code>ev_loop_new</code>.</p>
… …
209 <dd> 377 <dd>
210 <p>This function reinitialises the kernel state for backends that have 378 <p>This function reinitialises the kernel state for backends that have
211one. Despite the name, you can call it anytime, but it makes most sense 379one. Despite the name, you can call it anytime, but it makes most sense
212after forking, in either the parent or child process (or both, but that 380after forking, in either the parent or child process (or both, but that
213again makes little sense).</p> 381again makes little sense).</p>
214 <p>You <i>must</i> call this function after forking if and only if you want to 382 <p>You <i>must</i> call this function in the child process after forking if and
215use the event library in both processes. If you just fork+exec, you don't 383only if you want to use the event library in both processes. If you just
216have to call it.</p> 384fork+exec, you don't have to call it.</p>
217 <p>The function itself is quite fast and its usually not a problem to call 385 <p>The function itself is quite fast and it's usually not a problem to call
218it just in case after a fork. To make this easy, the function will fit in 386it just in case after a fork. To make this easy, the function will fit in
219quite nicely into a call to <code>pthread_atfork</code>:</p> 387quite nicely into a call to <code>pthread_atfork</code>:</p>
220<pre> pthread_atfork (0, 0, ev_default_fork); 388<pre> pthread_atfork (0, 0, ev_default_fork);
221 389
222</pre> 390</pre>
391 <p>At the moment, <code>EVBACKEND_SELECT</code> and <code>EVBACKEND_POLL</code> are safe to use
392without calling this function, so if you force one of those backends you
393do not need to care.</p>
223 </dd> 394 </dd>
224 <dt>ev_loop_fork (loop)</dt> 395 <dt>ev_loop_fork (loop)</dt>
225 <dd> 396 <dd>
226 <p>Like <code>ev_default_fork</code>, but acts on an event loop created by 397 <p>Like <code>ev_default_fork</code>, but acts on an event loop created by
227<code>ev_loop_new</code>. Yes, you have to call this on every allocated event loop 398<code>ev_loop_new</code>. Yes, you have to call this on every allocated event loop
228after fork, and how you do this is entirely your own problem.</p> 399after fork, and how you do this is entirely your own problem.</p>
229 </dd> 400 </dd>
230 <dt>unsigned int ev_method (loop)</dt> 401 <dt>unsigned int ev_backend (loop)</dt>
231 <dd> 402 <dd>
232 <p>Returns one of the <code>EVMETHOD_*</code> flags indicating the event backend in 403 <p>Returns one of the <code>EVBACKEND_*</code> flags indicating the event backend in
233use.</p> 404use.</p>
234 </dd> 405 </dd>
235 <dt>ev_tstamp = ev_now (loop)</dt> 406 <dt>ev_tstamp ev_now (loop)</dt>
236 <dd> 407 <dd>
237 <p>Returns the current &quot;event loop time&quot;, which is the time the event loop 408 <p>Returns the current &quot;event loop time&quot;, which is the time the event loop
238got events and started processing them. This timestamp does not change 409received events and started processing them. This timestamp does not
239as long as callbacks are being processed, and this is also the base time 410change as long as callbacks are being processed, and this is also the base
240used for relative timers. You can treat it as the timestamp of the event 411time used for relative timers. You can treat it as the timestamp of the
241occuring (or more correctly, the mainloop finding out about it).</p> 412event occuring (or more correctly, libev finding out about it).</p>
242 </dd> 413 </dd>
243 <dt>ev_loop (loop, int flags)</dt> 414 <dt>ev_loop (loop, int flags)</dt>
244 <dd> 415 <dd>
245 <p>Finally, this is it, the event handler. This function usually is called 416 <p>Finally, this is it, the event handler. This function usually is called
246after you initialised all your watchers and you want to start handling 417after you initialised all your watchers and you want to start handling
247events.</p> 418events.</p>
248 <p>If the flags argument is specified as 0, it will not return until either 419 <p>If the flags argument is specified as <code>0</code>, it will not return until
249no event watchers are active anymore or <code>ev_unloop</code> was called.</p> 420either no event watchers are active anymore or <code>ev_unloop</code> was called.</p>
421 <p>Please note that an explicit <code>ev_unloop</code> is usually better than
422relying on all watchers to be stopped when deciding when a program has
423finished (especially in interactive programs), but having a program that
424automatically loops as long as it has to and no longer by virtue of
425relying on its watchers stopping correctly is a thing of beauty.</p>
250 <p>A flags value of <code>EVLOOP_NONBLOCK</code> will look for new events, will handle 426 <p>A flags value of <code>EVLOOP_NONBLOCK</code> will look for new events, will handle
251those events and any outstanding ones, but will not block your process in 427those events and any outstanding ones, but will not block your process in
252case there are no events.</p> 428case there are no events and will return after one iteration of the loop.</p>
253 <p>A flags value of <code>EVLOOP_ONESHOT</code> will look for new events (waiting if 429 <p>A flags value of <code>EVLOOP_ONESHOT</code> will look for new events (waiting if
254neccessary) and will handle those and any outstanding ones. It will block 430neccessary) and will handle those and any outstanding ones. It will block
255your process until at least one new event arrives.</p> 431your process until at least one new event arrives, and will return after
256 <p>This flags value could be used to implement alternative looping 432one iteration of the loop. This is useful if you are waiting for some
257constructs, but the <code>prepare</code> and <code>check</code> watchers provide a better and 433external event in conjunction with something not expressible using other
258more generic mechanism.</p> 434libev watchers. However, a pair of <code>ev_prepare</code>/<code>ev_check</code> watchers is
435usually a better approach for this kind of thing.</p>
436 <p>Here are the gory details of what <code>ev_loop</code> does:</p>
437<pre> * If there are no active watchers (reference count is zero), return.
438 - Queue prepare watchers and then call all outstanding watchers.
439 - If we have been forked, recreate the kernel state.
440 - Update the kernel state with all outstanding changes.
441 - Update the &quot;event loop time&quot;.
442 - Calculate for how long to block.
443 - Block the process, waiting for any events.
444 - Queue all outstanding I/O (fd) events.
445 - Update the &quot;event loop time&quot; and do time jump handling.
446 - Queue all outstanding timers.
447 - Queue all outstanding periodics.
448 - If no events are pending now, queue all idle watchers.
449 - Queue all check watchers.
450 - Call all queued watchers in reverse order (i.e. check watchers first).
451 Signals and child watchers are implemented as I/O watchers, and will
452 be handled here by queueing them when their watcher gets executed.
453 - If ev_unloop has been called or EVLOOP_ONESHOT or EVLOOP_NONBLOCK
454 were used, return, otherwise continue with step *.
455
456</pre>
457 <p>Example: queue some jobs and then loop until no events are outsanding
458anymore.</p>
459<pre> ... queue jobs here, make sure they register event watchers as long
460 ... as they still have work to do (even an idle watcher will do..)
461 ev_loop (my_loop, 0);
462 ... jobs done. yeah!
463
464</pre>
259 </dd> 465 </dd>
260 <dt>ev_unloop (loop, how)</dt> 466 <dt>ev_unloop (loop, how)</dt>
261 <dd> 467 <dd>
262 <p>Can be used to make a call to <code>ev_loop</code> return early. The <code>how</code> argument 468 <p>Can be used to make a call to <code>ev_loop</code> return early (but only after it
469has processed all outstanding events). The <code>how</code> argument must be either
263must be either <code>EVUNLOOP_ONCE</code>, which will make the innermost <code>ev_loop</code> 470<code>EVUNLOOP_ONE</code>, which will make the innermost <code>ev_loop</code> call return, or
264call return, or <code>EVUNLOOP_ALL</code>, which will make all nested <code>ev_loop</code> 471<code>EVUNLOOP_ALL</code>, which will make all nested <code>ev_loop</code> calls return.</p>
265calls return.</p>
266 </dd> 472 </dd>
267 <dt>ev_ref (loop)</dt> 473 <dt>ev_ref (loop)</dt>
268 <dt>ev_unref (loop)</dt> 474 <dt>ev_unref (loop)</dt>
269 <dd> 475 <dd>
270 <p>Ref/unref can be used to add or remove a refcount on the event loop: Every 476 <p>Ref/unref can be used to add or remove a reference count on the event
271watcher keeps one reference. If you have a long-runing watcher you never 477loop: Every watcher keeps one reference, and as long as the reference
272unregister that should not keep ev_loop from running, ev_unref() after 478count is nonzero, <code>ev_loop</code> will not return on its own. If you have
273starting, and ev_ref() before stopping it. Libev itself uses this for 479a watcher you never unregister that should not keep <code>ev_loop</code> from
274example for its internal signal pipe: It is not visible to you as a user 480returning, ev_unref() after starting, and ev_ref() before stopping it. For
275and should not keep <code>ev_loop</code> from exiting if the work is done. It is 481example, libev itself uses this for its internal signal pipe: It is not
276also an excellent way to do this for generic recurring timers or from 482visible to the libev user and should not keep <code>ev_loop</code> from exiting if
277within third-party libraries. Just remember to unref after start and ref 483no event watchers registered by it are active. It is also an excellent
278before stop.</p> 484way to do this for generic recurring timers or from within third-party
485libraries. Just remember to <i>unref after start</i> and <i>ref before stop</i>.</p>
486 <p>Example: create a signal watcher, but keep it from keeping <code>ev_loop</code>
487running when nothing else is active.</p>
488<pre> struct dv_signal exitsig;
489 ev_signal_init (&amp;exitsig, sig_cb, SIGINT);
490 ev_signal_start (myloop, &amp;exitsig);
491 evf_unref (myloop);
492
493</pre>
494 <p>Example: for some weird reason, unregister the above signal handler again.</p>
495<pre> ev_ref (myloop);
496 ev_signal_stop (myloop, &amp;exitsig);
497
498</pre>
279 </dd> 499 </dd>
280</dl> 500</dl>
281 501
282</div> 502</div>
283<h1 id="ANATOMY_OF_A_WATCHER">ANATOMY OF A WATCHER</h1><p><a href="#TOP" class="toplink">Top</a></p> 503<h1 id="ANATOMY_OF_A_WATCHER">ANATOMY OF A WATCHER</h1><p><a href="#TOP" class="toplink">Top</a></p>
284<div id="ANATOMY_OF_A_WATCHER_CONTENT"> 504<div id="ANATOMY_OF_A_WATCHER_CONTENT">
285<p>A watcher is a structure that you create and register to record your 505<p>A watcher is a structure that you create and register to record your
286interest in some event. For instance, if you want to wait for STDIN to 506interest in some event. For instance, if you want to wait for STDIN to
287become readable, you would create an ev_io watcher for that:</p> 507become readable, you would create an <code>ev_io</code> watcher for that:</p>
288<pre> static void my_cb (struct ev_loop *loop, struct ev_io *w, int revents) 508<pre> static void my_cb (struct ev_loop *loop, struct ev_io *w, int revents)
289 { 509 {
290 ev_io_stop (w); 510 ev_io_stop (w);
291 ev_unloop (loop, EVUNLOOP_ALL); 511 ev_unloop (loop, EVUNLOOP_ALL);
292 } 512 }
… …
315with a watcher-specific start function (<code>ev_&lt;type&gt;_start (loop, watcher 535with a watcher-specific start function (<code>ev_&lt;type&gt;_start (loop, watcher
316*)</code>), and you can stop watching for events at any time by calling the 536*)</code>), and you can stop watching for events at any time by calling the
317corresponding stop function (<code>ev_&lt;type&gt;_stop (loop, watcher *)</code>.</p> 537corresponding stop function (<code>ev_&lt;type&gt;_stop (loop, watcher *)</code>.</p>
318<p>As long as your watcher is active (has been started but not stopped) you 538<p>As long as your watcher is active (has been started but not stopped) you
319must not touch the values stored in it. Most specifically you must never 539must not touch the values stored in it. Most specifically you must never
320reinitialise it or call its set method.</p> 540reinitialise it or call its set macro.</p>
321<p>You cna check wether an event is active by calling the <code>ev_is_active 541<p>You can check whether an event is active by calling the <code>ev_is_active
322(watcher *)</code> macro. To see wether an event is outstanding (but the 542(watcher *)</code> macro. To see whether an event is outstanding (but the
323callback for it has not been called yet) you cna use the <code>ev_is_pending 543callback for it has not been called yet) you can use the <code>ev_is_pending
324(watcher *)</code> macro.</p> 544(watcher *)</code> macro.</p>
325<p>Each and every callback receives the event loop pointer as first, the 545<p>Each and every callback receives the event loop pointer as first, the
326registered watcher structure as second, and a bitset of received events as 546registered watcher structure as second, and a bitset of received events as
327third argument.</p> 547third argument.</p>
328<p>The rceeived events usually include a single bit per event type received 548<p>The received events usually include a single bit per event type received
329(you can receive multiple events at the same time). The possible bit masks 549(you can receive multiple events at the same time). The possible bit masks
330are:</p> 550are:</p>
331<dl> 551<dl>
332 <dt>EV_READ</dt> 552 <dt><code>EV_READ</code></dt>
333 <dt>EV_WRITE</dt> 553 <dt><code>EV_WRITE</code></dt>
334 <dd> 554 <dd>
335 <p>The file descriptor in the ev_io watcher has become readable and/or 555 <p>The file descriptor in the <code>ev_io</code> watcher has become readable and/or
336writable.</p> 556writable.</p>
337 </dd> 557 </dd>
338 <dt>EV_TIMEOUT</dt> 558 <dt><code>EV_TIMEOUT</code></dt>
339 <dd>
340 <p>The ev_timer watcher has timed out.</p>
341 </dd> 559 <dd>
342 <dt>EV_PERIODIC</dt> 560 <p>The <code>ev_timer</code> watcher has timed out.</p>
343 <dd> 561 </dd>
344 <p>The ev_periodic watcher has timed out.</p> 562 <dt><code>EV_PERIODIC</code></dt>
345 </dd> 563 <dd>
346 <dt>EV_SIGNAL</dt> 564 <p>The <code>ev_periodic</code> watcher has timed out.</p>
347 <dd> 565 </dd>
566 <dt><code>EV_SIGNAL</code></dt>
567 <dd>
348 <p>The signal specified in the ev_signal watcher has been received by a thread.</p> 568 <p>The signal specified in the <code>ev_signal</code> watcher has been received by a thread.</p>
349 </dd>
350 <dt>EV_CHILD</dt>
351 <dd> 569 </dd>
570 <dt><code>EV_CHILD</code></dt>
571 <dd>
352 <p>The pid specified in the ev_child watcher has received a status change.</p> 572 <p>The pid specified in the <code>ev_child</code> watcher has received a status change.</p>
353 </dd>
354 <dt>EV_IDLE</dt>
355 <dd> 573 </dd>
574 <dt><code>EV_IDLE</code></dt>
575 <dd>
356 <p>The ev_idle watcher has determined that you have nothing better to do.</p> 576 <p>The <code>ev_idle</code> watcher has determined that you have nothing better to do.</p>
357 </dd>
358 <dt>EV_PREPARE</dt>
359 <dt>EV_CHECK</dt>
360 <dd> 577 </dd>
578 <dt><code>EV_PREPARE</code></dt>
579 <dt><code>EV_CHECK</code></dt>
580 <dd>
361 <p>All ev_prepare watchers are invoked just <i>before</i> <code>ev_loop</code> starts 581 <p>All <code>ev_prepare</code> watchers are invoked just <i>before</i> <code>ev_loop</code> starts
362to gather new events, and all ev_check watchers are invoked just after 582to gather new events, and all <code>ev_check</code> watchers are invoked just after
363<code>ev_loop</code> has gathered them, but before it invokes any callbacks for any 583<code>ev_loop</code> has gathered them, but before it invokes any callbacks for any
364received events. Callbacks of both watcher types can start and stop as 584received events. Callbacks of both watcher types can start and stop as
365many watchers as they want, and all of them will be taken into account 585many watchers as they want, and all of them will be taken into account
366(for example, a ev_prepare watcher might start an idle watcher to keep 586(for example, a <code>ev_prepare</code> watcher might start an idle watcher to keep
367<code>ev_loop</code> from blocking).</p> 587<code>ev_loop</code> from blocking).</p>
368 </dd> 588 </dd>
369 <dt>EV_ERROR</dt> 589 <dt><code>EV_ERROR</code></dt>
370 <dd> 590 <dd>
371 <p>An unspecified error has occured, the watcher has been stopped. This might 591 <p>An unspecified error has occured, the watcher has been stopped. This might
372happen because the watcher could not be properly started because libev 592happen because the watcher could not be properly started because libev
373ran out of memory, a file descriptor was found to be closed or any other 593ran out of memory, a file descriptor was found to be closed or any other
374problem. You best act on it by reporting the problem and somehow coping 594problem. You best act on it by reporting the problem and somehow coping
… …
383 603
384</div> 604</div>
385<h2 id="ASSOCIATING_CUSTOM_DATA_WITH_A_WATCH">ASSOCIATING CUSTOM DATA WITH A WATCHER</h2> 605<h2 id="ASSOCIATING_CUSTOM_DATA_WITH_A_WATCH">ASSOCIATING CUSTOM DATA WITH A WATCHER</h2>
386<div id="ASSOCIATING_CUSTOM_DATA_WITH_A_WATCH-2"> 606<div id="ASSOCIATING_CUSTOM_DATA_WITH_A_WATCH-2">
387<p>Each watcher has, by default, a member <code>void *data</code> that you can change 607<p>Each watcher has, by default, a member <code>void *data</code> that you can change
388and read at any time, libev will completely ignore it. This cna be used 608and read at any time, libev will completely ignore it. This can be used
389to associate arbitrary data with your watcher. If you need more data and 609to associate arbitrary data with your watcher. If you need more data and
390don't want to allocate memory and store a pointer to it in that data 610don't want to allocate memory and store a pointer to it in that data
391member, you can also &quot;subclass&quot; the watcher type and provide your own 611member, you can also &quot;subclass&quot; the watcher type and provide your own
392data:</p> 612data:</p>
393<pre> struct my_io 613<pre> struct my_io
… …
419<h1 id="WATCHER_TYPES">WATCHER TYPES</h1><p><a href="#TOP" class="toplink">Top</a></p> 639<h1 id="WATCHER_TYPES">WATCHER TYPES</h1><p><a href="#TOP" class="toplink">Top</a></p>
420<div id="WATCHER_TYPES_CONTENT"> 640<div id="WATCHER_TYPES_CONTENT">
421<p>This section describes each watcher in detail, but will not repeat 641<p>This section describes each watcher in detail, but will not repeat
422information given in the last section.</p> 642information given in the last section.</p>
423 643
644
645
646
647
424</div> 648</div>
425<h2 id="struct_ev_io_is_my_file_descriptor_r">struct ev_io - is my file descriptor readable or writable</h2> 649<h2 id="code_ev_io_code_is_this_file_descrip"><code>ev_io</code> - is this file descriptor readable or writable</h2>
426<div id="struct_ev_io_is_my_file_descriptor_r-2"> 650<div id="code_ev_io_code_is_this_file_descrip-2">
427<p>I/O watchers check wether a file descriptor is readable or writable 651<p>I/O watchers check whether a file descriptor is readable or writable
428in each iteration of the event loop (This behaviour is called 652in each iteration of the event loop (This behaviour is called
429level-triggering because you keep receiving events as long as the 653level-triggering because you keep receiving events as long as the
430condition persists. Remember you cna stop the watcher if you don't want to 654condition persists. Remember you can stop the watcher if you don't want to
431act on the event and neither want to receive future events).</p> 655act on the event and neither want to receive future events).</p>
656<p>In general you can register as many read and/or write event watchers per
657fd as you want (as long as you don't confuse yourself). Setting all file
658descriptors to non-blocking mode is also usually a good idea (but not
659required if you know what you are doing).</p>
660<p>You have to be careful with dup'ed file descriptors, though. Some backends
661(the linux epoll backend is a notable example) cannot handle dup'ed file
662descriptors correctly if you register interest in two or more fds pointing
663to the same underlying file/socket etc. description (that is, they share
664the same underlying &quot;file open&quot;).</p>
665<p>If you must do this, then force the use of a known-to-be-good backend
666(at the time of this writing, this includes only <code>EVBACKEND_SELECT</code> and
667<code>EVBACKEND_POLL</code>).</p>
432<dl> 668<dl>
433 <dt>ev_io_init (ev_io *, callback, int fd, int events)</dt> 669 <dt>ev_io_init (ev_io *, callback, int fd, int events)</dt>
434 <dt>ev_io_set (ev_io *, int fd, int events)</dt> 670 <dt>ev_io_set (ev_io *, int fd, int events)</dt>
435 <dd> 671 <dd>
436 <p>Configures an ev_io watcher. The fd is the file descriptor to rceeive 672 <p>Configures an <code>ev_io</code> watcher. The fd is the file descriptor to rceeive
437events for and events is either <code>EV_READ</code>, <code>EV_WRITE</code> or <code>EV_READ | 673events for and events is either <code>EV_READ</code>, <code>EV_WRITE</code> or <code>EV_READ |
438EV_WRITE</code> to receive the given events.</p> 674EV_WRITE</code> to receive the given events.</p>
675 <p>Please note that most of the more scalable backend mechanisms (for example
676epoll and solaris ports) can result in spurious readyness notifications
677for file descriptors, so you practically need to use non-blocking I/O (and
678treat callback invocation as hint only), or retest separately with a safe
679interface before doing I/O (XLib can do this), or force the use of either
680<code>EVBACKEND_SELECT</code> or <code>EVBACKEND_POLL</code>, which don't suffer from this
681problem. Also note that it is quite easy to have your callback invoked
682when the readyness condition is no longer valid even when employing
683typical ways of handling events, so its a good idea to use non-blocking
684I/O unconditionally.</p>
439 </dd> 685 </dd>
440</dl> 686</dl>
687<p>Example: call <code>stdin_readable_cb</code> when STDIN_FILENO has become, well
688readable, but only once. Since it is likely line-buffered, you could
689attempt to read a whole line in the callback:</p>
690<pre> static void
691 stdin_readable_cb (struct ev_loop *loop, struct ev_io *w, int revents)
692 {
693 ev_io_stop (loop, w);
694 .. read from stdin here (or from w-&gt;fd) and haqndle any I/O errors
695 }
441 696
697 ...
698 struct ev_loop *loop = ev_default_init (0);
699 struct ev_io stdin_readable;
700 ev_io_init (&amp;stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);
701 ev_io_start (loop, &amp;stdin_readable);
702 ev_loop (loop, 0);
703
704
705
706
707</pre>
708
442</div> 709</div>
443<h2 id="struct_ev_timer_relative_and_optiona">struct ev_timer - relative and optionally recurring timeouts</h2> 710<h2 id="code_ev_timer_code_relative_and_opti"><code>ev_timer</code> - relative and optionally recurring timeouts</h2>
444<div id="struct_ev_timer_relative_and_optiona-2"> 711<div id="code_ev_timer_code_relative_and_opti-2">
445<p>Timer watchers are simple relative timers that generate an event after a 712<p>Timer watchers are simple relative timers that generate an event after a
446given time, and optionally repeating in regular intervals after that.</p> 713given time, and optionally repeating in regular intervals after that.</p>
447<p>The timers are based on real time, that is, if you register an event that 714<p>The timers are based on real time, that is, if you register an event that
448times out after an hour and youreset your system clock to last years 715times out after an hour and you reset your system clock to last years
449time, it will still time out after (roughly) and hour. &quot;Roughly&quot; because 716time, it will still time out after (roughly) and hour. &quot;Roughly&quot; because
450detecting time jumps is hard, and soem inaccuracies are unavoidable (the 717detecting time jumps is hard, and some inaccuracies are unavoidable (the
451monotonic clock option helps a lot here).</p> 718monotonic clock option helps a lot here).</p>
719<p>The relative timeouts are calculated relative to the <code>ev_now ()</code>
720time. This is usually the right thing as this timestamp refers to the time
721of the event triggering whatever timeout you are modifying/starting. If
722you suspect event processing to be delayed and you <i>need</i> to base the timeout
723on the current time, use something like this to adjust for this:</p>
724<pre> ev_timer_set (&amp;timer, after + ev_now () - ev_time (), 0.);
725
726</pre>
727<p>The callback is guarenteed to be invoked only when its timeout has passed,
728but if multiple timers become ready during the same loop iteration then
729order of execution is undefined.</p>
452<dl> 730<dl>
453 <dt>ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)</dt> 731 <dt>ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)</dt>
454 <dt>ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)</dt> 732 <dt>ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)</dt>
455 <dd> 733 <dd>
456 <p>Configure the timer to trigger after <code>after</code> seconds. If <code>repeat</code> is 734 <p>Configure the timer to trigger after <code>after</code> seconds. If <code>repeat</code> is
… …
458timer will automatically be configured to trigger again <code>repeat</code> seconds 736timer will automatically be configured to trigger again <code>repeat</code> seconds
459later, again, and again, until stopped manually.</p> 737later, again, and again, until stopped manually.</p>
460 <p>The timer itself will do a best-effort at avoiding drift, that is, if you 738 <p>The timer itself will do a best-effort at avoiding drift, that is, if you
461configure a timer to trigger every 10 seconds, then it will trigger at 739configure a timer to trigger every 10 seconds, then it will trigger at
462exactly 10 second intervals. If, however, your program cannot keep up with 740exactly 10 second intervals. If, however, your program cannot keep up with
463the timer (ecause it takes longer than those 10 seconds to do stuff) the 741the timer (because it takes longer than those 10 seconds to do stuff) the
464timer will not fire more than once per event loop iteration.</p> 742timer will not fire more than once per event loop iteration.</p>
465 </dd> 743 </dd>
466 <dt>ev_timer_again (loop)</dt> 744 <dt>ev_timer_again (loop)</dt>
467 <dd> 745 <dd>
468 <p>This will act as if the timer timed out and restart it again if it is 746 <p>This will act as if the timer timed out and restart it again if it is
… …
472value), or reset the running timer to the repeat value.</p> 750value), or reset the running timer to the repeat value.</p>
473 <p>This sounds a bit complicated, but here is a useful and typical 751 <p>This sounds a bit complicated, but here is a useful and typical
474example: Imagine you have a tcp connection and you want a so-called idle 752example: Imagine you have a tcp connection and you want a so-called idle
475timeout, that is, you want to be called when there have been, say, 60 753timeout, that is, you want to be called when there have been, say, 60
476seconds of inactivity on the socket. The easiest way to do this is to 754seconds of inactivity on the socket. The easiest way to do this is to
477configure an ev_timer with after=repeat=60 and calling ev_timer_again each 755configure an <code>ev_timer</code> with after=repeat=60 and calling ev_timer_again each
478time you successfully read or write some data. If you go into an idle 756time you successfully read or write some data. If you go into an idle
479state where you do not expect data to travel on the socket, you can stop 757state where you do not expect data to travel on the socket, you can stop
480the timer, and again will automatically restart it if need be.</p> 758the timer, and again will automatically restart it if need be.</p>
481 </dd> 759 </dd>
482</dl> 760</dl>
761<p>Example: create a timer that fires after 60 seconds.</p>
762<pre> static void
763 one_minute_cb (struct ev_loop *loop, struct ev_timer *w, int revents)
764 {
765 .. one minute over, w is actually stopped right here
766 }
483 767
768 struct ev_timer mytimer;
769 ev_timer_init (&amp;mytimer, one_minute_cb, 60., 0.);
770 ev_timer_start (loop, &amp;mytimer);
771
772</pre>
773<p>Example: create a timeout timer that times out after 10 seconds of
774inactivity.</p>
775<pre> static void
776 timeout_cb (struct ev_loop *loop, struct ev_timer *w, int revents)
777 {
778 .. ten seconds without any activity
779 }
780
781 struct ev_timer mytimer;
782 ev_timer_init (&amp;mytimer, timeout_cb, 0., 10.); /* note, only repeat used */
783 ev_timer_again (&amp;mytimer); /* start timer */
784 ev_loop (loop, 0);
785
786 // and in some piece of code that gets executed on any &quot;activity&quot;:
787 // reset the timeout to start ticking again at 10 seconds
788 ev_timer_again (&amp;mytimer);
789
790
791
792
793</pre>
794
484</div> 795</div>
485<h2 id="ev_periodic_to_cron_or_not_to_cron_i">ev_periodic - to cron or not to cron it</h2> 796<h2 id="code_ev_periodic_code_to_cron_or_not"><code>ev_periodic</code> - to cron or not to cron</h2>
486<div id="ev_periodic_to_cron_or_not_to_cron_i-2"> 797<div id="code_ev_periodic_code_to_cron_or_not-2">
487<p>Periodic watchers are also timers of a kind, but they are very versatile 798<p>Periodic watchers are also timers of a kind, but they are very versatile
488(and unfortunately a bit complex).</p> 799(and unfortunately a bit complex).</p>
489<p>Unlike ev_timer's, they are not based on real time (or relative time) 800<p>Unlike <code>ev_timer</code>'s, they are not based on real time (or relative time)
490but on wallclock time (absolute time). You can tell a periodic watcher 801but on wallclock time (absolute time). You can tell a periodic watcher
491to trigger &quot;at&quot; some specific point in time. For example, if you tell a 802to trigger &quot;at&quot; some specific point in time. For example, if you tell a
492periodic watcher to trigger in 10 seconds (by specifiying e.g. c&lt;ev_now () 803periodic watcher to trigger in 10 seconds (by specifiying e.g. c&lt;ev_now ()
493+ 10.&gt;) and then reset your system clock to the last year, then it will 804+ 10.&gt;) and then reset your system clock to the last year, then it will
494take a year to trigger the event (unlike an ev_timer, which would trigger 805take a year to trigger the event (unlike an <code>ev_timer</code>, which would trigger
495roughly 10 seconds later and of course not if you reset your system time 806roughly 10 seconds later and of course not if you reset your system time
496again).</p> 807again).</p>
497<p>They can also be used to implement vastly more complex timers, such as 808<p>They can also be used to implement vastly more complex timers, such as
498triggering an event on eahc midnight, local time.</p> 809triggering an event on eahc midnight, local time.</p>
810<p>As with timers, the callback is guarenteed to be invoked only when the
811time (<code>at</code>) has been passed, but if multiple periodic timers become ready
812during the same loop iteration then order of execution is undefined.</p>
499<dl> 813<dl>
500 <dt>ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)</dt> 814 <dt>ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)</dt>
501 <dt>ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)</dt> 815 <dt>ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)</dt>
502 <dd> 816 <dd>
503 <p>Lots of arguments, lets sort it out... There are basically three modes of 817 <p>Lots of arguments, lets sort it out... There are basically three modes of
504operation, and we will explain them from simplest to complex:</p> 818operation, and we will explain them from simplest to complex:</p>
505
506
507
508
509 <p> 819 <p>
510 <dl> 820 <dl>
511 <dt>* absolute timer (interval = reschedule_cb = 0)</dt> 821 <dt>* absolute timer (interval = reschedule_cb = 0)</dt>
512 <dd> 822 <dd>
513 <p>In this configuration the watcher triggers an event at the wallclock time 823 <p>In this configuration the watcher triggers an event at the wallclock time
… …
525<pre> ev_periodic_set (&amp;periodic, 0., 3600., 0); 835<pre> ev_periodic_set (&amp;periodic, 0., 3600., 0);
526 836
527</pre> 837</pre>
528 <p>This doesn't mean there will always be 3600 seconds in between triggers, 838 <p>This doesn't mean there will always be 3600 seconds in between triggers,
529but only that the the callback will be called when the system time shows a 839but only that the the callback will be called when the system time shows a
530full hour (UTC), or more correct, when the system time is evenly divisible 840full hour (UTC), or more correctly, when the system time is evenly divisible
531by 3600.</p> 841by 3600.</p>
532 <p>Another way to think about it (for the mathematically inclined) is that 842 <p>Another way to think about it (for the mathematically inclined) is that
533ev_periodic will try to run the callback in this mode at the next possible 843<code>ev_periodic</code> will try to run the callback in this mode at the next possible
534time where <code>time = at (mod interval)</code>, regardless of any time jumps.</p> 844time where <code>time = at (mod interval)</code>, regardless of any time jumps.</p>
535 </dd> 845 </dd>
536 <dt>* manual reschedule mode (reschedule_cb = callback)</dt> 846 <dt>* manual reschedule mode (reschedule_cb = callback)</dt>
537 <dd> 847 <dd>
538 <p>In this mode the values for <code>interval</code> and <code>at</code> are both being 848 <p>In this mode the values for <code>interval</code> and <code>at</code> are both being
539ignored. Instead, each time the periodic watcher gets scheduled, the 849ignored. Instead, each time the periodic watcher gets scheduled, the
540reschedule callback will be called with the watcher as first, and the 850reschedule callback will be called with the watcher as first, and the
541current time as second argument.</p> 851current time as second argument.</p>
542 <p>NOTE: <i>This callback MUST NOT stop or destroy the periodic or any other 852 <p>NOTE: <i>This callback MUST NOT stop or destroy any periodic watcher,
543periodic watcher, ever, or make any event loop modificstions</i>. If you need 853ever, or make any event loop modifications</i>. If you need to stop it,
544to stop it, return 1e30 (or so, fudge fudge) and stop it afterwards.</p> 854return <code>now + 1e30</code> (or so, fudge fudge) and stop it afterwards (e.g. by
855starting a prepare watcher).</p>
545 <p>Its prototype is c&lt;ev_tstamp (*reschedule_cb)(struct ev_periodic *w, 856 <p>Its prototype is <code>ev_tstamp (*reschedule_cb)(struct ev_periodic *w,
546ev_tstamp now)&gt;, e.g.:</p> 857ev_tstamp now)</code>, e.g.:</p>
547<pre> static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now) 858<pre> static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now)
548 { 859 {
549 return now + 60.; 860 return now + 60.;
550 } 861 }
551 862
552</pre> 863</pre>
553 <p>It must return the next time to trigger, based on the passed time value 864 <p>It must return the next time to trigger, based on the passed time value
554(that is, the lowest time value larger than to the second argument). It 865(that is, the lowest time value larger than to the second argument). It
555will usually be called just before the callback will be triggered, but 866will usually be called just before the callback will be triggered, but
556might be called at other times, too.</p> 867might be called at other times, too.</p>
868 <p>NOTE: <i>This callback must always return a time that is later than the
869passed <code>now</code> value</i>. Not even <code>now</code> itself will do, it <i>must</i> be larger.</p>
557 <p>This can be used to create very complex timers, such as a timer that 870 <p>This can be used to create very complex timers, such as a timer that
558triggers on each midnight, local time. To do this, you would calculate the 871triggers on each midnight, local time. To do this, you would calculate the
559next midnight after <code>now</code> and return the timestamp value for this. How you do this 872next midnight after <code>now</code> and return the timestamp value for this. How
560is, again, up to you (but it is not trivial).</p> 873you do this is, again, up to you (but it is not trivial, which is the main
874reason I omitted it as an example).</p>
561 </dd> 875 </dd>
562 </dl> 876 </dl>
563 </p> 877 </p>
564 </dd> 878 </dd>
565 <dt>ev_periodic_again (loop, ev_periodic *)</dt> 879 <dt>ev_periodic_again (loop, ev_periodic *)</dt>
… …
568when you changed some parameters or the reschedule callback would return 882when you changed some parameters or the reschedule callback would return
569a different time than the last time it was called (e.g. in a crond like 883a different time than the last time it was called (e.g. in a crond like
570program when the crontabs have changed).</p> 884program when the crontabs have changed).</p>
571 </dd> 885 </dd>
572</dl> 886</dl>
887<p>Example: call a callback every hour, or, more precisely, whenever the
888system clock is divisible by 3600. The callback invocation times have
889potentially a lot of jittering, but good long-term stability.</p>
890<pre> static void
891 clock_cb (struct ev_loop *loop, struct ev_io *w, int revents)
892 {
893 ... its now a full hour (UTC, or TAI or whatever your clock follows)
894 }
573 895
896 struct ev_periodic hourly_tick;
897 ev_periodic_init (&amp;hourly_tick, clock_cb, 0., 3600., 0);
898 ev_periodic_start (loop, &amp;hourly_tick);
899
900</pre>
901<p>Example: the same as above, but use a reschedule callback to do it:</p>
902<pre> #include &lt;math.h&gt;
903
904 static ev_tstamp
905 my_scheduler_cb (struct ev_periodic *w, ev_tstamp now)
906 {
907 return fmod (now, 3600.) + 3600.;
908 }
909
910 ev_periodic_init (&amp;hourly_tick, clock_cb, 0., 0., my_scheduler_cb);
911
912</pre>
913<p>Example: call a callback every hour, starting now:</p>
914<pre> struct ev_periodic hourly_tick;
915 ev_periodic_init (&amp;hourly_tick, clock_cb,
916 fmod (ev_now (loop), 3600.), 3600., 0);
917 ev_periodic_start (loop, &amp;hourly_tick);
918
919
920
921
922</pre>
923
574</div> 924</div>
575<h2 id="ev_signal_signal_me_when_a_signal_ge">ev_signal - signal me when a signal gets signalled</h2> 925<h2 id="code_ev_signal_code_signal_me_when_a"><code>ev_signal</code> - signal me when a signal gets signalled</h2>
576<div id="ev_signal_signal_me_when_a_signal_ge-2"> 926<div id="code_ev_signal_code_signal_me_when_a-2">
577<p>Signal watchers will trigger an event when the process receives a specific 927<p>Signal watchers will trigger an event when the process receives a specific
578signal one or more times. Even though signals are very asynchronous, libev 928signal one or more times. Even though signals are very asynchronous, libev
579will try its best to deliver signals synchronously, i.e. as part of the 929will try it's best to deliver signals synchronously, i.e. as part of the
580normal event processing, like any other event.</p> 930normal event processing, like any other event.</p>
581<p>You cna configure as many watchers as you like per signal. Only when the 931<p>You can configure as many watchers as you like per signal. Only when the
582first watcher gets started will libev actually register a signal watcher 932first watcher gets started will libev actually register a signal watcher
583with the kernel (thus it coexists with your own signal handlers as long 933with the kernel (thus it coexists with your own signal handlers as long
584as you don't register any with libev). Similarly, when the last signal 934as you don't register any with libev). Similarly, when the last signal
585watcher for a signal is stopped libev will reset the signal handler to 935watcher for a signal is stopped libev will reset the signal handler to
586SIG_DFL (regardless of what it was set to before).</p> 936SIG_DFL (regardless of what it was set to before).</p>
… …
591 <p>Configures the watcher to trigger on the given signal number (usually one 941 <p>Configures the watcher to trigger on the given signal number (usually one
592of the <code>SIGxxx</code> constants).</p> 942of the <code>SIGxxx</code> constants).</p>
593 </dd> 943 </dd>
594</dl> 944</dl>
595 945
946
947
948
949
596</div> 950</div>
597<h2 id="ev_child_wait_for_pid_status_changes">ev_child - wait for pid status changes</h2> 951<h2 id="code_ev_child_code_wait_for_pid_stat"><code>ev_child</code> - wait for pid status changes</h2>
598<div id="ev_child_wait_for_pid_status_changes-2"> 952<div id="code_ev_child_code_wait_for_pid_stat-2">
599<p>Child watchers trigger when your process receives a SIGCHLD in response to 953<p>Child watchers trigger when your process receives a SIGCHLD in response to
600some child status changes (most typically when a child of yours dies).</p> 954some child status changes (most typically when a child of yours dies).</p>
601<dl> 955<dl>
602 <dt>ev_child_init (ev_child *, callback, int pid)</dt> 956 <dt>ev_child_init (ev_child *, callback, int pid)</dt>
603 <dt>ev_child_set (ev_child *, int pid)</dt> 957 <dt>ev_child_set (ev_child *, int pid)</dt>
604 <dd> 958 <dd>
605 <p>Configures the watcher to wait for status changes of process <code>pid</code> (or 959 <p>Configures the watcher to wait for status changes of process <code>pid</code> (or
606<i>any</i> process if <code>pid</code> is specified as <code>0</code>). The callback can look 960<i>any</i> process if <code>pid</code> is specified as <code>0</code>). The callback can look
607at the <code>rstatus</code> member of the <code>ev_child</code> watcher structure to see 961at the <code>rstatus</code> member of the <code>ev_child</code> watcher structure to see
608the status word (use the macros from <code>sys/wait.h</code>). The <code>rpid</code> member 962the status word (use the macros from <code>sys/wait.h</code> and see your systems
963<code>waitpid</code> documentation). The <code>rpid</code> member contains the pid of the
609contains the pid of the process causing the status change.</p> 964process causing the status change.</p>
610 </dd> 965 </dd>
611</dl> 966</dl>
967<p>Example: try to exit cleanly on SIGINT and SIGTERM.</p>
968<pre> static void
969 sigint_cb (struct ev_loop *loop, struct ev_signal *w, int revents)
970 {
971 ev_unloop (loop, EVUNLOOP_ALL);
972 }
612 973
974 struct ev_signal signal_watcher;
975 ev_signal_init (&amp;signal_watcher, sigint_cb, SIGINT);
976 ev_signal_start (loop, &amp;sigint_cb);
977
978
979
980
981</pre>
982
613</div> 983</div>
614<h2 id="ev_idle_when_you_ve_got_nothing_bett">ev_idle - when you've got nothing better to do</h2> 984<h2 id="code_ev_idle_code_when_you_ve_got_no"><code>ev_idle</code> - when you've got nothing better to do</h2>
615<div id="ev_idle_when_you_ve_got_nothing_bett-2"> 985<div id="code_ev_idle_code_when_you_ve_got_no-2">
616<p>Idle watchers trigger events when there are no other I/O or timer (or 986<p>Idle watchers trigger events when there are no other events are pending
617periodic) events pending. That is, as long as your process is busy 987(prepare, check and other idle watchers do not count). That is, as long
618handling sockets or timeouts it will not be called. But when your process 988as your process is busy handling sockets or timeouts (or even signals,
619is idle all idle watchers are being called again and again - until 989imagine) it will not be triggered. But when your process is idle all idle
990watchers are being called again and again, once per event loop iteration -
620stopped, that is, or your process receives more events.</p> 991until stopped, that is, or your process receives more events and becomes
992busy.</p>
621<p>The most noteworthy effect is that as long as any idle watchers are 993<p>The most noteworthy effect is that as long as any idle watchers are
622active, the process will not block when waiting for new events.</p> 994active, the process will not block when waiting for new events.</p>
623<p>Apart from keeping your process non-blocking (which is a useful 995<p>Apart from keeping your process non-blocking (which is a useful
624effect on its own sometimes), idle watchers are a good place to do 996effect on its own sometimes), idle watchers are a good place to do
625&quot;pseudo-background processing&quot;, or delay processing stuff to after the 997&quot;pseudo-background processing&quot;, or delay processing stuff to after the
… …
630 <p>Initialises and configures the idle watcher - it has no parameters of any 1002 <p>Initialises and configures the idle watcher - it has no parameters of any
631kind. There is a <code>ev_idle_set</code> macro, but using it is utterly pointless, 1003kind. There is a <code>ev_idle_set</code> macro, but using it is utterly pointless,
632believe me.</p> 1004believe me.</p>
633 </dd> 1005 </dd>
634</dl> 1006</dl>
1007<p>Example: dynamically allocate an <code>ev_idle</code>, start it, and in the
1008callback, free it. Alos, use no error checking, as usual.</p>
1009<pre> static void
1010 idle_cb (struct ev_loop *loop, struct ev_idle *w, int revents)
1011 {
1012 free (w);
1013 // now do something you wanted to do when the program has
1014 // no longer asnything immediate to do.
1015 }
635 1016
1017 struct ev_idle *idle_watcher = malloc (sizeof (struct ev_idle));
1018 ev_idle_init (idle_watcher, idle_cb);
1019 ev_idle_start (loop, idle_cb);
1020
1021
1022
1023
1024</pre>
1025
636</div> 1026</div>
637<h2 id="prepare_and_check_your_hooks_into_th">prepare and check - your hooks into the event loop</h2> 1027<h2 id="code_ev_prepare_code_and_code_ev_che"><code>ev_prepare</code> and <code>ev_check</code> - customise your event loop</h2>
638<div id="prepare_and_check_your_hooks_into_th-2"> 1028<div id="code_ev_prepare_code_and_code_ev_che-2">
639<p>Prepare and check watchers usually (but not always) are used in 1029<p>Prepare and check watchers are usually (but not always) used in tandem:
640tandom. Prepare watchers get invoked before the process blocks and check 1030prepare watchers get invoked before the process blocks and check watchers
641watchers afterwards.</p> 1031afterwards.</p>
642<p>Their main purpose is to integrate other event mechanisms into libev. This 1032<p>Their main purpose is to integrate other event mechanisms into libev and
643could be used, for example, to track variable changes, implement your own 1033their use is somewhat advanced. This could be used, for example, to track
644watchers, integrate net-snmp or a coroutine library and lots more.</p> 1034variable changes, implement your own watchers, integrate net-snmp or a
1035coroutine library and lots more.</p>
645<p>This is done by examining in each prepare call which file descriptors need 1036<p>This is done by examining in each prepare call which file descriptors need
646to be watched by the other library, registering ev_io watchers for them 1037to be watched by the other library, registering <code>ev_io</code> watchers for
647and starting an ev_timer watcher for any timeouts (many libraries provide 1038them and starting an <code>ev_timer</code> watcher for any timeouts (many libraries
648just this functionality). Then, in the check watcher you check for any 1039provide just this functionality). Then, in the check watcher you check for
649events that occured (by making your callbacks set soem flags for example) 1040any events that occured (by checking the pending status of all watchers
650and call back into the library.</p> 1041and stopping them) and call back into the library. The I/O and timer
1042callbacks will never actually be called (but must be valid nevertheless,
1043because you never know, you know?).</p>
651<p>As another example, the perl Coro module uses these hooks to integrate 1044<p>As another example, the Perl Coro module uses these hooks to integrate
652coroutines into libev programs, by yielding to other active coroutines 1045coroutines into libev programs, by yielding to other active coroutines
653during each prepare and only letting the process block if no coroutines 1046during each prepare and only letting the process block if no coroutines
654are ready to run.</p> 1047are ready to run (it's actually more complicated: it only runs coroutines
1048with priority higher than or equal to the event loop and one coroutine
1049of lower priority, but only once, using idle watchers to keep the event
1050loop from blocking if lower-priority coroutines are active, thus mapping
1051low-priority coroutines to idle/background tasks).</p>
655<dl> 1052<dl>
656 <dt>ev_prepare_init (ev_prepare *, callback)</dt> 1053 <dt>ev_prepare_init (ev_prepare *, callback)</dt>
657 <dt>ev_check_init (ev_check *, callback)</dt> 1054 <dt>ev_check_init (ev_check *, callback)</dt>
658 <dd> 1055 <dd>
659 <p>Initialises and configures the prepare or check watcher - they have no 1056 <p>Initialises and configures the prepare or check watcher - they have no
660parameters of any kind. There are <code>ev_prepare_set</code> and <code>ev_check_set</code> 1057parameters of any kind. There are <code>ev_prepare_set</code> and <code>ev_check_set</code>
661macros, but using them is utterly, utterly pointless.</p> 1058macros, but using them is utterly, utterly and completely pointless.</p>
662 </dd> 1059 </dd>
663</dl> 1060</dl>
1061<p>Example: *TODO*.</p>
1062
1063
1064
1065
1066
1067</div>
1068<h2 id="code_ev_embed_code_when_one_backend_"><code>ev_embed</code> - when one backend isn't enough</h2>
1069<div id="code_ev_embed_code_when_one_backend_-2">
1070<p>This is a rather advanced watcher type that lets you embed one event loop
1071into another.</p>
1072<p>There are primarily two reasons you would want that: work around bugs and
1073prioritise I/O.</p>
1074<p>As an example for a bug workaround, the kqueue backend might only support
1075sockets on some platform, so it is unusable as generic backend, but you
1076still want to make use of it because you have many sockets and it scales
1077so nicely. In this case, you would create a kqueue-based loop and embed it
1078into your default loop (which might use e.g. poll). Overall operation will
1079be a bit slower because first libev has to poll and then call kevent, but
1080at least you can use both at what they are best.</p>
1081<p>As for prioritising I/O: rarely you have the case where some fds have
1082to be watched and handled very quickly (with low latency), and even
1083priorities and idle watchers might have too much overhead. In this case
1084you would put all the high priority stuff in one loop and all the rest in
1085a second one, and embed the second one in the first.</p>
1086<p>As long as the watcher is started it will automatically handle events. The
1087callback will be invoked whenever some events have been handled. You can
1088set the callback to <code>0</code> to avoid having to specify one if you are not
1089interested in that.</p>
1090<p>Also, there have not currently been made special provisions for forking:
1091when you fork, you not only have to call <code>ev_loop_fork</code> on both loops,
1092but you will also have to stop and restart any <code>ev_embed</code> watchers
1093yourself.</p>
1094<p>Unfortunately, not all backends are embeddable, only the ones returned by
1095<code>ev_embeddable_backends</code> are, which, unfortunately, does not include any
1096portable one.</p>
1097<p>So when you want to use this feature you will always have to be prepared
1098that you cannot get an embeddable loop. The recommended way to get around
1099this is to have a separate variables for your embeddable loop, try to
1100create it, and if that fails, use the normal loop for everything:</p>
1101<pre> struct ev_loop *loop_hi = ev_default_init (0);
1102 struct ev_loop *loop_lo = 0;
1103 struct ev_embed embed;
1104
1105 // see if there is a chance of getting one that works
1106 // (remember that a flags value of 0 means autodetection)
1107 loop_lo = ev_embeddable_backends () &amp; ev_recommended_backends ()
1108 ? ev_loop_new (ev_embeddable_backends () &amp; ev_recommended_backends ())
1109 : 0;
1110
1111 // if we got one, then embed it, otherwise default to loop_hi
1112 if (loop_lo)
1113 {
1114 ev_embed_init (&amp;embed, 0, loop_lo);
1115 ev_embed_start (loop_hi, &amp;embed);
1116 }
1117 else
1118 loop_lo = loop_hi;
1119
1120</pre>
1121<dl>
1122 <dt>ev_embed_init (ev_embed *, callback, struct ev_loop *loop)</dt>
1123 <dt>ev_embed_set (ev_embed *, callback, struct ev_loop *loop)</dt>
1124 <dd>
1125 <p>Configures the watcher to embed the given loop, which must be embeddable.</p>
1126 </dd>
1127</dl>
1128
1129
1130
1131
664 1132
665</div> 1133</div>
666<h1 id="OTHER_FUNCTIONS">OTHER FUNCTIONS</h1><p><a href="#TOP" class="toplink">Top</a></p> 1134<h1 id="OTHER_FUNCTIONS">OTHER FUNCTIONS</h1><p><a href="#TOP" class="toplink">Top</a></p>
667<div id="OTHER_FUNCTIONS_CONTENT"> 1135<div id="OTHER_FUNCTIONS_CONTENT">
668<p>There are some other fucntions of possible interest. Described. Here. Now.</p> 1136<p>There are some other functions of possible interest. Described. Here. Now.</p>
669<dl> 1137<dl>
670 <dt>ev_once (loop, int fd, int events, ev_tstamp timeout, callback)</dt> 1138 <dt>ev_once (loop, int fd, int events, ev_tstamp timeout, callback)</dt>
671 <dd> 1139 <dd>
672 <p>This function combines a simple timer and an I/O watcher, calls your 1140 <p>This function combines a simple timer and an I/O watcher, calls your
673callback on whichever event happens first and automatically stop both 1141callback on whichever event happens first and automatically stop both
674watchers. This is useful if you want to wait for a single event on an fd 1142watchers. This is useful if you want to wait for a single event on an fd
675or timeout without havign to allocate/configure/start/stop/free one or 1143or timeout without having to allocate/configure/start/stop/free one or
676more watchers yourself.</p> 1144more watchers yourself.</p>
677 <p>If <code>fd</code> is less than 0, then no I/O watcher will be started and events is 1145 <p>If <code>fd</code> is less than 0, then no I/O watcher will be started and events
678ignored. Otherwise, an ev_io watcher for the given <code>fd</code> and <code>events</code> set 1146is being ignored. Otherwise, an <code>ev_io</code> watcher for the given <code>fd</code> and
679will be craeted and started.</p> 1147<code>events</code> set will be craeted and started.</p>
680 <p>If <code>timeout</code> is less than 0, then no timeout watcher will be 1148 <p>If <code>timeout</code> is less than 0, then no timeout watcher will be
681started. Otherwise an ev_timer watcher with after = <code>timeout</code> (and repeat 1149started. Otherwise an <code>ev_timer</code> watcher with after = <code>timeout</code> (and
682= 0) will be started.</p> 1150repeat = 0) will be started. While <code>0</code> is a valid timeout, it is of
1151dubious value.</p>
683 <p>The callback has the type <code>void (*cb)(int revents, void *arg)</code> and 1152 <p>The callback has the type <code>void (*cb)(int revents, void *arg)</code> and gets
684gets passed an events set (normally a combination of EV_ERROR, EV_READ, 1153passed an <code>revents</code> set like normal event callbacks (a combination of
685EV_WRITE or EV_TIMEOUT) and the <code>arg</code> value passed to <code>ev_once</code>:</p> 1154<code>EV_ERROR</code>, <code>EV_READ</code>, <code>EV_WRITE</code> or <code>EV_TIMEOUT</code>) and the <code>arg</code>
1155value passed to <code>ev_once</code>:</p>
686<pre> static void stdin_ready (int revents, void *arg) 1156<pre> static void stdin_ready (int revents, void *arg)
687 { 1157 {
688 if (revents &amp; EV_TIMEOUT) 1158 if (revents &amp; EV_TIMEOUT)
689 /* doh, nothing entered */ 1159 /* doh, nothing entered */;
690 else if (revents &amp; EV_READ) 1160 else if (revents &amp; EV_READ)
691 /* stdin might have data for us, joy! */ 1161 /* stdin might have data for us, joy! */;
692 } 1162 }
693 1163
694 ev_once (STDIN_FILENO, EV_READm 10., stdin_ready, 0); 1164 ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);
695 1165
696</pre> 1166</pre>
697 </dd> 1167 </dd>
698 <dt>ev_feed_event (loop, watcher, int events)</dt> 1168 <dt>ev_feed_event (loop, watcher, int events)</dt>
699 <dd> 1169 <dd>
700 <p>Feeds the given event set into the event loop, as if the specified event 1170 <p>Feeds the given event set into the event loop, as if the specified event
701has happened for the specified watcher (which must be a pointer to an 1171had happened for the specified watcher (which must be a pointer to an
702initialised but not necessarily active event watcher).</p> 1172initialised but not necessarily started event watcher).</p>
703 </dd> 1173 </dd>
704 <dt>ev_feed_fd_event (loop, int fd, int revents)</dt> 1174 <dt>ev_feed_fd_event (loop, int fd, int revents)</dt>
705 <dd> 1175 <dd>
706 <p>Feed an event on the given fd, as if a file descriptor backend detected it.</p> 1176 <p>Feed an event on the given fd, as if a file descriptor backend detected
1177the given events it.</p>
707 </dd> 1178 </dd>
708 <dt>ev_feed_signal_event (loop, int signum)</dt> 1179 <dt>ev_feed_signal_event (loop, int signum)</dt>
709 <dd> 1180 <dd>
710 <p>Feed an event as if the given signal occured (loop must be the default loop!).</p> 1181 <p>Feed an event as if the given signal occured (loop must be the default loop!).</p>
711 </dd> 1182 </dd>
712</dl> 1183</dl>
1184
1185
1186
1187
1188
1189</div>
1190<h1 id="LIBEVENT_EMULATION">LIBEVENT EMULATION</h1><p><a href="#TOP" class="toplink">Top</a></p>
1191<div id="LIBEVENT_EMULATION_CONTENT">
1192<p>Libev offers a compatibility emulation layer for libevent. It cannot
1193emulate the internals of libevent, so here are some usage hints:</p>
1194<dl>
1195 <dt>* Use it by including &lt;event.h&gt;, as usual.</dt>
1196 <dt>* The following members are fully supported: ev_base, ev_callback,
1197ev_arg, ev_fd, ev_res, ev_events.</dt>
1198 <dt>* Avoid using ev_flags and the EVLIST_*-macros, while it is
1199maintained by libev, it does not work exactly the same way as in libevent (consider
1200it a private API).</dt>
1201 <dt>* Priorities are not currently supported. Initialising priorities
1202will fail and all watchers will have the same priority, even though there
1203is an ev_pri field.</dt>
1204 <dt>* Other members are not supported.</dt>
1205 <dt>* The libev emulation is <i>not</i> ABI compatible to libevent, you need
1206to use the libev header file and library.</dt>
1207</dl>
1208
1209</div>
1210<h1 id="C_SUPPORT">C++ SUPPORT</h1><p><a href="#TOP" class="toplink">Top</a></p>
1211<div id="C_SUPPORT_CONTENT">
1212<p>TBD.</p>
713 1213
714</div> 1214</div>
715<h1 id="AUTHOR">AUTHOR</h1><p><a href="#TOP" class="toplink">Top</a></p> 1215<h1 id="AUTHOR">AUTHOR</h1><p><a href="#TOP" class="toplink">Top</a></p>
716<div id="AUTHOR_CONTENT"> 1216<div id="AUTHOR_CONTENT">
717<p>Marc Lehmann &lt;libev@schmorp.de&gt;.</p> 1217<p>Marc Lehmann &lt;libev@schmorp.de&gt;.</p>

Diff Legend

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