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

Comparing libev/ev.html (file contents):
Revision 1.7 by root, Mon Nov 12 08:16:02 2007 UTC vs.
Revision 1.59 by root, Wed Nov 28 17:32:24 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 09:16:01 2007" /> 9 <meta name="created" content="Wed Nov 28 18:32:11 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 -->
15<h3 id="TOP">Index</h3> 15<h3 id="TOP">Index</h3>
16 16
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="#EXAMPLE_PROGRAM">EXAMPLE PROGRAM</a></li>
19<li><a href="#DESCRIPTION">DESCRIPTION</a></li> 20<li><a href="#DESCRIPTION">DESCRIPTION</a></li>
20<li><a href="#FEATURES">FEATURES</a></li> 21<li><a href="#FEATURES">FEATURES</a></li>
21<li><a href="#CONVENTIONS">CONVENTIONS</a></li> 22<li><a href="#CONVENTIONS">CONVENTIONS</a></li>
23<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> 24<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> 25<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> 26<li><a href="#ANATOMY_OF_A_WATCHER">ANATOMY OF A WATCHER</a>
27<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> 28<li><a href="#ASSOCIATING_CUSTOM_DATA_WITH_A_WATCH">ASSOCIATING CUSTOM DATA WITH A WATCHER</a></li>
26</ul> 29</ul>
27</li> 30</li>
28<li><a href="#WATCHER_TYPES">WATCHER TYPES</a> 31<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> 32<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> 33<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> 34<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> 35<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> 36<li><a href="#code_ev_child_code_watch_out_for_pro"><code>ev_child</code> - watch out for process status changes</a></li>
37<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> 38<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> 39<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>
40<li><a href="#code_ev_embed_code_when_one_backend_"><code>ev_embed</code> - when one backend isn't enough...</a></li>
41<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> 42</ul>
37</li> 43</li>
38<li><a href="#OTHER_FUNCTIONS">OTHER FUNCTIONS</a></li> 44<li><a href="#OTHER_FUNCTIONS">OTHER FUNCTIONS</a></li>
45<li><a href="#LIBEVENT_EMULATION">LIBEVENT EMULATION</a></li>
46<li><a href="#C_SUPPORT">C++ SUPPORT</a></li>
47<li><a href="#MACRO_MAGIC">MACRO MAGIC</a></li>
48<li><a href="#EMBEDDING">EMBEDDING</a>
49<ul><li><a href="#FILESETS">FILESETS</a>
50<ul><li><a href="#CORE_EVENT_LOOP">CORE EVENT LOOP</a></li>
51<li><a href="#LIBEVENT_COMPATIBILITY_API">LIBEVENT COMPATIBILITY API</a></li>
52<li><a href="#AUTOCONF_SUPPORT">AUTOCONF SUPPORT</a></li>
53</ul>
54</li>
55<li><a href="#PREPROCESSOR_SYMBOLS_MACROS">PREPROCESSOR SYMBOLS/MACROS</a></li>
56<li><a href="#EXAMPLES">EXAMPLES</a></li>
57</ul>
58</li>
59<li><a href="#COMPLEXITIES">COMPLEXITIES</a></li>
39<li><a href="#AUTHOR">AUTHOR</a> 60<li><a href="#AUTHOR">AUTHOR</a>
40</li> 61</li>
41</ul><hr /> 62</ul><hr />
42<!-- INDEX END --> 63<!-- INDEX END -->
43 64
44<h1 id="NAME">NAME</h1><p><a href="#TOP" class="toplink">Top</a></p> 65<h1 id="NAME">NAME</h1>
45<div id="NAME_CONTENT"> 66<div id="NAME_CONTENT">
46<p>libev - a high performance full-featured event loop written in C</p> 67<p>libev - a high performance full-featured event loop written in C</p>
47 68
48</div> 69</div>
49<h1 id="SYNOPSIS">SYNOPSIS</h1><p><a href="#TOP" class="toplink">Top</a></p> 70<h1 id="SYNOPSIS">SYNOPSIS</h1>
50<div id="SYNOPSIS_CONTENT"> 71<div id="SYNOPSIS_CONTENT">
51<pre> #include &lt;ev.h&gt; 72<pre> #include &lt;ev.h&gt;
52 73
53</pre> 74</pre>
54 75
55</div> 76</div>
56<h1 id="DESCRIPTION">DESCRIPTION</h1><p><a href="#TOP" class="toplink">Top</a></p> 77<h1 id="EXAMPLE_PROGRAM">EXAMPLE PROGRAM</h1>
78<div id="EXAMPLE_PROGRAM_CONTENT">
79<pre> #include &lt;ev.h&gt;
80
81 ev_io stdin_watcher;
82 ev_timer timeout_watcher;
83
84 /* called when data readable on stdin */
85 static void
86 stdin_cb (EV_P_ struct ev_io *w, int revents)
87 {
88 /* puts (&quot;stdin ready&quot;); */
89 ev_io_stop (EV_A_ w); /* just a syntax example */
90 ev_unloop (EV_A_ EVUNLOOP_ALL); /* leave all loop calls */
91 }
92
93 static void
94 timeout_cb (EV_P_ struct ev_timer *w, int revents)
95 {
96 /* puts (&quot;timeout&quot;); */
97 ev_unloop (EV_A_ EVUNLOOP_ONE); /* leave one loop call */
98 }
99
100 int
101 main (void)
102 {
103 struct ev_loop *loop = ev_default_loop (0);
104
105 /* initialise an io watcher, then start it */
106 ev_io_init (&amp;stdin_watcher, stdin_cb, /*STDIN_FILENO*/ 0, EV_READ);
107 ev_io_start (loop, &amp;stdin_watcher);
108
109 /* simple non-repeating 5.5 second timeout */
110 ev_timer_init (&amp;timeout_watcher, timeout_cb, 5.5, 0.);
111 ev_timer_start (loop, &amp;timeout_watcher);
112
113 /* loop till timeout or data ready */
114 ev_loop (loop, 0);
115
116 return 0;
117 }
118
119</pre>
120
121</div>
122<h1 id="DESCRIPTION">DESCRIPTION</h1>
57<div id="DESCRIPTION_CONTENT"> 123<div id="DESCRIPTION_CONTENT">
58<p>Libev is an event loop: you register interest in certain events (such as a 124<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 125file descriptor being readable or a timeout occuring), and it will manage
60these event sources and provide your program with events.</p> 126these 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 127<p>To do this, it must take more or less complete control over your process
65watchers</i>, which are relatively small C structures you initialise with the 131watchers</i>, which are relatively small C structures you initialise with the
66details of the event, and then hand it over to libev by <i>starting</i> the 132details of the event, and then hand it over to libev by <i>starting</i> the
67watcher.</p> 133watcher.</p>
68 134
69</div> 135</div>
70<h1 id="FEATURES">FEATURES</h1><p><a href="#TOP" class="toplink">Top</a></p> 136<h1 id="FEATURES">FEATURES</h1>
71<div id="FEATURES_CONTENT"> 137<div id="FEATURES_CONTENT">
72<p>Libev supports select, poll, the linux-specific epoll and the bsd-specific 138<p>Libev supports <code>select</code>, <code>poll</code>, the Linux-specific <code>epoll</code>, the
73kqueue mechanisms for file descriptor events, relative timers, absolute 139BSD-specific <code>kqueue</code> and the Solaris-specific event port mechanisms
74timers with customised rescheduling, signal events, process status change 140for file descriptor events (<code>ev_io</code>), the Linux <code>inotify</code> interface
75events (related to SIGCHLD), and event watchers dealing with the event 141(for <code>ev_stat</code>), relative timers (<code>ev_timer</code>), absolute timers
76loop mechanism itself (idle, prepare and check watchers). It also is quite 142with customised rescheduling (<code>ev_periodic</code>), synchronous signals
143(<code>ev_signal</code>), process status change events (<code>ev_child</code>), and event
144watchers dealing with the event loop mechanism itself (<code>ev_idle</code>,
145<code>ev_embed</code>, <code>ev_prepare</code> and <code>ev_check</code> watchers) as well as
146file watchers (<code>ev_stat</code>) and even limited support for fork events
147(<code>ev_fork</code>).</p>
148<p>It also is quite fast (see this
77fast (see this <a href="http://libev.schmorp.de/bench.html">benchmark</a> comparing 149<a href="http://libev.schmorp.de/bench.html">benchmark</a> comparing it to libevent
78it to libevent for example).</p> 150for example).</p>
79 151
80</div> 152</div>
81<h1 id="CONVENTIONS">CONVENTIONS</h1><p><a href="#TOP" class="toplink">Top</a></p> 153<h1 id="CONVENTIONS">CONVENTIONS</h1>
82<div id="CONVENTIONS_CONTENT"> 154<div id="CONVENTIONS_CONTENT">
83<p>Libev is very configurable. In this manual the default configuration 155<p>Libev is very configurable. In this manual the default configuration will
84will be described, which supports multiple event loops. For more info 156be described, which supports multiple event loops. For more info about
85about various configuration options please have a look at the file 157various configuration options please have a look at <strong>EMBED</strong> section in
86<cite>README.embed</cite> in the libev distribution. If libev was configured without 158this manual. If libev was configured without support for multiple event
87support for multiple event loops, then all functions taking an initial 159loops, then all functions taking an initial argument of name <code>loop</code>
88argument of name <code>loop</code> (which is always of type <code>struct ev_loop *</code>) 160(which is always of type <code>struct ev_loop *</code>) will not have this argument.</p>
89will not have this argument.</p>
90 161
91</div> 162</div>
92<h1 id="TIME_AND_OTHER_GLOBAL_FUNCTIONS">TIME AND OTHER GLOBAL FUNCTIONS</h1><p><a href="#TOP" class="toplink">Top</a></p> 163<h1 id="TIME_REPRESENTATION">TIME REPRESENTATION</h1>
93<div id="TIME_AND_OTHER_GLOBAL_FUNCTIONS_CONT"> 164<div id="TIME_REPRESENTATION_CONTENT">
94<p>Libev represents time as a single floating point number, representing the 165<p>Libev represents time as a single floating point number, representing the
95(fractional) number of seconds since the (POSIX) epoch (somewhere near 166(fractional) number of seconds since the (POSIX) epoch (somewhere near
96the beginning of 1970, details are complicated, don't ask). This type is 167the beginning of 1970, details are complicated, don't ask). This type is
97called <code>ev_tstamp</code>, which is what you should use too. It usually aliases 168called <code>ev_tstamp</code>, which is what you should use too. It usually aliases
98to the double type in C.</p> 169to the <code>double</code> type in C, and when you need to do any calculations on
170it, you should treat it as such.</p>
171
172</div>
173<h1 id="GLOBAL_FUNCTIONS">GLOBAL FUNCTIONS</h1>
174<div id="GLOBAL_FUNCTIONS_CONTENT">
175<p>These functions can be called anytime, even before initialising the
176library in any way.</p>
99<dl> 177<dl>
100 <dt>ev_tstamp ev_time ()</dt> 178 <dt>ev_tstamp ev_time ()</dt>
101 <dd> 179 <dd>
102 <p>Returns the current time as libev would use it.</p> 180 <p>Returns the current time as libev would use it. Please note that the
181<code>ev_now</code> function is usually faster and also often returns the timestamp
182you actually want to know.</p>
103 </dd> 183 </dd>
104 <dt>int ev_version_major ()</dt> 184 <dt>int ev_version_major ()</dt>
105 <dt>int ev_version_minor ()</dt> 185 <dt>int ev_version_minor ()</dt>
106 <dd> 186 <dd>
107 <p>You can find out the major and minor version numbers of the library 187 <p>You can find out the major and minor version numbers of the library
108you linked against by calling the functions <code>ev_version_major</code> and 188you 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 189<code>ev_version_minor</code>. If you want, you can compare against the global
110symbols <code>EV_VERSION_MAJOR</code> and <code>EV_VERSION_MINOR</code>, which specify the 190symbols <code>EV_VERSION_MAJOR</code> and <code>EV_VERSION_MINOR</code>, which specify the
111version of the library your program was compiled against.</p> 191version of the library your program was compiled against.</p>
112 <p>Usually, its a good idea to terminate if the major versions mismatch, 192 <p>Usually, it's a good idea to terminate if the major versions mismatch,
113as this indicates an incompatible change. Minor versions are usually 193as this indicates an incompatible change. Minor versions are usually
114compatible to older versions, so a larger minor version alone is usually 194compatible to older versions, so a larger minor version alone is usually
115not a problem.</p> 195not a problem.</p>
196 <p>Example: Make sure we haven't accidentally been linked against the wrong
197version.</p>
198<pre> assert ((&quot;libev version mismatch&quot;,
199 ev_version_major () == EV_VERSION_MAJOR
200 &amp;&amp; ev_version_minor () &gt;= EV_VERSION_MINOR));
201
202</pre>
203 </dd>
204 <dt>unsigned int ev_supported_backends ()</dt>
205 <dd>
206 <p>Return the set of all backends (i.e. their corresponding <code>EV_BACKEND_*</code>
207value) compiled into this binary of libev (independent of their
208availability on the system you are running on). See <code>ev_default_loop</code> for
209a description of the set values.</p>
210 <p>Example: make sure we have the epoll method, because yeah this is cool and
211a must have and can we have a torrent of it please!!!11</p>
212<pre> assert ((&quot;sorry, no epoll, no sex&quot;,
213 ev_supported_backends () &amp; EVBACKEND_EPOLL));
214
215</pre>
216 </dd>
217 <dt>unsigned int ev_recommended_backends ()</dt>
218 <dd>
219 <p>Return the set of all backends compiled into this binary of libev and also
220recommended for this platform. This set is often smaller than the one
221returned by <code>ev_supported_backends</code>, as for example kqueue is broken on
222most BSDs and will not be autodetected unless you explicitly request it
223(assuming you know what you are doing). This is the set of backends that
224libev will probe for if you specify no backends explicitly.</p>
225 </dd>
226 <dt>unsigned int ev_embeddable_backends ()</dt>
227 <dd>
228 <p>Returns the set of backends that are embeddable in other event loops. This
229is the theoretical, all-platform, value. To find which backends
230might be supported on the current system, you would need to look at
231<code>ev_embeddable_backends () &amp; ev_supported_backends ()</code>, likewise for
232recommended ones.</p>
233 <p>See the description of <code>ev_embed</code> watchers for more info.</p>
116 </dd> 234 </dd>
117 <dt>ev_set_allocator (void *(*cb)(void *ptr, long size))</dt> 235 <dt>ev_set_allocator (void *(*cb)(void *ptr, long size))</dt>
118 <dd> 236 <dd>
119 <p>Sets the allocation function to use (the prototype is similar to the 237 <p>Sets the allocation function to use (the prototype is similar - the
120realloc C function, the semantics are identical). It is used to allocate 238semantics is identical - to the realloc C function). It is used to
121and free memory (no surprises here). If it returns zero when memory 239allocate and free memory (no surprises here). If it returns zero when
122needs to be allocated, the library might abort or take some potentially 240memory needs to be allocated, the library might abort or take some
123destructive action. The default is your system realloc function.</p> 241potentially destructive action. The default is your system realloc
242function.</p>
124 <p>You could override this function in high-availability programs to, say, 243 <p>You could override this function in high-availability programs to, say,
125free some memory if it cannot allocate memory, to use a special allocator, 244free some memory if it cannot allocate memory, to use a special allocator,
126or even to sleep a while and retry until some memory is available.</p> 245or even to sleep a while and retry until some memory is available.</p>
246 <p>Example: Replace the libev allocator with one that waits a bit and then
247retries).</p>
248<pre> static void *
249 persistent_realloc (void *ptr, size_t size)
250 {
251 for (;;)
252 {
253 void *newptr = realloc (ptr, size);
254
255 if (newptr)
256 return newptr;
257
258 sleep (60);
259 }
260 }
261
262 ...
263 ev_set_allocator (persistent_realloc);
264
265</pre>
127 </dd> 266 </dd>
128 <dt>ev_set_syserr_cb (void (*cb)(const char *msg));</dt> 267 <dt>ev_set_syserr_cb (void (*cb)(const char *msg));</dt>
129 <dd> 268 <dd>
130 <p>Set the callback function to call on a retryable syscall error (such 269 <p>Set the callback function to call on a retryable syscall error (such
131as failed select, poll, epoll_wait). The message is a printable string 270as failed select, poll, epoll_wait). The message is a printable string
132indicating the system call or subsystem causing the problem. If this 271indicating the system call or subsystem causing the problem. If this
133callback is set, then libev will expect it to remedy the sitution, no 272callback is set, then libev will expect it to remedy the sitution, no
134matter what, when it returns. That is, libev will generally retry the 273matter what, when it returns. That is, libev will generally retry the
135requested operation, or, if the condition doesn't go away, do bad stuff 274requested operation, or, if the condition doesn't go away, do bad stuff
136(such as abort).</p> 275(such as abort).</p>
276 <p>Example: This is basically the same thing that libev does internally, too.</p>
277<pre> static void
278 fatal_error (const char *msg)
279 {
280 perror (msg);
281 abort ();
282 }
283
284 ...
285 ev_set_syserr_cb (fatal_error);
286
287</pre>
137 </dd> 288 </dd>
138</dl> 289</dl>
139 290
140</div> 291</div>
141<h1 id="FUNCTIONS_CONTROLLING_THE_EVENT_LOOP">FUNCTIONS CONTROLLING THE EVENT LOOP</h1><p><a href="#TOP" class="toplink">Top</a></p> 292<h1 id="FUNCTIONS_CONTROLLING_THE_EVENT_LOOP">FUNCTIONS CONTROLLING THE EVENT LOOP</h1>
142<div id="FUNCTIONS_CONTROLLING_THE_EVENT_LOOP-2"> 293<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 294<p>An event loop is described by a <code>struct ev_loop *</code>. The library knows two
144types of such loops, the <i>default</i> loop, which supports signals and child 295types of such loops, the <i>default</i> loop, which supports signals and child
145events, and dynamically created loops which do not.</p> 296events, and dynamically created loops which do not.</p>
146<p>If you use threads, a common model is to run the default event loop 297<p>If you use threads, a common model is to run the default event loop
147in your main thread (or in a separate thrad) and for each thread you 298in your main thread (or in a separate thread) and for each thread you
148create, you also create another event loop. Libev itself does no locking 299create, you also create another event loop. Libev itself does no locking
149whatsoever, so if you mix calls to the same event loop in different 300whatsoever, so if you mix calls to the same event loop in different
150threads, make sure you lock (this is usually a bad idea, though, even if 301threads, make sure you lock (this is usually a bad idea, though, even if
151done correctly, because its hideous and inefficient).</p> 302done correctly, because it's hideous and inefficient).</p>
152<dl> 303<dl>
153 <dt>struct ev_loop *ev_default_loop (unsigned int flags)</dt> 304 <dt>struct ev_loop *ev_default_loop (unsigned int flags)</dt>
154 <dd> 305 <dd>
155 <p>This will initialise the default event loop if it hasn't been initialised 306 <p>This will initialise the default event loop if it hasn't been initialised
156yet and return it. If the default loop could not be initialised, returns 307yet and return it. If the default loop could not be initialised, returns
157false. If it already was initialised it simply returns it (and ignores the 308false. If it already was initialised it simply returns it (and ignores the
158flags).</p> 309flags. If that is troubling you, check <code>ev_backend ()</code> afterwards).</p>
159 <p>If you don't know what event loop to use, use the one returned from this 310 <p>If you don't know what event loop to use, use the one returned from this
160function.</p> 311function.</p>
161 <p>The flags argument can be used to specify special behaviour or specific 312 <p>The flags argument can be used to specify special behaviour or specific
162backends to use, and is usually specified as 0 (or EVFLAG_AUTO)</p> 313backends to use, and is usually specified as <code>0</code> (or <code>EVFLAG_AUTO</code>).</p>
163 <p>It supports the following flags:</p> 314 <p>The following flags are supported:</p>
164 <p> 315 <p>
165 <dl> 316 <dl>
166 <dt>EVFLAG_AUTO</dt> 317 <dt><code>EVFLAG_AUTO</code></dt>
167 <dd> 318 <dd>
168 <p>The default flags value. Use this if you have no clue (its the right 319 <p>The default flags value. Use this if you have no clue (it's the right
169thing, believe me).</p> 320thing, believe me).</p>
170 </dd> 321 </dd>
171 <dt>EVFLAG_NOENV</dt> 322 <dt><code>EVFLAG_NOENV</code></dt>
172 <dd> 323 <dd>
173 <p>If this flag bit is ored into the flag value then libev will <i>not</i> look 324 <p>If this flag bit is ored into the flag value (or the program runs setuid
174at the environment variable <code>LIBEV_FLAGS</code>. Otherwise (the default), this 325or setgid) then libev will <i>not</i> look at the environment variable
175environment variable will override the flags completely. This is useful 326<code>LIBEV_FLAGS</code>. Otherwise (the default), this environment variable will
327override the flags completely if it is found in the environment. This is
176to try out specific backends to tets their performance, or to work around 328useful to try out specific backends to test their performance, or to work
177bugs.</p> 329around bugs.</p>
178 </dd> 330 </dd>
179 <dt>EVMETHOD_SELECT portable select backend</dt> 331 <dt><code>EVBACKEND_SELECT</code> (value 1, portable select backend)</dt>
180 <dt>EVMETHOD_POLL poll backend (everywhere except windows)</dt>
181 <dt>EVMETHOD_EPOLL linux only</dt>
182 <dt>EVMETHOD_KQUEUE some bsds only</dt>
183 <dt>EVMETHOD_DEVPOLL solaris 8 only</dt>
184 <dt>EVMETHOD_PORT solaris 10 only</dt>
185 <dd> 332 <dd>
186 <p>If one or more of these are ored into the flags value, then only these 333 <p>This is your standard select(2) backend. Not <i>completely</i> standard, as
187backends will be tried (in the reverse order as given here). If one are 334libev tries to roll its own fd_set with no limits on the number of fds,
188specified, any backend will do.</p> 335but if that fails, expect a fairly low limit on the number of fds when
336using this backend. It doesn't scale too well (O(highest_fd)), but its usually
337the fastest backend for a low number of fds.</p>
338 </dd>
339 <dt><code>EVBACKEND_POLL</code> (value 2, poll backend, available everywhere except on windows)</dt>
340 <dd>
341 <p>And this is your standard poll(2) backend. It's more complicated than
342select, but handles sparse fds better and has no artificial limit on the
343number of fds you can use (except it will slow down considerably with a
344lot of inactive fds). It scales similarly to select, i.e. O(total_fds).</p>
345 </dd>
346 <dt><code>EVBACKEND_EPOLL</code> (value 4, Linux)</dt>
347 <dd>
348 <p>For few fds, this backend is a bit little slower than poll and select,
349but it scales phenomenally better. While poll and select usually scale like
350O(total_fds) where n is the total number of fds (or the highest fd), epoll scales
351either O(1) or O(active_fds).</p>
352 <p>While stopping and starting an I/O watcher in the same iteration will
353result in some caching, there is still a syscall per such incident
354(because the fd could point to a different file description now), so its
355best to avoid that. Also, dup()ed file descriptors might not work very
356well if you register events for both fds.</p>
357 <p>Please note that epoll sometimes generates spurious notifications, so you
358need to use non-blocking I/O or other means to avoid blocking when no data
359(or space) is available.</p>
360 </dd>
361 <dt><code>EVBACKEND_KQUEUE</code> (value 8, most BSD clones)</dt>
362 <dd>
363 <p>Kqueue deserves special mention, as at the time of this writing, it
364was broken on all BSDs except NetBSD (usually it doesn't work with
365anything but sockets and pipes, except on Darwin, where of course its
366completely useless). For this reason its not being &quot;autodetected&quot;
367unless you explicitly specify it explicitly in the flags (i.e. using
368<code>EVBACKEND_KQUEUE</code>).</p>
369 <p>It scales in the same way as the epoll backend, but the interface to the
370kernel is more efficient (which says nothing about its actual speed, of
371course). While starting and stopping an I/O watcher does not cause an
372extra syscall as with epoll, it still adds up to four event changes per
373incident, so its best to avoid that.</p>
374 </dd>
375 <dt><code>EVBACKEND_DEVPOLL</code> (value 16, Solaris 8)</dt>
376 <dd>
377 <p>This is not implemented yet (and might never be).</p>
378 </dd>
379 <dt><code>EVBACKEND_PORT</code> (value 32, Solaris 10)</dt>
380 <dd>
381 <p>This uses the Solaris 10 port mechanism. As with everything on Solaris,
382it's really slow, but it still scales very well (O(active_fds)).</p>
383 <p>Please note that solaris ports can result in a lot of spurious
384notifications, so you need to use non-blocking I/O or other means to avoid
385blocking when no data (or space) is available.</p>
386 </dd>
387 <dt><code>EVBACKEND_ALL</code></dt>
388 <dd>
389 <p>Try all backends (even potentially broken ones that wouldn't be tried
390with <code>EVFLAG_AUTO</code>). Since this is a mask, you can do stuff such as
391<code>EVBACKEND_ALL &amp; ~EVBACKEND_KQUEUE</code>.</p>
189 </dd> 392 </dd>
190 </dl> 393 </dl>
191 </p> 394 </p>
395 <p>If one or more of these are ored into the flags value, then only these
396backends will be tried (in the reverse order as given here). If none are
397specified, most compiled-in backend will be tried, usually in reverse
398order of their flag values :)</p>
399 <p>The most typical usage is like this:</p>
400<pre> if (!ev_default_loop (0))
401 fatal (&quot;could not initialise libev, bad $LIBEV_FLAGS in environment?&quot;);
402
403</pre>
404 <p>Restrict libev to the select and poll backends, and do not allow
405environment settings to be taken into account:</p>
406<pre> ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);
407
408</pre>
409 <p>Use whatever libev has to offer, but make sure that kqueue is used if
410available (warning, breaks stuff, best use only with your own private
411event loop and only if you know the OS supports your types of fds):</p>
412<pre> ev_default_loop (ev_recommended_backends () | EVBACKEND_KQUEUE);
413
414</pre>
192 </dd> 415 </dd>
193 <dt>struct ev_loop *ev_loop_new (unsigned int flags)</dt> 416 <dt>struct ev_loop *ev_loop_new (unsigned int flags)</dt>
194 <dd> 417 <dd>
195 <p>Similar to <code>ev_default_loop</code>, but always creates a new event loop that is 418 <p>Similar to <code>ev_default_loop</code>, but always creates a new event loop that is
196always distinct from the default loop. Unlike the default loop, it cannot 419always distinct from the default loop. Unlike the default loop, it cannot
197handle signal and child watchers, and attempts to do so will be greeted by 420handle signal and child watchers, and attempts to do so will be greeted by
198undefined behaviour (or a failed assertion if assertions are enabled).</p> 421undefined behaviour (or a failed assertion if assertions are enabled).</p>
422 <p>Example: Try to create a event loop that uses epoll and nothing else.</p>
423<pre> struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV);
424 if (!epoller)
425 fatal (&quot;no epoll found here, maybe it hides under your chair&quot;);
426
427</pre>
199 </dd> 428 </dd>
200 <dt>ev_default_destroy ()</dt> 429 <dt>ev_default_destroy ()</dt>
201 <dd> 430 <dd>
202 <p>Destroys the default loop again (frees all memory and kernel state 431 <p>Destroys the default loop again (frees all memory and kernel state
203etc.). This stops all registered event watchers (by not touching them in 432etc.). None of the active event watchers will be stopped in the normal
204any way whatsoever, although you cnanot rely on this :).</p> 433sense, so e.g. <code>ev_is_active</code> might still return true. It is your
434responsibility to either stop all watchers cleanly yoursef <i>before</i>
435calling this function, or cope with the fact afterwards (which is usually
436the easiest thing, youc na just ignore the watchers and/or <code>free ()</code> them
437for example).</p>
205 </dd> 438 </dd>
206 <dt>ev_loop_destroy (loop)</dt> 439 <dt>ev_loop_destroy (loop)</dt>
207 <dd> 440 <dd>
208 <p>Like <code>ev_default_destroy</code>, but destroys an event loop created by an 441 <p>Like <code>ev_default_destroy</code>, but destroys an event loop created by an
209earlier call to <code>ev_loop_new</code>.</p> 442earlier call to <code>ev_loop_new</code>.</p>
212 <dd> 445 <dd>
213 <p>This function reinitialises the kernel state for backends that have 446 <p>This function reinitialises the kernel state for backends that have
214one. Despite the name, you can call it anytime, but it makes most sense 447one. Despite the name, you can call it anytime, but it makes most sense
215after forking, in either the parent or child process (or both, but that 448after forking, in either the parent or child process (or both, but that
216again makes little sense).</p> 449again makes little sense).</p>
217 <p>You <i>must</i> call this function after forking if and only if you want to 450 <p>You <i>must</i> call this function in the child process after forking if and
218use the event library in both processes. If you just fork+exec, you don't 451only if you want to use the event library in both processes. If you just
219have to call it.</p> 452fork+exec, you don't have to call it.</p>
220 <p>The function itself is quite fast and its usually not a problem to call 453 <p>The function itself is quite fast and it's usually not a problem to call
221it just in case after a fork. To make this easy, the function will fit in 454it just in case after a fork. To make this easy, the function will fit in
222quite nicely into a call to <code>pthread_atfork</code>:</p> 455quite nicely into a call to <code>pthread_atfork</code>:</p>
223<pre> pthread_atfork (0, 0, ev_default_fork); 456<pre> pthread_atfork (0, 0, ev_default_fork);
224 457
225</pre> 458</pre>
459 <p>At the moment, <code>EVBACKEND_SELECT</code> and <code>EVBACKEND_POLL</code> are safe to use
460without calling this function, so if you force one of those backends you
461do not need to care.</p>
226 </dd> 462 </dd>
227 <dt>ev_loop_fork (loop)</dt> 463 <dt>ev_loop_fork (loop)</dt>
228 <dd> 464 <dd>
229 <p>Like <code>ev_default_fork</code>, but acts on an event loop created by 465 <p>Like <code>ev_default_fork</code>, but acts on an event loop created by
230<code>ev_loop_new</code>. Yes, you have to call this on every allocated event loop 466<code>ev_loop_new</code>. Yes, you have to call this on every allocated event loop
231after fork, and how you do this is entirely your own problem.</p> 467after fork, and how you do this is entirely your own problem.</p>
232 </dd> 468 </dd>
233 <dt>unsigned int ev_method (loop)</dt> 469 <dt>unsigned int ev_backend (loop)</dt>
234 <dd> 470 <dd>
235 <p>Returns one of the <code>EVMETHOD_*</code> flags indicating the event backend in 471 <p>Returns one of the <code>EVBACKEND_*</code> flags indicating the event backend in
236use.</p> 472use.</p>
237 </dd> 473 </dd>
238 <dt>ev_tstamp = ev_now (loop)</dt> 474 <dt>ev_tstamp ev_now (loop)</dt>
239 <dd> 475 <dd>
240 <p>Returns the current &quot;event loop time&quot;, which is the time the event loop 476 <p>Returns the current &quot;event loop time&quot;, which is the time the event loop
241got events and started processing them. This timestamp does not change 477received events and started processing them. This timestamp does not
242as long as callbacks are being processed, and this is also the base time 478change as long as callbacks are being processed, and this is also the base
243used for relative timers. You can treat it as the timestamp of the event 479time used for relative timers. You can treat it as the timestamp of the
244occuring (or more correctly, the mainloop finding out about it).</p> 480event occuring (or more correctly, libev finding out about it).</p>
245 </dd> 481 </dd>
246 <dt>ev_loop (loop, int flags)</dt> 482 <dt>ev_loop (loop, int flags)</dt>
247 <dd> 483 <dd>
248 <p>Finally, this is it, the event handler. This function usually is called 484 <p>Finally, this is it, the event handler. This function usually is called
249after you initialised all your watchers and you want to start handling 485after you initialised all your watchers and you want to start handling
250events.</p> 486events.</p>
251 <p>If the flags argument is specified as 0, it will not return until either 487 <p>If the flags argument is specified as <code>0</code>, it will not return until
252no event watchers are active anymore or <code>ev_unloop</code> was called.</p> 488either no event watchers are active anymore or <code>ev_unloop</code> was called.</p>
489 <p>Please note that an explicit <code>ev_unloop</code> is usually better than
490relying on all watchers to be stopped when deciding when a program has
491finished (especially in interactive programs), but having a program that
492automatically loops as long as it has to and no longer by virtue of
493relying on its watchers stopping correctly is a thing of beauty.</p>
253 <p>A flags value of <code>EVLOOP_NONBLOCK</code> will look for new events, will handle 494 <p>A flags value of <code>EVLOOP_NONBLOCK</code> will look for new events, will handle
254those events and any outstanding ones, but will not block your process in 495those events and any outstanding ones, but will not block your process in
255case there are no events.</p> 496case there are no events and will return after one iteration of the loop.</p>
256 <p>A flags value of <code>EVLOOP_ONESHOT</code> will look for new events (waiting if 497 <p>A flags value of <code>EVLOOP_ONESHOT</code> will look for new events (waiting if
257neccessary) and will handle those and any outstanding ones. It will block 498neccessary) and will handle those and any outstanding ones. It will block
258your process until at least one new event arrives.</p> 499your process until at least one new event arrives, and will return after
259 <p>This flags value could be used to implement alternative looping 500one iteration of the loop. This is useful if you are waiting for some
260constructs, but the <code>prepare</code> and <code>check</code> watchers provide a better and 501external event in conjunction with something not expressible using other
261more generic mechanism.</p> 502libev watchers. However, a pair of <code>ev_prepare</code>/<code>ev_check</code> watchers is
503usually a better approach for this kind of thing.</p>
504 <p>Here are the gory details of what <code>ev_loop</code> does:</p>
505<pre> * If there are no active watchers (reference count is zero), return.
506 - Queue prepare watchers and then call all outstanding watchers.
507 - If we have been forked, recreate the kernel state.
508 - Update the kernel state with all outstanding changes.
509 - Update the &quot;event loop time&quot;.
510 - Calculate for how long to block.
511 - Block the process, waiting for any events.
512 - Queue all outstanding I/O (fd) events.
513 - Update the &quot;event loop time&quot; and do time jump handling.
514 - Queue all outstanding timers.
515 - Queue all outstanding periodics.
516 - If no events are pending now, queue all idle watchers.
517 - Queue all check watchers.
518 - Call all queued watchers in reverse order (i.e. check watchers first).
519 Signals and child watchers are implemented as I/O watchers, and will
520 be handled here by queueing them when their watcher gets executed.
521 - If ev_unloop has been called or EVLOOP_ONESHOT or EVLOOP_NONBLOCK
522 were used, return, otherwise continue with step *.
523
524</pre>
525 <p>Example: Queue some jobs and then loop until no events are outsanding
526anymore.</p>
527<pre> ... queue jobs here, make sure they register event watchers as long
528 ... as they still have work to do (even an idle watcher will do..)
529 ev_loop (my_loop, 0);
530 ... jobs done. yeah!
531
532</pre>
262 </dd> 533 </dd>
263 <dt>ev_unloop (loop, how)</dt> 534 <dt>ev_unloop (loop, how)</dt>
264 <dd> 535 <dd>
265 <p>Can be used to make a call to <code>ev_loop</code> return early. The <code>how</code> argument 536 <p>Can be used to make a call to <code>ev_loop</code> return early (but only after it
537has processed all outstanding events). The <code>how</code> argument must be either
266must be either <code>EVUNLOOP_ONCE</code>, which will make the innermost <code>ev_loop</code> 538<code>EVUNLOOP_ONE</code>, which will make the innermost <code>ev_loop</code> call return, or
267call return, or <code>EVUNLOOP_ALL</code>, which will make all nested <code>ev_loop</code> 539<code>EVUNLOOP_ALL</code>, which will make all nested <code>ev_loop</code> calls return.</p>
268calls return.</p>
269 </dd> 540 </dd>
270 <dt>ev_ref (loop)</dt> 541 <dt>ev_ref (loop)</dt>
271 <dt>ev_unref (loop)</dt> 542 <dt>ev_unref (loop)</dt>
272 <dd> 543 <dd>
273 <p>Ref/unref can be used to add or remove a refcount on the event loop: Every 544 <p>Ref/unref can be used to add or remove a reference count on the event
274watcher keeps one reference. If you have a long-runing watcher you never 545loop: Every watcher keeps one reference, and as long as the reference
275unregister that should not keep ev_loop from running, ev_unref() after 546count is nonzero, <code>ev_loop</code> will not return on its own. If you have
276starting, and ev_ref() before stopping it. Libev itself uses this for 547a watcher you never unregister that should not keep <code>ev_loop</code> from
277example for its internal signal pipe: It is not visible to you as a user 548returning, ev_unref() after starting, and ev_ref() before stopping it. For
278and should not keep <code>ev_loop</code> from exiting if the work is done. It is 549example, libev itself uses this for its internal signal pipe: It is not
279also an excellent way to do this for generic recurring timers or from 550visible to the libev user and should not keep <code>ev_loop</code> from exiting if
280within third-party libraries. Just remember to unref after start and ref 551no event watchers registered by it are active. It is also an excellent
281before stop.</p> 552way to do this for generic recurring timers or from within third-party
553libraries. Just remember to <i>unref after start</i> and <i>ref before stop</i>.</p>
554 <p>Example: Create a signal watcher, but keep it from keeping <code>ev_loop</code>
555running when nothing else is active.</p>
556<pre> struct ev_signal exitsig;
557 ev_signal_init (&amp;exitsig, sig_cb, SIGINT);
558 ev_signal_start (loop, &amp;exitsig);
559 evf_unref (loop);
560
561</pre>
562 <p>Example: For some weird reason, unregister the above signal handler again.</p>
563<pre> ev_ref (loop);
564 ev_signal_stop (loop, &amp;exitsig);
565
566</pre>
282 </dd> 567 </dd>
283</dl> 568</dl>
284 569
570
571
572
573
285</div> 574</div>
286<h1 id="ANATOMY_OF_A_WATCHER">ANATOMY OF A WATCHER</h1><p><a href="#TOP" class="toplink">Top</a></p> 575<h1 id="ANATOMY_OF_A_WATCHER">ANATOMY OF A WATCHER</h1>
287<div id="ANATOMY_OF_A_WATCHER_CONTENT"> 576<div id="ANATOMY_OF_A_WATCHER_CONTENT">
288<p>A watcher is a structure that you create and register to record your 577<p>A watcher is a structure that you create and register to record your
289interest in some event. For instance, if you want to wait for STDIN to 578interest in some event. For instance, if you want to wait for STDIN to
290become readable, you would create an ev_io watcher for that:</p> 579become readable, you would create an <code>ev_io</code> watcher for that:</p>
291<pre> static void my_cb (struct ev_loop *loop, struct ev_io *w, int revents) 580<pre> static void my_cb (struct ev_loop *loop, struct ev_io *w, int revents)
292 { 581 {
293 ev_io_stop (w); 582 ev_io_stop (w);
294 ev_unloop (loop, EVUNLOOP_ALL); 583 ev_unloop (loop, EVUNLOOP_ALL);
295 } 584 }
318with a watcher-specific start function (<code>ev_&lt;type&gt;_start (loop, watcher 607with a watcher-specific start function (<code>ev_&lt;type&gt;_start (loop, watcher
319*)</code>), and you can stop watching for events at any time by calling the 608*)</code>), and you can stop watching for events at any time by calling the
320corresponding stop function (<code>ev_&lt;type&gt;_stop (loop, watcher *)</code>.</p> 609corresponding stop function (<code>ev_&lt;type&gt;_stop (loop, watcher *)</code>.</p>
321<p>As long as your watcher is active (has been started but not stopped) you 610<p>As long as your watcher is active (has been started but not stopped) you
322must not touch the values stored in it. Most specifically you must never 611must not touch the values stored in it. Most specifically you must never
323reinitialise it or call its set method.</p> 612reinitialise it or call its <code>set</code> macro.</p>
324<p>You cna check whether an event is active by calling the <code>ev_is_active
325(watcher *)</code> macro. To see whether an event is outstanding (but the
326callback for it has not been called yet) you cna use the <code>ev_is_pending
327(watcher *)</code> macro.</p>
328<p>Each and every callback receives the event loop pointer as first, the 613<p>Each and every callback receives the event loop pointer as first, the
329registered watcher structure as second, and a bitset of received events as 614registered watcher structure as second, and a bitset of received events as
330third argument.</p> 615third argument.</p>
331<p>The rceeived events usually include a single bit per event type received 616<p>The received events usually include a single bit per event type received
332(you can receive multiple events at the same time). The possible bit masks 617(you can receive multiple events at the same time). The possible bit masks
333are:</p> 618are:</p>
334<dl> 619<dl>
335 <dt>EV_READ</dt> 620 <dt><code>EV_READ</code></dt>
336 <dt>EV_WRITE</dt> 621 <dt><code>EV_WRITE</code></dt>
337 <dd> 622 <dd>
338 <p>The file descriptor in the ev_io watcher has become readable and/or 623 <p>The file descriptor in the <code>ev_io</code> watcher has become readable and/or
339writable.</p> 624writable.</p>
340 </dd> 625 </dd>
341 <dt>EV_TIMEOUT</dt> 626 <dt><code>EV_TIMEOUT</code></dt>
342 <dd>
343 <p>The ev_timer watcher has timed out.</p>
344 </dd> 627 <dd>
345 <dt>EV_PERIODIC</dt> 628 <p>The <code>ev_timer</code> watcher has timed out.</p>
346 <dd> 629 </dd>
347 <p>The ev_periodic watcher has timed out.</p> 630 <dt><code>EV_PERIODIC</code></dt>
348 </dd> 631 <dd>
349 <dt>EV_SIGNAL</dt> 632 <p>The <code>ev_periodic</code> watcher has timed out.</p>
350 <dd> 633 </dd>
634 <dt><code>EV_SIGNAL</code></dt>
635 <dd>
351 <p>The signal specified in the ev_signal watcher has been received by a thread.</p> 636 <p>The signal specified in the <code>ev_signal</code> watcher has been received by a thread.</p>
352 </dd>
353 <dt>EV_CHILD</dt>
354 <dd> 637 </dd>
638 <dt><code>EV_CHILD</code></dt>
639 <dd>
355 <p>The pid specified in the ev_child watcher has received a status change.</p> 640 <p>The pid specified in the <code>ev_child</code> watcher has received a status change.</p>
356 </dd>
357 <dt>EV_IDLE</dt>
358 <dd> 641 </dd>
642 <dt><code>EV_STAT</code></dt>
643 <dd>
644 <p>The path specified in the <code>ev_stat</code> watcher changed its attributes somehow.</p>
645 </dd>
646 <dt><code>EV_IDLE</code></dt>
647 <dd>
359 <p>The ev_idle watcher has determined that you have nothing better to do.</p> 648 <p>The <code>ev_idle</code> watcher has determined that you have nothing better to do.</p>
360 </dd>
361 <dt>EV_PREPARE</dt>
362 <dt>EV_CHECK</dt>
363 <dd> 649 </dd>
650 <dt><code>EV_PREPARE</code></dt>
651 <dt><code>EV_CHECK</code></dt>
652 <dd>
364 <p>All ev_prepare watchers are invoked just <i>before</i> <code>ev_loop</code> starts 653 <p>All <code>ev_prepare</code> watchers are invoked just <i>before</i> <code>ev_loop</code> starts
365to gather new events, and all ev_check watchers are invoked just after 654to gather new events, and all <code>ev_check</code> watchers are invoked just after
366<code>ev_loop</code> has gathered them, but before it invokes any callbacks for any 655<code>ev_loop</code> has gathered them, but before it invokes any callbacks for any
367received events. Callbacks of both watcher types can start and stop as 656received events. Callbacks of both watcher types can start and stop as
368many watchers as they want, and all of them will be taken into account 657many watchers as they want, and all of them will be taken into account
369(for example, a ev_prepare watcher might start an idle watcher to keep 658(for example, a <code>ev_prepare</code> watcher might start an idle watcher to keep
370<code>ev_loop</code> from blocking).</p> 659<code>ev_loop</code> from blocking).</p>
371 </dd> 660 </dd>
372 <dt>EV_ERROR</dt> 661 <dt><code>EV_EMBED</code></dt>
662 <dd>
663 <p>The embedded event loop specified in the <code>ev_embed</code> watcher needs attention.</p>
664 </dd>
665 <dt><code>EV_FORK</code></dt>
666 <dd>
667 <p>The event loop has been resumed in the child process after fork (see
668<code>ev_fork</code>).</p>
669 </dd>
670 <dt><code>EV_ERROR</code></dt>
373 <dd> 671 <dd>
374 <p>An unspecified error has occured, the watcher has been stopped. This might 672 <p>An unspecified error has occured, the watcher has been stopped. This might
375happen because the watcher could not be properly started because libev 673happen because the watcher could not be properly started because libev
376ran out of memory, a file descriptor was found to be closed or any other 674ran out of memory, a file descriptor was found to be closed or any other
377problem. You best act on it by reporting the problem and somehow coping 675problem. You best act on it by reporting the problem and somehow coping
383programs, though, so beware.</p> 681programs, though, so beware.</p>
384 </dd> 682 </dd>
385</dl> 683</dl>
386 684
387</div> 685</div>
686<h2 id="GENERIC_WATCHER_FUNCTIONS">GENERIC WATCHER FUNCTIONS</h2>
687<div id="GENERIC_WATCHER_FUNCTIONS_CONTENT">
688<p>In the following description, <code>TYPE</code> stands for the watcher type,
689e.g. <code>timer</code> for <code>ev_timer</code> watchers and <code>io</code> for <code>ev_io</code> watchers.</p>
690<dl>
691 <dt><code>ev_init</code> (ev_TYPE *watcher, callback)</dt>
692 <dd>
693 <p>This macro initialises the generic portion of a watcher. The contents
694of the watcher object can be arbitrary (so <code>malloc</code> will do). Only
695the generic parts of the watcher are initialised, you <i>need</i> to call
696the type-specific <code>ev_TYPE_set</code> macro afterwards to initialise the
697type-specific parts. For each type there is also a <code>ev_TYPE_init</code> macro
698which rolls both calls into one.</p>
699 <p>You can reinitialise a watcher at any time as long as it has been stopped
700(or never started) and there are no pending events outstanding.</p>
701 <p>The callback is always of type <code>void (*)(ev_loop *loop, ev_TYPE *watcher,
702int revents)</code>.</p>
703 </dd>
704 <dt><code>ev_TYPE_set</code> (ev_TYPE *, [args])</dt>
705 <dd>
706 <p>This macro initialises the type-specific parts of a watcher. You need to
707call <code>ev_init</code> at least once before you call this macro, but you can
708call <code>ev_TYPE_set</code> any number of times. You must not, however, call this
709macro on a watcher that is active (it can be pending, however, which is a
710difference to the <code>ev_init</code> macro).</p>
711 <p>Although some watcher types do not have type-specific arguments
712(e.g. <code>ev_prepare</code>) you still need to call its <code>set</code> macro.</p>
713 </dd>
714 <dt><code>ev_TYPE_init</code> (ev_TYPE *watcher, callback, [args])</dt>
715 <dd>
716 <p>This convinience macro rolls both <code>ev_init</code> and <code>ev_TYPE_set</code> macro
717calls into a single call. This is the most convinient method to initialise
718a watcher. The same limitations apply, of course.</p>
719 </dd>
720 <dt><code>ev_TYPE_start</code> (loop *, ev_TYPE *watcher)</dt>
721 <dd>
722 <p>Starts (activates) the given watcher. Only active watchers will receive
723events. If the watcher is already active nothing will happen.</p>
724 </dd>
725 <dt><code>ev_TYPE_stop</code> (loop *, ev_TYPE *watcher)</dt>
726 <dd>
727 <p>Stops the given watcher again (if active) and clears the pending
728status. It is possible that stopped watchers are pending (for example,
729non-repeating timers are being stopped when they become pending), but
730<code>ev_TYPE_stop</code> ensures that the watcher is neither active nor pending. If
731you want to free or reuse the memory used by the watcher it is therefore a
732good idea to always call its <code>ev_TYPE_stop</code> function.</p>
733 </dd>
734 <dt>bool ev_is_active (ev_TYPE *watcher)</dt>
735 <dd>
736 <p>Returns a true value iff the watcher is active (i.e. it has been started
737and not yet been stopped). As long as a watcher is active you must not modify
738it.</p>
739 </dd>
740 <dt>bool ev_is_pending (ev_TYPE *watcher)</dt>
741 <dd>
742 <p>Returns a true value iff the watcher is pending, (i.e. it has outstanding
743events but its callback has not yet been invoked). As long as a watcher
744is pending (but not active) you must not call an init function on it (but
745<code>ev_TYPE_set</code> is safe) and you must make sure the watcher is available to
746libev (e.g. you cnanot <code>free ()</code> it).</p>
747 </dd>
748 <dt>callback ev_cb (ev_TYPE *watcher)</dt>
749 <dd>
750 <p>Returns the callback currently set on the watcher.</p>
751 </dd>
752 <dt>ev_cb_set (ev_TYPE *watcher, callback)</dt>
753 <dd>
754 <p>Change the callback. You can change the callback at virtually any time
755(modulo threads).</p>
756 </dd>
757</dl>
758
759
760
761
762
763</div>
388<h2 id="ASSOCIATING_CUSTOM_DATA_WITH_A_WATCH">ASSOCIATING CUSTOM DATA WITH A WATCHER</h2> 764<h2 id="ASSOCIATING_CUSTOM_DATA_WITH_A_WATCH">ASSOCIATING CUSTOM DATA WITH A WATCHER</h2>
389<div id="ASSOCIATING_CUSTOM_DATA_WITH_A_WATCH-2"> 765<div id="ASSOCIATING_CUSTOM_DATA_WITH_A_WATCH-2">
390<p>Each watcher has, by default, a member <code>void *data</code> that you can change 766<p>Each watcher has, by default, a member <code>void *data</code> that you can change
391and read at any time, libev will completely ignore it. This cna be used 767and read at any time, libev will completely ignore it. This can be used
392to associate arbitrary data with your watcher. If you need more data and 768to associate arbitrary data with your watcher. If you need more data and
393don't want to allocate memory and store a pointer to it in that data 769don't want to allocate memory and store a pointer to it in that data
394member, you can also &quot;subclass&quot; the watcher type and provide your own 770member, you can also &quot;subclass&quot; the watcher type and provide your own
395data:</p> 771data:</p>
396<pre> struct my_io 772<pre> struct my_io
409 struct my_io *w = (struct my_io *)w_; 785 struct my_io *w = (struct my_io *)w_;
410 ... 786 ...
411 } 787 }
412 788
413</pre> 789</pre>
414<p>More interesting and less C-conformant ways of catsing your callback type 790<p>More interesting and less C-conformant ways of casting your callback type
415have been omitted....</p> 791instead have been omitted.</p>
792<p>Another common scenario is having some data structure with multiple
793watchers:</p>
794<pre> struct my_biggy
795 {
796 int some_data;
797 ev_timer t1;
798 ev_timer t2;
799 }
416 800
801</pre>
802<p>In this case getting the pointer to <code>my_biggy</code> is a bit more complicated,
803you need to use <code>offsetof</code>:</p>
804<pre> #include &lt;stddef.h&gt;
417 805
806 static void
807 t1_cb (EV_P_ struct ev_timer *w, int revents)
808 {
809 struct my_biggy big = (struct my_biggy *
810 (((char *)w) - offsetof (struct my_biggy, t1));
811 }
418 812
813 static void
814 t2_cb (EV_P_ struct ev_timer *w, int revents)
815 {
816 struct my_biggy big = (struct my_biggy *
817 (((char *)w) - offsetof (struct my_biggy, t2));
818 }
419 819
420 820
821
822
823</pre>
824
421</div> 825</div>
422<h1 id="WATCHER_TYPES">WATCHER TYPES</h1><p><a href="#TOP" class="toplink">Top</a></p> 826<h1 id="WATCHER_TYPES">WATCHER TYPES</h1>
423<div id="WATCHER_TYPES_CONTENT"> 827<div id="WATCHER_TYPES_CONTENT">
424<p>This section describes each watcher in detail, but will not repeat 828<p>This section describes each watcher in detail, but will not repeat
425information given in the last section.</p> 829information given in the last section. Any initialisation/set macros,
830functions and members specific to the watcher type are explained.</p>
831<p>Members are additionally marked with either <i>[read-only]</i>, meaning that,
832while the watcher is active, you can look at the member and expect some
833sensible content, but you must not modify it (you can modify it while the
834watcher is stopped to your hearts content), or <i>[read-write]</i>, which
835means you can expect it to have some sensible content while the watcher
836is active, but you can also modify it. Modifying it may not do something
837sensible or take immediate effect (or do anything at all), but libev will
838not crash or malfunction in any way.</p>
426 839
840
841
842
843
427</div> 844</div>
428<h2 id="struct_ev_io_is_my_file_descriptor_r">struct ev_io - is my file descriptor readable or writable</h2> 845<h2 id="code_ev_io_code_is_this_file_descrip"><code>ev_io</code> - is this file descriptor readable or writable?</h2>
429<div id="struct_ev_io_is_my_file_descriptor_r-2"> 846<div id="code_ev_io_code_is_this_file_descrip-2">
430<p>I/O watchers check whether a file descriptor is readable or writable 847<p>I/O watchers check whether a file descriptor is readable or writable
431in each iteration of the event loop (This behaviour is called 848in each iteration of the event loop, or, more precisely, when reading
432level-triggering because you keep receiving events as long as the 849would not block the process and writing would at least be able to write
433condition persists. Remember you cna stop the watcher if you don't want to 850some data. This behaviour is called level-triggering because you keep
434act on the event and neither want to receive future events).</p> 851receiving events as long as the condition persists. Remember you can stop
852the watcher if you don't want to act on the event and neither want to
853receive future events.</p>
854<p>In general you can register as many read and/or write event watchers per
855fd as you want (as long as you don't confuse yourself). Setting all file
856descriptors to non-blocking mode is also usually a good idea (but not
857required if you know what you are doing).</p>
858<p>You have to be careful with dup'ed file descriptors, though. Some backends
859(the linux epoll backend is a notable example) cannot handle dup'ed file
860descriptors correctly if you register interest in two or more fds pointing
861to the same underlying file/socket/etc. description (that is, they share
862the same underlying &quot;file open&quot;).</p>
863<p>If you must do this, then force the use of a known-to-be-good backend
864(at the time of this writing, this includes only <code>EVBACKEND_SELECT</code> and
865<code>EVBACKEND_POLL</code>).</p>
866<p>Another thing you have to watch out for is that it is quite easy to
867receive &quot;spurious&quot; readyness notifications, that is your callback might
868be called with <code>EV_READ</code> but a subsequent <code>read</code>(2) will actually block
869because there is no data. Not only are some backends known to create a
870lot of those (for example solaris ports), it is very easy to get into
871this situation even with a relatively standard program structure. Thus
872it is best to always use non-blocking I/O: An extra <code>read</code>(2) returning
873<code>EAGAIN</code> is far preferable to a program hanging until some data arrives.</p>
874<p>If you cannot run the fd in non-blocking mode (for example you should not
875play around with an Xlib connection), then you have to seperately re-test
876wether a file descriptor is really ready with a known-to-be good interface
877such as poll (fortunately in our Xlib example, Xlib already does this on
878its own, so its quite safe to use).</p>
435<dl> 879<dl>
436 <dt>ev_io_init (ev_io *, callback, int fd, int events)</dt> 880 <dt>ev_io_init (ev_io *, callback, int fd, int events)</dt>
437 <dt>ev_io_set (ev_io *, int fd, int events)</dt> 881 <dt>ev_io_set (ev_io *, int fd, int events)</dt>
438 <dd> 882 <dd>
439 <p>Configures an ev_io watcher. The fd is the file descriptor to rceeive 883 <p>Configures an <code>ev_io</code> watcher. The <code>fd</code> is the file descriptor to
440events for and events is either <code>EV_READ</code>, <code>EV_WRITE</code> or <code>EV_READ | 884rceeive events for and events is either <code>EV_READ</code>, <code>EV_WRITE</code> or
441EV_WRITE</code> to receive the given events.</p> 885<code>EV_READ | EV_WRITE</code> to receive the given events.</p>
886 </dd>
887 <dt>int fd [read-only]</dt>
888 <dd>
889 <p>The file descriptor being watched.</p>
890 </dd>
891 <dt>int events [read-only]</dt>
892 <dd>
893 <p>The events being watched.</p>
442 </dd> 894 </dd>
443</dl> 895</dl>
896<p>Example: Call <code>stdin_readable_cb</code> when STDIN_FILENO has become, well
897readable, but only once. Since it is likely line-buffered, you could
898attempt to read a whole line in the callback.</p>
899<pre> static void
900 stdin_readable_cb (struct ev_loop *loop, struct ev_io *w, int revents)
901 {
902 ev_io_stop (loop, w);
903 .. read from stdin here (or from w-&gt;fd) and haqndle any I/O errors
904 }
444 905
906 ...
907 struct ev_loop *loop = ev_default_init (0);
908 struct ev_io stdin_readable;
909 ev_io_init (&amp;stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);
910 ev_io_start (loop, &amp;stdin_readable);
911 ev_loop (loop, 0);
912
913
914
915
916</pre>
917
445</div> 918</div>
446<h2 id="struct_ev_timer_relative_and_optiona">struct ev_timer - relative and optionally recurring timeouts</h2> 919<h2 id="code_ev_timer_code_relative_and_opti"><code>ev_timer</code> - relative and optionally repeating timeouts</h2>
447<div id="struct_ev_timer_relative_and_optiona-2"> 920<div id="code_ev_timer_code_relative_and_opti-2">
448<p>Timer watchers are simple relative timers that generate an event after a 921<p>Timer watchers are simple relative timers that generate an event after a
449given time, and optionally repeating in regular intervals after that.</p> 922given time, and optionally repeating in regular intervals after that.</p>
450<p>The timers are based on real time, that is, if you register an event that 923<p>The timers are based on real time, that is, if you register an event that
451times out after an hour and youreset your system clock to last years 924times out after an hour and you reset your system clock to last years
452time, it will still time out after (roughly) and hour. &quot;Roughly&quot; because 925time, it will still time out after (roughly) and hour. &quot;Roughly&quot; because
453detecting time jumps is hard, and soem inaccuracies are unavoidable (the 926detecting time jumps is hard, and some inaccuracies are unavoidable (the
454monotonic clock option helps a lot here).</p> 927monotonic clock option helps a lot here).</p>
928<p>The relative timeouts are calculated relative to the <code>ev_now ()</code>
929time. This is usually the right thing as this timestamp refers to the time
930of the event triggering whatever timeout you are modifying/starting. If
931you suspect event processing to be delayed and you <i>need</i> to base the timeout
932on the current time, use something like this to adjust for this:</p>
933<pre> ev_timer_set (&amp;timer, after + ev_now () - ev_time (), 0.);
934
935</pre>
936<p>The callback is guarenteed to be invoked only when its timeout has passed,
937but if multiple timers become ready during the same loop iteration then
938order of execution is undefined.</p>
455<dl> 939<dl>
456 <dt>ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)</dt> 940 <dt>ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)</dt>
457 <dt>ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)</dt> 941 <dt>ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)</dt>
458 <dd> 942 <dd>
459 <p>Configure the timer to trigger after <code>after</code> seconds. If <code>repeat</code> is 943 <p>Configure the timer to trigger after <code>after</code> seconds. If <code>repeat</code> is
461timer will automatically be configured to trigger again <code>repeat</code> seconds 945timer will automatically be configured to trigger again <code>repeat</code> seconds
462later, again, and again, until stopped manually.</p> 946later, again, and again, until stopped manually.</p>
463 <p>The timer itself will do a best-effort at avoiding drift, that is, if you 947 <p>The timer itself will do a best-effort at avoiding drift, that is, if you
464configure a timer to trigger every 10 seconds, then it will trigger at 948configure a timer to trigger every 10 seconds, then it will trigger at
465exactly 10 second intervals. If, however, your program cannot keep up with 949exactly 10 second intervals. If, however, your program cannot keep up with
466the timer (ecause it takes longer than those 10 seconds to do stuff) the 950the timer (because it takes longer than those 10 seconds to do stuff) the
467timer will not fire more than once per event loop iteration.</p> 951timer will not fire more than once per event loop iteration.</p>
468 </dd> 952 </dd>
469 <dt>ev_timer_again (loop)</dt> 953 <dt>ev_timer_again (loop)</dt>
470 <dd> 954 <dd>
471 <p>This will act as if the timer timed out and restart it again if it is 955 <p>This will act as if the timer timed out and restart it again if it is
472repeating. The exact semantics are:</p> 956repeating. The exact semantics are:</p>
473 <p>If the timer is started but nonrepeating, stop it.</p> 957 <p>If the timer is started but nonrepeating, stop it.</p>
474 <p>If the timer is repeating, either start it if necessary (with the repeat 958 <p>If the timer is repeating, either start it if necessary (with the repeat
475value), or reset the running timer to the repeat value.</p> 959value), or reset the running timer to the repeat value.</p>
476 <p>This sounds a bit complicated, but here is a useful and typical 960 <p>This sounds a bit complicated, but here is a useful and typical
477example: Imagine you have a tcp connection and you want a so-called idle 961example: Imagine you have a tcp connection and you want a so-called
478timeout, that is, you want to be called when there have been, say, 60 962idle timeout, that is, you want to be called when there have been,
479seconds of inactivity on the socket. The easiest way to do this is to 963say, 60 seconds of inactivity on the socket. The easiest way to do
480configure an ev_timer with after=repeat=60 and calling ev_timer_again each 964this is to configure an <code>ev_timer</code> with <code>after</code>=<code>repeat</code>=<code>60</code> and calling
481time you successfully read or write some data. If you go into an idle 965<code>ev_timer_again</code> each time you successfully read or write some data. If
482state where you do not expect data to travel on the socket, you can stop 966you go into an idle state where you do not expect data to travel on the
483the timer, and again will automatically restart it if need be.</p> 967socket, you can stop the timer, and again will automatically restart it if
968need be.</p>
969 <p>You can also ignore the <code>after</code> value and <code>ev_timer_start</code> altogether
970and only ever use the <code>repeat</code> value:</p>
971<pre> ev_timer_init (timer, callback, 0., 5.);
972 ev_timer_again (loop, timer);
973 ...
974 timer-&gt;again = 17.;
975 ev_timer_again (loop, timer);
976 ...
977 timer-&gt;again = 10.;
978 ev_timer_again (loop, timer);
979
980</pre>
981 <p>This is more efficient then stopping/starting the timer eahc time you want
982to modify its timeout value.</p>
983 </dd>
984 <dt>ev_tstamp repeat [read-write]</dt>
985 <dd>
986 <p>The current <code>repeat</code> value. Will be used each time the watcher times out
987or <code>ev_timer_again</code> is called and determines the next timeout (if any),
988which is also when any modifications are taken into account.</p>
484 </dd> 989 </dd>
485</dl> 990</dl>
991<p>Example: Create a timer that fires after 60 seconds.</p>
992<pre> static void
993 one_minute_cb (struct ev_loop *loop, struct ev_timer *w, int revents)
994 {
995 .. one minute over, w is actually stopped right here
996 }
486 997
998 struct ev_timer mytimer;
999 ev_timer_init (&amp;mytimer, one_minute_cb, 60., 0.);
1000 ev_timer_start (loop, &amp;mytimer);
1001
1002</pre>
1003<p>Example: Create a timeout timer that times out after 10 seconds of
1004inactivity.</p>
1005<pre> static void
1006 timeout_cb (struct ev_loop *loop, struct ev_timer *w, int revents)
1007 {
1008 .. ten seconds without any activity
1009 }
1010
1011 struct ev_timer mytimer;
1012 ev_timer_init (&amp;mytimer, timeout_cb, 0., 10.); /* note, only repeat used */
1013 ev_timer_again (&amp;mytimer); /* start timer */
1014 ev_loop (loop, 0);
1015
1016 // and in some piece of code that gets executed on any &quot;activity&quot;:
1017 // reset the timeout to start ticking again at 10 seconds
1018 ev_timer_again (&amp;mytimer);
1019
1020
1021
1022
1023</pre>
1024
487</div> 1025</div>
488<h2 id="ev_periodic_to_cron_or_not_to_cron_i">ev_periodic - to cron or not to cron it</h2> 1026<h2 id="code_ev_periodic_code_to_cron_or_not"><code>ev_periodic</code> - to cron or not to cron?</h2>
489<div id="ev_periodic_to_cron_or_not_to_cron_i-2"> 1027<div id="code_ev_periodic_code_to_cron_or_not-2">
490<p>Periodic watchers are also timers of a kind, but they are very versatile 1028<p>Periodic watchers are also timers of a kind, but they are very versatile
491(and unfortunately a bit complex).</p> 1029(and unfortunately a bit complex).</p>
492<p>Unlike ev_timer's, they are not based on real time (or relative time) 1030<p>Unlike <code>ev_timer</code>'s, they are not based on real time (or relative time)
493but on wallclock time (absolute time). You can tell a periodic watcher 1031but on wallclock time (absolute time). You can tell a periodic watcher
494to trigger &quot;at&quot; some specific point in time. For example, if you tell a 1032to trigger &quot;at&quot; some specific point in time. For example, if you tell a
495periodic watcher to trigger in 10 seconds (by specifiying e.g. c&lt;ev_now () 1033periodic watcher to trigger in 10 seconds (by specifiying e.g. <code>ev_now ()
496+ 10.&gt;) and then reset your system clock to the last year, then it will 1034+ 10.</code>) and then reset your system clock to the last year, then it will
497take a year to trigger the event (unlike an ev_timer, which would trigger 1035take a year to trigger the event (unlike an <code>ev_timer</code>, which would trigger
498roughly 10 seconds later and of course not if you reset your system time 1036roughly 10 seconds later and of course not if you reset your system time
499again).</p> 1037again).</p>
500<p>They can also be used to implement vastly more complex timers, such as 1038<p>They can also be used to implement vastly more complex timers, such as
501triggering an event on eahc midnight, local time.</p> 1039triggering an event on eahc midnight, local time.</p>
1040<p>As with timers, the callback is guarenteed to be invoked only when the
1041time (<code>at</code>) has been passed, but if multiple periodic timers become ready
1042during the same loop iteration then order of execution is undefined.</p>
502<dl> 1043<dl>
503 <dt>ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)</dt> 1044 <dt>ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)</dt>
504 <dt>ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)</dt> 1045 <dt>ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)</dt>
505 <dd> 1046 <dd>
506 <p>Lots of arguments, lets sort it out... There are basically three modes of 1047 <p>Lots of arguments, lets sort it out... There are basically three modes of
507operation, and we will explain them from simplest to complex:</p> 1048operation, and we will explain them from simplest to complex:</p>
508
509
510
511
512 <p> 1049 <p>
513 <dl> 1050 <dl>
514 <dt>* absolute timer (interval = reschedule_cb = 0)</dt> 1051 <dt>* absolute timer (interval = reschedule_cb = 0)</dt>
515 <dd> 1052 <dd>
516 <p>In this configuration the watcher triggers an event at the wallclock time 1053 <p>In this configuration the watcher triggers an event at the wallclock time
528<pre> ev_periodic_set (&amp;periodic, 0., 3600., 0); 1065<pre> ev_periodic_set (&amp;periodic, 0., 3600., 0);
529 1066
530</pre> 1067</pre>
531 <p>This doesn't mean there will always be 3600 seconds in between triggers, 1068 <p>This doesn't mean there will always be 3600 seconds in between triggers,
532but only that the the callback will be called when the system time shows a 1069but only that the the callback will be called when the system time shows a
533full hour (UTC), or more correct, when the system time is evenly divisible 1070full hour (UTC), or more correctly, when the system time is evenly divisible
534by 3600.</p> 1071by 3600.</p>
535 <p>Another way to think about it (for the mathematically inclined) is that 1072 <p>Another way to think about it (for the mathematically inclined) is that
536ev_periodic will try to run the callback in this mode at the next possible 1073<code>ev_periodic</code> will try to run the callback in this mode at the next possible
537time where <code>time = at (mod interval)</code>, regardless of any time jumps.</p> 1074time where <code>time = at (mod interval)</code>, regardless of any time jumps.</p>
538 </dd> 1075 </dd>
539 <dt>* manual reschedule mode (reschedule_cb = callback)</dt> 1076 <dt>* manual reschedule mode (reschedule_cb = callback)</dt>
540 <dd> 1077 <dd>
541 <p>In this mode the values for <code>interval</code> and <code>at</code> are both being 1078 <p>In this mode the values for <code>interval</code> and <code>at</code> are both being
542ignored. Instead, each time the periodic watcher gets scheduled, the 1079ignored. Instead, each time the periodic watcher gets scheduled, the
543reschedule callback will be called with the watcher as first, and the 1080reschedule callback will be called with the watcher as first, and the
544current time as second argument.</p> 1081current time as second argument.</p>
545 <p>NOTE: <i>This callback MUST NOT stop or destroy the periodic or any other 1082 <p>NOTE: <i>This callback MUST NOT stop or destroy any periodic watcher,
546periodic watcher, ever, or make any event loop modificstions</i>. If you need 1083ever, or make any event loop modifications</i>. If you need to stop it,
547to stop it, return 1e30 (or so, fudge fudge) and stop it afterwards.</p> 1084return <code>now + 1e30</code> (or so, fudge fudge) and stop it afterwards (e.g. by
1085starting a prepare watcher).</p>
548 <p>Its prototype is c&lt;ev_tstamp (*reschedule_cb)(struct ev_periodic *w, 1086 <p>Its prototype is <code>ev_tstamp (*reschedule_cb)(struct ev_periodic *w,
549ev_tstamp now)&gt;, e.g.:</p> 1087ev_tstamp now)</code>, e.g.:</p>
550<pre> static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now) 1088<pre> static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now)
551 { 1089 {
552 return now + 60.; 1090 return now + 60.;
553 } 1091 }
554 1092
555</pre> 1093</pre>
556 <p>It must return the next time to trigger, based on the passed time value 1094 <p>It must return the next time to trigger, based on the passed time value
557(that is, the lowest time value larger than to the second argument). It 1095(that is, the lowest time value larger than to the second argument). It
558will usually be called just before the callback will be triggered, but 1096will usually be called just before the callback will be triggered, but
559might be called at other times, too.</p> 1097might be called at other times, too.</p>
1098 <p>NOTE: <i>This callback must always return a time that is later than the
1099passed <code>now</code> value</i>. Not even <code>now</code> itself will do, it <i>must</i> be larger.</p>
560 <p>This can be used to create very complex timers, such as a timer that 1100 <p>This can be used to create very complex timers, such as a timer that
561triggers on each midnight, local time. To do this, you would calculate the 1101triggers on each midnight, local time. To do this, you would calculate the
562next midnight after <code>now</code> and return the timestamp value for this. How you do this 1102next midnight after <code>now</code> and return the timestamp value for this. How
563is, again, up to you (but it is not trivial).</p> 1103you do this is, again, up to you (but it is not trivial, which is the main
1104reason I omitted it as an example).</p>
564 </dd> 1105 </dd>
565 </dl> 1106 </dl>
566 </p> 1107 </p>
567 </dd> 1108 </dd>
568 <dt>ev_periodic_again (loop, ev_periodic *)</dt> 1109 <dt>ev_periodic_again (loop, ev_periodic *)</dt>
570 <p>Simply stops and restarts the periodic watcher again. This is only useful 1111 <p>Simply stops and restarts the periodic watcher again. This is only useful
571when you changed some parameters or the reschedule callback would return 1112when you changed some parameters or the reschedule callback would return
572a different time than the last time it was called (e.g. in a crond like 1113a different time than the last time it was called (e.g. in a crond like
573program when the crontabs have changed).</p> 1114program when the crontabs have changed).</p>
574 </dd> 1115 </dd>
1116 <dt>ev_tstamp interval [read-write]</dt>
1117 <dd>
1118 <p>The current interval value. Can be modified any time, but changes only
1119take effect when the periodic timer fires or <code>ev_periodic_again</code> is being
1120called.</p>
1121 </dd>
1122 <dt>ev_tstamp (*reschedule_cb)(struct ev_periodic *w, ev_tstamp now) [read-write]</dt>
1123 <dd>
1124 <p>The current reschedule callback, or <code>0</code>, if this functionality is
1125switched off. Can be changed any time, but changes only take effect when
1126the periodic timer fires or <code>ev_periodic_again</code> is being called.</p>
1127 </dd>
575</dl> 1128</dl>
1129<p>Example: Call a callback every hour, or, more precisely, whenever the
1130system clock is divisible by 3600. The callback invocation times have
1131potentially a lot of jittering, but good long-term stability.</p>
1132<pre> static void
1133 clock_cb (struct ev_loop *loop, struct ev_io *w, int revents)
1134 {
1135 ... its now a full hour (UTC, or TAI or whatever your clock follows)
1136 }
576 1137
1138 struct ev_periodic hourly_tick;
1139 ev_periodic_init (&amp;hourly_tick, clock_cb, 0., 3600., 0);
1140 ev_periodic_start (loop, &amp;hourly_tick);
1141
1142</pre>
1143<p>Example: The same as above, but use a reschedule callback to do it:</p>
1144<pre> #include &lt;math.h&gt;
1145
1146 static ev_tstamp
1147 my_scheduler_cb (struct ev_periodic *w, ev_tstamp now)
1148 {
1149 return fmod (now, 3600.) + 3600.;
1150 }
1151
1152 ev_periodic_init (&amp;hourly_tick, clock_cb, 0., 0., my_scheduler_cb);
1153
1154</pre>
1155<p>Example: Call a callback every hour, starting now:</p>
1156<pre> struct ev_periodic hourly_tick;
1157 ev_periodic_init (&amp;hourly_tick, clock_cb,
1158 fmod (ev_now (loop), 3600.), 3600., 0);
1159 ev_periodic_start (loop, &amp;hourly_tick);
1160
1161
1162
1163
1164</pre>
1165
577</div> 1166</div>
578<h2 id="ev_signal_signal_me_when_a_signal_ge">ev_signal - signal me when a signal gets signalled</h2> 1167<h2 id="code_ev_signal_code_signal_me_when_a"><code>ev_signal</code> - signal me when a signal gets signalled!</h2>
579<div id="ev_signal_signal_me_when_a_signal_ge-2"> 1168<div id="code_ev_signal_code_signal_me_when_a-2">
580<p>Signal watchers will trigger an event when the process receives a specific 1169<p>Signal watchers will trigger an event when the process receives a specific
581signal one or more times. Even though signals are very asynchronous, libev 1170signal one or more times. Even though signals are very asynchronous, libev
582will try its best to deliver signals synchronously, i.e. as part of the 1171will try it's best to deliver signals synchronously, i.e. as part of the
583normal event processing, like any other event.</p> 1172normal event processing, like any other event.</p>
584<p>You cna configure as many watchers as you like per signal. Only when the 1173<p>You can configure as many watchers as you like per signal. Only when the
585first watcher gets started will libev actually register a signal watcher 1174first watcher gets started will libev actually register a signal watcher
586with the kernel (thus it coexists with your own signal handlers as long 1175with the kernel (thus it coexists with your own signal handlers as long
587as you don't register any with libev). Similarly, when the last signal 1176as you don't register any with libev). Similarly, when the last signal
588watcher for a signal is stopped libev will reset the signal handler to 1177watcher for a signal is stopped libev will reset the signal handler to
589SIG_DFL (regardless of what it was set to before).</p> 1178SIG_DFL (regardless of what it was set to before).</p>
592 <dt>ev_signal_set (ev_signal *, int signum)</dt> 1181 <dt>ev_signal_set (ev_signal *, int signum)</dt>
593 <dd> 1182 <dd>
594 <p>Configures the watcher to trigger on the given signal number (usually one 1183 <p>Configures the watcher to trigger on the given signal number (usually one
595of the <code>SIGxxx</code> constants).</p> 1184of the <code>SIGxxx</code> constants).</p>
596 </dd> 1185 </dd>
1186 <dt>int signum [read-only]</dt>
1187 <dd>
1188 <p>The signal the watcher watches out for.</p>
1189 </dd>
597</dl> 1190</dl>
598 1191
1192
1193
1194
1195
599</div> 1196</div>
600<h2 id="ev_child_wait_for_pid_status_changes">ev_child - wait for pid status changes</h2> 1197<h2 id="code_ev_child_code_watch_out_for_pro"><code>ev_child</code> - watch out for process status changes</h2>
601<div id="ev_child_wait_for_pid_status_changes-2"> 1198<div id="code_ev_child_code_watch_out_for_pro-2">
602<p>Child watchers trigger when your process receives a SIGCHLD in response to 1199<p>Child watchers trigger when your process receives a SIGCHLD in response to
603some child status changes (most typically when a child of yours dies).</p> 1200some child status changes (most typically when a child of yours dies).</p>
604<dl> 1201<dl>
605 <dt>ev_child_init (ev_child *, callback, int pid)</dt> 1202 <dt>ev_child_init (ev_child *, callback, int pid)</dt>
606 <dt>ev_child_set (ev_child *, int pid)</dt> 1203 <dt>ev_child_set (ev_child *, int pid)</dt>
607 <dd> 1204 <dd>
608 <p>Configures the watcher to wait for status changes of process <code>pid</code> (or 1205 <p>Configures the watcher to wait for status changes of process <code>pid</code> (or
609<i>any</i> process if <code>pid</code> is specified as <code>0</code>). The callback can look 1206<i>any</i> process if <code>pid</code> is specified as <code>0</code>). The callback can look
610at the <code>rstatus</code> member of the <code>ev_child</code> watcher structure to see 1207at the <code>rstatus</code> member of the <code>ev_child</code> watcher structure to see
611the status word (use the macros from <code>sys/wait.h</code>). The <code>rpid</code> member 1208the status word (use the macros from <code>sys/wait.h</code> and see your systems
1209<code>waitpid</code> documentation). The <code>rpid</code> member contains the pid of the
612contains the pid of the process causing the status change.</p> 1210process causing the status change.</p>
1211 </dd>
1212 <dt>int pid [read-only]</dt>
1213 <dd>
1214 <p>The process id this watcher watches out for, or <code>0</code>, meaning any process id.</p>
1215 </dd>
1216 <dt>int rpid [read-write]</dt>
1217 <dd>
1218 <p>The process id that detected a status change.</p>
1219 </dd>
1220 <dt>int rstatus [read-write]</dt>
1221 <dd>
1222 <p>The process exit/trace status caused by <code>rpid</code> (see your systems
1223<code>waitpid</code> and <code>sys/wait.h</code> documentation for details).</p>
613 </dd> 1224 </dd>
614</dl> 1225</dl>
1226<p>Example: Try to exit cleanly on SIGINT and SIGTERM.</p>
1227<pre> static void
1228 sigint_cb (struct ev_loop *loop, struct ev_signal *w, int revents)
1229 {
1230 ev_unloop (loop, EVUNLOOP_ALL);
1231 }
615 1232
1233 struct ev_signal signal_watcher;
1234 ev_signal_init (&amp;signal_watcher, sigint_cb, SIGINT);
1235 ev_signal_start (loop, &amp;sigint_cb);
1236
1237
1238
1239
1240</pre>
1241
616</div> 1242</div>
1243<h2 id="code_ev_stat_code_did_the_file_attri"><code>ev_stat</code> - did the file attributes just change?</h2>
1244<div id="code_ev_stat_code_did_the_file_attri-2">
1245<p>This watches a filesystem path for attribute changes. That is, it calls
1246<code>stat</code> regularly (or when the OS says it changed) and sees if it changed
1247compared to the last time, invoking the callback if it did.</p>
1248<p>The path does not need to exist: changing from &quot;path exists&quot; to &quot;path does
1249not exist&quot; is a status change like any other. The condition &quot;path does
1250not exist&quot; is signified by the <code>st_nlink</code> field being zero (which is
1251otherwise always forced to be at least one) and all the other fields of
1252the stat buffer having unspecified contents.</p>
1253<p>Since there is no standard to do this, the portable implementation simply
1254calls <code>stat (2)</code> regularly on the path to see if it changed somehow. You
1255can specify a recommended polling interval for this case. If you specify
1256a polling interval of <code>0</code> (highly recommended!) then a <i>suitable,
1257unspecified default</i> value will be used (which you can expect to be around
1258five seconds, although this might change dynamically). Libev will also
1259impose a minimum interval which is currently around <code>0.1</code>, but thats
1260usually overkill.</p>
1261<p>This watcher type is not meant for massive numbers of stat watchers,
1262as even with OS-supported change notifications, this can be
1263resource-intensive.</p>
1264<p>At the time of this writing, only the Linux inotify interface is
1265implemented (implementing kqueue support is left as an exercise for the
1266reader). Inotify will be used to give hints only and should not change the
1267semantics of <code>ev_stat</code> watchers, which means that libev sometimes needs
1268to fall back to regular polling again even with inotify, but changes are
1269usually detected immediately, and if the file exists there will be no
1270polling.</p>
1271<dl>
1272 <dt>ev_stat_init (ev_stat *, callback, const char *path, ev_tstamp interval)</dt>
1273 <dt>ev_stat_set (ev_stat *, const char *path, ev_tstamp interval)</dt>
1274 <dd>
1275 <p>Configures the watcher to wait for status changes of the given
1276<code>path</code>. The <code>interval</code> is a hint on how quickly a change is expected to
1277be detected and should normally be specified as <code>0</code> to let libev choose
1278a suitable value. The memory pointed to by <code>path</code> must point to the same
1279path for as long as the watcher is active.</p>
1280 <p>The callback will be receive <code>EV_STAT</code> when a change was detected,
1281relative to the attributes at the time the watcher was started (or the
1282last change was detected).</p>
1283 </dd>
1284 <dt>ev_stat_stat (ev_stat *)</dt>
1285 <dd>
1286 <p>Updates the stat buffer immediately with new values. If you change the
1287watched path in your callback, you could call this fucntion to avoid
1288detecting this change (while introducing a race condition). Can also be
1289useful simply to find out the new values.</p>
1290 </dd>
1291 <dt>ev_statdata attr [read-only]</dt>
1292 <dd>
1293 <p>The most-recently detected attributes of the file. Although the type is of
1294<code>ev_statdata</code>, this is usually the (or one of the) <code>struct stat</code> types
1295suitable for your system. If the <code>st_nlink</code> member is <code>0</code>, then there
1296was some error while <code>stat</code>ing the file.</p>
1297 </dd>
1298 <dt>ev_statdata prev [read-only]</dt>
1299 <dd>
1300 <p>The previous attributes of the file. The callback gets invoked whenever
1301<code>prev</code> != <code>attr</code>.</p>
1302 </dd>
1303 <dt>ev_tstamp interval [read-only]</dt>
1304 <dd>
1305 <p>The specified interval.</p>
1306 </dd>
1307 <dt>const char *path [read-only]</dt>
1308 <dd>
1309 <p>The filesystem path that is being watched.</p>
1310 </dd>
1311</dl>
1312<p>Example: Watch <code>/etc/passwd</code> for attribute changes.</p>
1313<pre> static void
1314 passwd_cb (struct ev_loop *loop, ev_stat *w, int revents)
1315 {
1316 /* /etc/passwd changed in some way */
1317 if (w-&gt;attr.st_nlink)
1318 {
1319 printf (&quot;passwd current size %ld\n&quot;, (long)w-&gt;attr.st_size);
1320 printf (&quot;passwd current atime %ld\n&quot;, (long)w-&gt;attr.st_mtime);
1321 printf (&quot;passwd current mtime %ld\n&quot;, (long)w-&gt;attr.st_mtime);
1322 }
1323 else
1324 /* you shalt not abuse printf for puts */
1325 puts (&quot;wow, /etc/passwd is not there, expect problems. &quot;
1326 &quot;if this is windows, they already arrived\n&quot;);
1327 }
1328
1329 ...
1330 ev_stat passwd;
1331
1332 ev_stat_init (&amp;passwd, passwd_cb, &quot;/etc/passwd&quot;);
1333 ev_stat_start (loop, &amp;passwd);
1334
1335
1336
1337
1338</pre>
1339
1340</div>
617<h2 id="ev_idle_when_you_ve_got_nothing_bett">ev_idle - when you've got nothing better to do</h2> 1341<h2 id="code_ev_idle_code_when_you_ve_got_no"><code>ev_idle</code> - when you've got nothing better to do...</h2>
618<div id="ev_idle_when_you_ve_got_nothing_bett-2"> 1342<div id="code_ev_idle_code_when_you_ve_got_no-2">
619<p>Idle watchers trigger events when there are no other I/O or timer (or 1343<p>Idle watchers trigger events when there are no other events are pending
620periodic) events pending. That is, as long as your process is busy 1344(prepare, check and other idle watchers do not count). That is, as long
621handling sockets or timeouts it will not be called. But when your process 1345as your process is busy handling sockets or timeouts (or even signals,
622is idle all idle watchers are being called again and again - until 1346imagine) it will not be triggered. But when your process is idle all idle
1347watchers are being called again and again, once per event loop iteration -
623stopped, that is, or your process receives more events.</p> 1348until stopped, that is, or your process receives more events and becomes
1349busy.</p>
624<p>The most noteworthy effect is that as long as any idle watchers are 1350<p>The most noteworthy effect is that as long as any idle watchers are
625active, the process will not block when waiting for new events.</p> 1351active, the process will not block when waiting for new events.</p>
626<p>Apart from keeping your process non-blocking (which is a useful 1352<p>Apart from keeping your process non-blocking (which is a useful
627effect on its own sometimes), idle watchers are a good place to do 1353effect on its own sometimes), idle watchers are a good place to do
628&quot;pseudo-background processing&quot;, or delay processing stuff to after the 1354&quot;pseudo-background processing&quot;, or delay processing stuff to after the
633 <p>Initialises and configures the idle watcher - it has no parameters of any 1359 <p>Initialises and configures the idle watcher - it has no parameters of any
634kind. There is a <code>ev_idle_set</code> macro, but using it is utterly pointless, 1360kind. There is a <code>ev_idle_set</code> macro, but using it is utterly pointless,
635believe me.</p> 1361believe me.</p>
636 </dd> 1362 </dd>
637</dl> 1363</dl>
1364<p>Example: Dynamically allocate an <code>ev_idle</code> watcher, start it, and in the
1365callback, free it. Also, use no error checking, as usual.</p>
1366<pre> static void
1367 idle_cb (struct ev_loop *loop, struct ev_idle *w, int revents)
1368 {
1369 free (w);
1370 // now do something you wanted to do when the program has
1371 // no longer asnything immediate to do.
1372 }
638 1373
1374 struct ev_idle *idle_watcher = malloc (sizeof (struct ev_idle));
1375 ev_idle_init (idle_watcher, idle_cb);
1376 ev_idle_start (loop, idle_cb);
1377
1378
1379
1380
1381</pre>
1382
639</div> 1383</div>
640<h2 id="prepare_and_check_your_hooks_into_th">prepare and check - your hooks into the event loop</h2> 1384<h2 id="code_ev_prepare_code_and_code_ev_che"><code>ev_prepare</code> and <code>ev_check</code> - customise your event loop!</h2>
641<div id="prepare_and_check_your_hooks_into_th-2"> 1385<div id="code_ev_prepare_code_and_code_ev_che-2">
642<p>Prepare and check watchers usually (but not always) are used in 1386<p>Prepare and check watchers are usually (but not always) used in tandem:
643tandom. Prepare watchers get invoked before the process blocks and check 1387prepare watchers get invoked before the process blocks and check watchers
644watchers afterwards.</p> 1388afterwards.</p>
1389<p>You <i>must not</i> call <code>ev_loop</code> or similar functions that enter
1390the current event loop from either <code>ev_prepare</code> or <code>ev_check</code>
1391watchers. Other loops than the current one are fine, however. The
1392rationale behind this is that you do not need to check for recursion in
1393those watchers, i.e. the sequence will always be <code>ev_prepare</code>, blocking,
1394<code>ev_check</code> so if you have one watcher of each kind they will always be
1395called in pairs bracketing the blocking call.</p>
645<p>Their main purpose is to integrate other event mechanisms into libev. This 1396<p>Their main purpose is to integrate other event mechanisms into libev and
646could be used, for example, to track variable changes, implement your own 1397their use is somewhat advanced. This could be used, for example, to track
647watchers, integrate net-snmp or a coroutine library and lots more.</p> 1398variable changes, implement your own watchers, integrate net-snmp or a
1399coroutine library and lots more. They are also occasionally useful if
1400you cache some data and want to flush it before blocking (for example,
1401in X programs you might want to do an <code>XFlush ()</code> in an <code>ev_prepare</code>
1402watcher).</p>
648<p>This is done by examining in each prepare call which file descriptors need 1403<p>This is done by examining in each prepare call which file descriptors need
649to be watched by the other library, registering ev_io watchers for them 1404to be watched by the other library, registering <code>ev_io</code> watchers for
650and starting an ev_timer watcher for any timeouts (many libraries provide 1405them and starting an <code>ev_timer</code> watcher for any timeouts (many libraries
651just this functionality). Then, in the check watcher you check for any 1406provide just this functionality). Then, in the check watcher you check for
652events that occured (by making your callbacks set soem flags for example) 1407any events that occured (by checking the pending status of all watchers
653and call back into the library.</p> 1408and stopping them) and call back into the library. The I/O and timer
1409callbacks will never actually be called (but must be valid nevertheless,
1410because you never know, you know?).</p>
654<p>As another example, the perl Coro module uses these hooks to integrate 1411<p>As another example, the Perl Coro module uses these hooks to integrate
655coroutines into libev programs, by yielding to other active coroutines 1412coroutines into libev programs, by yielding to other active coroutines
656during each prepare and only letting the process block if no coroutines 1413during each prepare and only letting the process block if no coroutines
657are ready to run.</p> 1414are ready to run (it's actually more complicated: it only runs coroutines
1415with priority higher than or equal to the event loop and one coroutine
1416of lower priority, but only once, using idle watchers to keep the event
1417loop from blocking if lower-priority coroutines are active, thus mapping
1418low-priority coroutines to idle/background tasks).</p>
658<dl> 1419<dl>
659 <dt>ev_prepare_init (ev_prepare *, callback)</dt> 1420 <dt>ev_prepare_init (ev_prepare *, callback)</dt>
660 <dt>ev_check_init (ev_check *, callback)</dt> 1421 <dt>ev_check_init (ev_check *, callback)</dt>
661 <dd> 1422 <dd>
662 <p>Initialises and configures the prepare or check watcher - they have no 1423 <p>Initialises and configures the prepare or check watcher - they have no
663parameters of any kind. There are <code>ev_prepare_set</code> and <code>ev_check_set</code> 1424parameters of any kind. There are <code>ev_prepare_set</code> and <code>ev_check_set</code>
664macros, but using them is utterly, utterly pointless.</p> 1425macros, but using them is utterly, utterly and completely pointless.</p>
665 </dd> 1426 </dd>
666</dl> 1427</dl>
1428<p>Example: To include a library such as adns, you would add IO watchers
1429and a timeout watcher in a prepare handler, as required by libadns, and
1430in a check watcher, destroy them and call into libadns. What follows is
1431pseudo-code only of course:</p>
1432<pre> static ev_io iow [nfd];
1433 static ev_timer tw;
667 1434
1435 static void
1436 io_cb (ev_loop *loop, ev_io *w, int revents)
1437 {
1438 // set the relevant poll flags
1439 // could also call adns_processreadable etc. here
1440 struct pollfd *fd = (struct pollfd *)w-&gt;data;
1441 if (revents &amp; EV_READ ) fd-&gt;revents |= fd-&gt;events &amp; POLLIN;
1442 if (revents &amp; EV_WRITE) fd-&gt;revents |= fd-&gt;events &amp; POLLOUT;
1443 }
1444
1445 // create io watchers for each fd and a timer before blocking
1446 static void
1447 adns_prepare_cb (ev_loop *loop, ev_prepare *w, int revents)
1448 {
1449 int timeout = 3600000;truct pollfd fds [nfd];
1450 // actual code will need to loop here and realloc etc.
1451 adns_beforepoll (ads, fds, &amp;nfd, &amp;timeout, timeval_from (ev_time ()));
1452
1453 /* the callback is illegal, but won't be called as we stop during check */
1454 ev_timer_init (&amp;tw, 0, timeout * 1e-3);
1455 ev_timer_start (loop, &amp;tw);
1456
1457 // create on ev_io per pollfd
1458 for (int i = 0; i &lt; nfd; ++i)
1459 {
1460 ev_io_init (iow + i, io_cb, fds [i].fd,
1461 ((fds [i].events &amp; POLLIN ? EV_READ : 0)
1462 | (fds [i].events &amp; POLLOUT ? EV_WRITE : 0)));
1463
1464 fds [i].revents = 0;
1465 iow [i].data = fds + i;
1466 ev_io_start (loop, iow + i);
1467 }
1468 }
1469
1470 // stop all watchers after blocking
1471 static void
1472 adns_check_cb (ev_loop *loop, ev_check *w, int revents)
1473 {
1474 ev_timer_stop (loop, &amp;tw);
1475
1476 for (int i = 0; i &lt; nfd; ++i)
1477 ev_io_stop (loop, iow + i);
1478
1479 adns_afterpoll (adns, fds, nfd, timeval_from (ev_now (loop));
1480 }
1481
1482
1483
1484
1485</pre>
1486
668</div> 1487</div>
669<h1 id="OTHER_FUNCTIONS">OTHER FUNCTIONS</h1><p><a href="#TOP" class="toplink">Top</a></p> 1488<h2 id="code_ev_embed_code_when_one_backend_"><code>ev_embed</code> - when one backend isn't enough...</h2>
1489<div id="code_ev_embed_code_when_one_backend_-2">
1490<p>This is a rather advanced watcher type that lets you embed one event loop
1491into another (currently only <code>ev_io</code> events are supported in the embedded
1492loop, other types of watchers might be handled in a delayed or incorrect
1493fashion and must not be used).</p>
1494<p>There are primarily two reasons you would want that: work around bugs and
1495prioritise I/O.</p>
1496<p>As an example for a bug workaround, the kqueue backend might only support
1497sockets on some platform, so it is unusable as generic backend, but you
1498still want to make use of it because you have many sockets and it scales
1499so nicely. In this case, you would create a kqueue-based loop and embed it
1500into your default loop (which might use e.g. poll). Overall operation will
1501be a bit slower because first libev has to poll and then call kevent, but
1502at least you can use both at what they are best.</p>
1503<p>As for prioritising I/O: rarely you have the case where some fds have
1504to be watched and handled very quickly (with low latency), and even
1505priorities and idle watchers might have too much overhead. In this case
1506you would put all the high priority stuff in one loop and all the rest in
1507a second one, and embed the second one in the first.</p>
1508<p>As long as the watcher is active, the callback will be invoked every time
1509there might be events pending in the embedded loop. The callback must then
1510call <code>ev_embed_sweep (mainloop, watcher)</code> to make a single sweep and invoke
1511their callbacks (you could also start an idle watcher to give the embedded
1512loop strictly lower priority for example). You can also set the callback
1513to <code>0</code>, in which case the embed watcher will automatically execute the
1514embedded loop sweep.</p>
1515<p>As long as the watcher is started it will automatically handle events. The
1516callback will be invoked whenever some events have been handled. You can
1517set the callback to <code>0</code> to avoid having to specify one if you are not
1518interested in that.</p>
1519<p>Also, there have not currently been made special provisions for forking:
1520when you fork, you not only have to call <code>ev_loop_fork</code> on both loops,
1521but you will also have to stop and restart any <code>ev_embed</code> watchers
1522yourself.</p>
1523<p>Unfortunately, not all backends are embeddable, only the ones returned by
1524<code>ev_embeddable_backends</code> are, which, unfortunately, does not include any
1525portable one.</p>
1526<p>So when you want to use this feature you will always have to be prepared
1527that you cannot get an embeddable loop. The recommended way to get around
1528this is to have a separate variables for your embeddable loop, try to
1529create it, and if that fails, use the normal loop for everything:</p>
1530<pre> struct ev_loop *loop_hi = ev_default_init (0);
1531 struct ev_loop *loop_lo = 0;
1532 struct ev_embed embed;
1533
1534 // see if there is a chance of getting one that works
1535 // (remember that a flags value of 0 means autodetection)
1536 loop_lo = ev_embeddable_backends () &amp; ev_recommended_backends ()
1537 ? ev_loop_new (ev_embeddable_backends () &amp; ev_recommended_backends ())
1538 : 0;
1539
1540 // if we got one, then embed it, otherwise default to loop_hi
1541 if (loop_lo)
1542 {
1543 ev_embed_init (&amp;embed, 0, loop_lo);
1544 ev_embed_start (loop_hi, &amp;embed);
1545 }
1546 else
1547 loop_lo = loop_hi;
1548
1549</pre>
1550<dl>
1551 <dt>ev_embed_init (ev_embed *, callback, struct ev_loop *embedded_loop)</dt>
1552 <dt>ev_embed_set (ev_embed *, callback, struct ev_loop *embedded_loop)</dt>
1553 <dd>
1554 <p>Configures the watcher to embed the given loop, which must be
1555embeddable. If the callback is <code>0</code>, then <code>ev_embed_sweep</code> will be
1556invoked automatically, otherwise it is the responsibility of the callback
1557to invoke it (it will continue to be called until the sweep has been done,
1558if you do not want thta, you need to temporarily stop the embed watcher).</p>
1559 </dd>
1560 <dt>ev_embed_sweep (loop, ev_embed *)</dt>
1561 <dd>
1562 <p>Make a single, non-blocking sweep over the embedded loop. This works
1563similarly to <code>ev_loop (embedded_loop, EVLOOP_NONBLOCK)</code>, but in the most
1564apropriate way for embedded loops.</p>
1565 </dd>
1566 <dt>struct ev_loop *loop [read-only]</dt>
1567 <dd>
1568 <p>The embedded event loop.</p>
1569 </dd>
1570</dl>
1571
1572
1573
1574
1575
1576</div>
1577<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>
1578<div id="code_ev_fork_code_the_audacity_to_re-2">
1579<p>Fork watchers are called when a <code>fork ()</code> was detected (usually because
1580whoever is a good citizen cared to tell libev about it by calling
1581<code>ev_default_fork</code> or <code>ev_loop_fork</code>). The invocation is done before the
1582event loop blocks next and before <code>ev_check</code> watchers are being called,
1583and only in the child after the fork. If whoever good citizen calling
1584<code>ev_default_fork</code> cheats and calls it in the wrong process, the fork
1585handlers will be invoked, too, of course.</p>
1586<dl>
1587 <dt>ev_fork_init (ev_signal *, callback)</dt>
1588 <dd>
1589 <p>Initialises and configures the fork watcher - it has no parameters of any
1590kind. There is a <code>ev_fork_set</code> macro, but using it is utterly pointless,
1591believe me.</p>
1592 </dd>
1593</dl>
1594
1595
1596
1597
1598
1599</div>
1600<h1 id="OTHER_FUNCTIONS">OTHER FUNCTIONS</h1>
670<div id="OTHER_FUNCTIONS_CONTENT"> 1601<div id="OTHER_FUNCTIONS_CONTENT">
671<p>There are some other fucntions of possible interest. Described. Here. Now.</p> 1602<p>There are some other functions of possible interest. Described. Here. Now.</p>
672<dl> 1603<dl>
673 <dt>ev_once (loop, int fd, int events, ev_tstamp timeout, callback)</dt> 1604 <dt>ev_once (loop, int fd, int events, ev_tstamp timeout, callback)</dt>
674 <dd> 1605 <dd>
675 <p>This function combines a simple timer and an I/O watcher, calls your 1606 <p>This function combines a simple timer and an I/O watcher, calls your
676callback on whichever event happens first and automatically stop both 1607callback on whichever event happens first and automatically stop both
677watchers. This is useful if you want to wait for a single event on an fd 1608watchers. This is useful if you want to wait for a single event on an fd
678or timeout without havign to allocate/configure/start/stop/free one or 1609or timeout without having to allocate/configure/start/stop/free one or
679more watchers yourself.</p> 1610more watchers yourself.</p>
680 <p>If <code>fd</code> is less than 0, then no I/O watcher will be started and events is 1611 <p>If <code>fd</code> is less than 0, then no I/O watcher will be started and events
681ignored. Otherwise, an ev_io watcher for the given <code>fd</code> and <code>events</code> set 1612is being ignored. Otherwise, an <code>ev_io</code> watcher for the given <code>fd</code> and
682will be craeted and started.</p> 1613<code>events</code> set will be craeted and started.</p>
683 <p>If <code>timeout</code> is less than 0, then no timeout watcher will be 1614 <p>If <code>timeout</code> is less than 0, then no timeout watcher will be
684started. Otherwise an ev_timer watcher with after = <code>timeout</code> (and repeat 1615started. Otherwise an <code>ev_timer</code> watcher with after = <code>timeout</code> (and
685= 0) will be started.</p> 1616repeat = 0) will be started. While <code>0</code> is a valid timeout, it is of
1617dubious value.</p>
686 <p>The callback has the type <code>void (*cb)(int revents, void *arg)</code> and 1618 <p>The callback has the type <code>void (*cb)(int revents, void *arg)</code> and gets
687gets passed an events set (normally a combination of EV_ERROR, EV_READ, 1619passed an <code>revents</code> set like normal event callbacks (a combination of
688EV_WRITE or EV_TIMEOUT) and the <code>arg</code> value passed to <code>ev_once</code>:</p> 1620<code>EV_ERROR</code>, <code>EV_READ</code>, <code>EV_WRITE</code> or <code>EV_TIMEOUT</code>) and the <code>arg</code>
1621value passed to <code>ev_once</code>:</p>
689<pre> static void stdin_ready (int revents, void *arg) 1622<pre> static void stdin_ready (int revents, void *arg)
690 { 1623 {
691 if (revents &amp; EV_TIMEOUT) 1624 if (revents &amp; EV_TIMEOUT)
692 /* doh, nothing entered */ 1625 /* doh, nothing entered */;
693 else if (revents &amp; EV_READ) 1626 else if (revents &amp; EV_READ)
694 /* stdin might have data for us, joy! */ 1627 /* stdin might have data for us, joy! */;
695 } 1628 }
696 1629
697 ev_once (STDIN_FILENO, EV_READm 10., stdin_ready, 0); 1630 ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);
698 1631
699</pre> 1632</pre>
700 </dd> 1633 </dd>
701 <dt>ev_feed_event (loop, watcher, int events)</dt> 1634 <dt>ev_feed_event (ev_loop *, watcher *, int revents)</dt>
702 <dd> 1635 <dd>
703 <p>Feeds the given event set into the event loop, as if the specified event 1636 <p>Feeds the given event set into the event loop, as if the specified event
704has happened for the specified watcher (which must be a pointer to an 1637had happened for the specified watcher (which must be a pointer to an
705initialised but not necessarily active event watcher).</p> 1638initialised but not necessarily started event watcher).</p>
706 </dd> 1639 </dd>
707 <dt>ev_feed_fd_event (loop, int fd, int revents)</dt> 1640 <dt>ev_feed_fd_event (ev_loop *, int fd, int revents)</dt>
708 <dd> 1641 <dd>
709 <p>Feed an event on the given fd, as if a file descriptor backend detected it.</p> 1642 <p>Feed an event on the given fd, as if a file descriptor backend detected
1643the given events it.</p>
710 </dd> 1644 </dd>
711 <dt>ev_feed_signal_event (loop, int signum)</dt> 1645 <dt>ev_feed_signal_event (ev_loop *loop, int signum)</dt>
712 <dd> 1646 <dd>
713 <p>Feed an event as if the given signal occured (loop must be the default loop!).</p> 1647 <p>Feed an event as if the given signal occured (<code>loop</code> must be the default
1648loop!).</p>
714 </dd> 1649 </dd>
715</dl> 1650</dl>
716 1651
1652
1653
1654
1655
717</div> 1656</div>
718<h1 id="AUTHOR">AUTHOR</h1><p><a href="#TOP" class="toplink">Top</a></p> 1657<h1 id="LIBEVENT_EMULATION">LIBEVENT EMULATION</h1>
1658<div id="LIBEVENT_EMULATION_CONTENT">
1659<p>Libev offers a compatibility emulation layer for libevent. It cannot
1660emulate the internals of libevent, so here are some usage hints:</p>
1661<dl>
1662 <dt>* Use it by including &lt;event.h&gt;, as usual.</dt>
1663 <dt>* The following members are fully supported: ev_base, ev_callback,
1664ev_arg, ev_fd, ev_res, ev_events.</dt>
1665 <dt>* Avoid using ev_flags and the EVLIST_*-macros, while it is
1666maintained by libev, it does not work exactly the same way as in libevent (consider
1667it a private API).</dt>
1668 <dt>* Priorities are not currently supported. Initialising priorities
1669will fail and all watchers will have the same priority, even though there
1670is an ev_pri field.</dt>
1671 <dt>* Other members are not supported.</dt>
1672 <dt>* The libev emulation is <i>not</i> ABI compatible to libevent, you need
1673to use the libev header file and library.</dt>
1674</dl>
1675
1676</div>
1677<h1 id="C_SUPPORT">C++ SUPPORT</h1>
1678<div id="C_SUPPORT_CONTENT">
1679<p>Libev comes with some simplistic wrapper classes for C++ that mainly allow
1680you to use some convinience methods to start/stop watchers and also change
1681the callback model to a model using method callbacks on objects.</p>
1682<p>To use it,</p>
1683<pre> #include &lt;ev++.h&gt;
1684
1685</pre>
1686<p>(it is not installed by default). This automatically includes <cite>ev.h</cite>
1687and puts all of its definitions (many of them macros) into the global
1688namespace. All C++ specific things are put into the <code>ev</code> namespace.</p>
1689<p>It should support all the same embedding options as <cite>ev.h</cite>, most notably
1690<code>EV_MULTIPLICITY</code>.</p>
1691<p>Here is a list of things available in the <code>ev</code> namespace:</p>
1692<dl>
1693 <dt><code>ev::READ</code>, <code>ev::WRITE</code> etc.</dt>
1694 <dd>
1695 <p>These are just enum values with the same values as the <code>EV_READ</code> etc.
1696macros from <cite>ev.h</cite>.</p>
1697 </dd>
1698 <dt><code>ev::tstamp</code>, <code>ev::now</code></dt>
1699 <dd>
1700 <p>Aliases to the same types/functions as with the <code>ev_</code> prefix.</p>
1701 </dd>
1702 <dt><code>ev::io</code>, <code>ev::timer</code>, <code>ev::periodic</code>, <code>ev::idle</code>, <code>ev::sig</code> etc.</dt>
1703 <dd>
1704 <p>For each <code>ev_TYPE</code> watcher in <cite>ev.h</cite> there is a corresponding class of
1705the same name in the <code>ev</code> namespace, with the exception of <code>ev_signal</code>
1706which is called <code>ev::sig</code> to avoid clashes with the <code>signal</code> macro
1707defines by many implementations.</p>
1708 <p>All of those classes have these methods:</p>
1709 <p>
1710 <dl>
1711 <dt>ev::TYPE::TYPE (object *, object::method *)</dt>
1712 <dt>ev::TYPE::TYPE (object *, object::method *, struct ev_loop *)</dt>
1713 <dt>ev::TYPE::~TYPE</dt>
1714 <dd>
1715 <p>The constructor takes a pointer to an object and a method pointer to
1716the event handler callback to call in this class. The constructor calls
1717<code>ev_init</code> for you, which means you have to call the <code>set</code> method
1718before starting it. If you do not specify a loop then the constructor
1719automatically associates the default loop with this watcher.</p>
1720 <p>The destructor automatically stops the watcher if it is active.</p>
1721 </dd>
1722 <dt>w-&gt;set (struct ev_loop *)</dt>
1723 <dd>
1724 <p>Associates a different <code>struct ev_loop</code> with this watcher. You can only
1725do this when the watcher is inactive (and not pending either).</p>
1726 </dd>
1727 <dt>w-&gt;set ([args])</dt>
1728 <dd>
1729 <p>Basically the same as <code>ev_TYPE_set</code>, with the same args. Must be
1730called at least once. Unlike the C counterpart, an active watcher gets
1731automatically stopped and restarted.</p>
1732 </dd>
1733 <dt>w-&gt;start ()</dt>
1734 <dd>
1735 <p>Starts the watcher. Note that there is no <code>loop</code> argument as the
1736constructor already takes the loop.</p>
1737 </dd>
1738 <dt>w-&gt;stop ()</dt>
1739 <dd>
1740 <p>Stops the watcher if it is active. Again, no <code>loop</code> argument.</p>
1741 </dd>
1742 <dt>w-&gt;again () <code>ev::timer</code>, <code>ev::periodic</code> only</dt>
1743 <dd>
1744 <p>For <code>ev::timer</code> and <code>ev::periodic</code>, this invokes the corresponding
1745<code>ev_TYPE_again</code> function.</p>
1746 </dd>
1747 <dt>w-&gt;sweep () <code>ev::embed</code> only</dt>
1748 <dd>
1749 <p>Invokes <code>ev_embed_sweep</code>.</p>
1750 </dd>
1751 <dt>w-&gt;update () <code>ev::stat</code> only</dt>
1752 <dd>
1753 <p>Invokes <code>ev_stat_stat</code>.</p>
1754 </dd>
1755 </dl>
1756 </p>
1757 </dd>
1758</dl>
1759<p>Example: Define a class with an IO and idle watcher, start one of them in
1760the constructor.</p>
1761<pre> class myclass
1762 {
1763 ev_io io; void io_cb (ev::io &amp;w, int revents);
1764 ev_idle idle void idle_cb (ev::idle &amp;w, int revents);
1765
1766 myclass ();
1767 }
1768
1769 myclass::myclass (int fd)
1770 : io (this, &amp;myclass::io_cb),
1771 idle (this, &amp;myclass::idle_cb)
1772 {
1773 io.start (fd, ev::READ);
1774 }
1775
1776
1777
1778
1779</pre>
1780
1781</div>
1782<h1 id="MACRO_MAGIC">MACRO MAGIC</h1>
1783<div id="MACRO_MAGIC_CONTENT">
1784<p>Libev can be compiled with a variety of options, the most fundemantal is
1785<code>EV_MULTIPLICITY</code>. This option determines wether (most) functions and
1786callbacks have an initial <code>struct ev_loop *</code> argument.</p>
1787<p>To make it easier to write programs that cope with either variant, the
1788following macros are defined:</p>
1789<dl>
1790 <dt><code>EV_A</code>, <code>EV_A_</code></dt>
1791 <dd>
1792 <p>This provides the loop <i>argument</i> for functions, if one is required (&quot;ev
1793loop argument&quot;). The <code>EV_A</code> form is used when this is the sole argument,
1794<code>EV_A_</code> is used when other arguments are following. Example:</p>
1795<pre> ev_unref (EV_A);
1796 ev_timer_add (EV_A_ watcher);
1797 ev_loop (EV_A_ 0);
1798
1799</pre>
1800 <p>It assumes the variable <code>loop</code> of type <code>struct ev_loop *</code> is in scope,
1801which is often provided by the following macro.</p>
1802 </dd>
1803 <dt><code>EV_P</code>, <code>EV_P_</code></dt>
1804 <dd>
1805 <p>This provides the loop <i>parameter</i> for functions, if one is required (&quot;ev
1806loop parameter&quot;). The <code>EV_P</code> form is used when this is the sole parameter,
1807<code>EV_P_</code> is used when other parameters are following. Example:</p>
1808<pre> // this is how ev_unref is being declared
1809 static void ev_unref (EV_P);
1810
1811 // this is how you can declare your typical callback
1812 static void cb (EV_P_ ev_timer *w, int revents)
1813
1814</pre>
1815 <p>It declares a parameter <code>loop</code> of type <code>struct ev_loop *</code>, quite
1816suitable for use with <code>EV_A</code>.</p>
1817 </dd>
1818 <dt><code>EV_DEFAULT</code>, <code>EV_DEFAULT_</code></dt>
1819 <dd>
1820 <p>Similar to the other two macros, this gives you the value of the default
1821loop, if multiple loops are supported (&quot;ev loop default&quot;).</p>
1822 </dd>
1823</dl>
1824<p>Example: Declare and initialise a check watcher, working regardless of
1825wether multiple loops are supported or not.</p>
1826<pre> static void
1827 check_cb (EV_P_ ev_timer *w, int revents)
1828 {
1829 ev_check_stop (EV_A_ w);
1830 }
1831
1832 ev_check check;
1833 ev_check_init (&amp;check, check_cb);
1834 ev_check_start (EV_DEFAULT_ &amp;check);
1835 ev_loop (EV_DEFAULT_ 0);
1836
1837
1838
1839
1840</pre>
1841
1842</div>
1843<h1 id="EMBEDDING">EMBEDDING</h1>
1844<div id="EMBEDDING_CONTENT">
1845<p>Libev can (and often is) directly embedded into host
1846applications. Examples of applications that embed it include the Deliantra
1847Game Server, the EV perl module, the GNU Virtual Private Ethernet (gvpe)
1848and rxvt-unicode.</p>
1849<p>The goal is to enable you to just copy the neecssary files into your
1850source directory without having to change even a single line in them, so
1851you can easily upgrade by simply copying (or having a checked-out copy of
1852libev somewhere in your source tree).</p>
1853
1854</div>
1855<h2 id="FILESETS">FILESETS</h2>
1856<div id="FILESETS_CONTENT">
1857<p>Depending on what features you need you need to include one or more sets of files
1858in your app.</p>
1859
1860</div>
1861<h3 id="CORE_EVENT_LOOP">CORE EVENT LOOP</h3>
1862<div id="CORE_EVENT_LOOP_CONTENT">
1863<p>To include only the libev core (all the <code>ev_*</code> functions), with manual
1864configuration (no autoconf):</p>
1865<pre> #define EV_STANDALONE 1
1866 #include &quot;ev.c&quot;
1867
1868</pre>
1869<p>This will automatically include <cite>ev.h</cite>, too, and should be done in a
1870single C source file only to provide the function implementations. To use
1871it, do the same for <cite>ev.h</cite> in all files wishing to use this API (best
1872done by writing a wrapper around <cite>ev.h</cite> that you can include instead and
1873where you can put other configuration options):</p>
1874<pre> #define EV_STANDALONE 1
1875 #include &quot;ev.h&quot;
1876
1877</pre>
1878<p>Both header files and implementation files can be compiled with a C++
1879compiler (at least, thats a stated goal, and breakage will be treated
1880as a bug).</p>
1881<p>You need the following files in your source tree, or in a directory
1882in your include path (e.g. in libev/ when using -Ilibev):</p>
1883<pre> ev.h
1884 ev.c
1885 ev_vars.h
1886 ev_wrap.h
1887
1888 ev_win32.c required on win32 platforms only
1889
1890 ev_select.c only when select backend is enabled (which is by default)
1891 ev_poll.c only when poll backend is enabled (disabled by default)
1892 ev_epoll.c only when the epoll backend is enabled (disabled by default)
1893 ev_kqueue.c only when the kqueue backend is enabled (disabled by default)
1894 ev_port.c only when the solaris port backend is enabled (disabled by default)
1895
1896</pre>
1897<p><cite>ev.c</cite> includes the backend files directly when enabled, so you only need
1898to compile this single file.</p>
1899
1900</div>
1901<h3 id="LIBEVENT_COMPATIBILITY_API">LIBEVENT COMPATIBILITY API</h3>
1902<div id="LIBEVENT_COMPATIBILITY_API_CONTENT">
1903<p>To include the libevent compatibility API, also include:</p>
1904<pre> #include &quot;event.c&quot;
1905
1906</pre>
1907<p>in the file including <cite>ev.c</cite>, and:</p>
1908<pre> #include &quot;event.h&quot;
1909
1910</pre>
1911<p>in the files that want to use the libevent API. This also includes <cite>ev.h</cite>.</p>
1912<p>You need the following additional files for this:</p>
1913<pre> event.h
1914 event.c
1915
1916</pre>
1917
1918</div>
1919<h3 id="AUTOCONF_SUPPORT">AUTOCONF SUPPORT</h3>
1920<div id="AUTOCONF_SUPPORT_CONTENT">
1921<p>Instead of using <code>EV_STANDALONE=1</code> and providing your config in
1922whatever way you want, you can also <code>m4_include([libev.m4])</code> in your
1923<cite>configure.ac</cite> and leave <code>EV_STANDALONE</code> undefined. <cite>ev.c</cite> will then
1924include <cite>config.h</cite> and configure itself accordingly.</p>
1925<p>For this of course you need the m4 file:</p>
1926<pre> libev.m4
1927
1928</pre>
1929
1930</div>
1931<h2 id="PREPROCESSOR_SYMBOLS_MACROS">PREPROCESSOR SYMBOLS/MACROS</h2>
1932<div id="PREPROCESSOR_SYMBOLS_MACROS_CONTENT">
1933<p>Libev can be configured via a variety of preprocessor symbols you have to define
1934before including any of its files. The default is not to build for multiplicity
1935and only include the select backend.</p>
1936<dl>
1937 <dt>EV_STANDALONE</dt>
1938 <dd>
1939 <p>Must always be <code>1</code> if you do not use autoconf configuration, which
1940keeps libev from including <cite>config.h</cite>, and it also defines dummy
1941implementations for some libevent functions (such as logging, which is not
1942supported). It will also not define any of the structs usually found in
1943<cite>event.h</cite> that are not directly supported by the libev core alone.</p>
1944 </dd>
1945 <dt>EV_USE_MONOTONIC</dt>
1946 <dd>
1947 <p>If defined to be <code>1</code>, libev will try to detect the availability of the
1948monotonic clock option at both compiletime and runtime. Otherwise no use
1949of the monotonic clock option will be attempted. If you enable this, you
1950usually have to link against librt or something similar. Enabling it when
1951the functionality isn't available is safe, though, althoguh you have
1952to make sure you link against any libraries where the <code>clock_gettime</code>
1953function is hiding in (often <cite>-lrt</cite>).</p>
1954 </dd>
1955 <dt>EV_USE_REALTIME</dt>
1956 <dd>
1957 <p>If defined to be <code>1</code>, libev will try to detect the availability of the
1958realtime clock option at compiletime (and assume its availability at
1959runtime if successful). Otherwise no use of the realtime clock option will
1960be attempted. This effectively replaces <code>gettimeofday</code> by <code>clock_get
1961(CLOCK_REALTIME, ...)</code> and will not normally affect correctness. See tzhe note about libraries
1962in the description of <code>EV_USE_MONOTONIC</code>, though.</p>
1963 </dd>
1964 <dt>EV_USE_SELECT</dt>
1965 <dd>
1966 <p>If undefined or defined to be <code>1</code>, libev will compile in support for the
1967<code>select</code>(2) backend. No attempt at autodetection will be done: if no
1968other method takes over, select will be it. Otherwise the select backend
1969will not be compiled in.</p>
1970 </dd>
1971 <dt>EV_SELECT_USE_FD_SET</dt>
1972 <dd>
1973 <p>If defined to <code>1</code>, then the select backend will use the system <code>fd_set</code>
1974structure. This is useful if libev doesn't compile due to a missing
1975<code>NFDBITS</code> or <code>fd_mask</code> definition or it misguesses the bitset layout on
1976exotic systems. This usually limits the range of file descriptors to some
1977low limit such as 1024 or might have other limitations (winsocket only
1978allows 64 sockets). The <code>FD_SETSIZE</code> macro, set before compilation, might
1979influence the size of the <code>fd_set</code> used.</p>
1980 </dd>
1981 <dt>EV_SELECT_IS_WINSOCKET</dt>
1982 <dd>
1983 <p>When defined to <code>1</code>, the select backend will assume that
1984select/socket/connect etc. don't understand file descriptors but
1985wants osf handles on win32 (this is the case when the select to
1986be used is the winsock select). This means that it will call
1987<code>_get_osfhandle</code> on the fd to convert it to an OS handle. Otherwise,
1988it is assumed that all these functions actually work on fds, even
1989on win32. Should not be defined on non-win32 platforms.</p>
1990 </dd>
1991 <dt>EV_USE_POLL</dt>
1992 <dd>
1993 <p>If defined to be <code>1</code>, libev will compile in support for the <code>poll</code>(2)
1994backend. Otherwise it will be enabled on non-win32 platforms. It
1995takes precedence over select.</p>
1996 </dd>
1997 <dt>EV_USE_EPOLL</dt>
1998 <dd>
1999 <p>If defined to be <code>1</code>, libev will compile in support for the Linux
2000<code>epoll</code>(7) backend. Its availability will be detected at runtime,
2001otherwise another method will be used as fallback. This is the
2002preferred backend for GNU/Linux systems.</p>
2003 </dd>
2004 <dt>EV_USE_KQUEUE</dt>
2005 <dd>
2006 <p>If defined to be <code>1</code>, libev will compile in support for the BSD style
2007<code>kqueue</code>(2) backend. Its actual availability will be detected at runtime,
2008otherwise another method will be used as fallback. This is the preferred
2009backend for BSD and BSD-like systems, although on most BSDs kqueue only
2010supports some types of fds correctly (the only platform we found that
2011supports ptys for example was NetBSD), so kqueue might be compiled in, but
2012not be used unless explicitly requested. The best way to use it is to find
2013out whether kqueue supports your type of fd properly and use an embedded
2014kqueue loop.</p>
2015 </dd>
2016 <dt>EV_USE_PORT</dt>
2017 <dd>
2018 <p>If defined to be <code>1</code>, libev will compile in support for the Solaris
201910 port style backend. Its availability will be detected at runtime,
2020otherwise another method will be used as fallback. This is the preferred
2021backend for Solaris 10 systems.</p>
2022 </dd>
2023 <dt>EV_USE_DEVPOLL</dt>
2024 <dd>
2025 <p>reserved for future expansion, works like the USE symbols above.</p>
2026 </dd>
2027 <dt>EV_USE_INOTIFY</dt>
2028 <dd>
2029 <p>If defined to be <code>1</code>, libev will compile in support for the Linux inotify
2030interface to speed up <code>ev_stat</code> watchers. Its actual availability will
2031be detected at runtime.</p>
2032 </dd>
2033 <dt>EV_H</dt>
2034 <dd>
2035 <p>The name of the <cite>ev.h</cite> header file used to include it. The default if
2036undefined is <code>&lt;ev.h&gt;</code> in <cite>event.h</cite> and <code>&quot;ev.h&quot;</code> in <cite>ev.c</cite>. This
2037can be used to virtually rename the <cite>ev.h</cite> header file in case of conflicts.</p>
2038 </dd>
2039 <dt>EV_CONFIG_H</dt>
2040 <dd>
2041 <p>If <code>EV_STANDALONE</code> isn't <code>1</code>, this variable can be used to override
2042<cite>ev.c</cite>'s idea of where to find the <cite>config.h</cite> file, similarly to
2043<code>EV_H</code>, above.</p>
2044 </dd>
2045 <dt>EV_EVENT_H</dt>
2046 <dd>
2047 <p>Similarly to <code>EV_H</code>, this macro can be used to override <cite>event.c</cite>'s idea
2048of how the <cite>event.h</cite> header can be found.</p>
2049 </dd>
2050 <dt>EV_PROTOTYPES</dt>
2051 <dd>
2052 <p>If defined to be <code>0</code>, then <cite>ev.h</cite> will not define any function
2053prototypes, but still define all the structs and other symbols. This is
2054occasionally useful if you want to provide your own wrapper functions
2055around libev functions.</p>
2056 </dd>
2057 <dt>EV_MULTIPLICITY</dt>
2058 <dd>
2059 <p>If undefined or defined to <code>1</code>, then all event-loop-specific functions
2060will have the <code>struct ev_loop *</code> as first argument, and you can create
2061additional independent event loops. Otherwise there will be no support
2062for multiple event loops and there is no first event loop pointer
2063argument. Instead, all functions act on the single default loop.</p>
2064 </dd>
2065 <dt>EV_PERIODIC_ENABLE</dt>
2066 <dd>
2067 <p>If undefined or defined to be <code>1</code>, then periodic timers are supported. If
2068defined to be <code>0</code>, then they are not. Disabling them saves a few kB of
2069code.</p>
2070 </dd>
2071 <dt>EV_EMBED_ENABLE</dt>
2072 <dd>
2073 <p>If undefined or defined to be <code>1</code>, then embed watchers are supported. If
2074defined to be <code>0</code>, then they are not.</p>
2075 </dd>
2076 <dt>EV_STAT_ENABLE</dt>
2077 <dd>
2078 <p>If undefined or defined to be <code>1</code>, then stat watchers are supported. If
2079defined to be <code>0</code>, then they are not.</p>
2080 </dd>
2081 <dt>EV_FORK_ENABLE</dt>
2082 <dd>
2083 <p>If undefined or defined to be <code>1</code>, then fork watchers are supported. If
2084defined to be <code>0</code>, then they are not.</p>
2085 </dd>
2086 <dt>EV_MINIMAL</dt>
2087 <dd>
2088 <p>If you need to shave off some kilobytes of code at the expense of some
2089speed, define this symbol to <code>1</code>. Currently only used for gcc to override
2090some inlining decisions, saves roughly 30% codesize of amd64.</p>
2091 </dd>
2092 <dt>EV_PID_HASHSIZE</dt>
2093 <dd>
2094 <p><code>ev_child</code> watchers use a small hash table to distribute workload by
2095pid. The default size is <code>16</code> (or <code>1</code> with <code>EV_MINIMAL</code>), usually more
2096than enough. If you need to manage thousands of children you might want to
2097increase this value (<i>must</i> be a power of two).</p>
2098 </dd>
2099 <dt>EV_INOTIFY_HASHSIZE</dt>
2100 <dd>
2101 <p><code>ev_staz</code> watchers use a small hash table to distribute workload by
2102inotify watch id. The default size is <code>16</code> (or <code>1</code> with <code>EV_MINIMAL</code>),
2103usually more than enough. If you need to manage thousands of <code>ev_stat</code>
2104watchers you might want to increase this value (<i>must</i> be a power of
2105two).</p>
2106 </dd>
2107 <dt>EV_COMMON</dt>
2108 <dd>
2109 <p>By default, all watchers have a <code>void *data</code> member. By redefining
2110this macro to a something else you can include more and other types of
2111members. You have to define it each time you include one of the files,
2112though, and it must be identical each time.</p>
2113 <p>For example, the perl EV module uses something like this:</p>
2114<pre> #define EV_COMMON \
2115 SV *self; /* contains this struct */ \
2116 SV *cb_sv, *fh /* note no trailing &quot;;&quot; */
2117
2118</pre>
2119 </dd>
2120 <dt>EV_CB_DECLARE (type)</dt>
2121 <dt>EV_CB_INVOKE (watcher, revents)</dt>
2122 <dt>ev_set_cb (ev, cb)</dt>
2123 <dd>
2124 <p>Can be used to change the callback member declaration in each watcher,
2125and the way callbacks are invoked and set. Must expand to a struct member
2126definition and a statement, respectively. See the <cite>ev.v</cite> header file for
2127their default definitions. One possible use for overriding these is to
2128avoid the <code>struct ev_loop *</code> as first argument in all cases, or to use
2129method calls instead of plain function calls in C++.</p>
2130
2131</div>
2132<h2 id="EXAMPLES">EXAMPLES</h2>
2133<div id="EXAMPLES_CONTENT">
2134 <p>For a real-world example of a program the includes libev
2135verbatim, you can have a look at the EV perl module
2136(<a href="http://software.schmorp.de/pkg/EV.html">http://software.schmorp.de/pkg/EV.html</a>). It has the libev files in
2137the <cite>libev/</cite> subdirectory and includes them in the <cite>EV/EVAPI.h</cite> (public
2138interface) and <cite>EV.xs</cite> (implementation) files. Only the <cite>EV.xs</cite> file
2139will be compiled. It is pretty complex because it provides its own header
2140file.</p>
2141 <p>The usage in rxvt-unicode is simpler. It has a <cite>ev_cpp.h</cite> header file
2142that everybody includes and which overrides some autoconf choices:</p>
2143<pre> #define EV_USE_POLL 0
2144 #define EV_MULTIPLICITY 0
2145 #define EV_PERIODICS 0
2146 #define EV_CONFIG_H &lt;config.h&gt;
2147
2148 #include &quot;ev++.h&quot;
2149
2150</pre>
2151 <p>And a <cite>ev_cpp.C</cite> implementation file that contains libev proper and is compiled:</p>
2152<pre> #include &quot;ev_cpp.h&quot;
2153 #include &quot;ev.c&quot;
2154
2155
2156
2157
2158</pre>
2159
2160</div>
2161<h1 id="COMPLEXITIES">COMPLEXITIES</h1>
2162<div id="COMPLEXITIES_CONTENT">
2163 <p>In this section the complexities of (many of) the algorithms used inside
2164libev will be explained. For complexity discussions about backends see the
2165documentation for <code>ev_default_init</code>.</p>
2166 <p>
2167 <dl>
2168 <dt>Starting and stopping timer/periodic watchers: O(log skipped_other_timers)</dt>
2169 <dt>Changing timer/periodic watchers (by autorepeat, again): O(log skipped_other_timers)</dt>
2170 <dt>Starting io/check/prepare/idle/signal/child watchers: O(1)</dt>
2171 <dt>Stopping check/prepare/idle watchers: O(1)</dt>
2172 <dt>Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % EV_PID_HASHSIZE))</dt>
2173 <dt>Finding the next timer per loop iteration: O(1)</dt>
2174 <dt>Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)</dt>
2175 <dt>Activating one watcher: O(1)</dt>
2176 </dl>
2177 </p>
2178
2179
2180
2181
2182
2183</div>
2184<h1 id="AUTHOR">AUTHOR</h1>
719<div id="AUTHOR_CONTENT"> 2185<div id="AUTHOR_CONTENT">
720<p>Marc Lehmann &lt;libev@schmorp.de&gt;.</p> 2186 <p>Marc Lehmann &lt;libev@schmorp.de&gt;.</p>
721 2187
722</div> 2188</div>
723</div></body> 2189</div></body>
724</html> 2190</html>

Diff Legend

Removed lines
+ Added lines
< Changed lines
> Changed lines