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

Comparing libev/ev.html (file contents):
Revision 1.72 by root, Sat Dec 8 15:30:29 2007 UTC vs.
Revision 1.77 by root, Wed Dec 12 04:53:58 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="Sat Dec 8 16:30:24 2007" /> 9 <meta name="created" content="Wed Dec 12 05:53:55 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 -->
… …
27<ul><li><a href="#GENERIC_WATCHER_FUNCTIONS">GENERIC WATCHER FUNCTIONS</a></li> 27<ul><li><a href="#GENERIC_WATCHER_FUNCTIONS">GENERIC WATCHER FUNCTIONS</a></li>
28<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>
29</ul> 29</ul>
30</li> 30</li>
31<li><a href="#WATCHER_TYPES">WATCHER TYPES</a> 31<li><a href="#WATCHER_TYPES">WATCHER TYPES</a>
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> 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>
33<ul><li><a href="#The_special_problem_of_disappearing_">The special problem of disappearing file descriptors</a></li>
34</ul>
35</li>
33<li><a href="#code_ev_timer_code_relative_and_opti"><code>ev_timer</code> - relative and optionally repeating timeouts</a></li> 36<li><a href="#code_ev_timer_code_relative_and_opti"><code>ev_timer</code> - relative and optionally repeating timeouts</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> 37<li><a href="#code_ev_periodic_code_to_cron_or_not"><code>ev_periodic</code> - to cron or not to cron?</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> 38<li><a href="#code_ev_signal_code_signal_me_when_a"><code>ev_signal</code> - signal me when a signal gets signalled!</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> 39<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> 40<li><a href="#code_ev_stat_code_did_the_file_attri"><code>ev_stat</code> - did the file attributes just change?</a></li>
… …
185you actually want to know.</p> 188you actually want to know.</p>
186 </dd> 189 </dd>
187 <dt>int ev_version_major ()</dt> 190 <dt>int ev_version_major ()</dt>
188 <dt>int ev_version_minor ()</dt> 191 <dt>int ev_version_minor ()</dt>
189 <dd> 192 <dd>
190 <p>You can find out the major and minor version numbers of the library 193 <p>You can find out the major and minor ABI version numbers of the library
191you linked against by calling the functions <code>ev_version_major</code> and 194you linked against by calling the functions <code>ev_version_major</code> and
192<code>ev_version_minor</code>. If you want, you can compare against the global 195<code>ev_version_minor</code>. If you want, you can compare against the global
193symbols <code>EV_VERSION_MAJOR</code> and <code>EV_VERSION_MINOR</code>, which specify the 196symbols <code>EV_VERSION_MAJOR</code> and <code>EV_VERSION_MINOR</code>, which specify the
194version of the library your program was compiled against.</p> 197version of the library your program was compiled against.</p>
198 <p>These version numbers refer to the ABI version of the library, not the
199release version.</p>
195 <p>Usually, it's a good idea to terminate if the major versions mismatch, 200 <p>Usually, it's a good idea to terminate if the major versions mismatch,
196as this indicates an incompatible change. Minor versions are usually 201as this indicates an incompatible change. Minor versions are usually
197compatible to older versions, so a larger minor version alone is usually 202compatible to older versions, so a larger minor version alone is usually
198not a problem.</p> 203not a problem.</p>
199 <p>Example: Make sure we haven't accidentally been linked against the wrong 204 <p>Example: Make sure we haven't accidentally been linked against the wrong
200version.</p> 205version.</p>
201<pre> assert ((&quot;libev version mismatch&quot;, 206<pre> assert ((&quot;libev version mismatch&quot;,
… …
529one iteration of the loop. This is useful if you are waiting for some 534one iteration of the loop. This is useful if you are waiting for some
530external event in conjunction with something not expressible using other 535external event in conjunction with something not expressible using other
531libev watchers. However, a pair of <code>ev_prepare</code>/<code>ev_check</code> watchers is 536libev watchers. However, a pair of <code>ev_prepare</code>/<code>ev_check</code> watchers is
532usually a better approach for this kind of thing.</p> 537usually a better approach for this kind of thing.</p>
533 <p>Here are the gory details of what <code>ev_loop</code> does:</p> 538 <p>Here are the gory details of what <code>ev_loop</code> does:</p>
539<pre> - Before the first iteration, call any pending watchers.
534<pre> * If there are no active watchers (reference count is zero), return. 540 * If there are no active watchers (reference count is zero), return.
535 - Queue prepare watchers and then call all outstanding watchers. 541 - Queue all prepare watchers and then call all outstanding watchers.
536 - If we have been forked, recreate the kernel state. 542 - If we have been forked, recreate the kernel state.
537 - Update the kernel state with all outstanding changes. 543 - Update the kernel state with all outstanding changes.
538 - Update the &quot;event loop time&quot;. 544 - Update the &quot;event loop time&quot;.
539 - Calculate for how long to block. 545 - Calculate for how long to block.
540 - Block the process, waiting for any events. 546 - Block the process, waiting for any events.
… …
938<p>If you cannot run the fd in non-blocking mode (for example you should not 944<p>If you cannot run the fd in non-blocking mode (for example you should not
939play around with an Xlib connection), then you have to seperately re-test 945play around with an Xlib connection), then you have to seperately re-test
940whether a file descriptor is really ready with a known-to-be good interface 946whether a file descriptor is really ready with a known-to-be good interface
941such as poll (fortunately in our Xlib example, Xlib already does this on 947such as poll (fortunately in our Xlib example, Xlib already does this on
942its own, so its quite safe to use).</p> 948its own, so its quite safe to use).</p>
949
950</div>
951<h3 id="The_special_problem_of_disappearing_">The special problem of disappearing file descriptors</h3>
952<div id="The_special_problem_of_disappearing_-2">
953<p>Some backends (e.g kqueue, epoll) need to be told about closing a file
954descriptor (either by calling <code>close</code> explicitly or by any other means,
955such as <code>dup</code>). The reason is that you register interest in some file
956descriptor, but when it goes away, the operating system will silently drop
957this interest. If another file descriptor with the same number then is
958registered with libev, there is no efficient way to see that this is, in
959fact, a different file descriptor.</p>
960<p>To avoid having to explicitly tell libev about such cases, libev follows
961the following policy: Each time <code>ev_io_set</code> is being called, libev
962will assume that this is potentially a new file descriptor, otherwise
963it is assumed that the file descriptor stays the same. That means that
964you <i>have</i> to call <code>ev_io_set</code> (or <code>ev_io_init</code>) when you change the
965descriptor even if the file descriptor number itself did not change.</p>
966<p>This is how one would do it normally anyway, the important point is that
967the libev application should not optimise around libev but should leave
968optimisations to libev.</p>
969
970
971
972
943<dl> 973<dl>
944 <dt>ev_io_init (ev_io *, callback, int fd, int events)</dt> 974 <dt>ev_io_init (ev_io *, callback, int fd, int events)</dt>
945 <dt>ev_io_set (ev_io *, int fd, int events)</dt> 975 <dt>ev_io_set (ev_io *, int fd, int events)</dt>
946 <dd> 976 <dd>
947 <p>Configures an <code>ev_io</code> watcher. The <code>fd</code> is the file descriptor to 977 <p>Configures an <code>ev_io</code> watcher. The <code>fd</code> is the file descriptor to
… …
1096but on wallclock time (absolute time). You can tell a periodic watcher 1126but on wallclock time (absolute time). You can tell a periodic watcher
1097to trigger &quot;at&quot; some specific point in time. For example, if you tell a 1127to trigger &quot;at&quot; some specific point in time. For example, if you tell a
1098periodic watcher to trigger in 10 seconds (by specifiying e.g. <code>ev_now () 1128periodic watcher to trigger in 10 seconds (by specifiying e.g. <code>ev_now ()
1099+ 10.</code>) and then reset your system clock to the last year, then it will 1129+ 10.</code>) and then reset your system clock to the last year, then it will
1100take a year to trigger the event (unlike an <code>ev_timer</code>, which would trigger 1130take a year to trigger the event (unlike an <code>ev_timer</code>, which would trigger
1101roughly 10 seconds later and of course not if you reset your system time 1131roughly 10 seconds later).</p>
1102again).</p>
1103<p>They can also be used to implement vastly more complex timers, such as 1132<p>They can also be used to implement vastly more complex timers, such as
1104triggering an event on eahc midnight, local time.</p> 1133triggering an event on each midnight, local time or other, complicated,
1134rules.</p>
1105<p>As with timers, the callback is guarenteed to be invoked only when the 1135<p>As with timers, the callback is guarenteed to be invoked only when the
1106time (<code>at</code>) has been passed, but if multiple periodic timers become ready 1136time (<code>at</code>) has been passed, but if multiple periodic timers become ready
1107during the same loop iteration then order of execution is undefined.</p> 1137during the same loop iteration then order of execution is undefined.</p>
1108<dl> 1138<dl>
1109 <dt>ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)</dt> 1139 <dt>ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)</dt>
… …
1111 <dd> 1141 <dd>
1112 <p>Lots of arguments, lets sort it out... There are basically three modes of 1142 <p>Lots of arguments, lets sort it out... There are basically three modes of
1113operation, and we will explain them from simplest to complex:</p> 1143operation, and we will explain them from simplest to complex:</p>
1114 <p> 1144 <p>
1115 <dl> 1145 <dl>
1116 <dt>* absolute timer (interval = reschedule_cb = 0)</dt> 1146 <dt>* absolute timer (at = time, interval = reschedule_cb = 0)</dt>
1117 <dd> 1147 <dd>
1118 <p>In this configuration the watcher triggers an event at the wallclock time 1148 <p>In this configuration the watcher triggers an event at the wallclock time
1119<code>at</code> and doesn't repeat. It will not adjust when a time jump occurs, 1149<code>at</code> and doesn't repeat. It will not adjust when a time jump occurs,
1120that is, if it is to be run at January 1st 2011 then it will run when the 1150that is, if it is to be run at January 1st 2011 then it will run when the
1121system time reaches or surpasses this time.</p> 1151system time reaches or surpasses this time.</p>
1122 </dd> 1152 </dd>
1123 <dt>* non-repeating interval timer (interval &gt; 0, reschedule_cb = 0)</dt> 1153 <dt>* non-repeating interval timer (at = offset, interval &gt; 0, reschedule_cb = 0)</dt>
1124 <dd> 1154 <dd>
1125 <p>In this mode the watcher will always be scheduled to time out at the next 1155 <p>In this mode the watcher will always be scheduled to time out at the next
1126<code>at + N * interval</code> time (for some integer N) and then repeat, regardless 1156<code>at + N * interval</code> time (for some integer N, which can also be negative)
1127of any time jumps.</p> 1157and then repeat, regardless of any time jumps.</p>
1128 <p>This can be used to create timers that do not drift with respect to system 1158 <p>This can be used to create timers that do not drift with respect to system
1129time:</p> 1159time:</p>
1130<pre> ev_periodic_set (&amp;periodic, 0., 3600., 0); 1160<pre> ev_periodic_set (&amp;periodic, 0., 3600., 0);
1131 1161
1132</pre> 1162</pre>
… …
1135full hour (UTC), or more correctly, when the system time is evenly divisible 1165full hour (UTC), or more correctly, when the system time is evenly divisible
1136by 3600.</p> 1166by 3600.</p>
1137 <p>Another way to think about it (for the mathematically inclined) is that 1167 <p>Another way to think about it (for the mathematically inclined) is that
1138<code>ev_periodic</code> will try to run the callback in this mode at the next possible 1168<code>ev_periodic</code> will try to run the callback in this mode at the next possible
1139time where <code>time = at (mod interval)</code>, regardless of any time jumps.</p> 1169time where <code>time = at (mod interval)</code>, regardless of any time jumps.</p>
1170 <p>For numerical stability it is preferable that the <code>at</code> value is near
1171<code>ev_now ()</code> (the current time), but there is no range requirement for
1172this value.</p>
1140 </dd> 1173 </dd>
1141 <dt>* manual reschedule mode (reschedule_cb = callback)</dt> 1174 <dt>* manual reschedule mode (at and interval ignored, reschedule_cb = callback)</dt>
1142 <dd> 1175 <dd>
1143 <p>In this mode the values for <code>interval</code> and <code>at</code> are both being 1176 <p>In this mode the values for <code>interval</code> and <code>at</code> are both being
1144ignored. Instead, each time the periodic watcher gets scheduled, the 1177ignored. Instead, each time the periodic watcher gets scheduled, the
1145reschedule callback will be called with the watcher as first, and the 1178reschedule callback will be called with the watcher as first, and the
1146current time as second argument.</p> 1179current time as second argument.</p>
1147 <p>NOTE: <i>This callback MUST NOT stop or destroy any periodic watcher, 1180 <p>NOTE: <i>This callback MUST NOT stop or destroy any periodic watcher,
1148ever, or make any event loop modifications</i>. If you need to stop it, 1181ever, or make any event loop modifications</i>. If you need to stop it,
1149return <code>now + 1e30</code> (or so, fudge fudge) and stop it afterwards (e.g. by 1182return <code>now + 1e30</code> (or so, fudge fudge) and stop it afterwards (e.g. by
1150starting a prepare watcher).</p> 1183starting an <code>ev_prepare</code> watcher, which is legal).</p>
1151 <p>Its prototype is <code>ev_tstamp (*reschedule_cb)(struct ev_periodic *w, 1184 <p>Its prototype is <code>ev_tstamp (*reschedule_cb)(struct ev_periodic *w,
1152ev_tstamp now)</code>, e.g.:</p> 1185ev_tstamp now)</code>, e.g.:</p>
1153<pre> static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now) 1186<pre> static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now)
1154 { 1187 {
1155 return now + 60.; 1188 return now + 60.;
… …
1175 <dd> 1208 <dd>
1176 <p>Simply stops and restarts the periodic watcher again. This is only useful 1209 <p>Simply stops and restarts the periodic watcher again. This is only useful
1177when you changed some parameters or the reschedule callback would return 1210when you changed some parameters or the reschedule callback would return
1178a different time than the last time it was called (e.g. in a crond like 1211a different time than the last time it was called (e.g. in a crond like
1179program when the crontabs have changed).</p> 1212program when the crontabs have changed).</p>
1213 </dd>
1214 <dt>ev_tstamp offset [read-write]</dt>
1215 <dd>
1216 <p>When repeating, this contains the offset value, otherwise this is the
1217absolute point in time (the <code>at</code> value passed to <code>ev_periodic_set</code>).</p>
1218 <p>Can be modified any time, but changes only take effect when the periodic
1219timer fires or <code>ev_periodic_again</code> is being called.</p>
1180 </dd> 1220 </dd>
1181 <dt>ev_tstamp interval [read-write]</dt> 1221 <dt>ev_tstamp interval [read-write]</dt>
1182 <dd> 1222 <dd>
1183 <p>The current interval value. Can be modified any time, but changes only 1223 <p>The current interval value. Can be modified any time, but changes only
1184take effect when the periodic timer fires or <code>ev_periodic_again</code> is being 1224take effect when the periodic timer fires or <code>ev_periodic_again</code> is being
… …
1483are ready to run (it's actually more complicated: it only runs coroutines 1523are ready to run (it's actually more complicated: it only runs coroutines
1484with priority higher than or equal to the event loop and one coroutine 1524with priority higher than or equal to the event loop and one coroutine
1485of lower priority, but only once, using idle watchers to keep the event 1525of lower priority, but only once, using idle watchers to keep the event
1486loop from blocking if lower-priority coroutines are active, thus mapping 1526loop from blocking if lower-priority coroutines are active, thus mapping
1487low-priority coroutines to idle/background tasks).</p> 1527low-priority coroutines to idle/background tasks).</p>
1528<p>It is recommended to give <code>ev_check</code> watchers highest (<code>EV_MAXPRI</code>)
1529priority, to ensure that they are being run before any other watchers
1530after the poll. Also, <code>ev_check</code> watchers (and <code>ev_prepare</code> watchers,
1531too) should not activate (&quot;feed&quot;) events into libev. While libev fully
1532supports this, they will be called before other <code>ev_check</code> watchers did
1533their job. As <code>ev_check</code> watchers are often used to embed other event
1534loops those other event loops might be in an unusable state until their
1535<code>ev_check</code> watcher ran (always remind yourself to coexist peacefully with
1536others).</p>
1488<dl> 1537<dl>
1489 <dt>ev_prepare_init (ev_prepare *, callback)</dt> 1538 <dt>ev_prepare_init (ev_prepare *, callback)</dt>
1490 <dt>ev_check_init (ev_check *, callback)</dt> 1539 <dt>ev_check_init (ev_check *, callback)</dt>
1491 <dd> 1540 <dd>
1492 <p>Initialises and configures the prepare or check watcher - they have no 1541 <p>Initialises and configures the prepare or check watcher - they have no

Diff Legend

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