1 .\" Automatically generated by Pod::Man v1.37, Pod::Parser v1.35
4 .\" ========================================================================
5 .de Sh \" Subsection heading
13 .de Sp \" Vertical space (when we can't use .PP)
17 .de Vb \" Begin verbatim text
22 .de Ve \" End verbatim text
26 .\" Set up some character translations and predefined strings. \*(-- will
27 .\" give an unbreakable dash, \*(PI will give pi, \*(L" will give a left
28 .\" double quote, and \*(R" will give a right double quote. | will give a
29 .\" real vertical bar. \*(C+ will give a nicer C++. Capital omega is used to
30 .\" do unbreakable dashes and therefore won't be available. \*(C` and \*(C'
31 .\" expand to `' in nroff, nothing in troff, for use with C<>.
33 .ds C+ C\v'-.1v'\h'-1p'\s-2+\h'-1p'+\s0\v'.1v'\h'-1p'
37 . if (\n(.H=4u)&(1m=24u) .ds -- \(*W\h'-12u'\(*W\h'-12u'-\" diablo 10 pitch
38 . if (\n(.H=4u)&(1m=20u) .ds -- \(*W\h'-12u'\(*W\h'-8u'-\" diablo 12 pitch
51 .\" If the F register is turned on, we'll generate index entries on stderr for
52 .\" titles (.TH), headers (.SH), subsections (.Sh), items (.Ip), and index
53 .\" entries marked with X<> in POD. Of course, you'll have to process the
54 .\" output yourself in some meaningful fashion.
57 . tm Index:\\$1\t\\n%\t"\\$2"
63 .\" For nroff, turn off justification. Always turn off hyphenation; it makes
64 .\" way too many mistakes in technical documents.
68 .\" Accent mark definitions (@(#)ms.acc 1.5 88/02/08 SMI; from UCB 4.2).
69 .\" Fear. Run. Save yourself. No user-serviceable parts.
70 . \" fudge factors for nroff and troff
79 . ds #H ((1u-(\\\\n(.fu%2u))*.13m)
85 . \" simple accents for nroff and troff
95 . ds ' \\k:\h'-(\\n(.wu*8/10-\*(#H)'\'\h"|\\n:u"
96 . ds ` \\k:\h'-(\\n(.wu*8/10-\*(#H)'\`\h'|\\n:u'
97 . ds ^ \\k:\h'-(\\n(.wu*10/11-\*(#H)'^\h'|\\n:u'
98 . ds , \\k:\h'-(\\n(.wu*8/10)',\h'|\\n:u'
99 . ds ~ \\k:\h'-(\\n(.wu-\*(#H-.1m)'~\h'|\\n:u'
100 . ds / \\k:\h'-(\\n(.wu*8/10-\*(#H)'\z\(sl\h'|\\n:u'
102 . \" troff and (daisy-wheel) nroff accents
103 .ds : \\k:\h'-(\\n(.wu*8/10-\*(#H+.1m+\*(#F)'\v'-\*(#V'\z.\h'.2m+\*(#F'.\h'|\\n:u'\v'\*(#V'
104 .ds 8 \h'\*(#H'\(*b\h'-\*(#H'
105 .ds o \\k:\h'-(\\n(.wu+\w'\(de'u-\*(#H)/2u'\v'-.3n'\*(#[\z\(de\v'.3n'\h'|\\n:u'\*(#]
106 .ds d- \h'\*(#H'\(pd\h'-\w'~'u'\v'-.25m'\f2\(hy\fP\v'.25m'\h'-\*(#H'
107 .ds D- D\\k:\h'-\w'D'u'\v'-.11m'\z\(hy\v'.11m'\h'|\\n:u'
108 .ds th \*(#[\v'.3m'\s+1I\s-1\v'-.3m'\h'-(\w'I'u*2/3)'\s-1o\s+1\*(#]
109 .ds Th \*(#[\s+2I\s-2\h'-\w'I'u*3/5'\v'-.3m'o\v'.3m'\*(#]
110 .ds ae a\h'-(\w'a'u*4/10)'e
111 .ds Ae A\h'-(\w'A'u*4/10)'E
112 . \" corrections for vroff
113 .if v .ds ~ \\k:\h'-(\\n(.wu*9/10-\*(#H)'\s-2\u~\d\s+2\h'|\\n:u'
114 .if v .ds ^ \\k:\h'-(\\n(.wu*10/11-\*(#H)'\v'-.4m'^\v'.4m'\h'|\\n:u'
115 . \" for low resolution devices (crt and lpr)
116 .if \n(.H>23 .if \n(.V>19 \
129 .\" ========================================================================
131 .IX Title ""<STANDARD INPUT>" 1"
132 .TH "<STANDARD INPUT>" 1 "2007-11-27" "perl v5.8.8" "User Contributed Perl Documentation"
134 libev \- a high performance full\-featured event loop written in C
136 .IX Header "SYNOPSIS"
141 .IX Header "DESCRIPTION"
142 Libev is an event loop: you register interest in certain events (such as a
143 file descriptor being readable or a timeout occuring), and it will manage
144 these event sources and provide your program with events.
146 To do this, it must take more or less complete control over your process
147 (or thread) by executing the \fIevent loop\fR handler, and will then
148 communicate events via a callback mechanism.
150 You register interest in certain events by registering so-called \fIevent
151 watchers\fR, which are relatively small C structures you initialise with the
152 details of the event, and then hand it over to libev by \fIstarting\fR the
155 .IX Header "FEATURES"
156 Libev supports select, poll, the linux-specific epoll and the bsd-specific
157 kqueue mechanisms for file descriptor events, relative timers, absolute
158 timers with customised rescheduling, signal events, process status change
159 events (related to \s-1SIGCHLD\s0), and event watchers dealing with the event
160 loop mechanism itself (idle, prepare and check watchers). It also is quite
161 fast (see this benchmark comparing
162 it to libevent for example).
164 .IX Header "CONVENTIONS"
165 Libev is very configurable. In this manual the default configuration
166 will be described, which supports multiple event loops. For more info
167 about various configuration options please have a look at the file
168 \&\fI\s-1README\s0.embed\fR in the libev distribution. If libev was configured without
169 support for multiple event loops, then all functions taking an initial
170 argument of name \f(CW\*(C`loop\*(C'\fR (which is always of type \f(CW\*(C`struct ev_loop *\*(C'\fR)
171 will not have this argument.
172 .SH "TIME REPRESENTATION"
173 .IX Header "TIME REPRESENTATION"
174 Libev represents time as a single floating point number, representing the
175 (fractional) number of seconds since the (\s-1POSIX\s0) epoch (somewhere near
176 the beginning of 1970, details are complicated, don't ask). This type is
177 called \f(CW\*(C`ev_tstamp\*(C'\fR, which is what you should use too. It usually aliases
178 to the \f(CW\*(C`double\*(C'\fR type in C, and when you need to do any calculations on
179 it, you should treat it as such.
180 .SH "GLOBAL FUNCTIONS"
181 .IX Header "GLOBAL FUNCTIONS"
182 These functions can be called anytime, even before initialising the
184 .IP "ev_tstamp ev_time ()" 4
185 .IX Item "ev_tstamp ev_time ()"
186 Returns the current time as libev would use it. Please note that the
187 \&\f(CW\*(C`ev_now\*(C'\fR function is usually faster and also often returns the timestamp
188 you actually want to know.
189 .IP "int ev_version_major ()" 4
190 .IX Item "int ev_version_major ()"
192 .IP "int ev_version_minor ()" 4
193 .IX Item "int ev_version_minor ()"
195 You can find out the major and minor version numbers of the library
196 you linked against by calling the functions \f(CW\*(C`ev_version_major\*(C'\fR and
197 \&\f(CW\*(C`ev_version_minor\*(C'\fR. If you want, you can compare against the global
198 symbols \f(CW\*(C`EV_VERSION_MAJOR\*(C'\fR and \f(CW\*(C`EV_VERSION_MINOR\*(C'\fR, which specify the
199 version of the library your program was compiled against.
201 Usually, it's a good idea to terminate if the major versions mismatch,
202 as this indicates an incompatible change. Minor versions are usually
203 compatible to older versions, so a larger minor version alone is usually
206 Example: make sure we haven't accidentally been linked against the wrong
210 \& assert (("libev version mismatch",
211 \& ev_version_major () == EV_VERSION_MAJOR
212 \& && ev_version_minor () >= EV_VERSION_MINOR));
214 .IP "unsigned int ev_supported_backends ()" 4
215 .IX Item "unsigned int ev_supported_backends ()"
216 Return the set of all backends (i.e. their corresponding \f(CW\*(C`EV_BACKEND_*\*(C'\fR
217 value) compiled into this binary of libev (independent of their
218 availability on the system you are running on). See \f(CW\*(C`ev_default_loop\*(C'\fR for
219 a description of the set values.
221 Example: make sure we have the epoll method, because yeah this is cool and
222 a must have and can we have a torrent of it please!!!11
225 \& assert (("sorry, no epoll, no sex",
226 \& ev_supported_backends () & EVBACKEND_EPOLL));
228 .IP "unsigned int ev_recommended_backends ()" 4
229 .IX Item "unsigned int ev_recommended_backends ()"
230 Return the set of all backends compiled into this binary of libev and also
231 recommended for this platform. This set is often smaller than the one
232 returned by \f(CW\*(C`ev_supported_backends\*(C'\fR, as for example kqueue is broken on
233 most BSDs and will not be autodetected unless you explicitly request it
234 (assuming you know what you are doing). This is the set of backends that
235 libev will probe for if you specify no backends explicitly.
236 .IP "unsigned int ev_embeddable_backends ()" 4
237 .IX Item "unsigned int ev_embeddable_backends ()"
238 Returns the set of backends that are embeddable in other event loops. This
239 is the theoretical, all\-platform, value. To find which backends
240 might be supported on the current system, you would need to look at
241 \&\f(CW\*(C`ev_embeddable_backends () & ev_supported_backends ()\*(C'\fR, likewise for
244 See the description of \f(CW\*(C`ev_embed\*(C'\fR watchers for more info.
245 .IP "ev_set_allocator (void *(*cb)(void *ptr, long size))" 4
246 .IX Item "ev_set_allocator (void *(*cb)(void *ptr, long size))"
247 Sets the allocation function to use (the prototype is similar to the
248 realloc C function, the semantics are identical). It is used to allocate
249 and free memory (no surprises here). If it returns zero when memory
250 needs to be allocated, the library might abort or take some potentially
251 destructive action. The default is your system realloc function.
253 You could override this function in high-availability programs to, say,
254 free some memory if it cannot allocate memory, to use a special allocator,
255 or even to sleep a while and retry until some memory is available.
257 Example: replace the libev allocator with one that waits a bit and then
258 retries: better than mine).
262 \& persistent_realloc (void *ptr, long size)
266 \& void *newptr = realloc (ptr, size);
282 \& ev_set_allocator (persistent_realloc);
284 .IP "ev_set_syserr_cb (void (*cb)(const char *msg));" 4
285 .IX Item "ev_set_syserr_cb (void (*cb)(const char *msg));"
286 Set the callback function to call on a retryable syscall error (such
287 as failed select, poll, epoll_wait). The message is a printable string
288 indicating the system call or subsystem causing the problem. If this
289 callback is set, then libev will expect it to remedy the sitution, no
290 matter what, when it returns. That is, libev will generally retry the
291 requested operation, or, if the condition doesn't go away, do bad stuff
294 Example: do the same thing as libev does internally:
298 \& fatal_error (const char *msg)
307 \& ev_set_syserr_cb (fatal_error);
309 .SH "FUNCTIONS CONTROLLING THE EVENT LOOP"
310 .IX Header "FUNCTIONS CONTROLLING THE EVENT LOOP"
311 An event loop is described by a \f(CW\*(C`struct ev_loop *\*(C'\fR. The library knows two
312 types of such loops, the \fIdefault\fR loop, which supports signals and child
313 events, and dynamically created loops which do not.
315 If you use threads, a common model is to run the default event loop
316 in your main thread (or in a separate thread) and for each thread you
317 create, you also create another event loop. Libev itself does no locking
318 whatsoever, so if you mix calls to the same event loop in different
319 threads, make sure you lock (this is usually a bad idea, though, even if
320 done correctly, because it's hideous and inefficient).
321 .IP "struct ev_loop *ev_default_loop (unsigned int flags)" 4
322 .IX Item "struct ev_loop *ev_default_loop (unsigned int flags)"
323 This will initialise the default event loop if it hasn't been initialised
324 yet and return it. If the default loop could not be initialised, returns
325 false. If it already was initialised it simply returns it (and ignores the
326 flags. If that is troubling you, check \f(CW\*(C`ev_backend ()\*(C'\fR afterwards).
328 If you don't know what event loop to use, use the one returned from this
331 The flags argument can be used to specify special behaviour or specific
332 backends to use, and is usually specified as \f(CW0\fR (or \f(CW\*(C`EVFLAG_AUTO\*(C'\fR).
334 The following flags are supported:
336 .ie n .IP """EVFLAG_AUTO""" 4
337 .el .IP "\f(CWEVFLAG_AUTO\fR" 4
338 .IX Item "EVFLAG_AUTO"
339 The default flags value. Use this if you have no clue (it's the right
341 .ie n .IP """EVFLAG_NOENV""" 4
342 .el .IP "\f(CWEVFLAG_NOENV\fR" 4
343 .IX Item "EVFLAG_NOENV"
344 If this flag bit is ored into the flag value (or the program runs setuid
345 or setgid) then libev will \fInot\fR look at the environment variable
346 \&\f(CW\*(C`LIBEV_FLAGS\*(C'\fR. Otherwise (the default), this environment variable will
347 override the flags completely if it is found in the environment. This is
348 useful to try out specific backends to test their performance, or to work
350 .ie n .IP """EVBACKEND_SELECT"" (value 1, portable select backend)" 4
351 .el .IP "\f(CWEVBACKEND_SELECT\fR (value 1, portable select backend)" 4
352 .IX Item "EVBACKEND_SELECT (value 1, portable select backend)"
353 This is your standard \fIselect\fR\|(2) backend. Not \fIcompletely\fR standard, as
354 libev tries to roll its own fd_set with no limits on the number of fds,
355 but if that fails, expect a fairly low limit on the number of fds when
356 using this backend. It doesn't scale too well (O(highest_fd)), but its usually
357 the fastest backend for a low number of fds.
358 .ie n .IP """EVBACKEND_POLL"" (value 2, poll backend, available everywhere except on windows)" 4
359 .el .IP "\f(CWEVBACKEND_POLL\fR (value 2, poll backend, available everywhere except on windows)" 4
360 .IX Item "EVBACKEND_POLL (value 2, poll backend, available everywhere except on windows)"
361 And this is your standard \fIpoll\fR\|(2) backend. It's more complicated than
362 select, but handles sparse fds better and has no artificial limit on the
363 number of fds you can use (except it will slow down considerably with a
364 lot of inactive fds). It scales similarly to select, i.e. O(total_fds).
365 .ie n .IP """EVBACKEND_EPOLL"" (value 4, Linux)" 4
366 .el .IP "\f(CWEVBACKEND_EPOLL\fR (value 4, Linux)" 4
367 .IX Item "EVBACKEND_EPOLL (value 4, Linux)"
368 For few fds, this backend is a bit little slower than poll and select,
369 but it scales phenomenally better. While poll and select usually scale like
370 O(total_fds) where n is the total number of fds (or the highest fd), epoll scales
371 either O(1) or O(active_fds).
373 While stopping and starting an I/O watcher in the same iteration will
374 result in some caching, there is still a syscall per such incident
375 (because the fd could point to a different file description now), so its
376 best to avoid that. Also, \fIdup()\fRed file descriptors might not work very
377 well if you register events for both fds.
379 Please note that epoll sometimes generates spurious notifications, so you
380 need to use non-blocking I/O or other means to avoid blocking when no data
381 (or space) is available.
382 .ie n .IP """EVBACKEND_KQUEUE"" (value 8, most \s-1BSD\s0 clones)" 4
383 .el .IP "\f(CWEVBACKEND_KQUEUE\fR (value 8, most \s-1BSD\s0 clones)" 4
384 .IX Item "EVBACKEND_KQUEUE (value 8, most BSD clones)"
385 Kqueue deserves special mention, as at the time of this writing, it
386 was broken on all BSDs except NetBSD (usually it doesn't work with
387 anything but sockets and pipes, except on Darwin, where of course its
388 completely useless). For this reason its not being \*(L"autodetected\*(R"
389 unless you explicitly specify it explicitly in the flags (i.e. using
390 \&\f(CW\*(C`EVBACKEND_KQUEUE\*(C'\fR).
392 It scales in the same way as the epoll backend, but the interface to the
393 kernel is more efficient (which says nothing about its actual speed, of
394 course). While starting and stopping an I/O watcher does not cause an
395 extra syscall as with epoll, it still adds up to four event changes per
396 incident, so its best to avoid that.
397 .ie n .IP """EVBACKEND_DEVPOLL"" (value 16, Solaris 8)" 4
398 .el .IP "\f(CWEVBACKEND_DEVPOLL\fR (value 16, Solaris 8)" 4
399 .IX Item "EVBACKEND_DEVPOLL (value 16, Solaris 8)"
400 This is not implemented yet (and might never be).
401 .ie n .IP """EVBACKEND_PORT"" (value 32, Solaris 10)" 4
402 .el .IP "\f(CWEVBACKEND_PORT\fR (value 32, Solaris 10)" 4
403 .IX Item "EVBACKEND_PORT (value 32, Solaris 10)"
404 This uses the Solaris 10 port mechanism. As with everything on Solaris,
405 it's really slow, but it still scales very well (O(active_fds)).
407 Please note that solaris ports can result in a lot of spurious
408 notifications, so you need to use non-blocking I/O or other means to avoid
409 blocking when no data (or space) is available.
410 .ie n .IP """EVBACKEND_ALL""" 4
411 .el .IP "\f(CWEVBACKEND_ALL\fR" 4
412 .IX Item "EVBACKEND_ALL"
413 Try all backends (even potentially broken ones that wouldn't be tried
414 with \f(CW\*(C`EVFLAG_AUTO\*(C'\fR). Since this is a mask, you can do stuff such as
415 \&\f(CW\*(C`EVBACKEND_ALL & ~EVBACKEND_KQUEUE\*(C'\fR.
419 If one or more of these are ored into the flags value, then only these
420 backends will be tried (in the reverse order as given here). If none are
421 specified, most compiled-in backend will be tried, usually in reverse
422 order of their flag values :)
424 The most typical usage is like this:
427 \& if (!ev_default_loop (0))
428 \& fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
431 Restrict libev to the select and poll backends, and do not allow
432 environment settings to be taken into account:
435 \& ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);
438 Use whatever libev has to offer, but make sure that kqueue is used if
439 available (warning, breaks stuff, best use only with your own private
440 event loop and only if you know the \s-1OS\s0 supports your types of fds):
443 \& ev_default_loop (ev_recommended_backends () | EVBACKEND_KQUEUE);
446 .IP "struct ev_loop *ev_loop_new (unsigned int flags)" 4
447 .IX Item "struct ev_loop *ev_loop_new (unsigned int flags)"
448 Similar to \f(CW\*(C`ev_default_loop\*(C'\fR, but always creates a new event loop that is
449 always distinct from the default loop. Unlike the default loop, it cannot
450 handle signal and child watchers, and attempts to do so will be greeted by
451 undefined behaviour (or a failed assertion if assertions are enabled).
453 Example: try to create a event loop that uses epoll and nothing else.
456 \& struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV);
458 \& fatal ("no epoll found here, maybe it hides under your chair");
460 .IP "ev_default_destroy ()" 4
461 .IX Item "ev_default_destroy ()"
462 Destroys the default loop again (frees all memory and kernel state
463 etc.). None of the active event watchers will be stopped in the normal
464 sense, so e.g. \f(CW\*(C`ev_is_active\*(C'\fR might still return true. It is your
465 responsibility to either stop all watchers cleanly yoursef \fIbefore\fR
466 calling this function, or cope with the fact afterwards (which is usually
467 the easiest thing, youc na just ignore the watchers and/or \f(CW\*(C`free ()\*(C'\fR them
469 .IP "ev_loop_destroy (loop)" 4
470 .IX Item "ev_loop_destroy (loop)"
471 Like \f(CW\*(C`ev_default_destroy\*(C'\fR, but destroys an event loop created by an
472 earlier call to \f(CW\*(C`ev_loop_new\*(C'\fR.
473 .IP "ev_default_fork ()" 4
474 .IX Item "ev_default_fork ()"
475 This function reinitialises the kernel state for backends that have
476 one. Despite the name, you can call it anytime, but it makes most sense
477 after forking, in either the parent or child process (or both, but that
478 again makes little sense).
480 You \fImust\fR call this function in the child process after forking if and
481 only if you want to use the event library in both processes. If you just
482 fork+exec, you don't have to call it.
484 The function itself is quite fast and it's usually not a problem to call
485 it just in case after a fork. To make this easy, the function will fit in
486 quite nicely into a call to \f(CW\*(C`pthread_atfork\*(C'\fR:
489 \& pthread_atfork (0, 0, ev_default_fork);
492 At the moment, \f(CW\*(C`EVBACKEND_SELECT\*(C'\fR and \f(CW\*(C`EVBACKEND_POLL\*(C'\fR are safe to use
493 without calling this function, so if you force one of those backends you
495 .IP "ev_loop_fork (loop)" 4
496 .IX Item "ev_loop_fork (loop)"
497 Like \f(CW\*(C`ev_default_fork\*(C'\fR, but acts on an event loop created by
498 \&\f(CW\*(C`ev_loop_new\*(C'\fR. Yes, you have to call this on every allocated event loop
499 after fork, and how you do this is entirely your own problem.
500 .IP "unsigned int ev_backend (loop)" 4
501 .IX Item "unsigned int ev_backend (loop)"
502 Returns one of the \f(CW\*(C`EVBACKEND_*\*(C'\fR flags indicating the event backend in
504 .IP "ev_tstamp ev_now (loop)" 4
505 .IX Item "ev_tstamp ev_now (loop)"
506 Returns the current \*(L"event loop time\*(R", which is the time the event loop
507 received events and started processing them. This timestamp does not
508 change as long as callbacks are being processed, and this is also the base
509 time used for relative timers. You can treat it as the timestamp of the
510 event occuring (or more correctly, libev finding out about it).
511 .IP "ev_loop (loop, int flags)" 4
512 .IX Item "ev_loop (loop, int flags)"
513 Finally, this is it, the event handler. This function usually is called
514 after you initialised all your watchers and you want to start handling
517 If the flags argument is specified as \f(CW0\fR, it will not return until
518 either no event watchers are active anymore or \f(CW\*(C`ev_unloop\*(C'\fR was called.
520 Please note that an explicit \f(CW\*(C`ev_unloop\*(C'\fR is usually better than
521 relying on all watchers to be stopped when deciding when a program has
522 finished (especially in interactive programs), but having a program that
523 automatically loops as long as it has to and no longer by virtue of
524 relying on its watchers stopping correctly is a thing of beauty.
526 A flags value of \f(CW\*(C`EVLOOP_NONBLOCK\*(C'\fR will look for new events, will handle
527 those events and any outstanding ones, but will not block your process in
528 case there are no events and will return after one iteration of the loop.
530 A flags value of \f(CW\*(C`EVLOOP_ONESHOT\*(C'\fR will look for new events (waiting if
531 neccessary) and will handle those and any outstanding ones. It will block
532 your process until at least one new event arrives, and will return after
533 one iteration of the loop. This is useful if you are waiting for some
534 external event in conjunction with something not expressible using other
535 libev watchers. However, a pair of \f(CW\*(C`ev_prepare\*(C'\fR/\f(CW\*(C`ev_check\*(C'\fR watchers is
536 usually a better approach for this kind of thing.
538 Here are the gory details of what \f(CW\*(C`ev_loop\*(C'\fR does:
541 \& * If there are no active watchers (reference count is zero), return.
542 \& - Queue prepare watchers and then call all outstanding watchers.
543 \& - If we have been forked, recreate the kernel state.
544 \& - Update the kernel state with all outstanding changes.
545 \& - Update the "event loop time".
546 \& - Calculate for how long to block.
547 \& - Block the process, waiting for any events.
548 \& - Queue all outstanding I/O (fd) events.
549 \& - Update the "event loop time" and do time jump handling.
550 \& - Queue all outstanding timers.
551 \& - Queue all outstanding periodics.
552 \& - If no events are pending now, queue all idle watchers.
553 \& - Queue all check watchers.
554 \& - Call all queued watchers in reverse order (i.e. check watchers first).
555 \& Signals and child watchers are implemented as I/O watchers, and will
556 \& be handled here by queueing them when their watcher gets executed.
557 \& - If ev_unloop has been called or EVLOOP_ONESHOT or EVLOOP_NONBLOCK
558 \& were used, return, otherwise continue with step *.
561 Example: queue some jobs and then loop until no events are outsanding
565 \& ... queue jobs here, make sure they register event watchers as long
566 \& ... as they still have work to do (even an idle watcher will do..)
567 \& ev_loop (my_loop, 0);
568 \& ... jobs done. yeah!
570 .IP "ev_unloop (loop, how)" 4
571 .IX Item "ev_unloop (loop, how)"
572 Can be used to make a call to \f(CW\*(C`ev_loop\*(C'\fR return early (but only after it
573 has processed all outstanding events). The \f(CW\*(C`how\*(C'\fR argument must be either
574 \&\f(CW\*(C`EVUNLOOP_ONE\*(C'\fR, which will make the innermost \f(CW\*(C`ev_loop\*(C'\fR call return, or
575 \&\f(CW\*(C`EVUNLOOP_ALL\*(C'\fR, which will make all nested \f(CW\*(C`ev_loop\*(C'\fR calls return.
576 .IP "ev_ref (loop)" 4
577 .IX Item "ev_ref (loop)"
579 .IP "ev_unref (loop)" 4
580 .IX Item "ev_unref (loop)"
582 Ref/unref can be used to add or remove a reference count on the event
583 loop: Every watcher keeps one reference, and as long as the reference
584 count is nonzero, \f(CW\*(C`ev_loop\*(C'\fR will not return on its own. If you have
585 a watcher you never unregister that should not keep \f(CW\*(C`ev_loop\*(C'\fR from
586 returning, \fIev_unref()\fR after starting, and \fIev_ref()\fR before stopping it. For
587 example, libev itself uses this for its internal signal pipe: It is not
588 visible to the libev user and should not keep \f(CW\*(C`ev_loop\*(C'\fR from exiting if
589 no event watchers registered by it are active. It is also an excellent
590 way to do this for generic recurring timers or from within third-party
591 libraries. Just remember to \fIunref after start\fR and \fIref before stop\fR.
593 Example: create a signal watcher, but keep it from keeping \f(CW\*(C`ev_loop\*(C'\fR
594 running when nothing else is active.
597 \& struct dv_signal exitsig;
598 \& ev_signal_init (&exitsig, sig_cb, SIGINT);
599 \& ev_signal_start (myloop, &exitsig);
600 \& evf_unref (myloop);
603 Example: for some weird reason, unregister the above signal handler again.
607 \& ev_signal_stop (myloop, &exitsig);
609 .SH "ANATOMY OF A WATCHER"
610 .IX Header "ANATOMY OF A WATCHER"
611 A watcher is a structure that you create and register to record your
612 interest in some event. For instance, if you want to wait for \s-1STDIN\s0 to
613 become readable, you would create an \f(CW\*(C`ev_io\*(C'\fR watcher for that:
616 \& static void my_cb (struct ev_loop *loop, struct ev_io *w, int revents)
619 \& ev_unloop (loop, EVUNLOOP_ALL);
624 \& struct ev_loop *loop = ev_default_loop (0);
625 \& struct ev_io stdin_watcher;
626 \& ev_init (&stdin_watcher, my_cb);
627 \& ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ);
628 \& ev_io_start (loop, &stdin_watcher);
629 \& ev_loop (loop, 0);
632 As you can see, you are responsible for allocating the memory for your
633 watcher structures (and it is usually a bad idea to do this on the stack,
634 although this can sometimes be quite valid).
636 Each watcher structure must be initialised by a call to \f(CW\*(C`ev_init
637 (watcher *, callback)\*(C'\fR, which expects a callback to be provided. This
638 callback gets invoked each time the event occurs (or, in the case of io
639 watchers, each time the event loop detects that the file descriptor given
640 is readable and/or writable).
642 Each watcher type has its own \f(CW\*(C`ev_<type>_set (watcher *, ...)\*(C'\fR macro
643 with arguments specific to this watcher type. There is also a macro
644 to combine initialisation and setting in one call: \f(CW\*(C`ev_<type>_init
645 (watcher *, callback, ...)\*(C'\fR.
647 To make the watcher actually watch out for events, you have to start it
648 with a watcher-specific start function (\f(CW\*(C`ev_<type>_start (loop, watcher
649 *)\*(C'\fR), and you can stop watching for events at any time by calling the
650 corresponding stop function (\f(CW\*(C`ev_<type>_stop (loop, watcher *)\*(C'\fR.
652 As long as your watcher is active (has been started but not stopped) you
653 must not touch the values stored in it. Most specifically you must never
654 reinitialise it or call its \f(CW\*(C`set\*(C'\fR macro.
656 Each and every callback receives the event loop pointer as first, the
657 registered watcher structure as second, and a bitset of received events as
660 The received events usually include a single bit per event type received
661 (you can receive multiple events at the same time). The possible bit masks
663 .ie n .IP """EV_READ""" 4
664 .el .IP "\f(CWEV_READ\fR" 4
667 .ie n .IP """EV_WRITE""" 4
668 .el .IP "\f(CWEV_WRITE\fR" 4
671 The file descriptor in the \f(CW\*(C`ev_io\*(C'\fR watcher has become readable and/or
673 .ie n .IP """EV_TIMEOUT""" 4
674 .el .IP "\f(CWEV_TIMEOUT\fR" 4
675 .IX Item "EV_TIMEOUT"
676 The \f(CW\*(C`ev_timer\*(C'\fR watcher has timed out.
677 .ie n .IP """EV_PERIODIC""" 4
678 .el .IP "\f(CWEV_PERIODIC\fR" 4
679 .IX Item "EV_PERIODIC"
680 The \f(CW\*(C`ev_periodic\*(C'\fR watcher has timed out.
681 .ie n .IP """EV_SIGNAL""" 4
682 .el .IP "\f(CWEV_SIGNAL\fR" 4
684 The signal specified in the \f(CW\*(C`ev_signal\*(C'\fR watcher has been received by a thread.
685 .ie n .IP """EV_CHILD""" 4
686 .el .IP "\f(CWEV_CHILD\fR" 4
688 The pid specified in the \f(CW\*(C`ev_child\*(C'\fR watcher has received a status change.
689 .ie n .IP """EV_STAT""" 4
690 .el .IP "\f(CWEV_STAT\fR" 4
692 The path specified in the \f(CW\*(C`ev_stat\*(C'\fR watcher changed its attributes somehow.
693 .ie n .IP """EV_IDLE""" 4
694 .el .IP "\f(CWEV_IDLE\fR" 4
696 The \f(CW\*(C`ev_idle\*(C'\fR watcher has determined that you have nothing better to do.
697 .ie n .IP """EV_PREPARE""" 4
698 .el .IP "\f(CWEV_PREPARE\fR" 4
699 .IX Item "EV_PREPARE"
701 .ie n .IP """EV_CHECK""" 4
702 .el .IP "\f(CWEV_CHECK\fR" 4
705 All \f(CW\*(C`ev_prepare\*(C'\fR watchers are invoked just \fIbefore\fR \f(CW\*(C`ev_loop\*(C'\fR starts
706 to gather new events, and all \f(CW\*(C`ev_check\*(C'\fR watchers are invoked just after
707 \&\f(CW\*(C`ev_loop\*(C'\fR has gathered them, but before it invokes any callbacks for any
708 received events. Callbacks of both watcher types can start and stop as
709 many watchers as they want, and all of them will be taken into account
710 (for example, a \f(CW\*(C`ev_prepare\*(C'\fR watcher might start an idle watcher to keep
711 \&\f(CW\*(C`ev_loop\*(C'\fR from blocking).
712 .ie n .IP """EV_ERROR""" 4
713 .el .IP "\f(CWEV_ERROR\fR" 4
715 An unspecified error has occured, the watcher has been stopped. This might
716 happen because the watcher could not be properly started because libev
717 ran out of memory, a file descriptor was found to be closed or any other
718 problem. You best act on it by reporting the problem and somehow coping
719 with the watcher being stopped.
721 Libev will usually signal a few \*(L"dummy\*(R" events together with an error,
722 for example it might indicate that a fd is readable or writable, and if
723 your callbacks is well-written it can just attempt the operation and cope
724 with the error from \fIread()\fR or \fIwrite()\fR. This will not work in multithreaded
725 programs, though, so beware.
726 .Sh "\s-1GENERIC\s0 \s-1WATCHER\s0 \s-1FUNCTIONS\s0"
727 .IX Subsection "GENERIC WATCHER FUNCTIONS"
728 In the following description, \f(CW\*(C`TYPE\*(C'\fR stands for the watcher type,
729 e.g. \f(CW\*(C`timer\*(C'\fR for \f(CW\*(C`ev_timer\*(C'\fR watchers and \f(CW\*(C`io\*(C'\fR for \f(CW\*(C`ev_io\*(C'\fR watchers.
730 .ie n .IP """ev_init"" (ev_TYPE *watcher, callback)" 4
731 .el .IP "\f(CWev_init\fR (ev_TYPE *watcher, callback)" 4
732 .IX Item "ev_init (ev_TYPE *watcher, callback)"
733 This macro initialises the generic portion of a watcher. The contents
734 of the watcher object can be arbitrary (so \f(CW\*(C`malloc\*(C'\fR will do). Only
735 the generic parts of the watcher are initialised, you \fIneed\fR to call
736 the type-specific \f(CW\*(C`ev_TYPE_set\*(C'\fR macro afterwards to initialise the
737 type-specific parts. For each type there is also a \f(CW\*(C`ev_TYPE_init\*(C'\fR macro
738 which rolls both calls into one.
740 You can reinitialise a watcher at any time as long as it has been stopped
741 (or never started) and there are no pending events outstanding.
743 The callback is always of type \f(CW\*(C`void (*)(ev_loop *loop, ev_TYPE *watcher,
744 int revents)\*(C'\fR.
745 .ie n .IP """ev_TYPE_set"" (ev_TYPE *, [args])" 4
746 .el .IP "\f(CWev_TYPE_set\fR (ev_TYPE *, [args])" 4
747 .IX Item "ev_TYPE_set (ev_TYPE *, [args])"
748 This macro initialises the type-specific parts of a watcher. You need to
749 call \f(CW\*(C`ev_init\*(C'\fR at least once before you call this macro, but you can
750 call \f(CW\*(C`ev_TYPE_set\*(C'\fR any number of times. You must not, however, call this
751 macro on a watcher that is active (it can be pending, however, which is a
752 difference to the \f(CW\*(C`ev_init\*(C'\fR macro).
754 Although some watcher types do not have type-specific arguments
755 (e.g. \f(CW\*(C`ev_prepare\*(C'\fR) you still need to call its \f(CW\*(C`set\*(C'\fR macro.
756 .ie n .IP """ev_TYPE_init"" (ev_TYPE *watcher, callback, [args])" 4
757 .el .IP "\f(CWev_TYPE_init\fR (ev_TYPE *watcher, callback, [args])" 4
758 .IX Item "ev_TYPE_init (ev_TYPE *watcher, callback, [args])"
759 This convinience macro rolls both \f(CW\*(C`ev_init\*(C'\fR and \f(CW\*(C`ev_TYPE_set\*(C'\fR macro
760 calls into a single call. This is the most convinient method to initialise
761 a watcher. The same limitations apply, of course.
762 .ie n .IP """ev_TYPE_start"" (loop *, ev_TYPE *watcher)" 4
763 .el .IP "\f(CWev_TYPE_start\fR (loop *, ev_TYPE *watcher)" 4
764 .IX Item "ev_TYPE_start (loop *, ev_TYPE *watcher)"
765 Starts (activates) the given watcher. Only active watchers will receive
766 events. If the watcher is already active nothing will happen.
767 .ie n .IP """ev_TYPE_stop"" (loop *, ev_TYPE *watcher)" 4
768 .el .IP "\f(CWev_TYPE_stop\fR (loop *, ev_TYPE *watcher)" 4
769 .IX Item "ev_TYPE_stop (loop *, ev_TYPE *watcher)"
770 Stops the given watcher again (if active) and clears the pending
771 status. It is possible that stopped watchers are pending (for example,
772 non-repeating timers are being stopped when they become pending), but
773 \&\f(CW\*(C`ev_TYPE_stop\*(C'\fR ensures that the watcher is neither active nor pending. If
774 you want to free or reuse the memory used by the watcher it is therefore a
775 good idea to always call its \f(CW\*(C`ev_TYPE_stop\*(C'\fR function.
776 .IP "bool ev_is_active (ev_TYPE *watcher)" 4
777 .IX Item "bool ev_is_active (ev_TYPE *watcher)"
778 Returns a true value iff the watcher is active (i.e. it has been started
779 and not yet been stopped). As long as a watcher is active you must not modify
781 .IP "bool ev_is_pending (ev_TYPE *watcher)" 4
782 .IX Item "bool ev_is_pending (ev_TYPE *watcher)"
783 Returns a true value iff the watcher is pending, (i.e. it has outstanding
784 events but its callback has not yet been invoked). As long as a watcher
785 is pending (but not active) you must not call an init function on it (but
786 \&\f(CW\*(C`ev_TYPE_set\*(C'\fR is safe) and you must make sure the watcher is available to
787 libev (e.g. you cnanot \f(CW\*(C`free ()\*(C'\fR it).
788 .IP "callback = ev_cb (ev_TYPE *watcher)" 4
789 .IX Item "callback = ev_cb (ev_TYPE *watcher)"
790 Returns the callback currently set on the watcher.
791 .IP "ev_cb_set (ev_TYPE *watcher, callback)" 4
792 .IX Item "ev_cb_set (ev_TYPE *watcher, callback)"
793 Change the callback. You can change the callback at virtually any time
795 .Sh "\s-1ASSOCIATING\s0 \s-1CUSTOM\s0 \s-1DATA\s0 \s-1WITH\s0 A \s-1WATCHER\s0"
796 .IX Subsection "ASSOCIATING CUSTOM DATA WITH A WATCHER"
797 Each watcher has, by default, a member \f(CW\*(C`void *data\*(C'\fR that you can change
798 and read at any time, libev will completely ignore it. This can be used
799 to associate arbitrary data with your watcher. If you need more data and
800 don't want to allocate memory and store a pointer to it in that data
801 member, you can also \*(L"subclass\*(R" the watcher type and provide your own
810 \& struct whatever *mostinteresting;
814 And since your callback will be called with a pointer to the watcher, you
815 can cast it back to your own type:
818 \& static void my_cb (struct ev_loop *loop, struct ev_io *w_, int revents)
820 \& struct my_io *w = (struct my_io *)w_;
825 More interesting and less C\-conformant ways of catsing your callback type
826 have been omitted....
828 .IX Header "WATCHER TYPES"
829 This section describes each watcher in detail, but will not repeat
830 information given in the last section. Any initialisation/set macros,
831 functions and members specific to the watcher type are explained.
833 Members are additionally marked with either \fI[read\-only]\fR, meaning that,
834 while the watcher is active, you can look at the member and expect some
835 sensible content, but you must not modify it (you can modify it while the
836 watcher is stopped to your hearts content), or \fI[read\-write]\fR, which
837 means you can expect it to have some sensible content while the watcher
838 is active, but you can also modify it. Modifying it may not do something
839 sensible or take immediate effect (or do anything at all), but libev will
840 not crash or malfunction in any way.
841 .ie n .Sh """ev_io"" \- is this file descriptor readable or writable?"
842 .el .Sh "\f(CWev_io\fP \- is this file descriptor readable or writable?"
843 .IX Subsection "ev_io - is this file descriptor readable or writable?"
844 I/O watchers check whether a file descriptor is readable or writable
845 in each iteration of the event loop, or, more precisely, when reading
846 would not block the process and writing would at least be able to write
847 some data. This behaviour is called level-triggering because you keep
848 receiving events as long as the condition persists. Remember you can stop
849 the watcher if you don't want to act on the event and neither want to
850 receive future events.
852 In general you can register as many read and/or write event watchers per
853 fd as you want (as long as you don't confuse yourself). Setting all file
854 descriptors to non-blocking mode is also usually a good idea (but not
855 required if you know what you are doing).
857 You have to be careful with dup'ed file descriptors, though. Some backends
858 (the linux epoll backend is a notable example) cannot handle dup'ed file
859 descriptors correctly if you register interest in two or more fds pointing
860 to the same underlying file/socket/etc. description (that is, they share
861 the same underlying \*(L"file open\*(R").
863 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 \f(CW\*(C`EVBACKEND_SELECT\*(C'\fR and
865 \&\f(CW\*(C`EVBACKEND_POLL\*(C'\fR).
867 Another thing you have to watch out for is that it is quite easy to
868 receive \*(L"spurious\*(R" readyness notifications, that is your callback might
869 be called with \f(CW\*(C`EV_READ\*(C'\fR but a subsequent \f(CW\*(C`read\*(C'\fR(2) will actually block
870 because there is no data. Not only are some backends known to create a
871 lot of those (for example solaris ports), it is very easy to get into
872 this situation even with a relatively standard program structure. Thus
873 it is best to always use non-blocking I/O: An extra \f(CW\*(C`read\*(C'\fR(2) returning
874 \&\f(CW\*(C`EAGAIN\*(C'\fR is far preferable to a program hanging until some data arrives.
876 If you cannot run the fd in non-blocking mode (for example you should not
877 play around with an Xlib connection), then you have to seperately re-test
878 wether a file descriptor is really ready with a known-to-be good interface
879 such as poll (fortunately in our Xlib example, Xlib already does this on
880 its own, so its quite safe to use).
881 .IP "ev_io_init (ev_io *, callback, int fd, int events)" 4
882 .IX Item "ev_io_init (ev_io *, callback, int fd, int events)"
884 .IP "ev_io_set (ev_io *, int fd, int events)" 4
885 .IX Item "ev_io_set (ev_io *, int fd, int events)"
887 Configures an \f(CW\*(C`ev_io\*(C'\fR watcher. The \f(CW\*(C`fd\*(C'\fR is the file descriptor to
888 rceeive events for and events is either \f(CW\*(C`EV_READ\*(C'\fR, \f(CW\*(C`EV_WRITE\*(C'\fR or
889 \&\f(CW\*(C`EV_READ | EV_WRITE\*(C'\fR to receive the given events.
890 .IP "int fd [read\-only]" 4
891 .IX Item "int fd [read-only]"
892 The file descriptor being watched.
893 .IP "int events [read\-only]" 4
894 .IX Item "int events [read-only]"
895 The events being watched.
897 Example: call \f(CW\*(C`stdin_readable_cb\*(C'\fR when \s-1STDIN_FILENO\s0 has become, well
898 readable, but only once. Since it is likely line\-buffered, you could
899 attempt to read a whole line in the callback:
903 \& stdin_readable_cb (struct ev_loop *loop, struct ev_io *w, int revents)
905 \& ev_io_stop (loop, w);
906 \& .. read from stdin here (or from w->fd) and haqndle any I/O errors
912 \& struct ev_loop *loop = ev_default_init (0);
913 \& struct ev_io stdin_readable;
914 \& ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);
915 \& ev_io_start (loop, &stdin_readable);
916 \& ev_loop (loop, 0);
918 .ie n .Sh """ev_timer"" \- relative and optionally repeating timeouts"
919 .el .Sh "\f(CWev_timer\fP \- relative and optionally repeating timeouts"
920 .IX Subsection "ev_timer - relative and optionally repeating timeouts"
921 Timer watchers are simple relative timers that generate an event after a
922 given time, and optionally repeating in regular intervals after that.
924 The timers are based on real time, that is, if you register an event that
925 times out after an hour and you reset your system clock to last years
926 time, it will still time out after (roughly) and hour. \*(L"Roughly\*(R" because
927 detecting time jumps is hard, and some inaccuracies are unavoidable (the
928 monotonic clock option helps a lot here).
930 The relative timeouts are calculated relative to the \f(CW\*(C`ev_now ()\*(C'\fR
931 time. This is usually the right thing as this timestamp refers to the time
932 of the event triggering whatever timeout you are modifying/starting. If
933 you suspect event processing to be delayed and you \fIneed\fR to base the timeout
934 on the current time, use something like this to adjust for this:
937 \& ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);
940 The callback is guarenteed to be invoked only when its timeout has passed,
941 but if multiple timers become ready during the same loop iteration then
942 order of execution is undefined.
943 .IP "ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)" 4
944 .IX Item "ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)"
946 .IP "ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)" 4
947 .IX Item "ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)"
949 Configure the timer to trigger after \f(CW\*(C`after\*(C'\fR seconds. If \f(CW\*(C`repeat\*(C'\fR is
950 \&\f(CW0.\fR, then it will automatically be stopped. If it is positive, then the
951 timer will automatically be configured to trigger again \f(CW\*(C`repeat\*(C'\fR seconds
952 later, again, and again, until stopped manually.
954 The timer itself will do a best-effort at avoiding drift, that is, if you
955 configure a timer to trigger every 10 seconds, then it will trigger at
956 exactly 10 second intervals. If, however, your program cannot keep up with
957 the timer (because it takes longer than those 10 seconds to do stuff) the
958 timer will not fire more than once per event loop iteration.
959 .IP "ev_timer_again (loop)" 4
960 .IX Item "ev_timer_again (loop)"
961 This will act as if the timer timed out and restart it again if it is
962 repeating. The exact semantics are:
964 If the timer is started but nonrepeating, stop it.
966 If the timer is repeating, either start it if necessary (with the repeat
967 value), or reset the running timer to the repeat value.
969 This sounds a bit complicated, but here is a useful and typical
970 example: Imagine you have a tcp connection and you want a so-called
971 idle timeout, that is, you want to be called when there have been,
972 say, 60 seconds of inactivity on the socket. The easiest way to do
973 this is to configure an \f(CW\*(C`ev_timer\*(C'\fR with \f(CW\*(C`after\*(C'\fR=\f(CW\*(C`repeat\*(C'\fR=\f(CW60\fR and calling
974 \&\f(CW\*(C`ev_timer_again\*(C'\fR each time you successfully read or write some data. If
975 you go into an idle state where you do not expect data to travel on the
976 socket, you can stop the timer, and again will automatically restart it if
979 You can also ignore the \f(CW\*(C`after\*(C'\fR value and \f(CW\*(C`ev_timer_start\*(C'\fR altogether
980 and only ever use the \f(CW\*(C`repeat\*(C'\fR value:
983 \& ev_timer_init (timer, callback, 0., 5.);
984 \& ev_timer_again (loop, timer);
986 \& timer->again = 17.;
987 \& ev_timer_again (loop, timer);
989 \& timer->again = 10.;
990 \& ev_timer_again (loop, timer);
993 This is more efficient then stopping/starting the timer eahc time you want
994 to modify its timeout value.
995 .IP "ev_tstamp repeat [read\-write]" 4
996 .IX Item "ev_tstamp repeat [read-write]"
997 The current \f(CW\*(C`repeat\*(C'\fR value. Will be used each time the watcher times out
998 or \f(CW\*(C`ev_timer_again\*(C'\fR is called and determines the next timeout (if any),
999 which is also when any modifications are taken into account.
1001 Example: create a timer that fires after 60 seconds.
1005 \& one_minute_cb (struct ev_loop *loop, struct ev_timer *w, int revents)
1007 \& .. one minute over, w is actually stopped right here
1012 \& struct ev_timer mytimer;
1013 \& ev_timer_init (&mytimer, one_minute_cb, 60., 0.);
1014 \& ev_timer_start (loop, &mytimer);
1017 Example: create a timeout timer that times out after 10 seconds of
1022 \& timeout_cb (struct ev_loop *loop, struct ev_timer *w, int revents)
1024 \& .. ten seconds without any activity
1029 \& struct ev_timer mytimer;
1030 \& ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */
1031 \& ev_timer_again (&mytimer); /* start timer */
1032 \& ev_loop (loop, 0);
1036 \& // and in some piece of code that gets executed on any "activity":
1037 \& // reset the timeout to start ticking again at 10 seconds
1038 \& ev_timer_again (&mytimer);
1040 .ie n .Sh """ev_periodic"" \- to cron or not to cron?"
1041 .el .Sh "\f(CWev_periodic\fP \- to cron or not to cron?"
1042 .IX Subsection "ev_periodic - to cron or not to cron?"
1043 Periodic watchers are also timers of a kind, but they are very versatile
1044 (and unfortunately a bit complex).
1046 Unlike \f(CW\*(C`ev_timer\*(C'\fR's, they are not based on real time (or relative time)
1047 but on wallclock time (absolute time). You can tell a periodic watcher
1048 to trigger \*(L"at\*(R" some specific point in time. For example, if you tell a
1049 periodic watcher to trigger in 10 seconds (by specifiying e.g. \f(CW\*(C`ev_now ()
1050 + 10.\*(C'\fR) and then reset your system clock to the last year, then it will
1051 take a year to trigger the event (unlike an \f(CW\*(C`ev_timer\*(C'\fR, which would trigger
1052 roughly 10 seconds later and of course not if you reset your system time
1055 They can also be used to implement vastly more complex timers, such as
1056 triggering an event on eahc midnight, local time.
1058 As with timers, the callback is guarenteed to be invoked only when the
1059 time (\f(CW\*(C`at\*(C'\fR) has been passed, but if multiple periodic timers become ready
1060 during the same loop iteration then order of execution is undefined.
1061 .IP "ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)" 4
1062 .IX Item "ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)"
1064 .IP "ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)" 4
1065 .IX Item "ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)"
1067 Lots of arguments, lets sort it out... There are basically three modes of
1068 operation, and we will explain them from simplest to complex:
1070 .IP "* absolute timer (interval = reschedule_cb = 0)" 4
1071 .IX Item "absolute timer (interval = reschedule_cb = 0)"
1072 In this configuration the watcher triggers an event at the wallclock time
1073 \&\f(CW\*(C`at\*(C'\fR and doesn't repeat. It will not adjust when a time jump occurs,
1074 that is, if it is to be run at January 1st 2011 then it will run when the
1075 system time reaches or surpasses this time.
1076 .IP "* non-repeating interval timer (interval > 0, reschedule_cb = 0)" 4
1077 .IX Item "non-repeating interval timer (interval > 0, reschedule_cb = 0)"
1078 In this mode the watcher will always be scheduled to time out at the next
1079 \&\f(CW\*(C`at + N * interval\*(C'\fR time (for some integer N) and then repeat, regardless
1082 This can be used to create timers that do not drift with respect to system
1086 \& ev_periodic_set (&periodic, 0., 3600., 0);
1089 This doesn't mean there will always be 3600 seconds in between triggers,
1090 but only that the the callback will be called when the system time shows a
1091 full hour (\s-1UTC\s0), or more correctly, when the system time is evenly divisible
1094 Another way to think about it (for the mathematically inclined) is that
1095 \&\f(CW\*(C`ev_periodic\*(C'\fR will try to run the callback in this mode at the next possible
1096 time where \f(CW\*(C`time = at (mod interval)\*(C'\fR, regardless of any time jumps.
1097 .IP "* manual reschedule mode (reschedule_cb = callback)" 4
1098 .IX Item "manual reschedule mode (reschedule_cb = callback)"
1099 In this mode the values for \f(CW\*(C`interval\*(C'\fR and \f(CW\*(C`at\*(C'\fR are both being
1100 ignored. Instead, each time the periodic watcher gets scheduled, the
1101 reschedule callback will be called with the watcher as first, and the
1102 current time as second argument.
1104 \&\s-1NOTE:\s0 \fIThis callback \s-1MUST\s0 \s-1NOT\s0 stop or destroy any periodic watcher,
1105 ever, or make any event loop modifications\fR. If you need to stop it,
1106 return \f(CW\*(C`now + 1e30\*(C'\fR (or so, fudge fudge) and stop it afterwards (e.g. by
1107 starting a prepare watcher).
1109 Its prototype is \f(CW\*(C`ev_tstamp (*reschedule_cb)(struct ev_periodic *w,
1110 ev_tstamp now)\*(C'\fR, e.g.:
1113 \& static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now)
1115 \& return now + 60.;
1119 It must return the next time to trigger, based on the passed time value
1120 (that is, the lowest time value larger than to the second argument). It
1121 will usually be called just before the callback will be triggered, but
1122 might be called at other times, too.
1124 \&\s-1NOTE:\s0 \fIThis callback must always return a time that is later than the
1125 passed \f(CI\*(C`now\*(C'\fI value\fR. Not even \f(CW\*(C`now\*(C'\fR itself will do, it \fImust\fR be larger.
1127 This can be used to create very complex timers, such as a timer that
1128 triggers on each midnight, local time. To do this, you would calculate the
1129 next midnight after \f(CW\*(C`now\*(C'\fR and return the timestamp value for this. How
1130 you do this is, again, up to you (but it is not trivial, which is the main
1131 reason I omitted it as an example).
1135 .IP "ev_periodic_again (loop, ev_periodic *)" 4
1136 .IX Item "ev_periodic_again (loop, ev_periodic *)"
1137 Simply stops and restarts the periodic watcher again. This is only useful
1138 when you changed some parameters or the reschedule callback would return
1139 a different time than the last time it was called (e.g. in a crond like
1140 program when the crontabs have changed).
1141 .IP "ev_tstamp interval [read\-write]" 4
1142 .IX Item "ev_tstamp interval [read-write]"
1143 The current interval value. Can be modified any time, but changes only
1144 take effect when the periodic timer fires or \f(CW\*(C`ev_periodic_again\*(C'\fR is being
1146 .IP "ev_tstamp (*reschedule_cb)(struct ev_periodic *w, ev_tstamp now) [read\-write]" 4
1147 .IX Item "ev_tstamp (*reschedule_cb)(struct ev_periodic *w, ev_tstamp now) [read-write]"
1148 The current reschedule callback, or \f(CW0\fR, if this functionality is
1149 switched off. Can be changed any time, but changes only take effect when
1150 the periodic timer fires or \f(CW\*(C`ev_periodic_again\*(C'\fR is being called.
1152 Example: call a callback every hour, or, more precisely, whenever the
1153 system clock is divisible by 3600. The callback invocation times have
1154 potentially a lot of jittering, but good long-term stability.
1158 \& clock_cb (struct ev_loop *loop, struct ev_io *w, int revents)
1160 \& ... its now a full hour (UTC, or TAI or whatever your clock follows)
1165 \& struct ev_periodic hourly_tick;
1166 \& ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);
1167 \& ev_periodic_start (loop, &hourly_tick);
1170 Example: the same as above, but use a reschedule callback to do it:
1173 \& #include <math.h>
1178 \& my_scheduler_cb (struct ev_periodic *w, ev_tstamp now)
1180 \& return fmod (now, 3600.) + 3600.;
1185 \& ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);
1188 Example: call a callback every hour, starting now:
1191 \& struct ev_periodic hourly_tick;
1192 \& ev_periodic_init (&hourly_tick, clock_cb,
1193 \& fmod (ev_now (loop), 3600.), 3600., 0);
1194 \& ev_periodic_start (loop, &hourly_tick);
1196 .ie n .Sh """ev_signal"" \- signal me when a signal gets signalled!"
1197 .el .Sh "\f(CWev_signal\fP \- signal me when a signal gets signalled!"
1198 .IX Subsection "ev_signal - signal me when a signal gets signalled!"
1199 Signal watchers will trigger an event when the process receives a specific
1200 signal one or more times. Even though signals are very asynchronous, libev
1201 will try it's best to deliver signals synchronously, i.e. as part of the
1202 normal event processing, like any other event.
1204 You can configure as many watchers as you like per signal. Only when the
1205 first watcher gets started will libev actually register a signal watcher
1206 with the kernel (thus it coexists with your own signal handlers as long
1207 as you don't register any with libev). Similarly, when the last signal
1208 watcher for a signal is stopped libev will reset the signal handler to
1209 \&\s-1SIG_DFL\s0 (regardless of what it was set to before).
1210 .IP "ev_signal_init (ev_signal *, callback, int signum)" 4
1211 .IX Item "ev_signal_init (ev_signal *, callback, int signum)"
1213 .IP "ev_signal_set (ev_signal *, int signum)" 4
1214 .IX Item "ev_signal_set (ev_signal *, int signum)"
1216 Configures the watcher to trigger on the given signal number (usually one
1217 of the \f(CW\*(C`SIGxxx\*(C'\fR constants).
1218 .IP "int signum [read\-only]" 4
1219 .IX Item "int signum [read-only]"
1220 The signal the watcher watches out for.
1221 .ie n .Sh """ev_child"" \- watch out for process status changes"
1222 .el .Sh "\f(CWev_child\fP \- watch out for process status changes"
1223 .IX Subsection "ev_child - watch out for process status changes"
1224 Child watchers trigger when your process receives a \s-1SIGCHLD\s0 in response to
1225 some child status changes (most typically when a child of yours dies).
1226 .IP "ev_child_init (ev_child *, callback, int pid)" 4
1227 .IX Item "ev_child_init (ev_child *, callback, int pid)"
1229 .IP "ev_child_set (ev_child *, int pid)" 4
1230 .IX Item "ev_child_set (ev_child *, int pid)"
1232 Configures the watcher to wait for status changes of process \f(CW\*(C`pid\*(C'\fR (or
1233 \&\fIany\fR process if \f(CW\*(C`pid\*(C'\fR is specified as \f(CW0\fR). The callback can look
1234 at the \f(CW\*(C`rstatus\*(C'\fR member of the \f(CW\*(C`ev_child\*(C'\fR watcher structure to see
1235 the status word (use the macros from \f(CW\*(C`sys/wait.h\*(C'\fR and see your systems
1236 \&\f(CW\*(C`waitpid\*(C'\fR documentation). The \f(CW\*(C`rpid\*(C'\fR member contains the pid of the
1237 process causing the status change.
1238 .IP "int pid [read\-only]" 4
1239 .IX Item "int pid [read-only]"
1240 The process id this watcher watches out for, or \f(CW0\fR, meaning any process id.
1241 .IP "int rpid [read\-write]" 4
1242 .IX Item "int rpid [read-write]"
1243 The process id that detected a status change.
1244 .IP "int rstatus [read\-write]" 4
1245 .IX Item "int rstatus [read-write]"
1246 The process exit/trace status caused by \f(CW\*(C`rpid\*(C'\fR (see your systems
1247 \&\f(CW\*(C`waitpid\*(C'\fR and \f(CW\*(C`sys/wait.h\*(C'\fR documentation for details).
1249 Example: try to exit cleanly on \s-1SIGINT\s0 and \s-1SIGTERM\s0.
1253 \& sigint_cb (struct ev_loop *loop, struct ev_signal *w, int revents)
1255 \& ev_unloop (loop, EVUNLOOP_ALL);
1260 \& struct ev_signal signal_watcher;
1261 \& ev_signal_init (&signal_watcher, sigint_cb, SIGINT);
1262 \& ev_signal_start (loop, &sigint_cb);
1264 .ie n .Sh """ev_stat"" \- did the file attributes just change?"
1265 .el .Sh "\f(CWev_stat\fP \- did the file attributes just change?"
1266 .IX Subsection "ev_stat - did the file attributes just change?"
1267 This watches a filesystem path for attribute changes. That is, it calls
1268 \&\f(CW\*(C`stat\*(C'\fR regularly (or when the \s-1OS\s0 says it changed) and sees if it changed
1269 compared to the last time, invoking the callback if it did.
1271 The path does not need to exist: changing from \*(L"path exists\*(R" to \*(L"path does
1272 not exist\*(R" is a status change like any other. The condition \*(L"path does
1273 not exist\*(R" is signified by the \f(CW\*(C`st_nlink\*(C'\fR field being zero (which is
1274 otherwise always forced to be at least one) and all the other fields of
1275 the stat buffer having unspecified contents.
1277 Since there is no standard to do this, the portable implementation simply
1278 calls \f(CW\*(C`stat (2)\*(C'\fR regulalry on the path to see if it changed somehow. You
1279 can specify a recommended polling interval for this case. If you specify
1280 a polling interval of \f(CW0\fR (highly recommended!) then a \fIsuitable,
1281 unspecified default\fR value will be used (which you can expect to be around
1282 five seconds, although this might change dynamically). Libev will also
1283 impose a minimum interval which is currently around \f(CW0.1\fR, but thats
1286 This watcher type is not meant for massive numbers of stat watchers,
1287 as even with OS-supported change notifications, this can be
1288 resource\-intensive.
1290 At the time of this writing, no specific \s-1OS\s0 backends are implemented, but
1291 if demand increases, at least a kqueue and inotify backend will be added.
1292 .IP "ev_stat_init (ev_stat *, callback, const char *path, ev_tstamp interval)" 4
1293 .IX Item "ev_stat_init (ev_stat *, callback, const char *path, ev_tstamp interval)"
1295 .IP "ev_stat_set (ev_stat *, const char *path, ev_tstamp interval)" 4
1296 .IX Item "ev_stat_set (ev_stat *, const char *path, ev_tstamp interval)"
1298 Configures the watcher to wait for status changes of the given
1299 \&\f(CW\*(C`path\*(C'\fR. The \f(CW\*(C`interval\*(C'\fR is a hint on how quickly a change is expected to
1300 be detected and should normally be specified as \f(CW0\fR to let libev choose
1301 a suitable value. The memory pointed to by \f(CW\*(C`path\*(C'\fR must point to the same
1302 path for as long as the watcher is active.
1304 The callback will be receive \f(CW\*(C`EV_STAT\*(C'\fR when a change was detected,
1305 relative to the attributes at the time the watcher was started (or the
1306 last change was detected).
1307 .IP "ev_stat_stat (ev_stat *)" 4
1308 .IX Item "ev_stat_stat (ev_stat *)"
1309 Updates the stat buffer immediately with new values. If you change the
1310 watched path in your callback, you could call this fucntion to avoid
1311 detecting this change (while introducing a race condition). Can also be
1312 useful simply to find out the new values.
1313 .IP "ev_statdata attr [read\-only]" 4
1314 .IX Item "ev_statdata attr [read-only]"
1315 The most-recently detected attributes of the file. Although the type is of
1316 \&\f(CW\*(C`ev_statdata\*(C'\fR, this is usually the (or one of the) \f(CW\*(C`struct stat\*(C'\fR types
1317 suitable for your system. If the \f(CW\*(C`st_nlink\*(C'\fR member is \f(CW0\fR, then there
1318 was some error while \f(CW\*(C`stat\*(C'\fRing the file.
1319 .IP "ev_statdata prev [read\-only]" 4
1320 .IX Item "ev_statdata prev [read-only]"
1321 The previous attributes of the file. The callback gets invoked whenever
1322 \&\f(CW\*(C`prev\*(C'\fR != \f(CW\*(C`attr\*(C'\fR.
1323 .IP "ev_tstamp interval [read\-only]" 4
1324 .IX Item "ev_tstamp interval [read-only]"
1325 The specified interval.
1326 .IP "const char *path [read\-only]" 4
1327 .IX Item "const char *path [read-only]"
1328 The filesystem path that is being watched.
1330 Example: Watch \f(CW\*(C`/etc/passwd\*(C'\fR for attribute changes.
1334 \& passwd_cb (struct ev_loop *loop, ev_stat *w, int revents)
1336 \& /* /etc/passwd changed in some way */
1337 \& if (w->attr.st_nlink)
1339 \& printf ("passwd current size %ld\en", (long)w->attr.st_size);
1340 \& printf ("passwd current atime %ld\en", (long)w->attr.st_mtime);
1341 \& printf ("passwd current mtime %ld\en", (long)w->attr.st_mtime);
1344 \& /* you shalt not abuse printf for puts */
1345 \& puts ("wow, /etc/passwd is not there, expect problems. "
1346 \& "if this is windows, they already arrived\en");
1356 \& ev_stat_init (&passwd, passwd_cb, "/etc/passwd");
1357 \& ev_stat_start (loop, &passwd);
1359 .ie n .Sh """ev_idle"" \- when you've got nothing better to do..."
1360 .el .Sh "\f(CWev_idle\fP \- when you've got nothing better to do..."
1361 .IX Subsection "ev_idle - when you've got nothing better to do..."
1362 Idle watchers trigger events when there are no other events are pending
1363 (prepare, check and other idle watchers do not count). That is, as long
1364 as your process is busy handling sockets or timeouts (or even signals,
1365 imagine) it will not be triggered. But when your process is idle all idle
1366 watchers are being called again and again, once per event loop iteration \-
1367 until stopped, that is, or your process receives more events and becomes
1370 The most noteworthy effect is that as long as any idle watchers are
1371 active, the process will not block when waiting for new events.
1373 Apart from keeping your process non-blocking (which is a useful
1374 effect on its own sometimes), idle watchers are a good place to do
1375 \&\*(L"pseudo\-background processing\*(R", or delay processing stuff to after the
1376 event loop has handled all outstanding events.
1377 .IP "ev_idle_init (ev_signal *, callback)" 4
1378 .IX Item "ev_idle_init (ev_signal *, callback)"
1379 Initialises and configures the idle watcher \- it has no parameters of any
1380 kind. There is a \f(CW\*(C`ev_idle_set\*(C'\fR macro, but using it is utterly pointless,
1383 Example: dynamically allocate an \f(CW\*(C`ev_idle\*(C'\fR, start it, and in the
1384 callback, free it. Alos, use no error checking, as usual.
1388 \& idle_cb (struct ev_loop *loop, struct ev_idle *w, int revents)
1391 \& // now do something you wanted to do when the program has
1392 \& // no longer asnything immediate to do.
1397 \& struct ev_idle *idle_watcher = malloc (sizeof (struct ev_idle));
1398 \& ev_idle_init (idle_watcher, idle_cb);
1399 \& ev_idle_start (loop, idle_cb);
1401 .ie n .Sh """ev_prepare""\fP and \f(CW""ev_check"" \- customise your event loop!"
1402 .el .Sh "\f(CWev_prepare\fP and \f(CWev_check\fP \- customise your event loop!"
1403 .IX Subsection "ev_prepare and ev_check - customise your event loop!"
1404 Prepare and check watchers are usually (but not always) used in tandem:
1405 prepare watchers get invoked before the process blocks and check watchers
1408 You \fImust not\fR call \f(CW\*(C`ev_loop\*(C'\fR or similar functions that enter
1409 the current event loop from either \f(CW\*(C`ev_prepare\*(C'\fR or \f(CW\*(C`ev_check\*(C'\fR
1410 watchers. Other loops than the current one are fine, however. The
1411 rationale behind this is that you do not need to check for recursion in
1412 those watchers, i.e. the sequence will always be \f(CW\*(C`ev_prepare\*(C'\fR, blocking,
1413 \&\f(CW\*(C`ev_check\*(C'\fR so if you have one watcher of each kind they will always be
1414 called in pairs bracketing the blocking call.
1416 Their main purpose is to integrate other event mechanisms into libev and
1417 their use is somewhat advanced. This could be used, for example, to track
1418 variable changes, implement your own watchers, integrate net-snmp or a
1419 coroutine library and lots more. They are also occasionally useful if
1420 you cache some data and want to flush it before blocking (for example,
1421 in X programs you might want to do an \f(CW\*(C`XFlush ()\*(C'\fR in an \f(CW\*(C`ev_prepare\*(C'\fR
1424 This is done by examining in each prepare call which file descriptors need
1425 to be watched by the other library, registering \f(CW\*(C`ev_io\*(C'\fR watchers for
1426 them and starting an \f(CW\*(C`ev_timer\*(C'\fR watcher for any timeouts (many libraries
1427 provide just this functionality). Then, in the check watcher you check for
1428 any events that occured (by checking the pending status of all watchers
1429 and stopping them) and call back into the library. The I/O and timer
1430 callbacks will never actually be called (but must be valid nevertheless,
1431 because you never know, you know?).
1433 As another example, the Perl Coro module uses these hooks to integrate
1434 coroutines into libev programs, by yielding to other active coroutines
1435 during each prepare and only letting the process block if no coroutines
1436 are ready to run (it's actually more complicated: it only runs coroutines
1437 with priority higher than or equal to the event loop and one coroutine
1438 of lower priority, but only once, using idle watchers to keep the event
1439 loop from blocking if lower-priority coroutines are active, thus mapping
1440 low-priority coroutines to idle/background tasks).
1441 .IP "ev_prepare_init (ev_prepare *, callback)" 4
1442 .IX Item "ev_prepare_init (ev_prepare *, callback)"
1444 .IP "ev_check_init (ev_check *, callback)" 4
1445 .IX Item "ev_check_init (ev_check *, callback)"
1447 Initialises and configures the prepare or check watcher \- they have no
1448 parameters of any kind. There are \f(CW\*(C`ev_prepare_set\*(C'\fR and \f(CW\*(C`ev_check_set\*(C'\fR
1449 macros, but using them is utterly, utterly and completely pointless.
1451 Example: To include a library such as adns, you would add \s-1IO\s0 watchers
1452 and a timeout watcher in a prepare handler, as required by libadns, and
1453 in a check watcher, destroy them and call into libadns. What follows is
1454 pseudo-code only of course:
1457 \& static ev_io iow [nfd];
1458 \& static ev_timer tw;
1463 \& io_cb (ev_loop *loop, ev_io *w, int revents)
1465 \& // set the relevant poll flags
1466 \& // could also call adns_processreadable etc. here
1467 \& struct pollfd *fd = (struct pollfd *)w->data;
1468 \& if (revents & EV_READ ) fd->revents |= fd->events & POLLIN;
1469 \& if (revents & EV_WRITE) fd->revents |= fd->events & POLLOUT;
1474 \& // create io watchers for each fd and a timer before blocking
1476 \& adns_prepare_cb (ev_loop *loop, ev_prepare *w, int revents)
1478 \& int timeout = 3600000;truct pollfd fds [nfd];
1479 \& // actual code will need to loop here and realloc etc.
1480 \& adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));
1484 \& /* the callback is illegal, but won't be called as we stop during check */
1485 \& ev_timer_init (&tw, 0, timeout * 1e-3);
1486 \& ev_timer_start (loop, &tw);
1490 \& // create on ev_io per pollfd
1491 \& for (int i = 0; i < nfd; ++i)
1493 \& ev_io_init (iow + i, io_cb, fds [i].fd,
1494 \& ((fds [i].events & POLLIN ? EV_READ : 0)
1495 \& | (fds [i].events & POLLOUT ? EV_WRITE : 0)));
1499 \& fds [i].revents = 0;
1500 \& iow [i].data = fds + i;
1501 \& ev_io_start (loop, iow + i);
1507 \& // stop all watchers after blocking
1509 \& adns_check_cb (ev_loop *loop, ev_check *w, int revents)
1511 \& ev_timer_stop (loop, &tw);
1515 \& for (int i = 0; i < nfd; ++i)
1516 \& ev_io_stop (loop, iow + i);
1520 \& adns_afterpoll (adns, fds, nfd, timeval_from (ev_now (loop));
1523 .ie n .Sh """ev_embed"" \- when one backend isn't enough..."
1524 .el .Sh "\f(CWev_embed\fP \- when one backend isn't enough..."
1525 .IX Subsection "ev_embed - when one backend isn't enough..."
1526 This is a rather advanced watcher type that lets you embed one event loop
1527 into another (currently only \f(CW\*(C`ev_io\*(C'\fR events are supported in the embedded
1528 loop, other types of watchers might be handled in a delayed or incorrect
1529 fashion and must not be used).
1531 There are primarily two reasons you would want that: work around bugs and
1534 As an example for a bug workaround, the kqueue backend might only support
1535 sockets on some platform, so it is unusable as generic backend, but you
1536 still want to make use of it because you have many sockets and it scales
1537 so nicely. In this case, you would create a kqueue-based loop and embed it
1538 into your default loop (which might use e.g. poll). Overall operation will
1539 be a bit slower because first libev has to poll and then call kevent, but
1540 at least you can use both at what they are best.
1542 As for prioritising I/O: rarely you have the case where some fds have
1543 to be watched and handled very quickly (with low latency), and even
1544 priorities and idle watchers might have too much overhead. In this case
1545 you would put all the high priority stuff in one loop and all the rest in
1546 a second one, and embed the second one in the first.
1548 As long as the watcher is active, the callback will be invoked every time
1549 there might be events pending in the embedded loop. The callback must then
1550 call \f(CW\*(C`ev_embed_sweep (mainloop, watcher)\*(C'\fR to make a single sweep and invoke
1551 their callbacks (you could also start an idle watcher to give the embedded
1552 loop strictly lower priority for example). You can also set the callback
1553 to \f(CW0\fR, in which case the embed watcher will automatically execute the
1554 embedded loop sweep.
1556 As long as the watcher is started it will automatically handle events. The
1557 callback will be invoked whenever some events have been handled. You can
1558 set the callback to \f(CW0\fR to avoid having to specify one if you are not
1561 Also, there have not currently been made special provisions for forking:
1562 when you fork, you not only have to call \f(CW\*(C`ev_loop_fork\*(C'\fR on both loops,
1563 but you will also have to stop and restart any \f(CW\*(C`ev_embed\*(C'\fR watchers
1566 Unfortunately, not all backends are embeddable, only the ones returned by
1567 \&\f(CW\*(C`ev_embeddable_backends\*(C'\fR are, which, unfortunately, does not include any
1570 So when you want to use this feature you will always have to be prepared
1571 that you cannot get an embeddable loop. The recommended way to get around
1572 this is to have a separate variables for your embeddable loop, try to
1573 create it, and if that fails, use the normal loop for everything:
1576 \& struct ev_loop *loop_hi = ev_default_init (0);
1577 \& struct ev_loop *loop_lo = 0;
1578 \& struct ev_embed embed;
1582 \& // see if there is a chance of getting one that works
1583 \& // (remember that a flags value of 0 means autodetection)
1584 \& loop_lo = ev_embeddable_backends () & ev_recommended_backends ()
1585 \& ? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())
1590 \& // if we got one, then embed it, otherwise default to loop_hi
1593 \& ev_embed_init (&embed, 0, loop_lo);
1594 \& ev_embed_start (loop_hi, &embed);
1597 \& loop_lo = loop_hi;
1599 .IP "ev_embed_init (ev_embed *, callback, struct ev_loop *embedded_loop)" 4
1600 .IX Item "ev_embed_init (ev_embed *, callback, struct ev_loop *embedded_loop)"
1602 .IP "ev_embed_set (ev_embed *, callback, struct ev_loop *embedded_loop)" 4
1603 .IX Item "ev_embed_set (ev_embed *, callback, struct ev_loop *embedded_loop)"
1605 Configures the watcher to embed the given loop, which must be
1606 embeddable. If the callback is \f(CW0\fR, then \f(CW\*(C`ev_embed_sweep\*(C'\fR will be
1607 invoked automatically, otherwise it is the responsibility of the callback
1608 to invoke it (it will continue to be called until the sweep has been done,
1609 if you do not want thta, you need to temporarily stop the embed watcher).
1610 .IP "ev_embed_sweep (loop, ev_embed *)" 4
1611 .IX Item "ev_embed_sweep (loop, ev_embed *)"
1612 Make a single, non-blocking sweep over the embedded loop. This works
1613 similarly to \f(CW\*(C`ev_loop (embedded_loop, EVLOOP_NONBLOCK)\*(C'\fR, but in the most
1614 apropriate way for embedded loops.
1615 .IP "struct ev_loop *loop [read\-only]" 4
1616 .IX Item "struct ev_loop *loop [read-only]"
1617 The embedded event loop.
1618 .SH "OTHER FUNCTIONS"
1619 .IX Header "OTHER FUNCTIONS"
1620 There are some other functions of possible interest. Described. Here. Now.
1621 .IP "ev_once (loop, int fd, int events, ev_tstamp timeout, callback)" 4
1622 .IX Item "ev_once (loop, int fd, int events, ev_tstamp timeout, callback)"
1623 This function combines a simple timer and an I/O watcher, calls your
1624 callback on whichever event happens first and automatically stop both
1625 watchers. This is useful if you want to wait for a single event on an fd
1626 or timeout without having to allocate/configure/start/stop/free one or
1627 more watchers yourself.
1629 If \f(CW\*(C`fd\*(C'\fR is less than 0, then no I/O watcher will be started and events
1630 is being ignored. Otherwise, an \f(CW\*(C`ev_io\*(C'\fR watcher for the given \f(CW\*(C`fd\*(C'\fR and
1631 \&\f(CW\*(C`events\*(C'\fR set will be craeted and started.
1633 If \f(CW\*(C`timeout\*(C'\fR is less than 0, then no timeout watcher will be
1634 started. Otherwise an \f(CW\*(C`ev_timer\*(C'\fR watcher with after = \f(CW\*(C`timeout\*(C'\fR (and
1635 repeat = 0) will be started. While \f(CW0\fR is a valid timeout, it is of
1638 The callback has the type \f(CW\*(C`void (*cb)(int revents, void *arg)\*(C'\fR and gets
1639 passed an \f(CW\*(C`revents\*(C'\fR set like normal event callbacks (a combination of
1640 \&\f(CW\*(C`EV_ERROR\*(C'\fR, \f(CW\*(C`EV_READ\*(C'\fR, \f(CW\*(C`EV_WRITE\*(C'\fR or \f(CW\*(C`EV_TIMEOUT\*(C'\fR) and the \f(CW\*(C`arg\*(C'\fR
1641 value passed to \f(CW\*(C`ev_once\*(C'\fR:
1644 \& static void stdin_ready (int revents, void *arg)
1646 \& if (revents & EV_TIMEOUT)
1647 \& /* doh, nothing entered */;
1648 \& else if (revents & EV_READ)
1649 \& /* stdin might have data for us, joy! */;
1654 \& ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);
1656 .IP "ev_feed_event (ev_loop *, watcher *, int revents)" 4
1657 .IX Item "ev_feed_event (ev_loop *, watcher *, int revents)"
1658 Feeds the given event set into the event loop, as if the specified event
1659 had happened for the specified watcher (which must be a pointer to an
1660 initialised but not necessarily started event watcher).
1661 .IP "ev_feed_fd_event (ev_loop *, int fd, int revents)" 4
1662 .IX Item "ev_feed_fd_event (ev_loop *, int fd, int revents)"
1663 Feed an event on the given fd, as if a file descriptor backend detected
1664 the given events it.
1665 .IP "ev_feed_signal_event (ev_loop *loop, int signum)" 4
1666 .IX Item "ev_feed_signal_event (ev_loop *loop, int signum)"
1667 Feed an event as if the given signal occured (\f(CW\*(C`loop\*(C'\fR must be the default
1669 .SH "LIBEVENT EMULATION"
1670 .IX Header "LIBEVENT EMULATION"
1671 Libev offers a compatibility emulation layer for libevent. It cannot
1672 emulate the internals of libevent, so here are some usage hints:
1673 .IP "* Use it by including <event.h>, as usual." 4
1674 .IX Item "Use it by including <event.h>, as usual."
1676 .IP "* The following members are fully supported: ev_base, ev_callback, ev_arg, ev_fd, ev_res, ev_events." 4
1677 .IX Item "The following members are fully supported: ev_base, ev_callback, ev_arg, ev_fd, ev_res, ev_events."
1678 .IP "* Avoid using ev_flags and the EVLIST_*\-macros, while it is maintained by libev, it does not work exactly the same way as in libevent (consider it a private \s-1API\s0)." 4
1679 .IX Item "Avoid using ev_flags and the EVLIST_*-macros, while it is maintained by libev, it does not work exactly the same way as in libevent (consider it a private API)."
1680 .IP "* Priorities are not currently supported. Initialising priorities will fail and all watchers will have the same priority, even though there is an ev_pri field." 4
1681 .IX Item "Priorities are not currently supported. Initialising priorities will fail and all watchers will have the same priority, even though there is an ev_pri field."
1682 .IP "* Other members are not supported." 4
1683 .IX Item "Other members are not supported."
1684 .IP "* The libev emulation is \fInot\fR \s-1ABI\s0 compatible to libevent, you need to use the libev header file and library." 4
1685 .IX Item "The libev emulation is not ABI compatible to libevent, you need to use the libev header file and library."
1688 .IX Header " SUPPORT"
1689 Libev comes with some simplistic wrapper classes for \*(C+ that mainly allow
1690 you to use some convinience methods to start/stop watchers and also change
1691 the callback model to a model using method callbacks on objects.
1696 \& #include <ev++.h>
1699 (it is not installed by default). This automatically includes \fIev.h\fR
1700 and puts all of its definitions (many of them macros) into the global
1701 namespace. All \*(C+ specific things are put into the \f(CW\*(C`ev\*(C'\fR namespace.
1703 It should support all the same embedding options as \fIev.h\fR, most notably
1704 \&\f(CW\*(C`EV_MULTIPLICITY\*(C'\fR.
1706 Here is a list of things available in the \f(CW\*(C`ev\*(C'\fR namespace:
1707 .ie n .IP """ev::READ""\fR, \f(CW""ev::WRITE"" etc." 4
1708 .el .IP "\f(CWev::READ\fR, \f(CWev::WRITE\fR etc." 4
1709 .IX Item "ev::READ, ev::WRITE etc."
1710 These are just enum values with the same values as the \f(CW\*(C`EV_READ\*(C'\fR etc.
1711 macros from \fIev.h\fR.
1712 .ie n .IP """ev::tstamp""\fR, \f(CW""ev::now""" 4
1713 .el .IP "\f(CWev::tstamp\fR, \f(CWev::now\fR" 4
1714 .IX Item "ev::tstamp, ev::now"
1715 Aliases to the same types/functions as with the \f(CW\*(C`ev_\*(C'\fR prefix.
1716 .ie n .IP """ev::io""\fR, \f(CW""ev::timer""\fR, \f(CW""ev::periodic""\fR, \f(CW""ev::idle""\fR, \f(CW""ev::sig"" etc." 4
1717 .el .IP "\f(CWev::io\fR, \f(CWev::timer\fR, \f(CWev::periodic\fR, \f(CWev::idle\fR, \f(CWev::sig\fR etc." 4
1718 .IX Item "ev::io, ev::timer, ev::periodic, ev::idle, ev::sig etc."
1719 For each \f(CW\*(C`ev_TYPE\*(C'\fR watcher in \fIev.h\fR there is a corresponding class of
1720 the same name in the \f(CW\*(C`ev\*(C'\fR namespace, with the exception of \f(CW\*(C`ev_signal\*(C'\fR
1721 which is called \f(CW\*(C`ev::sig\*(C'\fR to avoid clashes with the \f(CW\*(C`signal\*(C'\fR macro
1722 defines by many implementations.
1724 All of those classes have these methods:
1726 .IP "ev::TYPE::TYPE (object *, object::method *)" 4
1727 .IX Item "ev::TYPE::TYPE (object *, object::method *)"
1729 .IP "ev::TYPE::TYPE (object *, object::method *, struct ev_loop *)" 4
1730 .IX Item "ev::TYPE::TYPE (object *, object::method *, struct ev_loop *)"
1731 .IP "ev::TYPE::~TYPE" 4
1732 .IX Item "ev::TYPE::~TYPE"
1734 The constructor takes a pointer to an object and a method pointer to
1735 the event handler callback to call in this class. The constructor calls
1736 \&\f(CW\*(C`ev_init\*(C'\fR for you, which means you have to call the \f(CW\*(C`set\*(C'\fR method
1737 before starting it. If you do not specify a loop then the constructor
1738 automatically associates the default loop with this watcher.
1740 The destructor automatically stops the watcher if it is active.
1741 .IP "w\->set (struct ev_loop *)" 4
1742 .IX Item "w->set (struct ev_loop *)"
1743 Associates a different \f(CW\*(C`struct ev_loop\*(C'\fR with this watcher. You can only
1744 do this when the watcher is inactive (and not pending either).
1745 .IP "w\->set ([args])" 4
1746 .IX Item "w->set ([args])"
1747 Basically the same as \f(CW\*(C`ev_TYPE_set\*(C'\fR, with the same args. Must be
1748 called at least once. Unlike the C counterpart, an active watcher gets
1749 automatically stopped and restarted.
1750 .IP "w\->start ()" 4
1751 .IX Item "w->start ()"
1752 Starts the watcher. Note that there is no \f(CW\*(C`loop\*(C'\fR argument as the
1753 constructor already takes the loop.
1755 .IX Item "w->stop ()"
1756 Stops the watcher if it is active. Again, no \f(CW\*(C`loop\*(C'\fR argument.
1757 .ie n .IP "w\->again () ""ev::timer""\fR, \f(CW""ev::periodic"" only" 4
1758 .el .IP "w\->again () \f(CWev::timer\fR, \f(CWev::periodic\fR only" 4
1759 .IX Item "w->again () ev::timer, ev::periodic only"
1760 For \f(CW\*(C`ev::timer\*(C'\fR and \f(CW\*(C`ev::periodic\*(C'\fR, this invokes the corresponding
1761 \&\f(CW\*(C`ev_TYPE_again\*(C'\fR function.
1762 .ie n .IP "w\->sweep () ""ev::embed"" only" 4
1763 .el .IP "w\->sweep () \f(CWev::embed\fR only" 4
1764 .IX Item "w->sweep () ev::embed only"
1765 Invokes \f(CW\*(C`ev_embed_sweep\*(C'\fR.
1766 .ie n .IP "w\->update () ""ev::stat"" only" 4
1767 .el .IP "w\->update () \f(CWev::stat\fR only" 4
1768 .IX Item "w->update () ev::stat only"
1769 Invokes \f(CW\*(C`ev_stat_stat\*(C'\fR.
1774 Example: Define a class with an \s-1IO\s0 and idle watcher, start one of them in
1780 \& ev_io io; void io_cb (ev::io &w, int revents);
1781 \& ev_idle idle void idle_cb (ev::idle &w, int revents);
1790 \& myclass::myclass (int fd)
1791 \& : io (this, &myclass::io_cb),
1792 \& idle (this, &myclass::idle_cb)
1794 \& io.start (fd, ev::READ);
1798 .IX Header "EMBEDDING"
1799 Libev can (and often is) directly embedded into host
1800 applications. Examples of applications that embed it include the Deliantra
1801 Game Server, the \s-1EV\s0 perl module, the \s-1GNU\s0 Virtual Private Ethernet (gvpe)
1804 The goal is to enable you to just copy the neecssary files into your
1805 source directory without having to change even a single line in them, so
1806 you can easily upgrade by simply copying (or having a checked-out copy of
1807 libev somewhere in your source tree).
1808 .Sh "\s-1FILESETS\s0"
1809 .IX Subsection "FILESETS"
1810 Depending on what features you need you need to include one or more sets of files
1813 \fI\s-1CORE\s0 \s-1EVENT\s0 \s-1LOOP\s0\fR
1814 .IX Subsection "CORE EVENT LOOP"
1816 To include only the libev core (all the \f(CW\*(C`ev_*\*(C'\fR functions), with manual
1817 configuration (no autoconf):
1820 \& #define EV_STANDALONE 1
1824 This will automatically include \fIev.h\fR, too, and should be done in a
1825 single C source file only to provide the function implementations. To use
1826 it, do the same for \fIev.h\fR in all files wishing to use this \s-1API\s0 (best
1827 done by writing a wrapper around \fIev.h\fR that you can include instead and
1828 where you can put other configuration options):
1831 \& #define EV_STANDALONE 1
1835 Both header files and implementation files can be compiled with a \*(C+
1836 compiler (at least, thats a stated goal, and breakage will be treated
1839 You need the following files in your source tree, or in a directory
1840 in your include path (e.g. in libev/ when using \-Ilibev):
1850 \& ev_win32.c required on win32 platforms only
1854 \& ev_select.c only when select backend is enabled (which is by default)
1855 \& ev_poll.c only when poll backend is enabled (disabled by default)
1856 \& ev_epoll.c only when the epoll backend is enabled (disabled by default)
1857 \& ev_kqueue.c only when the kqueue backend is enabled (disabled by default)
1858 \& ev_port.c only when the solaris port backend is enabled (disabled by default)
1861 \&\fIev.c\fR includes the backend files directly when enabled, so you only need
1862 to compile this single file.
1864 \fI\s-1LIBEVENT\s0 \s-1COMPATIBILITY\s0 \s-1API\s0\fR
1865 .IX Subsection "LIBEVENT COMPATIBILITY API"
1867 To include the libevent compatibility \s-1API\s0, also include:
1870 \& #include "event.c"
1873 in the file including \fIev.c\fR, and:
1876 \& #include "event.h"
1879 in the files that want to use the libevent \s-1API\s0. This also includes \fIev.h\fR.
1881 You need the following additional files for this:
1888 \fI\s-1AUTOCONF\s0 \s-1SUPPORT\s0\fR
1889 .IX Subsection "AUTOCONF SUPPORT"
1891 Instead of using \f(CW\*(C`EV_STANDALONE=1\*(C'\fR and providing your config in
1892 whatever way you want, you can also \f(CW\*(C`m4_include([libev.m4])\*(C'\fR in your
1893 \&\fIconfigure.ac\fR and leave \f(CW\*(C`EV_STANDALONE\*(C'\fR undefined. \fIev.c\fR will then
1894 include \fIconfig.h\fR and configure itself accordingly.
1896 For this of course you need the m4 file:
1901 .Sh "\s-1PREPROCESSOR\s0 \s-1SYMBOLS/MACROS\s0"
1902 .IX Subsection "PREPROCESSOR SYMBOLS/MACROS"
1903 Libev can be configured via a variety of preprocessor symbols you have to define
1904 before including any of its files. The default is not to build for multiplicity
1905 and only include the select backend.
1906 .IP "\s-1EV_STANDALONE\s0" 4
1907 .IX Item "EV_STANDALONE"
1908 Must always be \f(CW1\fR if you do not use autoconf configuration, which
1909 keeps libev from including \fIconfig.h\fR, and it also defines dummy
1910 implementations for some libevent functions (such as logging, which is not
1911 supported). It will also not define any of the structs usually found in
1912 \&\fIevent.h\fR that are not directly supported by the libev core alone.
1913 .IP "\s-1EV_USE_MONOTONIC\s0" 4
1914 .IX Item "EV_USE_MONOTONIC"
1915 If defined to be \f(CW1\fR, libev will try to detect the availability of the
1916 monotonic clock option at both compiletime and runtime. Otherwise no use
1917 of the monotonic clock option will be attempted. If you enable this, you
1918 usually have to link against librt or something similar. Enabling it when
1919 the functionality isn't available is safe, though, althoguh you have
1920 to make sure you link against any libraries where the \f(CW\*(C`clock_gettime\*(C'\fR
1921 function is hiding in (often \fI\-lrt\fR).
1922 .IP "\s-1EV_USE_REALTIME\s0" 4
1923 .IX Item "EV_USE_REALTIME"
1924 If defined to be \f(CW1\fR, libev will try to detect the availability of the
1925 realtime clock option at compiletime (and assume its availability at
1926 runtime if successful). Otherwise no use of the realtime clock option will
1927 be attempted. This effectively replaces \f(CW\*(C`gettimeofday\*(C'\fR by \f(CW\*(C`clock_get
1928 (CLOCK_REALTIME, ...)\*(C'\fR and will not normally affect correctness. See tzhe note about libraries
1929 in the description of \f(CW\*(C`EV_USE_MONOTONIC\*(C'\fR, though.
1930 .IP "\s-1EV_USE_SELECT\s0" 4
1931 .IX Item "EV_USE_SELECT"
1932 If undefined or defined to be \f(CW1\fR, libev will compile in support for the
1933 \&\f(CW\*(C`select\*(C'\fR(2) backend. No attempt at autodetection will be done: if no
1934 other method takes over, select will be it. Otherwise the select backend
1935 will not be compiled in.
1936 .IP "\s-1EV_SELECT_USE_FD_SET\s0" 4
1937 .IX Item "EV_SELECT_USE_FD_SET"
1938 If defined to \f(CW1\fR, then the select backend will use the system \f(CW\*(C`fd_set\*(C'\fR
1939 structure. This is useful if libev doesn't compile due to a missing
1940 \&\f(CW\*(C`NFDBITS\*(C'\fR or \f(CW\*(C`fd_mask\*(C'\fR definition or it misguesses the bitset layout on
1941 exotic systems. This usually limits the range of file descriptors to some
1942 low limit such as 1024 or might have other limitations (winsocket only
1943 allows 64 sockets). The \f(CW\*(C`FD_SETSIZE\*(C'\fR macro, set before compilation, might
1944 influence the size of the \f(CW\*(C`fd_set\*(C'\fR used.
1945 .IP "\s-1EV_SELECT_IS_WINSOCKET\s0" 4
1946 .IX Item "EV_SELECT_IS_WINSOCKET"
1947 When defined to \f(CW1\fR, the select backend will assume that
1948 select/socket/connect etc. don't understand file descriptors but
1949 wants osf handles on win32 (this is the case when the select to
1950 be used is the winsock select). This means that it will call
1951 \&\f(CW\*(C`_get_osfhandle\*(C'\fR on the fd to convert it to an \s-1OS\s0 handle. Otherwise,
1952 it is assumed that all these functions actually work on fds, even
1953 on win32. Should not be defined on non\-win32 platforms.
1954 .IP "\s-1EV_USE_POLL\s0" 4
1955 .IX Item "EV_USE_POLL"
1956 If defined to be \f(CW1\fR, libev will compile in support for the \f(CW\*(C`poll\*(C'\fR(2)
1957 backend. Otherwise it will be enabled on non\-win32 platforms. It
1958 takes precedence over select.
1959 .IP "\s-1EV_USE_EPOLL\s0" 4
1960 .IX Item "EV_USE_EPOLL"
1961 If defined to be \f(CW1\fR, libev will compile in support for the Linux
1962 \&\f(CW\*(C`epoll\*(C'\fR(7) backend. Its availability will be detected at runtime,
1963 otherwise another method will be used as fallback. This is the
1964 preferred backend for GNU/Linux systems.
1965 .IP "\s-1EV_USE_KQUEUE\s0" 4
1966 .IX Item "EV_USE_KQUEUE"
1967 If defined to be \f(CW1\fR, libev will compile in support for the \s-1BSD\s0 style
1968 \&\f(CW\*(C`kqueue\*(C'\fR(2) backend. Its actual availability will be detected at runtime,
1969 otherwise another method will be used as fallback. This is the preferred
1970 backend for \s-1BSD\s0 and BSD-like systems, although on most BSDs kqueue only
1971 supports some types of fds correctly (the only platform we found that
1972 supports ptys for example was NetBSD), so kqueue might be compiled in, but
1973 not be used unless explicitly requested. The best way to use it is to find
1974 out whether kqueue supports your type of fd properly and use an embedded
1976 .IP "\s-1EV_USE_PORT\s0" 4
1977 .IX Item "EV_USE_PORT"
1978 If defined to be \f(CW1\fR, libev will compile in support for the Solaris
1979 10 port style backend. Its availability will be detected at runtime,
1980 otherwise another method will be used as fallback. This is the preferred
1981 backend for Solaris 10 systems.
1982 .IP "\s-1EV_USE_DEVPOLL\s0" 4
1983 .IX Item "EV_USE_DEVPOLL"
1984 reserved for future expansion, works like the \s-1USE\s0 symbols above.
1987 The name of the \fIev.h\fR header file used to include it. The default if
1988 undefined is \f(CW\*(C`<ev.h>\*(C'\fR in \fIevent.h\fR and \f(CW"ev.h"\fR in \fIev.c\fR. This
1989 can be used to virtually rename the \fIev.h\fR header file in case of conflicts.
1990 .IP "\s-1EV_CONFIG_H\s0" 4
1991 .IX Item "EV_CONFIG_H"
1992 If \f(CW\*(C`EV_STANDALONE\*(C'\fR isn't \f(CW1\fR, this variable can be used to override
1993 \&\fIev.c\fR's idea of where to find the \fIconfig.h\fR file, similarly to
1994 \&\f(CW\*(C`EV_H\*(C'\fR, above.
1995 .IP "\s-1EV_EVENT_H\s0" 4
1996 .IX Item "EV_EVENT_H"
1997 Similarly to \f(CW\*(C`EV_H\*(C'\fR, this macro can be used to override \fIevent.c\fR's idea
1998 of how the \fIevent.h\fR header can be found.
1999 .IP "\s-1EV_PROTOTYPES\s0" 4
2000 .IX Item "EV_PROTOTYPES"
2001 If defined to be \f(CW0\fR, then \fIev.h\fR will not define any function
2002 prototypes, but still define all the structs and other symbols. This is
2003 occasionally useful if you want to provide your own wrapper functions
2004 around libev functions.
2005 .IP "\s-1EV_MULTIPLICITY\s0" 4
2006 .IX Item "EV_MULTIPLICITY"
2007 If undefined or defined to \f(CW1\fR, then all event-loop-specific functions
2008 will have the \f(CW\*(C`struct ev_loop *\*(C'\fR as first argument, and you can create
2009 additional independent event loops. Otherwise there will be no support
2010 for multiple event loops and there is no first event loop pointer
2011 argument. Instead, all functions act on the single default loop.
2012 .IP "\s-1EV_PERIODIC_ENABLE\s0" 4
2013 .IX Item "EV_PERIODIC_ENABLE"
2014 If undefined or defined to be \f(CW1\fR, then periodic timers are supported. If
2015 defined to be \f(CW0\fR, then they are not. Disabling them saves a few kB of
2017 .IP "\s-1EV_EMBED_ENABLE\s0" 4
2018 .IX Item "EV_EMBED_ENABLE"
2019 If undefined or defined to be \f(CW1\fR, then embed watchers are supported. If
2020 defined to be \f(CW0\fR, then they are not.
2021 .IP "\s-1EV_STAT_ENABLE\s0" 4
2022 .IX Item "EV_STAT_ENABLE"
2023 If undefined or defined to be \f(CW1\fR, then stat watchers are supported. If
2024 defined to be \f(CW0\fR, then they are not.
2025 .IP "\s-1EV_MINIMAL\s0" 4
2026 .IX Item "EV_MINIMAL"
2027 If you need to shave off some kilobytes of code at the expense of some
2028 speed, define this symbol to \f(CW1\fR. Currently only used for gcc to override
2029 some inlining decisions, saves roughly 30% codesize of amd64.
2030 .IP "\s-1EV_COMMON\s0" 4
2031 .IX Item "EV_COMMON"
2032 By default, all watchers have a \f(CW\*(C`void *data\*(C'\fR member. By redefining
2033 this macro to a something else you can include more and other types of
2034 members. You have to define it each time you include one of the files,
2035 though, and it must be identical each time.
2037 For example, the perl \s-1EV\s0 module uses something like this:
2040 \& #define EV_COMMON \e
2041 \& SV *self; /* contains this struct */ \e
2042 \& SV *cb_sv, *fh /* note no trailing ";" */
2044 .IP "\s-1EV_CB_DECLARE\s0 (type)" 4
2045 .IX Item "EV_CB_DECLARE (type)"
2047 .IP "\s-1EV_CB_INVOKE\s0 (watcher, revents)" 4
2048 .IX Item "EV_CB_INVOKE (watcher, revents)"
2049 .IP "ev_set_cb (ev, cb)" 4
2050 .IX Item "ev_set_cb (ev, cb)"
2052 Can be used to change the callback member declaration in each watcher,
2053 and the way callbacks are invoked and set. Must expand to a struct member
2054 definition and a statement, respectively. See the \fIev.v\fR header file for
2055 their default definitions. One possible use for overriding these is to
2056 avoid the \f(CW\*(C`struct ev_loop *\*(C'\fR as first argument in all cases, or to use
2057 method calls instead of plain function calls in \*(C+.
2058 .Sh "\s-1EXAMPLES\s0"
2059 .IX Subsection "EXAMPLES"
2060 For a real-world example of a program the includes libev
2061 verbatim, you can have a look at the \s-1EV\s0 perl module
2062 (<http://software.schmorp.de/pkg/EV.html>). It has the libev files in
2063 the \fIlibev/\fR subdirectory and includes them in the \fI\s-1EV/EVAPI\s0.h\fR (public
2064 interface) and \fI\s-1EV\s0.xs\fR (implementation) files. Only the \fI\s-1EV\s0.xs\fR file
2065 will be compiled. It is pretty complex because it provides its own header
2068 The usage in rxvt-unicode is simpler. It has a \fIev_cpp.h\fR header file
2069 that everybody includes and which overrides some autoconf choices:
2072 \& #define EV_USE_POLL 0
2073 \& #define EV_MULTIPLICITY 0
2074 \& #define EV_PERIODICS 0
2075 \& #define EV_CONFIG_H <config.h>
2079 \& #include "ev++.h"
2082 And a \fIev_cpp.C\fR implementation file that contains libev proper and is compiled:
2085 \& #include "ev_cpp.h"
2089 .IX Header "COMPLEXITIES"
2090 In this section the complexities of (many of) the algorithms used inside
2091 libev will be explained. For complexity discussions about backends see the
2092 documentation for \f(CW\*(C`ev_default_init\*(C'\fR.
2094 .IP "Starting and stopping timer/periodic watchers: O(log skipped_other_timers)" 4
2095 .IX Item "Starting and stopping timer/periodic watchers: O(log skipped_other_timers)"
2097 .IP "Changing timer/periodic watchers (by autorepeat, again): O(log skipped_other_timers)" 4
2098 .IX Item "Changing timer/periodic watchers (by autorepeat, again): O(log skipped_other_timers)"
2099 .IP "Starting io/check/prepare/idle/signal/child watchers: O(1)" 4
2100 .IX Item "Starting io/check/prepare/idle/signal/child watchers: O(1)"
2101 .IP "Stopping check/prepare/idle watchers: O(1)" 4
2102 .IX Item "Stopping check/prepare/idle watchers: O(1)"
2103 .IP "Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % 16))" 4
2104 .IX Item "Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % 16))"
2105 .IP "Finding the next timer per loop iteration: O(1)" 4
2106 .IX Item "Finding the next timer per loop iteration: O(1)"
2107 .IP "Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)" 4
2108 .IX Item "Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)"
2109 .IP "Activating one watcher: O(1)" 4
2110 .IX Item "Activating one watcher: O(1)"
2116 Marc Lehmann <libev@schmorp.de>.