ViewVC Help
View File | Revision Log | Show Annotations | Download File
Revision: 1.469
Committed: Sat Jun 3 08:53:03 2023 UTC (13 months, 2 weeks ago) by root
Branch: MAIN
Changes since 1.468: +35 -19 lines
Log Message:
*** empty log message ***

File Contents

# Content
1 =encoding utf-8
3 =head1 NAME
5 libev - a high performance full-featured event loop written in C
7 =head1 SYNOPSIS
9 #include <ev.h>
13 // a single header file is required
14 #include <ev.h>
16 #include <stdio.h> // for puts
18 // every watcher type has its own typedef'd struct
19 // with the name ev_TYPE
20 ev_io stdin_watcher;
21 ev_timer timeout_watcher;
23 // all watcher callbacks have a similar signature
24 // this callback is called when data is readable on stdin
25 static void
26 stdin_cb (EV_P_ ev_io *w, int revents)
27 {
28 puts ("stdin ready");
29 // for one-shot events, one must manually stop the watcher
30 // with its corresponding stop function.
31 ev_io_stop (EV_A_ w);
33 // this causes all nested ev_run's to stop iterating
34 ev_break (EV_A_ EVBREAK_ALL);
35 }
37 // another callback, this time for a time-out
38 static void
39 timeout_cb (EV_P_ ev_timer *w, int revents)
40 {
41 puts ("timeout");
42 // this causes the innermost ev_run to stop iterating
43 ev_break (EV_A_ EVBREAK_ONE);
44 }
46 int
47 main (void)
48 {
49 // use the default event loop unless you have special needs
50 struct ev_loop *loop = EV_DEFAULT;
52 // initialise an io watcher, then start it
53 // this one will watch for stdin to become readable
54 ev_io_init (&stdin_watcher, stdin_cb, /*STDIN_FILENO*/ 0, EV_READ);
55 ev_io_start (loop, &stdin_watcher);
57 // initialise a timer watcher, then start it
58 // simple non-repeating 5.5 second timeout
59 ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);
60 ev_timer_start (loop, &timeout_watcher);
62 // now wait for events to arrive
63 ev_run (loop, 0);
65 // break was called, so exit
66 return 0;
67 }
71 This document documents the libev software package.
73 The newest version of this document is also available as an html-formatted
74 web page you might find easier to navigate when reading it for the first
75 time: L<>.
77 While this document tries to be as complete as possible in documenting
78 libev, its usage and the rationale behind its design, it is not a tutorial
79 on event-based programming, nor will it introduce event-based programming
80 with libev.
82 Familiarity with event based programming techniques in general is assumed
83 throughout this document.
87 This manual tries to be very detailed, but unfortunately, this also makes
88 it very long. If you just want to know the basics of libev, I suggest
89 reading L</ANATOMY OF A WATCHER>, then the L</EXAMPLE PROGRAM> above and
90 look up the missing functions in L</GLOBAL FUNCTIONS> and the C<ev_io> and
91 C<ev_timer> sections in L</WATCHER TYPES>.
93 =head1 ABOUT LIBEV
95 Libev is an event loop: you register interest in certain events (such as a
96 file descriptor being readable or a timeout occurring), and it will manage
97 these event sources and provide your program with events.
99 To do this, it must take more or less complete control over your process
100 (or thread) by executing the I<event loop> handler, and will then
101 communicate events via a callback mechanism.
103 You register interest in certain events by registering so-called I<event
104 watchers>, which are relatively small C structures you initialise with the
105 details of the event, and then hand it over to libev by I<starting> the
106 watcher.
108 =head2 FEATURES
110 Libev supports C<select>, C<poll>, the Linux-specific aio and C<epoll>
111 interfaces, the BSD-specific C<kqueue> and the Solaris-specific event port
112 mechanisms for file descriptor events (C<ev_io>), the Linux C<inotify>
113 interface (for C<ev_stat>), Linux eventfd/signalfd (for faster and cleaner
114 inter-thread wakeup (C<ev_async>)/signal handling (C<ev_signal>)) relative
115 timers (C<ev_timer>), absolute timers with customised rescheduling
116 (C<ev_periodic>), synchronous signals (C<ev_signal>), process status
117 change events (C<ev_child>), and event watchers dealing with the event
118 loop mechanism itself (C<ev_idle>, C<ev_embed>, C<ev_prepare> and
119 C<ev_check> watchers) as well as file watchers (C<ev_stat>) and even
120 limited support for fork events (C<ev_fork>).
122 It also is quite fast (see this
123 L<benchmark|> comparing it to libevent
124 for example).
126 =head2 CONVENTIONS
128 Libev is very configurable. In this manual the default (and most common)
129 configuration will be described, which supports multiple event loops. For
130 more info about various configuration options please have a look at
131 B<EMBED> section in this manual. If libev was configured without support
132 for multiple event loops, then all functions taking an initial argument of
133 name C<loop> (which is always of type C<struct ev_loop *>) will not have
134 this argument.
138 Libev represents time as a single floating point number, representing
139 the (fractional) number of seconds since the (POSIX) epoch (in practice
140 somewhere near the beginning of 1970, details are complicated, don't
141 ask). This type is called C<ev_tstamp>, which is what you should use
142 too. It usually aliases to the C<double> type in C. When you need to do
143 any calculations on it, you should treat it as some floating point value.
145 Unlike the name component C<stamp> might indicate, it is also used for
146 time differences (e.g. delays) throughout libev.
150 Libev knows three classes of errors: operating system errors, usage errors
151 and internal errors (bugs).
153 When libev catches an operating system error it cannot handle (for example
154 a system call indicating a condition libev cannot fix), it calls the callback
155 set via C<ev_set_syserr_cb>, which is supposed to fix the problem or
156 abort. The default is to print a diagnostic message and to call C<abort
157 ()>.
159 When libev detects a usage error such as a negative timer interval, then
160 it will print a diagnostic message and abort (via the C<assert> mechanism,
161 so C<NDEBUG> will disable this checking): these are programming errors in
162 the libev caller and need to be fixed there.
164 Via the C<EV_FREQUENT> macro you can compile in and/or enable extensive
165 consistency checking code inside libev that can be used to check for
166 internal inconsistencies, suually caused by application bugs.
168 Libev also has a few internal error-checking C<assert>ions. These do not
169 trigger under normal circumstances, as they indicate either a bug in libev
170 or worse.
175 These functions can be called anytime, even before initialising the
176 library in any way.
178 =over 4
180 =item ev_tstamp ev_time ()
182 Returns the current time as libev would use it. Please note that the
183 C<ev_now> function is usually faster and also often returns the timestamp
184 you actually want to know. Also interesting is the combination of
185 C<ev_now_update> and C<ev_now>.
187 =item ev_sleep (ev_tstamp interval)
189 Sleep for the given interval: The current thread will be blocked
190 until either it is interrupted or the given time interval has
191 passed (approximately - it might return a bit earlier even if not
192 interrupted). Returns immediately if C<< interval <= 0 >>.
194 Basically this is a sub-second-resolution C<sleep ()>.
196 The range of the C<interval> is limited - libev only guarantees to work
197 with sleep times of up to one day (C<< interval <= 86400 >>).
199 =item int ev_version_major ()
201 =item int ev_version_minor ()
203 You can find out the major and minor ABI version numbers of the library
204 you linked against by calling the functions C<ev_version_major> and
205 C<ev_version_minor>. If you want, you can compare against the global
206 symbols C<EV_VERSION_MAJOR> and C<EV_VERSION_MINOR>, which specify the
207 version of the library your program was compiled against.
209 These version numbers refer to the ABI version of the library, not the
210 release version.
212 Usually, it's a good idea to terminate if the major versions mismatch,
213 as this indicates an incompatible change. Minor versions are usually
214 compatible to older versions, so a larger minor version alone is usually
215 not a problem.
217 Example: Make sure we haven't accidentally been linked against the wrong
218 version (note, however, that this will not detect other ABI mismatches,
219 such as LFS or reentrancy).
221 assert (("libev version mismatch",
222 ev_version_major () == EV_VERSION_MAJOR
223 && ev_version_minor () >= EV_VERSION_MINOR));
225 =item unsigned int ev_supported_backends ()
227 Return the set of all backends (i.e. their corresponding C<EV_BACKEND_*>
228 value) compiled into this binary of libev (independent of their
229 availability on the system you are running on). See C<ev_default_loop> for
230 a description of the set values.
232 Example: make sure we have the epoll method, because yeah this is cool and
233 a must have and can we have a torrent of it please!!!11
235 assert (("sorry, no epoll, no sex",
236 ev_supported_backends () & EVBACKEND_EPOLL));
238 =item unsigned int ev_recommended_backends ()
240 Return the set of all backends compiled into this binary of libev and
241 also recommended for this platform, meaning it will work for most file
242 descriptor types. This set is often smaller than the one returned by
243 C<ev_supported_backends>, as for example kqueue is broken on most BSDs
244 and will not be auto-detected unless you explicitly request it (assuming
245 you know what you are doing). This is the set of backends that libev will
246 probe for if you specify no backends explicitly.
248 =item unsigned int ev_embeddable_backends ()
250 Returns the set of backends that are embeddable in other event loops. This
251 value is platform-specific but can include backends not available on the
252 current system. To find which embeddable backends might be supported on
253 the current system, you would need to look at C<ev_embeddable_backends ()
254 & ev_supported_backends ()>, likewise for recommended ones.
256 See the description of C<ev_embed> watchers for more info.
258 =item ev_set_allocator (void *(*cb)(void *ptr, long size) throw ())
260 Sets the allocation function to use (the prototype is similar - the
261 semantics are identical to the C<realloc> C89/SuS/POSIX function). It is
262 used to allocate and free memory (no surprises here). If it returns zero
263 when memory needs to be allocated (C<size != 0>), the library might abort
264 or take some potentially destructive action.
266 Since some systems (at least OpenBSD and Darwin) fail to implement
267 correct C<realloc> semantics, libev will use a wrapper around the system
268 C<realloc> and C<free> functions by default.
270 You could override this function in high-availability programs to, say,
271 free some memory if it cannot allocate memory, to use a special allocator,
272 or even to sleep a while and retry until some memory is available.
274 Example: The following is the C<realloc> function that libev itself uses
275 which should work with C<realloc> and C<free> functions of all kinds and
276 is probably a good basis for your own implementation.
278 static void *
279 ev_realloc_emul (void *ptr, long size) EV_NOEXCEPT
280 {
281 if (size)
282 return realloc (ptr, size);
284 free (ptr);
285 return 0;
286 }
288 Example: Replace the libev allocator with one that waits a bit and then
289 retries.
291 static void *
292 persistent_realloc (void *ptr, size_t size)
293 {
294 if (!size)
295 {
296 free (ptr);
297 return 0;
298 }
300 for (;;)
301 {
302 void *newptr = realloc (ptr, size);
304 if (newptr)
305 return newptr;
307 sleep (60);
308 }
309 }
311 ...
312 ev_set_allocator (persistent_realloc);
314 =item ev_set_syserr_cb (void (*cb)(const char *msg) throw ())
316 Set the callback function to call on a retryable system call error (such
317 as failed select, poll, epoll_wait). The message is a printable string
318 indicating the system call or subsystem causing the problem. If this
319 callback is set, then libev will expect it to remedy the situation, no
320 matter what, when it returns. That is, libev will generally retry the
321 requested operation, or, if the condition doesn't go away, do bad stuff
322 (such as abort).
324 Example: This is basically the same thing that libev does internally, too.
326 static void
327 fatal_error (const char *msg)
328 {
329 perror (msg);
330 abort ();
331 }
333 ...
334 ev_set_syserr_cb (fatal_error);
336 =item ev_feed_signal (int signum)
338 This function can be used to "simulate" a signal receive. It is completely
339 safe to call this function at any time, from any context, including signal
340 handlers or random threads.
342 Its main use is to customise signal handling in your process, especially
343 in the presence of threads. For example, you could block signals
344 by default in all threads (and specifying C<EVFLAG_NOSIGMASK> when
345 creating any loops), and in one thread, use C<sigwait> or any other
346 mechanism to wait for signals, then "deliver" them to libev by calling
347 C<ev_feed_signal>.
349 =back
353 An event loop is described by a C<struct ev_loop *> (the C<struct> is
354 I<not> optional in this case unless libev 3 compatibility is disabled, as
355 libev 3 had an C<ev_loop> function colliding with the struct name).
357 The library knows two types of such loops, the I<default> loop, which
358 supports child process events, and dynamically created event loops which
359 do not.
361 =over 4
363 =item struct ev_loop *ev_default_loop (unsigned int flags)
365 This returns the "default" event loop object, which is what you should
366 normally use when you just need "the event loop". Event loop objects and
367 the C<flags> parameter are described in more detail in the entry for
368 C<ev_loop_new>.
370 If the default loop is already initialised then this function simply
371 returns it (and ignores the flags. If that is troubling you, check
372 C<ev_backend ()> afterwards). Otherwise it will create it with the given
373 flags, which should almost always be C<0>, unless the caller is also the
374 one calling C<ev_run> or otherwise qualifies as "the main program".
376 If you don't know what event loop to use, use the one returned from this
377 function (or via the C<EV_DEFAULT> macro).
379 Note that this function is I<not> thread-safe, so if you want to use it
380 from multiple threads, you have to employ some kind of mutex (note also
381 that this case is unlikely, as loops cannot be shared easily between
382 threads anyway).
384 The default loop is the only loop that can handle C<ev_child> watchers,
385 and to do this, it always registers a handler for C<SIGCHLD>. If this is
386 a problem for your application you can either create a dynamic loop with
387 C<ev_loop_new> which doesn't do that, or you can simply overwrite the
388 C<SIGCHLD> signal handler I<after> calling C<ev_default_init>.
390 Example: This is the most typical usage.
392 if (!ev_default_loop (0))
393 fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
395 Example: Restrict libev to the select and poll backends, and do not allow
396 environment settings to be taken into account:
400 =item struct ev_loop *ev_loop_new (unsigned int flags)
402 This will create and initialise a new event loop object. If the loop
403 could not be initialised, returns false.
405 This function is thread-safe, and one common way to use libev with
406 threads is indeed to create one loop per thread, and using the default
407 loop in the "main" or "initial" thread.
409 The flags argument can be used to specify special behaviour or specific
410 backends to use, and is usually specified as C<0> (or C<EVFLAG_AUTO>).
412 The following flags are supported:
414 =over 4
416 =item C<EVFLAG_AUTO>
418 The default flags value. Use this if you have no clue (it's the right
419 thing, believe me).
421 =item C<EVFLAG_NOENV>
423 If this flag bit is or'ed into the flag value (or the program runs setuid
424 or setgid) then libev will I<not> look at the environment variable
425 C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will
426 override the flags completely if it is found in the environment. This is
427 useful to try out specific backends to test their performance, to work
428 around bugs, or to make libev threadsafe (accessing environment variables
429 cannot be done in a threadsafe way, but usually it works if no other
430 thread modifies them).
434 Instead of calling C<ev_loop_fork> manually after a fork, you can also
435 make libev check for a fork in each iteration by enabling this flag.
437 This works by calling C<getpid ()> on every iteration of the loop,
438 and thus this might slow down your event loop if you do a lot of loop
439 iterations and little real work, but is usually not noticeable (on my
440 GNU/Linux system for example, C<getpid> is actually a simple 5-insn
441 sequence without a system call and thus I<very> fast, but my GNU/Linux
442 system also has C<pthread_atfork> which is even faster). (Update: glibc
443 versions 2.25 apparently removed the C<getpid> optimisation again).
445 The big advantage of this flag is that you can forget about fork (and
446 forget about forgetting to tell libev about forking, although you still
447 have to ignore C<SIGPIPE>) when you use this flag.
449 This flag setting cannot be overridden or specified in the C<LIBEV_FLAGS>
450 environment variable.
454 When this flag is specified, then libev will not attempt to use the
455 I<inotify> API for its C<ev_stat> watchers. Apart from debugging and
456 testing, this flag can be useful to conserve inotify file descriptors, as
457 otherwise each loop using C<ev_stat> watchers consumes one inotify handle.
461 When this flag is specified, then libev will attempt to use the
462 I<signalfd> API for its C<ev_signal> (and C<ev_child>) watchers. This API
463 delivers signals synchronously, which makes it both faster and might make
464 it possible to get the queued signal data. It can also simplify signal
465 handling with threads, as long as you properly block signals in your
466 threads that are not interested in handling them.
468 Signalfd will not be used by default as this changes your signal mask, and
469 there are a lot of shoddy libraries and programs (glib's threadpool for
470 example) that can't properly initialise their signal masks.
474 When this flag is specified, then libev will avoid to modify the signal
475 mask. Specifically, this means you have to make sure signals are unblocked
476 when you want to receive them.
478 This behaviour is useful when you want to do your own signal handling, or
479 want to handle signals only in specific threads and want to avoid libev
480 unblocking the signals.
482 It's also required by POSIX in a threaded program, as libev calls
483 C<sigprocmask>, whose behaviour is officially unspecified.
487 When this flag is specified, the libev will avoid using a C<timerfd> to
488 detect time jumps. It will still be able to detect time jumps, but takes
489 longer and has a lower accuracy in doing so, but saves a file descriptor
490 per loop.
492 The current implementation only tries to use a C<timerfd> when the first
493 C<ev_periodic> watcher is started and falls back on other methods if it
494 cannot be created, but this behaviour might change in the future.
496 =item C<EVBACKEND_SELECT> (value 1, portable select backend)
498 This is your standard select(2) backend. Not I<completely> standard, as
499 libev tries to roll its own fd_set with no limits on the number of fds,
500 but if that fails, expect a fairly low limit on the number of fds when
501 using this backend. It doesn't scale too well (O(highest_fd)), but its
502 usually the fastest backend for a low number of (low-numbered :) fds.
504 To get good performance out of this backend you need a high amount of
505 parallelism (most of the file descriptors should be busy). If you are
506 writing a server, you should C<accept ()> in a loop to accept as many
507 connections as possible during one iteration. You might also want to have
508 a look at C<ev_set_io_collect_interval ()> to increase the amount of
509 readiness notifications you get per iteration.
511 This backend maps C<EV_READ> to the C<readfds> set and C<EV_WRITE> to the
512 C<writefds> set (and to work around Microsoft Windows bugs, also onto the
513 C<exceptfds> set on that platform).
515 =item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows)
517 And this is your standard poll(2) backend. It's more complicated
518 than select, but handles sparse fds better and has no artificial
519 limit on the number of fds you can use (except it will slow down
520 considerably with a lot of inactive fds). It scales similarly to select,
521 i.e. O(total_fds). See the entry for C<EVBACKEND_SELECT>, above, for
522 performance tips.
524 This backend maps C<EV_READ> to C<POLLIN | POLLERR | POLLHUP>, and
527 =item C<EVBACKEND_EPOLL> (value 4, Linux)
529 Use the Linux-specific epoll(7) interface (for both pre- and post-2.6.9
530 kernels).
532 For few fds, this backend is a bit little slower than poll and select, but
533 it scales phenomenally better. While poll and select usually scale like
534 O(total_fds) where total_fds is the total number of fds (or the highest
535 fd), epoll scales either O(1) or O(active_fds).
537 The epoll mechanism deserves honorable mention as the most misdesigned
538 of the more advanced event mechanisms: mere annoyances include silently
539 dropping file descriptors, requiring a system call per change per file
540 descriptor (and unnecessary guessing of parameters), problems with dup,
541 returning before the timeout value, resulting in additional iterations
542 (and only giving 5ms accuracy while select on the same platform gives
543 0.1ms) and so on. The biggest issue is fork races, however - if a program
544 forks then I<both> parent and child process have to recreate the epoll
545 set, which can take considerable time (one syscall per file descriptor)
546 and is of course hard to detect.
548 Epoll is also notoriously buggy - embedding epoll fds I<should> work,
549 but of course I<doesn't>, and epoll just loves to report events for
550 totally I<different> file descriptors (even already closed ones, so
551 one cannot even remove them from the set) than registered in the set
552 (especially on SMP systems). Libev tries to counter these spurious
553 notifications by employing an additional generation counter and comparing
554 that against the events to filter out spurious ones, recreating the set
555 when required. Epoll also erroneously rounds down timeouts, but gives you
556 no way to know when and by how much, so sometimes you have to busy-wait
557 because epoll returns immediately despite a nonzero timeout. And last
558 not least, it also refuses to work with some file descriptors which work
559 perfectly fine with C<select> (files, many character devices...).
561 Epoll is truly the train wreck among event poll mechanisms, a frankenpoll,
562 cobbled together in a hurry, no thought to design or interaction with
563 others. Oh, the pain, will it ever stop...
565 While stopping, setting and starting an I/O watcher in the same iteration
566 will result in some caching, there is still a system call per such
567 incident (because the same I<file descriptor> could point to a different
568 I<file description> now), so its best to avoid that. Also, C<dup ()>'ed
569 file descriptors might not work very well if you register events for both
570 file descriptors.
572 Best performance from this backend is achieved by not unregistering all
573 watchers for a file descriptor until it has been closed, if possible,
574 i.e. keep at least one watcher active per fd at all times. Stopping and
575 starting a watcher (without re-setting it) also usually doesn't cause
576 extra overhead. A fork can both result in spurious notifications as well
577 as in libev having to destroy and recreate the epoll object, which can
578 take considerable time and thus should be avoided.
580 All this means that, in practice, C<EVBACKEND_SELECT> can be as fast or
581 faster than epoll for maybe up to a hundred file descriptors, depending on
582 the usage. So sad.
584 While nominally embeddable in other event loops, this feature is broken in
585 a lot of kernel revisions, but probably(!) works in current versions.
587 This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
590 =item C<EVBACKEND_LINUXAIO> (value 64, Linux)
592 Use the Linux-specific Linux AIO (I<not> C<< aio(7) >> but C<<
593 io_submit(2) >>) event interface available in post-4.18 kernels (but libev
594 only tries to use it in 4.19+).
596 This is another Linux train wreck of an event interface.
598 If this backend works for you (as of this writing, it was very
599 experimental), it is the best event interface available on Linux and might
600 be well worth enabling it - if it isn't available in your kernel this will
601 be detected and this backend will be skipped.
603 This backend can batch oneshot requests and supports a user-space ring
604 buffer to receive events. It also doesn't suffer from most of the design
605 problems of epoll (such as not being able to remove event sources from
606 the epoll set), and generally sounds too good to be true. Because, this
607 being the Linux kernel, of course it suffers from a whole new set of
608 limitations, forcing you to fall back to epoll, inheriting all its design
609 issues.
611 For one, it is not easily embeddable (but probably could be done using
612 an event fd at some extra overhead). It also is subject to a system wide
613 limit that can be configured in F</proc/sys/fs/aio-max-nr>. If no AIO
614 requests are left, this backend will be skipped during initialisation, and
615 will switch to epoll when the loop is active.
617 Most problematic in practice, however, is that not all file descriptors
618 work with it. For example, in Linux 5.1, TCP sockets, pipes, event fds,
619 files, F</dev/null> and many others are supported, but ttys do not work
620 properly (a known bug that the kernel developers don't care about, see
621 L<>), so this is not
622 (yet?) a generic event polling interface.
624 Overall, it seems the Linux developers just don't want it to have a
625 generic event handling mechanism other than C<select> or C<poll>.
627 To work around all these problem, the current version of libev uses its
628 epoll backend as a fallback for file descriptor types that do not work. Or
629 falls back completely to epoll if the kernel acts up.
631 This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
634 =item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)
636 Kqueue deserves special mention, as at the time this backend was
637 implemented, it was broken on all BSDs except NetBSD (usually it doesn't
638 work reliably with anything but sockets and pipes, except on Darwin,
639 where of course it's completely useless). Unlike epoll, however, whose
640 brokenness is by design, these kqueue bugs can be (and mostly have been)
641 fixed without API changes to existing programs. For this reason it's not
642 being "auto-detected" on all platforms unless you explicitly specify it
643 in the flags (i.e. using C<EVBACKEND_KQUEUE>) or libev was compiled on a
644 known-to-be-good (-enough) system like NetBSD.
646 You still can embed kqueue into a normal poll or select backend and use it
647 only for sockets (after having made sure that sockets work with kqueue on
648 the target platform). See C<ev_embed> watchers for more info.
650 It scales in the same way as the epoll backend, but the interface to the
651 kernel is more efficient (which says nothing about its actual speed, of
652 course). While stopping, setting and starting an I/O watcher does never
653 cause an extra system call as with C<EVBACKEND_EPOLL>, it still adds up to
654 two event changes per incident. Support for C<fork ()> is very bad (you
655 might have to leak fds on fork, but it's more sane than epoll) and it
656 drops fds silently in similarly hard-to-detect cases.
658 This backend usually performs well under most conditions.
660 While nominally embeddable in other event loops, this doesn't work
661 everywhere, so you might need to test for this. And since it is broken
662 almost everywhere, you should only use it when you have a lot of sockets
663 (for which it usually works), by embedding it into another event loop
664 (e.g. C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> (but C<poll> is of course
665 also broken on OS X)) and, did I mention it, using it only for sockets.
667 This backend maps C<EV_READ> into an C<EVFILT_READ> kevent with
668 C<NOTE_EOF>, and C<EV_WRITE> into an C<EVFILT_WRITE> kevent with
669 C<NOTE_EOF>.
671 =item C<EVBACKEND_DEVPOLL> (value 16, Solaris 8)
673 This is not implemented yet (and might never be, unless you send me an
674 implementation). According to reports, C</dev/poll> only supports sockets
675 and is not embeddable, which would limit the usefulness of this backend
676 immensely.
678 =item C<EVBACKEND_PORT> (value 32, Solaris 10)
680 This uses the Solaris 10 event port mechanism. As with everything on
681 Solaris, it's really slow, but it still scales very well (O(active_fds)).
683 While this backend scales well, it requires one system call per active
684 file descriptor per loop iteration. For small and medium numbers of file
685 descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend
686 might perform better.
688 On the positive side, this backend actually performed fully to
689 specification in all tests and is fully embeddable, which is a rare feat
690 among the OS-specific backends (I vastly prefer correctness over speed
691 hacks).
693 On the negative side, the interface is I<bizarre> - so bizarre that
694 even sun itself gets it wrong in their code examples: The event polling
695 function sometimes returns events to the caller even though an error
696 occurred, but with no indication whether it has done so or not (yes, it's
697 even documented that way) - deadly for edge-triggered interfaces where you
698 absolutely have to know whether an event occurred or not because you have
699 to re-arm the watcher.
701 Fortunately libev seems to be able to work around these idiocies.
703 This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
706 =item C<EVBACKEND_ALL>
708 Try all backends (even potentially broken ones that wouldn't be tried
709 with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as
712 It is definitely not recommended to use this flag, use whatever
713 C<ev_recommended_backends ()> returns, or simply do not specify a backend
714 at all.
718 Not a backend at all, but a mask to select all backend bits from a
719 C<flags> value, in case you want to mask out any backends from a flags
720 value (e.g. when modifying the C<LIBEV_FLAGS> environment variable).
722 =back
724 If one or more of the backend flags are or'ed into the flags value,
725 then only these backends will be tried (in the reverse order as listed
726 here). If none are specified, all backends in C<ev_recommended_backends
727 ()> will be tried.
729 Example: Try to create a event loop that uses epoll and nothing else.
731 struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV);
732 if (!epoller)
733 fatal ("no epoll found here, maybe it hides under your chair");
735 Example: Use whatever libev has to offer, but make sure that kqueue is
736 used if available.
738 struct ev_loop *loop = ev_loop_new (ev_recommended_backends () | EVBACKEND_KQUEUE);
740 Example: Similarly, on linux, you mgiht want to take advantage of the
741 linux aio backend if possible, but fall back to something else if that
742 isn't available.
744 struct ev_loop *loop = ev_loop_new (ev_recommended_backends () | EVBACKEND_LINUXAIO);
746 =item ev_loop_destroy (loop)
748 Destroys an event loop object (frees all memory and kernel state
749 etc.). None of the active event watchers will be stopped in the normal
750 sense, so e.g. C<ev_is_active> might still return true. It is your
751 responsibility to either stop all watchers cleanly yourself I<before>
752 calling this function, or cope with the fact afterwards (which is usually
753 the easiest thing, you can just ignore the watchers and/or C<free ()> them
754 for example).
756 Note that certain global state, such as signal state (and installed signal
757 handlers), will not be freed by this function, and related watchers (such
758 as signal and child watchers) would need to be stopped manually.
760 This function is normally used on loop objects allocated by
761 C<ev_loop_new>, but it can also be used on the default loop returned by
762 C<ev_default_loop>, in which case it is not thread-safe.
764 Note that it is not advisable to call this function on the default loop
765 except in the rare occasion where you really need to free its resources.
766 If you need dynamically allocated loops it is better to use C<ev_loop_new>
767 and C<ev_loop_destroy>.
769 =item ev_loop_fork (loop)
771 This function sets a flag that causes subsequent C<ev_run> iterations
772 to reinitialise the kernel state for backends that have one. Despite
773 the name, you can call it anytime you are allowed to start or stop
774 watchers (except inside an C<ev_prepare> callback), but it makes most
775 sense after forking, in the child process. You I<must> call it (or use
776 C<EVFLAG_FORKCHECK>) in the child before resuming or calling C<ev_run>.
778 In addition, if you want to reuse a loop (via this function or
779 C<EVFLAG_FORKCHECK>), you I<also> have to ignore C<SIGPIPE>.
781 Again, you I<have> to call it on I<any> loop that you want to re-use after
782 a fork, I<even if you do not plan to use the loop in the parent>. This is
783 because some kernel interfaces *cough* I<kqueue> *cough* do funny things
784 during fork.
786 On the other hand, you only need to call this function in the child
787 process if and only if you want to use the event loop in the child. If
788 you just fork+exec or create a new loop in the child, you don't have to
789 call it at all (in fact, C<epoll> is so badly broken that it makes a
790 difference, but libev will usually detect this case on its own and do a
791 costly reset of the backend).
793 The function itself is quite fast and it's usually not a problem to call
794 it just in case after a fork.
796 Example: Automate calling C<ev_loop_fork> on the default loop when
797 using pthreads.
799 static void
800 post_fork_child (void)
801 {
802 ev_loop_fork (EV_DEFAULT);
803 }
805 ...
806 pthread_atfork (0, 0, post_fork_child);
808 =item int ev_is_default_loop (loop)
810 Returns true when the given loop is, in fact, the default loop, and false
811 otherwise.
813 =item unsigned int ev_iteration (loop)
815 Returns the current iteration count for the event loop, which is identical
816 to the number of times libev did poll for new events. It starts at C<0>
817 and happily wraps around with enough iterations.
819 This value can sometimes be useful as a generation counter of sorts (it
820 "ticks" the number of loop iterations), as it roughly corresponds with
821 C<ev_prepare> and C<ev_check> calls - and is incremented between the
822 prepare and check phases.
824 =item unsigned int ev_depth (loop)
826 Returns the number of times C<ev_run> was entered minus the number of
827 times C<ev_run> was exited normally, in other words, the recursion depth.
829 Outside C<ev_run>, this number is zero. In a callback, this number is
830 C<1>, unless C<ev_run> was invoked recursively (or from another thread),
831 in which case it is higher.
833 Leaving C<ev_run> abnormally (setjmp/longjmp, cancelling the thread,
834 throwing an exception etc.), doesn't count as "exit" - consider this
835 as a hint to avoid such ungentleman-like behaviour unless it's really
836 convenient, in which case it is fully supported.
838 =item unsigned int ev_backend (loop)
840 Returns one of the C<EVBACKEND_*> flags indicating the event backend in
841 use.
843 =item ev_tstamp ev_now (loop)
845 Returns the current "event loop time", which is the time the event loop
846 received events and started processing them. This timestamp does not
847 change as long as callbacks are being processed, and this is also the base
848 time used for relative timers. You can treat it as the timestamp of the
849 event occurring (or more correctly, libev finding out about it).
851 =item ev_now_update (loop)
853 Establishes the current time by querying the kernel, updating the time
854 returned by C<ev_now ()> in the progress. This is a costly operation and
855 is usually done automatically within C<ev_run ()>.
857 This function is rarely useful, but when some event callback runs for a
858 very long time without entering the event loop, updating libev's idea of
859 the current time is a good idea.
861 See also L</The special problem of time updates> in the C<ev_timer> section.
863 =item ev_suspend (loop)
865 =item ev_resume (loop)
867 These two functions suspend and resume an event loop, for use when the
868 loop is not used for a while and timeouts should not be processed.
870 A typical use case would be an interactive program such as a game: When
871 the user presses C<^Z> to suspend the game and resumes it an hour later it
872 would be best to handle timeouts as if no time had actually passed while
873 the program was suspended. This can be achieved by calling C<ev_suspend>
874 in your C<SIGTSTP> handler, sending yourself a C<SIGSTOP> and calling
875 C<ev_resume> directly afterwards to resume timer processing.
877 Effectively, all C<ev_timer> watchers will be delayed by the time spend
878 between C<ev_suspend> and C<ev_resume>, and all C<ev_periodic> watchers
879 will be rescheduled (that is, they will lose any events that would have
880 occurred while suspended).
882 After calling C<ev_suspend> you B<must not> call I<any> function on the
883 given loop other than C<ev_resume>, and you B<must not> call C<ev_resume>
884 without a previous call to C<ev_suspend>.
886 Calling C<ev_suspend>/C<ev_resume> has the side effect of updating the
887 event loop time (see C<ev_now_update>).
889 =item bool ev_run (loop, int flags)
891 Finally, this is it, the event handler. This function usually is called
892 after you have initialised all your watchers and you want to start
893 handling events. It will ask the operating system for any new events, call
894 the watcher callbacks, and then repeat the whole process indefinitely: This
895 is why event loops are called I<loops>.
897 If the flags argument is specified as C<0>, it will keep handling events
898 until either no event watchers are active anymore or C<ev_break> was
899 called.
901 The return value is false if there are no more active watchers (which
902 usually means "all jobs done" or "deadlock"), and true in all other cases
903 (which usually means " you should call C<ev_run> again").
905 Please note that an explicit C<ev_break> is usually better than
906 relying on all watchers to be stopped when deciding when a program has
907 finished (especially in interactive programs), but having a program
908 that automatically loops as long as it has to and no longer by virtue
909 of relying on its watchers stopping correctly, that is truly a thing of
910 beauty.
912 This function is I<mostly> exception-safe - you can break out of a
913 C<ev_run> call by calling C<longjmp> in a callback, throwing a C++
914 exception and so on. This does not decrement the C<ev_depth> value, nor
915 will it clear any outstanding C<EVBREAK_ONE> breaks.
917 A flags value of C<EVRUN_NOWAIT> will look for new events, will handle
918 those events and any already outstanding ones, but will not wait and
919 block your process in case there are no events and will return after one
920 iteration of the loop. This is sometimes useful to poll and handle new
921 events while doing lengthy calculations, to keep the program responsive.
923 A flags value of C<EVRUN_ONCE> will look for new events (waiting if
924 necessary) and will handle those and any already outstanding ones. It
925 will block your process until at least one new event arrives (which could
926 be an event internal to libev itself, so there is no guarantee that a
927 user-registered callback will be called), and will return after one
928 iteration of the loop.
930 This is useful if you are waiting for some external event in conjunction
931 with something not expressible using other libev watchers (i.e. "roll your
932 own C<ev_run>"). However, a pair of C<ev_prepare>/C<ev_check> watchers is
933 usually a better approach for this kind of thing.
935 Here are the gory details of what C<ev_run> does (this is for your
936 understanding, not a guarantee that things will work exactly like this in
937 future versions):
939 - Increment loop depth.
940 - Reset the ev_break status.
941 - Before the first iteration, call any pending watchers.
942 LOOP:
943 - If EVFLAG_FORKCHECK was used, check for a fork.
944 - If a fork was detected (by any means), queue and call all fork watchers.
945 - Queue and call all prepare watchers.
946 - If ev_break was called, goto FINISH.
947 - If we have been forked, detach and recreate the kernel state
948 as to not disturb the other process.
949 - Update the kernel state with all outstanding changes.
950 - Update the "event loop time" (ev_now ()).
951 - Calculate for how long to sleep or block, if at all
952 (active idle watchers, EVRUN_NOWAIT or not having
953 any active watchers at all will result in not sleeping).
954 - Sleep if the I/O and timer collect interval say so.
955 - Increment loop iteration counter.
956 - Block the process, waiting for any events.
957 - Queue all outstanding I/O (fd) events.
958 - Update the "event loop time" (ev_now ()), and do time jump adjustments.
959 - Queue all expired timers.
960 - Queue all expired periodics.
961 - Queue all idle watchers with priority higher than that of pending events.
962 - Queue all check watchers.
963 - Call all queued watchers in reverse order (i.e. check watchers first).
964 Signals, async and child watchers are implemented as I/O watchers, and
965 will be handled here by queueing them when their watcher gets executed.
966 - If ev_break has been called, or EVRUN_ONCE or EVRUN_NOWAIT
967 were used, or there are no active watchers, goto FINISH, otherwise
968 continue with step LOOP.
970 - Reset the ev_break status iff it was EVBREAK_ONE.
971 - Decrement the loop depth.
972 - Return.
974 Example: Queue some jobs and then loop until no events are outstanding
975 anymore.
977 ... queue jobs here, make sure they register event watchers as long
978 ... as they still have work to do (even an idle watcher will do..)
979 ev_run (my_loop, 0);
980 ... jobs done or somebody called break. yeah!
982 =item ev_break (loop, how)
984 Can be used to make a call to C<ev_run> return early (but only after it
985 has processed all outstanding events). The C<how> argument must be either
986 C<EVBREAK_ONE>, which will make the innermost C<ev_run> call return, or
987 C<EVBREAK_ALL>, which will make all nested C<ev_run> calls return.
989 This "break state" will be cleared on the next call to C<ev_run>.
991 It is safe to call C<ev_break> from outside any C<ev_run> calls, too, in
992 which case it will have no effect.
994 =item ev_ref (loop)
996 =item ev_unref (loop)
998 Ref/unref can be used to add or remove a reference count on the event
999 loop: Every watcher keeps one reference, and as long as the reference
1000 count is nonzero, C<ev_run> will not return on its own.
1002 This is useful when you have a watcher that you never intend to
1003 unregister, but that nevertheless should not keep C<ev_run> from
1004 returning. In such a case, call C<ev_unref> after starting, and C<ev_ref>
1005 before stopping it.
1007 As an example, libev itself uses this for its internal signal pipe: It
1008 is not visible to the libev user and should not keep C<ev_run> from
1009 exiting if no event watchers registered by it are active. It is also an
1010 excellent way to do this for generic recurring timers or from within
1011 third-party libraries. Just remember to I<unref after start> and I<ref
1012 before stop> (but only if the watcher wasn't active before, or was active
1013 before, respectively. Note also that libev might stop watchers itself
1014 (e.g. non-repeating timers) in which case you have to C<ev_ref>
1015 in the callback).
1017 Example: Create a signal watcher, but keep it from keeping C<ev_run>
1018 running when nothing else is active.
1020 ev_signal exitsig;
1021 ev_signal_init (&exitsig, sig_cb, SIGINT);
1022 ev_signal_start (loop, &exitsig);
1023 ev_unref (loop);
1025 Example: For some weird reason, unregister the above signal handler again.
1027 ev_ref (loop);
1028 ev_signal_stop (loop, &exitsig);
1030 =item ev_set_io_collect_interval (loop, ev_tstamp interval)
1032 =item ev_set_timeout_collect_interval (loop, ev_tstamp interval)
1034 These advanced functions influence the time that libev will spend waiting
1035 for events. Both time intervals are by default C<0>, meaning that libev
1036 will try to invoke timer/periodic callbacks and I/O callbacks with minimum
1037 latency.
1039 Setting these to a higher value (the C<interval> I<must> be >= C<0>)
1040 allows libev to delay invocation of I/O and timer/periodic callbacks
1041 to increase efficiency of loop iterations (or to increase power-saving
1042 opportunities).
1044 The idea is that sometimes your program runs just fast enough to handle
1045 one (or very few) event(s) per loop iteration. While this makes the
1046 program responsive, it also wastes a lot of CPU time to poll for new
1047 events, especially with backends like C<select ()> which have a high
1048 overhead for the actual polling but can deliver many events at once.
1050 By setting a higher I<io collect interval> you allow libev to spend more
1051 time collecting I/O events, so you can handle more events per iteration,
1052 at the cost of increasing latency. Timeouts (both C<ev_periodic> and
1053 C<ev_timer>) will not be affected. Setting this to a non-null value will
1054 introduce an additional C<ev_sleep ()> call into most loop iterations. The
1055 sleep time ensures that libev will not poll for I/O events more often then
1056 once per this interval, on average (as long as the host time resolution is
1057 good enough).
1059 Likewise, by setting a higher I<timeout collect interval> you allow libev
1060 to spend more time collecting timeouts, at the expense of increased
1061 latency/jitter/inexactness (the watcher callback will be called
1062 later). C<ev_io> watchers will not be affected. Setting this to a non-null
1063 value will not introduce any overhead in libev.
1065 Many (busy) programs can usually benefit by setting the I/O collect
1066 interval to a value near C<0.1> or so, which is often enough for
1067 interactive servers (of course not for games), likewise for timeouts. It
1068 usually doesn't make much sense to set it to a lower value than C<0.01>,
1069 as this approaches the timing granularity of most systems. Note that if
1070 you do transactions with the outside world and you can't increase the
1071 parallelity, then this setting will limit your transaction rate (if you
1072 need to poll once per transaction and the I/O collect interval is 0.01,
1073 then you can't do more than 100 transactions per second).
1075 Setting the I<timeout collect interval> can improve the opportunity for
1076 saving power, as the program will "bundle" timer callback invocations that
1077 are "near" in time together, by delaying some, thus reducing the number of
1078 times the process sleeps and wakes up again. Another useful technique to
1079 reduce iterations/wake-ups is to use C<ev_periodic> watchers and make sure
1080 they fire on, say, one-second boundaries only.
1082 Example: we only need 0.1s timeout granularity, and we wish not to poll
1083 more often than 100 times per second:
1085 ev_set_timeout_collect_interval (EV_DEFAULT_UC_ 0.1);
1086 ev_set_io_collect_interval (EV_DEFAULT_UC_ 0.01);
1088 =item ev_invoke_pending (loop)
1090 This call will simply invoke all pending watchers while resetting their
1091 pending state. Normally, C<ev_run> does this automatically when required,
1092 but when overriding the invoke callback this call comes handy. This
1093 function can be invoked from a watcher - this can be useful for example
1094 when you want to do some lengthy calculation and want to pass further
1095 event handling to another thread (you still have to make sure only one
1096 thread executes within C<ev_invoke_pending> or C<ev_run> of course).
1098 =item int ev_pending_count (loop)
1100 Returns the number of pending watchers - zero indicates that no watchers
1101 are pending.
1103 =item ev_set_invoke_pending_cb (loop, void (*invoke_pending_cb)(EV_P))
1105 This overrides the invoke pending functionality of the loop: Instead of
1106 invoking all pending watchers when there are any, C<ev_run> will call
1107 this callback instead. This is useful, for example, when you want to
1108 invoke the actual watchers inside another context (another thread etc.).
1110 If you want to reset the callback, use C<ev_invoke_pending> as new
1111 callback.
1113 =item ev_set_loop_release_cb (loop, void (*release)(EV_P) throw (), void (*acquire)(EV_P) throw ())
1115 Sometimes you want to share the same loop between multiple threads. This
1116 can be done relatively simply by putting mutex_lock/unlock calls around
1117 each call to a libev function.
1119 However, C<ev_run> can run an indefinite time, so it is not feasible
1120 to wait for it to return. One way around this is to wake up the event
1121 loop via C<ev_break> and C<ev_async_send>, another way is to set these
1122 I<release> and I<acquire> callbacks on the loop.
1124 When set, then C<release> will be called just before the thread is
1125 suspended waiting for new events, and C<acquire> is called just
1126 afterwards.
1128 Ideally, C<release> will just call your mutex_unlock function, and
1129 C<acquire> will just call the mutex_lock function again.
1131 While event loop modifications are allowed between invocations of
1132 C<release> and C<acquire> (that's their only purpose after all), no
1133 modifications done will affect the event loop, i.e. adding watchers will
1134 have no effect on the set of file descriptors being watched, or the time
1135 waited. Use an C<ev_async> watcher to wake up C<ev_run> when you want it
1136 to take note of any changes you made.
1138 In theory, threads executing C<ev_run> will be async-cancel safe between
1139 invocations of C<release> and C<acquire>.
1141 See also the locking example in the C<THREADS> section later in this
1142 document.
1144 =item ev_set_userdata (loop, void *data)
1146 =item void *ev_userdata (loop)
1148 Set and retrieve a single C<void *> associated with a loop. When
1149 C<ev_set_userdata> has never been called, then C<ev_userdata> returns
1150 C<0>.
1152 These two functions can be used to associate arbitrary data with a loop,
1153 and are intended solely for the C<invoke_pending_cb>, C<release> and
1154 C<acquire> callbacks described above, but of course can be (ab-)used for
1155 any other purpose as well.
1157 =item ev_verify (loop)
1159 This function only does something when C<EV_VERIFY> support has been
1160 compiled in, which is the default for non-minimal builds. It tries to go
1161 through all internal structures and checks them for validity. If anything
1162 is found to be inconsistent, it will print an error message to standard
1163 error and call C<abort ()>.
1165 This can be used to catch bugs inside libev itself: under normal
1166 circumstances, this function will never abort as of course libev keeps its
1167 data structures consistent.
1169 =back
1174 In the following description, uppercase C<TYPE> in names stands for the
1175 watcher type, e.g. C<ev_TYPE_start> can mean C<ev_timer_start> for timer
1176 watchers and C<ev_io_start> for I/O watchers.
1178 A watcher is an opaque structure that you allocate and register to record
1179 your interest in some event. To make a concrete example, imagine you want
1180 to wait for STDIN to become readable, you would create an C<ev_io> watcher
1181 for that:
1183 static void my_cb (struct ev_loop *loop, ev_io *w, int revents)
1184 {
1185 ev_io_stop (w);
1186 ev_break (loop, EVBREAK_ALL);
1187 }
1189 struct ev_loop *loop = ev_default_loop (0);
1191 ev_io stdin_watcher;
1193 ev_init (&stdin_watcher, my_cb);
1194 ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ);
1195 ev_io_start (loop, &stdin_watcher);
1197 ev_run (loop, 0);
1199 As you can see, you are responsible for allocating the memory for your
1200 watcher structures (and it is I<usually> a bad idea to do this on the
1201 stack).
1203 Each watcher has an associated watcher structure (called C<struct ev_TYPE>
1204 or simply C<ev_TYPE>, as typedefs are provided for all watcher structs).
1206 Each watcher structure must be initialised by a call to C<ev_init (watcher
1207 *, callback)>, which expects a callback to be provided. This callback is
1208 invoked each time the event occurs (or, in the case of I/O watchers, each
1209 time the event loop detects that the file descriptor given is readable
1210 and/or writable).
1212 Each watcher type further has its own C<< ev_TYPE_set (watcher *, ...) >>
1213 macro to configure it, with arguments specific to the watcher type. There
1214 is also a macro to combine initialisation and setting in one call: C<<
1215 ev_TYPE_init (watcher *, callback, ...) >>.
1217 To make the watcher actually watch out for events, you have to start it
1218 with a watcher-specific start function (C<< ev_TYPE_start (loop, watcher
1219 *) >>), and you can stop watching for events at any time by calling the
1220 corresponding stop function (C<< ev_TYPE_stop (loop, watcher *) >>.
1222 As long as your watcher is active (has been started but not stopped) you
1223 must not touch the values stored in it except when explicitly documented
1224 otherwise. Most specifically you must never reinitialise it or call its
1225 C<ev_TYPE_set> macro.
1227 Each and every callback receives the event loop pointer as first, the
1228 registered watcher structure as second, and a bitset of received events as
1229 third argument.
1231 The received events usually include a single bit per event type received
1232 (you can receive multiple events at the same time). The possible bit masks
1233 are:
1235 =over 4
1237 =item C<EV_READ>
1239 =item C<EV_WRITE>
1241 The file descriptor in the C<ev_io> watcher has become readable and/or
1242 writable.
1244 =item C<EV_TIMER>
1246 The C<ev_timer> watcher has timed out.
1248 =item C<EV_PERIODIC>
1250 The C<ev_periodic> watcher has timed out.
1252 =item C<EV_SIGNAL>
1254 The signal specified in the C<ev_signal> watcher has been received by a thread.
1256 =item C<EV_CHILD>
1258 The pid specified in the C<ev_child> watcher has received a status change.
1260 =item C<EV_STAT>
1262 The path specified in the C<ev_stat> watcher changed its attributes somehow.
1264 =item C<EV_IDLE>
1266 The C<ev_idle> watcher has determined that you have nothing better to do.
1268 =item C<EV_PREPARE>
1270 =item C<EV_CHECK>
1272 All C<ev_prepare> watchers are invoked just I<before> C<ev_run> starts to
1273 gather new events, and all C<ev_check> watchers are queued (not invoked)
1274 just after C<ev_run> has gathered them, but before it queues any callbacks
1275 for any received events. That means C<ev_prepare> watchers are the last
1276 watchers invoked before the event loop sleeps or polls for new events, and
1277 C<ev_check> watchers will be invoked before any other watchers of the same
1278 or lower priority within an event loop iteration.
1280 Callbacks of both watcher types can start and stop as many watchers as
1281 they want, and all of them will be taken into account (for example, a
1282 C<ev_prepare> watcher might start an idle watcher to keep C<ev_run> from
1283 blocking).
1285 =item C<EV_EMBED>
1287 The embedded event loop specified in the C<ev_embed> watcher needs attention.
1289 =item C<EV_FORK>
1291 The event loop has been resumed in the child process after fork (see
1292 C<ev_fork>).
1294 =item C<EV_CLEANUP>
1296 The event loop is about to be destroyed (see C<ev_cleanup>).
1298 =item C<EV_ASYNC>
1300 The given async watcher has been asynchronously notified (see C<ev_async>).
1302 =item C<EV_CUSTOM>
1304 Not ever sent (or otherwise used) by libev itself, but can be freely used
1305 by libev users to signal watchers (e.g. via C<ev_feed_event>).
1307 =item C<EV_ERROR>
1309 An unspecified error has occurred, the watcher has been stopped. This might
1310 happen because the watcher could not be properly started because libev
1311 ran out of memory, a file descriptor was found to be closed or any other
1312 problem. Libev considers these application bugs.
1314 You best act on it by reporting the problem and somehow coping with the
1315 watcher being stopped. Note that well-written programs should not receive
1316 an error ever, so when your watcher receives it, this usually indicates a
1317 bug in your program.
1319 Libev will usually signal a few "dummy" events together with an error, for
1320 example it might indicate that a fd is readable or writable, and if your
1321 callbacks is well-written it can just attempt the operation and cope with
1322 the error from read() or write(). This will not work in multi-threaded
1323 programs, though, as the fd could already be closed and reused for another
1324 thing, so beware.
1326 =back
1330 =over 4
1332 =item C<ev_init> (ev_TYPE *watcher, callback)
1334 This macro initialises the generic portion of a watcher. The contents
1335 of the watcher object can be arbitrary (so C<malloc> will do). Only
1336 the generic parts of the watcher are initialised, you I<need> to call
1337 the type-specific C<ev_TYPE_set> macro afterwards to initialise the
1338 type-specific parts. For each type there is also a C<ev_TYPE_init> macro
1339 which rolls both calls into one.
1341 You can reinitialise a watcher at any time as long as it has been stopped
1342 (or never started) and there are no pending events outstanding.
1344 The callback is always of type C<void (*)(struct ev_loop *loop, ev_TYPE *watcher,
1345 int revents)>.
1347 Example: Initialise an C<ev_io> watcher in two steps.
1349 ev_io w;
1350 ev_init (&w, my_cb);
1351 ev_io_set (&w, STDIN_FILENO, EV_READ);
1353 =item C<ev_TYPE_set> (ev_TYPE *watcher, [args])
1355 This macro initialises the type-specific parts of a watcher. You need to
1356 call C<ev_init> at least once before you call this macro, but you can
1357 call C<ev_TYPE_set> any number of times. You must not, however, call this
1358 macro on a watcher that is active (it can be pending, however, which is a
1359 difference to the C<ev_init> macro).
1361 Although some watcher types do not have type-specific arguments
1362 (e.g. C<ev_prepare>) you still need to call its C<set> macro.
1364 See C<ev_init>, above, for an example.
1366 =item C<ev_TYPE_init> (ev_TYPE *watcher, callback, [args])
1368 This convenience macro rolls both C<ev_init> and C<ev_TYPE_set> macro
1369 calls into a single call. This is the most convenient method to initialise
1370 a watcher. The same limitations apply, of course.
1372 Example: Initialise and set an C<ev_io> watcher in one step.
1374 ev_io_init (&w, my_cb, STDIN_FILENO, EV_READ);
1376 =item C<ev_TYPE_start> (loop, ev_TYPE *watcher)
1378 Starts (activates) the given watcher. Only active watchers will receive
1379 events. If the watcher is already active nothing will happen.
1381 Example: Start the C<ev_io> watcher that is being abused as example in this
1382 whole section.
1384 ev_io_start (EV_DEFAULT_UC, &w);
1386 =item C<ev_TYPE_stop> (loop, ev_TYPE *watcher)
1388 Stops the given watcher if active, and clears the pending status (whether
1389 the watcher was active or not).
1391 It is possible that stopped watchers are pending - for example,
1392 non-repeating timers are being stopped when they become pending - but
1393 calling C<ev_TYPE_stop> ensures that the watcher is neither active nor
1394 pending. If you want to free or reuse the memory used by the watcher it is
1395 therefore a good idea to always call its C<ev_TYPE_stop> function.
1397 =item bool ev_is_active (ev_TYPE *watcher)
1399 Returns a true value iff the watcher is active (i.e. it has been started
1400 and not yet been stopped). As long as a watcher is active you must not modify
1401 it unless documented otherwise.
1403 Obviously, it is safe to call this on an active watcher, or actually any
1404 watcher that is initialised.
1406 =item bool ev_is_pending (ev_TYPE *watcher)
1408 Returns a true value iff the watcher is pending, (i.e. it has outstanding
1409 events but its callback has not yet been invoked). As long as a watcher
1410 is pending (but not active) you must not call an init function on it (but
1411 C<ev_TYPE_set> is safe), you must not change its priority, and you must
1412 make sure the watcher is available to libev (e.g. you cannot C<free ()>
1413 it).
1415 It is safe to call this on any watcher in any state as long as it is
1416 initialised.
1418 =item callback ev_cb (ev_TYPE *watcher)
1420 Returns the callback currently set on the watcher.
1422 =item ev_set_cb (ev_TYPE *watcher, callback)
1424 Change the callback. You can change the callback at virtually any time
1425 (modulo threads).
1427 =item ev_set_priority (ev_TYPE *watcher, int priority)
1429 =item int ev_priority (ev_TYPE *watcher)
1431 Set and query the priority of the watcher. The priority is a small
1432 integer between C<EV_MAXPRI> (default: C<2>) and C<EV_MINPRI>
1433 (default: C<-2>). Pending watchers with higher priority will be invoked
1434 before watchers with lower priority, but priority will not keep watchers
1435 from being executed (except for C<ev_idle> watchers).
1437 If you need to suppress invocation when higher priority events are pending
1438 you need to look at C<ev_idle> watchers, which provide this functionality.
1440 You I<must not> change the priority of a watcher as long as it is active
1441 or pending. Reading the priority with C<ev_priority> is fine in any state.
1443 Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is
1444 fine, as long as you do not mind that the priority value you query might
1445 or might not have been clamped to the valid range.
1447 The default priority used by watchers when no priority has been set is
1448 always C<0>, which is supposed to not be too high and not be too low :).
1450 See L</WATCHER PRIORITY MODELS>, below, for a more thorough treatment of
1451 priorities.
1453 =item ev_invoke (loop, ev_TYPE *watcher, int revents)
1455 Invoke the C<watcher> with the given C<loop> and C<revents>. Neither
1456 C<loop> nor C<revents> need to be valid as long as the watcher callback
1457 can deal with that fact, as both are simply passed through to the
1458 callback.
1460 =item int ev_clear_pending (loop, ev_TYPE *watcher)
1462 If the watcher is pending, this function clears its pending status and
1463 returns its C<revents> bitset (as if its callback was invoked). If the
1464 watcher isn't pending it does nothing and returns C<0>.
1466 Sometimes it can be useful to "poll" a watcher instead of waiting for its
1467 callback to be invoked, which can be accomplished with this function.
1469 =item ev_feed_event (loop, ev_TYPE *watcher, int revents)
1471 Feeds the given event set into the event loop, as if the specified event
1472 had happened for the specified watcher (which must be a pointer to an
1473 initialised but not necessarily started event watcher, though it can be
1474 active). Obviously you must not free the watcher as long as it has pending
1475 events.
1477 Stopping the watcher, letting libev invoke it, or calling
1478 C<ev_clear_pending> will clear the pending event, even if the watcher was
1479 not started in the first place.
1481 See also C<ev_feed_fd_event> and C<ev_feed_signal_event> for related
1482 functions that do not need a watcher.
1484 =back
1489 =head2 WATCHER STATES
1491 There are various watcher states mentioned throughout this manual -
1492 active, pending and so on. In this section these states and the rules to
1493 transition between them will be described in more detail - and while these
1494 rules might look complicated, they usually do "the right thing".
1496 =over 4
1498 =item initialised
1500 Before a watcher can be registered with the event loop it has to be
1501 initialised. This can be done with a call to C<ev_TYPE_init>, or calls to
1502 C<ev_init> followed by the watcher-specific C<ev_TYPE_set> function.
1504 In this state it is simply some block of memory that is suitable for
1505 use in an event loop. It can be moved around, freed, reused etc. at
1506 will - as long as you either keep the memory contents intact, or call
1507 C<ev_TYPE_init> again.
1509 =item started/running/active
1511 Once a watcher has been started with a call to C<ev_TYPE_start> it becomes
1512 property of the event loop, and is actively waiting for events. While in
1513 this state it cannot be accessed (except in a few documented ways, such as
1514 stoping it), moved, freed or anything else - the only legal thing is to
1515 keep a pointer to it, and call libev functions on it that are documented
1516 to work on active watchers.
1518 As a rule of thumb, before accessing a member or calling any function on
1519 a watcher, it should be stopped (or freshly initialised). If that is not
1520 convenient, you can check the documentation for that function or member to
1521 see if it is safe to use on an active watcher.
1523 =item pending
1525 If a watcher is active and libev determines that an event it is interested
1526 in has occurred (such as a timer expiring), it will become pending. It
1527 will stay in this pending state until either it is explicitly stopped or
1528 its callback is about to be invoked, so it is not normally pending inside
1529 the watcher callback.
1531 Generally, the watcher might or might not be active while it is pending
1532 (for example, an expired non-repeating timer can be pending but no longer
1533 active). If it is pending but not active, it can be freely accessed (e.g.
1534 by calling C<ev_TYPE_set>), but it is still property of the event loop at
1535 this time, so cannot be moved, freed or reused. And if it is active the
1536 rules described in the previous item still apply.
1538 Explicitly stopping a watcher will also clear the pending state
1539 unconditionally, so it is safe to stop a watcher and then free it.
1541 It is also possible to feed an event on a watcher that is not active (e.g.
1542 via C<ev_feed_event>), in which case it becomes pending without being
1543 active.
1545 =item stopped
1547 A watcher can be stopped implicitly by libev (in which case it might still
1548 be pending), or explicitly by calling its C<ev_TYPE_stop> function. The
1549 latter will clear any pending state the watcher might be in, regardless
1550 of whether it was active or not, so stopping a watcher explicitly before
1551 freeing it is often a good idea.
1553 While stopped (and not pending) the watcher is essentially in the
1554 initialised state, that is, it can be reused, moved, modified in any way
1555 you wish (but when you trash the memory block, you need to C<ev_TYPE_init>
1556 it again).
1558 =back
1562 Many event loops support I<watcher priorities>, which are usually small
1563 integers that influence the ordering of event callback invocation
1564 between watchers in some way, all else being equal.
1566 In libev, watcher priorities can be set using C<ev_set_priority>. See its
1567 description for the more technical details such as the actual priority
1568 range.
1570 There are two common ways how these these priorities are being interpreted
1571 by event loops:
1573 In the more common lock-out model, higher priorities "lock out" invocation
1574 of lower priority watchers, which means as long as higher priority
1575 watchers receive events, lower priority watchers are not being invoked.
1577 The less common only-for-ordering model uses priorities solely to order
1578 callback invocation within a single event loop iteration: Higher priority
1579 watchers are invoked before lower priority ones, but they all get invoked
1580 before polling for new events.
1582 Libev uses the second (only-for-ordering) model for all its watchers
1583 except for idle watchers (which use the lock-out model).
1585 The rationale behind this is that implementing the lock-out model for
1586 watchers is not well supported by most kernel interfaces, and most event
1587 libraries will just poll for the same events again and again as long as
1588 their callbacks have not been executed, which is very inefficient in the
1589 common case of one high-priority watcher locking out a mass of lower
1590 priority ones.
1592 Static (ordering) priorities are most useful when you have two or more
1593 watchers handling the same resource: a typical usage example is having an
1594 C<ev_io> watcher to receive data, and an associated C<ev_timer> to handle
1595 timeouts. Under load, data might be received while the program handles
1596 other jobs, but since timers normally get invoked first, the timeout
1597 handler will be executed before checking for data. In that case, giving
1598 the timer a lower priority than the I/O watcher ensures that I/O will be
1599 handled first even under adverse conditions (which is usually, but not
1600 always, what you want).
1602 Since idle watchers use the "lock-out" model, meaning that idle watchers
1603 will only be executed when no same or higher priority watchers have
1604 received events, they can be used to implement the "lock-out" model when
1605 required.
1607 For example, to emulate how many other event libraries handle priorities,
1608 you can associate an C<ev_idle> watcher to each such watcher, and in
1609 the normal watcher callback, you just start the idle watcher. The real
1610 processing is done in the idle watcher callback. This causes libev to
1611 continuously poll and process kernel event data for the watcher, but when
1612 the lock-out case is known to be rare (which in turn is rare :), this is
1613 workable.
1615 Usually, however, the lock-out model implemented that way will perform
1616 miserably under the type of load it was designed to handle. In that case,
1617 it might be preferable to stop the real watcher before starting the
1618 idle watcher, so the kernel will not have to process the event in case
1619 the actual processing will be delayed for considerable time.
1621 Here is an example of an I/O watcher that should run at a strictly lower
1622 priority than the default, and which should only process data when no
1623 other events are pending:
1625 ev_idle idle; // actual processing watcher
1626 ev_io io; // actual event watcher
1628 static void
1629 io_cb (EV_P_ ev_io *w, int revents)
1630 {
1631 // stop the I/O watcher, we received the event, but
1632 // are not yet ready to handle it.
1633 ev_io_stop (EV_A_ w);
1635 // start the idle watcher to handle the actual event.
1636 // it will not be executed as long as other watchers
1637 // with the default priority are receiving events.
1638 ev_idle_start (EV_A_ &idle);
1639 }
1641 static void
1642 idle_cb (EV_P_ ev_idle *w, int revents)
1643 {
1644 // actual processing
1645 read (STDIN_FILENO, ...);
1647 // have to start the I/O watcher again, as
1648 // we have handled the event
1649 ev_io_start (EV_P_ &io);
1650 }
1652 // initialisation
1653 ev_idle_init (&idle, idle_cb);
1654 ev_io_init (&io, io_cb, STDIN_FILENO, EV_READ);
1655 ev_io_start (EV_DEFAULT_ &io);
1657 In the "real" world, it might also be beneficial to start a timer, so that
1658 low-priority connections can not be locked out forever under load. This
1659 enables your program to keep a lower latency for important connections
1660 during short periods of high load, while not completely locking out less
1661 important ones.
1664 =head1 WATCHER TYPES
1666 This section describes each watcher in detail, but will not repeat
1667 information given in the last section. Any initialisation/set macros,
1668 functions and members specific to the watcher type are explained.
1670 Most members are additionally marked with either I<[read-only]>, meaning
1671 that, while the watcher is active, you can look at the member and expect
1672 some sensible content, but you must not modify it (you can modify it while
1673 the watcher is stopped to your hearts content), or I<[read-write]>, which
1674 means you can expect it to have some sensible content while the watcher is
1675 active, but you can also modify it (within the same thread as the event
1676 loop, i.e. without creating data races). Modifying it may not do something
1677 sensible or take immediate effect (or do anything at all), but libev will
1678 not crash or malfunction in any way.
1680 In any case, the documentation for each member will explain what the
1681 effects are, and if there are any additional access restrictions.
1683 =head2 C<ev_io> - is this file descriptor readable or writable?
1685 I/O watchers check whether a file descriptor is readable or writable
1686 in each iteration of the event loop, or, more precisely, when reading
1687 would not block the process and writing would at least be able to write
1688 some data. This behaviour is called level-triggering because you keep
1689 receiving events as long as the condition persists. Remember you can stop
1690 the watcher if you don't want to act on the event and neither want to
1691 receive future events.
1693 In general you can register as many read and/or write event watchers per
1694 fd as you want (as long as you don't confuse yourself). Setting all file
1695 descriptors to non-blocking mode is also usually a good idea (but not
1696 required if you know what you are doing).
1698 Another thing you have to watch out for is that it is quite easy to
1699 receive "spurious" readiness notifications, that is, your callback might
1700 be called with C<EV_READ> but a subsequent C<read>(2) will actually block
1701 because there is no data. It is very easy to get into this situation even
1702 with a relatively standard program structure. Thus it is best to always
1703 use non-blocking I/O: An extra C<read>(2) returning C<EAGAIN> is far
1704 preferable to a program hanging until some data arrives.
1706 If you cannot run the fd in non-blocking mode (for example you should
1707 not play around with an Xlib connection), then you have to separately
1708 re-test whether a file descriptor is really ready with a known-to-be good
1709 interface such as poll (fortunately in the case of Xlib, it already does
1710 this on its own, so its quite safe to use). Some people additionally
1711 use C<SIGALRM> and an interval timer, just to be sure you won't block
1712 indefinitely.
1714 But really, best use non-blocking mode.
1716 =head3 The special problem of disappearing file descriptors
1718 Some backends (e.g. kqueue, epoll, linuxaio) need to be told about closing
1719 a file descriptor (either due to calling C<close> explicitly or any other
1720 means, such as C<dup2>). The reason is that you register interest in some
1721 file descriptor, but when it goes away, the operating system will silently
1722 drop this interest. If another file descriptor with the same number then
1723 is registered with libev, there is no efficient way to see that this is,
1724 in fact, a different file descriptor.
1726 To avoid having to explicitly tell libev about such cases, libev follows
1727 the following policy: Each time C<ev_io_set> is being called, libev
1728 will assume that this is potentially a new file descriptor, otherwise
1729 it is assumed that the file descriptor stays the same. That means that
1730 you I<have> to call C<ev_io_set> (or C<ev_io_init>) when you change the
1731 descriptor even if the file descriptor number itself did not change.
1733 This is how one would do it normally anyway, the important point is that
1734 the libev application should not optimise around libev but should leave
1735 optimisations to libev.
1737 =head3 The special problem of dup'ed file descriptors
1739 Some backends (e.g. epoll), cannot register events for file descriptors,
1740 but only events for the underlying file descriptions. That means when you
1741 have C<dup ()>'ed file descriptors or weirder constellations, and register
1742 events for them, only one file descriptor might actually receive events.
1744 There is no workaround possible except not registering events
1745 for potentially C<dup ()>'ed file descriptors, or to resort to
1748 =head3 The special problem of files
1750 Many people try to use C<select> (or libev) on file descriptors
1751 representing files, and expect it to become ready when their program
1752 doesn't block on disk accesses (which can take a long time on their own).
1754 However, this cannot ever work in the "expected" way - you get a readiness
1755 notification as soon as the kernel knows whether and how much data is
1756 there, and in the case of open files, that's always the case, so you
1757 always get a readiness notification instantly, and your read (or possibly
1758 write) will still block on the disk I/O.
1760 Another way to view it is that in the case of sockets, pipes, character
1761 devices and so on, there is another party (the sender) that delivers data
1762 on its own, but in the case of files, there is no such thing: the disk
1763 will not send data on its own, simply because it doesn't know what you
1764 wish to read - you would first have to request some data.
1766 Since files are typically not-so-well supported by advanced notification
1767 mechanism, libev tries hard to emulate POSIX behaviour with respect
1768 to files, even though you should not use it. The reason for this is
1769 convenience: sometimes you want to watch STDIN or STDOUT, which is
1770 usually a tty, often a pipe, but also sometimes files or special devices
1771 (for example, C<epoll> on Linux works with F</dev/random> but not with
1772 F</dev/urandom>), and even though the file might better be served with
1773 asynchronous I/O instead of with non-blocking I/O, it is still useful when
1774 it "just works" instead of freezing.
1776 So avoid file descriptors pointing to files when you know it (e.g. use
1777 libeio), but use them when it is convenient, e.g. for STDIN/STDOUT, or
1778 when you rarely read from a file instead of from a socket, and want to
1779 reuse the same code path.
1781 =head3 The special problem of fork
1783 Some backends (epoll, kqueue, linuxaio, iouring) do not support C<fork ()>
1784 at all or exhibit useless behaviour. Libev fully supports fork, but needs
1785 to be told about it in the child if you want to continue to use it in the
1786 child.
1788 To support fork in your child processes, you have to call C<ev_loop_fork
1789 ()> after a fork in the child, enable C<EVFLAG_FORKCHECK>, or resort to
1792 =head3 The special problem of SIGPIPE
1794 While not really specific to libev, it is easy to forget about C<SIGPIPE>:
1795 when writing to a pipe whose other end has been closed, your program gets
1796 sent a SIGPIPE, which, by default, aborts your program. For most programs
1797 this is sensible behaviour, for daemons, this is usually undesirable.
1799 So when you encounter spurious, unexplained daemon exits, make sure you
1800 ignore SIGPIPE (and maybe make sure you log the exit status of your daemon
1801 somewhere, as that would have given you a big clue).
1803 =head3 The special problem of accept()ing when you can't
1805 Many implementations of the POSIX C<accept> function (for example,
1806 found in post-2004 Linux) have the peculiar behaviour of not removing a
1807 connection from the pending queue in all error cases.
1809 For example, larger servers often run out of file descriptors (because
1810 of resource limits), causing C<accept> to fail with C<ENFILE> but not
1811 rejecting the connection, leading to libev signalling readiness on
1812 the next iteration again (the connection still exists after all), and
1813 typically causing the program to loop at 100% CPU usage.
1815 Unfortunately, the set of errors that cause this issue differs between
1816 operating systems, there is usually little the app can do to remedy the
1817 situation, and no known thread-safe method of removing the connection to
1818 cope with overload is known (to me).
1820 One of the easiest ways to handle this situation is to just ignore it
1821 - when the program encounters an overload, it will just loop until the
1822 situation is over. While this is a form of busy waiting, no OS offers an
1823 event-based way to handle this situation, so it's the best one can do.
1825 A better way to handle the situation is to log any errors other than
1826 C<EAGAIN> and C<EWOULDBLOCK>, making sure not to flood the log with such
1827 messages, and continue as usual, which at least gives the user an idea of
1828 what could be wrong ("raise the ulimit!"). For extra points one could stop
1829 the C<ev_io> watcher on the listening fd "for a while", which reduces CPU
1830 usage.
1832 If your program is single-threaded, then you could also keep a dummy file
1833 descriptor for overload situations (e.g. by opening F</dev/null>), and
1834 when you run into C<ENFILE> or C<EMFILE>, close it, run C<accept>,
1835 close that fd, and create a new dummy fd. This will gracefully refuse
1836 clients under typical overload conditions.
1838 The last way to handle it is to simply log the error and C<exit>, as
1839 is often done with C<malloc> failures, but this results in an easy
1840 opportunity for a DoS attack.
1842 =head3 Watcher-Specific Functions
1844 =over 4
1846 =item ev_io_init (ev_io *, callback, int fd, int events)
1848 =item ev_io_set (ev_io *, int fd, int events)
1850 Configures an C<ev_io> watcher. The C<fd> is the file descriptor to
1851 receive events for and C<events> is either C<EV_READ>, C<EV_WRITE>, both
1852 C<EV_READ | EV_WRITE> or C<0>, to express the desire to receive the given
1853 events.
1855 Note that setting the C<events> to C<0> and starting the watcher is
1856 supported, but not specially optimized - if your program sometimes happens
1857 to generate this combination this is fine, but if it is easy to avoid
1858 starting an io watcher watching for no events you should do so.
1860 =item ev_io_modify (ev_io *, int events)
1862 Similar to C<ev_io_set>, but only changes the requested events. Using this
1863 might be faster with some backends, as libev can assume that the C<fd>
1864 still refers to the same underlying file description, something it cannot
1865 do when using C<ev_io_set>.
1867 =item int fd [no-modify]
1869 The file descriptor being watched. While it can be read at any time, you
1870 must not modify this member even when the watcher is stopped - always use
1871 C<ev_io_set> for that.
1873 =item int events [no-modify]
1875 The set of events the fd is being watched for, among other flags. Remember
1876 that this is a bit set - to test for C<EV_READ>, use C<< w->events &
1877 EV_READ >>, and similarly for C<EV_WRITE>.
1879 As with C<fd>, you must not modify this member even when the watcher is
1880 stopped, always use C<ev_io_set> or C<ev_io_modify> for that.
1882 =back
1884 =head3 Examples
1886 Example: Call C<stdin_readable_cb> when STDIN_FILENO has become, well
1887 readable, but only once. Since it is likely line-buffered, you could
1888 attempt to read a whole line in the callback.
1890 static void
1891 stdin_readable_cb (struct ev_loop *loop, ev_io *w, int revents)
1892 {
1893 ev_io_stop (loop, w);
1894 .. read from stdin here (or from w->fd) and handle any I/O errors
1895 }
1897 ...
1898 struct ev_loop *loop = ev_default_init (0);
1899 ev_io stdin_readable;
1900 ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);
1901 ev_io_start (loop, &stdin_readable);
1902 ev_run (loop, 0);
1905 =head2 C<ev_timer> - relative and optionally repeating timeouts
1907 Timer watchers are simple relative timers that generate an event after a
1908 given time, and optionally repeating in regular intervals after that.
1910 The timers are based on real time, that is, if you register an event that
1911 times out after an hour and you reset your system clock to January last
1912 year, it will still time out after (roughly) one hour. "Roughly" because
1913 detecting time jumps is hard, and some inaccuracies are unavoidable (the
1914 monotonic clock option helps a lot here).
1916 The callback is guaranteed to be invoked only I<after> its timeout has
1917 passed (not I<at>, so on systems with very low-resolution clocks this
1918 might introduce a small delay, see "the special problem of being too
1919 early", below). If multiple timers become ready during the same loop
1920 iteration then the ones with earlier time-out values are invoked before
1921 ones of the same priority with later time-out values (but this is no
1922 longer true when a callback calls C<ev_run> recursively).
1924 =head3 Be smart about timeouts
1926 Many real-world problems involve some kind of timeout, usually for error
1927 recovery. A typical example is an HTTP request - if the other side hangs,
1928 you want to raise some error after a while.
1930 What follows are some ways to handle this problem, from obvious and
1931 inefficient to smart and efficient.
1933 In the following, a 60 second activity timeout is assumed - a timeout that
1934 gets reset to 60 seconds each time there is activity (e.g. each time some
1935 data or other life sign was received).
1937 =over 4
1939 =item 1. Use a timer and stop, reinitialise and start it on activity.
1941 This is the most obvious, but not the most simple way: In the beginning,
1942 start the watcher:
1944 ev_timer_init (timer, callback, 60., 0.);
1945 ev_timer_start (loop, timer);
1947 Then, each time there is some activity, C<ev_timer_stop> it, initialise it
1948 and start it again:
1950 ev_timer_stop (loop, timer);
1951 ev_timer_set (timer, 60., 0.);
1952 ev_timer_start (loop, timer);
1954 This is relatively simple to implement, but means that each time there is
1955 some activity, libev will first have to remove the timer from its internal
1956 data structure and then add it again. Libev tries to be fast, but it's
1957 still not a constant-time operation.
1959 =item 2. Use a timer and re-start it with C<ev_timer_again> inactivity.
1961 This is the easiest way, and involves using C<ev_timer_again> instead of
1962 C<ev_timer_start>.
1964 To implement this, configure an C<ev_timer> with a C<repeat> value
1965 of C<60> and then call C<ev_timer_again> at start and each time you
1966 successfully read or write some data. If you go into an idle state where
1967 you do not expect data to travel on the socket, you can C<ev_timer_stop>
1968 the timer, and C<ev_timer_again> will automatically restart it if need be.
1970 That means you can ignore both the C<ev_timer_start> function and the
1971 C<after> argument to C<ev_timer_set>, and only ever use the C<repeat>
1972 member and C<ev_timer_again>.
1974 At start:
1976 ev_init (timer, callback);
1977 timer->repeat = 60.;
1978 ev_timer_again (loop, timer);
1980 Each time there is some activity:
1982 ev_timer_again (loop, timer);
1984 It is even possible to change the time-out on the fly, regardless of
1985 whether the watcher is active or not:
1987 timer->repeat = 30.;
1988 ev_timer_again (loop, timer);
1990 This is slightly more efficient then stopping/starting the timer each time
1991 you want to modify its timeout value, as libev does not have to completely
1992 remove and re-insert the timer from/into its internal data structure.
1994 It is, however, even simpler than the "obvious" way to do it.
1996 =item 3. Let the timer time out, but then re-arm it as required.
1998 This method is more tricky, but usually most efficient: Most timeouts are
1999 relatively long compared to the intervals between other activity - in
2000 our example, within 60 seconds, there are usually many I/O events with
2001 associated activity resets.
2003 In this case, it would be more efficient to leave the C<ev_timer> alone,
2004 but remember the time of last activity, and check for a real timeout only
2005 within the callback:
2007 ev_tstamp timeout = 60.;
2008 ev_tstamp last_activity; // time of last activity
2009 ev_timer timer;
2011 static void
2012 callback (EV_P_ ev_timer *w, int revents)
2013 {
2014 // calculate when the timeout would happen
2015 ev_tstamp after = last_activity - ev_now (EV_A) + timeout;
2017 // if negative, it means we the timeout already occurred
2018 if (after < 0.)
2019 {
2020 // timeout occurred, take action
2021 }
2022 else
2023 {
2024 // callback was invoked, but there was some recent
2025 // activity. simply restart the timer to time out
2026 // after "after" seconds, which is the earliest time
2027 // the timeout can occur.
2028 ev_timer_set (w, after, 0.);
2029 ev_timer_start (EV_A_ w);
2030 }
2031 }
2033 To summarise the callback: first calculate in how many seconds the
2034 timeout will occur (by calculating the absolute time when it would occur,
2035 C<last_activity + timeout>, and subtracting the current time, C<ev_now
2036 (EV_A)> from that).
2038 If this value is negative, then we are already past the timeout, i.e. we
2039 timed out, and need to do whatever is needed in this case.
2041 Otherwise, we now the earliest time at which the timeout would trigger,
2042 and simply start the timer with this timeout value.
2044 In other words, each time the callback is invoked it will check whether
2045 the timeout occurred. If not, it will simply reschedule itself to check
2046 again at the earliest time it could time out. Rinse. Repeat.
2048 This scheme causes more callback invocations (about one every 60 seconds
2049 minus half the average time between activity), but virtually no calls to
2050 libev to change the timeout.
2052 To start the machinery, simply initialise the watcher and set
2053 C<last_activity> to the current time (meaning there was some activity just
2054 now), then call the callback, which will "do the right thing" and start
2055 the timer:
2057 last_activity = ev_now (EV_A);
2058 ev_init (&timer, callback);
2059 callback (EV_A_ &timer, 0);
2061 When there is some activity, simply store the current time in
2062 C<last_activity>, no libev calls at all:
2064 if (activity detected)
2065 last_activity = ev_now (EV_A);
2067 When your timeout value changes, then the timeout can be changed by simply
2068 providing a new value, stopping the timer and calling the callback, which
2069 will again do the right thing (for example, time out immediately :).
2071 timeout = new_value;
2072 ev_timer_stop (EV_A_ &timer);
2073 callback (EV_A_ &timer, 0);
2075 This technique is slightly more complex, but in most cases where the
2076 time-out is unlikely to be triggered, much more efficient.
2078 =item 4. Wee, just use a double-linked list for your timeouts.
2080 If there is not one request, but many thousands (millions...), all
2081 employing some kind of timeout with the same timeout value, then one can
2082 do even better:
2084 When starting the timeout, calculate the timeout value and put the timeout
2085 at the I<end> of the list.
2087 Then use an C<ev_timer> to fire when the timeout at the I<beginning> of
2088 the list is expected to fire (for example, using the technique #3).
2090 When there is some activity, remove the timer from the list, recalculate
2091 the timeout, append it to the end of the list again, and make sure to
2092 update the C<ev_timer> if it was taken from the beginning of the list.
2094 This way, one can manage an unlimited number of timeouts in O(1) time for
2095 starting, stopping and updating the timers, at the expense of a major
2096 complication, and having to use a constant timeout. The constant timeout
2097 ensures that the list stays sorted.
2099 =back
2101 So which method the best?
2103 Method #2 is a simple no-brain-required solution that is adequate in most
2104 situations. Method #3 requires a bit more thinking, but handles many cases
2105 better, and isn't very complicated either. In most case, choosing either
2106 one is fine, with #3 being better in typical situations.
2108 Method #1 is almost always a bad idea, and buys you nothing. Method #4 is
2109 rather complicated, but extremely efficient, something that really pays
2110 off after the first million or so of active timers, i.e. it's usually
2111 overkill :)
2113 =head3 The special problem of being too early
2115 If you ask a timer to call your callback after three seconds, then
2116 you expect it to be invoked after three seconds - but of course, this
2117 cannot be guaranteed to infinite precision. Less obviously, it cannot be
2118 guaranteed to any precision by libev - imagine somebody suspending the
2119 process with a STOP signal for a few hours for example.
2121 So, libev tries to invoke your callback as soon as possible I<after> the
2122 delay has occurred, but cannot guarantee this.
2124 A less obvious failure mode is calling your callback too early: many event
2125 loops compare timestamps with a "elapsed delay >= requested delay", but
2126 this can cause your callback to be invoked much earlier than you would
2127 expect.
2129 To see why, imagine a system with a clock that only offers full second
2130 resolution (think windows if you can't come up with a broken enough OS
2131 yourself). If you schedule a one-second timer at the time 500.9, then the
2132 event loop will schedule your timeout to elapse at a system time of 500
2133 (500.9 truncated to the resolution) + 1, or 501.
2135 If an event library looks at the timeout 0.1s later, it will see "501 >=
2136 501" and invoke the callback 0.1s after it was started, even though a
2137 one-second delay was requested - this is being "too early", despite best
2138 intentions.
2140 This is the reason why libev will never invoke the callback if the elapsed
2141 delay equals the requested delay, but only when the elapsed delay is
2142 larger than the requested delay. In the example above, libev would only invoke
2143 the callback at system time 502, or 1.1s after the timer was started.
2145 So, while libev cannot guarantee that your callback will be invoked
2146 exactly when requested, it I<can> and I<does> guarantee that the requested
2147 delay has actually elapsed, or in other words, it always errs on the "too
2148 late" side of things.
2150 =head3 The special problem of time updates
2152 Establishing the current time is a costly operation (it usually takes
2153 at least one system call): EV therefore updates its idea of the current
2154 time only before and after C<ev_run> collects new events, which causes a
2155 growing difference between C<ev_now ()> and C<ev_time ()> when handling
2156 lots of events in one iteration.
2158 The relative timeouts are calculated relative to the C<ev_now ()>
2159 time. This is usually the right thing as this timestamp refers to the time
2160 of the event triggering whatever timeout you are modifying/starting. If
2161 you suspect event processing to be delayed and you I<need> to base the
2162 timeout on the current time, use something like the following to adjust
2163 for it:
2165 ev_timer_set (&timer, after + (ev_time () - ev_now ()), 0.);
2167 If the event loop is suspended for a long time, you can also force an
2168 update of the time returned by C<ev_now ()> by calling C<ev_now_update
2169 ()>, although that will push the event time of all outstanding events
2170 further into the future.
2172 =head3 The special problem of unsynchronised clocks
2174 Modern systems have a variety of clocks - libev itself uses the normal
2175 "wall clock" clock and, if available, the monotonic clock (to avoid time
2176 jumps).
2178 Neither of these clocks is synchronised with each other or any other clock
2179 on the system, so C<ev_time ()> might return a considerably different time
2180 than C<gettimeofday ()> or C<time ()>. On a GNU/Linux system, for example,
2181 a call to C<gettimeofday> might return a second count that is one higher
2182 than a directly following call to C<time>.
2184 The moral of this is to only compare libev-related timestamps with
2185 C<ev_time ()> and C<ev_now ()>, at least if you want better precision than
2186 a second or so.
2188 One more problem arises due to this lack of synchronisation: if libev uses
2189 the system monotonic clock and you compare timestamps from C<ev_time>
2190 or C<ev_now> from when you started your timer and when your callback is
2191 invoked, you will find that sometimes the callback is a bit "early".
2193 This is because C<ev_timer>s work in real time, not wall clock time, so
2194 libev makes sure your callback is not invoked before the delay happened,
2195 I<measured according to the real time>, not the system clock.
2197 If your timeouts are based on a physical timescale (e.g. "time out this
2198 connection after 100 seconds") then this shouldn't bother you as it is
2199 exactly the right behaviour.
2201 If you want to compare wall clock/system timestamps to your timers, then
2202 you need to use C<ev_periodic>s, as these are based on the wall clock
2203 time, where your comparisons will always generate correct results.
2205 =head3 The special problems of suspended animation
2207 When you leave the server world it is quite customary to hit machines that
2208 can suspend/hibernate - what happens to the clocks during such a suspend?
2210 Some quick tests made with a Linux 2.6.28 indicate that a suspend freezes
2211 all processes, while the clocks (C<times>, C<CLOCK_MONOTONIC>) continue
2212 to run until the system is suspended, but they will not advance while the
2213 system is suspended. That means, on resume, it will be as if the program
2214 was frozen for a few seconds, but the suspend time will not be counted
2215 towards C<ev_timer> when a monotonic clock source is used. The real time
2216 clock advanced as expected, but if it is used as sole clocksource, then a
2217 long suspend would be detected as a time jump by libev, and timers would
2218 be adjusted accordingly.
2220 I would not be surprised to see different behaviour in different between
2221 operating systems, OS versions or even different hardware.
2223 The other form of suspend (job control, or sending a SIGSTOP) will see a
2224 time jump in the monotonic clocks and the realtime clock. If the program
2225 is suspended for a very long time, and monotonic clock sources are in use,
2226 then you can expect C<ev_timer>s to expire as the full suspension time
2227 will be counted towards the timers. When no monotonic clock source is in
2228 use, then libev will again assume a timejump and adjust accordingly.
2230 It might be beneficial for this latter case to call C<ev_suspend>
2231 and C<ev_resume> in code that handles C<SIGTSTP>, to at least get
2232 deterministic behaviour in this case (you can do nothing against
2233 C<SIGSTOP>).
2235 =head3 Watcher-Specific Functions and Data Members
2237 =over 4
2239 =item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)
2241 =item ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)
2243 Configure the timer to trigger after C<after> seconds (fractional and
2244 negative values are supported). If C<repeat> is C<0.>, then it will
2245 automatically be stopped once the timeout is reached. If it is positive,
2246 then the timer will automatically be configured to trigger again C<repeat>
2247 seconds later, again, and again, until stopped manually.
2249 The timer itself will do a best-effort at avoiding drift, that is, if
2250 you configure a timer to trigger every 10 seconds, then it will normally
2251 trigger at exactly 10 second intervals. If, however, your program cannot
2252 keep up with the timer (because it takes longer than those 10 seconds to
2253 do stuff) the timer will not fire more than once per event loop iteration.
2255 =item ev_timer_again (loop, ev_timer *)
2257 This will act as if the timer timed out, and restarts it again if it is
2258 repeating. It basically works like calling C<ev_timer_stop>, updating the
2259 timeout to the C<repeat> value and calling C<ev_timer_start>.
2261 The exact semantics are as in the following rules, all of which will be
2262 applied to the watcher:
2264 =over 4
2266 =item If the timer is pending, the pending status is always cleared.
2268 =item If the timer is started but non-repeating, stop it (as if it timed
2269 out, without invoking it).
2271 =item If the timer is repeating, make the C<repeat> value the new timeout
2272 and start the timer, if necessary.
2274 =back
2276 This sounds a bit complicated, see L</Be smart about timeouts>, above, for a
2277 usage example.
2279 =item ev_tstamp ev_timer_remaining (loop, ev_timer *)
2281 Returns the remaining time until a timer fires. If the timer is active,
2282 then this time is relative to the current event loop time, otherwise it's
2283 the timeout value currently configured.
2285 That is, after an C<ev_timer_set (w, 5, 7)>, C<ev_timer_remaining> returns
2286 C<5>. When the timer is started and one second passes, C<ev_timer_remaining>
2287 will return C<4>. When the timer expires and is restarted, it will return
2288 roughly C<7> (likely slightly less as callback invocation takes some time,
2289 too), and so on.
2291 =item ev_tstamp repeat [read-write]
2293 The current C<repeat> value. Will be used each time the watcher times out
2294 or C<ev_timer_again> is called, and determines the next timeout (if any),
2295 which is also when any modifications are taken into account.
2297 =back
2299 =head3 Examples
2301 Example: Create a timer that fires after 60 seconds.
2303 static void
2304 one_minute_cb (struct ev_loop *loop, ev_timer *w, int revents)
2305 {
2306 .. one minute over, w is actually stopped right here
2307 }
2309 ev_timer mytimer;
2310 ev_timer_init (&mytimer, one_minute_cb, 60., 0.);
2311 ev_timer_start (loop, &mytimer);
2313 Example: Create a timeout timer that times out after 10 seconds of
2314 inactivity.
2316 static void
2317 timeout_cb (struct ev_loop *loop, ev_timer *w, int revents)
2318 {
2319 .. ten seconds without any activity
2320 }
2322 ev_timer mytimer;
2323 ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */
2324 ev_timer_again (&mytimer); /* start timer */
2325 ev_run (loop, 0);
2327 // and in some piece of code that gets executed on any "activity":
2328 // reset the timeout to start ticking again at 10 seconds
2329 ev_timer_again (&mytimer);
2332 =head2 C<ev_periodic> - to cron or not to cron?
2334 Periodic watchers are also timers of a kind, but they are very versatile
2335 (and unfortunately a bit complex).
2337 Unlike C<ev_timer>, periodic watchers are not based on real time (or
2338 relative time, the physical time that passes) but on wall clock time
2339 (absolute time, the thing you can read on your calendar or clock). The
2340 difference is that wall clock time can run faster or slower than real
2341 time, and time jumps are not uncommon (e.g. when you adjust your
2342 wrist-watch).
2344 You can tell a periodic watcher to trigger after some specific point
2345 in time: for example, if you tell a periodic watcher to trigger "in 10
2346 seconds" (by specifying e.g. C<ev_now () + 10.>, that is, an absolute time
2347 not a delay) and then reset your system clock to January of the previous
2348 year, then it will take a year or more to trigger the event (unlike an
2349 C<ev_timer>, which would still trigger roughly 10 seconds after starting
2350 it, as it uses a relative timeout).
2352 C<ev_periodic> watchers can also be used to implement vastly more complex
2353 timers, such as triggering an event on each "midnight, local time", or
2354 other complicated rules. This cannot easily be done with C<ev_timer>
2355 watchers, as those cannot react to time jumps.
2357 As with timers, the callback is guaranteed to be invoked only when the
2358 point in time where it is supposed to trigger has passed. If multiple
2359 timers become ready during the same loop iteration then the ones with
2360 earlier time-out values are invoked before ones with later time-out values
2361 (but this is no longer true when a callback calls C<ev_run> recursively).
2363 =head3 Watcher-Specific Functions and Data Members
2365 =over 4
2367 =item ev_periodic_init (ev_periodic *, callback, ev_tstamp offset, ev_tstamp interval, reschedule_cb)
2369 =item ev_periodic_set (ev_periodic *, ev_tstamp offset, ev_tstamp interval, reschedule_cb)
2371 Lots of arguments, let's sort it out... There are basically three modes of
2372 operation, and we will explain them from simplest to most complex:
2374 =over 4
2376 =item * absolute timer (offset = absolute time, interval = 0, reschedule_cb = 0)
2378 In this configuration the watcher triggers an event after the wall clock
2379 time C<offset> has passed. It will not repeat and will not adjust when a
2380 time jump occurs, that is, if it is to be run at January 1st 2011 then it
2381 will be stopped and invoked when the system clock reaches or surpasses
2382 this point in time.
2384 =item * repeating interval timer (offset = offset within interval, interval > 0, reschedule_cb = 0)
2386 In this mode the watcher will always be scheduled to time out at the next
2387 C<offset + N * interval> time (for some integer N, which can also be
2388 negative) and then repeat, regardless of any time jumps. The C<offset>
2389 argument is merely an offset into the C<interval> periods.
2391 This can be used to create timers that do not drift with respect to the
2392 system clock, for example, here is an C<ev_periodic> that triggers each
2393 hour, on the hour (with respect to UTC):
2395 ev_periodic_set (&periodic, 0., 3600., 0);
2397 This doesn't mean there will always be 3600 seconds in between triggers,
2398 but only that the callback will be called when the system time shows a
2399 full hour (UTC), or more correctly, when the system time is evenly divisible
2400 by 3600.
2402 Another way to think about it (for the mathematically inclined) is that
2403 C<ev_periodic> will try to run the callback in this mode at the next possible
2404 time where C<time = offset (mod interval)>, regardless of any time jumps.
2406 The C<interval> I<MUST> be positive, and for numerical stability, the
2407 interval value should be higher than C<1/8192> (which is around 100
2408 microseconds) and C<offset> should be higher than C<0> and should have
2409 at most a similar magnitude as the current time (say, within a factor of
2410 ten). Typical values for offset are, in fact, C<0> or something between
2411 C<0> and C<interval>, which is also the recommended range.
2413 Note also that there is an upper limit to how often a timer can fire (CPU
2414 speed for example), so if C<interval> is very small then timing stability
2415 will of course deteriorate. Libev itself tries to be exact to be about one
2416 millisecond (if the OS supports it and the machine is fast enough).
2418 =item * manual reschedule mode (offset ignored, interval ignored, reschedule_cb = callback)
2420 In this mode the values for C<interval> and C<offset> are both being
2421 ignored. Instead, each time the periodic watcher gets scheduled, the
2422 reschedule callback will be called with the watcher as first, and the
2423 current time as second argument.
2425 NOTE: I<This callback MUST NOT stop or destroy any periodic watcher, ever,
2426 or make ANY other event loop modifications whatsoever, unless explicitly
2427 allowed by documentation here>.
2429 If you need to stop it, return C<now + 1e30> (or so, fudge fudge) and stop
2430 it afterwards (e.g. by starting an C<ev_prepare> watcher, which is the
2431 only event loop modification you are allowed to do).
2433 The callback prototype is C<ev_tstamp (*reschedule_cb)(ev_periodic
2434 *w, ev_tstamp now)>, e.g.:
2436 static ev_tstamp
2437 my_rescheduler (ev_periodic *w, ev_tstamp now)
2438 {
2439 return now + 60.;
2440 }
2442 It must return the next time to trigger, based on the passed time value
2443 (that is, the lowest time value larger than to the second argument). It
2444 will usually be called just before the callback will be triggered, but
2445 might be called at other times, too.
2447 NOTE: I<< This callback must always return a time that is higher than or
2448 equal to the passed C<now> value >>.
2450 This can be used to create very complex timers, such as a timer that
2451 triggers on "next midnight, local time". To do this, you would calculate
2452 the next midnight after C<now> and return the timestamp value for
2453 this. Here is a (completely untested, no error checking) example on how to
2454 do this:
2456 #include <time.h>
2458 static ev_tstamp
2459 my_rescheduler (ev_periodic *w, ev_tstamp now)
2460 {
2461 time_t tnow = (time_t)now;
2462 struct tm tm;
2463 localtime_r (&tnow, &tm);
2465 tm.tm_sec = tm.tm_min = tm.tm_hour = 0; // midnight current day
2466 ++tm.tm_mday; // midnight next day
2468 return mktime (&tm);
2469 }
2471 Note: this code might run into trouble on days that have more then two
2472 midnights (beginning and end).
2474 =back
2476 =item ev_periodic_again (loop, ev_periodic *)
2478 Simply stops and restarts the periodic watcher again. This is only useful
2479 when you changed some parameters or the reschedule callback would return
2480 a different time than the last time it was called (e.g. in a crond like
2481 program when the crontabs have changed).
2483 =item ev_tstamp ev_periodic_at (ev_periodic *)
2485 When active, returns the absolute time that the watcher is supposed
2486 to trigger next. This is not the same as the C<offset> argument to
2487 C<ev_periodic_set>, but indeed works even in interval and manual
2488 rescheduling modes.
2490 =item ev_tstamp offset [read-write]
2492 When repeating, this contains the offset value, otherwise this is the
2493 absolute point in time (the C<offset> value passed to C<ev_periodic_set>,
2494 although libev might modify this value for better numerical stability).
2496 Can be modified any time, but changes only take effect when the periodic
2497 timer fires or C<ev_periodic_again> is being called.
2499 =item ev_tstamp interval [read-write]
2501 The current interval value. Can be modified any time, but changes only
2502 take effect when the periodic timer fires or C<ev_periodic_again> is being
2503 called.
2505 =item ev_tstamp (*reschedule_cb)(ev_periodic *w, ev_tstamp now) [read-write]
2507 The current reschedule callback, or C<0>, if this functionality is
2508 switched off. Can be changed any time, but changes only take effect when
2509 the periodic timer fires or C<ev_periodic_again> is being called.
2511 =back
2513 =head3 Examples
2515 Example: Call a callback every hour, or, more precisely, whenever the
2516 system time is divisible by 3600. The callback invocation times have
2517 potentially a lot of jitter, but good long-term stability.
2519 static void
2520 clock_cb (struct ev_loop *loop, ev_periodic *w, int revents)
2521 {
2522 ... its now a full hour (UTC, or TAI or whatever your clock follows)
2523 }
2525 ev_periodic hourly_tick;
2526 ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);
2527 ev_periodic_start (loop, &hourly_tick);
2529 Example: The same as above, but use a reschedule callback to do it:
2531 #include <math.h>
2533 static ev_tstamp
2534 my_scheduler_cb (ev_periodic *w, ev_tstamp now)
2535 {
2536 return now + (3600. - fmod (now, 3600.));
2537 }
2539 ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);
2541 Example: Call a callback every hour, starting now:
2543 ev_periodic hourly_tick;
2544 ev_periodic_init (&hourly_tick, clock_cb,
2545 fmod (ev_now (loop), 3600.), 3600., 0);
2546 ev_periodic_start (loop, &hourly_tick);
2549 =head2 C<ev_signal> - signal me when a signal gets signalled!
2551 Signal watchers will trigger an event when the process receives a specific
2552 signal one or more times. Even though signals are very asynchronous, libev
2553 will try its best to deliver signals synchronously, i.e. as part of the
2554 normal event processing, like any other event.
2556 If you want signals to be delivered truly asynchronously, just use
2557 C<sigaction> as you would do without libev and forget about sharing
2558 the signal. You can even use C<ev_async> from a signal handler to
2559 synchronously wake up an event loop.
2561 You can configure as many watchers as you like for the same signal, but
2562 only within the same loop, i.e. you can watch for C<SIGINT> in your
2563 default loop and for C<SIGIO> in another loop, but you cannot watch for
2564 C<SIGINT> in both the default loop and another loop at the same time. At
2565 the moment, C<SIGCHLD> is permanently tied to the default loop.
2567 Only after the first watcher for a signal is started will libev actually
2568 register something with the kernel. It thus coexists with your own signal
2569 handlers as long as you don't register any with libev for the same signal.
2571 If possible and supported, libev will install its handlers with
2572 C<SA_RESTART> (or equivalent) behaviour enabled, so system calls should
2573 not be unduly interrupted. If you have a problem with system calls getting
2574 interrupted by signals you can block all signals in an C<ev_check> watcher
2575 and unblock them in an C<ev_prepare> watcher.
2577 =head3 The special problem of inheritance over fork/execve/pthread_create
2579 Both the signal mask (C<sigprocmask>) and the signal disposition
2580 (C<sigaction>) are unspecified after starting a signal watcher (and after
2581 stopping it again), that is, libev might or might not block the signal,
2582 and might or might not set or restore the installed signal handler (but
2585 While this does not matter for the signal disposition (libev never
2586 sets signals to C<SIG_IGN>, so handlers will be reset to C<SIG_DFL> on
2587 C<execve>), this matters for the signal mask: many programs do not expect
2588 certain signals to be blocked.
2590 This means that before calling C<exec> (from the child) you should reset
2591 the signal mask to whatever "default" you expect (all clear is a good
2592 choice usually).
2594 The simplest way to ensure that the signal mask is reset in the child is
2595 to install a fork handler with C<pthread_atfork> that resets it. That will
2596 catch fork calls done by libraries (such as the libc) as well.
2598 In current versions of libev, the signal will not be blocked indefinitely
2599 unless you use the C<signalfd> API (C<EV_SIGNALFD>). While this reduces
2600 the window of opportunity for problems, it will not go away, as libev
2601 I<has> to modify the signal mask, at least temporarily.
2603 So I can't stress this enough: I<If you do not reset your signal mask when
2604 you expect it to be empty, you have a race condition in your code>. This
2605 is not a libev-specific thing, this is true for most event libraries.
2607 =head3 The special problem of threads signal handling
2609 POSIX threads has problematic signal handling semantics, specifically,
2610 a lot of functionality (sigfd, sigwait etc.) only really works if all
2611 threads in a process block signals, which is hard to achieve.
2613 When you want to use sigwait (or mix libev signal handling with your own
2614 for the same signals), you can tackle this problem by globally blocking
2615 all signals before creating any threads (or creating them with a fully set
2616 sigprocmask) and also specifying the C<EVFLAG_NOSIGMASK> when creating
2617 loops. Then designate one thread as "signal receiver thread" which handles
2618 these signals. You can pass on any signals that libev might be interested
2619 in by calling C<ev_feed_signal>.
2621 =head3 Watcher-Specific Functions and Data Members
2623 =over 4
2625 =item ev_signal_init (ev_signal *, callback, int signum)
2627 =item ev_signal_set (ev_signal *, int signum)
2629 Configures the watcher to trigger on the given signal number (usually one
2630 of the C<SIGxxx> constants).
2632 =item int signum [read-only]
2634 The signal the watcher watches out for.
2636 =back
2638 =head3 Examples
2640 Example: Try to exit cleanly on SIGINT.
2642 static void
2643 sigint_cb (struct ev_loop *loop, ev_signal *w, int revents)
2644 {
2645 ev_break (loop, EVBREAK_ALL);
2646 }
2648 ev_signal signal_watcher;
2649 ev_signal_init (&signal_watcher, sigint_cb, SIGINT);
2650 ev_signal_start (loop, &signal_watcher);
2653 =head2 C<ev_child> - watch out for process status changes
2655 Child watchers trigger when your process receives a SIGCHLD in response to
2656 some child status changes (most typically when a child of yours dies or
2657 exits). It is permissible to install a child watcher I<after> the child
2658 has been forked (which implies it might have already exited), as long
2659 as the event loop isn't entered (or is continued from a watcher), i.e.,
2660 forking and then immediately registering a watcher for the child is fine,
2661 but forking and registering a watcher a few event loop iterations later or
2662 in the next callback invocation is not.
2664 Only the default event loop is capable of handling signals, and therefore
2665 you can only register child watchers in the default event loop.
2667 Due to some design glitches inside libev, child watchers will always be
2668 handled at maximum priority (their priority is set to C<EV_MAXPRI> by
2669 libev)
2671 =head3 Process Interaction
2673 Libev grabs C<SIGCHLD> as soon as the default event loop is
2674 initialised. This is necessary to guarantee proper behaviour even if the
2675 first child watcher is started after the child exits. The occurrence
2676 of C<SIGCHLD> is recorded asynchronously, but child reaping is done
2677 synchronously as part of the event loop processing. Libev always reaps all
2678 children, even ones not watched.
2680 =head3 Overriding the Built-In Processing
2682 Libev offers no special support for overriding the built-in child
2683 processing, but if your application collides with libev's default child
2684 handler, you can override it easily by installing your own handler for
2685 C<SIGCHLD> after initialising the default loop, and making sure the
2686 default loop never gets destroyed. You are encouraged, however, to use an
2687 event-based approach to child reaping and thus use libev's support for
2688 that, so other libev users can use C<ev_child> watchers freely.
2690 =head3 Stopping the Child Watcher
2692 Currently, the child watcher never gets stopped, even when the
2693 child terminates, so normally one needs to stop the watcher in the
2694 callback. Future versions of libev might stop the watcher automatically
2695 when a child exit is detected (calling C<ev_child_stop> twice is not a
2696 problem).
2698 =head3 Watcher-Specific Functions and Data Members
2700 =over 4
2702 =item ev_child_init (ev_child *, callback, int pid, int trace)
2704 =item ev_child_set (ev_child *, int pid, int trace)
2706 Configures the watcher to wait for status changes of process C<pid> (or
2707 I<any> process if C<pid> is specified as C<0>). The callback can look
2708 at the C<rstatus> member of the C<ev_child> watcher structure to see
2709 the status word (use the macros from C<sys/wait.h> and see your systems
2710 C<waitpid> documentation). The C<rpid> member contains the pid of the
2711 process causing the status change. C<trace> must be either C<0> (only
2712 activate the watcher when the process terminates) or C<1> (additionally
2713 activate the watcher when the process is stopped or continued).
2715 =item int pid [read-only]
2717 The process id this watcher watches out for, or C<0>, meaning any process id.
2719 =item int rpid [read-write]
2721 The process id that detected a status change.
2723 =item int rstatus [read-write]
2725 The process exit/trace status caused by C<rpid> (see your systems
2726 C<waitpid> and C<sys/wait.h> documentation for details).
2728 =back
2730 =head3 Examples
2732 Example: C<fork()> a new process and install a child handler to wait for
2733 its completion.
2735 ev_child cw;
2737 static void
2738 child_cb (EV_P_ ev_child *w, int revents)
2739 {
2740 ev_child_stop (EV_A_ w);
2741 printf ("process %d exited with status %x\n", w->rpid, w->rstatus);
2742 }
2744 pid_t pid = fork ();
2746 if (pid < 0)
2747 // error
2748 else if (pid == 0)
2749 {
2750 // the forked child executes here
2751 exit (1);
2752 }
2753 else
2754 {
2755 ev_child_init (&cw, child_cb, pid, 0);
2756 ev_child_start (EV_DEFAULT_ &cw);
2757 }
2760 =head2 C<ev_stat> - did the file attributes just change?
2762 This watches a file system path for attribute changes. That is, it calls
2763 C<stat> on that path in regular intervals (or when the OS says it changed)
2764 and sees if it changed compared to the last time, invoking the callback
2765 if it did. Starting the watcher C<stat>'s the file, so only changes that
2766 happen after the watcher has been started will be reported.
2768 The path does not need to exist: changing from "path exists" to "path does
2769 not exist" is a status change like any other. The condition "path does not
2770 exist" (or more correctly "path cannot be stat'ed") is signified by the
2771 C<st_nlink> field being zero (which is otherwise always forced to be at
2772 least one) and all the other fields of the stat buffer having unspecified
2773 contents.
2775 The path I<must not> end in a slash or contain special components such as
2776 C<.> or C<..>. The path I<should> be absolute: If it is relative and
2777 your working directory changes, then the behaviour is undefined.
2779 Since there is no portable change notification interface available, the
2780 portable implementation simply calls C<stat(2)> regularly on the path
2781 to see if it changed somehow. You can specify a recommended polling
2782 interval for this case. If you specify a polling interval of C<0> (highly
2783 recommended!) then a I<suitable, unspecified default> value will be used
2784 (which you can expect to be around five seconds, although this might
2785 change dynamically). Libev will also impose a minimum interval which is
2786 currently around C<0.1>, but that's usually overkill.
2788 This watcher type is not meant for massive numbers of stat watchers,
2789 as even with OS-supported change notifications, this can be
2790 resource-intensive.
2792 At the time of this writing, the only OS-specific interface implemented
2793 is the Linux inotify interface (implementing kqueue support is left as an
2794 exercise for the reader. Note, however, that the author sees no way of
2795 implementing C<ev_stat> semantics with kqueue, except as a hint).
2797 =head3 ABI Issues (Largefile Support)
2799 Libev by default (unless the user overrides this) uses the default
2800 compilation environment, which means that on systems with large file
2801 support disabled by default, you get the 32 bit version of the stat
2802 structure. When using the library from programs that change the ABI to
2803 use 64 bit file offsets the programs will fail. In that case you have to
2804 compile libev with the same flags to get binary compatibility. This is
2805 obviously the case with any flags that change the ABI, but the problem is
2806 most noticeably displayed with ev_stat and large file support.
2808 The solution for this is to lobby your distribution maker to make large
2809 file interfaces available by default (as e.g. FreeBSD does) and not
2810 optional. Libev cannot simply switch on large file support because it has
2811 to exchange stat structures with application programs compiled using the
2812 default compilation environment.
2814 =head3 Inotify and Kqueue
2816 When C<inotify (7)> support has been compiled into libev and present at
2817 runtime, it will be used to speed up change detection where possible. The
2818 inotify descriptor will be created lazily when the first C<ev_stat>
2819 watcher is being started.
2821 Inotify presence does not change the semantics of C<ev_stat> watchers
2822 except that changes might be detected earlier, and in some cases, to avoid
2823 making regular C<stat> calls. Even in the presence of inotify support
2824 there are many cases where libev has to resort to regular C<stat> polling,
2825 but as long as kernel 2.6.25 or newer is used (2.6.24 and older have too
2826 many bugs), the path exists (i.e. stat succeeds), and the path resides on
2827 a local filesystem (libev currently assumes only ext2/3, jfs, reiserfs and
2828 xfs are fully working) libev usually gets away without polling.
2830 There is no support for kqueue, as apparently it cannot be used to
2831 implement this functionality, due to the requirement of having a file
2832 descriptor open on the object at all times, and detecting renames, unlinks
2833 etc. is difficult.
2835 =head3 C<stat ()> is a synchronous operation
2837 Libev doesn't normally do any kind of I/O itself, and so is not blocking
2838 the process. The exception are C<ev_stat> watchers - those call C<stat
2839 ()>, which is a synchronous operation.
2841 For local paths, this usually doesn't matter: unless the system is very
2842 busy or the intervals between stat's are large, a stat call will be fast,
2843 as the path data is usually in memory already (except when starting the
2844 watcher).
2846 For networked file systems, calling C<stat ()> can block an indefinite
2847 time due to network issues, and even under good conditions, a stat call
2848 often takes multiple milliseconds.
2850 Therefore, it is best to avoid using C<ev_stat> watchers on networked
2851 paths, although this is fully supported by libev.
2853 =head3 The special problem of stat time resolution
2855 The C<stat ()> system call only supports full-second resolution portably,
2856 and even on systems where the resolution is higher, most file systems
2857 still only support whole seconds.
2859 That means that, if the time is the only thing that changes, you can
2860 easily miss updates: on the first update, C<ev_stat> detects a change and
2861 calls your callback, which does something. When there is another update
2862 within the same second, C<ev_stat> will be unable to detect unless the
2863 stat data does change in other ways (e.g. file size).
2865 The solution to this is to delay acting on a change for slightly more
2866 than a second (or till slightly after the next full second boundary), using
2867 a roughly one-second-delay C<ev_timer> (e.g. C<ev_timer_set (w, 0., 1.02);
2868 ev_timer_again (loop, w)>).
2870 The C<.02> offset is added to work around small timing inconsistencies
2871 of some operating systems (where the second counter of the current time
2872 might be be delayed. One such system is the Linux kernel, where a call to
2873 C<gettimeofday> might return a timestamp with a full second later than
2874 a subsequent C<time> call - if the equivalent of C<time ()> is used to
2875 update file times then there will be a small window where the kernel uses
2876 the previous second to update file times but libev might already execute
2877 the timer callback).
2879 =head3 Watcher-Specific Functions and Data Members
2881 =over 4
2883 =item ev_stat_init (ev_stat *, callback, const char *path, ev_tstamp interval)
2885 =item ev_stat_set (ev_stat *, const char *path, ev_tstamp interval)
2887 Configures the watcher to wait for status changes of the given
2888 C<path>. The C<interval> is a hint on how quickly a change is expected to
2889 be detected and should normally be specified as C<0> to let libev choose
2890 a suitable value. The memory pointed to by C<path> must point to the same
2891 path for as long as the watcher is active.
2893 The callback will receive an C<EV_STAT> event when a change was detected,
2894 relative to the attributes at the time the watcher was started (or the
2895 last change was detected).
2897 =item ev_stat_stat (loop, ev_stat *)
2899 Updates the stat buffer immediately with new values. If you change the
2900 watched path in your callback, you could call this function to avoid
2901 detecting this change (while introducing a race condition if you are not
2902 the only one changing the path). Can also be useful simply to find out the
2903 new values.
2905 =item ev_statdata attr [read-only]
2907 The most-recently detected attributes of the file. Although the type is
2908 C<ev_statdata>, this is usually the (or one of the) C<struct stat> types
2909 suitable for your system, but you can only rely on the POSIX-standardised
2910 members to be present. If the C<st_nlink> member is C<0>, then there was
2911 some error while C<stat>ing the file.
2913 =item ev_statdata prev [read-only]
2915 The previous attributes of the file. The callback gets invoked whenever
2916 C<prev> != C<attr>, or, more precisely, one or more of these members
2917 differ: C<st_dev>, C<st_ino>, C<st_mode>, C<st_nlink>, C<st_uid>,
2918 C<st_gid>, C<st_rdev>, C<st_size>, C<st_atime>, C<st_mtime>, C<st_ctime>.
2920 =item ev_tstamp interval [read-only]
2922 The specified interval.
2924 =item const char *path [read-only]
2926 The file system path that is being watched.
2928 =back
2930 =head3 Examples
2932 Example: Watch C</etc/passwd> for attribute changes.
2934 static void
2935 passwd_cb (struct ev_loop *loop, ev_stat *w, int revents)
2936 {
2937 /* /etc/passwd changed in some way */
2938 if (w->attr.st_nlink)
2939 {
2940 printf ("passwd current size %ld\n", (long)w->attr.st_size);
2941 printf ("passwd current atime %ld\n", (long)w->attr.st_mtime);
2942 printf ("passwd current mtime %ld\n", (long)w->attr.st_mtime);
2943 }
2944 else
2945 /* you shalt not abuse printf for puts */
2946 puts ("wow, /etc/passwd is not there, expect problems. "
2947 "if this is windows, they already arrived\n");
2948 }
2950 ...
2951 ev_stat passwd;
2953 ev_stat_init (&passwd, passwd_cb, "/etc/passwd", 0.);
2954 ev_stat_start (loop, &passwd);
2956 Example: Like above, but additionally use a one-second delay so we do not
2957 miss updates (however, frequent updates will delay processing, too, so
2958 one might do the work both on C<ev_stat> callback invocation I<and> on
2959 C<ev_timer> callback invocation).
2961 static ev_stat passwd;
2962 static ev_timer timer;
2964 static void
2965 timer_cb (EV_P_ ev_timer *w, int revents)
2966 {
2967 ev_timer_stop (EV_A_ w);
2969 /* now it's one second after the most recent passwd change */
2970 }
2972 static void
2973 stat_cb (EV_P_ ev_stat *w, int revents)
2974 {
2975 /* reset the one-second timer */
2976 ev_timer_again (EV_A_ &timer);
2977 }
2979 ...
2980 ev_stat_init (&passwd, stat_cb, "/etc/passwd", 0.);
2981 ev_stat_start (loop, &passwd);
2982 ev_timer_init (&timer, timer_cb, 0., 1.02);
2985 =head2 C<ev_idle> - when you've got nothing better to do...
2987 Idle watchers trigger events when no other events of the same or higher
2988 priority are pending (prepare, check and other idle watchers do not count
2989 as receiving "events").
2991 That is, as long as your process is busy handling sockets or timeouts
2992 (or even signals, imagine) of the same or higher priority it will not be
2993 triggered. But when your process is idle (or only lower-priority watchers
2994 are pending), the idle watchers are being called once per event loop
2995 iteration - until stopped, that is, or your process receives more events
2996 and becomes busy again with higher priority stuff.
2998 The most noteworthy effect is that as long as any idle watchers are
2999 active, the process will not block when waiting for new events.
3001 Apart from keeping your process non-blocking (which is a useful
3002 effect on its own sometimes), idle watchers are a good place to do
3003 "pseudo-background processing", or delay processing stuff to after the
3004 event loop has handled all outstanding events.
3006 =head3 Abusing an C<ev_idle> watcher for its side-effect
3008 As long as there is at least one active idle watcher, libev will never
3009 sleep unnecessarily. Or in other words, it will loop as fast as possible.
3010 For this to work, the idle watcher doesn't need to be invoked at all - the
3011 lowest priority will do.
3013 This mode of operation can be useful together with an C<ev_check> watcher,
3014 to do something on each event loop iteration - for example to balance load
3015 between different connections.
3017 See L</Abusing an ev_check watcher for its side-effect> for a longer
3018 example.
3020 =head3 Watcher-Specific Functions and Data Members
3022 =over 4
3024 =item ev_idle_init (ev_idle *, callback)
3026 Initialises and configures the idle watcher - it has no parameters of any
3027 kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,
3028 believe me.
3030 =back
3032 =head3 Examples
3034 Example: Dynamically allocate an C<ev_idle> watcher, start it, and in the
3035 callback, free it. Also, use no error checking, as usual.
3037 static void
3038 idle_cb (struct ev_loop *loop, ev_idle *w, int revents)
3039 {
3040 // stop the watcher
3041 ev_idle_stop (loop, w);
3043 // now we can free it
3044 free (w);
3046 // now do something you wanted to do when the program has
3047 // no longer anything immediate to do.
3048 }
3050 ev_idle *idle_watcher = malloc (sizeof (ev_idle));
3051 ev_idle_init (idle_watcher, idle_cb);
3052 ev_idle_start (loop, idle_watcher);
3055 =head2 C<ev_prepare> and C<ev_check> - customise your event loop!
3057 Prepare and check watchers are often (but not always) used in pairs:
3058 prepare watchers get invoked before the process blocks and check watchers
3059 afterwards.
3061 You I<must not> call C<ev_run> (or similar functions that enter the
3062 current event loop) or C<ev_loop_fork> from either C<ev_prepare> or
3063 C<ev_check> watchers. Other loops than the current one are fine,
3064 however. The rationale behind this is that you do not need to check
3065 for recursion in those watchers, i.e. the sequence will always be
3066 C<ev_prepare>, blocking, C<ev_check> so if you have one watcher of each
3067 kind they will always be called in pairs bracketing the blocking call.
3069 Their main purpose is to integrate other event mechanisms into libev and
3070 their use is somewhat advanced. They could be used, for example, to track
3071 variable changes, implement your own watchers, integrate net-snmp or a
3072 coroutine library and lots more. They are also occasionally useful if
3073 you cache some data and want to flush it before blocking (for example,
3074 in X programs you might want to do an C<XFlush ()> in an C<ev_prepare>
3075 watcher).
3077 This is done by examining in each prepare call which file descriptors
3078 need to be watched by the other library, registering C<ev_io> watchers
3079 for them and starting an C<ev_timer> watcher for any timeouts (many
3080 libraries provide exactly this functionality). Then, in the check watcher,
3081 you check for any events that occurred (by checking the pending status
3082 of all watchers and stopping them) and call back into the library. The
3083 I/O and timer callbacks will never actually be called (but must be valid
3084 nevertheless, because you never know, you know?).
3086 As another example, the Perl Coro module uses these hooks to integrate
3087 coroutines into libev programs, by yielding to other active coroutines
3088 during each prepare and only letting the process block if no coroutines
3089 are ready to run (it's actually more complicated: it only runs coroutines
3090 with priority higher than or equal to the event loop and one coroutine
3091 of lower priority, but only once, using idle watchers to keep the event
3092 loop from blocking if lower-priority coroutines are active, thus mapping
3093 low-priority coroutines to idle/background tasks).
3095 When used for this purpose, it is recommended to give C<ev_check> watchers
3096 highest (C<EV_MAXPRI>) priority, to ensure that they are being run before
3097 any other watchers after the poll (this doesn't matter for C<ev_prepare>
3098 watchers).
3100 Also, C<ev_check> watchers (and C<ev_prepare> watchers, too) should not
3101 activate ("feed") events into libev. While libev fully supports this, they
3102 might get executed before other C<ev_check> watchers did their job. As
3103 C<ev_check> watchers are often used to embed other (non-libev) event
3104 loops those other event loops might be in an unusable state until their
3105 C<ev_check> watcher ran (always remind yourself to coexist peacefully with
3106 others).
3108 =head3 Abusing an C<ev_check> watcher for its side-effect
3110 C<ev_check> (and less often also C<ev_prepare>) watchers can also be
3111 useful because they are called once per event loop iteration. For
3112 example, if you want to handle a large number of connections fairly, you
3113 normally only do a bit of work for each active connection, and if there
3114 is more work to do, you wait for the next event loop iteration, so other
3115 connections have a chance of making progress.
3117 Using an C<ev_check> watcher is almost enough: it will be called on the
3118 next event loop iteration. However, that isn't as soon as possible -
3119 without external events, your C<ev_check> watcher will not be invoked.
3121 This is where C<ev_idle> watchers come in handy - all you need is a
3122 single global idle watcher that is active as long as you have one active
3123 C<ev_check> watcher. The C<ev_idle> watcher makes sure the event loop
3124 will not sleep, and the C<ev_check> watcher makes sure a callback gets
3125 invoked. Neither watcher alone can do that.
3127 =head3 Watcher-Specific Functions and Data Members
3129 =over 4
3131 =item ev_prepare_init (ev_prepare *, callback)
3133 =item ev_check_init (ev_check *, callback)
3135 Initialises and configures the prepare or check watcher - they have no
3136 parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>
3137 macros, but using them is utterly, utterly, utterly and completely
3138 pointless.
3140 =back
3142 =head3 Examples
3144 There are a number of principal ways to embed other event loops or modules
3145 into libev. Here are some ideas on how to include libadns into libev
3146 (there is a Perl module named C<EV::ADNS> that does this, which you could
3147 use as a working example. Another Perl module named C<EV::Glib> embeds a
3148 Glib main context into libev, and finally, C<Glib::EV> embeds EV into the
3149 Glib event loop).
3151 Method 1: Add IO watchers and a timeout watcher in a prepare handler,
3152 and in a check watcher, destroy them and call into libadns. What follows
3153 is pseudo-code only of course. This requires you to either use a low
3154 priority for the check watcher or use C<ev_clear_pending> explicitly, as
3155 the callbacks for the IO/timeout watchers might not have been called yet.
3157 static ev_io iow [nfd];
3158 static ev_timer tw;
3160 static void
3161 io_cb (struct ev_loop *loop, ev_io *w, int revents)
3162 {
3163 }
3165 // create io watchers for each fd and a timer before blocking
3166 static void
3167 adns_prepare_cb (struct ev_loop *loop, ev_prepare *w, int revents)
3168 {
3169 int timeout = 3600000;
3170 struct pollfd fds [nfd];
3171 // actual code will need to loop here and realloc etc.
3172 adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));
3174 /* the callback is illegal, but won't be called as we stop during check */
3175 ev_timer_init (&tw, 0, timeout * 1e-3, 0.);
3176 ev_timer_start (loop, &tw);
3178 // create one ev_io per pollfd
3179 for (int i = 0; i < nfd; ++i)
3180 {
3181 ev_io_init (iow + i, io_cb, fds [i].fd,
3182 ((fds [i].events & POLLIN ? EV_READ : 0)
3183 | (fds [i].events & POLLOUT ? EV_WRITE : 0)));
3185 fds [i].revents = 0;
3186 ev_io_start (loop, iow + i);
3187 }
3188 }
3190 // stop all watchers after blocking
3191 static void
3192 adns_check_cb (struct ev_loop *loop, ev_check *w, int revents)
3193 {
3194 ev_timer_stop (loop, &tw);
3196 for (int i = 0; i < nfd; ++i)
3197 {
3198 // set the relevant poll flags
3199 // could also call adns_processreadable etc. here
3200 struct pollfd *fd = fds + i;
3201 int revents = ev_clear_pending (iow + i);
3202 if (revents & EV_READ ) fd->revents |= fd->events & POLLIN;
3203 if (revents & EV_WRITE) fd->revents |= fd->events & POLLOUT;
3205 // now stop the watcher
3206 ev_io_stop (loop, iow + i);
3207 }
3209 adns_afterpoll (adns, fds, nfd, timeval_from (ev_now (loop));
3210 }
3212 Method 2: This would be just like method 1, but you run C<adns_afterpoll>
3213 in the prepare watcher and would dispose of the check watcher.
3215 Method 3: If the module to be embedded supports explicit event
3216 notification (libadns does), you can also make use of the actual watcher
3217 callbacks, and only destroy/create the watchers in the prepare watcher.
3219 static void
3220 timer_cb (EV_P_ ev_timer *w, int revents)
3221 {
3222 adns_state ads = (adns_state)w->data;
3223 update_now (EV_A);
3225 adns_processtimeouts (ads, &tv_now);
3226 }
3228 static void
3229 io_cb (EV_P_ ev_io *w, int revents)
3230 {
3231 adns_state ads = (adns_state)w->data;
3232 update_now (EV_A);
3234 if (revents & EV_READ ) adns_processreadable (ads, w->fd, &tv_now);
3235 if (revents & EV_WRITE) adns_processwriteable (ads, w->fd, &tv_now);
3236 }
3238 // do not ever call adns_afterpoll
3240 Method 4: Do not use a prepare or check watcher because the module you
3241 want to embed is not flexible enough to support it. Instead, you can
3242 override their poll function. The drawback with this solution is that the
3243 main loop is now no longer controllable by EV. The C<Glib::EV> module uses
3244 this approach, effectively embedding EV as a client into the horrible
3245 libglib event loop.
3247 static gint
3248 event_poll_func (GPollFD *fds, guint nfds, gint timeout)
3249 {
3250 int got_events = 0;
3252 for (n = 0; n < nfds; ++n)
3253 // create/start io watcher that sets the relevant bits in fds[n] and increment got_events
3255 if (timeout >= 0)
3256 // create/start timer
3258 // poll
3259 ev_run (EV_A_ 0);
3261 // stop timer again
3262 if (timeout >= 0)
3263 ev_timer_stop (EV_A_ &to);
3265 // stop io watchers again - their callbacks should have set
3266 for (n = 0; n < nfds; ++n)
3267 ev_io_stop (EV_A_ iow [n]);
3269 return got_events;
3270 }
3273 =head2 C<ev_embed> - when one backend isn't enough...
3275 This is a rather advanced watcher type that lets you embed one event loop
3276 into another (currently only C<ev_io> events are supported in the embedded
3277 loop, other types of watchers might be handled in a delayed or incorrect
3278 fashion and must not be used).
3280 There are primarily two reasons you would want that: work around bugs and
3281 prioritise I/O.
3283 As an example for a bug workaround, the kqueue backend might only support
3284 sockets on some platform, so it is unusable as generic backend, but you
3285 still want to make use of it because you have many sockets and it scales
3286 so nicely. In this case, you would create a kqueue-based loop and embed
3287 it into your default loop (which might use e.g. poll). Overall operation
3288 will be a bit slower because first libev has to call C<poll> and then
3289 C<kevent>, but at least you can use both mechanisms for what they are
3290 best: C<kqueue> for scalable sockets and C<poll> if you want it to work :)
3292 As for prioritising I/O: under rare circumstances you have the case where
3293 some fds have to be watched and handled very quickly (with low latency),
3294 and even priorities and idle watchers might have too much overhead. In
3295 this case you would put all the high priority stuff in one loop and all
3296 the rest in a second one, and embed the second one in the first.
3298 As long as the watcher is active, the callback will be invoked every
3299 time there might be events pending in the embedded loop. The callback
3300 must then call C<ev_embed_sweep (mainloop, watcher)> to make a single
3301 sweep and invoke their callbacks (the callback doesn't need to invoke the
3302 C<ev_embed_sweep> function directly, it could also start an idle watcher
3303 to give the embedded loop strictly lower priority for example).
3305 You can also set the callback to C<0>, in which case the embed watcher
3306 will automatically execute the embedded loop sweep whenever necessary.
3308 Fork detection will be handled transparently while the C<ev_embed> watcher
3309 is active, i.e., the embedded loop will automatically be forked when the
3310 embedding loop forks. In other cases, the user is responsible for calling
3311 C<ev_loop_fork> on the embedded loop.
3313 Unfortunately, not all backends are embeddable: only the ones returned by
3314 C<ev_embeddable_backends> are, which, unfortunately, does not include any
3315 portable one.
3317 So when you want to use this feature you will always have to be prepared
3318 that you cannot get an embeddable loop. The recommended way to get around
3319 this is to have a separate variables for your embeddable loop, try to
3320 create it, and if that fails, use the normal loop for everything.
3322 =head3 C<ev_embed> and fork
3324 While the C<ev_embed> watcher is running, forks in the embedding loop will
3325 automatically be applied to the embedded loop as well, so no special
3326 fork handling is required in that case. When the watcher is not running,
3327 however, it is still the task of the libev user to call C<ev_loop_fork ()>
3328 as applicable.
3330 =head3 Watcher-Specific Functions and Data Members
3332 =over 4
3334 =item ev_embed_init (ev_embed *, callback, struct ev_loop *embedded_loop)
3336 =item ev_embed_set (ev_embed *, struct ev_loop *embedded_loop)
3338 Configures the watcher to embed the given loop, which must be
3339 embeddable. If the callback is C<0>, then C<ev_embed_sweep> will be
3340 invoked automatically, otherwise it is the responsibility of the callback
3341 to invoke it (it will continue to be called until the sweep has been done,
3342 if you do not want that, you need to temporarily stop the embed watcher).
3344 =item ev_embed_sweep (loop, ev_embed *)
3346 Make a single, non-blocking sweep over the embedded loop. This works
3347 similarly to C<ev_run (embedded_loop, EVRUN_NOWAIT)>, but in the most
3348 appropriate way for embedded loops.
3350 =item struct ev_loop *other [read-only]
3352 The embedded event loop.
3354 =back
3356 =head3 Examples
3358 Example: Try to get an embeddable event loop and embed it into the default
3359 event loop. If that is not possible, use the default loop. The default
3360 loop is stored in C<loop_hi>, while the embeddable loop is stored in
3361 C<loop_lo> (which is C<loop_hi> in the case no embeddable loop can be
3362 used).
3364 struct ev_loop *loop_hi = ev_default_init (0);
3365 struct ev_loop *loop_lo = 0;
3366 ev_embed embed;
3368 // see if there is a chance of getting one that works
3369 // (remember that a flags value of 0 means autodetection)
3370 loop_lo = ev_embeddable_backends () & ev_recommended_backends ()
3371 ? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())
3372 : 0;
3374 // if we got one, then embed it, otherwise default to loop_hi
3375 if (loop_lo)
3376 {
3377 ev_embed_init (&embed, 0, loop_lo);
3378 ev_embed_start (loop_hi, &embed);
3379 }
3380 else
3381 loop_lo = loop_hi;
3383 Example: Check if kqueue is available but not recommended and create
3384 a kqueue backend for use with sockets (which usually work with any
3385 kqueue implementation). Store the kqueue/socket-only event loop in
3386 C<loop_socket>. (One might optionally use C<EVFLAG_NOENV>, too).
3388 struct ev_loop *loop = ev_default_init (0);
3389 struct ev_loop *loop_socket = 0;
3390 ev_embed embed;
3392 if (ev_supported_backends () & ~ev_recommended_backends () & EVBACKEND_KQUEUE)
3393 if ((loop_socket = ev_loop_new (EVBACKEND_KQUEUE))
3394 {
3395 ev_embed_init (&embed, 0, loop_socket);
3396 ev_embed_start (loop, &embed);
3397 }
3399 if (!loop_socket)
3400 loop_socket = loop;
3402 // now use loop_socket for all sockets, and loop for everything else
3405 =head2 C<ev_fork> - the audacity to resume the event loop after a fork
3407 Fork watchers are called when a C<fork ()> was detected (usually because
3408 whoever is a good citizen cared to tell libev about it by calling
3409 C<ev_loop_fork>). The invocation is done before the event loop blocks next
3410 and before C<ev_check> watchers are being called, and only in the child
3411 after the fork. If whoever good citizen calling C<ev_default_fork> cheats
3412 and calls it in the wrong process, the fork handlers will be invoked, too,
3413 of course.
3415 =head3 The special problem of life after fork - how is it possible?
3417 Most uses of C<fork ()> consist of forking, then some simple calls to set
3418 up/change the process environment, followed by a call to C<exec()>. This
3419 sequence should be handled by libev without any problems.
3421 This changes when the application actually wants to do event handling
3422 in the child, or both parent in child, in effect "continuing" after the
3423 fork.
3425 The default mode of operation (for libev, with application help to detect
3426 forks) is to duplicate all the state in the child, as would be expected
3427 when I<either> the parent I<or> the child process continues.
3429 When both processes want to continue using libev, then this is usually the
3430 wrong result. In that case, usually one process (typically the parent) is
3431 supposed to continue with all watchers in place as before, while the other
3432 process typically wants to start fresh, i.e. without any active watchers.
3434 The cleanest and most efficient way to achieve that with libev is to
3435 simply create a new event loop, which of course will be "empty", and
3436 use that for new watchers. This has the advantage of not touching more
3437 memory than necessary, and thus avoiding the copy-on-write, and the
3438 disadvantage of having to use multiple event loops (which do not support
3439 signal watchers).
3441 When this is not possible, or you want to use the default loop for
3442 other reasons, then in the process that wants to start "fresh", call
3443 C<ev_loop_destroy (EV_DEFAULT)> followed by C<ev_default_loop (...)>.
3444 Destroying the default loop will "orphan" (not stop) all registered
3445 watchers, so you have to be careful not to execute code that modifies
3446 those watchers. Note also that in that case, you have to re-register any
3447 signal watchers.
3449 =head3 Watcher-Specific Functions and Data Members
3451 =over 4
3453 =item ev_fork_init (ev_fork *, callback)
3455 Initialises and configures the fork watcher - it has no parameters of any
3456 kind. There is a C<ev_fork_set> macro, but using it is utterly pointless,
3457 really.
3459 =back
3462 =head2 C<ev_cleanup> - even the best things end
3464 Cleanup watchers are called just before the event loop is being destroyed
3465 by a call to C<ev_loop_destroy>.
3467 While there is no guarantee that the event loop gets destroyed, cleanup
3468 watchers provide a convenient method to install cleanup hooks for your
3469 program, worker threads and so on - you just to make sure to destroy the
3470 loop when you want them to be invoked.
3472 Cleanup watchers are invoked in the same way as any other watcher. Unlike
3473 all other watchers, they do not keep a reference to the event loop (which
3474 makes a lot of sense if you think about it). Like all other watchers, you
3475 can call libev functions in the callback, except C<ev_cleanup_start>.
3477 =head3 Watcher-Specific Functions and Data Members
3479 =over 4
3481 =item ev_cleanup_init (ev_cleanup *, callback)
3483 Initialises and configures the cleanup watcher - it has no parameters of
3484 any kind. There is a C<ev_cleanup_set> macro, but using it is utterly
3485 pointless, I assure you.
3487 =back
3489 Example: Register an atexit handler to destroy the default loop, so any
3490 cleanup functions are called.
3492 static void
3493 program_exits (void)
3494 {
3495 ev_loop_destroy (EV_DEFAULT_UC);
3496 }
3498 ...
3499 atexit (program_exits);
3502 =head2 C<ev_async> - how to wake up an event loop
3504 In general, you cannot use an C<ev_loop> from multiple threads or other
3505 asynchronous sources such as signal handlers (as opposed to multiple event
3506 loops - those are of course safe to use in different threads).
3508 Sometimes, however, you need to wake up an event loop you do not control,
3509 for example because it belongs to another thread. This is what C<ev_async>
3510 watchers do: as long as the C<ev_async> watcher is active, you can signal
3511 it by calling C<ev_async_send>, which is thread- and signal safe.
3513 This functionality is very similar to C<ev_signal> watchers, as signals,
3514 too, are asynchronous in nature, and signals, too, will be compressed
3515 (i.e. the number of callback invocations may be less than the number of
3516 C<ev_async_send> calls). In fact, you could use signal watchers as a kind
3517 of "global async watchers" by using a watcher on an otherwise unused
3518 signal, and C<ev_feed_signal> to signal this watcher from another thread,
3519 even without knowing which loop owns the signal.
3521 =head3 Queueing
3523 C<ev_async> does not support queueing of data in any way. The reason
3524 is that the author does not know of a simple (or any) algorithm for a
3525 multiple-writer-single-reader queue that works in all cases and doesn't
3526 need elaborate support such as pthreads or unportable memory access
3527 semantics.
3529 That means that if you want to queue data, you have to provide your own
3530 queue. But at least I can tell you how to implement locking around your
3531 queue:
3533 =over 4
3535 =item queueing from a signal handler context
3537 To implement race-free queueing, you simply add to the queue in the signal
3538 handler but you block the signal handler in the watcher callback. Here is
3539 an example that does that for some fictitious SIGUSR1 handler:
3541 static ev_async mysig;
3543 static void
3544 sigusr1_handler (void)
3545 {
3546 sometype data;
3548 // no locking etc.
3549 queue_put (data);
3550 ev_async_send (EV_DEFAULT_ &mysig);
3551 }
3553 static void
3554 mysig_cb (EV_P_ ev_async *w, int revents)
3555 {
3556 sometype data;
3557 sigset_t block, prev;
3559 sigemptyset (&block);
3560 sigaddset (&block, SIGUSR1);
3561 sigprocmask (SIG_BLOCK, &block, &prev);
3563 while (queue_get (&data))
3564 process (data);
3566 if (sigismember (&prev, SIGUSR1)
3567 sigprocmask (SIG_UNBLOCK, &block, 0);
3568 }
3570 (Note: pthreads in theory requires you to use C<pthread_setmask>
3571 instead of C<sigprocmask> when you use threads, but libev doesn't do it
3572 either...).
3574 =item queueing from a thread context
3576 The strategy for threads is different, as you cannot (easily) block
3577 threads but you can easily preempt them, so to queue safely you need to
3578 employ a traditional mutex lock, such as in this pthread example:
3580 static ev_async mysig;
3581 static pthread_mutex_t mymutex = PTHREAD_MUTEX_INITIALIZER;
3583 static void
3584 otherthread (void)
3585 {
3586 // only need to lock the actual queueing operation
3587 pthread_mutex_lock (&mymutex);
3588 queue_put (data);
3589 pthread_mutex_unlock (&mymutex);
3591 ev_async_send (EV_DEFAULT_ &mysig);
3592 }
3594 static void
3595 mysig_cb (EV_P_ ev_async *w, int revents)
3596 {
3597 pthread_mutex_lock (&mymutex);
3599 while (queue_get (&data))
3600 process (data);
3602 pthread_mutex_unlock (&mymutex);
3603 }
3605 =back
3608 =head3 Watcher-Specific Functions and Data Members
3610 =over 4
3612 =item ev_async_init (ev_async *, callback)
3614 Initialises and configures the async watcher - it has no parameters of any
3615 kind. There is a C<ev_async_set> macro, but using it is utterly pointless,
3616 trust me.
3618 =item ev_async_send (loop, ev_async *)
3620 Sends/signals/activates the given C<ev_async> watcher, that is, feeds
3621 an C<EV_ASYNC> event on the watcher into the event loop, and instantly
3622 returns.
3624 Unlike C<ev_feed_event>, this call is safe to do from other threads,
3625 signal or similar contexts (see the discussion of C<EV_ATOMIC_T> in the
3626 embedding section below on what exactly this means).
3628 Note that, as with other watchers in libev, multiple events might get
3629 compressed into a single callback invocation (another way to look at
3630 this is that C<ev_async> watchers are level-triggered: they are set on
3631 C<ev_async_send>, reset when the event loop detects that).
3633 This call incurs the overhead of at most one extra system call per event
3634 loop iteration, if the event loop is blocked, and no syscall at all if
3635 the event loop (or your program) is processing events. That means that
3636 repeated calls are basically free (there is no need to avoid calls for
3637 performance reasons) and that the overhead becomes smaller (typically
3638 zero) under load.
3640 =item bool = ev_async_pending (ev_async *)
3642 Returns a non-zero value when C<ev_async_send> has been called on the
3643 watcher but the event has not yet been processed (or even noted) by the
3644 event loop.
3646 C<ev_async_send> sets a flag in the watcher and wakes up the loop. When
3647 the loop iterates next and checks for the watcher to have become active,
3648 it will reset the flag again. C<ev_async_pending> can be used to very
3649 quickly check whether invoking the loop might be a good idea.
3651 Not that this does I<not> check whether the watcher itself is pending,
3652 only whether it has been requested to make this watcher pending: there
3653 is a time window between the event loop checking and resetting the async
3654 notification, and the callback being invoked.
3656 =back
3661 There are some other functions of possible interest. Described. Here. Now.
3663 =over 4
3665 =item ev_once (loop, int fd, int events, ev_tstamp timeout, callback, arg)
3667 This function combines a simple timer and an I/O watcher, calls your
3668 callback on whichever event happens first and automatically stops both
3669 watchers. This is useful if you want to wait for a single event on an fd
3670 or timeout without having to allocate/configure/start/stop/free one or
3671 more watchers yourself.
3673 If C<fd> is less than 0, then no I/O watcher will be started and the
3674 C<events> argument is being ignored. Otherwise, an C<ev_io> watcher for
3675 the given C<fd> and C<events> set will be created and started.
3677 If C<timeout> is less than 0, then no timeout watcher will be
3678 started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and
3679 repeat = 0) will be started. C<0> is a valid timeout.
3681 The callback has the type C<void (*cb)(int revents, void *arg)> and is
3682 passed an C<revents> set like normal event callbacks (a combination of
3683 C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMER>) and the C<arg>
3684 value passed to C<ev_once>. Note that it is possible to receive I<both>
3685 a timeout and an io event at the same time - you probably should give io
3686 events precedence.
3688 Example: wait up to ten seconds for data to appear on STDIN_FILENO.
3690 static void stdin_ready (int revents, void *arg)
3691 {
3692 if (revents & EV_READ)
3693 /* stdin might have data for us, joy! */;
3694 else if (revents & EV_TIMER)
3695 /* doh, nothing entered */;
3696 }
3698 ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);
3700 =item ev_feed_fd_event (loop, int fd, int revents)
3702 Feed an event on the given fd, as if a file descriptor backend detected
3703 the given events.
3705 =item ev_feed_signal_event (loop, int signum)
3707 Feed an event as if the given signal occurred. See also C<ev_feed_signal>,
3708 which is async-safe.
3710 =back
3715 This section explains some common idioms that are not immediately
3716 obvious. Note that examples are sprinkled over the whole manual, and this
3717 section only contains stuff that wouldn't fit anywhere else.
3721 Each watcher has, by default, a C<void *data> member that you can read
3722 or modify at any time: libev will completely ignore it. This can be used
3723 to associate arbitrary data with your watcher. If you need more data and
3724 don't want to allocate memory separately and store a pointer to it in that
3725 data member, you can also "subclass" the watcher type and provide your own
3726 data:
3728 struct my_io
3729 {
3730 ev_io io;
3731 int otherfd;
3732 void *somedata;
3733 struct whatever *mostinteresting;
3734 };
3736 ...
3737 struct my_io w;
3738 ev_io_init (&, my_cb, fd, EV_READ);
3740 And since your callback will be called with a pointer to the watcher, you
3741 can cast it back to your own type:
3743 static void my_cb (struct ev_loop *loop, ev_io *w_, int revents)
3744 {
3745 struct my_io *w = (struct my_io *)w_;
3746 ...
3747 }
3749 More interesting and less C-conformant ways of casting your callback
3750 function type instead have been omitted.
3754 Another common scenario is to use some data structure with multiple
3755 embedded watchers, in effect creating your own watcher that combines
3756 multiple libev event sources into one "super-watcher":
3758 struct my_biggy
3759 {
3760 int some_data;
3761 ev_timer t1;
3762 ev_timer t2;
3763 }
3765 In this case getting the pointer to C<my_biggy> is a bit more
3766 complicated: Either you store the address of your C<my_biggy> struct in
3767 the C<data> member of the watcher (for woozies or C++ coders), or you need
3768 to use some pointer arithmetic using C<offsetof> inside your watchers (for
3769 real programmers):
3771 #include <stddef.h>
3773 static void
3774 t1_cb (EV_P_ ev_timer *w, int revents)
3775 {
3776 struct my_biggy big = (struct my_biggy *)
3777 (((char *)w) - offsetof (struct my_biggy, t1));
3778 }
3780 static void
3781 t2_cb (EV_P_ ev_timer *w, int revents)
3782 {
3783 struct my_biggy big = (struct my_biggy *)
3784 (((char *)w) - offsetof (struct my_biggy, t2));
3785 }
3789 Often you have structures like this in event-based programs:
3791 callback ()
3792 {
3793 free (request);
3794 }
3796 request = start_new_request (..., callback);
3798 The intent is to start some "lengthy" operation. The C<request> could be
3799 used to cancel the operation, or do other things with it.
3801 It's not uncommon to have code paths in C<start_new_request> that
3802 immediately invoke the callback, for example, to report errors. Or you add
3803 some caching layer that finds that it can skip the lengthy aspects of the
3804 operation and simply invoke the callback with the result.
3806 The problem here is that this will happen I<before> C<start_new_request>
3807 has returned, so C<request> is not set.
3809 Even if you pass the request by some safer means to the callback, you
3810 might want to do something to the request after starting it, such as
3811 canceling it, which probably isn't working so well when the callback has
3812 already been invoked.
3814 A common way around all these issues is to make sure that
3815 C<start_new_request> I<always> returns before the callback is invoked. If
3816 C<start_new_request> immediately knows the result, it can artificially
3817 delay invoking the callback by using a C<prepare> or C<idle> watcher for
3818 example, or more sneakily, by reusing an existing (stopped) watcher and
3819 pushing it into the pending queue:
3821 ev_set_cb (watcher, callback);
3822 ev_feed_event (EV_A_ watcher, 0);
3824 This way, C<start_new_request> can safely return before the callback is
3825 invoked, while not delaying callback invocation too much.
3829 Often (especially in GUI toolkits) there are places where you have
3830 I<modal> interaction, which is most easily implemented by recursively
3831 invoking C<ev_run>.
3833 This brings the problem of exiting - a callback might want to finish the
3834 main C<ev_run> call, but not the nested one (e.g. user clicked "Quit", but
3835 a modal "Are you sure?" dialog is still waiting), or just the nested one
3836 and not the main one (e.g. user clocked "Ok" in a modal dialog), or some
3837 other combination: In these cases, a simple C<ev_break> will not work.
3839 The solution is to maintain "break this loop" variable for each C<ev_run>
3840 invocation, and use a loop around C<ev_run> until the condition is
3841 triggered, using C<EVRUN_ONCE>:
3843 // main loop
3844 int exit_main_loop = 0;
3846 while (!exit_main_loop)
3847 ev_run (EV_DEFAULT_ EVRUN_ONCE);
3849 // in a modal watcher
3850 int exit_nested_loop = 0;
3852 while (!exit_nested_loop)
3853 ev_run (EV_A_ EVRUN_ONCE);
3855 To exit from any of these loops, just set the corresponding exit variable:
3857 // exit modal loop
3858 exit_nested_loop = 1;
3860 // exit main program, after modal loop is finished
3861 exit_main_loop = 1;
3863 // exit both
3864 exit_main_loop = exit_nested_loop = 1;
3868 Here is a fictitious example of how to run an event loop in a different
3869 thread from where callbacks are being invoked and watchers are
3870 created/added/removed.
3872 For a real-world example, see the C<EV::Loop::Async> perl module,
3873 which uses exactly this technique (which is suited for many high-level
3874 languages).
3876 The example uses a pthread mutex to protect the loop data, a condition
3877 variable to wait for callback invocations, an async watcher to notify the
3878 event loop thread and an unspecified mechanism to wake up the main thread.
3880 First, you need to associate some data with the event loop:
3882 typedef struct {
3883 pthread_mutex_t lock; /* global loop lock */
3884 pthread_t tid;
3885 pthread_cond_t invoke_cv;
3886 ev_async async_w;
3887 } userdata;
3889 void prepare_loop (EV_P)
3890 {
3891 // for simplicity, we use a static userdata struct.
3892 static userdata u;
3894 ev_async_init (&u.async_w, async_cb);
3895 ev_async_start (EV_A_ &u.async_w);
3897 pthread_mutex_init (&u.lock, 0);
3898 pthread_cond_init (&u.invoke_cv, 0);
3900 // now associate this with the loop
3901 ev_set_userdata (EV_A_ &u);
3902 ev_set_invoke_pending_cb (EV_A_ l_invoke);
3903 ev_set_loop_release_cb (EV_A_ l_release, l_acquire);
3905 // then create the thread running ev_run
3906 pthread_create (&u.tid, 0, l_run, EV_A);
3907 }
3909 The callback for the C<ev_async> watcher does nothing: the watcher is used
3910 solely to wake up the event loop so it takes notice of any new watchers
3911 that might have been added:
3913 static void
3914 async_cb (EV_P_ ev_async *w, int revents)
3915 {
3916 // just used for the side effects
3917 }
3919 The C<l_release> and C<l_acquire> callbacks simply unlock/lock the mutex
3920 protecting the loop data, respectively.
3922 static void
3923 l_release (EV_P)
3924 {
3925 userdata *u = ev_userdata (EV_A);
3926 pthread_mutex_unlock (&u->lock);
3927 }
3929 static void
3930 l_acquire (EV_P)
3931 {
3932 userdata *u = ev_userdata (EV_A);
3933 pthread_mutex_lock (&u->lock);
3934 }
3936 The event loop thread first acquires the mutex, and then jumps straight
3937 into C<ev_run>:
3939 void *
3940 l_run (void *thr_arg)
3941 {
3942 struct ev_loop *loop = (struct ev_loop *)thr_arg;
3944 l_acquire (EV_A);
3945 pthread_setcanceltype (PTHREAD_CANCEL_ASYNCHRONOUS, 0);
3946 ev_run (EV_A_ 0);
3947 l_release (EV_A);
3949 return 0;
3950 }
3952 Instead of invoking all pending watchers, the C<l_invoke> callback will
3953 signal the main thread via some unspecified mechanism (signals? pipe
3954 writes? C<Async::Interrupt>?) and then waits until all pending watchers
3955 have been called (in a while loop because a) spurious wakeups are possible
3956 and b) skipping inter-thread-communication when there are no pending
3957 watchers is very beneficial):
3959 static void
3960 l_invoke (EV_P)
3961 {
3962 userdata *u = ev_userdata (EV_A);
3964 while (ev_pending_count (EV_A))
3965 {
3966 wake_up_other_thread_in_some_magic_or_not_so_magic_way ();
3967 pthread_cond_wait (&u->invoke_cv, &u->lock);
3968 }
3969 }
3971 Now, whenever the main thread gets told to invoke pending watchers, it
3972 will grab the lock, call C<ev_invoke_pending> and then signal the loop
3973 thread to continue:
3975 static void
3976 real_invoke_pending (EV_P)
3977 {
3978 userdata *u = ev_userdata (EV_A);
3980 pthread_mutex_lock (&u->lock);
3981 ev_invoke_pending (EV_A);
3982 pthread_cond_signal (&u->invoke_cv);
3983 pthread_mutex_unlock (&u->lock);
3984 }
3986 Whenever you want to start/stop a watcher or do other modifications to an
3987 event loop, you will now have to lock:
3989 ev_timer timeout_watcher;
3990 userdata *u = ev_userdata (EV_A);
3992 ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);
3994 pthread_mutex_lock (&u->lock);
3995 ev_timer_start (EV_A_ &timeout_watcher);
3996 ev_async_send (EV_A_ &u->async_w);
3997 pthread_mutex_unlock (&u->lock);
3999 Note that sending the C<ev_async> watcher is required because otherwise
4000 an event loop currently blocking in the kernel will have no knowledge
4001 about the newly added timer. By waking up the loop it will pick up any new
4002 watchers in the next event loop iteration.
4006 While the overhead of a callback that e.g. schedules a thread is small, it
4007 is still an overhead. If you embed libev, and your main usage is with some
4008 kind of threads or coroutines, you might want to customise libev so that
4009 doesn't need callbacks anymore.
4011 Imagine you have coroutines that you can switch to using a function
4012 C<switch_to (coro)>, that libev runs in a coroutine called C<libev_coro>
4013 and that due to some magic, the currently active coroutine is stored in a
4014 global called C<current_coro>. Then you can build your own "wait for libev
4015 event" primitive by changing C<EV_CB_DECLARE> and C<EV_CB_INVOKE> (note
4016 the differing C<;> conventions):
4018 #define EV_CB_DECLARE(type) struct my_coro *cb;
4019 #define EV_CB_INVOKE(watcher) switch_to ((watcher)->cb)
4021 That means instead of having a C callback function, you store the
4022 coroutine to switch to in each watcher, and instead of having libev call
4023 your callback, you instead have it switch to that coroutine.
4025 A coroutine might now wait for an event with a function called
4026 C<wait_for_event>. (the watcher needs to be started, as always, but it doesn't
4027 matter when, or whether the watcher is active or not when this function is
4028 called):
4030 void
4031 wait_for_event (ev_watcher *w)
4032 {
4033 ev_set_cb (w, current_coro);
4034 switch_to (libev_coro);
4035 }
4037 That basically suspends the coroutine inside C<wait_for_event> and
4038 continues the libev coroutine, which, when appropriate, switches back to
4039 this or any other coroutine.
4041 You can do similar tricks if you have, say, threads with an event queue -
4042 instead of storing a coroutine, you store the queue object and instead of
4043 switching to a coroutine, you push the watcher onto the queue and notify
4044 any waiters.
4046 To embed libev, see L</EMBEDDING>, but in short, it's easiest to create two
4047 files, F<my_ev.h> and F<my_ev.c> that include the respective libev files:
4049 // my_ev.h
4050 #define EV_CB_DECLARE(type) struct my_coro *cb;
4051 #define EV_CB_INVOKE(watcher) switch_to ((watcher)->cb)
4052 #include "../libev/ev.h"
4054 // my_ev.c
4055 #define EV_H "my_ev.h"
4056 #include "../libev/ev.c"
4058 And then use F<my_ev.h> when you would normally use F<ev.h>, and compile
4059 F<my_ev.c> into your project. When properly specifying include paths, you
4060 can even use F<ev.h> as header file name directly.
4065 Libev offers a compatibility emulation layer for libevent. It cannot
4066 emulate the internals of libevent, so here are some usage hints:
4068 =over 4
4070 =item * Only the libevent-1.4.1-beta API is being emulated.
4072 This was the newest libevent version available when libev was implemented,
4073 and is still mostly unchanged in 2010.
4075 =item * Use it by including <event.h>, as usual.
4077 =item * The following members are fully supported: ev_base, ev_callback,
4078 ev_arg, ev_fd, ev_res, ev_events.
4080 =item * Avoid using ev_flags and the EVLIST_*-macros, while it is
4081 maintained by libev, it does not work exactly the same way as in libevent (consider
4082 it a private API).
4084 =item * Priorities are not currently supported. Initialising priorities
4085 will fail and all watchers will have the same priority, even though there
4086 is an ev_pri field.
4088 =item * In libevent, the last base created gets the signals, in libev, the
4089 base that registered the signal gets the signals.
4091 =item * Other members are not supported.
4093 =item * The libev emulation is I<not> ABI compatible to libevent, you need
4094 to use the libev header file and library.
4096 =back
4098 =head1 C++ SUPPORT
4100 =head2 C API
4102 The normal C API should work fine when used from C++: both ev.h and the
4103 libev sources can be compiled as C++. Therefore, code that uses the C API
4104 will work fine.
4106 Proper exception specifications might have to be added to callbacks passed
4107 to libev: exceptions may be thrown only from watcher callbacks, all other
4108 callbacks (allocator, syserr, loop acquire/release and periodic reschedule
4109 callbacks) must not throw exceptions, and might need a C<noexcept>
4110 specification. If you have code that needs to be compiled as both C and
4111 C++ you can use the C<EV_NOEXCEPT> macro for this:
4113 static void
4114 fatal_error (const char *msg) EV_NOEXCEPT
4115 {
4116 perror (msg);
4117 abort ();
4118 }
4120 ...
4121 ev_set_syserr_cb (fatal_error);
4123 The only API functions that can currently throw exceptions are C<ev_run>,
4124 C<ev_invoke>, C<ev_invoke_pending> and C<ev_loop_destroy> (the latter
4125 because it runs cleanup watchers).
4127 Throwing exceptions in watcher callbacks is only supported if libev itself
4128 is compiled with a C++ compiler or your C and C++ environments allow
4129 throwing exceptions through C libraries (most do).
4131 =head2 C++ API
4133 Libev comes with some simplistic wrapper classes for C++ that mainly allow
4134 you to use some convenience methods to start/stop watchers and also change
4135 the callback model to a model using method callbacks on objects.
4137 To use it,
4139 #include <ev++.h>
4141 This automatically includes F<ev.h> and puts all of its definitions (many
4142 of them macros) into the global namespace. All C++ specific things are
4143 put into the C<ev> namespace. It should support all the same embedding
4144 options as F<ev.h>, most notably C<EV_MULTIPLICITY>.
4146 Care has been taken to keep the overhead low. The only data member the C++
4147 classes add (compared to plain C-style watchers) is the event loop pointer
4148 that the watcher is associated with (or no additional members at all if
4149 you disable C<EV_MULTIPLICITY> when embedding libev).
4151 Currently, functions, static and non-static member functions and classes
4152 with C<operator ()> can be used as callbacks. Other types should be easy
4153 to add as long as they only need one additional pointer for context. If
4154 you need support for other types of functors please contact the author
4155 (preferably after implementing it).
4157 For all this to work, your C++ compiler either has to use the same calling
4158 conventions as your C compiler (for static member functions), or you have
4159 to embed libev and compile libev itself as C++.
4161 Here is a list of things available in the C<ev> namespace:
4163 =over 4
4165 =item C<ev::READ>, C<ev::WRITE> etc.
4167 These are just enum values with the same values as the C<EV_READ> etc.
4168 macros from F<ev.h>.
4170 =item C<ev::tstamp>, C<ev::now>
4172 Aliases to the same types/functions as with the C<ev_> prefix.
4174 =item C<ev::io>, C<ev::timer>, C<ev::periodic>, C<ev::idle>, C<ev::sig> etc.
4176 For each C<ev_TYPE> watcher in F<ev.h> there is a corresponding class of
4177 the same name in the C<ev> namespace, with the exception of C<ev_signal>
4178 which is called C<ev::sig> to avoid clashes with the C<signal> macro
4179 defined by many implementations.
4181 All of those classes have these methods:
4183 =over 4
4185 =item ev::TYPE::TYPE ()
4187 =item ev::TYPE::TYPE (loop)
4189 =item ev::TYPE::~TYPE
4191 The constructor (optionally) takes an event loop to associate the watcher
4192 with. If it is omitted, it will use C<EV_DEFAULT>.
4194 The constructor calls C<ev_init> for you, which means you have to call the
4195 C<set> method before starting it.
4197 It will not set a callback, however: You have to call the templated C<set>
4198 method to set a callback before you can start the watcher.
4200 (The reason why you have to use a method is a limitation in C++ which does
4201 not allow explicit template arguments for constructors).
4203 The destructor automatically stops the watcher if it is active.
4205 =item w->set<class, &class::method> (object *)
4207 This method sets the callback method to call. The method has to have a
4208 signature of C<void (*)(ev_TYPE &, int)>, it receives the watcher as
4209 first argument and the C<revents> as second. The object must be given as
4210 parameter and is stored in the C<data> member of the watcher.
4212 This method synthesizes efficient thunking code to call your method from
4213 the C callback that libev requires. If your compiler can inline your
4214 callback (i.e. it is visible to it at the place of the C<set> call and
4215 your compiler is good :), then the method will be fully inlined into the
4216 thunking function, making it as fast as a direct C callback.
4218 Example: simple class declaration and watcher initialisation
4220 struct myclass
4221 {
4222 void io_cb (ev::io &w, int revents) { }
4223 }
4225 myclass obj;
4226 ev::io iow;
4227 iow.set <myclass, &myclass::io_cb> (&obj);
4229 =item w->set (object *)
4231 This is a variation of a method callback - leaving out the method to call
4232 will default the method to C<operator ()>, which makes it possible to use
4233 functor objects without having to manually specify the C<operator ()> all
4234 the time. Incidentally, you can then also leave out the template argument
4235 list.
4237 The C<operator ()> method prototype must be C<void operator ()(watcher &w,
4238 int revents)>.
4240 See the method-C<set> above for more details.
4242 Example: use a functor object as callback.
4244 struct myfunctor
4245 {
4246 void operator() (ev::io &w, int revents)
4247 {
4248 ...
4249 }
4250 }
4252 myfunctor f;
4254 ev::io w;
4255 w.set (&f);
4257 =item w->set<function> (void *data = 0)
4259 Also sets a callback, but uses a static method or plain function as
4260 callback. The optional C<data> argument will be stored in the watcher's
4261 C<data> member and is free for you to use.
4263 The prototype of the C<function> must be C<void (*)(ev::TYPE &w, int)>.
4265 See the method-C<set> above for more details.
4267 Example: Use a plain function as callback.
4269 static void io_cb (ev::io &w, int revents) { }
4270 iow.set <io_cb> ();
4272 =item w->set (loop)
4274 Associates a different C<struct ev_loop> with this watcher. You can only
4275 do this when the watcher is inactive (and not pending either).
4277 =item w->set ([arguments])
4279 Basically the same as C<ev_TYPE_set> (except for C<ev::embed> watchers>),
4280 with the same arguments. Either this method or a suitable start method
4281 must be called at least once. Unlike the C counterpart, an active watcher
4282 gets automatically stopped and restarted when reconfiguring it with this
4283 method.
4285 For C<ev::embed> watchers this method is called C<set_embed>, to avoid
4286 clashing with the C<set (loop)> method.
4288 For C<ev::io> watchers there is an additional C<set> method that acepts a
4289 new event mask only, and internally calls C<ev_io_modify>.
4291 =item w->start ()
4293 Starts the watcher. Note that there is no C<loop> argument, as the
4294 constructor already stores the event loop.
4296 =item w->start ([arguments])
4298 Instead of calling C<set> and C<start> methods separately, it is often
4299 convenient to wrap them in one call. Uses the same type of arguments as
4300 the configure C<set> method of the watcher.
4302 =item w->stop ()
4304 Stops the watcher if it is active. Again, no C<loop> argument.
4306 =item w->again () (C<ev::timer>, C<ev::periodic> only)
4308 For C<ev::timer> and C<ev::periodic>, this invokes the corresponding
4309 C<ev_TYPE_again> function.
4311 =item w->sweep () (C<ev::embed> only)
4313 Invokes C<ev_embed_sweep>.
4315 =item w->update () (C<ev::stat> only)
4317 Invokes C<ev_stat_stat>.
4319 =back
4321 =back
4323 Example: Define a class with two I/O and idle watchers, start the I/O
4324 watchers in the constructor.
4326 class myclass
4327 {
4328 ev::io io ; void io_cb (ev::io &w, int revents);
4329 ev::io io2 ; void io2_cb (ev::io &w, int revents);
4330 ev::idle idle; void idle_cb (ev::idle &w, int revents);
4332 myclass (int fd)
4333 {
4334 io .set <myclass, &myclass::io_cb > (this);
4335 io2 .set <myclass, &myclass::io2_cb > (this);
4336 idle.set <myclass, &myclass::idle_cb> (this);
4338 io.set (fd, ev::WRITE); // configure the watcher
4339 io.start (); // start it whenever convenient
4341 io2.start (fd, ev::READ); // set + start in one call
4342 }
4343 };
4348 Libev does not offer other language bindings itself, but bindings for a
4349 number of languages exist in the form of third-party packages. If you know
4350 any interesting language binding in addition to the ones listed here, drop
4351 me a note.
4353 =over 4
4355 =item Perl
4357 The EV module implements the full libev API and is actually used to test
4358 libev. EV is developed together with libev. Apart from the EV core module,
4359 there are additional modules that implement libev-compatible interfaces
4360 to C<libadns> (C<EV::ADNS>, but C<AnyEvent::DNS> is preferred nowadays),
4361 C<Net::SNMP> (C<Net::SNMP::EV>) and the C<libglib> event core (C<Glib::EV>
4362 and C<EV::Glib>).
4364 It can be found and installed via CPAN, its homepage is at
4365 L<>.
4367 =item Python
4369 Python bindings can be found at L<>. It
4370 seems to be quite complete and well-documented.
4372 =item Ruby
4374 Tony Arcieri has written a ruby extension that offers access to a subset
4375 of the libev API and adds file handle abstractions, asynchronous DNS and
4376 more on top of it. It can be found via gem servers. Its homepage is at
4377 L<>.
4379 Roger Pack reports that using the link order C<-lws2_32 -lmsvcrt-ruby-190>
4380 makes rev work even on mingw.
4382 =item Haskell
4384 A haskell binding to libev is available at
4385 L<>.
4387 =item D
4389 Leandro Lucarella has written a D language binding (F<ev.d>) for libev, to
4390 be found at L<>.
4392 =item Ocaml
4394 Erkki Seppala has written Ocaml bindings for libev, to be found at
4395 L<>.
4397 =item Lua
4399 Brian Maher has written a partial interface to libev for lua (at the
4400 time of this writing, only C<ev_io> and C<ev_timer>), to be found at
4401 L<>.
4403 =item Javascript
4405 Node.js (L<>) uses libev as the underlying event library.
4407 =item Others
4409 There are others, and I stopped counting.
4411 =back
4414 =head1 MACRO MAGIC
4416 Libev can be compiled with a variety of options, the most fundamental
4417 of which is C<EV_MULTIPLICITY>. This option determines whether (most)
4418 functions and callbacks have an initial C<struct ev_loop *> argument.
4420 To make it easier to write programs that cope with either variant, the
4421 following macros are defined:
4423 =over 4
4425 =item C<EV_A>, C<EV_A_>
4427 This provides the loop I<argument> for functions, if one is required ("ev
4428 loop argument"). The C<EV_A> form is used when this is the sole argument,
4429 C<EV_A_> is used when other arguments are following. Example:
4431 ev_unref (EV_A);
4432 ev_timer_add (EV_A_ watcher);
4433 ev_run (EV_A_ 0);
4435 It assumes the variable C<loop> of type C<struct ev_loop *> is in scope,
4436 which is often provided by the following macro.
4438 =item C<EV_P>, C<EV_P_>
4440 This provides the loop I<parameter> for functions, if one is required ("ev
4441 loop parameter"). The C<EV_P> form is used when this is the sole parameter,
4442 C<EV_P_> is used when other parameters are following. Example:
4444 // this is how ev_unref is being declared
4445 static void ev_unref (EV_P);
4447 // this is how you can declare your typical callback
4448 static void cb (EV_P_ ev_timer *w, int revents)
4450 It declares a parameter C<loop> of type C<struct ev_loop *>, quite
4451 suitable for use with C<EV_A>.
4453 =item C<EV_DEFAULT>, C<EV_DEFAULT_>
4455 Similar to the other two macros, this gives you the value of the default
4456 loop, if multiple loops are supported ("ev loop default"). The default loop
4457 will be initialised if it isn't already initialised.
4459 For non-multiplicity builds, these macros do nothing, so you always have
4460 to initialise the loop somewhere.
4464 Usage identical to C<EV_DEFAULT> and C<EV_DEFAULT_>, but requires that the
4465 default loop has been initialised (C<UC> == unchecked). Their behaviour
4466 is undefined when the default loop has not been initialised by a previous
4467 execution of C<EV_DEFAULT>, C<EV_DEFAULT_> or C<ev_default_init (...)>.
4469 It is often prudent to use C<EV_DEFAULT> when initialising the first
4470 watcher in a function but use C<EV_DEFAULT_UC> afterwards.
4472 =back
4474 Example: Declare and initialise a check watcher, utilising the above
4475 macros so it will work regardless of whether multiple loops are supported
4476 or not.
4478 static void
4479 check_cb (EV_P_ ev_timer *w, int revents)
4480 {
4481 ev_check_stop (EV_A_ w);
4482 }
4484 ev_check check;
4485 ev_check_init (&check, check_cb);
4486 ev_check_start (EV_DEFAULT_ &check);
4487 ev_run (EV_DEFAULT_ 0);
4489 =head1 EMBEDDING
4491 Libev can (and often is) directly embedded into host
4492 applications. Examples of applications that embed it include the Deliantra
4493 Game Server, the EV perl module, the GNU Virtual Private Ethernet (gvpe)
4494 and rxvt-unicode.
4496 The goal is to enable you to just copy the necessary files into your
4497 source directory without having to change even a single line in them, so
4498 you can easily upgrade by simply copying (or having a checked-out copy of
4499 libev somewhere in your source tree).
4501 =head2 FILESETS
4503 Depending on what features you need you need to include one or more sets of files
4504 in your application.
4506 =head3 CORE EVENT LOOP
4508 To include only the libev core (all the C<ev_*> functions), with manual
4509 configuration (no autoconf):
4511 #define EV_STANDALONE 1
4512 #include "ev.c"
4514 This will automatically include F<ev.h>, too, and should be done in a
4515 single C source file only to provide the function implementations. To use
4516 it, do the same for F<ev.h> in all files wishing to use this API (best
4517 done by writing a wrapper around F<ev.h> that you can include instead and
4518 where you can put other configuration options):
4520 #define EV_STANDALONE 1
4521 #include "ev.h"
4523 Both header files and implementation files can be compiled with a C++
4524 compiler (at least, that's a stated goal, and breakage will be treated
4525 as a bug).
4527 You need the following files in your source tree, or in a directory
4528 in your include path (e.g. in libev/ when using -Ilibev):
4530 ev.h
4531 ev.c
4532 ev_vars.h
4533 ev_wrap.h
4535 ev_win32.c required on win32 platforms only
4537 ev_select.c only when select backend is enabled
4538 ev_poll.c only when poll backend is enabled
4539 ev_epoll.c only when the epoll backend is enabled
4540 ev_linuxaio.c only when the linux aio backend is enabled
4541 ev_iouring.c only when the linux io_uring backend is enabled
4542 ev_kqueue.c only when the kqueue backend is enabled
4543 ev_port.c only when the solaris port backend is enabled
4545 F<ev.c> includes the backend files directly when enabled, so you only need
4546 to compile this single file.
4550 To include the libevent compatibility API, also include:
4552 #include "event.c"
4554 in the file including F<ev.c>, and:
4556 #include "event.h"
4558 in the files that want to use the libevent API. This also includes F<ev.h>.
4560 You need the following additional files for this:
4562 event.h
4563 event.c
4567 Instead of using C<EV_STANDALONE=1> and providing your configuration in
4568 whatever way you want, you can also C<m4_include([libev.m4])> in your
4569 F<> and leave C<EV_STANDALONE> undefined. F<ev.c> will then
4570 include F<config.h> and configure itself accordingly.
4572 For this of course you need the m4 file:
4574 libev.m4
4578 Libev can be configured via a variety of preprocessor symbols you have to
4579 define before including (or compiling) any of its files. The default in
4580 the absence of autoconf is documented for every option.
4582 Symbols marked with "(h)" do not change the ABI, and can have different
4583 values when compiling libev vs. including F<ev.h>, so it is permissible
4584 to redefine them before including F<ev.h> without breaking compatibility
4585 to a compiled library. All other symbols change the ABI, which means all
4586 users of libev and the libev code itself must be compiled with compatible
4587 settings.
4589 =over 4
4591 =item EV_COMPAT3 (h)
4593 Backwards compatibility is a major concern for libev. This is why this
4594 release of libev comes with wrappers for the functions and symbols that
4595 have been renamed between libev version 3 and 4.
4597 You can disable these wrappers (to test compatibility with future
4598 versions) by defining C<EV_COMPAT3> to C<0> when compiling your
4599 sources. This has the additional advantage that you can drop the C<struct>
4600 from C<struct ev_loop> declarations, as libev will provide an C<ev_loop>
4601 typedef in that case.
4603 In some future version, the default for C<EV_COMPAT3> will become C<0>,
4604 and in some even more future version the compatibility code will be
4605 removed completely.
4607 =item EV_STANDALONE (h)
4609 Must always be C<1> if you do not use autoconf configuration, which
4610 keeps libev from including F<config.h>, and it also defines dummy
4611 implementations for some libevent functions (such as logging, which is not
4612 supported). It will also not define any of the structs usually found in
4613 F<event.h> that are not directly supported by the libev core alone.
4615 In standalone mode, libev will still try to automatically deduce the
4616 configuration, but has to be more conservative.
4618 =item EV_USE_FLOOR
4620 If defined to be C<1>, libev will use the C<floor ()> function for its
4621 periodic reschedule calculations, otherwise libev will fall back on a
4622 portable (slower) implementation. If you enable this, you usually have to
4623 link against libm or something equivalent. Enabling this when the C<floor>
4624 function is not available will fail, so the safe default is to not enable
4625 this.
4629 If defined to be C<1>, libev will try to detect the availability of the
4630 monotonic clock option at both compile time and runtime. Otherwise no
4631 use of the monotonic clock option will be attempted. If you enable this,
4632 you usually have to link against librt or something similar. Enabling it
4633 when the functionality isn't available is safe, though, although you have
4634 to make sure you link against any libraries where the C<clock_gettime>
4635 function is hiding in (often F<-lrt>). See also C<EV_USE_CLOCK_SYSCALL>.
4637 =item EV_USE_REALTIME
4639 If defined to be C<1>, libev will try to detect the availability of the
4640 real-time clock option at compile time (and assume its availability
4641 at runtime if successful). Otherwise no use of the real-time clock
4642 option will be attempted. This effectively replaces C<gettimeofday>
4643 by C<clock_get (CLOCK_REALTIME, ...)> and will not normally affect
4644 correctness. See the note about libraries in the description of
4645 C<EV_USE_MONOTONIC>, though. Defaults to the opposite value of
4650 If defined to be C<1>, libev will try to use a direct syscall instead
4651 of calling the system-provided C<clock_gettime> function. This option
4652 exists because on GNU/Linux, C<clock_gettime> is in C<librt>, but C<librt>
4653 unconditionally pulls in C<libpthread>, slowing down single-threaded
4654 programs needlessly. Using a direct syscall is slightly slower (in
4655 theory), because no optimised vdso implementation can be used, but avoids
4656 the pthread dependency. Defaults to C<1> on GNU/Linux with glibc 2.x or
4657 higher, as it simplifies linking (no need for C<-lrt>).
4661 If defined to be C<1>, libev will assume that C<nanosleep ()> is available
4662 and will use it for delays. Otherwise it will use C<select ()>.
4664 =item EV_USE_EVENTFD
4666 If defined to be C<1>, then libev will assume that C<eventfd ()> is
4667 available and will probe for kernel support at runtime. This will improve
4668 C<ev_signal> and C<ev_async> performance and reduce resource consumption.
4669 If undefined, it will be enabled if the headers indicate GNU/Linux + Glibc
4670 2.7 or newer, otherwise disabled.
4672 =item EV_USE_SIGNALFD
4674 If defined to be C<1>, then libev will assume that C<signalfd ()> is
4675 available and will probe for kernel support at runtime. This enables
4676 the use of EVFLAG_SIGNALFD for faster and simpler signal handling. If
4677 undefined, it will be enabled if the headers indicate GNU/Linux + Glibc
4678 2.7 or newer, otherwise disabled.
4680 =item EV_USE_TIMERFD
4682 If defined to be C<1>, then libev will assume that C<timerfd ()> is
4683 available and will probe for kernel support at runtime. This allows
4684 libev to detect time jumps accurately. If undefined, it will be enabled
4685 if the headers indicate GNU/Linux + Glibc 2.8 or newer and define
4686 C<TFD_TIMER_CANCEL_ON_SET>, otherwise disabled.
4688 =item EV_USE_EVENTFD
4690 If defined to be C<1>, then libev will assume that C<eventfd ()> is
4691 available and will probe for kernel support at runtime. This will improve
4692 C<ev_signal> and C<ev_async> performance and reduce resource consumption.
4693 If undefined, it will be enabled if the headers indicate GNU/Linux + Glibc
4694 2.7 or newer, otherwise disabled.
4696 =item EV_USE_SELECT
4698 If undefined or defined to be C<1>, libev will compile in support for the
4699 C<select>(2) backend. No attempt at auto-detection will be done: if no
4700 other method takes over, select will be it. Otherwise the select backend
4701 will not be compiled in.
4705 If defined to C<1>, then the select backend will use the system C<fd_set>
4706 structure. This is useful if libev doesn't compile due to a missing
4707 C<NFDBITS> or C<fd_mask> definition or it mis-guesses the bitset layout
4708 on exotic systems. This usually limits the range of file descriptors to
4709 some low limit such as 1024 or might have other limitations (winsocket
4710 only allows 64 sockets). The C<FD_SETSIZE> macro, set before compilation,
4711 configures the maximum size of the C<fd_set>.
4715 When defined to C<1>, the select backend will assume that
4716 select/socket/connect etc. don't understand file descriptors but
4717 wants osf handles on win32 (this is the case when the select to
4718 be used is the winsock select). This means that it will call
4719 C<_get_osfhandle> on the fd to convert it to an OS handle. Otherwise,
4720 it is assumed that all these functions actually work on fds, even
4721 on win32. Should not be defined on non-win32 platforms.
4723 =item EV_FD_TO_WIN32_HANDLE(fd)
4725 If C<EV_SELECT_IS_WINSOCKET> is enabled, then libev needs a way to map
4726 file descriptors to socket handles. When not defining this symbol (the
4727 default), then libev will call C<_get_osfhandle>, which is usually
4728 correct. In some cases, programs use their own file descriptor management,
4729 in which case they can provide this function to map fds to socket handles.
4731 =item EV_WIN32_HANDLE_TO_FD(handle)
4733 If C<EV_SELECT_IS_WINSOCKET> then libev maps handles to file descriptors
4734 using the standard C<_open_osfhandle> function. For programs implementing
4735 their own fd to handle mapping, overwriting this function makes it easier
4736 to do so. This can be done by defining this macro to an appropriate value.
4738 =item EV_WIN32_CLOSE_FD(fd)
4740 If programs implement their own fd to handle mapping on win32, then this
4741 macro can be used to override the C<close> function, useful to unregister
4742 file descriptors again. Note that the replacement function has to close
4743 the underlying OS handle.
4747 If defined to be C<1>, libev will use C<WSASocket> to create its internal
4748 communication socket, which works better in some environments. Otherwise,
4749 the normal C<socket> function will be used, which works better in other
4750 environments.
4752 =item EV_USE_POLL
4754 If defined to be C<1>, libev will compile in support for the C<poll>(2)
4755 backend. Otherwise it will be enabled on non-win32 platforms. It
4756 takes precedence over select.
4758 =item EV_USE_EPOLL
4760 If defined to be C<1>, libev will compile in support for the Linux
4761 C<epoll>(7) backend. Its availability will be detected at runtime,
4762 otherwise another method will be used as fallback. This is the preferred
4763 backend for GNU/Linux systems. If undefined, it will be enabled if the
4764 headers indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.
4766 =item EV_USE_LINUXAIO
4768 If defined to be C<1>, libev will compile in support for the Linux aio
4769 backend (C<EV_USE_EPOLL> must also be enabled). If undefined, it will be
4770 enabled on linux, otherwise disabled.
4772 =item EV_USE_IOURING
4774 If defined to be C<1>, libev will compile in support for the Linux
4775 io_uring backend (C<EV_USE_EPOLL> must also be enabled). Due to it's
4776 current limitations it has to be requested explicitly. If undefined, it
4777 will be enabled on linux, otherwise disabled.
4779 =item EV_USE_KQUEUE
4781 If defined to be C<1>, libev will compile in support for the BSD style
4782 C<kqueue>(2) backend. Its actual availability will be detected at runtime,
4783 otherwise another method will be used as fallback. This is the preferred
4784 backend for BSD and BSD-like systems, although on most BSDs kqueue only
4785 supports some types of fds correctly (the only platform we found that
4786 supports ptys for example was NetBSD), so kqueue might be compiled in, but
4787 not be used unless explicitly requested. The best way to use it is to find
4788 out whether kqueue supports your type of fd properly and use an embedded
4789 kqueue loop.
4791 =item EV_USE_PORT
4793 If defined to be C<1>, libev will compile in support for the Solaris
4794 10 port style backend. Its availability will be detected at runtime,
4795 otherwise another method will be used as fallback. This is the preferred
4796 backend for Solaris 10 systems.
4798 =item EV_USE_DEVPOLL
4800 Reserved for future expansion, works like the USE symbols above.
4802 =item EV_USE_INOTIFY
4804 If defined to be C<1>, libev will compile in support for the Linux inotify
4805 interface to speed up C<ev_stat> watchers. Its actual availability will
4806 be detected at runtime. If undefined, it will be enabled if the headers
4807 indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.
4809 =item EV_NO_SMP
4811 If defined to be C<1>, libev will assume that memory is always coherent
4812 between threads, that is, threads can be used, but threads never run on
4813 different cpus (or different cpu cores). This reduces dependencies
4814 and makes libev faster.
4816 =item EV_NO_THREADS
4818 If defined to be C<1>, libev will assume that it will never be called from
4819 different threads (that includes signal handlers), which is a stronger
4820 assumption than C<EV_NO_SMP>, above. This reduces dependencies and makes
4821 libev faster.
4823 =item EV_ATOMIC_T
4825 Libev requires an integer type (suitable for storing C<0> or C<1>) whose
4826 access is atomic with respect to other threads or signal contexts. No
4827 such type is easily found in the C language, so you can provide your own
4828 type that you know is safe for your purposes. It is used both for signal
4829 handler "locking" as well as for signal and thread safety in C<ev_async>
4830 watchers.
4832 In the absence of this define, libev will use C<sig_atomic_t volatile>
4833 (from F<signal.h>), which is usually good enough on most platforms.
4835 =item EV_H (h)
4837 The name of the F<ev.h> header file used to include it. The default if
4838 undefined is C<"ev.h"> in F<event.h>, F<ev.c> and F<ev++.h>. This can be
4839 used to virtually rename the F<ev.h> header file in case of conflicts.
4841 =item EV_CONFIG_H (h)
4843 If C<EV_STANDALONE> isn't C<1>, this variable can be used to override
4844 F<ev.c>'s idea of where to find the F<config.h> file, similarly to
4845 C<EV_H>, above.
4847 =item EV_EVENT_H (h)
4849 Similarly to C<EV_H>, this macro can be used to override F<event.c>'s idea
4850 of how the F<event.h> header can be found, the default is C<"event.h">.
4852 =item EV_PROTOTYPES (h)
4854 If defined to be C<0>, then F<ev.h> will not define any function
4855 prototypes, but still define all the structs and other symbols. This is
4856 occasionally useful if you want to provide your own wrapper functions
4857 around libev functions.
4861 If undefined or defined to C<1>, then all event-loop-specific functions
4862 will have the C<struct ev_loop *> as first argument, and you can create
4863 additional independent event loops. Otherwise there will be no support
4864 for multiple event loops and there is no first event loop pointer
4865 argument. Instead, all functions act on the single default loop.
4867 Note that C<EV_DEFAULT> and C<EV_DEFAULT_> will no longer provide a
4868 default loop when multiplicity is switched off - you always have to
4869 initialise the loop manually in this case.
4871 =item EV_MINPRI
4873 =item EV_MAXPRI
4875 The range of allowed priorities. C<EV_MINPRI> must be smaller or equal to
4876 C<EV_MAXPRI>, but otherwise there are no non-obvious limitations. You can
4877 provide for more priorities by overriding those symbols (usually defined
4878 to be C<-2> and C<2>, respectively).
4880 When doing priority-based operations, libev usually has to linearly search
4881 all the priorities, so having many of them (hundreds) uses a lot of space
4882 and time, so using the defaults of five priorities (-2 .. +2) is usually
4883 fine.
4885 If your embedding application does not need any priorities, defining these
4886 both to C<0> will save some memory and CPU.
4892 If undefined or defined to be C<1> (and the platform supports it), then
4893 the respective watcher type is supported. If defined to be C<0>, then it
4894 is not. Disabling watcher types mainly saves code size.
4896 =item EV_FEATURES
4898 If you need to shave off some kilobytes of code at the expense of some
4899 speed (but with the full API), you can define this symbol to request
4900 certain subsets of functionality. The default is to enable all features
4901 that can be enabled on the platform.
4903 A typical way to use this symbol is to define it to C<0> (or to a bitset
4904 with some broad features you want) and then selectively re-enable
4905 additional parts you want, for example if you want everything minimal,
4906 but multiple event loop support, async and child watchers and the poll
4907 backend, use this:
4909 #define EV_FEATURES 0
4910 #define EV_MULTIPLICITY 1
4911 #define EV_USE_POLL 1
4912 #define EV_CHILD_ENABLE 1
4913 #define EV_ASYNC_ENABLE 1
4915 The actual value is a bitset, it can be a combination of the following
4916 values (by default, all of these are enabled):
4918 =over 4
4920 =item C<1> - faster/larger code
4922 Use larger code to speed up some operations.
4924 Currently this is used to override some inlining decisions (enlarging the
4925 code size by roughly 30% on amd64).
4927 When optimising for size, use of compiler flags such as C<-Os> with
4928 gcc is recommended, as well as C<-DNDEBUG>, as libev contains a number of
4929 assertions.
4931 The default is off when C<__OPTIMIZE_SIZE__> is defined by your compiler
4932 (e.g. gcc with C<-Os>).
4934 =item C<2> - faster/larger data structures
4936 Replaces the small 2-heap for timer management by a faster 4-heap, larger
4937 hash table sizes and so on. This will usually further increase code size
4938 and can additionally have an effect on the size of data structures at
4939 runtime.
4941 The default is off when C<__OPTIMIZE_SIZE__> is defined by your compiler
4942 (e.g. gcc with C<-Os>).
4944 =item C<4> - full API configuration
4946 This enables priorities (sets C<EV_MAXPRI>=2 and C<EV_MINPRI>=-2), and
4947 enables multiplicity (C<EV_MULTIPLICITY>=1).
4949 =item C<8> - full API
4951 This enables a lot of the "lesser used" API functions. See C<ev.h> for
4952 details on which parts of the API are still available without this
4953 feature, and do not complain if this subset changes over time.
4955 =item C<16> - enable all optional watcher types
4957 Enables all optional watcher types. If you want to selectively enable