ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/libev/ev.pod

Comparing libev/ev.pod (file contents):
Revision 1.320 by root, Fri Oct 22 10:48:54 2010 UTC vs.
Revision 1.448 by root, Sun Jun 23 02:02:24 2019 UTC

		1	=encoding utf-8
		2
1	=head1 NAME	3	=head1 NAME
2		4
3	libev - a high performance full-featured event loop written in C	5	libev - a high performance full-featured event loop written in C
4		6
5	=head1 SYNOPSIS	7	=head1 SYNOPSIS
…		…
43		45
44	int	46	int
45	main (void)	47	main (void)
46	{	48	{
47	// use the default event loop unless you have special needs	49	// use the default event loop unless you have special needs
48	struct ev_loop *loop = ev_default_loop (0);	50	struct ev_loop *loop = EV_DEFAULT;
49		51
50	// initialise an io watcher, then start it	52	// initialise an io watcher, then start it
51	// this one will watch for stdin to become readable	53	// this one will watch for stdin to become readable
52	ev_io_init (&stdin_watcher, stdin_cb, /*STDIN_FILENO*/ 0, EV_READ);	54	ev_io_init (&stdin_watcher, stdin_cb, /*STDIN_FILENO*/ 0, EV_READ);
53	ev_io_start (loop, &stdin_watcher);	55	ev_io_start (loop, &stdin_watcher);
…		…
58	ev_timer_start (loop, &timeout_watcher);	60	ev_timer_start (loop, &timeout_watcher);
59		61
60	// now wait for events to arrive	62	// now wait for events to arrive
61	ev_run (loop, 0);	63	ev_run (loop, 0);
62		64
63	// unloop was called, so exit	65	// break was called, so exit
64	return 0;	66	return 0;
65	}	67	}
66		68
67	=head1 ABOUT THIS DOCUMENT	69	=head1 ABOUT THIS DOCUMENT
68		70
…		…
78	with libev.	80	with libev.
79		81
80	Familiarity with event based programming techniques in general is assumed	82	Familiarity with event based programming techniques in general is assumed
81	throughout this document.	83	throughout this document.
82		84
		85	=head1 WHAT TO READ WHEN IN A HURRY
		86
		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>.
		92
83	=head1 ABOUT LIBEV	93	=head1 ABOUT LIBEV
84		94
85	Libev is an event loop: you register interest in certain events (such as a	95	Libev is an event loop: you register interest in certain events (such as a
86	file descriptor being readable or a timeout occurring), and it will manage	96	file descriptor being readable or a timeout occurring), and it will manage
87	these event sources and provide your program with events.	97	these event sources and provide your program with events.
…		…
95	details of the event, and then hand it over to libev by I<starting> the	105	details of the event, and then hand it over to libev by I<starting> the
96	watcher.	106	watcher.
97		107
98	=head2 FEATURES	108	=head2 FEATURES
99		109
100	Libev supports C<select>, C<poll>, the Linux-specific C<epoll>, the	110	Libev supports C<select>, C<poll>, the Linux-specific aio and C<epoll>
101	BSD-specific C<kqueue> and the Solaris-specific event port mechanisms	111	interfaces, the BSD-specific C<kqueue> and the Solaris-specific event port
102	for file descriptor events (C<ev_io>), the Linux C<inotify> interface	112	mechanisms for file descriptor events (C<ev_io>), the Linux C<inotify>
103	(for C<ev_stat>), Linux eventfd/signalfd (for faster and cleaner	113	interface (for C<ev_stat>), Linux eventfd/signalfd (for faster and cleaner
104	inter-thread wakeup (C<ev_async>)/signal handling (C<ev_signal>)) relative	114	inter-thread wakeup (C<ev_async>)/signal handling (C<ev_signal>)) relative
105	timers (C<ev_timer>), absolute timers with customised rescheduling	115	timers (C<ev_timer>), absolute timers with customised rescheduling
106	(C<ev_periodic>), synchronous signals (C<ev_signal>), process status	116	(C<ev_periodic>), synchronous signals (C<ev_signal>), process status
107	change events (C<ev_child>), and event watchers dealing with the event	117	change events (C<ev_child>), and event watchers dealing with the event
108	loop mechanism itself (C<ev_idle>, C<ev_embed>, C<ev_prepare> and	118	loop mechanism itself (C<ev_idle>, C<ev_embed>, C<ev_prepare> and
…		…
165		175
166	=item ev_tstamp ev_time ()	176	=item ev_tstamp ev_time ()
167		177
168	Returns the current time as libev would use it. Please note that the	178	Returns the current time as libev would use it. Please note that the
169	C<ev_now> function is usually faster and also often returns the timestamp	179	C<ev_now> function is usually faster and also often returns the timestamp
170	you actually want to know. Also interetsing is the combination of	180	you actually want to know. Also interesting is the combination of
171	C<ev_update_now> and C<ev_now>.	181	C<ev_now_update> and C<ev_now>.
172		182
173	=item ev_sleep (ev_tstamp interval)	183	=item ev_sleep (ev_tstamp interval)
174		184
175	Sleep for the given interval: The current thread will be blocked until	185	Sleep for the given interval: The current thread will be blocked
176	either it is interrupted or the given time interval has passed. Basically	186	until either it is interrupted or the given time interval has
		187	passed (approximately - it might return a bit earlier even if not
		188	interrupted). Returns immediately if C<< interval <= 0 >>.
		189
177	this is a sub-second-resolution C<sleep ()>.	190	Basically this is a sub-second-resolution C<sleep ()>.
		191
		192	The range of the C<interval> is limited - libev only guarantees to work
		193	with sleep times of up to one day (C<< interval <= 86400 >>).
178		194
179	=item int ev_version_major ()	195	=item int ev_version_major ()
180		196
181	=item int ev_version_minor ()	197	=item int ev_version_minor ()
182		198
…		…
233	the current system, you would need to look at C<ev_embeddable_backends ()	249	the current system, you would need to look at C<ev_embeddable_backends ()
234	& ev_supported_backends ()>, likewise for recommended ones.	250	& ev_supported_backends ()>, likewise for recommended ones.
235		251
236	See the description of C<ev_embed> watchers for more info.	252	See the description of C<ev_embed> watchers for more info.
237		253
238	=item ev_set_allocator (void *(*cb)(void *ptr, long size)) [NOT REENTRANT]	254	=item ev_set_allocator (void *(*cb)(void *ptr, long size) throw ())
239		255
240	Sets the allocation function to use (the prototype is similar - the	256	Sets the allocation function to use (the prototype is similar - the
241	semantics are identical to the C<realloc> C89/SuS/POSIX function). It is	257	semantics are identical to the C<realloc> C89/SuS/POSIX function). It is
242	used to allocate and free memory (no surprises here). If it returns zero	258	used to allocate and free memory (no surprises here). If it returns zero
243	when memory needs to be allocated (C<size != 0>), the library might abort	259	when memory needs to be allocated (C<size != 0>), the library might abort
…		…
249		265
250	You could override this function in high-availability programs to, say,	266	You could override this function in high-availability programs to, say,
251	free some memory if it cannot allocate memory, to use a special allocator,	267	free some memory if it cannot allocate memory, to use a special allocator,
252	or even to sleep a while and retry until some memory is available.	268	or even to sleep a while and retry until some memory is available.
253		269
		270	Example: The following is the C<realloc> function that libev itself uses
		271	which should work with C<realloc> and C<free> functions of all kinds and
		272	is probably a good basis for your own implementation.
		273
		274	static void *
		275	ev_realloc_emul (void *ptr, long size) EV_NOEXCEPT
		276	{
		277	if (size)
		278	return realloc (ptr, size);
		279
		280	free (ptr);
		281	return 0;
		282	}
		283
254	Example: Replace the libev allocator with one that waits a bit and then	284	Example: Replace the libev allocator with one that waits a bit and then
255	retries (example requires a standards-compliant C<realloc>).	285	retries.
256		286
257	static void *	287	static void *
258	persistent_realloc (void *ptr, size_t size)	288	persistent_realloc (void *ptr, size_t size)
259	{	289	{
		290	if (!size)
		291	{
		292	free (ptr);
		293	return 0;
		294	}
		295
260	for (;;)	296	for (;;)
261	{	297	{
262	void *newptr = realloc (ptr, size);	298	void *newptr = realloc (ptr, size);
263		299
264	if (newptr)	300	if (newptr)
…		…
269	}	305	}
270		306
271	...	307	...
272	ev_set_allocator (persistent_realloc);	308	ev_set_allocator (persistent_realloc);
273		309
274	=item ev_set_syserr_cb (void (*cb)(const char *msg)); [NOT REENTRANT]	310	=item ev_set_syserr_cb (void (*cb)(const char *msg) throw ())
275		311
276	Set the callback function to call on a retryable system call error (such	312	Set the callback function to call on a retryable system call error (such
277	as failed select, poll, epoll_wait). The message is a printable string	313	as failed select, poll, epoll_wait). The message is a printable string
278	indicating the system call or subsystem causing the problem. If this	314	indicating the system call or subsystem causing the problem. If this
279	callback is set, then libev will expect it to remedy the situation, no	315	callback is set, then libev will expect it to remedy the situation, no
…		…
291	}	327	}
292		328
293	...	329	...
294	ev_set_syserr_cb (fatal_error);	330	ev_set_syserr_cb (fatal_error);
295		331
		332	=item ev_feed_signal (int signum)
		333
		334	This function can be used to "simulate" a signal receive. It is completely
		335	safe to call this function at any time, from any context, including signal
		336	handlers or random threads.
		337
		338	Its main use is to customise signal handling in your process, especially
		339	in the presence of threads. For example, you could block signals
		340	by default in all threads (and specifying C<EVFLAG_NOSIGMASK> when
		341	creating any loops), and in one thread, use C<sigwait> or any other
		342	mechanism to wait for signals, then "deliver" them to libev by calling
		343	C<ev_feed_signal>.
		344
296	=back	345	=back
297		346
298	=head1 FUNCTIONS CONTROLLING THE EVENT LOOP	347	=head1 FUNCTIONS CONTROLLING EVENT LOOPS
299		348
300	An event loop is described by a C<struct ev_loop *> (the C<struct> is	349	An event loop is described by a C<struct ev_loop *> (the C<struct> is
301	I<not> optional in this case unless libev 3 compatibility is disabled, as	350	I<not> optional in this case unless libev 3 compatibility is disabled, as
302	libev 3 had an C<ev_loop> function colliding with the struct name).	351	libev 3 had an C<ev_loop> function colliding with the struct name).
303		352
304	The library knows two types of such loops, the I<default> loop, which	353	The library knows two types of such loops, the I<default> loop, which
305	supports signals and child events, and dynamically created event loops	354	supports child process events, and dynamically created event loops which
306	which do not.	355	do not.
307		356
308	=over 4	357	=over 4
309		358
310	=item struct ev_loop *ev_default_loop (unsigned int flags)	359	=item struct ev_loop *ev_default_loop (unsigned int flags)
311		360
312	This will initialise the default event loop if it hasn't been initialised	361	This returns the "default" event loop object, which is what you should
313	yet and return it. If the default loop could not be initialised, returns	362	normally use when you just need "the event loop". Event loop objects and
314	false. If it already was initialised it simply returns it (and ignores the	363	the C<flags> parameter are described in more detail in the entry for
315	flags. If that is troubling you, check C<ev_backend ()> afterwards).	364	C<ev_loop_new>.
		365
		366	If the default loop is already initialised then this function simply
		367	returns it (and ignores the flags. If that is troubling you, check
		368	C<ev_backend ()> afterwards). Otherwise it will create it with the given
		369	flags, which should almost always be C<0>, unless the caller is also the
		370	one calling C<ev_run> or otherwise qualifies as "the main program".
316		371
317	If you don't know what event loop to use, use the one returned from this	372	If you don't know what event loop to use, use the one returned from this
318	function.	373	function (or via the C<EV_DEFAULT> macro).
319		374
320	Note that this function is I<not> thread-safe, so if you want to use it	375	Note that this function is I<not> thread-safe, so if you want to use it
321	from multiple threads, you have to lock (note also that this is unlikely,	376	from multiple threads, you have to employ some kind of mutex (note also
322	as loops cannot be shared easily between threads anyway).	377	that this case is unlikely, as loops cannot be shared easily between
		378	threads anyway).
323		379
324	The default loop is the only loop that can handle C<ev_signal> and	380	The default loop is the only loop that can handle C<ev_child> watchers,
325	C<ev_child> watchers, and to do this, it always registers a handler	381	and to do this, it always registers a handler for C<SIGCHLD>. If this is
326	for C<SIGCHLD>. If this is a problem for your application you can either	382	a problem for your application you can either create a dynamic loop with
327	create a dynamic loop with C<ev_loop_new> that doesn't do that, or you	383	C<ev_loop_new> which doesn't do that, or you can simply overwrite the
328	can simply overwrite the C<SIGCHLD> signal handler I<after> calling	384	C<SIGCHLD> signal handler I<after> calling C<ev_default_init>.
329	C<ev_default_init>.	385
		386	Example: This is the most typical usage.
		387
		388	if (!ev_default_loop (0))
		389	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
		390
		391	Example: Restrict libev to the select and poll backends, and do not allow
		392	environment settings to be taken into account:
		393
		394	ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);
		395
		396	=item struct ev_loop *ev_loop_new (unsigned int flags)
		397
		398	This will create and initialise a new event loop object. If the loop
		399	could not be initialised, returns false.
		400
		401	This function is thread-safe, and one common way to use libev with
		402	threads is indeed to create one loop per thread, and using the default
		403	loop in the "main" or "initial" thread.
330		404
331	The flags argument can be used to specify special behaviour or specific	405	The flags argument can be used to specify special behaviour or specific
332	backends to use, and is usually specified as C<0> (or C<EVFLAG_AUTO>).	406	backends to use, and is usually specified as C<0> (or C<EVFLAG_AUTO>).
333		407
334	The following flags are supported:	408	The following flags are supported:
…		…
344		418
345	If this flag bit is or'ed into the flag value (or the program runs setuid	419	If this flag bit is or'ed into the flag value (or the program runs setuid
346	or setgid) then libev will I<not> look at the environment variable	420	or setgid) then libev will I<not> look at the environment variable
347	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will	421	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will
348	override the flags completely if it is found in the environment. This is	422	override the flags completely if it is found in the environment. This is
349	useful to try out specific backends to test their performance, or to work	423	useful to try out specific backends to test their performance, to work
350	around bugs.	424	around bugs, or to make libev threadsafe (accessing environment variables
		425	cannot be done in a threadsafe way, but usually it works if no other
		426	thread modifies them).
351		427
352	=item C<EVFLAG_FORKCHECK>	428	=item C<EVFLAG_FORKCHECK>
353		429
354	Instead of calling C<ev_loop_fork> manually after a fork, you can also	430	Instead of calling C<ev_loop_fork> manually after a fork, you can also
355	make libev check for a fork in each iteration by enabling this flag.	431	make libev check for a fork in each iteration by enabling this flag.
356		432
357	This works by calling C<getpid ()> on every iteration of the loop,	433	This works by calling C<getpid ()> on every iteration of the loop,
358	and thus this might slow down your event loop if you do a lot of loop	434	and thus this might slow down your event loop if you do a lot of loop
359	iterations and little real work, but is usually not noticeable (on my	435	iterations and little real work, but is usually not noticeable (on my
360	GNU/Linux system for example, C<getpid> is actually a simple 5-insn sequence	436	GNU/Linux system for example, C<getpid> is actually a simple 5-insn
361	without a system call and thus I<very> fast, but my GNU/Linux system also has	437	sequence without a system call and thus I<very> fast, but my GNU/Linux
362	C<pthread_atfork> which is even faster).	438	system also has C<pthread_atfork> which is even faster). (Update: glibc
		439	versions 2.25 apparently removed the C<getpid> optimisation again).
363		440
364	The big advantage of this flag is that you can forget about fork (and	441	The big advantage of this flag is that you can forget about fork (and
365	forget about forgetting to tell libev about forking) when you use this	442	forget about forgetting to tell libev about forking, although you still
366	flag.	443	have to ignore C<SIGPIPE>) when you use this flag.
367		444
368	This flag setting cannot be overridden or specified in the C<LIBEV_FLAGS>	445	This flag setting cannot be overridden or specified in the C<LIBEV_FLAGS>
369	environment variable.	446	environment variable.
370		447
371	=item C<EVFLAG_NOINOTIFY>	448	=item C<EVFLAG_NOINOTIFY>
372		449
373	When this flag is specified, then libev will not attempt to use the	450	When this flag is specified, then libev will not attempt to use the
374	I<inotify> API for it's C<ev_stat> watchers. Apart from debugging and	451	I<inotify> API for its C<ev_stat> watchers. Apart from debugging and
375	testing, this flag can be useful to conserve inotify file descriptors, as	452	testing, this flag can be useful to conserve inotify file descriptors, as
376	otherwise each loop using C<ev_stat> watchers consumes one inotify handle.	453	otherwise each loop using C<ev_stat> watchers consumes one inotify handle.
377		454
378	=item C<EVFLAG_SIGNALFD>	455	=item C<EVFLAG_SIGNALFD>
379		456
380	When this flag is specified, then libev will attempt to use the	457	When this flag is specified, then libev will attempt to use the
381	I<signalfd> API for it's C<ev_signal> (and C<ev_child>) watchers. This API	458	I<signalfd> API for its C<ev_signal> (and C<ev_child>) watchers. This API
382	delivers signals synchronously, which makes it both faster and might make	459	delivers signals synchronously, which makes it both faster and might make
383	it possible to get the queued signal data. It can also simplify signal	460	it possible to get the queued signal data. It can also simplify signal
384	handling with threads, as long as you properly block signals in your	461	handling with threads, as long as you properly block signals in your
385	threads that are not interested in handling them.	462	threads that are not interested in handling them.
386		463
387	Signalfd will not be used by default as this changes your signal mask, and	464	Signalfd will not be used by default as this changes your signal mask, and
388	there are a lot of shoddy libraries and programs (glib's threadpool for	465	there are a lot of shoddy libraries and programs (glib's threadpool for
389	example) that can't properly initialise their signal masks.	466	example) that can't properly initialise their signal masks.
		467
		468	=item C<EVFLAG_NOSIGMASK>
		469
		470	When this flag is specified, then libev will avoid to modify the signal
		471	mask. Specifically, this means you have to make sure signals are unblocked
		472	when you want to receive them.
		473
		474	This behaviour is useful when you want to do your own signal handling, or
		475	want to handle signals only in specific threads and want to avoid libev
		476	unblocking the signals.
		477
		478	It's also required by POSIX in a threaded program, as libev calls
		479	C<sigprocmask>, whose behaviour is officially unspecified.
		480
		481	This flag's behaviour will become the default in future versions of libev.
390		482
391	=item C<EVBACKEND_SELECT> (value 1, portable select backend)	483	=item C<EVBACKEND_SELECT> (value 1, portable select backend)
392		484
393	This is your standard select(2) backend. Not I<completely> standard, as	485	This is your standard select(2) backend. Not I<completely> standard, as
394	libev tries to roll its own fd_set with no limits on the number of fds,	486	libev tries to roll its own fd_set with no limits on the number of fds,
…		…
422	=item C<EVBACKEND_EPOLL> (value 4, Linux)	514	=item C<EVBACKEND_EPOLL> (value 4, Linux)
423		515
424	Use the linux-specific epoll(7) interface (for both pre- and post-2.6.9	516	Use the linux-specific epoll(7) interface (for both pre- and post-2.6.9
425	kernels).	517	kernels).
426		518
427	For few fds, this backend is a bit little slower than poll and select,	519	For few fds, this backend is a bit little slower than poll and select, but
428	but it scales phenomenally better. While poll and select usually scale	520	it scales phenomenally better. While poll and select usually scale like
429	like O(total_fds) where n is the total number of fds (or the highest fd),	521	O(total_fds) where total_fds is the total number of fds (or the highest
430	epoll scales either O(1) or O(active_fds).	522	fd), epoll scales either O(1) or O(active_fds).
431		523
432	The epoll mechanism deserves honorable mention as the most misdesigned	524	The epoll mechanism deserves honorable mention as the most misdesigned
433	of the more advanced event mechanisms: mere annoyances include silently	525	of the more advanced event mechanisms: mere annoyances include silently
434	dropping file descriptors, requiring a system call per change per file	526	dropping file descriptors, requiring a system call per change per file
435	descriptor (and unnecessary guessing of parameters), problems with dup and	527	descriptor (and unnecessary guessing of parameters), problems with dup,
		528	returning before the timeout value, resulting in additional iterations
		529	(and only giving 5ms accuracy while select on the same platform gives
436	so on. The biggest issue is fork races, however - if a program forks then	530	0.1ms) and so on. The biggest issue is fork races, however - if a program
437	I<both> parent and child process have to recreate the epoll set, which can	531	forks then I<both> parent and child process have to recreate the epoll
438	take considerable time (one syscall per file descriptor) and is of course	532	set, which can take considerable time (one syscall per file descriptor)
439	hard to detect.	533	and is of course hard to detect.
440		534
441	Epoll is also notoriously buggy - embedding epoll fds I<should> work, but	535	Epoll is also notoriously buggy - embedding epoll fds I<should> work,
442	of course I<doesn't>, and epoll just loves to report events for totally	536	but of course I<doesn't>, and epoll just loves to report events for
443	I<different> file descriptors (even already closed ones, so one cannot	537	totally I<different> file descriptors (even already closed ones, so
444	even remove them from the set) than registered in the set (especially	538	one cannot even remove them from the set) than registered in the set
445	on SMP systems). Libev tries to counter these spurious notifications by	539	(especially on SMP systems). Libev tries to counter these spurious
446	employing an additional generation counter and comparing that against the	540	notifications by employing an additional generation counter and comparing
447	events to filter out spurious ones, recreating the set when required. Last	541	that against the events to filter out spurious ones, recreating the set
		542	when required. Epoll also erroneously rounds down timeouts, but gives you
		543	no way to know when and by how much, so sometimes you have to busy-wait
		544	because epoll returns immediately despite a nonzero timeout. And last
448	not least, it also refuses to work with some file descriptors which work	545	not least, it also refuses to work with some file descriptors which work
449	perfectly fine with C<select> (files, many character devices...).	546	perfectly fine with C<select> (files, many character devices...).
		547
		548	Epoll is truly the train wreck among event poll mechanisms, a frankenpoll,
		549	cobbled together in a hurry, no thought to design or interaction with
		550	others. Oh, the pain, will it ever stop...
450		551
451	While stopping, setting and starting an I/O watcher in the same iteration	552	While stopping, setting and starting an I/O watcher in the same iteration
452	will result in some caching, there is still a system call per such	553	will result in some caching, there is still a system call per such
453	incident (because the same I<file descriptor> could point to a different	554	incident (because the same I<file descriptor> could point to a different
454	I<file description> now), so its best to avoid that. Also, C<dup ()>'ed	555	I<file description> now), so its best to avoid that. Also, C<dup ()>'ed
…		…
466	All this means that, in practice, C<EVBACKEND_SELECT> can be as fast or	567	All this means that, in practice, C<EVBACKEND_SELECT> can be as fast or
467	faster than epoll for maybe up to a hundred file descriptors, depending on	568	faster than epoll for maybe up to a hundred file descriptors, depending on
468	the usage. So sad.	569	the usage. So sad.
469		570
470	While nominally embeddable in other event loops, this feature is broken in	571	While nominally embeddable in other event loops, this feature is broken in
471	all kernel versions tested so far.	572	a lot of kernel revisions, but probably(!) works in current versions.
		573
		574	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
		575	C<EVBACKEND_POLL>.
		576
		577	=item C<EVBACKEND_LINUXAIO> (value 64, Linux)
		578
		579	Use the linux-specific linux aio (I<not> C<< aio(7) >> but C<<
		580	io_submit(2) >>) event interface available in post-4.18 kernels.
		581
		582	If this backend works for you (as of this writing, it was very
		583	experimental), it is the best event interface available on linux and might
		584	be well worth enabling it - if it isn't available in your kernel this will
		585	be detected and this backend will be skipped.
		586
		587	This backend can batch oneshot requests and supports a user-space ring
		588	buffer to receive events. It also doesn't suffer from most of the design
		589	problems of epoll (such as not being able to remove event sources from
		590	the epoll set), and generally sounds too good to be true. Because, this
		591	being the linux kernel, of course it suffers from a whole new set of
		592	limitations.
		593
		594	For one, it is not easily embeddable (but probably could be done using
		595	an event fd at some extra overhead). It also is subject to a system wide
		596	limit that can be configured in F</proc/sys/fs/aio-max-nr> - each loop
		597	currently requires C<61> of this number. If no aio requests are left, this
		598	backend will be skipped during initialisation.
		599
		600	Most problematic in practise, however, is that not all file descriptors
		601	work with it. For example, in linux 5.1, tcp sockets, pipes, event fds,
		602	files, F</dev/null> and a few others are supported, but ttys do not work
		603	(probably because of a bug), so this is not (yet?) a generic event polling
		604	interface.
		605
		606	To work around this latter problem, the current version of libev uses
		607	epoll as a fallback for file deescriptor types that do not work. Epoll
		608	is used in, kind of, slow mode that hopefully avoids most of its design
		609	problems.
472		610
473	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as	611	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
474	C<EVBACKEND_POLL>.	612	C<EVBACKEND_POLL>.
475		613
476	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)	614	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)
…		…
491		629
492	It scales in the same way as the epoll backend, but the interface to the	630	It scales in the same way as the epoll backend, but the interface to the
493	kernel is more efficient (which says nothing about its actual speed, of	631	kernel is more efficient (which says nothing about its actual speed, of
494	course). While stopping, setting and starting an I/O watcher does never	632	course). While stopping, setting and starting an I/O watcher does never
495	cause an extra system call as with C<EVBACKEND_EPOLL>, it still adds up to	633	cause an extra system call as with C<EVBACKEND_EPOLL>, it still adds up to
496	two event changes per incident. Support for C<fork ()> is very bad (but	634	two event changes per incident. Support for C<fork ()> is very bad (you
497	sane, unlike epoll) and it drops fds silently in similarly hard-to-detect	635	might have to leak fd's on fork, but it's more sane than epoll) and it
498	cases	636	drops fds silently in similarly hard-to-detect cases.
499		637
500	This backend usually performs well under most conditions.	638	This backend usually performs well under most conditions.
501		639
502	While nominally embeddable in other event loops, this doesn't work	640	While nominally embeddable in other event loops, this doesn't work
503	everywhere, so you might need to test for this. And since it is broken	641	everywhere, so you might need to test for this. And since it is broken
…		…
520	=item C<EVBACKEND_PORT> (value 32, Solaris 10)	658	=item C<EVBACKEND_PORT> (value 32, Solaris 10)
521		659
522	This uses the Solaris 10 event port mechanism. As with everything on Solaris,	660	This uses the Solaris 10 event port mechanism. As with everything on Solaris,
523	it's really slow, but it still scales very well (O(active_fds)).	661	it's really slow, but it still scales very well (O(active_fds)).
524		662
525	Please note that Solaris event ports can deliver a lot of spurious
526	notifications, so you need to use non-blocking I/O or other means to avoid
527	blocking when no data (or space) is available.
528
529	While this backend scales well, it requires one system call per active	663	While this backend scales well, it requires one system call per active
530	file descriptor per loop iteration. For small and medium numbers of file	664	file descriptor per loop iteration. For small and medium numbers of file
531	descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend	665	descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend
532	might perform better.	666	might perform better.
533		667
534	On the positive side, with the exception of the spurious readiness	668	On the positive side, this backend actually performed fully to
535	notifications, this backend actually performed fully to specification
536	in all tests and is fully embeddable, which is a rare feat among the	669	specification in all tests and is fully embeddable, which is a rare feat
537	OS-specific backends (I vastly prefer correctness over speed hacks).	670	among the OS-specific backends (I vastly prefer correctness over speed
		671	hacks).
		672
		673	On the negative side, the interface is I<bizarre> - so bizarre that
		674	even sun itself gets it wrong in their code examples: The event polling
		675	function sometimes returns events to the caller even though an error
		676	occurred, but with no indication whether it has done so or not (yes, it's
		677	even documented that way) - deadly for edge-triggered interfaces where you
		678	absolutely have to know whether an event occurred or not because you have
		679	to re-arm the watcher.
		680
		681	Fortunately libev seems to be able to work around these idiocies.
538		682
539	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as	683	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
540	C<EVBACKEND_POLL>.	684	C<EVBACKEND_POLL>.
541		685
542	=item C<EVBACKEND_ALL>	686	=item C<EVBACKEND_ALL>
543		687
544	Try all backends (even potentially broken ones that wouldn't be tried	688	Try all backends (even potentially broken ones that wouldn't be tried
545	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as	689	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as
546	C<EVBACKEND_ALL & ~EVBACKEND_KQUEUE>.	690	C<EVBACKEND_ALL & ~EVBACKEND_KQUEUE>.
547		691
548	It is definitely not recommended to use this flag.	692	It is definitely not recommended to use this flag, use whatever
		693	C<ev_recommended_backends ()> returns, or simply do not specify a backend
		694	at all.
		695
		696	=item C<EVBACKEND_MASK>
		697
		698	Not a backend at all, but a mask to select all backend bits from a
		699	C<flags> value, in case you want to mask out any backends from a flags
		700	value (e.g. when modifying the C<LIBEV_FLAGS> environment variable).
549		701
550	=back	702	=back
551		703
552	If one or more of the backend flags are or'ed into the flags value,	704	If one or more of the backend flags are or'ed into the flags value,
553	then only these backends will be tried (in the reverse order as listed	705	then only these backends will be tried (in the reverse order as listed
554	here). If none are specified, all backends in C<ev_recommended_backends	706	here). If none are specified, all backends in C<ev_recommended_backends
555	()> will be tried.	707	()> will be tried.
556		708
557	Example: This is the most typical usage.
558
559	if (!ev_default_loop (0))
560	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
561
562	Example: Restrict libev to the select and poll backends, and do not allow
563	environment settings to be taken into account:
564
565	ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);
566
567	Example: Use whatever libev has to offer, but make sure that kqueue is
568	used if available (warning, breaks stuff, best use only with your own
569	private event loop and only if you know the OS supports your types of
570	fds):
571
572	ev_default_loop (ev_recommended_backends () | EVBACKEND_KQUEUE);
573
574	=item struct ev_loop *ev_loop_new (unsigned int flags)
575
576	Similar to C<ev_default_loop>, but always creates a new event loop that is
577	always distinct from the default loop.
578
579	Note that this function I<is> thread-safe, and one common way to use
580	libev with threads is indeed to create one loop per thread, and using the
581	default loop in the "main" or "initial" thread.
582
583	Example: Try to create a event loop that uses epoll and nothing else.	709	Example: Try to create a event loop that uses epoll and nothing else.
584		710
585	struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV);	711	struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV);
586	if (!epoller)	712	if (!epoller)
587	fatal ("no epoll found here, maybe it hides under your chair");	713	fatal ("no epoll found here, maybe it hides under your chair");
588		714
		715	Example: Use whatever libev has to offer, but make sure that kqueue is
		716	used if available.
		717
		718	struct ev_loop *loop = ev_loop_new (ev_recommended_backends () | EVBACKEND_KQUEUE);
		719
		720	Example: Similarly, on linux, you mgiht want to take advantage of the
		721	linux aio backend if possible, but fall back to something else if that
		722	isn't available.
		723
		724	struct ev_loop *loop = ev_loop_new (ev_recommended_backends () | EVBACKEND_LINUXAIO);
		725
589	=item ev_default_destroy ()	726	=item ev_loop_destroy (loop)
590		727
591	Destroys the default loop (frees all memory and kernel state etc.). None	728	Destroys an event loop object (frees all memory and kernel state
592	of the active event watchers will be stopped in the normal sense, so	729	etc.). None of the active event watchers will be stopped in the normal
593	e.g. C<ev_is_active> might still return true. It is your responsibility to	730	sense, so e.g. C<ev_is_active> might still return true. It is your
594	either stop all watchers cleanly yourself I<before> calling this function,	731	responsibility to either stop all watchers cleanly yourself I<before>
595	or cope with the fact afterwards (which is usually the easiest thing, you	732	calling this function, or cope with the fact afterwards (which is usually
596	can just ignore the watchers and/or C<free ()> them for example).	733	the easiest thing, you can just ignore the watchers and/or C<free ()> them
		734	for example).
597		735
598	Note that certain global state, such as signal state (and installed signal	736	Note that certain global state, such as signal state (and installed signal
599	handlers), will not be freed by this function, and related watchers (such	737	handlers), will not be freed by this function, and related watchers (such
600	as signal and child watchers) would need to be stopped manually.	738	as signal and child watchers) would need to be stopped manually.
601		739
602	In general it is not advisable to call this function except in the	740	This function is normally used on loop objects allocated by
603	rare occasion where you really need to free e.g. the signal handling	741	C<ev_loop_new>, but it can also be used on the default loop returned by
		742	C<ev_default_loop>, in which case it is not thread-safe.
		743
		744	Note that it is not advisable to call this function on the default loop
		745	except in the rare occasion where you really need to free its resources.
604	pipe fds. If you need dynamically allocated loops it is better to use	746	If you need dynamically allocated loops it is better to use C<ev_loop_new>
605	C<ev_loop_new> and C<ev_loop_destroy>.	747	and C<ev_loop_destroy>.
606		748
607	=item ev_loop_destroy (loop)	749	=item ev_loop_fork (loop)
608
609	Like C<ev_default_destroy>, but destroys an event loop created by an
610	earlier call to C<ev_loop_new>.
611
612	=item ev_default_fork ()
613		750
614	This function sets a flag that causes subsequent C<ev_run> iterations	751	This function sets a flag that causes subsequent C<ev_run> iterations
615	to reinitialise the kernel state for backends that have one. Despite the	752	to reinitialise the kernel state for backends that have one. Despite
616	name, you can call it anytime, but it makes most sense after forking, in	753	the name, you can call it anytime you are allowed to start or stop
617	the child process (or both child and parent, but that again makes little	754	watchers (except inside an C<ev_prepare> callback), but it makes most
618	sense). You I<must> call it in the child before using any of the libev	755	sense after forking, in the child process. You I<must> call it (or use
619	functions, and it will only take effect at the next C<ev_run> iteration.	756	C<EVFLAG_FORKCHECK>) in the child before resuming or calling C<ev_run>.
620		757
		758	In addition, if you want to reuse a loop (via this function or
		759	C<EVFLAG_FORKCHECK>), you I<also> have to ignore C<SIGPIPE>.
		760
621	Again, you I<have> to call it on I<any> loop that you want to re-use after	761	Again, you I<have> to call it on I<any> loop that you want to re-use after
622	a fork, I<even if you do not plan to use the loop in the parent>. This is	762	a fork, I<even if you do not plan to use the loop in the parent>. This is
623	because some kernel interfaces *cough* I<kqueue> *cough* do funny things	763	because some kernel interfaces *cough* I<kqueue> *cough* do funny things
624	during fork.	764	during fork.
625		765
626	On the other hand, you only need to call this function in the child	766	On the other hand, you only need to call this function in the child
…		…
629	call it at all (in fact, C<epoll> is so badly broken that it makes a	769	call it at all (in fact, C<epoll> is so badly broken that it makes a
630	difference, but libev will usually detect this case on its own and do a	770	difference, but libev will usually detect this case on its own and do a
631	costly reset of the backend).	771	costly reset of the backend).
632		772
633	The function itself is quite fast and it's usually not a problem to call	773	The function itself is quite fast and it's usually not a problem to call
634	it just in case after a fork. To make this easy, the function will fit in	774	it just in case after a fork.
635	quite nicely into a call to C<pthread_atfork>:
636		775
		776	Example: Automate calling C<ev_loop_fork> on the default loop when
		777	using pthreads.
		778
		779	static void
		780	post_fork_child (void)
		781	{
		782	ev_loop_fork (EV_DEFAULT);
		783	}
		784
		785	...
637	pthread_atfork (0, 0, ev_default_fork);	786	pthread_atfork (0, 0, post_fork_child);
638
639	=item ev_loop_fork (loop)
640
641	Like C<ev_default_fork>, but acts on an event loop created by
642	C<ev_loop_new>. Yes, you have to call this on every allocated event loop
643	after fork that you want to re-use in the child, and how you keep track of
644	them is entirely your own problem.
645		787
646	=item int ev_is_default_loop (loop)	788	=item int ev_is_default_loop (loop)
647		789
648	Returns true when the given loop is, in fact, the default loop, and false	790	Returns true when the given loop is, in fact, the default loop, and false
649	otherwise.	791	otherwise.
…		…
660	prepare and check phases.	802	prepare and check phases.
661		803
662	=item unsigned int ev_depth (loop)	804	=item unsigned int ev_depth (loop)
663		805
664	Returns the number of times C<ev_run> was entered minus the number of	806	Returns the number of times C<ev_run> was entered minus the number of
665	times C<ev_run> was exited, in other words, the recursion depth.	807	times C<ev_run> was exited normally, in other words, the recursion depth.
666		808
667	Outside C<ev_run>, this number is zero. In a callback, this number is	809	Outside C<ev_run>, this number is zero. In a callback, this number is
668	C<1>, unless C<ev_run> was invoked recursively (or from another thread),	810	C<1>, unless C<ev_run> was invoked recursively (or from another thread),
669	in which case it is higher.	811	in which case it is higher.
670		812
671	Leaving C<ev_run> abnormally (setjmp/longjmp, cancelling the thread	813	Leaving C<ev_run> abnormally (setjmp/longjmp, cancelling the thread,
672	etc.), doesn't count as "exit" - consider this as a hint to avoid such	814	throwing an exception etc.), doesn't count as "exit" - consider this
673	ungentleman-like behaviour unless it's really convenient.	815	as a hint to avoid such ungentleman-like behaviour unless it's really
		816	convenient, in which case it is fully supported.
674		817
675	=item unsigned int ev_backend (loop)	818	=item unsigned int ev_backend (loop)
676		819
677	Returns one of the C<EVBACKEND_*> flags indicating the event backend in	820	Returns one of the C<EVBACKEND_*> flags indicating the event backend in
678	use.	821	use.
…		…
693		836
694	This function is rarely useful, but when some event callback runs for a	837	This function is rarely useful, but when some event callback runs for a
695	very long time without entering the event loop, updating libev's idea of	838	very long time without entering the event loop, updating libev's idea of
696	the current time is a good idea.	839	the current time is a good idea.
697		840
698	See also L<The special problem of time updates> in the C<ev_timer> section.	841	See also L</The special problem of time updates> in the C<ev_timer> section.
699		842
700	=item ev_suspend (loop)	843	=item ev_suspend (loop)
701		844
702	=item ev_resume (loop)	845	=item ev_resume (loop)
703		846
…		…
721	without a previous call to C<ev_suspend>.	864	without a previous call to C<ev_suspend>.
722		865
723	Calling C<ev_suspend>/C<ev_resume> has the side effect of updating the	866	Calling C<ev_suspend>/C<ev_resume> has the side effect of updating the
724	event loop time (see C<ev_now_update>).	867	event loop time (see C<ev_now_update>).
725		868
726	=item ev_run (loop, int flags)	869	=item bool ev_run (loop, int flags)
727		870
728	Finally, this is it, the event handler. This function usually is called	871	Finally, this is it, the event handler. This function usually is called
729	after you have initialised all your watchers and you want to start	872	after you have initialised all your watchers and you want to start
730	handling events. It will ask the operating system for any new events, call	873	handling events. It will ask the operating system for any new events, call
731	the watcher callbacks, an then repeat the whole process indefinitely: This	874	the watcher callbacks, and then repeat the whole process indefinitely: This
732	is why event loops are called I<loops>.	875	is why event loops are called I<loops>.
733		876
734	If the flags argument is specified as C<0>, it will keep handling events	877	If the flags argument is specified as C<0>, it will keep handling events
735	until either no event watchers are active anymore or C<ev_break> was	878	until either no event watchers are active anymore or C<ev_break> was
736	called.	879	called.
		880
		881	The return value is false if there are no more active watchers (which
		882	usually means "all jobs done" or "deadlock"), and true in all other cases
		883	(which usually means " you should call C<ev_run> again").
737		884
738	Please note that an explicit C<ev_break> is usually better than	885	Please note that an explicit C<ev_break> is usually better than
739	relying on all watchers to be stopped when deciding when a program has	886	relying on all watchers to be stopped when deciding when a program has
740	finished (especially in interactive programs), but having a program	887	finished (especially in interactive programs), but having a program
741	that automatically loops as long as it has to and no longer by virtue	888	that automatically loops as long as it has to and no longer by virtue
742	of relying on its watchers stopping correctly, that is truly a thing of	889	of relying on its watchers stopping correctly, that is truly a thing of
743	beauty.	890	beauty.
744		891
		892	This function is I<mostly> exception-safe - you can break out of a
		893	C<ev_run> call by calling C<longjmp> in a callback, throwing a C++
		894	exception and so on. This does not decrement the C<ev_depth> value, nor
		895	will it clear any outstanding C<EVBREAK_ONE> breaks.
		896
745	A flags value of C<EVRUN_NOWAIT> will look for new events, will handle	897	A flags value of C<EVRUN_NOWAIT> will look for new events, will handle
746	those events and any already outstanding ones, but will not wait and	898	those events and any already outstanding ones, but will not wait and
747	block your process in case there are no events and will return after one	899	block your process in case there are no events and will return after one
748	iteration of the loop. This is sometimes useful to poll and handle new	900	iteration of the loop. This is sometimes useful to poll and handle new
749	events while doing lengthy calculations, to keep the program responsive.	901	events while doing lengthy calculations, to keep the program responsive.
…		…
758	This is useful if you are waiting for some external event in conjunction	910	This is useful if you are waiting for some external event in conjunction
759	with something not expressible using other libev watchers (i.e. "roll your	911	with something not expressible using other libev watchers (i.e. "roll your
760	own C<ev_run>"). However, a pair of C<ev_prepare>/C<ev_check> watchers is	912	own C<ev_run>"). However, a pair of C<ev_prepare>/C<ev_check> watchers is
761	usually a better approach for this kind of thing.	913	usually a better approach for this kind of thing.
762		914
763	Here are the gory details of what C<ev_run> does:	915	Here are the gory details of what C<ev_run> does (this is for your
		916	understanding, not a guarantee that things will work exactly like this in
		917	future versions):
764		918
765	- Increment loop depth.	919	- Increment loop depth.
766	- Reset the ev_break status.	920	- Reset the ev_break status.
767	- Before the first iteration, call any pending watchers.	921	- Before the first iteration, call any pending watchers.
768	LOOP:	922	LOOP:
…		…
801	anymore.	955	anymore.
802		956
803	... queue jobs here, make sure they register event watchers as long	957	... queue jobs here, make sure they register event watchers as long
804	... as they still have work to do (even an idle watcher will do..)	958	... as they still have work to do (even an idle watcher will do..)
805	ev_run (my_loop, 0);	959	ev_run (my_loop, 0);
806	... jobs done or somebody called unloop. yeah!	960	... jobs done or somebody called break. yeah!
807		961
808	=item ev_break (loop, how)	962	=item ev_break (loop, how)
809		963
810	Can be used to make a call to C<ev_run> return early (but only after it	964	Can be used to make a call to C<ev_run> return early (but only after it
811	has processed all outstanding events). The C<how> argument must be either	965	has processed all outstanding events). The C<how> argument must be either
812	C<EVBREAK_ONE>, which will make the innermost C<ev_run> call return, or	966	C<EVBREAK_ONE>, which will make the innermost C<ev_run> call return, or
813	C<EVBREAK_ALL>, which will make all nested C<ev_run> calls return.	967	C<EVBREAK_ALL>, which will make all nested C<ev_run> calls return.
814		968
815	This "unloop state" will be cleared when entering C<ev_run> again.	969	This "break state" will be cleared on the next call to C<ev_run>.
816		970
817	It is safe to call C<ev_break> from outside any C<ev_run> calls. ##TODO##	971	It is safe to call C<ev_break> from outside any C<ev_run> calls, too, in
		972	which case it will have no effect.
818		973
819	=item ev_ref (loop)	974	=item ev_ref (loop)
820		975
821	=item ev_unref (loop)	976	=item ev_unref (loop)
822		977
…		…
843	running when nothing else is active.	998	running when nothing else is active.
844		999
845	ev_signal exitsig;	1000	ev_signal exitsig;
846	ev_signal_init (&exitsig, sig_cb, SIGINT);	1001	ev_signal_init (&exitsig, sig_cb, SIGINT);
847	ev_signal_start (loop, &exitsig);	1002	ev_signal_start (loop, &exitsig);
848	evf_unref (loop);	1003	ev_unref (loop);
849		1004
850	Example: For some weird reason, unregister the above signal handler again.	1005	Example: For some weird reason, unregister the above signal handler again.
851		1006
852	ev_ref (loop);	1007	ev_ref (loop);
853	ev_signal_stop (loop, &exitsig);	1008	ev_signal_stop (loop, &exitsig);
…		…
873	overhead for the actual polling but can deliver many events at once.	1028	overhead for the actual polling but can deliver many events at once.
874		1029
875	By setting a higher I<io collect interval> you allow libev to spend more	1030	By setting a higher I<io collect interval> you allow libev to spend more
876	time collecting I/O events, so you can handle more events per iteration,	1031	time collecting I/O events, so you can handle more events per iteration,
877	at the cost of increasing latency. Timeouts (both C<ev_periodic> and	1032	at the cost of increasing latency. Timeouts (both C<ev_periodic> and
878	C<ev_timer>) will be not affected. Setting this to a non-null value will	1033	C<ev_timer>) will not be affected. Setting this to a non-null value will
879	introduce an additional C<ev_sleep ()> call into most loop iterations. The	1034	introduce an additional C<ev_sleep ()> call into most loop iterations. The
880	sleep time ensures that libev will not poll for I/O events more often then	1035	sleep time ensures that libev will not poll for I/O events more often then
881	once per this interval, on average.	1036	once per this interval, on average (as long as the host time resolution is
		1037	good enough).
882		1038
883	Likewise, by setting a higher I<timeout collect interval> you allow libev	1039	Likewise, by setting a higher I<timeout collect interval> you allow libev
884	to spend more time collecting timeouts, at the expense of increased	1040	to spend more time collecting timeouts, at the expense of increased
885	latency/jitter/inexactness (the watcher callback will be called	1041	latency/jitter/inexactness (the watcher callback will be called
886	later). C<ev_io> watchers will not be affected. Setting this to a non-null	1042	later). C<ev_io> watchers will not be affected. Setting this to a non-null
…		…
932	invoke the actual watchers inside another context (another thread etc.).	1088	invoke the actual watchers inside another context (another thread etc.).
933		1089
934	If you want to reset the callback, use C<ev_invoke_pending> as new	1090	If you want to reset the callback, use C<ev_invoke_pending> as new
935	callback.	1091	callback.
936		1092
937	=item ev_set_loop_release_cb (loop, void (*release)(EV_P), void (*acquire)(EV_P))	1093	=item ev_set_loop_release_cb (loop, void (*release)(EV_P) throw (), void (*acquire)(EV_P) throw ())
938		1094
939	Sometimes you want to share the same loop between multiple threads. This	1095	Sometimes you want to share the same loop between multiple threads. This
940	can be done relatively simply by putting mutex_lock/unlock calls around	1096	can be done relatively simply by putting mutex_lock/unlock calls around
941	each call to a libev function.	1097	each call to a libev function.
942		1098
943	However, C<ev_run> can run an indefinite time, so it is not feasible	1099	However, C<ev_run> can run an indefinite time, so it is not feasible
944	to wait for it to return. One way around this is to wake up the event	1100	to wait for it to return. One way around this is to wake up the event
945	loop via C<ev_break> and C<av_async_send>, another way is to set these	1101	loop via C<ev_break> and C<ev_async_send>, another way is to set these
946	I<release> and I<acquire> callbacks on the loop.	1102	I<release> and I<acquire> callbacks on the loop.
947		1103
948	When set, then C<release> will be called just before the thread is	1104	When set, then C<release> will be called just before the thread is
949	suspended waiting for new events, and C<acquire> is called just	1105	suspended waiting for new events, and C<acquire> is called just
950	afterwards.	1106	afterwards.
…		…
965	See also the locking example in the C<THREADS> section later in this	1121	See also the locking example in the C<THREADS> section later in this
966	document.	1122	document.
967		1123
968	=item ev_set_userdata (loop, void *data)	1124	=item ev_set_userdata (loop, void *data)
969		1125
970	=item ev_userdata (loop)	1126	=item void *ev_userdata (loop)
971		1127
972	Set and retrieve a single C<void *> associated with a loop. When	1128	Set and retrieve a single C<void *> associated with a loop. When
973	C<ev_set_userdata> has never been called, then C<ev_userdata> returns	1129	C<ev_set_userdata> has never been called, then C<ev_userdata> returns
974	C<0.>	1130	C<0>.
975		1131
976	These two functions can be used to associate arbitrary data with a loop,	1132	These two functions can be used to associate arbitrary data with a loop,
977	and are intended solely for the C<invoke_pending_cb>, C<release> and	1133	and are intended solely for the C<invoke_pending_cb>, C<release> and
978	C<acquire> callbacks described above, but of course can be (ab-)used for	1134	C<acquire> callbacks described above, but of course can be (ab-)used for
979	any other purpose as well.	1135	any other purpose as well.
…		…
1090		1246
1091	=item C<EV_PREPARE>	1247	=item C<EV_PREPARE>
1092		1248
1093	=item C<EV_CHECK>	1249	=item C<EV_CHECK>
1094		1250
1095	All C<ev_prepare> watchers are invoked just I<before> C<ev_run> starts	1251	All C<ev_prepare> watchers are invoked just I<before> C<ev_run> starts to
1096	to gather new events, and all C<ev_check> watchers are invoked just after	1252	gather new events, and all C<ev_check> watchers are queued (not invoked)
1097	C<ev_run> has gathered them, but before it invokes any callbacks for any	1253	just after C<ev_run> has gathered them, but before it queues any callbacks
		1254	for any received events. That means C<ev_prepare> watchers are the last
		1255	watchers invoked before the event loop sleeps or polls for new events, and
		1256	C<ev_check> watchers will be invoked before any other watchers of the same
		1257	or lower priority within an event loop iteration.
		1258
1098	received events. Callbacks of both watcher types can start and stop as	1259	Callbacks of both watcher types can start and stop as many watchers as
1099	many watchers as they want, and all of them will be taken into account	1260	they want, and all of them will be taken into account (for example, a
1100	(for example, a C<ev_prepare> watcher might start an idle watcher to keep	1261	C<ev_prepare> watcher might start an idle watcher to keep C<ev_run> from
1101	C<ev_run> from blocking).	1262	blocking).
1102		1263
1103	=item C<EV_EMBED>	1264	=item C<EV_EMBED>
1104		1265
1105	The embedded event loop specified in the C<ev_embed> watcher needs attention.	1266	The embedded event loop specified in the C<ev_embed> watcher needs attention.
1106		1267
1107	=item C<EV_FORK>	1268	=item C<EV_FORK>
1108		1269
1109	The event loop has been resumed in the child process after fork (see	1270	The event loop has been resumed in the child process after fork (see
1110	C<ev_fork>).	1271	C<ev_fork>).
		1272
		1273	=item C<EV_CLEANUP>
		1274
		1275	The event loop is about to be destroyed (see C<ev_cleanup>).
1111		1276
1112	=item C<EV_ASYNC>	1277	=item C<EV_ASYNC>
1113		1278
1114	The given async watcher has been asynchronously notified (see C<ev_async>).	1279	The given async watcher has been asynchronously notified (see C<ev_async>).
1115		1280
…		…
1137	programs, though, as the fd could already be closed and reused for another	1302	programs, though, as the fd could already be closed and reused for another
1138	thing, so beware.	1303	thing, so beware.
1139		1304
1140	=back	1305	=back
1141		1306
		1307	=head2 GENERIC WATCHER FUNCTIONS
		1308
		1309	=over 4
		1310
		1311	=item C<ev_init> (ev_TYPE *watcher, callback)
		1312
		1313	This macro initialises the generic portion of a watcher. The contents
		1314	of the watcher object can be arbitrary (so C<malloc> will do). Only
		1315	the generic parts of the watcher are initialised, you I<need> to call
		1316	the type-specific C<ev_TYPE_set> macro afterwards to initialise the
		1317	type-specific parts. For each type there is also a C<ev_TYPE_init> macro
		1318	which rolls both calls into one.
		1319
		1320	You can reinitialise a watcher at any time as long as it has been stopped
		1321	(or never started) and there are no pending events outstanding.
		1322
		1323	The callback is always of type C<void (*)(struct ev_loop *loop, ev_TYPE *watcher,
		1324	int revents)>.
		1325
		1326	Example: Initialise an C<ev_io> watcher in two steps.
		1327
		1328	ev_io w;
		1329	ev_init (&w, my_cb);
		1330	ev_io_set (&w, STDIN_FILENO, EV_READ);
		1331
		1332	=item C<ev_TYPE_set> (ev_TYPE *watcher, [args])
		1333
		1334	This macro initialises the type-specific parts of a watcher. You need to
		1335	call C<ev_init> at least once before you call this macro, but you can
		1336	call C<ev_TYPE_set> any number of times. You must not, however, call this
		1337	macro on a watcher that is active (it can be pending, however, which is a
		1338	difference to the C<ev_init> macro).
		1339
		1340	Although some watcher types do not have type-specific arguments
		1341	(e.g. C<ev_prepare>) you still need to call its C<set> macro.
		1342
		1343	See C<ev_init>, above, for an example.
		1344
		1345	=item C<ev_TYPE_init> (ev_TYPE *watcher, callback, [args])
		1346
		1347	This convenience macro rolls both C<ev_init> and C<ev_TYPE_set> macro
		1348	calls into a single call. This is the most convenient method to initialise
		1349	a watcher. The same limitations apply, of course.
		1350
		1351	Example: Initialise and set an C<ev_io> watcher in one step.
		1352
		1353	ev_io_init (&w, my_cb, STDIN_FILENO, EV_READ);
		1354
		1355	=item C<ev_TYPE_start> (loop, ev_TYPE *watcher)
		1356
		1357	Starts (activates) the given watcher. Only active watchers will receive
		1358	events. If the watcher is already active nothing will happen.
		1359
		1360	Example: Start the C<ev_io> watcher that is being abused as example in this
		1361	whole section.
		1362
		1363	ev_io_start (EV_DEFAULT_UC, &w);
		1364
		1365	=item C<ev_TYPE_stop> (loop, ev_TYPE *watcher)
		1366
		1367	Stops the given watcher if active, and clears the pending status (whether
		1368	the watcher was active or not).
		1369
		1370	It is possible that stopped watchers are pending - for example,
		1371	non-repeating timers are being stopped when they become pending - but
		1372	calling C<ev_TYPE_stop> ensures that the watcher is neither active nor
		1373	pending. If you want to free or reuse the memory used by the watcher it is
		1374	therefore a good idea to always call its C<ev_TYPE_stop> function.
		1375
		1376	=item bool ev_is_active (ev_TYPE *watcher)
		1377
		1378	Returns a true value iff the watcher is active (i.e. it has been started
		1379	and not yet been stopped). As long as a watcher is active you must not modify
		1380	it.
		1381
		1382	=item bool ev_is_pending (ev_TYPE *watcher)
		1383
		1384	Returns a true value iff the watcher is pending, (i.e. it has outstanding
		1385	events but its callback has not yet been invoked). As long as a watcher
		1386	is pending (but not active) you must not call an init function on it (but
		1387	C<ev_TYPE_set> is safe), you must not change its priority, and you must
		1388	make sure the watcher is available to libev (e.g. you cannot C<free ()>
		1389	it).
		1390
		1391	=item callback ev_cb (ev_TYPE *watcher)
		1392
		1393	Returns the callback currently set on the watcher.
		1394
		1395	=item ev_set_cb (ev_TYPE *watcher, callback)
		1396
		1397	Change the callback. You can change the callback at virtually any time
		1398	(modulo threads).
		1399
		1400	=item ev_set_priority (ev_TYPE *watcher, int priority)
		1401
		1402	=item int ev_priority (ev_TYPE *watcher)
		1403
		1404	Set and query the priority of the watcher. The priority is a small
		1405	integer between C<EV_MAXPRI> (default: C<2>) and C<EV_MINPRI>
		1406	(default: C<-2>). Pending watchers with higher priority will be invoked
		1407	before watchers with lower priority, but priority will not keep watchers
		1408	from being executed (except for C<ev_idle> watchers).
		1409
		1410	If you need to suppress invocation when higher priority events are pending
		1411	you need to look at C<ev_idle> watchers, which provide this functionality.
		1412
		1413	You I<must not> change the priority of a watcher as long as it is active or
		1414	pending.
		1415
		1416	Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is
		1417	fine, as long as you do not mind that the priority value you query might
		1418	or might not have been clamped to the valid range.
		1419
		1420	The default priority used by watchers when no priority has been set is
		1421	always C<0>, which is supposed to not be too high and not be too low :).
		1422
		1423	See L</WATCHER PRIORITY MODELS>, below, for a more thorough treatment of
		1424	priorities.
		1425
		1426	=item ev_invoke (loop, ev_TYPE *watcher, int revents)
		1427
		1428	Invoke the C<watcher> with the given C<loop> and C<revents>. Neither
		1429	C<loop> nor C<revents> need to be valid as long as the watcher callback
		1430	can deal with that fact, as both are simply passed through to the
		1431	callback.
		1432
		1433	=item int ev_clear_pending (loop, ev_TYPE *watcher)
		1434
		1435	If the watcher is pending, this function clears its pending status and
		1436	returns its C<revents> bitset (as if its callback was invoked). If the
		1437	watcher isn't pending it does nothing and returns C<0>.
		1438
		1439	Sometimes it can be useful to "poll" a watcher instead of waiting for its
		1440	callback to be invoked, which can be accomplished with this function.
		1441
		1442	=item ev_feed_event (loop, ev_TYPE *watcher, int revents)
		1443
		1444	Feeds the given event set into the event loop, as if the specified event
		1445	had happened for the specified watcher (which must be a pointer to an
		1446	initialised but not necessarily started event watcher). Obviously you must
		1447	not free the watcher as long as it has pending events.
		1448
		1449	Stopping the watcher, letting libev invoke it, or calling
		1450	C<ev_clear_pending> will clear the pending event, even if the watcher was
		1451	not started in the first place.
		1452
		1453	See also C<ev_feed_fd_event> and C<ev_feed_signal_event> for related
		1454	functions that do not need a watcher.
		1455
		1456	=back
		1457
		1458	See also the L</ASSOCIATING CUSTOM DATA WITH A WATCHER> and L</BUILDING YOUR
		1459	OWN COMPOSITE WATCHERS> idioms.
		1460
1142	=head2 WATCHER STATES	1461	=head2 WATCHER STATES
1143		1462
1144	There are various watcher states mentioned throughout this manual -	1463	There are various watcher states mentioned throughout this manual -
1145	active, pending and so on. In this section these states and the rules to	1464	active, pending and so on. In this section these states and the rules to
1146	transition between them will be described in more detail - and while these	1465	transition between them will be described in more detail - and while these
1147	rules might look complicated, they usually do "the right thing".	1466	rules might look complicated, they usually do "the right thing".
1148		1467
1149	=over 4	1468	=over 4
1150		1469
1151	=item initialiased	1470	=item initialised
1152		1471
1153	Before a watcher can be registered with the event looop it has to be	1472	Before a watcher can be registered with the event loop it has to be
1154	initialised. This can be done with a call to C<ev_TYPE_init>, or calls to	1473	initialised. This can be done with a call to C<ev_TYPE_init>, or calls to
1155	C<ev_init> followed by the watcher-specific C<ev_TYPE_set> function.	1474	C<ev_init> followed by the watcher-specific C<ev_TYPE_set> function.
1156		1475
1157	In this state it is simply some block of memory that is suitable for use	1476	In this state it is simply some block of memory that is suitable for
1158	in an event loop. It can be moved around, freed, reused etc. at will.	1477	use in an event loop. It can be moved around, freed, reused etc. at
		1478	will - as long as you either keep the memory contents intact, or call
		1479	C<ev_TYPE_init> again.
1159		1480
1160	=item started/running/active	1481	=item started/running/active
1161		1482
1162	Once a watcher has been started with a call to C<ev_TYPE_start> it becomes	1483	Once a watcher has been started with a call to C<ev_TYPE_start> it becomes
1163	property of the event loop, and is actively waiting for events. While in	1484	property of the event loop, and is actively waiting for events. While in
…		…
1191	latter will clear any pending state the watcher might be in, regardless	1512	latter will clear any pending state the watcher might be in, regardless
1192	of whether it was active or not, so stopping a watcher explicitly before	1513	of whether it was active or not, so stopping a watcher explicitly before
1193	freeing it is often a good idea.	1514	freeing it is often a good idea.
1194		1515
1195	While stopped (and not pending) the watcher is essentially in the	1516	While stopped (and not pending) the watcher is essentially in the
1196	initialised state, that is it can be reused, moved, modified in any way	1517	initialised state, that is, it can be reused, moved, modified in any way
1197	you wish.	1518	you wish (but when you trash the memory block, you need to C<ev_TYPE_init>
		1519	it again).
1198		1520
1199	=back	1521	=back
1200
1201	=head2 GENERIC WATCHER FUNCTIONS
1202
1203	=over 4
1204
1205	=item C<ev_init> (ev_TYPE *watcher, callback)
1206
1207	This macro initialises the generic portion of a watcher. The contents
1208	of the watcher object can be arbitrary (so C<malloc> will do). Only
1209	the generic parts of the watcher are initialised, you I<need> to call
1210	the type-specific C<ev_TYPE_set> macro afterwards to initialise the
1211	type-specific parts. For each type there is also a C<ev_TYPE_init> macro
1212	which rolls both calls into one.
1213
1214	You can reinitialise a watcher at any time as long as it has been stopped
1215	(or never started) and there are no pending events outstanding.
1216
1217	The callback is always of type C<void (*)(struct ev_loop *loop, ev_TYPE *watcher,
1218	int revents)>.
1219
1220	Example: Initialise an C<ev_io> watcher in two steps.
1221
1222	ev_io w;
1223	ev_init (&w, my_cb);
1224	ev_io_set (&w, STDIN_FILENO, EV_READ);
1225
1226	=item C<ev_TYPE_set> (ev_TYPE *watcher, [args])
1227
1228	This macro initialises the type-specific parts of a watcher. You need to
1229	call C<ev_init> at least once before you call this macro, but you can
1230	call C<ev_TYPE_set> any number of times. You must not, however, call this
1231	macro on a watcher that is active (it can be pending, however, which is a
1232	difference to the C<ev_init> macro).
1233
1234	Although some watcher types do not have type-specific arguments
1235	(e.g. C<ev_prepare>) you still need to call its C<set> macro.
1236
1237	See C<ev_init>, above, for an example.
1238
1239	=item C<ev_TYPE_init> (ev_TYPE *watcher, callback, [args])
1240
1241	This convenience macro rolls both C<ev_init> and C<ev_TYPE_set> macro
1242	calls into a single call. This is the most convenient method to initialise
1243	a watcher. The same limitations apply, of course.
1244
1245	Example: Initialise and set an C<ev_io> watcher in one step.
1246
1247	ev_io_init (&w, my_cb, STDIN_FILENO, EV_READ);
1248
1249	=item C<ev_TYPE_start> (loop, ev_TYPE *watcher)
1250
1251	Starts (activates) the given watcher. Only active watchers will receive
1252	events. If the watcher is already active nothing will happen.
1253
1254	Example: Start the C<ev_io> watcher that is being abused as example in this
1255	whole section.
1256
1257	ev_io_start (EV_DEFAULT_UC, &w);
1258
1259	=item C<ev_TYPE_stop> (loop, ev_TYPE *watcher)
1260
1261	Stops the given watcher if active, and clears the pending status (whether
1262	the watcher was active or not).
1263
1264	It is possible that stopped watchers are pending - for example,
1265	non-repeating timers are being stopped when they become pending - but
1266	calling C<ev_TYPE_stop> ensures that the watcher is neither active nor
1267	pending. If you want to free or reuse the memory used by the watcher it is
1268	therefore a good idea to always call its C<ev_TYPE_stop> function.
1269
1270	=item bool ev_is_active (ev_TYPE *watcher)
1271
1272	Returns a true value iff the watcher is active (i.e. it has been started
1273	and not yet been stopped). As long as a watcher is active you must not modify
1274	it.
1275
1276	=item bool ev_is_pending (ev_TYPE *watcher)
1277
1278	Returns a true value iff the watcher is pending, (i.e. it has outstanding
1279	events but its callback has not yet been invoked). As long as a watcher
1280	is pending (but not active) you must not call an init function on it (but
1281	C<ev_TYPE_set> is safe), you must not change its priority, and you must
1282	make sure the watcher is available to libev (e.g. you cannot C<free ()>
1283	it).
1284
1285	=item callback ev_cb (ev_TYPE *watcher)
1286
1287	Returns the callback currently set on the watcher.
1288
1289	=item ev_cb_set (ev_TYPE *watcher, callback)
1290
1291	Change the callback. You can change the callback at virtually any time
1292	(modulo threads).
1293
1294	=item ev_set_priority (ev_TYPE *watcher, int priority)
1295
1296	=item int ev_priority (ev_TYPE *watcher)
1297
1298	Set and query the priority of the watcher. The priority is a small
1299	integer between C<EV_MAXPRI> (default: C<2>) and C<EV_MINPRI>
1300	(default: C<-2>). Pending watchers with higher priority will be invoked
1301	before watchers with lower priority, but priority will not keep watchers
1302	from being executed (except for C<ev_idle> watchers).
1303
1304	If you need to suppress invocation when higher priority events are pending
1305	you need to look at C<ev_idle> watchers, which provide this functionality.
1306
1307	You I<must not> change the priority of a watcher as long as it is active or
1308	pending.
1309
1310	Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is
1311	fine, as long as you do not mind that the priority value you query might
1312	or might not have been clamped to the valid range.
1313
1314	The default priority used by watchers when no priority has been set is
1315	always C<0>, which is supposed to not be too high and not be too low :).
1316
1317	See L<WATCHER PRIORITY MODELS>, below, for a more thorough treatment of
1318	priorities.
1319
1320	=item ev_invoke (loop, ev_TYPE *watcher, int revents)
1321
1322	Invoke the C<watcher> with the given C<loop> and C<revents>. Neither
1323	C<loop> nor C<revents> need to be valid as long as the watcher callback
1324	can deal with that fact, as both are simply passed through to the
1325	callback.
1326
1327	=item int ev_clear_pending (loop, ev_TYPE *watcher)
1328
1329	If the watcher is pending, this function clears its pending status and
1330	returns its C<revents> bitset (as if its callback was invoked). If the
1331	watcher isn't pending it does nothing and returns C<0>.
1332
1333	Sometimes it can be useful to "poll" a watcher instead of waiting for its
1334	callback to be invoked, which can be accomplished with this function.
1335
1336	=item ev_feed_event (loop, ev_TYPE *watcher, int revents)
1337
1338	Feeds the given event set into the event loop, as if the specified event
1339	had happened for the specified watcher (which must be a pointer to an
1340	initialised but not necessarily started event watcher). Obviously you must
1341	not free the watcher as long as it has pending events.
1342
1343	Stopping the watcher, letting libev invoke it, or calling
1344	C<ev_clear_pending> will clear the pending event, even if the watcher was
1345	not started in the first place.
1346
1347	See also C<ev_feed_fd_event> and C<ev_feed_signal_event> for related
1348	functions that do not need a watcher.
1349
1350	=back
1351
1352
1353	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER
1354
1355	Each watcher has, by default, a member C<void *data> that you can change
1356	and read at any time: libev will completely ignore it. This can be used
1357	to associate arbitrary data with your watcher. If you need more data and
1358	don't want to allocate memory and store a pointer to it in that data
1359	member, you can also "subclass" the watcher type and provide your own
1360	data:
1361
1362	struct my_io
1363	{
1364	ev_io io;
1365	int otherfd;
1366	void *somedata;
1367	struct whatever *mostinteresting;
1368	};
1369
1370	...
1371	struct my_io w;
1372	ev_io_init (&w.io, my_cb, fd, EV_READ);
1373
1374	And since your callback will be called with a pointer to the watcher, you
1375	can cast it back to your own type:
1376
1377	static void my_cb (struct ev_loop *loop, ev_io *w_, int revents)
1378	{
1379	struct my_io *w = (struct my_io *)w_;
1380	...
1381	}
1382
1383	More interesting and less C-conformant ways of casting your callback type
1384	instead have been omitted.
1385
1386	Another common scenario is to use some data structure with multiple
1387	embedded watchers:
1388
1389	struct my_biggy
1390	{
1391	int some_data;
1392	ev_timer t1;
1393	ev_timer t2;
1394	}
1395
1396	In this case getting the pointer to C<my_biggy> is a bit more
1397	complicated: Either you store the address of your C<my_biggy> struct
1398	in the C<data> member of the watcher (for woozies), or you need to use
1399	some pointer arithmetic using C<offsetof> inside your watchers (for real
1400	programmers):
1401
1402	#include <stddef.h>
1403
1404	static void
1405	t1_cb (EV_P_ ev_timer *w, int revents)
1406	{
1407	struct my_biggy big = (struct my_biggy *)
1408	(((char *)w) - offsetof (struct my_biggy, t1));
1409	}
1410
1411	static void
1412	t2_cb (EV_P_ ev_timer *w, int revents)
1413	{
1414	struct my_biggy big = (struct my_biggy *)
1415	(((char *)w) - offsetof (struct my_biggy, t2));
1416	}
1417		1522
1418	=head2 WATCHER PRIORITY MODELS	1523	=head2 WATCHER PRIORITY MODELS
1419		1524
1420	Many event loops support I<watcher priorities>, which are usually small	1525	Many event loops support I<watcher priorities>, which are usually small
1421	integers that influence the ordering of event callback invocation	1526	integers that influence the ordering of event callback invocation
…		…
1548	In general you can register as many read and/or write event watchers per	1653	In general you can register as many read and/or write event watchers per
1549	fd as you want (as long as you don't confuse yourself). Setting all file	1654	fd as you want (as long as you don't confuse yourself). Setting all file
1550	descriptors to non-blocking mode is also usually a good idea (but not	1655	descriptors to non-blocking mode is also usually a good idea (but not
1551	required if you know what you are doing).	1656	required if you know what you are doing).
1552		1657
1553	If you cannot use non-blocking mode, then force the use of a
1554	known-to-be-good backend (at the time of this writing, this includes only
1555	C<EVBACKEND_SELECT> and C<EVBACKEND_POLL>). The same applies to file
1556	descriptors for which non-blocking operation makes no sense (such as
1557	files) - libev doesn't guarantee any specific behaviour in that case.
1558
1559	Another thing you have to watch out for is that it is quite easy to	1658	Another thing you have to watch out for is that it is quite easy to
1560	receive "spurious" readiness notifications, that is your callback might	1659	receive "spurious" readiness notifications, that is, your callback might
1561	be called with C<EV_READ> but a subsequent C<read>(2) will actually block	1660	be called with C<EV_READ> but a subsequent C<read>(2) will actually block
1562	because there is no data. Not only are some backends known to create a	1661	because there is no data. It is very easy to get into this situation even
1563	lot of those (for example Solaris ports), it is very easy to get into	1662	with a relatively standard program structure. Thus it is best to always
1564	this situation even with a relatively standard program structure. Thus	1663	use non-blocking I/O: An extra C<read>(2) returning C<EAGAIN> is far
1565	it is best to always use non-blocking I/O: An extra C<read>(2) returning
1566	C<EAGAIN> is far preferable to a program hanging until some data arrives.	1664	preferable to a program hanging until some data arrives.
1567		1665
1568	If you cannot run the fd in non-blocking mode (for example you should	1666	If you cannot run the fd in non-blocking mode (for example you should
1569	not play around with an Xlib connection), then you have to separately	1667	not play around with an Xlib connection), then you have to separately
1570	re-test whether a file descriptor is really ready with a known-to-be good	1668	re-test whether a file descriptor is really ready with a known-to-be good
1571	interface such as poll (fortunately in our Xlib example, Xlib already	1669	interface such as poll (fortunately in the case of Xlib, it already does
1572	does this on its own, so its quite safe to use). Some people additionally	1670	this on its own, so its quite safe to use). Some people additionally
1573	use C<SIGALRM> and an interval timer, just to be sure you won't block	1671	use C<SIGALRM> and an interval timer, just to be sure you won't block
1574	indefinitely.	1672	indefinitely.
1575		1673
1576	But really, best use non-blocking mode.	1674	But really, best use non-blocking mode.
1577		1675
1578	=head3 The special problem of disappearing file descriptors	1676	=head3 The special problem of disappearing file descriptors
1579		1677
1580	Some backends (e.g. kqueue, epoll) need to be told about closing a file	1678	Some backends (e.g. kqueue, epoll, linuxaio) need to be told about closing
1581	descriptor (either due to calling C<close> explicitly or any other means,	1679	a file descriptor (either due to calling C<close> explicitly or any other
1582	such as C<dup2>). The reason is that you register interest in some file	1680	means, such as C<dup2>). The reason is that you register interest in some
1583	descriptor, but when it goes away, the operating system will silently drop	1681	file descriptor, but when it goes away, the operating system will silently
1584	this interest. If another file descriptor with the same number then is	1682	drop this interest. If another file descriptor with the same number then
1585	registered with libev, there is no efficient way to see that this is, in	1683	is registered with libev, there is no efficient way to see that this is,
1586	fact, a different file descriptor.	1684	in fact, a different file descriptor.
1587		1685
1588	To avoid having to explicitly tell libev about such cases, libev follows	1686	To avoid having to explicitly tell libev about such cases, libev follows
1589	the following policy: Each time C<ev_io_set> is being called, libev	1687	the following policy: Each time C<ev_io_set> is being called, libev
1590	will assume that this is potentially a new file descriptor, otherwise	1688	will assume that this is potentially a new file descriptor, otherwise
1591	it is assumed that the file descriptor stays the same. That means that	1689	it is assumed that the file descriptor stays the same. That means that
…		…
1605		1703
1606	There is no workaround possible except not registering events	1704	There is no workaround possible except not registering events
1607	for potentially C<dup ()>'ed file descriptors, or to resort to	1705	for potentially C<dup ()>'ed file descriptors, or to resort to
1608	C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>.	1706	C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>.
1609		1707
		1708	=head3 The special problem of files
		1709
		1710	Many people try to use C<select> (or libev) on file descriptors
		1711	representing files, and expect it to become ready when their program
		1712	doesn't block on disk accesses (which can take a long time on their own).
		1713
		1714	However, this cannot ever work in the "expected" way - you get a readiness
		1715	notification as soon as the kernel knows whether and how much data is
		1716	there, and in the case of open files, that's always the case, so you
		1717	always get a readiness notification instantly, and your read (or possibly
		1718	write) will still block on the disk I/O.
		1719
		1720	Another way to view it is that in the case of sockets, pipes, character
		1721	devices and so on, there is another party (the sender) that delivers data
		1722	on its own, but in the case of files, there is no such thing: the disk
		1723	will not send data on its own, simply because it doesn't know what you
		1724	wish to read - you would first have to request some data.
		1725
		1726	Since files are typically not-so-well supported by advanced notification
		1727	mechanism, libev tries hard to emulate POSIX behaviour with respect
		1728	to files, even though you should not use it. The reason for this is
		1729	convenience: sometimes you want to watch STDIN or STDOUT, which is
		1730	usually a tty, often a pipe, but also sometimes files or special devices
		1731	(for example, C<epoll> on Linux works with F</dev/random> but not with
		1732	F</dev/urandom>), and even though the file might better be served with
		1733	asynchronous I/O instead of with non-blocking I/O, it is still useful when
		1734	it "just works" instead of freezing.
		1735
		1736	So avoid file descriptors pointing to files when you know it (e.g. use
		1737	libeio), but use them when it is convenient, e.g. for STDIN/STDOUT, or
		1738	when you rarely read from a file instead of from a socket, and want to
		1739	reuse the same code path.
		1740
1610	=head3 The special problem of fork	1741	=head3 The special problem of fork
1611		1742
1612	Some backends (epoll, kqueue) do not support C<fork ()> at all or exhibit	1743	Some backends (epoll, kqueue, probably linuxaio) do not support C<fork ()>
1613	useless behaviour. Libev fully supports fork, but needs to be told about	1744	at all or exhibit useless behaviour. Libev fully supports fork, but needs
1614	it in the child.	1745	to be told about it in the child if you want to continue to use it in the
		1746	child.
1615		1747
1616	To support fork in your programs, you either have to call	1748	To support fork in your child processes, you have to call C<ev_loop_fork
1617	C<ev_default_fork ()> or C<ev_loop_fork ()> after a fork in the child,	1749	()> after a fork in the child, enable C<EVFLAG_FORKCHECK>, or resort to
1618	enable C<EVFLAG_FORKCHECK>, or resort to C<EVBACKEND_SELECT> or	1750	C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>.
1619	C<EVBACKEND_POLL>.
1620		1751
1621	=head3 The special problem of SIGPIPE	1752	=head3 The special problem of SIGPIPE
1622		1753
1623	While not really specific to libev, it is easy to forget about C<SIGPIPE>:	1754	While not really specific to libev, it is easy to forget about C<SIGPIPE>:
1624	when writing to a pipe whose other end has been closed, your program gets	1755	when writing to a pipe whose other end has been closed, your program gets
…		…
1722	detecting time jumps is hard, and some inaccuracies are unavoidable (the	1853	detecting time jumps is hard, and some inaccuracies are unavoidable (the
1723	monotonic clock option helps a lot here).	1854	monotonic clock option helps a lot here).
1724		1855
1725	The callback is guaranteed to be invoked only I<after> its timeout has	1856	The callback is guaranteed to be invoked only I<after> its timeout has
1726	passed (not I<at>, so on systems with very low-resolution clocks this	1857	passed (not I<at>, so on systems with very low-resolution clocks this
1727	might introduce a small delay). If multiple timers become ready during the	1858	might introduce a small delay, see "the special problem of being too
		1859	early", below). If multiple timers become ready during the same loop
1728	same loop iteration then the ones with earlier time-out values are invoked	1860	iteration then the ones with earlier time-out values are invoked before
1729	before ones of the same priority with later time-out values (but this is	1861	ones of the same priority with later time-out values (but this is no
1730	no longer true when a callback calls C<ev_run> recursively).	1862	longer true when a callback calls C<ev_run> recursively).
1731		1863
1732	=head3 Be smart about timeouts	1864	=head3 Be smart about timeouts
1733		1865
1734	Many real-world problems involve some kind of timeout, usually for error	1866	Many real-world problems involve some kind of timeout, usually for error
1735	recovery. A typical example is an HTTP request - if the other side hangs,	1867	recovery. A typical example is an HTTP request - if the other side hangs,
…		…
1810		1942
1811	In this case, it would be more efficient to leave the C<ev_timer> alone,	1943	In this case, it would be more efficient to leave the C<ev_timer> alone,
1812	but remember the time of last activity, and check for a real timeout only	1944	but remember the time of last activity, and check for a real timeout only
1813	within the callback:	1945	within the callback:
1814		1946
		1947	ev_tstamp timeout = 60.;
1815	ev_tstamp last_activity; // time of last activity	1948	ev_tstamp last_activity; // time of last activity
		1949	ev_timer timer;
1816		1950
1817	static void	1951	static void
1818	callback (EV_P_ ev_timer *w, int revents)	1952	callback (EV_P_ ev_timer *w, int revents)
1819	{	1953	{
1820	ev_tstamp now = ev_now (EV_A);	1954	// calculate when the timeout would happen
1821	ev_tstamp timeout = last_activity + 60.;	1955	ev_tstamp after = last_activity - ev_now (EV_A) + timeout;
1822		1956
1823	// if last_activity + 60. is older than now, we did time out	1957	// if negative, it means we the timeout already occurred
1824	if (timeout < now)	1958	if (after < 0.)
1825	{	1959	{
1826	// timeout occurred, take action	1960	// timeout occurred, take action
1827	}	1961	}
1828	else	1962	else
1829	{	1963	{
1830	// callback was invoked, but there was some activity, re-arm	1964	// callback was invoked, but there was some recent
1831	// the watcher to fire in last_activity + 60, which is	1965	// activity. simply restart the timer to time out
1832	// guaranteed to be in the future, so "again" is positive:	1966	// after "after" seconds, which is the earliest time
1833	w->repeat = timeout - now;	1967	// the timeout can occur.
		1968	ev_timer_set (w, after, 0.);
1834	ev_timer_again (EV_A_ w);	1969	ev_timer_start (EV_A_ w);
1835	}	1970	}
1836	}	1971	}
1837		1972
1838	To summarise the callback: first calculate the real timeout (defined	1973	To summarise the callback: first calculate in how many seconds the
1839	as "60 seconds after the last activity"), then check if that time has	1974	timeout will occur (by calculating the absolute time when it would occur,
1840	been reached, which means something I<did>, in fact, time out. Otherwise	1975	C<last_activity + timeout>, and subtracting the current time, C<ev_now
1841	the callback was invoked too early (C<timeout> is in the future), so	1976	(EV_A)> from that).
1842	re-schedule the timer to fire at that future time, to see if maybe we have
1843	a timeout then.
1844		1977
1845	Note how C<ev_timer_again> is used, taking advantage of the	1978	If this value is negative, then we are already past the timeout, i.e. we
1846	C<ev_timer_again> optimisation when the timer is already running.	1979	timed out, and need to do whatever is needed in this case.
		1980
		1981	Otherwise, we now the earliest time at which the timeout would trigger,
		1982	and simply start the timer with this timeout value.
		1983
		1984	In other words, each time the callback is invoked it will check whether
		1985	the timeout occurred. If not, it will simply reschedule itself to check
		1986	again at the earliest time it could time out. Rinse. Repeat.
1847		1987
1848	This scheme causes more callback invocations (about one every 60 seconds	1988	This scheme causes more callback invocations (about one every 60 seconds
1849	minus half the average time between activity), but virtually no calls to	1989	minus half the average time between activity), but virtually no calls to
1850	libev to change the timeout.	1990	libev to change the timeout.
1851		1991
1852	To start the timer, simply initialise the watcher and set C<last_activity>	1992	To start the machinery, simply initialise the watcher and set
1853	to the current time (meaning we just have some activity :), then call the	1993	C<last_activity> to the current time (meaning there was some activity just
1854	callback, which will "do the right thing" and start the timer:	1994	now), then call the callback, which will "do the right thing" and start
		1995	the timer:
1855		1996
		1997	last_activity = ev_now (EV_A);
1856	ev_init (timer, callback);	1998	ev_init (&timer, callback);
1857	last_activity = ev_now (loop);	1999	callback (EV_A_ &timer, 0);
1858	callback (loop, timer, EV_TIMER);
1859		2000
1860	And when there is some activity, simply store the current time in	2001	When there is some activity, simply store the current time in
1861	C<last_activity>, no libev calls at all:	2002	C<last_activity>, no libev calls at all:
1862		2003
		2004	if (activity detected)
1863	last_activity = ev_now (loop);	2005	last_activity = ev_now (EV_A);
		2006
		2007	When your timeout value changes, then the timeout can be changed by simply
		2008	providing a new value, stopping the timer and calling the callback, which
		2009	will again do the right thing (for example, time out immediately :).
		2010
		2011	timeout = new_value;
		2012	ev_timer_stop (EV_A_ &timer);
		2013	callback (EV_A_ &timer, 0);
1864		2014
1865	This technique is slightly more complex, but in most cases where the	2015	This technique is slightly more complex, but in most cases where the
1866	time-out is unlikely to be triggered, much more efficient.	2016	time-out is unlikely to be triggered, much more efficient.
1867
1868	Changing the timeout is trivial as well (if it isn't hard-coded in the
1869	callback :) - just change the timeout and invoke the callback, which will
1870	fix things for you.
1871		2017
1872	=item 4. Wee, just use a double-linked list for your timeouts.	2018	=item 4. Wee, just use a double-linked list for your timeouts.
1873		2019
1874	If there is not one request, but many thousands (millions...), all	2020	If there is not one request, but many thousands (millions...), all
1875	employing some kind of timeout with the same timeout value, then one can	2021	employing some kind of timeout with the same timeout value, then one can
…		…
1902	Method #1 is almost always a bad idea, and buys you nothing. Method #4 is	2048	Method #1 is almost always a bad idea, and buys you nothing. Method #4 is
1903	rather complicated, but extremely efficient, something that really pays	2049	rather complicated, but extremely efficient, something that really pays
1904	off after the first million or so of active timers, i.e. it's usually	2050	off after the first million or so of active timers, i.e. it's usually
1905	overkill :)	2051	overkill :)
1906		2052
		2053	=head3 The special problem of being too early
		2054
		2055	If you ask a timer to call your callback after three seconds, then
		2056	you expect it to be invoked after three seconds - but of course, this
		2057	cannot be guaranteed to infinite precision. Less obviously, it cannot be
		2058	guaranteed to any precision by libev - imagine somebody suspending the
		2059	process with a STOP signal for a few hours for example.
		2060
		2061	So, libev tries to invoke your callback as soon as possible I<after> the
		2062	delay has occurred, but cannot guarantee this.
		2063
		2064	A less obvious failure mode is calling your callback too early: many event
		2065	loops compare timestamps with a "elapsed delay >= requested delay", but
		2066	this can cause your callback to be invoked much earlier than you would
		2067	expect.
		2068
		2069	To see why, imagine a system with a clock that only offers full second
		2070	resolution (think windows if you can't come up with a broken enough OS
		2071	yourself). If you schedule a one-second timer at the time 500.9, then the
		2072	event loop will schedule your timeout to elapse at a system time of 500
		2073	(500.9 truncated to the resolution) + 1, or 501.
		2074
		2075	If an event library looks at the timeout 0.1s later, it will see "501 >=
		2076	501" and invoke the callback 0.1s after it was started, even though a
		2077	one-second delay was requested - this is being "too early", despite best
		2078	intentions.
		2079
		2080	This is the reason why libev will never invoke the callback if the elapsed
		2081	delay equals the requested delay, but only when the elapsed delay is
		2082	larger than the requested delay. In the example above, libev would only invoke
		2083	the callback at system time 502, or 1.1s after the timer was started.
		2084
		2085	So, while libev cannot guarantee that your callback will be invoked
		2086	exactly when requested, it I<can> and I<does> guarantee that the requested
		2087	delay has actually elapsed, or in other words, it always errs on the "too
		2088	late" side of things.
		2089
1907	=head3 The special problem of time updates	2090	=head3 The special problem of time updates
1908		2091
1909	Establishing the current time is a costly operation (it usually takes at	2092	Establishing the current time is a costly operation (it usually takes
1910	least two system calls): EV therefore updates its idea of the current	2093	at least one system call): EV therefore updates its idea of the current
1911	time only before and after C<ev_run> collects new events, which causes a	2094	time only before and after C<ev_run> collects new events, which causes a
1912	growing difference between C<ev_now ()> and C<ev_time ()> when handling	2095	growing difference between C<ev_now ()> and C<ev_time ()> when handling
1913	lots of events in one iteration.	2096	lots of events in one iteration.
1914		2097
1915	The relative timeouts are calculated relative to the C<ev_now ()>	2098	The relative timeouts are calculated relative to the C<ev_now ()>
1916	time. This is usually the right thing as this timestamp refers to the time	2099	time. This is usually the right thing as this timestamp refers to the time
1917	of the event triggering whatever timeout you are modifying/starting. If	2100	of the event triggering whatever timeout you are modifying/starting. If
1918	you suspect event processing to be delayed and you I<need> to base the	2101	you suspect event processing to be delayed and you I<need> to base the
1919	timeout on the current time, use something like this to adjust for this:	2102	timeout on the current time, use something like the following to adjust
		2103	for it:
1920		2104
1921	ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);	2105	ev_timer_set (&timer, after + (ev_time () - ev_now ()), 0.);
1922		2106
1923	If the event loop is suspended for a long time, you can also force an	2107	If the event loop is suspended for a long time, you can also force an
1924	update of the time returned by C<ev_now ()> by calling C<ev_now_update	2108	update of the time returned by C<ev_now ()> by calling C<ev_now_update
1925	()>.	2109	()>, although that will push the event time of all outstanding events
		2110	further into the future.
		2111
		2112	=head3 The special problem of unsynchronised clocks
		2113
		2114	Modern systems have a variety of clocks - libev itself uses the normal
		2115	"wall clock" clock and, if available, the monotonic clock (to avoid time
		2116	jumps).
		2117
		2118	Neither of these clocks is synchronised with each other or any other clock
		2119	on the system, so C<ev_time ()> might return a considerably different time
		2120	than C<gettimeofday ()> or C<time ()>. On a GNU/Linux system, for example,
		2121	a call to C<gettimeofday> might return a second count that is one higher
		2122	than a directly following call to C<time>.
		2123
		2124	The moral of this is to only compare libev-related timestamps with
		2125	C<ev_time ()> and C<ev_now ()>, at least if you want better precision than
		2126	a second or so.
		2127
		2128	One more problem arises due to this lack of synchronisation: if libev uses
		2129	the system monotonic clock and you compare timestamps from C<ev_time>
		2130	or C<ev_now> from when you started your timer and when your callback is
		2131	invoked, you will find that sometimes the callback is a bit "early".
		2132
		2133	This is because C<ev_timer>s work in real time, not wall clock time, so
		2134	libev makes sure your callback is not invoked before the delay happened,
		2135	I<measured according to the real time>, not the system clock.
		2136
		2137	If your timeouts are based on a physical timescale (e.g. "time out this
		2138	connection after 100 seconds") then this shouldn't bother you as it is
		2139	exactly the right behaviour.
		2140
		2141	If you want to compare wall clock/system timestamps to your timers, then
		2142	you need to use C<ev_periodic>s, as these are based on the wall clock
		2143	time, where your comparisons will always generate correct results.
1926		2144
1927	=head3 The special problems of suspended animation	2145	=head3 The special problems of suspended animation
1928		2146
1929	When you leave the server world it is quite customary to hit machines that	2147	When you leave the server world it is quite customary to hit machines that
1930	can suspend/hibernate - what happens to the clocks during such a suspend?	2148	can suspend/hibernate - what happens to the clocks during such a suspend?
…		…
1960		2178
1961	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)	2179	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)
1962		2180
1963	=item ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)	2181	=item ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)
1964		2182
1965	Configure the timer to trigger after C<after> seconds. If C<repeat>	2183	Configure the timer to trigger after C<after> seconds (fractional and
1966	is C<0.>, then it will automatically be stopped once the timeout is	2184	negative values are supported). If C<repeat> is C<0.>, then it will
1967	reached. If it is positive, then the timer will automatically be	2185	automatically be stopped once the timeout is reached. If it is positive,
1968	configured to trigger again C<repeat> seconds later, again, and again,	2186	then the timer will automatically be configured to trigger again C<repeat>
1969	until stopped manually.	2187	seconds later, again, and again, until stopped manually.
1970		2188
1971	The timer itself will do a best-effort at avoiding drift, that is, if	2189	The timer itself will do a best-effort at avoiding drift, that is, if
1972	you configure a timer to trigger every 10 seconds, then it will normally	2190	you configure a timer to trigger every 10 seconds, then it will normally
1973	trigger at exactly 10 second intervals. If, however, your program cannot	2191	trigger at exactly 10 second intervals. If, however, your program cannot
1974	keep up with the timer (because it takes longer than those 10 seconds to	2192	keep up with the timer (because it takes longer than those 10 seconds to
1975	do stuff) the timer will not fire more than once per event loop iteration.	2193	do stuff) the timer will not fire more than once per event loop iteration.
1976		2194
1977	=item ev_timer_again (loop, ev_timer *)	2195	=item ev_timer_again (loop, ev_timer *)
1978		2196
1979	This will act as if the timer timed out and restart it again if it is	2197	This will act as if the timer timed out, and restarts it again if it is
1980	repeating. The exact semantics are:	2198	repeating. It basically works like calling C<ev_timer_stop>, updating the
		2199	timeout to the C<repeat> value and calling C<ev_timer_start>.
1981		2200
		2201	The exact semantics are as in the following rules, all of which will be
		2202	applied to the watcher:
		2203
		2204	=over 4
		2205
1982	If the timer is pending, its pending status is cleared.	2206	=item If the timer is pending, the pending status is always cleared.
1983		2207
1984	If the timer is started but non-repeating, stop it (as if it timed out).	2208	=item If the timer is started but non-repeating, stop it (as if it timed
		2209	out, without invoking it).
1985		2210
1986	If the timer is repeating, either start it if necessary (with the	2211	=item If the timer is repeating, make the C<repeat> value the new timeout
1987	C<repeat> value), or reset the running timer to the C<repeat> value.	2212	and start the timer, if necessary.
1988		2213
		2214	=back
		2215
1989	This sounds a bit complicated, see L<Be smart about timeouts>, above, for a	2216	This sounds a bit complicated, see L</Be smart about timeouts>, above, for a
1990	usage example.	2217	usage example.
1991		2218
1992	=item ev_tstamp ev_timer_remaining (loop, ev_timer *)	2219	=item ev_tstamp ev_timer_remaining (loop, ev_timer *)
1993		2220
1994	Returns the remaining time until a timer fires. If the timer is active,	2221	Returns the remaining time until a timer fires. If the timer is active,
…		…
2047	Periodic watchers are also timers of a kind, but they are very versatile	2274	Periodic watchers are also timers of a kind, but they are very versatile
2048	(and unfortunately a bit complex).	2275	(and unfortunately a bit complex).
2049		2276
2050	Unlike C<ev_timer>, periodic watchers are not based on real time (or	2277	Unlike C<ev_timer>, periodic watchers are not based on real time (or
2051	relative time, the physical time that passes) but on wall clock time	2278	relative time, the physical time that passes) but on wall clock time
2052	(absolute time, the thing you can read on your calender or clock). The	2279	(absolute time, the thing you can read on your calendar or clock). The
2053	difference is that wall clock time can run faster or slower than real	2280	difference is that wall clock time can run faster or slower than real
2054	time, and time jumps are not uncommon (e.g. when you adjust your	2281	time, and time jumps are not uncommon (e.g. when you adjust your
2055	wrist-watch).	2282	wrist-watch).
2056		2283
2057	You can tell a periodic watcher to trigger after some specific point	2284	You can tell a periodic watcher to trigger after some specific point
…		…
2062	C<ev_timer>, which would still trigger roughly 10 seconds after starting	2289	C<ev_timer>, which would still trigger roughly 10 seconds after starting
2063	it, as it uses a relative timeout).	2290	it, as it uses a relative timeout).
2064		2291
2065	C<ev_periodic> watchers can also be used to implement vastly more complex	2292	C<ev_periodic> watchers can also be used to implement vastly more complex
2066	timers, such as triggering an event on each "midnight, local time", or	2293	timers, such as triggering an event on each "midnight, local time", or
2067	other complicated rules. This cannot be done with C<ev_timer> watchers, as	2294	other complicated rules. This cannot easily be done with C<ev_timer>
2068	those cannot react to time jumps.	2295	watchers, as those cannot react to time jumps.
2069		2296
2070	As with timers, the callback is guaranteed to be invoked only when the	2297	As with timers, the callback is guaranteed to be invoked only when the
2071	point in time where it is supposed to trigger has passed. If multiple	2298	point in time where it is supposed to trigger has passed. If multiple
2072	timers become ready during the same loop iteration then the ones with	2299	timers become ready during the same loop iteration then the ones with
2073	earlier time-out values are invoked before ones with later time-out values	2300	earlier time-out values are invoked before ones with later time-out values
…		…
2114		2341
2115	Another way to think about it (for the mathematically inclined) is that	2342	Another way to think about it (for the mathematically inclined) is that
2116	C<ev_periodic> will try to run the callback in this mode at the next possible	2343	C<ev_periodic> will try to run the callback in this mode at the next possible
2117	time where C<time = offset (mod interval)>, regardless of any time jumps.	2344	time where C<time = offset (mod interval)>, regardless of any time jumps.
2118		2345
2119	For numerical stability it is preferable that the C<offset> value is near	2346	The C<interval> I<MUST> be positive, and for numerical stability, the
2120	C<ev_now ()> (the current time), but there is no range requirement for	2347	interval value should be higher than C<1/8192> (which is around 100
2121	this value, and in fact is often specified as zero.	2348	microseconds) and C<offset> should be higher than C<0> and should have
		2349	at most a similar magnitude as the current time (say, within a factor of
		2350	ten). Typical values for offset are, in fact, C<0> or something between
		2351	C<0> and C<interval>, which is also the recommended range.
2122		2352
2123	Note also that there is an upper limit to how often a timer can fire (CPU	2353	Note also that there is an upper limit to how often a timer can fire (CPU
2124	speed for example), so if C<interval> is very small then timing stability	2354	speed for example), so if C<interval> is very small then timing stability
2125	will of course deteriorate. Libev itself tries to be exact to be about one	2355	will of course deteriorate. Libev itself tries to be exact to be about one
2126	millisecond (if the OS supports it and the machine is fast enough).	2356	millisecond (if the OS supports it and the machine is fast enough).
…		…
2156		2386
2157	NOTE: I<< This callback must always return a time that is higher than or	2387	NOTE: I<< This callback must always return a time that is higher than or
2158	equal to the passed C<now> value >>.	2388	equal to the passed C<now> value >>.
2159		2389
2160	This can be used to create very complex timers, such as a timer that	2390	This can be used to create very complex timers, such as a timer that
2161	triggers on "next midnight, local time". To do this, you would calculate the	2391	triggers on "next midnight, local time". To do this, you would calculate
2162	next midnight after C<now> and return the timestamp value for this. How	2392	the next midnight after C<now> and return the timestamp value for
2163	you do this is, again, up to you (but it is not trivial, which is the main	2393	this. Here is a (completely untested, no error checking) example on how to
2164	reason I omitted it as an example).	2394	do this:
		2395
		2396	#include <time.h>
		2397
		2398	static ev_tstamp
		2399	my_rescheduler (ev_periodic *w, ev_tstamp now)
		2400	{
		2401	time_t tnow = (time_t)now;
		2402	struct tm tm;
		2403	localtime_r (&tnow, &tm);
		2404
		2405	tm.tm_sec = tm.tm_min = tm.tm_hour = 0; // midnight current day
		2406	++tm.tm_mday; // midnight next day
		2407
		2408	return mktime (&tm);
		2409	}
		2410
		2411	Note: this code might run into trouble on days that have more then two
		2412	midnights (beginning and end).
2165		2413
2166	=back	2414	=back
2167		2415
2168	=item ev_periodic_again (loop, ev_periodic *)	2416	=item ev_periodic_again (loop, ev_periodic *)
2169		2417
…		…
2234		2482
2235	ev_periodic hourly_tick;	2483	ev_periodic hourly_tick;
2236	ev_periodic_init (&hourly_tick, clock_cb,	2484	ev_periodic_init (&hourly_tick, clock_cb,
2237	fmod (ev_now (loop), 3600.), 3600., 0);	2485	fmod (ev_now (loop), 3600.), 3600., 0);
2238	ev_periodic_start (loop, &hourly_tick);	2486	ev_periodic_start (loop, &hourly_tick);
2239		2487
2240		2488
2241	=head2 C<ev_signal> - signal me when a signal gets signalled!	2489	=head2 C<ev_signal> - signal me when a signal gets signalled!
2242		2490
2243	Signal watchers will trigger an event when the process receives a specific	2491	Signal watchers will trigger an event when the process receives a specific
2244	signal one or more times. Even though signals are very asynchronous, libev	2492	signal one or more times. Even though signals are very asynchronous, libev
2245	will try it's best to deliver signals synchronously, i.e. as part of the	2493	will try its best to deliver signals synchronously, i.e. as part of the
2246	normal event processing, like any other event.	2494	normal event processing, like any other event.
2247		2495
2248	If you want signals to be delivered truly asynchronously, just use	2496	If you want signals to be delivered truly asynchronously, just use
2249	C<sigaction> as you would do without libev and forget about sharing	2497	C<sigaction> as you would do without libev and forget about sharing
2250	the signal. You can even use C<ev_async> from a signal handler to	2498	the signal. You can even use C<ev_async> from a signal handler to
…		…
2254	only within the same loop, i.e. you can watch for C<SIGINT> in your	2502	only within the same loop, i.e. you can watch for C<SIGINT> in your
2255	default loop and for C<SIGIO> in another loop, but you cannot watch for	2503	default loop and for C<SIGIO> in another loop, but you cannot watch for
2256	C<SIGINT> in both the default loop and another loop at the same time. At	2504	C<SIGINT> in both the default loop and another loop at the same time. At
2257	the moment, C<SIGCHLD> is permanently tied to the default loop.	2505	the moment, C<SIGCHLD> is permanently tied to the default loop.
2258		2506
2259	When the first watcher gets started will libev actually register something	2507	Only after the first watcher for a signal is started will libev actually
2260	with the kernel (thus it coexists with your own signal handlers as long as	2508	register something with the kernel. It thus coexists with your own signal
2261	you don't register any with libev for the same signal).	2509	handlers as long as you don't register any with libev for the same signal.
2262		2510
2263	If possible and supported, libev will install its handlers with	2511	If possible and supported, libev will install its handlers with
2264	C<SA_RESTART> (or equivalent) behaviour enabled, so system calls should	2512	C<SA_RESTART> (or equivalent) behaviour enabled, so system calls should
2265	not be unduly interrupted. If you have a problem with system calls getting	2513	not be unduly interrupted. If you have a problem with system calls getting
2266	interrupted by signals you can block all signals in an C<ev_check> watcher	2514	interrupted by signals you can block all signals in an C<ev_check> watcher
…		…
2269	=head3 The special problem of inheritance over fork/execve/pthread_create	2517	=head3 The special problem of inheritance over fork/execve/pthread_create
2270		2518
2271	Both the signal mask (C<sigprocmask>) and the signal disposition	2519	Both the signal mask (C<sigprocmask>) and the signal disposition
2272	(C<sigaction>) are unspecified after starting a signal watcher (and after	2520	(C<sigaction>) are unspecified after starting a signal watcher (and after
2273	stopping it again), that is, libev might or might not block the signal,	2521	stopping it again), that is, libev might or might not block the signal,
2274	and might or might not set or restore the installed signal handler.	2522	and might or might not set or restore the installed signal handler (but
		2523	see C<EVFLAG_NOSIGMASK>).
2275		2524
2276	While this does not matter for the signal disposition (libev never	2525	While this does not matter for the signal disposition (libev never
2277	sets signals to C<SIG_IGN>, so handlers will be reset to C<SIG_DFL> on	2526	sets signals to C<SIG_IGN>, so handlers will be reset to C<SIG_DFL> on
2278	C<execve>), this matters for the signal mask: many programs do not expect	2527	C<execve>), this matters for the signal mask: many programs do not expect
2279	certain signals to be blocked.	2528	certain signals to be blocked.
…		…
2292	I<has> to modify the signal mask, at least temporarily.	2541	I<has> to modify the signal mask, at least temporarily.
2293		2542
2294	So I can't stress this enough: I<If you do not reset your signal mask when	2543	So I can't stress this enough: I<If you do not reset your signal mask when
2295	you expect it to be empty, you have a race condition in your code>. This	2544	you expect it to be empty, you have a race condition in your code>. This
2296	is not a libev-specific thing, this is true for most event libraries.	2545	is not a libev-specific thing, this is true for most event libraries.
		2546
		2547	=head3 The special problem of threads signal handling
		2548
		2549	POSIX threads has problematic signal handling semantics, specifically,
		2550	a lot of functionality (sigfd, sigwait etc.) only really works if all
		2551	threads in a process block signals, which is hard to achieve.
		2552
		2553	When you want to use sigwait (or mix libev signal handling with your own
		2554	for the same signals), you can tackle this problem by globally blocking
		2555	all signals before creating any threads (or creating them with a fully set
		2556	sigprocmask) and also specifying the C<EVFLAG_NOSIGMASK> when creating
		2557	loops. Then designate one thread as "signal receiver thread" which handles
		2558	these signals. You can pass on any signals that libev might be interested
		2559	in by calling C<ev_feed_signal>.
2297		2560
2298	=head3 Watcher-Specific Functions and Data Members	2561	=head3 Watcher-Specific Functions and Data Members
2299		2562
2300	=over 4	2563	=over 4
2301		2564
…		…
2436		2699
2437	=head2 C<ev_stat> - did the file attributes just change?	2700	=head2 C<ev_stat> - did the file attributes just change?
2438		2701
2439	This watches a file system path for attribute changes. That is, it calls	2702	This watches a file system path for attribute changes. That is, it calls
2440	C<stat> on that path in regular intervals (or when the OS says it changed)	2703	C<stat> on that path in regular intervals (or when the OS says it changed)
2441	and sees if it changed compared to the last time, invoking the callback if	2704	and sees if it changed compared to the last time, invoking the callback
2442	it did.	2705	if it did. Starting the watcher C<stat>'s the file, so only changes that
		2706	happen after the watcher has been started will be reported.
2443		2707
2444	The path does not need to exist: changing from "path exists" to "path does	2708	The path does not need to exist: changing from "path exists" to "path does
2445	not exist" is a status change like any other. The condition "path does not	2709	not exist" is a status change like any other. The condition "path does not
2446	exist" (or more correctly "path cannot be stat'ed") is signified by the	2710	exist" (or more correctly "path cannot be stat'ed") is signified by the
2447	C<st_nlink> field being zero (which is otherwise always forced to be at	2711	C<st_nlink> field being zero (which is otherwise always forced to be at
…		…
2677	Apart from keeping your process non-blocking (which is a useful	2941	Apart from keeping your process non-blocking (which is a useful
2678	effect on its own sometimes), idle watchers are a good place to do	2942	effect on its own sometimes), idle watchers are a good place to do
2679	"pseudo-background processing", or delay processing stuff to after the	2943	"pseudo-background processing", or delay processing stuff to after the
2680	event loop has handled all outstanding events.	2944	event loop has handled all outstanding events.
2681		2945
		2946	=head3 Abusing an C<ev_idle> watcher for its side-effect
		2947
		2948	As long as there is at least one active idle watcher, libev will never
		2949	sleep unnecessarily. Or in other words, it will loop as fast as possible.
		2950	For this to work, the idle watcher doesn't need to be invoked at all - the
		2951	lowest priority will do.
		2952
		2953	This mode of operation can be useful together with an C<ev_check> watcher,
		2954	to do something on each event loop iteration - for example to balance load
		2955	between different connections.
		2956
		2957	See L</Abusing an ev_check watcher for its side-effect> for a longer
		2958	example.
		2959
2682	=head3 Watcher-Specific Functions and Data Members	2960	=head3 Watcher-Specific Functions and Data Members
2683		2961
2684	=over 4	2962	=over 4
2685		2963
2686	=item ev_idle_init (ev_idle *, callback)	2964	=item ev_idle_init (ev_idle *, callback)
…		…
2697	callback, free it. Also, use no error checking, as usual.	2975	callback, free it. Also, use no error checking, as usual.
2698		2976
2699	static void	2977	static void
2700	idle_cb (struct ev_loop *loop, ev_idle *w, int revents)	2978	idle_cb (struct ev_loop *loop, ev_idle *w, int revents)
2701	{	2979	{
		2980	// stop the watcher
		2981	ev_idle_stop (loop, w);
		2982
		2983	// now we can free it
2702	free (w);	2984	free (w);
		2985
2703	// now do something you wanted to do when the program has	2986	// now do something you wanted to do when the program has
2704	// no longer anything immediate to do.	2987	// no longer anything immediate to do.
2705	}	2988	}
2706		2989
2707	ev_idle *idle_watcher = malloc (sizeof (ev_idle));	2990	ev_idle *idle_watcher = malloc (sizeof (ev_idle));
…		…
2709	ev_idle_start (loop, idle_watcher);	2992	ev_idle_start (loop, idle_watcher);
2710		2993
2711		2994
2712	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!	2995	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!
2713		2996
2714	Prepare and check watchers are usually (but not always) used in pairs:	2997	Prepare and check watchers are often (but not always) used in pairs:
2715	prepare watchers get invoked before the process blocks and check watchers	2998	prepare watchers get invoked before the process blocks and check watchers
2716	afterwards.	2999	afterwards.
2717		3000
2718	You I<must not> call C<ev_run> or similar functions that enter	3001	You I<must not> call C<ev_run> (or similar functions that enter the
2719	the current event loop from either C<ev_prepare> or C<ev_check>	3002	current event loop) or C<ev_loop_fork> from either C<ev_prepare> or
2720	watchers. Other loops than the current one are fine, however. The	3003	C<ev_check> watchers. Other loops than the current one are fine,
2721	rationale behind this is that you do not need to check for recursion in	3004	however. The rationale behind this is that you do not need to check
2722	those watchers, i.e. the sequence will always be C<ev_prepare>, blocking,	3005	for recursion in those watchers, i.e. the sequence will always be
2723	C<ev_check> so if you have one watcher of each kind they will always be	3006	C<ev_prepare>, blocking, C<ev_check> so if you have one watcher of each
2724	called in pairs bracketing the blocking call.	3007	kind they will always be called in pairs bracketing the blocking call.
2725		3008
2726	Their main purpose is to integrate other event mechanisms into libev and	3009	Their main purpose is to integrate other event mechanisms into libev and
2727	their use is somewhat advanced. They could be used, for example, to track	3010	their use is somewhat advanced. They could be used, for example, to track
2728	variable changes, implement your own watchers, integrate net-snmp or a	3011	variable changes, implement your own watchers, integrate net-snmp or a
2729	coroutine library and lots more. They are also occasionally useful if	3012	coroutine library and lots more. They are also occasionally useful if
…		…
2747	with priority higher than or equal to the event loop and one coroutine	3030	with priority higher than or equal to the event loop and one coroutine
2748	of lower priority, but only once, using idle watchers to keep the event	3031	of lower priority, but only once, using idle watchers to keep the event
2749	loop from blocking if lower-priority coroutines are active, thus mapping	3032	loop from blocking if lower-priority coroutines are active, thus mapping
2750	low-priority coroutines to idle/background tasks).	3033	low-priority coroutines to idle/background tasks).
2751		3034
2752	It is recommended to give C<ev_check> watchers highest (C<EV_MAXPRI>)	3035	When used for this purpose, it is recommended to give C<ev_check> watchers
2753	priority, to ensure that they are being run before any other watchers	3036	highest (C<EV_MAXPRI>) priority, to ensure that they are being run before
2754	after the poll (this doesn't matter for C<ev_prepare> watchers).	3037	any other watchers after the poll (this doesn't matter for C<ev_prepare>
		3038	watchers).
2755		3039
2756	Also, C<ev_check> watchers (and C<ev_prepare> watchers, too) should not	3040	Also, C<ev_check> watchers (and C<ev_prepare> watchers, too) should not
2757	activate ("feed") events into libev. While libev fully supports this, they	3041	activate ("feed") events into libev. While libev fully supports this, they
2758	might get executed before other C<ev_check> watchers did their job. As	3042	might get executed before other C<ev_check> watchers did their job. As
2759	C<ev_check> watchers are often used to embed other (non-libev) event	3043	C<ev_check> watchers are often used to embed other (non-libev) event
2760	loops those other event loops might be in an unusable state until their	3044	loops those other event loops might be in an unusable state until their
2761	C<ev_check> watcher ran (always remind yourself to coexist peacefully with	3045	C<ev_check> watcher ran (always remind yourself to coexist peacefully with
2762	others).	3046	others).
		3047
		3048	=head3 Abusing an C<ev_check> watcher for its side-effect
		3049
		3050	C<ev_check> (and less often also C<ev_prepare>) watchers can also be
		3051	useful because they are called once per event loop iteration. For
		3052	example, if you want to handle a large number of connections fairly, you
		3053	normally only do a bit of work for each active connection, and if there
		3054	is more work to do, you wait for the next event loop iteration, so other
		3055	connections have a chance of making progress.
		3056
		3057	Using an C<ev_check> watcher is almost enough: it will be called on the
		3058	next event loop iteration. However, that isn't as soon as possible -
		3059	without external events, your C<ev_check> watcher will not be invoked.
		3060
		3061	This is where C<ev_idle> watchers come in handy - all you need is a
		3062	single global idle watcher that is active as long as you have one active
		3063	C<ev_check> watcher. The C<ev_idle> watcher makes sure the event loop
		3064	will not sleep, and the C<ev_check> watcher makes sure a callback gets
		3065	invoked. Neither watcher alone can do that.
2763		3066
2764	=head3 Watcher-Specific Functions and Data Members	3067	=head3 Watcher-Specific Functions and Data Members
2765		3068
2766	=over 4	3069	=over 4
2767		3070
…		…
2968		3271
2969	=over 4	3272	=over 4
2970		3273
2971	=item ev_embed_init (ev_embed *, callback, struct ev_loop *embedded_loop)	3274	=item ev_embed_init (ev_embed *, callback, struct ev_loop *embedded_loop)
2972		3275
2973	=item ev_embed_set (ev_embed *, callback, struct ev_loop *embedded_loop)	3276	=item ev_embed_set (ev_embed *, struct ev_loop *embedded_loop)
2974		3277
2975	Configures the watcher to embed the given loop, which must be	3278	Configures the watcher to embed the given loop, which must be
2976	embeddable. If the callback is C<0>, then C<ev_embed_sweep> will be	3279	embeddable. If the callback is C<0>, then C<ev_embed_sweep> will be
2977	invoked automatically, otherwise it is the responsibility of the callback	3280	invoked automatically, otherwise it is the responsibility of the callback
2978	to invoke it (it will continue to be called until the sweep has been done,	3281	to invoke it (it will continue to be called until the sweep has been done,
…		…
2999	used).	3302	used).
3000		3303
3001	struct ev_loop *loop_hi = ev_default_init (0);	3304	struct ev_loop *loop_hi = ev_default_init (0);
3002	struct ev_loop *loop_lo = 0;	3305	struct ev_loop *loop_lo = 0;
3003	ev_embed embed;	3306	ev_embed embed;
3004		3307
3005	// see if there is a chance of getting one that works	3308	// see if there is a chance of getting one that works
3006	// (remember that a flags value of 0 means autodetection)	3309	// (remember that a flags value of 0 means autodetection)
3007	loop_lo = ev_embeddable_backends () & ev_recommended_backends ()	3310	loop_lo = ev_embeddable_backends () & ev_recommended_backends ()
3008	? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())	3311	? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())
3009	: 0;	3312	: 0;
…		…
3023	C<loop_socket>. (One might optionally use C<EVFLAG_NOENV>, too).	3326	C<loop_socket>. (One might optionally use C<EVFLAG_NOENV>, too).
3024		3327
3025	struct ev_loop *loop = ev_default_init (0);	3328	struct ev_loop *loop = ev_default_init (0);
3026	struct ev_loop *loop_socket = 0;	3329	struct ev_loop *loop_socket = 0;
3027	ev_embed embed;	3330	ev_embed embed;
3028		3331
3029	if (ev_supported_backends () & ~ev_recommended_backends () & EVBACKEND_KQUEUE)	3332	if (ev_supported_backends () & ~ev_recommended_backends () & EVBACKEND_KQUEUE)
3030	if ((loop_socket = ev_loop_new (EVBACKEND_KQUEUE))	3333	if ((loop_socket = ev_loop_new (EVBACKEND_KQUEUE))
3031	{	3334	{
3032	ev_embed_init (&embed, 0, loop_socket);	3335	ev_embed_init (&embed, 0, loop_socket);
3033	ev_embed_start (loop, &embed);	3336	ev_embed_start (loop, &embed);
…		…
3041		3344
3042	=head2 C<ev_fork> - the audacity to resume the event loop after a fork	3345	=head2 C<ev_fork> - the audacity to resume the event loop after a fork
3043		3346
3044	Fork watchers are called when a C<fork ()> was detected (usually because	3347	Fork watchers are called when a C<fork ()> was detected (usually because
3045	whoever is a good citizen cared to tell libev about it by calling	3348	whoever is a good citizen cared to tell libev about it by calling
3046	C<ev_default_fork> or C<ev_loop_fork>). The invocation is done before the	3349	C<ev_loop_fork>). The invocation is done before the event loop blocks next
3047	event loop blocks next and before C<ev_check> watchers are being called,	3350	and before C<ev_check> watchers are being called, and only in the child
3048	and only in the child after the fork. If whoever good citizen calling	3351	after the fork. If whoever good citizen calling C<ev_default_fork> cheats
3049	C<ev_default_fork> cheats and calls it in the wrong process, the fork	3352	and calls it in the wrong process, the fork handlers will be invoked, too,
3050	handlers will be invoked, too, of course.	3353	of course.
3051		3354
3052	=head3 The special problem of life after fork - how is it possible?	3355	=head3 The special problem of life after fork - how is it possible?
3053		3356
3054	Most uses of C<fork()> consist of forking, then some simple calls to set	3357	Most uses of C<fork ()> consist of forking, then some simple calls to set
3055	up/change the process environment, followed by a call to C<exec()>. This	3358	up/change the process environment, followed by a call to C<exec()>. This
3056	sequence should be handled by libev without any problems.	3359	sequence should be handled by libev without any problems.
3057		3360
3058	This changes when the application actually wants to do event handling	3361	This changes when the application actually wants to do event handling
3059	in the child, or both parent in child, in effect "continuing" after the	3362	in the child, or both parent in child, in effect "continuing" after the
…		…
3075	disadvantage of having to use multiple event loops (which do not support	3378	disadvantage of having to use multiple event loops (which do not support
3076	signal watchers).	3379	signal watchers).
3077		3380
3078	When this is not possible, or you want to use the default loop for	3381	When this is not possible, or you want to use the default loop for
3079	other reasons, then in the process that wants to start "fresh", call	3382	other reasons, then in the process that wants to start "fresh", call
3080	C<ev_default_destroy ()> followed by C<ev_default_loop (...)>. Destroying	3383	C<ev_loop_destroy (EV_DEFAULT)> followed by C<ev_default_loop (...)>.
3081	the default loop will "orphan" (not stop) all registered watchers, so you	3384	Destroying the default loop will "orphan" (not stop) all registered
3082	have to be careful not to execute code that modifies those watchers. Note	3385	watchers, so you have to be careful not to execute code that modifies
3083	also that in that case, you have to re-register any signal watchers.	3386	those watchers. Note also that in that case, you have to re-register any
		3387	signal watchers.
3084		3388
3085	=head3 Watcher-Specific Functions and Data Members	3389	=head3 Watcher-Specific Functions and Data Members
3086		3390
3087	=over 4	3391	=over 4
3088		3392
3089	=item ev_fork_init (ev_signal *, callback)	3393	=item ev_fork_init (ev_fork *, callback)
3090		3394
3091	Initialises and configures the fork watcher - it has no parameters of any	3395	Initialises and configures the fork watcher - it has no parameters of any
3092	kind. There is a C<ev_fork_set> macro, but using it is utterly pointless,	3396	kind. There is a C<ev_fork_set> macro, but using it is utterly pointless,
3093	believe me.	3397	really.
3094		3398
3095	=back	3399	=back
3096		3400
3097		3401
		3402	=head2 C<ev_cleanup> - even the best things end
		3403
		3404	Cleanup watchers are called just before the event loop is being destroyed
		3405	by a call to C<ev_loop_destroy>.
		3406
		3407	While there is no guarantee that the event loop gets destroyed, cleanup
		3408	watchers provide a convenient method to install cleanup hooks for your
		3409	program, worker threads and so on - you just to make sure to destroy the
		3410	loop when you want them to be invoked.
		3411
		3412	Cleanup watchers are invoked in the same way as any other watcher. Unlike
		3413	all other watchers, they do not keep a reference to the event loop (which
		3414	makes a lot of sense if you think about it). Like all other watchers, you
		3415	can call libev functions in the callback, except C<ev_cleanup_start>.
		3416
		3417	=head3 Watcher-Specific Functions and Data Members
		3418
		3419	=over 4
		3420
		3421	=item ev_cleanup_init (ev_cleanup *, callback)
		3422
		3423	Initialises and configures the cleanup watcher - it has no parameters of
		3424	any kind. There is a C<ev_cleanup_set> macro, but using it is utterly
		3425	pointless, I assure you.
		3426
		3427	=back
		3428
		3429	Example: Register an atexit handler to destroy the default loop, so any
		3430	cleanup functions are called.
		3431
		3432	static void
		3433	program_exits (void)
		3434	{
		3435	ev_loop_destroy (EV_DEFAULT_UC);
		3436	}
		3437
		3438	...
		3439	atexit (program_exits);
		3440
		3441
3098	=head2 C<ev_async> - how to wake up an event loop	3442	=head2 C<ev_async> - how to wake up an event loop
3099		3443
3100	In general, you cannot use an C<ev_run> from multiple threads or other	3444	In general, you cannot use an C<ev_loop> from multiple threads or other
3101	asynchronous sources such as signal handlers (as opposed to multiple event	3445	asynchronous sources such as signal handlers (as opposed to multiple event
3102	loops - those are of course safe to use in different threads).	3446	loops - those are of course safe to use in different threads).
3103		3447
3104	Sometimes, however, you need to wake up an event loop you do not control,	3448	Sometimes, however, you need to wake up an event loop you do not control,
3105	for example because it belongs to another thread. This is what C<ev_async>	3449	for example because it belongs to another thread. This is what C<ev_async>
…		…
3107	it by calling C<ev_async_send>, which is thread- and signal safe.	3451	it by calling C<ev_async_send>, which is thread- and signal safe.
3108		3452
3109	This functionality is very similar to C<ev_signal> watchers, as signals,	3453	This functionality is very similar to C<ev_signal> watchers, as signals,
3110	too, are asynchronous in nature, and signals, too, will be compressed	3454	too, are asynchronous in nature, and signals, too, will be compressed
3111	(i.e. the number of callback invocations may be less than the number of	3455	(i.e. the number of callback invocations may be less than the number of
3112	C<ev_async_sent> calls).	3456	C<ev_async_send> calls). In fact, you could use signal watchers as a kind
3113		3457	of "global async watchers" by using a watcher on an otherwise unused
3114	Unlike C<ev_signal> watchers, C<ev_async> works with any event loop, not	3458	signal, and C<ev_feed_signal> to signal this watcher from another thread,
3115	just the default loop.	3459	even without knowing which loop owns the signal.
3116		3460
3117	=head3 Queueing	3461	=head3 Queueing
3118		3462
3119	C<ev_async> does not support queueing of data in any way. The reason	3463	C<ev_async> does not support queueing of data in any way. The reason
3120	is that the author does not know of a simple (or any) algorithm for a	3464	is that the author does not know of a simple (or any) algorithm for a
…		…
3212	trust me.	3556	trust me.
3213		3557
3214	=item ev_async_send (loop, ev_async *)	3558	=item ev_async_send (loop, ev_async *)
3215		3559
3216	Sends/signals/activates the given C<ev_async> watcher, that is, feeds	3560	Sends/signals/activates the given C<ev_async> watcher, that is, feeds
3217	an C<EV_ASYNC> event on the watcher into the event loop. Unlike	3561	an C<EV_ASYNC> event on the watcher into the event loop, and instantly
		3562	returns.
		3563
3218	C<ev_feed_event>, this call is safe to do from other threads, signal or	3564	Unlike C<ev_feed_event>, this call is safe to do from other threads,
3219	similar contexts (see the discussion of C<EV_ATOMIC_T> in the embedding	3565	signal or similar contexts (see the discussion of C<EV_ATOMIC_T> in the
3220	section below on what exactly this means).	3566	embedding section below on what exactly this means).
3221		3567
3222	Note that, as with other watchers in libev, multiple events might get	3568	Note that, as with other watchers in libev, multiple events might get
3223	compressed into a single callback invocation (another way to look at this	3569	compressed into a single callback invocation (another way to look at
3224	is that C<ev_async> watchers are level-triggered, set on C<ev_async_send>,	3570	this is that C<ev_async> watchers are level-triggered: they are set on
3225	reset when the event loop detects that).	3571	C<ev_async_send>, reset when the event loop detects that).
3226		3572
3227	This call incurs the overhead of a system call only once per event loop	3573	This call incurs the overhead of at most one extra system call per event
3228	iteration, so while the overhead might be noticeable, it doesn't apply to	3574	loop iteration, if the event loop is blocked, and no syscall at all if
3229	repeated calls to C<ev_async_send> for the same event loop.	3575	the event loop (or your program) is processing events. That means that
		3576	repeated calls are basically free (there is no need to avoid calls for
		3577	performance reasons) and that the overhead becomes smaller (typically
		3578	zero) under load.
3230		3579
3231	=item bool = ev_async_pending (ev_async *)	3580	=item bool = ev_async_pending (ev_async *)
3232		3581
3233	Returns a non-zero value when C<ev_async_send> has been called on the	3582	Returns a non-zero value when C<ev_async_send> has been called on the
3234	watcher but the event has not yet been processed (or even noted) by the	3583	watcher but the event has not yet been processed (or even noted) by the
…		…
3251		3600
3252	There are some other functions of possible interest. Described. Here. Now.	3601	There are some other functions of possible interest. Described. Here. Now.
3253		3602
3254	=over 4	3603	=over 4
3255		3604
3256	=item ev_once (loop, int fd, int events, ev_tstamp timeout, callback)	3605	=item ev_once (loop, int fd, int events, ev_tstamp timeout, callback, arg)
3257		3606
3258	This function combines a simple timer and an I/O watcher, calls your	3607	This function combines a simple timer and an I/O watcher, calls your
3259	callback on whichever event happens first and automatically stops both	3608	callback on whichever event happens first and automatically stops both
3260	watchers. This is useful if you want to wait for a single event on an fd	3609	watchers. This is useful if you want to wait for a single event on an fd
3261	or timeout without having to allocate/configure/start/stop/free one or	3610	or timeout without having to allocate/configure/start/stop/free one or
…		…
3289	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);	3638	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);
3290		3639
3291	=item ev_feed_fd_event (loop, int fd, int revents)	3640	=item ev_feed_fd_event (loop, int fd, int revents)
3292		3641
3293	Feed an event on the given fd, as if a file descriptor backend detected	3642	Feed an event on the given fd, as if a file descriptor backend detected
3294	the given events it.	3643	the given events.
3295		3644
3296	=item ev_feed_signal_event (loop, int signum)	3645	=item ev_feed_signal_event (loop, int signum)
3297		3646
3298	Feed an event as if the given signal occurred (C<loop> must be the default	3647	Feed an event as if the given signal occurred. See also C<ev_feed_signal>,
3299	loop!).	3648	which is async-safe.
3300		3649
3301	=back	3650	=back
		3651
		3652
		3653	=head1 COMMON OR USEFUL IDIOMS (OR BOTH)
		3654
		3655	This section explains some common idioms that are not immediately
		3656	obvious. Note that examples are sprinkled over the whole manual, and this
		3657	section only contains stuff that wouldn't fit anywhere else.
		3658
		3659	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER
		3660
		3661	Each watcher has, by default, a C<void *data> member that you can read
		3662	or modify at any time: libev will completely ignore it. This can be used
		3663	to associate arbitrary data with your watcher. If you need more data and
		3664	don't want to allocate memory separately and store a pointer to it in that
		3665	data member, you can also "subclass" the watcher type and provide your own
		3666	data:
		3667
		3668	struct my_io
		3669	{
		3670	ev_io io;
		3671	int otherfd;
		3672	void *somedata;
		3673	struct whatever *mostinteresting;
		3674	};
		3675
		3676	...
		3677	struct my_io w;
		3678	ev_io_init (&w.io, my_cb, fd, EV_READ);
		3679
		3680	And since your callback will be called with a pointer to the watcher, you
		3681	can cast it back to your own type:
		3682
		3683	static void my_cb (struct ev_loop *loop, ev_io *w_, int revents)
		3684	{
		3685	struct my_io *w = (struct my_io *)w_;
		3686	...
		3687	}
		3688
		3689	More interesting and less C-conformant ways of casting your callback
		3690	function type instead have been omitted.
		3691
		3692	=head2 BUILDING YOUR OWN COMPOSITE WATCHERS
		3693
		3694	Another common scenario is to use some data structure with multiple
		3695	embedded watchers, in effect creating your own watcher that combines
		3696	multiple libev event sources into one "super-watcher":
		3697
		3698	struct my_biggy
		3699	{
		3700	int some_data;
		3701	ev_timer t1;
		3702	ev_timer t2;
		3703	}
		3704
		3705	In this case getting the pointer to C<my_biggy> is a bit more
		3706	complicated: Either you store the address of your C<my_biggy> struct in
		3707	the C<data> member of the watcher (for woozies or C++ coders), or you need
		3708	to use some pointer arithmetic using C<offsetof> inside your watchers (for
		3709	real programmers):
		3710
		3711	#include <stddef.h>
		3712
		3713	static void
		3714	t1_cb (EV_P_ ev_timer *w, int revents)
		3715	{
		3716	struct my_biggy big = (struct my_biggy *)
		3717	(((char *)w) - offsetof (struct my_biggy, t1));
		3718	}
		3719
		3720	static void
		3721	t2_cb (EV_P_ ev_timer *w, int revents)
		3722	{
		3723	struct my_biggy big = (struct my_biggy *)
		3724	(((char *)w) - offsetof (struct my_biggy, t2));
		3725	}
		3726
		3727	=head2 AVOIDING FINISHING BEFORE RETURNING
		3728
		3729	Often you have structures like this in event-based programs:
		3730
		3731	callback ()
		3732	{
		3733	free (request);
		3734	}
		3735
		3736	request = start_new_request (..., callback);
		3737
		3738	The intent is to start some "lengthy" operation. The C<request> could be
		3739	used to cancel the operation, or do other things with it.
		3740
		3741	It's not uncommon to have code paths in C<start_new_request> that
		3742	immediately invoke the callback, for example, to report errors. Or you add
		3743	some caching layer that finds that it can skip the lengthy aspects of the
		3744	operation and simply invoke the callback with the result.
		3745
		3746	The problem here is that this will happen I<before> C<start_new_request>
		3747	has returned, so C<request> is not set.
		3748
		3749	Even if you pass the request by some safer means to the callback, you
		3750	might want to do something to the request after starting it, such as
		3751	canceling it, which probably isn't working so well when the callback has
		3752	already been invoked.
		3753
		3754	A common way around all these issues is to make sure that
		3755	C<start_new_request> I<always> returns before the callback is invoked. If
		3756	C<start_new_request> immediately knows the result, it can artificially
		3757	delay invoking the callback by using a C<prepare> or C<idle> watcher for
		3758	example, or more sneakily, by reusing an existing (stopped) watcher and
		3759	pushing it into the pending queue:
		3760
		3761	ev_set_cb (watcher, callback);
		3762	ev_feed_event (EV_A_ watcher, 0);
		3763
		3764	This way, C<start_new_request> can safely return before the callback is
		3765	invoked, while not delaying callback invocation too much.
		3766
		3767	=head2 MODEL/NESTED EVENT LOOP INVOCATIONS AND EXIT CONDITIONS
		3768
		3769	Often (especially in GUI toolkits) there are places where you have
		3770	I<modal> interaction, which is most easily implemented by recursively
		3771	invoking C<ev_run>.
		3772
		3773	This brings the problem of exiting - a callback might want to finish the
		3774	main C<ev_run> call, but not the nested one (e.g. user clicked "Quit", but
		3775	a modal "Are you sure?" dialog is still waiting), or just the nested one
		3776	and not the main one (e.g. user clocked "Ok" in a modal dialog), or some
		3777	other combination: In these cases, a simple C<ev_break> will not work.
		3778
		3779	The solution is to maintain "break this loop" variable for each C<ev_run>
		3780	invocation, and use a loop around C<ev_run> until the condition is
		3781	triggered, using C<EVRUN_ONCE>:
		3782
		3783	// main loop
		3784	int exit_main_loop = 0;
		3785
		3786	while (!exit_main_loop)
		3787	ev_run (EV_DEFAULT_ EVRUN_ONCE);
		3788
		3789	// in a modal watcher
		3790	int exit_nested_loop = 0;
		3791
		3792	while (!exit_nested_loop)
		3793	ev_run (EV_A_ EVRUN_ONCE);
		3794
		3795	To exit from any of these loops, just set the corresponding exit variable:
		3796
		3797	// exit modal loop
		3798	exit_nested_loop = 1;
		3799
		3800	// exit main program, after modal loop is finished
		3801	exit_main_loop = 1;
		3802
		3803	// exit both
		3804	exit_main_loop = exit_nested_loop = 1;
		3805
		3806	=head2 THREAD LOCKING EXAMPLE
		3807
		3808	Here is a fictitious example of how to run an event loop in a different
		3809	thread from where callbacks are being invoked and watchers are
		3810	created/added/removed.
		3811
		3812	For a real-world example, see the C<EV::Loop::Async> perl module,
		3813	which uses exactly this technique (which is suited for many high-level
		3814	languages).
		3815
		3816	The example uses a pthread mutex to protect the loop data, a condition
		3817	variable to wait for callback invocations, an async watcher to notify the
		3818	event loop thread and an unspecified mechanism to wake up the main thread.
		3819
		3820	First, you need to associate some data with the event loop:
		3821
		3822	typedef struct {
		3823	mutex_t lock; /* global loop lock */
		3824	ev_async async_w;
		3825	thread_t tid;
		3826	cond_t invoke_cv;
		3827	} userdata;
		3828
		3829	void prepare_loop (EV_P)
		3830	{
		3831	// for simplicity, we use a static userdata struct.
		3832	static userdata u;
		3833
		3834	ev_async_init (&u->async_w, async_cb);
		3835	ev_async_start (EV_A_ &u->async_w);
		3836
		3837	pthread_mutex_init (&u->lock, 0);
		3838	pthread_cond_init (&u->invoke_cv, 0);
		3839
		3840	// now associate this with the loop
		3841	ev_set_userdata (EV_A_ u);
		3842	ev_set_invoke_pending_cb (EV_A_ l_invoke);
		3843	ev_set_loop_release_cb (EV_A_ l_release, l_acquire);
		3844
		3845	// then create the thread running ev_run
		3846	pthread_create (&u->tid, 0, l_run, EV_A);
		3847	}
		3848
		3849	The callback for the C<ev_async> watcher does nothing: the watcher is used
		3850	solely to wake up the event loop so it takes notice of any new watchers
		3851	that might have been added:
		3852
		3853	static void
		3854	async_cb (EV_P_ ev_async *w, int revents)
		3855	{
		3856	// just used for the side effects
		3857	}
		3858
		3859	The C<l_release> and C<l_acquire> callbacks simply unlock/lock the mutex
		3860	protecting the loop data, respectively.
		3861
		3862	static void
		3863	l_release (EV_P)
		3864	{
		3865	userdata *u = ev_userdata (EV_A);
		3866	pthread_mutex_unlock (&u->lock);
		3867	}
		3868
		3869	static void
		3870	l_acquire (EV_P)
		3871	{
		3872	userdata *u = ev_userdata (EV_A);
		3873	pthread_mutex_lock (&u->lock);
		3874	}
		3875
		3876	The event loop thread first acquires the mutex, and then jumps straight
		3877	into C<ev_run>:
		3878
		3879	void *
		3880	l_run (void *thr_arg)
		3881	{
		3882	struct ev_loop *loop = (struct ev_loop *)thr_arg;
		3883
		3884	l_acquire (EV_A);
		3885	pthread_setcanceltype (PTHREAD_CANCEL_ASYNCHRONOUS, 0);
		3886	ev_run (EV_A_ 0);
		3887	l_release (EV_A);
		3888
		3889	return 0;
		3890	}
		3891
		3892	Instead of invoking all pending watchers, the C<l_invoke> callback will
		3893	signal the main thread via some unspecified mechanism (signals? pipe
		3894	writes? C<Async::Interrupt>?) and then waits until all pending watchers
		3895	have been called (in a while loop because a) spurious wakeups are possible
		3896	and b) skipping inter-thread-communication when there are no pending
		3897	watchers is very beneficial):
		3898
		3899	static void
		3900	l_invoke (EV_P)
		3901	{
		3902	userdata *u = ev_userdata (EV_A);
		3903
		3904	while (ev_pending_count (EV_A))
		3905	{
		3906	wake_up_other_thread_in_some_magic_or_not_so_magic_way ();
		3907	pthread_cond_wait (&u->invoke_cv, &u->lock);
		3908	}
		3909	}
		3910
		3911	Now, whenever the main thread gets told to invoke pending watchers, it
		3912	will grab the lock, call C<ev_invoke_pending> and then signal the loop
		3913	thread to continue:
		3914
		3915	static void
		3916	real_invoke_pending (EV_P)
		3917	{
		3918	userdata *u = ev_userdata (EV_A);
		3919
		3920	pthread_mutex_lock (&u->lock);
		3921	ev_invoke_pending (EV_A);
		3922	pthread_cond_signal (&u->invoke_cv);
		3923	pthread_mutex_unlock (&u->lock);
		3924	}
		3925
		3926	Whenever you want to start/stop a watcher or do other modifications to an
		3927	event loop, you will now have to lock:
		3928
		3929	ev_timer timeout_watcher;
		3930	userdata *u = ev_userdata (EV_A);
		3931
		3932	ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);
		3933
		3934	pthread_mutex_lock (&u->lock);
		3935	ev_timer_start (EV_A_ &timeout_watcher);
		3936	ev_async_send (EV_A_ &u->async_w);
		3937	pthread_mutex_unlock (&u->lock);
		3938
		3939	Note that sending the C<ev_async> watcher is required because otherwise
		3940	an event loop currently blocking in the kernel will have no knowledge
		3941	about the newly added timer. By waking up the loop it will pick up any new
		3942	watchers in the next event loop iteration.
		3943
		3944	=head2 THREADS, COROUTINES, CONTINUATIONS, QUEUES... INSTEAD OF CALLBACKS
		3945
		3946	While the overhead of a callback that e.g. schedules a thread is small, it
		3947	is still an overhead. If you embed libev, and your main usage is with some
		3948	kind of threads or coroutines, you might want to customise libev so that
		3949	doesn't need callbacks anymore.
		3950
		3951	Imagine you have coroutines that you can switch to using a function
		3952	C<switch_to (coro)>, that libev runs in a coroutine called C<libev_coro>
		3953	and that due to some magic, the currently active coroutine is stored in a
		3954	global called C<current_coro>. Then you can build your own "wait for libev
		3955	event" primitive by changing C<EV_CB_DECLARE> and C<EV_CB_INVOKE> (note
		3956	the differing C<;> conventions):
		3957
		3958	#define EV_CB_DECLARE(type) struct my_coro *cb;
		3959	#define EV_CB_INVOKE(watcher) switch_to ((watcher)->cb)
		3960
		3961	That means instead of having a C callback function, you store the
		3962	coroutine to switch to in each watcher, and instead of having libev call
		3963	your callback, you instead have it switch to that coroutine.
		3964
		3965	A coroutine might now wait for an event with a function called
		3966	C<wait_for_event>. (the watcher needs to be started, as always, but it doesn't
		3967	matter when, or whether the watcher is active or not when this function is
		3968	called):
		3969
		3970	void
		3971	wait_for_event (ev_watcher *w)
		3972	{
		3973	ev_set_cb (w, current_coro);
		3974	switch_to (libev_coro);
		3975	}
		3976
		3977	That basically suspends the coroutine inside C<wait_for_event> and
		3978	continues the libev coroutine, which, when appropriate, switches back to
		3979	this or any other coroutine.
		3980
		3981	You can do similar tricks if you have, say, threads with an event queue -
		3982	instead of storing a coroutine, you store the queue object and instead of
		3983	switching to a coroutine, you push the watcher onto the queue and notify
		3984	any waiters.
		3985
		3986	To embed libev, see L</EMBEDDING>, but in short, it's easiest to create two
		3987	files, F<my_ev.h> and F<my_ev.c> that include the respective libev files:
		3988
		3989	// my_ev.h
		3990	#define EV_CB_DECLARE(type) struct my_coro *cb;
		3991	#define EV_CB_INVOKE(watcher) switch_to ((watcher)->cb)
		3992	#include "../libev/ev.h"
		3993
		3994	// my_ev.c
		3995	#define EV_H "my_ev.h"
		3996	#include "../libev/ev.c"
		3997
		3998	And then use F<my_ev.h> when you would normally use F<ev.h>, and compile
		3999	F<my_ev.c> into your project. When properly specifying include paths, you
		4000	can even use F<ev.h> as header file name directly.
3302		4001
3303		4002
3304	=head1 LIBEVENT EMULATION	4003	=head1 LIBEVENT EMULATION
3305		4004
3306	Libev offers a compatibility emulation layer for libevent. It cannot	4005	Libev offers a compatibility emulation layer for libevent. It cannot
3307	emulate the internals of libevent, so here are some usage hints:	4006	emulate the internals of libevent, so here are some usage hints:
3308		4007
3309	=over 4	4008	=over 4
		4009
		4010	=item * Only the libevent-1.4.1-beta API is being emulated.
		4011
		4012	This was the newest libevent version available when libev was implemented,
		4013	and is still mostly unchanged in 2010.
3310		4014
3311	=item * Use it by including <event.h>, as usual.	4015	=item * Use it by including <event.h>, as usual.
3312		4016
3313	=item * The following members are fully supported: ev_base, ev_callback,	4017	=item * The following members are fully supported: ev_base, ev_callback,
3314	ev_arg, ev_fd, ev_res, ev_events.	4018	ev_arg, ev_fd, ev_res, ev_events.
…		…
3320	=item * Priorities are not currently supported. Initialising priorities	4024	=item * Priorities are not currently supported. Initialising priorities
3321	will fail and all watchers will have the same priority, even though there	4025	will fail and all watchers will have the same priority, even though there
3322	is an ev_pri field.	4026	is an ev_pri field.
3323		4027
3324	=item * In libevent, the last base created gets the signals, in libev, the	4028	=item * In libevent, the last base created gets the signals, in libev, the
3325	first base created (== the default loop) gets the signals.	4029	base that registered the signal gets the signals.
3326		4030
3327	=item * Other members are not supported.	4031	=item * Other members are not supported.
3328		4032
3329	=item * The libev emulation is I<not> ABI compatible to libevent, you need	4033	=item * The libev emulation is I<not> ABI compatible to libevent, you need
3330	to use the libev header file and library.	4034	to use the libev header file and library.
3331		4035
3332	=back	4036	=back
3333		4037
3334	=head1 C++ SUPPORT	4038	=head1 C++ SUPPORT
		4039
		4040	=head2 C API
		4041
		4042	The normal C API should work fine when used from C++: both ev.h and the
		4043	libev sources can be compiled as C++. Therefore, code that uses the C API
		4044	will work fine.
		4045
		4046	Proper exception specifications might have to be added to callbacks passed
		4047	to libev: exceptions may be thrown only from watcher callbacks, all other
		4048	callbacks (allocator, syserr, loop acquire/release and periodic reschedule
		4049	callbacks) must not throw exceptions, and might need a C<noexcept>
		4050	specification. If you have code that needs to be compiled as both C and
		4051	C++ you can use the C<EV_NOEXCEPT> macro for this:
		4052
		4053	static void
		4054	fatal_error (const char *msg) EV_NOEXCEPT
		4055	{
		4056	perror (msg);
		4057	abort ();
		4058	}
		4059
		4060	...
		4061	ev_set_syserr_cb (fatal_error);
		4062
		4063	The only API functions that can currently throw exceptions are C<ev_run>,
		4064	C<ev_invoke>, C<ev_invoke_pending> and C<ev_loop_destroy> (the latter
		4065	because it runs cleanup watchers).
		4066
		4067	Throwing exceptions in watcher callbacks is only supported if libev itself
		4068	is compiled with a C++ compiler or your C and C++ environments allow
		4069	throwing exceptions through C libraries (most do).
		4070
		4071	=head2 C++ API
3335		4072
3336	Libev comes with some simplistic wrapper classes for C++ that mainly allow	4073	Libev comes with some simplistic wrapper classes for C++ that mainly allow
3337	you to use some convenience methods to start/stop watchers and also change	4074	you to use some convenience methods to start/stop watchers and also change
3338	the callback model to a model using method callbacks on objects.	4075	the callback model to a model using method callbacks on objects.
3339		4076
3340	To use it,	4077	To use it,
3341		4078
3342	#include <ev++.h>	4079	#include <ev++.h>
3343		4080
3344	This automatically includes F<ev.h> and puts all of its definitions (many	4081	This automatically includes F<ev.h> and puts all of its definitions (many
3345	of them macros) into the global namespace. All C++ specific things are	4082	of them macros) into the global namespace. All C++ specific things are
3346	put into the C<ev> namespace. It should support all the same embedding	4083	put into the C<ev> namespace. It should support all the same embedding
…		…
3349	Care has been taken to keep the overhead low. The only data member the C++	4086	Care has been taken to keep the overhead low. The only data member the C++
3350	classes add (compared to plain C-style watchers) is the event loop pointer	4087	classes add (compared to plain C-style watchers) is the event loop pointer
3351	that the watcher is associated with (or no additional members at all if	4088	that the watcher is associated with (or no additional members at all if
3352	you disable C<EV_MULTIPLICITY> when embedding libev).	4089	you disable C<EV_MULTIPLICITY> when embedding libev).
3353		4090
3354	Currently, functions, and static and non-static member functions can be	4091	Currently, functions, static and non-static member functions and classes
3355	used as callbacks. Other types should be easy to add as long as they only	4092	with C<operator ()> can be used as callbacks. Other types should be easy
3356	need one additional pointer for context. If you need support for other	4093	to add as long as they only need one additional pointer for context. If
3357	types of functors please contact the author (preferably after implementing	4094	you need support for other types of functors please contact the author
3358	it).	4095	(preferably after implementing it).
		4096
		4097	For all this to work, your C++ compiler either has to use the same calling
		4098	conventions as your C compiler (for static member functions), or you have
		4099	to embed libev and compile libev itself as C++.
3359		4100
3360	Here is a list of things available in the C<ev> namespace:	4101	Here is a list of things available in the C<ev> namespace:
3361		4102
3362	=over 4	4103	=over 4
3363		4104
…		…
3373	=item C<ev::io>, C<ev::timer>, C<ev::periodic>, C<ev::idle>, C<ev::sig> etc.	4114	=item C<ev::io>, C<ev::timer>, C<ev::periodic>, C<ev::idle>, C<ev::sig> etc.
3374		4115
3375	For each C<ev_TYPE> watcher in F<ev.h> there is a corresponding class of	4116	For each C<ev_TYPE> watcher in F<ev.h> there is a corresponding class of
3376	the same name in the C<ev> namespace, with the exception of C<ev_signal>	4117	the same name in the C<ev> namespace, with the exception of C<ev_signal>
3377	which is called C<ev::sig> to avoid clashes with the C<signal> macro	4118	which is called C<ev::sig> to avoid clashes with the C<signal> macro
3378	defines by many implementations.	4119	defined by many implementations.
3379		4120
3380	All of those classes have these methods:	4121	All of those classes have these methods:
3381		4122
3382	=over 4	4123	=over 4
3383		4124
…		…
3445	void operator() (ev::io &w, int revents)	4186	void operator() (ev::io &w, int revents)
3446	{	4187	{
3447	...	4188	...
3448	}	4189	}
3449	}	4190	}
3450		4191
3451	myfunctor f;	4192	myfunctor f;
3452		4193
3453	ev::io w;	4194	ev::io w;
3454	w.set (&f);	4195	w.set (&f);
3455		4196
…		…
3473	Associates a different C<struct ev_loop> with this watcher. You can only	4214	Associates a different C<struct ev_loop> with this watcher. You can only
3474	do this when the watcher is inactive (and not pending either).	4215	do this when the watcher is inactive (and not pending either).
3475		4216
3476	=item w->set ([arguments])	4217	=item w->set ([arguments])
3477		4218
3478	Basically the same as C<ev_TYPE_set>, with the same arguments. Either this	4219	Basically the same as C<ev_TYPE_set> (except for C<ev::embed> watchers>),
3479	method or a suitable start method must be called at least once. Unlike the	4220	with the same arguments. Either this method or a suitable start method
3480	C counterpart, an active watcher gets automatically stopped and restarted	4221	must be called at least once. Unlike the C counterpart, an active watcher
3481	when reconfiguring it with this method.	4222	gets automatically stopped and restarted when reconfiguring it with this
		4223	method.
		4224
		4225	For C<ev::embed> watchers this method is called C<set_embed>, to avoid
		4226	clashing with the C<set (loop)> method.
3482		4227
3483	=item w->start ()	4228	=item w->start ()
3484		4229
3485	Starts the watcher. Note that there is no C<loop> argument, as the	4230	Starts the watcher. Note that there is no C<loop> argument, as the
3486	constructor already stores the event loop.	4231	constructor already stores the event loop.
…		…
3516	watchers in the constructor.	4261	watchers in the constructor.
3517		4262
3518	class myclass	4263	class myclass
3519	{	4264	{
3520	ev::io io ; void io_cb (ev::io &w, int revents);	4265	ev::io io ; void io_cb (ev::io &w, int revents);
3521	ev::io2 io2 ; void io2_cb (ev::io &w, int revents);	4266	ev::io io2 ; void io2_cb (ev::io &w, int revents);
3522	ev::idle idle; void idle_cb (ev::idle &w, int revents);	4267	ev::idle idle; void idle_cb (ev::idle &w, int revents);
3523		4268
3524	myclass (int fd)	4269	myclass (int fd)
3525	{	4270	{
3526	io .set <myclass, &myclass::io_cb > (this);	4271	io .set <myclass, &myclass::io_cb > (this);
…		…
3577	L<http://hackage.haskell.org/cgi-bin/hackage-scripts/package/hlibev>.	4322	L<http://hackage.haskell.org/cgi-bin/hackage-scripts/package/hlibev>.
3578		4323
3579	=item D	4324	=item D
3580		4325
3581	Leandro Lucarella has written a D language binding (F<ev.d>) for libev, to	4326	Leandro Lucarella has written a D language binding (F<ev.d>) for libev, to
3582	be found at L<http://proj.llucax.com.ar/wiki/evd>.	4327	be found at L<http://www.llucax.com.ar/proj/ev.d/index.html>.
3583		4328
3584	=item Ocaml	4329	=item Ocaml
3585		4330
3586	Erkki Seppala has written Ocaml bindings for libev, to be found at	4331	Erkki Seppala has written Ocaml bindings for libev, to be found at
3587	L<http://modeemi.cs.tut.fi/~flux/software/ocaml-ev/>.	4332	L<http://modeemi.cs.tut.fi/~flux/software/ocaml-ev/>.
…		…
3590		4335
3591	Brian Maher has written a partial interface to libev for lua (at the	4336	Brian Maher has written a partial interface to libev for lua (at the
3592	time of this writing, only C<ev_io> and C<ev_timer>), to be found at	4337	time of this writing, only C<ev_io> and C<ev_timer>), to be found at
3593	L<http://github.com/brimworks/lua-ev>.	4338	L<http://github.com/brimworks/lua-ev>.
3594		4339
		4340	=item Javascript
		4341
		4342	Node.js (L<http://nodejs.org>) uses libev as the underlying event library.
		4343
		4344	=item Others
		4345
		4346	There are others, and I stopped counting.
		4347
3595	=back	4348	=back
3596		4349
3597		4350
3598	=head1 MACRO MAGIC	4351	=head1 MACRO MAGIC
3599		4352
…		…
3635	suitable for use with C<EV_A>.	4388	suitable for use with C<EV_A>.
3636		4389
3637	=item C<EV_DEFAULT>, C<EV_DEFAULT_>	4390	=item C<EV_DEFAULT>, C<EV_DEFAULT_>
3638		4391
3639	Similar to the other two macros, this gives you the value of the default	4392	Similar to the other two macros, this gives you the value of the default
3640	loop, if multiple loops are supported ("ev loop default").	4393	loop, if multiple loops are supported ("ev loop default"). The default loop
		4394	will be initialised if it isn't already initialised.
		4395
		4396	For non-multiplicity builds, these macros do nothing, so you always have
		4397	to initialise the loop somewhere.
3641		4398
3642	=item C<EV_DEFAULT_UC>, C<EV_DEFAULT_UC_>	4399	=item C<EV_DEFAULT_UC>, C<EV_DEFAULT_UC_>
3643		4400
3644	Usage identical to C<EV_DEFAULT> and C<EV_DEFAULT_>, but requires that the	4401	Usage identical to C<EV_DEFAULT> and C<EV_DEFAULT_>, but requires that the
3645	default loop has been initialised (C<UC> == unchecked). Their behaviour	4402	default loop has been initialised (C<UC> == unchecked). Their behaviour
…		…
3712	ev_vars.h	4469	ev_vars.h
3713	ev_wrap.h	4470	ev_wrap.h
3714		4471
3715	ev_win32.c required on win32 platforms only	4472	ev_win32.c required on win32 platforms only
3716		4473
3717	ev_select.c only when select backend is enabled (which is enabled by default)	4474	ev_select.c only when select backend is enabled
3718	ev_poll.c only when poll backend is enabled (disabled by default)	4475	ev_poll.c only when poll backend is enabled
3719	ev_epoll.c only when the epoll backend is enabled (disabled by default)	4476	ev_epoll.c only when the epoll backend is enabled
		4477	ev_linuxaio.c only when the linux aio backend is enabled
3720	ev_kqueue.c only when the kqueue backend is enabled (disabled by default)	4478	ev_kqueue.c only when the kqueue backend is enabled
3721	ev_port.c only when the solaris port backend is enabled (disabled by default)	4479	ev_port.c only when the solaris port backend is enabled
3722		4480
3723	F<ev.c> includes the backend files directly when enabled, so you only need	4481	F<ev.c> includes the backend files directly when enabled, so you only need
3724	to compile this single file.	4482	to compile this single file.
3725		4483
3726	=head3 LIBEVENT COMPATIBILITY API	4484	=head3 LIBEVENT COMPATIBILITY API
…		…
3790	supported). It will also not define any of the structs usually found in	4548	supported). It will also not define any of the structs usually found in
3791	F<event.h> that are not directly supported by the libev core alone.	4549	F<event.h> that are not directly supported by the libev core alone.
3792		4550
3793	In standalone mode, libev will still try to automatically deduce the	4551	In standalone mode, libev will still try to automatically deduce the
3794	configuration, but has to be more conservative.	4552	configuration, but has to be more conservative.
		4553
		4554	=item EV_USE_FLOOR
		4555
		4556	If defined to be C<1>, libev will use the C<floor ()> function for its
		4557	periodic reschedule calculations, otherwise libev will fall back on a
		4558	portable (slower) implementation. If you enable this, you usually have to
		4559	link against libm or something equivalent. Enabling this when the C<floor>
		4560	function is not available will fail, so the safe default is to not enable
		4561	this.
3795		4562
3796	=item EV_USE_MONOTONIC	4563	=item EV_USE_MONOTONIC
3797		4564
3798	If defined to be C<1>, libev will try to detect the availability of the	4565	If defined to be C<1>, libev will try to detect the availability of the
3799	monotonic clock option at both compile time and runtime. Otherwise no	4566	monotonic clock option at both compile time and runtime. Otherwise no
…		…
3885	If programs implement their own fd to handle mapping on win32, then this	4652	If programs implement their own fd to handle mapping on win32, then this
3886	macro can be used to override the C<close> function, useful to unregister	4653	macro can be used to override the C<close> function, useful to unregister
3887	file descriptors again. Note that the replacement function has to close	4654	file descriptors again. Note that the replacement function has to close
3888	the underlying OS handle.	4655	the underlying OS handle.
3889		4656
		4657	=item EV_USE_WSASOCKET
		4658
		4659	If defined to be C<1>, libev will use C<WSASocket> to create its internal
		4660	communication socket, which works better in some environments. Otherwise,
		4661	the normal C<socket> function will be used, which works better in other
		4662	environments.
		4663
3890	=item EV_USE_POLL	4664	=item EV_USE_POLL
3891		4665
3892	If defined to be C<1>, libev will compile in support for the C<poll>(2)	4666	If defined to be C<1>, libev will compile in support for the C<poll>(2)
3893	backend. Otherwise it will be enabled on non-win32 platforms. It	4667	backend. Otherwise it will be enabled on non-win32 platforms. It
3894	takes precedence over select.	4668	takes precedence over select.
…		…
3898	If defined to be C<1>, libev will compile in support for the Linux	4672	If defined to be C<1>, libev will compile in support for the Linux
3899	C<epoll>(7) backend. Its availability will be detected at runtime,	4673	C<epoll>(7) backend. Its availability will be detected at runtime,
3900	otherwise another method will be used as fallback. This is the preferred	4674	otherwise another method will be used as fallback. This is the preferred
3901	backend for GNU/Linux systems. If undefined, it will be enabled if the	4675	backend for GNU/Linux systems. If undefined, it will be enabled if the
3902	headers indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.	4676	headers indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.
		4677
		4678	=item EV_USE_LINUXAIO
		4679
		4680	If defined to be C<1>, libev will compile in support for the Linux
		4681	aio backend. Due to it's currenbt limitations it has to be requested
		4682	explicitly. If undefined, it will be enabled on linux, otherwise
		4683	disabled.
3903		4684
3904	=item EV_USE_KQUEUE	4685	=item EV_USE_KQUEUE
3905		4686
3906	If defined to be C<1>, libev will compile in support for the BSD style	4687	If defined to be C<1>, libev will compile in support for the BSD style
3907	C<kqueue>(2) backend. Its actual availability will be detected at runtime,	4688	C<kqueue>(2) backend. Its actual availability will be detected at runtime,
…		…
3929	If defined to be C<1>, libev will compile in support for the Linux inotify	4710	If defined to be C<1>, libev will compile in support for the Linux inotify
3930	interface to speed up C<ev_stat> watchers. Its actual availability will	4711	interface to speed up C<ev_stat> watchers. Its actual availability will
3931	be detected at runtime. If undefined, it will be enabled if the headers	4712	be detected at runtime. If undefined, it will be enabled if the headers
3932	indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.	4713	indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.
3933		4714
		4715	=item EV_NO_SMP
		4716
		4717	If defined to be C<1>, libev will assume that memory is always coherent
		4718	between threads, that is, threads can be used, but threads never run on
		4719	different cpus (or different cpu cores). This reduces dependencies
		4720	and makes libev faster.
		4721
		4722	=item EV_NO_THREADS
		4723
		4724	If defined to be C<1>, libev will assume that it will never be called from
		4725	different threads (that includes signal handlers), which is a stronger
		4726	assumption than C<EV_NO_SMP>, above. This reduces dependencies and makes
		4727	libev faster.
		4728
3934	=item EV_ATOMIC_T	4729	=item EV_ATOMIC_T
3935		4730
3936	Libev requires an integer type (suitable for storing C<0> or C<1>) whose	4731	Libev requires an integer type (suitable for storing C<0> or C<1>) whose
3937	access is atomic with respect to other threads or signal contexts. No such	4732	access is atomic with respect to other threads or signal contexts. No
3938	type is easily found in the C language, so you can provide your own type	4733	such type is easily found in the C language, so you can provide your own
3939	that you know is safe for your purposes. It is used both for signal handler "locking"	4734	type that you know is safe for your purposes. It is used both for signal
3940	as well as for signal and thread safety in C<ev_async> watchers.	4735	handler "locking" as well as for signal and thread safety in C<ev_async>
		4736	watchers.
3941		4737
3942	In the absence of this define, libev will use C<sig_atomic_t volatile>	4738	In the absence of this define, libev will use C<sig_atomic_t volatile>
3943	(from F<signal.h>), which is usually good enough on most platforms.	4739	(from F<signal.h>), which is usually good enough on most platforms.
3944		4740
3945	=item EV_H (h)	4741	=item EV_H (h)
…		…
3972	will have the C<struct ev_loop *> as first argument, and you can create	4768	will have the C<struct ev_loop *> as first argument, and you can create
3973	additional independent event loops. Otherwise there will be no support	4769	additional independent event loops. Otherwise there will be no support
3974	for multiple event loops and there is no first event loop pointer	4770	for multiple event loops and there is no first event loop pointer
3975	argument. Instead, all functions act on the single default loop.	4771	argument. Instead, all functions act on the single default loop.
3976		4772
		4773	Note that C<EV_DEFAULT> and C<EV_DEFAULT_> will no longer provide a
		4774	default loop when multiplicity is switched off - you always have to
		4775	initialise the loop manually in this case.
		4776
3977	=item EV_MINPRI	4777	=item EV_MINPRI
3978		4778
3979	=item EV_MAXPRI	4779	=item EV_MAXPRI
3980		4780
3981	The range of allowed priorities. C<EV_MINPRI> must be smaller or equal to	4781	The range of allowed priorities. C<EV_MINPRI> must be smaller or equal to
…		…
4017	#define EV_USE_POLL 1	4817	#define EV_USE_POLL 1
4018	#define EV_CHILD_ENABLE 1	4818	#define EV_CHILD_ENABLE 1
4019	#define EV_ASYNC_ENABLE 1	4819	#define EV_ASYNC_ENABLE 1
4020		4820
4021	The actual value is a bitset, it can be a combination of the following	4821	The actual value is a bitset, it can be a combination of the following
4022	values:	4822	values (by default, all of these are enabled):
4023		4823
4024	=over 4	4824	=over 4
4025		4825
4026	=item C<1> - faster/larger code	4826	=item C<1> - faster/larger code
4027		4827
…		…
4031	code size by roughly 30% on amd64).	4831	code size by roughly 30% on amd64).
4032		4832
4033	When optimising for size, use of compiler flags such as C<-Os> with	4833	When optimising for size, use of compiler flags such as C<-Os> with
4034	gcc is recommended, as well as C<-DNDEBUG>, as libev contains a number of	4834	gcc is recommended, as well as C<-DNDEBUG>, as libev contains a number of
4035	assertions.	4835	assertions.
		4836
		4837	The default is off when C<__OPTIMIZE_SIZE__> is defined by your compiler
		4838	(e.g. gcc with C<-Os>).
4036		4839
4037	=item C<2> - faster/larger data structures	4840	=item C<2> - faster/larger data structures
4038		4841
4039	Replaces the small 2-heap for timer management by a faster 4-heap, larger	4842	Replaces the small 2-heap for timer management by a faster 4-heap, larger
4040	hash table sizes and so on. This will usually further increase code size	4843	hash table sizes and so on. This will usually further increase code size
4041	and can additionally have an effect on the size of data structures at	4844	and can additionally have an effect on the size of data structures at
4042	runtime.	4845	runtime.
4043		4846
		4847	The default is off when C<__OPTIMIZE_SIZE__> is defined by your compiler
		4848	(e.g. gcc with C<-Os>).
		4849
4044	=item C<4> - full API configuration	4850	=item C<4> - full API configuration
4045		4851
4046	This enables priorities (sets C<EV_MAXPRI>=2 and C<EV_MINPRI>=-2), and	4852	This enables priorities (sets C<EV_MAXPRI>=2 and C<EV_MINPRI>=-2), and
4047	enables multiplicity (C<EV_MULTIPLICITY>=1).	4853	enables multiplicity (C<EV_MULTIPLICITY>=1).
4048		4854
…		…
4078		4884
4079	With an intelligent-enough linker (gcc+binutils are intelligent enough	4885	With an intelligent-enough linker (gcc+binutils are intelligent enough
4080	when you use C<-Wl,--gc-sections -ffunction-sections>) functions unused by	4886	when you use C<-Wl,--gc-sections -ffunction-sections>) functions unused by
4081	your program might be left out as well - a binary starting a timer and an	4887	your program might be left out as well - a binary starting a timer and an
4082	I/O watcher then might come out at only 5Kb.	4888	I/O watcher then might come out at only 5Kb.
		4889
		4890	=item EV_API_STATIC
		4891
		4892	If this symbol is defined (by default it is not), then all identifiers
		4893	will have static linkage. This means that libev will not export any
		4894	identifiers, and you cannot link against libev anymore. This can be useful
		4895	when you embed libev, only want to use libev functions in a single file,
		4896	and do not want its identifiers to be visible.
		4897
		4898	To use this, define C<EV_API_STATIC> and include F<ev.c> in the file that
		4899	wants to use libev.
		4900
		4901	This option only works when libev is compiled with a C compiler, as C++
		4902	doesn't support the required declaration syntax.
4083		4903
4084	=item EV_AVOID_STDIO	4904	=item EV_AVOID_STDIO
4085		4905
4086	If this is set to C<1> at compiletime, then libev will avoid using stdio	4906	If this is set to C<1> at compiletime, then libev will avoid using stdio
4087	functions (printf, scanf, perror etc.). This will increase the code size	4907	functions (printf, scanf, perror etc.). This will increase the code size
…		…
4231	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:	5051	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:
4232		5052
4233	#include "ev_cpp.h"	5053	#include "ev_cpp.h"
4234	#include "ev.c"	5054	#include "ev.c"
4235		5055
4236	=head1 INTERACTION WITH OTHER PROGRAMS OR LIBRARIES	5056	=head1 INTERACTION WITH OTHER PROGRAMS, LIBRARIES OR THE ENVIRONMENT
4237		5057
4238	=head2 THREADS AND COROUTINES	5058	=head2 THREADS AND COROUTINES
4239		5059
4240	=head3 THREADS	5060	=head3 THREADS
4241		5061
…		…
4292	default loop and triggering an C<ev_async> watcher from the default loop	5112	default loop and triggering an C<ev_async> watcher from the default loop
4293	watcher callback into the event loop interested in the signal.	5113	watcher callback into the event loop interested in the signal.
4294		5114
4295	=back	5115	=back
4296		5116
4297	=head4 THREAD LOCKING EXAMPLE	5117	See also L</THREAD LOCKING EXAMPLE>.
4298
4299	Here is a fictitious example of how to run an event loop in a different
4300	thread than where callbacks are being invoked and watchers are
4301	created/added/removed.
4302
4303	For a real-world example, see the C<EV::Loop::Async> perl module,
4304	which uses exactly this technique (which is suited for many high-level
4305	languages).
4306
4307	The example uses a pthread mutex to protect the loop data, a condition
4308	variable to wait for callback invocations, an async watcher to notify the
4309	event loop thread and an unspecified mechanism to wake up the main thread.
4310
4311	First, you need to associate some data with the event loop:
4312
4313	typedef struct {
4314	mutex_t lock; /* global loop lock */
4315	ev_async async_w;
4316	thread_t tid;
4317	cond_t invoke_cv;
4318	} userdata;
4319
4320	void prepare_loop (EV_P)
4321	{
4322	// for simplicity, we use a static userdata struct.
4323	static userdata u;
4324
4325	ev_async_init (&u->async_w, async_cb);
4326	ev_async_start (EV_A_ &u->async_w);
4327
4328	pthread_mutex_init (&u->lock, 0);
4329	pthread_cond_init (&u->invoke_cv, 0);
4330
4331	// now associate this with the loop
4332	ev_set_userdata (EV_A_ u);
4333	ev_set_invoke_pending_cb (EV_A_ l_invoke);
4334	ev_set_loop_release_cb (EV_A_ l_release, l_acquire);
4335
4336	// then create the thread running ev_loop
4337	pthread_create (&u->tid, 0, l_run, EV_A);
4338	}
4339
4340	The callback for the C<ev_async> watcher does nothing: the watcher is used
4341	solely to wake up the event loop so it takes notice of any new watchers
4342	that might have been added:
4343
4344	static void
4345	async_cb (EV_P_ ev_async *w, int revents)
4346	{
4347	// just used for the side effects
4348	}
4349
4350	The C<l_release> and C<l_acquire> callbacks simply unlock/lock the mutex
4351	protecting the loop data, respectively.
4352
4353	static void
4354	l_release (EV_P)
4355	{
4356	userdata *u = ev_userdata (EV_A);
4357	pthread_mutex_unlock (&u->lock);
4358	}
4359
4360	static void
4361	l_acquire (EV_P)
4362	{
4363	userdata *u = ev_userdata (EV_A);
4364	pthread_mutex_lock (&u->lock);
4365	}
4366
4367	The event loop thread first acquires the mutex, and then jumps straight
4368	into C<ev_run>:
4369
4370	void *
4371	l_run (void *thr_arg)
4372	{
4373	struct ev_loop *loop = (struct ev_loop *)thr_arg;
4374
4375	l_acquire (EV_A);
4376	pthread_setcanceltype (PTHREAD_CANCEL_ASYNCHRONOUS, 0);
4377	ev_run (EV_A_ 0);
4378	l_release (EV_A);
4379
4380	return 0;
4381	}
4382
4383	Instead of invoking all pending watchers, the C<l_invoke> callback will
4384	signal the main thread via some unspecified mechanism (signals? pipe
4385	writes? C<Async::Interrupt>?) and then waits until all pending watchers
4386	have been called (in a while loop because a) spurious wakeups are possible
4387	and b) skipping inter-thread-communication when there are no pending
4388	watchers is very beneficial):
4389
4390	static void
4391	l_invoke (EV_P)
4392	{
4393	userdata *u = ev_userdata (EV_A);
4394
4395	while (ev_pending_count (EV_A))
4396	{
4397	wake_up_other_thread_in_some_magic_or_not_so_magic_way ();
4398	pthread_cond_wait (&u->invoke_cv, &u->lock);
4399	}
4400	}
4401
4402	Now, whenever the main thread gets told to invoke pending watchers, it
4403	will grab the lock, call C<ev_invoke_pending> and then signal the loop
4404	thread to continue:
4405
4406	static void
4407	real_invoke_pending (EV_P)
4408	{
4409	userdata *u = ev_userdata (EV_A);
4410
4411	pthread_mutex_lock (&u->lock);
4412	ev_invoke_pending (EV_A);
4413	pthread_cond_signal (&u->invoke_cv);
4414	pthread_mutex_unlock (&u->lock);
4415	}
4416
4417	Whenever you want to start/stop a watcher or do other modifications to an
4418	event loop, you will now have to lock:
4419
4420	ev_timer timeout_watcher;
4421	userdata *u = ev_userdata (EV_A);
4422
4423	ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);
4424
4425	pthread_mutex_lock (&u->lock);
4426	ev_timer_start (EV_A_ &timeout_watcher);
4427	ev_async_send (EV_A_ &u->async_w);
4428	pthread_mutex_unlock (&u->lock);
4429
4430	Note that sending the C<ev_async> watcher is required because otherwise
4431	an event loop currently blocking in the kernel will have no knowledge
4432	about the newly added timer. By waking up the loop it will pick up any new
4433	watchers in the next event loop iteration.
4434		5118
4435	=head3 COROUTINES	5119	=head3 COROUTINES
4436		5120
4437	Libev is very accommodating to coroutines ("cooperative threads"):	5121	Libev is very accommodating to coroutines ("cooperative threads"):
4438	libev fully supports nesting calls to its functions from different	5122	libev fully supports nesting calls to its functions from different
…		…
4603	requires, and its I/O model is fundamentally incompatible with the POSIX	5287	requires, and its I/O model is fundamentally incompatible with the POSIX
4604	model. Libev still offers limited functionality on this platform in	5288	model. Libev still offers limited functionality on this platform in
4605	the form of the C<EVBACKEND_SELECT> backend, and only supports socket	5289	the form of the C<EVBACKEND_SELECT> backend, and only supports socket
4606	descriptors. This only applies when using Win32 natively, not when using	5290	descriptors. This only applies when using Win32 natively, not when using
4607	e.g. cygwin. Actually, it only applies to the microsofts own compilers,	5291	e.g. cygwin. Actually, it only applies to the microsofts own compilers,
4608	as every compielr comes with a slightly differently broken/incompatible	5292	as every compiler comes with a slightly differently broken/incompatible
4609	environment.	5293	environment.
4610		5294
4611	Lifting these limitations would basically require the full	5295	Lifting these limitations would basically require the full
4612	re-implementation of the I/O system. If you are into this kind of thing,	5296	re-implementation of the I/O system. If you are into this kind of thing,
4613	then note that glib does exactly that for you in a very portable way (note	5297	then note that glib does exactly that for you in a very portable way (note
…		…
4707	structure (guaranteed by POSIX but not by ISO C for example), but it also	5391	structure (guaranteed by POSIX but not by ISO C for example), but it also
4708	assumes that the same (machine) code can be used to call any watcher	5392	assumes that the same (machine) code can be used to call any watcher
4709	callback: The watcher callbacks have different type signatures, but libev	5393	callback: The watcher callbacks have different type signatures, but libev
4710	calls them using an C<ev_watcher *> internally.	5394	calls them using an C<ev_watcher *> internally.
4711		5395
		5396	=item null pointers and integer zero are represented by 0 bytes
		5397
		5398	Libev uses C<memset> to initialise structs and arrays to C<0> bytes, and
		5399	relies on this setting pointers and integers to null.
		5400
		5401	=item pointer accesses must be thread-atomic
		5402
		5403	Accessing a pointer value must be atomic, it must both be readable and
		5404	writable in one piece - this is the case on all current architectures.
		5405
4712	=item C<sig_atomic_t volatile> must be thread-atomic as well	5406	=item C<sig_atomic_t volatile> must be thread-atomic as well
4713		5407
4714	The type C<sig_atomic_t volatile> (or whatever is defined as	5408	The type C<sig_atomic_t volatile> (or whatever is defined as
4715	C<EV_ATOMIC_T>) must be atomic with respect to accesses from different	5409	C<EV_ATOMIC_T>) must be atomic with respect to accesses from different
4716	threads. This is not part of the specification for C<sig_atomic_t>, but is	5410	threads. This is not part of the specification for C<sig_atomic_t>, but is
…		…
4724	thread" or will block signals process-wide, both behaviours would	5418	thread" or will block signals process-wide, both behaviours would
4725	be compatible with libev. Interaction between C<sigprocmask> and	5419	be compatible with libev. Interaction between C<sigprocmask> and
4726	C<pthread_sigmask> could complicate things, however.	5420	C<pthread_sigmask> could complicate things, however.
4727		5421
4728	The most portable way to handle signals is to block signals in all threads	5422	The most portable way to handle signals is to block signals in all threads
4729	except the initial one, and run the default loop in the initial thread as	5423	except the initial one, and run the signal handling loop in the initial
4730	well.	5424	thread as well.
4731		5425
4732	=item C<long> must be large enough for common memory allocation sizes	5426	=item C<long> must be large enough for common memory allocation sizes
4733		5427
4734	To improve portability and simplify its API, libev uses C<long> internally	5428	To improve portability and simplify its API, libev uses C<long> internally
4735	instead of C<size_t> when allocating its data structures. On non-POSIX	5429	instead of C<size_t> when allocating its data structures. On non-POSIX
…		…
4741		5435
4742	The type C<double> is used to represent timestamps. It is required to	5436	The type C<double> is used to represent timestamps. It is required to
4743	have at least 51 bits of mantissa (and 9 bits of exponent), which is	5437	have at least 51 bits of mantissa (and 9 bits of exponent), which is
4744	good enough for at least into the year 4000 with millisecond accuracy	5438	good enough for at least into the year 4000 with millisecond accuracy
4745	(the design goal for libev). This requirement is overfulfilled by	5439	(the design goal for libev). This requirement is overfulfilled by
4746	implementations using IEEE 754, which is basically all existing ones. With	5440	implementations using IEEE 754, which is basically all existing ones.
		5441
4747	IEEE 754 doubles, you get microsecond accuracy until at least 2200.	5442	With IEEE 754 doubles, you get microsecond accuracy until at least the
		5443	year 2255 (and millisecond accuracy till the year 287396 - by then, libev
		5444	is either obsolete or somebody patched it to use C<long double> or
		5445	something like that, just kidding).
4748		5446
4749	=back	5447	=back
4750		5448
4751	If you know of other additional requirements drop me a note.	5449	If you know of other additional requirements drop me a note.
4752		5450
…		…
4814	=item Processing ev_async_send: O(number_of_async_watchers)	5512	=item Processing ev_async_send: O(number_of_async_watchers)
4815		5513
4816	=item Processing signals: O(max_signal_number)	5514	=item Processing signals: O(max_signal_number)
4817		5515
4818	Sending involves a system call I<iff> there were no other C<ev_async_send>	5516	Sending involves a system call I<iff> there were no other C<ev_async_send>
4819	calls in the current loop iteration. Checking for async and signal events	5517	calls in the current loop iteration and the loop is currently
		5518	blocked. Checking for async and signal events involves iterating over all
4820	involves iterating over all running async watchers or all signal numbers.	5519	running async watchers or all signal numbers.
4821		5520
4822	=back	5521	=back
4823		5522
4824		5523
4825	=head1 PORTING FROM LIBEV 3.X TO 4.X	5524	=head1 PORTING FROM LIBEV 3.X TO 4.X
4826		5525
4827	The major version 4 introduced some minor incompatible changes to the API.	5526	The major version 4 introduced some incompatible changes to the API.
4828		5527
4829	At the moment, the C<ev.h> header file tries to implement superficial	5528	At the moment, the C<ev.h> header file provides compatibility definitions
4830	compatibility, so most programs should still compile. Those might be	5529	for all changes, so most programs should still compile. The compatibility
4831	removed in later versions of libev, so better update early than late.	5530	layer might be removed in later versions of libev, so better update to the
		5531	new API early than late.
4832		5532
4833	=over 4	5533	=over 4
		5534
		5535	=item C<EV_COMPAT3> backwards compatibility mechanism
		5536
		5537	The backward compatibility mechanism can be controlled by
		5538	C<EV_COMPAT3>. See L</"PREPROCESSOR SYMBOLS/MACROS"> in the L</EMBEDDING>
		5539	section.
		5540
		5541	=item C<ev_default_destroy> and C<ev_default_fork> have been removed
		5542
		5543	These calls can be replaced easily by their C<ev_loop_xxx> counterparts:
		5544
		5545	ev_loop_destroy (EV_DEFAULT_UC);
		5546	ev_loop_fork (EV_DEFAULT);
4834		5547
4835	=item function/symbol renames	5548	=item function/symbol renames
4836		5549
4837	A number of functions and symbols have been renamed:	5550	A number of functions and symbols have been renamed:
4838		5551
…		…
4857	ev_loop> anymore and C<EV_TIMER> now follows the same naming scheme	5570	ev_loop> anymore and C<EV_TIMER> now follows the same naming scheme
4858	as all other watcher types. Note that C<ev_loop_fork> is still called	5571	as all other watcher types. Note that C<ev_loop_fork> is still called
4859	C<ev_loop_fork> because it would otherwise clash with the C<ev_fork>	5572	C<ev_loop_fork> because it would otherwise clash with the C<ev_fork>
4860	typedef.	5573	typedef.
4861		5574
4862	=item C<EV_COMPAT3> backwards compatibility mechanism
4863
4864	The backward compatibility mechanism can be controlled by
4865	C<EV_COMPAT3>. See L<PREPROCESSOR SYMBOLS/MACROS> in the L<EMBEDDING>
4866	section.
4867
4868	=item C<EV_MINIMAL> mechanism replaced by C<EV_FEATURES>	5575	=item C<EV_MINIMAL> mechanism replaced by C<EV_FEATURES>
4869		5576
4870	The preprocessor symbol C<EV_MINIMAL> has been replaced by a different	5577	The preprocessor symbol C<EV_MINIMAL> has been replaced by a different
4871	mechanism, C<EV_FEATURES>. Programs using C<EV_MINIMAL> usually compile	5578	mechanism, C<EV_FEATURES>. Programs using C<EV_MINIMAL> usually compile
4872	and work, but the library code will of course be larger.	5579	and work, but the library code will of course be larger.
…		…
4879	=over 4	5586	=over 4
4880		5587
4881	=item active	5588	=item active
4882		5589
4883	A watcher is active as long as it has been started and not yet stopped.	5590	A watcher is active as long as it has been started and not yet stopped.
4884	See L<WATCHER STATES> for details.	5591	See L</WATCHER STATES> for details.
4885		5592
4886	=item application	5593	=item application
4887		5594
4888	In this document, an application is whatever is using libev.	5595	In this document, an application is whatever is using libev.
4889		5596
…		…
4925	watchers and events.	5632	watchers and events.
4926		5633
4927	=item pending	5634	=item pending
4928		5635
4929	A watcher is pending as soon as the corresponding event has been	5636	A watcher is pending as soon as the corresponding event has been
4930	detected. See L<WATCHER STATES> for details.	5637	detected. See L</WATCHER STATES> for details.
4931		5638
4932	=item real time	5639	=item real time
4933		5640
4934	The physical time that is observed. It is apparently strictly monotonic :)	5641	The physical time that is observed. It is apparently strictly monotonic :)
4935		5642
4936	=item wall-clock time	5643	=item wall-clock time
4937		5644
4938	The time and date as shown on clocks. Unlike real time, it can actually	5645	The time and date as shown on clocks. Unlike real time, it can actually
4939	be wrong and jump forwards and backwards, e.g. when the you adjust your	5646	be wrong and jump forwards and backwards, e.g. when you adjust your
4940	clock.	5647	clock.
4941		5648
4942	=item watcher	5649	=item watcher
4943		5650
4944	A data structure that describes interest in certain events. Watchers need	5651	A data structure that describes interest in certain events. Watchers need
…		…
4946		5653
4947	=back	5654	=back
4948		5655
4949	=head1 AUTHOR	5656	=head1 AUTHOR
4950		5657
4951	Marc Lehmann <libev@schmorp.de>, with repeated corrections by Mikael Magnusson.	5658	Marc Lehmann <libev@schmorp.de>, with repeated corrections by Mikael
		5659	Magnusson and Emanuele Giaquinta, and minor corrections by many others.
4952		5660

Diff Legend

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