ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/libev/ev.pod

Comparing libev/ev.pod (file contents):
Revision 1.351 by root, Mon Jan 10 14:24:26 2011 UTC vs.
Revision 1.469 by root, Sat Jun 3 08:53:03 2023 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
…		…
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
…		…
82		84
83	=head1 WHAT TO READ WHEN IN A HURRY	85	=head1 WHAT TO READ WHEN IN A HURRY
84		86
85	This manual tries to be very detailed, but unfortunately, this also makes	87	This manual tries to be very detailed, but unfortunately, this also makes
86	it very long. If you just want to know the basics of libev, I suggest	88	it very long. If you just want to know the basics of libev, I suggest
87	reading L<ANATOMY OF A WATCHER>, then the L<EXAMPLE PROGRAM> above and	89	reading L</ANATOMY OF A WATCHER>, then the L</EXAMPLE PROGRAM> above and
88	look up the missing functions in L<GLOBAL FUNCTIONS> and the C<ev_io> and	90	look up the missing functions in L</GLOBAL FUNCTIONS> and the C<ev_io> and
89	C<ev_timer> sections in L<WATCHER TYPES>.	91	C<ev_timer> sections in L</WATCHER TYPES>.
90		92
91	=head1 ABOUT LIBEV	93	=head1 ABOUT LIBEV
92		94
93	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
94	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
…		…
103	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
104	watcher.	106	watcher.
105		107
106	=head2 FEATURES	108	=head2 FEATURES
107		109
108	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>
109	BSD-specific C<kqueue> and the Solaris-specific event port mechanisms	111	interfaces, the BSD-specific C<kqueue> and the Solaris-specific event port
110	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>
111	(for C<ev_stat>), Linux eventfd/signalfd (for faster and cleaner	113	interface (for C<ev_stat>), Linux eventfd/signalfd (for faster and cleaner
112	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
113	timers (C<ev_timer>), absolute timers with customised rescheduling	115	timers (C<ev_timer>), absolute timers with customised rescheduling
114	(C<ev_periodic>), synchronous signals (C<ev_signal>), process status	116	(C<ev_periodic>), synchronous signals (C<ev_signal>), process status
115	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
116	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
…		…
157	When libev detects a usage error such as a negative timer interval, then	159	When libev detects a usage error such as a negative timer interval, then
158	it will print a diagnostic message and abort (via the C<assert> mechanism,	160	it will print a diagnostic message and abort (via the C<assert> mechanism,
159	so C<NDEBUG> will disable this checking): these are programming errors in	161	so C<NDEBUG> will disable this checking): these are programming errors in
160	the libev caller and need to be fixed there.	162	the libev caller and need to be fixed there.
161		163
		164	Via the C<EV_FREQUENT> macro you can compile in and/or enable extensive
		165	consistency checking code inside libev that can be used to check for
		166	internal inconsistencies, suually caused by application bugs.
		167
162	Libev also has a few internal error-checking C<assert>ions, and also has	168	Libev also has a few internal error-checking C<assert>ions. These do not
163	extensive consistency checking code. These do not trigger under normal
164	circumstances, as they indicate either a bug in libev or worse.	169	trigger under normal circumstances, as they indicate either a bug in libev
		170	or worse.
165		171
166		172
167	=head1 GLOBAL FUNCTIONS	173	=head1 GLOBAL FUNCTIONS
168		174
169	These functions can be called anytime, even before initialising the	175	These functions can be called anytime, even before initialising the
…		…
174	=item ev_tstamp ev_time ()	180	=item ev_tstamp ev_time ()
175		181
176	Returns the current time as libev would use it. Please note that the	182	Returns the current time as libev would use it. Please note that the
177	C<ev_now> function is usually faster and also often returns the timestamp	183	C<ev_now> function is usually faster and also often returns the timestamp
178	you actually want to know. Also interesting is the combination of	184	you actually want to know. Also interesting is the combination of
179	C<ev_update_now> and C<ev_now>.	185	C<ev_now_update> and C<ev_now>.
180		186
181	=item ev_sleep (ev_tstamp interval)	187	=item ev_sleep (ev_tstamp interval)
182		188
183	Sleep for the given interval: The current thread will be blocked until	189	Sleep for the given interval: The current thread will be blocked
184	either it is interrupted or the given time interval has passed. Basically	190	until either it is interrupted or the given time interval has
		191	passed (approximately - it might return a bit earlier even if not
		192	interrupted). Returns immediately if C<< interval <= 0 >>.
		193
185	this is a sub-second-resolution C<sleep ()>.	194	Basically this is a sub-second-resolution C<sleep ()>.
		195
		196	The range of the C<interval> is limited - libev only guarantees to work
		197	with sleep times of up to one day (C<< interval <= 86400 >>).
186		198
187	=item int ev_version_major ()	199	=item int ev_version_major ()
188		200
189	=item int ev_version_minor ()	201	=item int ev_version_minor ()
190		202
…		…
241	the current system, you would need to look at C<ev_embeddable_backends ()	253	the current system, you would need to look at C<ev_embeddable_backends ()
242	& ev_supported_backends ()>, likewise for recommended ones.	254	& ev_supported_backends ()>, likewise for recommended ones.
243		255
244	See the description of C<ev_embed> watchers for more info.	256	See the description of C<ev_embed> watchers for more info.
245		257
246	=item ev_set_allocator (void *(*cb)(void *ptr, long size))	258	=item ev_set_allocator (void *(*cb)(void *ptr, long size) throw ())
247		259
248	Sets the allocation function to use (the prototype is similar - the	260	Sets the allocation function to use (the prototype is similar - the
249	semantics are identical to the C<realloc> C89/SuS/POSIX function). It is	261	semantics are identical to the C<realloc> C89/SuS/POSIX function). It is
250	used to allocate and free memory (no surprises here). If it returns zero	262	used to allocate and free memory (no surprises here). If it returns zero
251	when memory needs to be allocated (C<size != 0>), the library might abort	263	when memory needs to be allocated (C<size != 0>), the library might abort
…		…
257		269
258	You could override this function in high-availability programs to, say,	270	You could override this function in high-availability programs to, say,
259	free some memory if it cannot allocate memory, to use a special allocator,	271	free some memory if it cannot allocate memory, to use a special allocator,
260	or even to sleep a while and retry until some memory is available.	272	or even to sleep a while and retry until some memory is available.
261		273
		274	Example: The following is the C<realloc> function that libev itself uses
		275	which should work with C<realloc> and C<free> functions of all kinds and
		276	is probably a good basis for your own implementation.
		277
		278	static void *
		279	ev_realloc_emul (void *ptr, long size) EV_NOEXCEPT
		280	{
		281	if (size)
		282	return realloc (ptr, size);
		283
		284	free (ptr);
		285	return 0;
		286	}
		287
262	Example: Replace the libev allocator with one that waits a bit and then	288	Example: Replace the libev allocator with one that waits a bit and then
263	retries (example requires a standards-compliant C<realloc>).	289	retries.
264		290
265	static void *	291	static void *
266	persistent_realloc (void *ptr, size_t size)	292	persistent_realloc (void *ptr, size_t size)
267	{	293	{
		294	if (!size)
		295	{
		296	free (ptr);
		297	return 0;
		298	}
		299
268	for (;;)	300	for (;;)
269	{	301	{
270	void *newptr = realloc (ptr, size);	302	void *newptr = realloc (ptr, size);
271		303
272	if (newptr)	304	if (newptr)
…		…
277	}	309	}
278		310
279	...	311	...
280	ev_set_allocator (persistent_realloc);	312	ev_set_allocator (persistent_realloc);
281		313
282	=item ev_set_syserr_cb (void (*cb)(const char *msg))	314	=item ev_set_syserr_cb (void (*cb)(const char *msg) throw ())
283		315
284	Set the callback function to call on a retryable system call error (such	316	Set the callback function to call on a retryable system call error (such
285	as failed select, poll, epoll_wait). The message is a printable string	317	as failed select, poll, epoll_wait). The message is a printable string
286	indicating the system call or subsystem causing the problem. If this	318	indicating the system call or subsystem causing the problem. If this
287	callback is set, then libev will expect it to remedy the situation, no	319	callback is set, then libev will expect it to remedy the situation, no
…		…
390		422
391	If this flag bit is or'ed into the flag value (or the program runs setuid	423	If this flag bit is or'ed into the flag value (or the program runs setuid
392	or setgid) then libev will I<not> look at the environment variable	424	or setgid) then libev will I<not> look at the environment variable
393	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will	425	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will
394	override the flags completely if it is found in the environment. This is	426	override the flags completely if it is found in the environment. This is
395	useful to try out specific backends to test their performance, or to work	427	useful to try out specific backends to test their performance, to work
396	around bugs.	428	around bugs, or to make libev threadsafe (accessing environment variables
		429	cannot be done in a threadsafe way, but usually it works if no other
		430	thread modifies them).
397		431
398	=item C<EVFLAG_FORKCHECK>	432	=item C<EVFLAG_FORKCHECK>
399		433
400	Instead of calling C<ev_loop_fork> manually after a fork, you can also	434	Instead of calling C<ev_loop_fork> manually after a fork, you can also
401	make libev check for a fork in each iteration by enabling this flag.	435	make libev check for a fork in each iteration by enabling this flag.
402		436
403	This works by calling C<getpid ()> on every iteration of the loop,	437	This works by calling C<getpid ()> on every iteration of the loop,
404	and thus this might slow down your event loop if you do a lot of loop	438	and thus this might slow down your event loop if you do a lot of loop
405	iterations and little real work, but is usually not noticeable (on my	439	iterations and little real work, but is usually not noticeable (on my
406	GNU/Linux system for example, C<getpid> is actually a simple 5-insn sequence	440	GNU/Linux system for example, C<getpid> is actually a simple 5-insn
407	without a system call and thus I<very> fast, but my GNU/Linux system also has	441	sequence without a system call and thus I<very> fast, but my GNU/Linux
408	C<pthread_atfork> which is even faster).	442	system also has C<pthread_atfork> which is even faster). (Update: glibc
		443	versions 2.25 apparently removed the C<getpid> optimisation again).
409		444
410	The big advantage of this flag is that you can forget about fork (and	445	The big advantage of this flag is that you can forget about fork (and
411	forget about forgetting to tell libev about forking) when you use this	446	forget about forgetting to tell libev about forking, although you still
412	flag.	447	have to ignore C<SIGPIPE>) when you use this flag.
413		448
414	This flag setting cannot be overridden or specified in the C<LIBEV_FLAGS>	449	This flag setting cannot be overridden or specified in the C<LIBEV_FLAGS>
415	environment variable.	450	environment variable.
416		451
417	=item C<EVFLAG_NOINOTIFY>	452	=item C<EVFLAG_NOINOTIFY>
…		…
435	example) that can't properly initialise their signal masks.	470	example) that can't properly initialise their signal masks.
436		471
437	=item C<EVFLAG_NOSIGMASK>	472	=item C<EVFLAG_NOSIGMASK>
438		473
439	When this flag is specified, then libev will avoid to modify the signal	474	When this flag is specified, then libev will avoid to modify the signal
440	mask. Specifically, this means you ahve to make sure signals are unblocked	475	mask. Specifically, this means you have to make sure signals are unblocked
441	when you want to receive them.	476	when you want to receive them.
442		477
443	This behaviour is useful when you want to do your own signal handling, or	478	This behaviour is useful when you want to do your own signal handling, or
444	want to handle signals only in specific threads and want to avoid libev	479	want to handle signals only in specific threads and want to avoid libev
445	unblocking the signals.	480	unblocking the signals.
446		481
447	This flag's behaviour will become the default in future versions of libev.	482	It's also required by POSIX in a threaded program, as libev calls
		483	C<sigprocmask>, whose behaviour is officially unspecified.
		484
		485	=item C<EVFLAG_NOTIMERFD>
		486
		487	When this flag is specified, the libev will avoid using a C<timerfd> to
		488	detect time jumps. It will still be able to detect time jumps, but takes
		489	longer and has a lower accuracy in doing so, but saves a file descriptor
		490	per loop.
		491
		492	The current implementation only tries to use a C<timerfd> when the first
		493	C<ev_periodic> watcher is started and falls back on other methods if it
		494	cannot be created, but this behaviour might change in the future.
448		495
449	=item C<EVBACKEND_SELECT> (value 1, portable select backend)	496	=item C<EVBACKEND_SELECT> (value 1, portable select backend)
450		497
451	This is your standard select(2) backend. Not I<completely> standard, as	498	This is your standard select(2) backend. Not I<completely> standard, as
452	libev tries to roll its own fd_set with no limits on the number of fds,	499	libev tries to roll its own fd_set with no limits on the number of fds,
…		…
477	This backend maps C<EV_READ> to C<POLLIN | POLLERR | POLLHUP>, and	524	This backend maps C<EV_READ> to C<POLLIN | POLLERR | POLLHUP>, and
478	C<EV_WRITE> to C<POLLOUT | POLLERR | POLLHUP>.	525	C<EV_WRITE> to C<POLLOUT | POLLERR | POLLHUP>.
479		526
480	=item C<EVBACKEND_EPOLL> (value 4, Linux)	527	=item C<EVBACKEND_EPOLL> (value 4, Linux)
481		528
482	Use the linux-specific epoll(7) interface (for both pre- and post-2.6.9	529	Use the Linux-specific epoll(7) interface (for both pre- and post-2.6.9
483	kernels).	530	kernels).
484		531
485	For few fds, this backend is a bit little slower than poll and select,	532	For few fds, this backend is a bit little slower than poll and select, but
486	but it scales phenomenally better. While poll and select usually scale	533	it scales phenomenally better. While poll and select usually scale like
487	like O(total_fds) where n is the total number of fds (or the highest fd),	534	O(total_fds) where total_fds is the total number of fds (or the highest
488	epoll scales either O(1) or O(active_fds).	535	fd), epoll scales either O(1) or O(active_fds).
489		536
490	The epoll mechanism deserves honorable mention as the most misdesigned	537	The epoll mechanism deserves honorable mention as the most misdesigned
491	of the more advanced event mechanisms: mere annoyances include silently	538	of the more advanced event mechanisms: mere annoyances include silently
492	dropping file descriptors, requiring a system call per change per file	539	dropping file descriptors, requiring a system call per change per file
493	descriptor (and unnecessary guessing of parameters), problems with dup,	540	descriptor (and unnecessary guessing of parameters), problems with dup,
…		…
496	0.1ms) and so on. The biggest issue is fork races, however - if a program	543	0.1ms) and so on. The biggest issue is fork races, however - if a program
497	forks then I<both> parent and child process have to recreate the epoll	544	forks then I<both> parent and child process have to recreate the epoll
498	set, which can take considerable time (one syscall per file descriptor)	545	set, which can take considerable time (one syscall per file descriptor)
499	and is of course hard to detect.	546	and is of course hard to detect.
500		547
501	Epoll is also notoriously buggy - embedding epoll fds I<should> work, but	548	Epoll is also notoriously buggy - embedding epoll fds I<should> work,
502	of course I<doesn't>, and epoll just loves to report events for totally	549	but of course I<doesn't>, and epoll just loves to report events for
503	I<different> file descriptors (even already closed ones, so one cannot	550	totally I<different> file descriptors (even already closed ones, so
504	even remove them from the set) than registered in the set (especially	551	one cannot even remove them from the set) than registered in the set
505	on SMP systems). Libev tries to counter these spurious notifications by	552	(especially on SMP systems). Libev tries to counter these spurious
506	employing an additional generation counter and comparing that against the	553	notifications by employing an additional generation counter and comparing
507	events to filter out spurious ones, recreating the set when required. Last	554	that against the events to filter out spurious ones, recreating the set
		555	when required. Epoll also erroneously rounds down timeouts, but gives you
		556	no way to know when and by how much, so sometimes you have to busy-wait
		557	because epoll returns immediately despite a nonzero timeout. And last
508	not least, it also refuses to work with some file descriptors which work	558	not least, it also refuses to work with some file descriptors which work
509	perfectly fine with C<select> (files, many character devices...).	559	perfectly fine with C<select> (files, many character devices...).
510		560
511	Epoll is truly the train wreck analog among event poll mechanisms.	561	Epoll is truly the train wreck among event poll mechanisms, a frankenpoll,
		562	cobbled together in a hurry, no thought to design or interaction with
		563	others. Oh, the pain, will it ever stop...
512		564
513	While stopping, setting and starting an I/O watcher in the same iteration	565	While stopping, setting and starting an I/O watcher in the same iteration
514	will result in some caching, there is still a system call per such	566	will result in some caching, there is still a system call per such
515	incident (because the same I<file descriptor> could point to a different	567	incident (because the same I<file descriptor> could point to a different
516	I<file description> now), so its best to avoid that. Also, C<dup ()>'ed	568	I<file description> now), so its best to avoid that. Also, C<dup ()>'ed
…		…
528	All this means that, in practice, C<EVBACKEND_SELECT> can be as fast or	580	All this means that, in practice, C<EVBACKEND_SELECT> can be as fast or
529	faster than epoll for maybe up to a hundred file descriptors, depending on	581	faster than epoll for maybe up to a hundred file descriptors, depending on
530	the usage. So sad.	582	the usage. So sad.
531		583
532	While nominally embeddable in other event loops, this feature is broken in	584	While nominally embeddable in other event loops, this feature is broken in
533	all kernel versions tested so far.	585	a lot of kernel revisions, but probably(!) works in current versions.
534		586
535	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as	587	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
536	C<EVBACKEND_POLL>.	588	C<EVBACKEND_POLL>.
537		589
		590	=item C<EVBACKEND_LINUXAIO> (value 64, Linux)
		591
		592	Use the Linux-specific Linux AIO (I<not> C<< aio(7) >> but C<<
		593	io_submit(2) >>) event interface available in post-4.18 kernels (but libev
		594	only tries to use it in 4.19+).
		595
		596	This is another Linux train wreck of an event interface.
		597
		598	If this backend works for you (as of this writing, it was very
		599	experimental), it is the best event interface available on Linux and might
		600	be well worth enabling it - if it isn't available in your kernel this will
		601	be detected and this backend will be skipped.
		602
		603	This backend can batch oneshot requests and supports a user-space ring
		604	buffer to receive events. It also doesn't suffer from most of the design
		605	problems of epoll (such as not being able to remove event sources from
		606	the epoll set), and generally sounds too good to be true. Because, this
		607	being the Linux kernel, of course it suffers from a whole new set of
		608	limitations, forcing you to fall back to epoll, inheriting all its design
		609	issues.
		610
		611	For one, it is not easily embeddable (but probably could be done using
		612	an event fd at some extra overhead). It also is subject to a system wide
		613	limit that can be configured in F</proc/sys/fs/aio-max-nr>. If no AIO
		614	requests are left, this backend will be skipped during initialisation, and
		615	will switch to epoll when the loop is active.
		616
		617	Most problematic in practice, however, is that not all file descriptors
		618	work with it. For example, in Linux 5.1, TCP sockets, pipes, event fds,
		619	files, F</dev/null> and many others are supported, but ttys do not work
		620	properly (a known bug that the kernel developers don't care about, see
		621	L<https://lore.kernel.org/patchwork/patch/1047453/>), so this is not
		622	(yet?) a generic event polling interface.
		623
		624	Overall, it seems the Linux developers just don't want it to have a
		625	generic event handling mechanism other than C<select> or C<poll>.
		626
		627	To work around all these problem, the current version of libev uses its
		628	epoll backend as a fallback for file descriptor types that do not work. Or
		629	falls back completely to epoll if the kernel acts up.
		630
		631	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
		632	C<EVBACKEND_POLL>.
		633
538	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)	634	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)
539		635
540	Kqueue deserves special mention, as at the time of this writing, it	636	Kqueue deserves special mention, as at the time this backend was
541	was broken on all BSDs except NetBSD (usually it doesn't work reliably	637	implemented, it was broken on all BSDs except NetBSD (usually it doesn't
542	with anything but sockets and pipes, except on Darwin, where of course	638	work reliably with anything but sockets and pipes, except on Darwin,
543	it's completely useless). Unlike epoll, however, whose brokenness	639	where of course it's completely useless). Unlike epoll, however, whose
544	is by design, these kqueue bugs can (and eventually will) be fixed	640	brokenness is by design, these kqueue bugs can be (and mostly have been)
545	without API changes to existing programs. For this reason it's not being	641	fixed without API changes to existing programs. For this reason it's not
546	"auto-detected" unless you explicitly specify it in the flags (i.e. using	642	being "auto-detected" on all platforms unless you explicitly specify it
547	C<EVBACKEND_KQUEUE>) or libev was compiled on a known-to-be-good (-enough)	643	in the flags (i.e. using C<EVBACKEND_KQUEUE>) or libev was compiled on a
548	system like NetBSD.	644	known-to-be-good (-enough) system like NetBSD.
549		645
550	You still can embed kqueue into a normal poll or select backend and use it	646	You still can embed kqueue into a normal poll or select backend and use it
551	only for sockets (after having made sure that sockets work with kqueue on	647	only for sockets (after having made sure that sockets work with kqueue on
552	the target platform). See C<ev_embed> watchers for more info.	648	the target platform). See C<ev_embed> watchers for more info.
553		649
554	It scales in the same way as the epoll backend, but the interface to the	650	It scales in the same way as the epoll backend, but the interface to the
555	kernel is more efficient (which says nothing about its actual speed, of	651	kernel is more efficient (which says nothing about its actual speed, of
556	course). While stopping, setting and starting an I/O watcher does never	652	course). While stopping, setting and starting an I/O watcher does never
557	cause an extra system call as with C<EVBACKEND_EPOLL>, it still adds up to	653	cause an extra system call as with C<EVBACKEND_EPOLL>, it still adds up to
558	two event changes per incident. Support for C<fork ()> is very bad (but	654	two event changes per incident. Support for C<fork ()> is very bad (you
559	sane, unlike epoll) and it drops fds silently in similarly hard-to-detect	655	might have to leak fds on fork, but it's more sane than epoll) and it
560	cases	656	drops fds silently in similarly hard-to-detect cases.
561		657
562	This backend usually performs well under most conditions.	658	This backend usually performs well under most conditions.
563		659
564	While nominally embeddable in other event loops, this doesn't work	660	While nominally embeddable in other event loops, this doesn't work
565	everywhere, so you might need to test for this. And since it is broken	661	everywhere, so you might need to test for this. And since it is broken
…		…
579	and is not embeddable, which would limit the usefulness of this backend	675	and is not embeddable, which would limit the usefulness of this backend
580	immensely.	676	immensely.
581		677
582	=item C<EVBACKEND_PORT> (value 32, Solaris 10)	678	=item C<EVBACKEND_PORT> (value 32, Solaris 10)
583		679
584	This uses the Solaris 10 event port mechanism. As with everything on Solaris,	680	This uses the Solaris 10 event port mechanism. As with everything on
585	it's really slow, but it still scales very well (O(active_fds)).	681	Solaris, it's really slow, but it still scales very well (O(active_fds)).
586		682
587	While this backend scales well, it requires one system call per active	683	While this backend scales well, it requires one system call per active
588	file descriptor per loop iteration. For small and medium numbers of file	684	file descriptor per loop iteration. For small and medium numbers of file
589	descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend	685	descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend
590	might perform better.	686	might perform better.
…		…
592	On the positive side, this backend actually performed fully to	688	On the positive side, this backend actually performed fully to
593	specification in all tests and is fully embeddable, which is a rare feat	689	specification in all tests and is fully embeddable, which is a rare feat
594	among the OS-specific backends (I vastly prefer correctness over speed	690	among the OS-specific backends (I vastly prefer correctness over speed
595	hacks).	691	hacks).
596		692
597	On the negative side, the interface is I<bizarre>, with the event polling	693	On the negative side, the interface is I<bizarre> - so bizarre that
		694	even sun itself gets it wrong in their code examples: The event polling
598	function sometimes returning events to the caller even though an error	695	function sometimes returns events to the caller even though an error
599	occured, but with no indication whether it has done so or not (yes, it's	696	occurred, but with no indication whether it has done so or not (yes, it's
600	even documented that way) - deadly for edge-triggered interfaces, but	697	even documented that way) - deadly for edge-triggered interfaces where you
		698	absolutely have to know whether an event occurred or not because you have
		699	to re-arm the watcher.
		700
601	fortunately libev seems to be able to work around it.	701	Fortunately libev seems to be able to work around these idiocies.
602		702
603	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as	703	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
604	C<EVBACKEND_POLL>.	704	C<EVBACKEND_POLL>.
605		705
606	=item C<EVBACKEND_ALL>	706	=item C<EVBACKEND_ALL>
…		…
634		734
635	Example: Use whatever libev has to offer, but make sure that kqueue is	735	Example: Use whatever libev has to offer, but make sure that kqueue is
636	used if available.	736	used if available.
637		737
638	struct ev_loop *loop = ev_loop_new (ev_recommended_backends () | EVBACKEND_KQUEUE);	738	struct ev_loop *loop = ev_loop_new (ev_recommended_backends () | EVBACKEND_KQUEUE);
		739
		740	Example: Similarly, on linux, you mgiht want to take advantage of the
		741	linux aio backend if possible, but fall back to something else if that
		742	isn't available.
		743
		744	struct ev_loop *loop = ev_loop_new (ev_recommended_backends () | EVBACKEND_LINUXAIO);
639		745
640	=item ev_loop_destroy (loop)	746	=item ev_loop_destroy (loop)
641		747
642	Destroys an event loop object (frees all memory and kernel state	748	Destroys an event loop object (frees all memory and kernel state
643	etc.). None of the active event watchers will be stopped in the normal	749	etc.). None of the active event watchers will be stopped in the normal
…		…
660	If you need dynamically allocated loops it is better to use C<ev_loop_new>	766	If you need dynamically allocated loops it is better to use C<ev_loop_new>
661	and C<ev_loop_destroy>.	767	and C<ev_loop_destroy>.
662		768
663	=item ev_loop_fork (loop)	769	=item ev_loop_fork (loop)
664		770
665	This function sets a flag that causes subsequent C<ev_run> iterations to	771	This function sets a flag that causes subsequent C<ev_run> iterations
666	reinitialise the kernel state for backends that have one. Despite the	772	to reinitialise the kernel state for backends that have one. Despite
667	name, you can call it anytime, but it makes most sense after forking, in	773	the name, you can call it anytime you are allowed to start or stop
668	the child process. You I<must> call it (or use C<EVFLAG_FORKCHECK>) in the	774	watchers (except inside an C<ev_prepare> callback), but it makes most
		775	sense after forking, in the child process. You I<must> call it (or use
669	child before resuming or calling C<ev_run>.	776	C<EVFLAG_FORKCHECK>) in the child before resuming or calling C<ev_run>.
670		777
		778	In addition, if you want to reuse a loop (via this function or
		779	C<EVFLAG_FORKCHECK>), you I<also> have to ignore C<SIGPIPE>.
		780
671	Again, you I<have> to call it on I<any> loop that you want to re-use after	781	Again, you I<have> to call it on I<any> loop that you want to re-use after
672	a fork, I<even if you do not plan to use the loop in the parent>. This is	782	a fork, I<even if you do not plan to use the loop in the parent>. This is
673	because some kernel interfaces *cough* I<kqueue> *cough* do funny things	783	because some kernel interfaces *cough* I<kqueue> *cough* do funny things
674	during fork.	784	during fork.
675		785
676	On the other hand, you only need to call this function in the child	786	On the other hand, you only need to call this function in the child
…		…
746		856
747	This function is rarely useful, but when some event callback runs for a	857	This function is rarely useful, but when some event callback runs for a
748	very long time without entering the event loop, updating libev's idea of	858	very long time without entering the event loop, updating libev's idea of
749	the current time is a good idea.	859	the current time is a good idea.
750		860
751	See also L<The special problem of time updates> in the C<ev_timer> section.	861	See also L</The special problem of time updates> in the C<ev_timer> section.
752		862
753	=item ev_suspend (loop)	863	=item ev_suspend (loop)
754		864
755	=item ev_resume (loop)	865	=item ev_resume (loop)
756		866
…		…
774	without a previous call to C<ev_suspend>.	884	without a previous call to C<ev_suspend>.
775		885
776	Calling C<ev_suspend>/C<ev_resume> has the side effect of updating the	886	Calling C<ev_suspend>/C<ev_resume> has the side effect of updating the
777	event loop time (see C<ev_now_update>).	887	event loop time (see C<ev_now_update>).
778		888
779	=item ev_run (loop, int flags)	889	=item bool ev_run (loop, int flags)
780		890
781	Finally, this is it, the event handler. This function usually is called	891	Finally, this is it, the event handler. This function usually is called
782	after you have initialised all your watchers and you want to start	892	after you have initialised all your watchers and you want to start
783	handling events. It will ask the operating system for any new events, call	893	handling events. It will ask the operating system for any new events, call
784	the watcher callbacks, an then repeat the whole process indefinitely: This	894	the watcher callbacks, and then repeat the whole process indefinitely: This
785	is why event loops are called I<loops>.	895	is why event loops are called I<loops>.
786		896
787	If the flags argument is specified as C<0>, it will keep handling events	897	If the flags argument is specified as C<0>, it will keep handling events
788	until either no event watchers are active anymore or C<ev_break> was	898	until either no event watchers are active anymore or C<ev_break> was
789	called.	899	called.
		900
		901	The return value is false if there are no more active watchers (which
		902	usually means "all jobs done" or "deadlock"), and true in all other cases
		903	(which usually means " you should call C<ev_run> again").
790		904
791	Please note that an explicit C<ev_break> is usually better than	905	Please note that an explicit C<ev_break> is usually better than
792	relying on all watchers to be stopped when deciding when a program has	906	relying on all watchers to be stopped when deciding when a program has
793	finished (especially in interactive programs), but having a program	907	finished (especially in interactive programs), but having a program
794	that automatically loops as long as it has to and no longer by virtue	908	that automatically loops as long as it has to and no longer by virtue
795	of relying on its watchers stopping correctly, that is truly a thing of	909	of relying on its watchers stopping correctly, that is truly a thing of
796	beauty.	910	beauty.
797		911
798	This function is also I<mostly> exception-safe - you can break out of	912	This function is I<mostly> exception-safe - you can break out of a
799	a C<ev_run> call by calling C<longjmp> in a callback, throwing a C++	913	C<ev_run> call by calling C<longjmp> in a callback, throwing a C++
800	exception and so on. This does not decrement the C<ev_depth> value, nor	914	exception and so on. This does not decrement the C<ev_depth> value, nor
801	will it clear any outstanding C<EVBREAK_ONE> breaks.	915	will it clear any outstanding C<EVBREAK_ONE> breaks.
802		916
803	A flags value of C<EVRUN_NOWAIT> will look for new events, will handle	917	A flags value of C<EVRUN_NOWAIT> will look for new events, will handle
804	those events and any already outstanding ones, but will not wait and	918	those events and any already outstanding ones, but will not wait and
…		…
816	This is useful if you are waiting for some external event in conjunction	930	This is useful if you are waiting for some external event in conjunction
817	with something not expressible using other libev watchers (i.e. "roll your	931	with something not expressible using other libev watchers (i.e. "roll your
818	own C<ev_run>"). However, a pair of C<ev_prepare>/C<ev_check> watchers is	932	own C<ev_run>"). However, a pair of C<ev_prepare>/C<ev_check> watchers is
819	usually a better approach for this kind of thing.	933	usually a better approach for this kind of thing.
820		934
821	Here are the gory details of what C<ev_run> does:	935	Here are the gory details of what C<ev_run> does (this is for your
		936	understanding, not a guarantee that things will work exactly like this in
		937	future versions):
822		938
823	- Increment loop depth.	939	- Increment loop depth.
824	- Reset the ev_break status.	940	- Reset the ev_break status.
825	- Before the first iteration, call any pending watchers.	941	- Before the first iteration, call any pending watchers.
826	LOOP:	942	LOOP:
…		…
843	- Queue all expired timers.	959	- Queue all expired timers.
844	- Queue all expired periodics.	960	- Queue all expired periodics.
845	- Queue all idle watchers with priority higher than that of pending events.	961	- Queue all idle watchers with priority higher than that of pending events.
846	- Queue all check watchers.	962	- Queue all check watchers.
847	- Call all queued watchers in reverse order (i.e. check watchers first).	963	- Call all queued watchers in reverse order (i.e. check watchers first).
848	Signals and child watchers are implemented as I/O watchers, and will	964	Signals, async and child watchers are implemented as I/O watchers, and
849	be handled here by queueing them when their watcher gets executed.	965	will be handled here by queueing them when their watcher gets executed.
850	- If ev_break has been called, or EVRUN_ONCE or EVRUN_NOWAIT	966	- If ev_break has been called, or EVRUN_ONCE or EVRUN_NOWAIT
851	were used, or there are no active watchers, goto FINISH, otherwise	967	were used, or there are no active watchers, goto FINISH, otherwise
852	continue with step LOOP.	968	continue with step LOOP.
853	FINISH:	969	FINISH:
854	- Reset the ev_break status iff it was EVBREAK_ONE.	970	- Reset the ev_break status iff it was EVBREAK_ONE.
…		…
859	anymore.	975	anymore.
860		976
861	... queue jobs here, make sure they register event watchers as long	977	... queue jobs here, make sure they register event watchers as long
862	... as they still have work to do (even an idle watcher will do..)	978	... as they still have work to do (even an idle watcher will do..)
863	ev_run (my_loop, 0);	979	ev_run (my_loop, 0);
864	... jobs done or somebody called unloop. yeah!	980	... jobs done or somebody called break. yeah!
865		981
866	=item ev_break (loop, how)	982	=item ev_break (loop, how)
867		983
868	Can be used to make a call to C<ev_run> return early (but only after it	984	Can be used to make a call to C<ev_run> return early (but only after it
869	has processed all outstanding events). The C<how> argument must be either	985	has processed all outstanding events). The C<how> argument must be either
…		…
932	overhead for the actual polling but can deliver many events at once.	1048	overhead for the actual polling but can deliver many events at once.
933		1049
934	By setting a higher I<io collect interval> you allow libev to spend more	1050	By setting a higher I<io collect interval> you allow libev to spend more
935	time collecting I/O events, so you can handle more events per iteration,	1051	time collecting I/O events, so you can handle more events per iteration,
936	at the cost of increasing latency. Timeouts (both C<ev_periodic> and	1052	at the cost of increasing latency. Timeouts (both C<ev_periodic> and
937	C<ev_timer>) will be not affected. Setting this to a non-null value will	1053	C<ev_timer>) will not be affected. Setting this to a non-null value will
938	introduce an additional C<ev_sleep ()> call into most loop iterations. The	1054	introduce an additional C<ev_sleep ()> call into most loop iterations. The
939	sleep time ensures that libev will not poll for I/O events more often then	1055	sleep time ensures that libev will not poll for I/O events more often then
940	once per this interval, on average.	1056	once per this interval, on average (as long as the host time resolution is
		1057	good enough).
941		1058
942	Likewise, by setting a higher I<timeout collect interval> you allow libev	1059	Likewise, by setting a higher I<timeout collect interval> you allow libev
943	to spend more time collecting timeouts, at the expense of increased	1060	to spend more time collecting timeouts, at the expense of increased
944	latency/jitter/inexactness (the watcher callback will be called	1061	latency/jitter/inexactness (the watcher callback will be called
945	later). C<ev_io> watchers will not be affected. Setting this to a non-null	1062	later). C<ev_io> watchers will not be affected. Setting this to a non-null
…		…
991	invoke the actual watchers inside another context (another thread etc.).	1108	invoke the actual watchers inside another context (another thread etc.).
992		1109
993	If you want to reset the callback, use C<ev_invoke_pending> as new	1110	If you want to reset the callback, use C<ev_invoke_pending> as new
994	callback.	1111	callback.
995		1112
996	=item ev_set_loop_release_cb (loop, void (*release)(EV_P), void (*acquire)(EV_P))	1113	=item ev_set_loop_release_cb (loop, void (*release)(EV_P) throw (), void (*acquire)(EV_P) throw ())
997		1114
998	Sometimes you want to share the same loop between multiple threads. This	1115	Sometimes you want to share the same loop between multiple threads. This
999	can be done relatively simply by putting mutex_lock/unlock calls around	1116	can be done relatively simply by putting mutex_lock/unlock calls around
1000	each call to a libev function.	1117	each call to a libev function.
1001		1118
1002	However, C<ev_run> can run an indefinite time, so it is not feasible	1119	However, C<ev_run> can run an indefinite time, so it is not feasible
1003	to wait for it to return. One way around this is to wake up the event	1120	to wait for it to return. One way around this is to wake up the event
1004	loop via C<ev_break> and C<av_async_send>, another way is to set these	1121	loop via C<ev_break> and C<ev_async_send>, another way is to set these
1005	I<release> and I<acquire> callbacks on the loop.	1122	I<release> and I<acquire> callbacks on the loop.
1006		1123
1007	When set, then C<release> will be called just before the thread is	1124	When set, then C<release> will be called just before the thread is
1008	suspended waiting for new events, and C<acquire> is called just	1125	suspended waiting for new events, and C<acquire> is called just
1009	afterwards.	1126	afterwards.
…		…
1101	with a watcher-specific start function (C<< ev_TYPE_start (loop, watcher	1218	with a watcher-specific start function (C<< ev_TYPE_start (loop, watcher
1102	*) >>), and you can stop watching for events at any time by calling the	1219	*) >>), and you can stop watching for events at any time by calling the
1103	corresponding stop function (C<< ev_TYPE_stop (loop, watcher *) >>.	1220	corresponding stop function (C<< ev_TYPE_stop (loop, watcher *) >>.
1104		1221
1105	As long as your watcher is active (has been started but not stopped) you	1222	As long as your watcher is active (has been started but not stopped) you
1106	must not touch the values stored in it. Most specifically you must never	1223	must not touch the values stored in it except when explicitly documented
1107	reinitialise it or call its C<ev_TYPE_set> macro.	1224	otherwise. Most specifically you must never reinitialise it or call its
		1225	C<ev_TYPE_set> macro.
1108		1226
1109	Each and every callback receives the event loop pointer as first, the	1227	Each and every callback receives the event loop pointer as first, the
1110	registered watcher structure as second, and a bitset of received events as	1228	registered watcher structure as second, and a bitset of received events as
1111	third argument.	1229	third argument.
1112		1230
…		…
1149		1267
1150	=item C<EV_PREPARE>	1268	=item C<EV_PREPARE>
1151		1269
1152	=item C<EV_CHECK>	1270	=item C<EV_CHECK>
1153		1271
1154	All C<ev_prepare> watchers are invoked just I<before> C<ev_run> starts	1272	All C<ev_prepare> watchers are invoked just I<before> C<ev_run> starts to
1155	to gather new events, and all C<ev_check> watchers are invoked just after	1273	gather new events, and all C<ev_check> watchers are queued (not invoked)
1156	C<ev_run> has gathered them, but before it invokes any callbacks for any	1274	just after C<ev_run> has gathered them, but before it queues any callbacks
		1275	for any received events. That means C<ev_prepare> watchers are the last
		1276	watchers invoked before the event loop sleeps or polls for new events, and
		1277	C<ev_check> watchers will be invoked before any other watchers of the same
		1278	or lower priority within an event loop iteration.
		1279
1157	received events. Callbacks of both watcher types can start and stop as	1280	Callbacks of both watcher types can start and stop as many watchers as
1158	many watchers as they want, and all of them will be taken into account	1281	they want, and all of them will be taken into account (for example, a
1159	(for example, a C<ev_prepare> watcher might start an idle watcher to keep	1282	C<ev_prepare> watcher might start an idle watcher to keep C<ev_run> from
1160	C<ev_run> from blocking).	1283	blocking).
1161		1284
1162	=item C<EV_EMBED>	1285	=item C<EV_EMBED>
1163		1286
1164	The embedded event loop specified in the C<ev_embed> watcher needs attention.	1287	The embedded event loop specified in the C<ev_embed> watcher needs attention.
1165		1288
…		…
1273		1396
1274	=item bool ev_is_active (ev_TYPE *watcher)	1397	=item bool ev_is_active (ev_TYPE *watcher)
1275		1398
1276	Returns a true value iff the watcher is active (i.e. it has been started	1399	Returns a true value iff the watcher is active (i.e. it has been started
1277	and not yet been stopped). As long as a watcher is active you must not modify	1400	and not yet been stopped). As long as a watcher is active you must not modify
1278	it.	1401	it unless documented otherwise.
		1402
		1403	Obviously, it is safe to call this on an active watcher, or actually any
		1404	watcher that is initialised.
1279		1405
1280	=item bool ev_is_pending (ev_TYPE *watcher)	1406	=item bool ev_is_pending (ev_TYPE *watcher)
1281		1407
1282	Returns a true value iff the watcher is pending, (i.e. it has outstanding	1408	Returns a true value iff the watcher is pending, (i.e. it has outstanding
1283	events but its callback has not yet been invoked). As long as a watcher	1409	events but its callback has not yet been invoked). As long as a watcher
1284	is pending (but not active) you must not call an init function on it (but	1410	is pending (but not active) you must not call an init function on it (but
1285	C<ev_TYPE_set> is safe), you must not change its priority, and you must	1411	C<ev_TYPE_set> is safe), you must not change its priority, and you must
1286	make sure the watcher is available to libev (e.g. you cannot C<free ()>	1412	make sure the watcher is available to libev (e.g. you cannot C<free ()>
1287	it).	1413	it).
1288		1414
		1415	It is safe to call this on any watcher in any state as long as it is
		1416	initialised.
		1417
1289	=item callback ev_cb (ev_TYPE *watcher)	1418	=item callback ev_cb (ev_TYPE *watcher)
1290		1419
1291	Returns the callback currently set on the watcher.	1420	Returns the callback currently set on the watcher.
1292		1421
1293	=item ev_cb_set (ev_TYPE *watcher, callback)	1422	=item ev_set_cb (ev_TYPE *watcher, callback)
1294		1423
1295	Change the callback. You can change the callback at virtually any time	1424	Change the callback. You can change the callback at virtually any time
1296	(modulo threads).	1425	(modulo threads).
1297		1426
1298	=item ev_set_priority (ev_TYPE *watcher, int priority)	1427	=item ev_set_priority (ev_TYPE *watcher, int priority)
…		…
1306	from being executed (except for C<ev_idle> watchers).	1435	from being executed (except for C<ev_idle> watchers).
1307		1436
1308	If you need to suppress invocation when higher priority events are pending	1437	If you need to suppress invocation when higher priority events are pending
1309	you need to look at C<ev_idle> watchers, which provide this functionality.	1438	you need to look at C<ev_idle> watchers, which provide this functionality.
1310		1439
1311	You I<must not> change the priority of a watcher as long as it is active or	1440	You I<must not> change the priority of a watcher as long as it is active
1312	pending.	1441	or pending. Reading the priority with C<ev_priority> is fine in any state.
1313		1442
1314	Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is	1443	Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is
1315	fine, as long as you do not mind that the priority value you query might	1444	fine, as long as you do not mind that the priority value you query might
1316	or might not have been clamped to the valid range.	1445	or might not have been clamped to the valid range.
1317		1446
1318	The default priority used by watchers when no priority has been set is	1447	The default priority used by watchers when no priority has been set is
1319	always C<0>, which is supposed to not be too high and not be too low :).	1448	always C<0>, which is supposed to not be too high and not be too low :).
1320		1449
1321	See L<WATCHER PRIORITY MODELS>, below, for a more thorough treatment of	1450	See L</WATCHER PRIORITY MODELS>, below, for a more thorough treatment of
1322	priorities.	1451	priorities.
1323		1452
1324	=item ev_invoke (loop, ev_TYPE *watcher, int revents)	1453	=item ev_invoke (loop, ev_TYPE *watcher, int revents)
1325		1454
1326	Invoke the C<watcher> with the given C<loop> and C<revents>. Neither	1455	Invoke the C<watcher> with the given C<loop> and C<revents>. Neither
…		…
1339		1468
1340	=item ev_feed_event (loop, ev_TYPE *watcher, int revents)	1469	=item ev_feed_event (loop, ev_TYPE *watcher, int revents)
1341		1470
1342	Feeds the given event set into the event loop, as if the specified event	1471	Feeds the given event set into the event loop, as if the specified event
1343	had happened for the specified watcher (which must be a pointer to an	1472	had happened for the specified watcher (which must be a pointer to an
1344	initialised but not necessarily started event watcher). Obviously you must	1473	initialised but not necessarily started event watcher, though it can be
1345	not free the watcher as long as it has pending events.	1474	active). Obviously you must not free the watcher as long as it has pending
		1475	events.
1346		1476
1347	Stopping the watcher, letting libev invoke it, or calling	1477	Stopping the watcher, letting libev invoke it, or calling
1348	C<ev_clear_pending> will clear the pending event, even if the watcher was	1478	C<ev_clear_pending> will clear the pending event, even if the watcher was
1349	not started in the first place.	1479	not started in the first place.
1350		1480
1351	See also C<ev_feed_fd_event> and C<ev_feed_signal_event> for related	1481	See also C<ev_feed_fd_event> and C<ev_feed_signal_event> for related
1352	functions that do not need a watcher.	1482	functions that do not need a watcher.
1353		1483
1354	=back	1484	=back
1355		1485
1356	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER	1486	See also the L</ASSOCIATING CUSTOM DATA WITH A WATCHER> and L</BUILDING YOUR
1357		1487	OWN COMPOSITE WATCHERS> idioms.
1358	Each watcher has, by default, a member C<void *data> that you can change
1359	and read at any time: libev will completely ignore it. This can be used
1360	to associate arbitrary data with your watcher. If you need more data and
1361	don't want to allocate memory and store a pointer to it in that data
1362	member, you can also "subclass" the watcher type and provide your own
1363	data:
1364
1365	struct my_io
1366	{
1367	ev_io io;
1368	int otherfd;
1369	void *somedata;
1370	struct whatever *mostinteresting;
1371	};
1372
1373	...
1374	struct my_io w;
1375	ev_io_init (&w.io, my_cb, fd, EV_READ);
1376
1377	And since your callback will be called with a pointer to the watcher, you
1378	can cast it back to your own type:
1379
1380	static void my_cb (struct ev_loop *loop, ev_io *w_, int revents)
1381	{
1382	struct my_io *w = (struct my_io *)w_;
1383	...
1384	}
1385
1386	More interesting and less C-conformant ways of casting your callback type
1387	instead have been omitted.
1388
1389	Another common scenario is to use some data structure with multiple
1390	embedded watchers:
1391
1392	struct my_biggy
1393	{
1394	int some_data;
1395	ev_timer t1;
1396	ev_timer t2;
1397	}
1398
1399	In this case getting the pointer to C<my_biggy> is a bit more
1400	complicated: Either you store the address of your C<my_biggy> struct
1401	in the C<data> member of the watcher (for woozies), or you need to use
1402	some pointer arithmetic using C<offsetof> inside your watchers (for real
1403	programmers):
1404
1405	#include <stddef.h>
1406
1407	static void
1408	t1_cb (EV_P_ ev_timer *w, int revents)
1409	{
1410	struct my_biggy big = (struct my_biggy *)
1411	(((char *)w) - offsetof (struct my_biggy, t1));
1412	}
1413
1414	static void
1415	t2_cb (EV_P_ ev_timer *w, int revents)
1416	{
1417	struct my_biggy big = (struct my_biggy *)
1418	(((char *)w) - offsetof (struct my_biggy, t2));
1419	}
1420		1488
1421	=head2 WATCHER STATES	1489	=head2 WATCHER STATES
1422		1490
1423	There are various watcher states mentioned throughout this manual -	1491	There are various watcher states mentioned throughout this manual -
1424	active, pending and so on. In this section these states and the rules to	1492	active, pending and so on. In this section these states and the rules to
1425	transition between them will be described in more detail - and while these	1493	transition between them will be described in more detail - and while these
1426	rules might look complicated, they usually do "the right thing".	1494	rules might look complicated, they usually do "the right thing".
1427		1495
1428	=over 4	1496	=over 4
1429		1497
1430	=item initialiased	1498	=item initialised
1431		1499
1432	Before a watcher can be registered with the event looop it has to be	1500	Before a watcher can be registered with the event loop it has to be
1433	initialised. This can be done with a call to C<ev_TYPE_init>, or calls to	1501	initialised. This can be done with a call to C<ev_TYPE_init>, or calls to
1434	C<ev_init> followed by the watcher-specific C<ev_TYPE_set> function.	1502	C<ev_init> followed by the watcher-specific C<ev_TYPE_set> function.
1435		1503
1436	In this state it is simply some block of memory that is suitable for use	1504	In this state it is simply some block of memory that is suitable for
1437	in an event loop. It can be moved around, freed, reused etc. at will.	1505	use in an event loop. It can be moved around, freed, reused etc. at
		1506	will - as long as you either keep the memory contents intact, or call
		1507	C<ev_TYPE_init> again.
1438		1508
1439	=item started/running/active	1509	=item started/running/active
1440		1510
1441	Once a watcher has been started with a call to C<ev_TYPE_start> it becomes	1511	Once a watcher has been started with a call to C<ev_TYPE_start> it becomes
1442	property of the event loop, and is actively waiting for events. While in	1512	property of the event loop, and is actively waiting for events. While in
1443	this state it cannot be accessed (except in a few documented ways), moved,	1513	this state it cannot be accessed (except in a few documented ways, such as
1444	freed or anything else - the only legal thing is to keep a pointer to it,	1514	stoping it), moved, freed or anything else - the only legal thing is to
1445	and call libev functions on it that are documented to work on active watchers.	1515	keep a pointer to it, and call libev functions on it that are documented
		1516	to work on active watchers.
		1517
		1518	As a rule of thumb, before accessing a member or calling any function on
		1519	a watcher, it should be stopped (or freshly initialised). If that is not
		1520	convenient, you can check the documentation for that function or member to
		1521	see if it is safe to use on an active watcher.
1446		1522
1447	=item pending	1523	=item pending
1448		1524
1449	If a watcher is active and libev determines that an event it is interested	1525	If a watcher is active and libev determines that an event it is interested
1450	in has occurred (such as a timer expiring), it will become pending. It will	1526	in has occurred (such as a timer expiring), it will become pending. It
1451	stay in this pending state until either it is stopped or its callback is	1527	will stay in this pending state until either it is explicitly stopped or
1452	about to be invoked, so it is not normally pending inside the watcher	1528	its callback is about to be invoked, so it is not normally pending inside
1453	callback.	1529	the watcher callback.
1454		1530
1455	The watcher might or might not be active while it is pending (for example,	1531	Generally, the watcher might or might not be active while it is pending
1456	an expired non-repeating timer can be pending but no longer active). If it	1532	(for example, an expired non-repeating timer can be pending but no longer
1457	is stopped, it can be freely accessed (e.g. by calling C<ev_TYPE_set>),	1533	active). If it is pending but not active, it can be freely accessed (e.g.
1458	but it is still property of the event loop at this time, so cannot be	1534	by calling C<ev_TYPE_set>), but it is still property of the event loop at
1459	moved, freed or reused. And if it is active the rules described in the	1535	this time, so cannot be moved, freed or reused. And if it is active the
1460	previous item still apply.	1536	rules described in the previous item still apply.
		1537
		1538	Explicitly stopping a watcher will also clear the pending state
		1539	unconditionally, so it is safe to stop a watcher and then free it.
1461		1540
1462	It is also possible to feed an event on a watcher that is not active (e.g.	1541	It is also possible to feed an event on a watcher that is not active (e.g.
1463	via C<ev_feed_event>), in which case it becomes pending without being	1542	via C<ev_feed_event>), in which case it becomes pending without being
1464	active.	1543	active.
1465		1544
…		…
1470	latter will clear any pending state the watcher might be in, regardless	1549	latter will clear any pending state the watcher might be in, regardless
1471	of whether it was active or not, so stopping a watcher explicitly before	1550	of whether it was active or not, so stopping a watcher explicitly before
1472	freeing it is often a good idea.	1551	freeing it is often a good idea.
1473		1552
1474	While stopped (and not pending) the watcher is essentially in the	1553	While stopped (and not pending) the watcher is essentially in the
1475	initialised state, that is it can be reused, moved, modified in any way	1554	initialised state, that is, it can be reused, moved, modified in any way
1476	you wish.	1555	you wish (but when you trash the memory block, you need to C<ev_TYPE_init>
		1556	it again).
1477		1557
1478	=back	1558	=back
1479		1559
1480	=head2 WATCHER PRIORITY MODELS	1560	=head2 WATCHER PRIORITY MODELS
1481		1561
1482	Many event loops support I<watcher priorities>, which are usually small	1562	Many event loops support I<watcher priorities>, which are usually small
1483	integers that influence the ordering of event callback invocation	1563	integers that influence the ordering of event callback invocation
1484	between watchers in some way, all else being equal.	1564	between watchers in some way, all else being equal.
1485		1565
1486	In libev, Watcher priorities can be set using C<ev_set_priority>. See its	1566	In libev, watcher priorities can be set using C<ev_set_priority>. See its
1487	description for the more technical details such as the actual priority	1567	description for the more technical details such as the actual priority
1488	range.	1568	range.
1489		1569
1490	There are two common ways how these these priorities are being interpreted	1570	There are two common ways how these these priorities are being interpreted
1491	by event loops:	1571	by event loops:
…		…
1585		1665
1586	This section describes each watcher in detail, but will not repeat	1666	This section describes each watcher in detail, but will not repeat
1587	information given in the last section. Any initialisation/set macros,	1667	information given in the last section. Any initialisation/set macros,
1588	functions and members specific to the watcher type are explained.	1668	functions and members specific to the watcher type are explained.
1589		1669
1590	Members are additionally marked with either I<[read-only]>, meaning that,	1670	Most members are additionally marked with either I<[read-only]>, meaning
1591	while the watcher is active, you can look at the member and expect some	1671	that, while the watcher is active, you can look at the member and expect
1592	sensible content, but you must not modify it (you can modify it while the	1672	some sensible content, but you must not modify it (you can modify it while
1593	watcher is stopped to your hearts content), or I<[read-write]>, which	1673	the watcher is stopped to your hearts content), or I<[read-write]>, which
1594	means you can expect it to have some sensible content while the watcher	1674	means you can expect it to have some sensible content while the watcher is
1595	is active, but you can also modify it. Modifying it may not do something	1675	active, but you can also modify it (within the same thread as the event
		1676	loop, i.e. without creating data races). Modifying it may not do something
1596	sensible or take immediate effect (or do anything at all), but libev will	1677	sensible or take immediate effect (or do anything at all), but libev will
1597	not crash or malfunction in any way.	1678	not crash or malfunction in any way.
1598		1679
		1680	In any case, the documentation for each member will explain what the
		1681	effects are, and if there are any additional access restrictions.
1599		1682
1600	=head2 C<ev_io> - is this file descriptor readable or writable?	1683	=head2 C<ev_io> - is this file descriptor readable or writable?
1601		1684
1602	I/O watchers check whether a file descriptor is readable or writable	1685	I/O watchers check whether a file descriptor is readable or writable
1603	in each iteration of the event loop, or, more precisely, when reading	1686	in each iteration of the event loop, or, more precisely, when reading
…		…
1610	In general you can register as many read and/or write event watchers per	1693	In general you can register as many read and/or write event watchers per
1611	fd as you want (as long as you don't confuse yourself). Setting all file	1694	fd as you want (as long as you don't confuse yourself). Setting all file
1612	descriptors to non-blocking mode is also usually a good idea (but not	1695	descriptors to non-blocking mode is also usually a good idea (but not
1613	required if you know what you are doing).	1696	required if you know what you are doing).
1614		1697
1615	If you cannot use non-blocking mode, then force the use of a
1616	known-to-be-good backend (at the time of this writing, this includes only
1617	C<EVBACKEND_SELECT> and C<EVBACKEND_POLL>). The same applies to file
1618	descriptors for which non-blocking operation makes no sense (such as
1619	files) - libev doesn't guarantee any specific behaviour in that case.
1620
1621	Another thing you have to watch out for is that it is quite easy to	1698	Another thing you have to watch out for is that it is quite easy to
1622	receive "spurious" readiness notifications, that is your callback might	1699	receive "spurious" readiness notifications, that is, your callback might
1623	be called with C<EV_READ> but a subsequent C<read>(2) will actually block	1700	be called with C<EV_READ> but a subsequent C<read>(2) will actually block
1624	because there is no data. Not only are some backends known to create a	1701	because there is no data. It is very easy to get into this situation even
1625	lot of those (for example Solaris ports), it is very easy to get into	1702	with a relatively standard program structure. Thus it is best to always
1626	this situation even with a relatively standard program structure. Thus	1703	use non-blocking I/O: An extra C<read>(2) returning C<EAGAIN> is far
1627	it is best to always use non-blocking I/O: An extra C<read>(2) returning
1628	C<EAGAIN> is far preferable to a program hanging until some data arrives.	1704	preferable to a program hanging until some data arrives.
1629		1705
1630	If you cannot run the fd in non-blocking mode (for example you should	1706	If you cannot run the fd in non-blocking mode (for example you should
1631	not play around with an Xlib connection), then you have to separately	1707	not play around with an Xlib connection), then you have to separately
1632	re-test whether a file descriptor is really ready with a known-to-be good	1708	re-test whether a file descriptor is really ready with a known-to-be good
1633	interface such as poll (fortunately in our Xlib example, Xlib already	1709	interface such as poll (fortunately in the case of Xlib, it already does
1634	does this on its own, so its quite safe to use). Some people additionally	1710	this on its own, so its quite safe to use). Some people additionally
1635	use C<SIGALRM> and an interval timer, just to be sure you won't block	1711	use C<SIGALRM> and an interval timer, just to be sure you won't block
1636	indefinitely.	1712	indefinitely.
1637		1713
1638	But really, best use non-blocking mode.	1714	But really, best use non-blocking mode.
1639		1715
1640	=head3 The special problem of disappearing file descriptors	1716	=head3 The special problem of disappearing file descriptors
1641		1717
1642	Some backends (e.g. kqueue, epoll) need to be told about closing a file	1718	Some backends (e.g. kqueue, epoll, linuxaio) need to be told about closing
1643	descriptor (either due to calling C<close> explicitly or any other means,	1719	a file descriptor (either due to calling C<close> explicitly or any other
1644	such as C<dup2>). The reason is that you register interest in some file	1720	means, such as C<dup2>). The reason is that you register interest in some
1645	descriptor, but when it goes away, the operating system will silently drop	1721	file descriptor, but when it goes away, the operating system will silently
1646	this interest. If another file descriptor with the same number then is	1722	drop this interest. If another file descriptor with the same number then
1647	registered with libev, there is no efficient way to see that this is, in	1723	is registered with libev, there is no efficient way to see that this is,
1648	fact, a different file descriptor.	1724	in fact, a different file descriptor.
1649		1725
1650	To avoid having to explicitly tell libev about such cases, libev follows	1726	To avoid having to explicitly tell libev about such cases, libev follows
1651	the following policy: Each time C<ev_io_set> is being called, libev	1727	the following policy: Each time C<ev_io_set> is being called, libev
1652	will assume that this is potentially a new file descriptor, otherwise	1728	will assume that this is potentially a new file descriptor, otherwise
1653	it is assumed that the file descriptor stays the same. That means that	1729	it is assumed that the file descriptor stays the same. That means that
…		…
1667		1743
1668	There is no workaround possible except not registering events	1744	There is no workaround possible except not registering events
1669	for potentially C<dup ()>'ed file descriptors, or to resort to	1745	for potentially C<dup ()>'ed file descriptors, or to resort to
1670	C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>.	1746	C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>.
1671		1747
		1748	=head3 The special problem of files
		1749
		1750	Many people try to use C<select> (or libev) on file descriptors
		1751	representing files, and expect it to become ready when their program
		1752	doesn't block on disk accesses (which can take a long time on their own).
		1753
		1754	However, this cannot ever work in the "expected" way - you get a readiness
		1755	notification as soon as the kernel knows whether and how much data is
		1756	there, and in the case of open files, that's always the case, so you
		1757	always get a readiness notification instantly, and your read (or possibly
		1758	write) will still block on the disk I/O.
		1759
		1760	Another way to view it is that in the case of sockets, pipes, character
		1761	devices and so on, there is another party (the sender) that delivers data
		1762	on its own, but in the case of files, there is no such thing: the disk
		1763	will not send data on its own, simply because it doesn't know what you
		1764	wish to read - you would first have to request some data.
		1765
		1766	Since files are typically not-so-well supported by advanced notification
		1767	mechanism, libev tries hard to emulate POSIX behaviour with respect
		1768	to files, even though you should not use it. The reason for this is
		1769	convenience: sometimes you want to watch STDIN or STDOUT, which is
		1770	usually a tty, often a pipe, but also sometimes files or special devices
		1771	(for example, C<epoll> on Linux works with F</dev/random> but not with
		1772	F</dev/urandom>), and even though the file might better be served with
		1773	asynchronous I/O instead of with non-blocking I/O, it is still useful when
		1774	it "just works" instead of freezing.
		1775
		1776	So avoid file descriptors pointing to files when you know it (e.g. use
		1777	libeio), but use them when it is convenient, e.g. for STDIN/STDOUT, or
		1778	when you rarely read from a file instead of from a socket, and want to
		1779	reuse the same code path.
		1780
1672	=head3 The special problem of fork	1781	=head3 The special problem of fork
1673		1782
1674	Some backends (epoll, kqueue) do not support C<fork ()> at all or exhibit	1783	Some backends (epoll, kqueue, linuxaio, iouring) do not support C<fork ()>
1675	useless behaviour. Libev fully supports fork, but needs to be told about	1784	at all or exhibit useless behaviour. Libev fully supports fork, but needs
1676	it in the child.	1785	to be told about it in the child if you want to continue to use it in the
		1786	child.
1677		1787
1678	To support fork in your programs, you either have to call	1788	To support fork in your child processes, you have to call C<ev_loop_fork
1679	C<ev_default_fork ()> or C<ev_loop_fork ()> after a fork in the child,	1789	()> after a fork in the child, enable C<EVFLAG_FORKCHECK>, or resort to
1680	enable C<EVFLAG_FORKCHECK>, or resort to C<EVBACKEND_SELECT> or	1790	C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>.
1681	C<EVBACKEND_POLL>.
1682		1791
1683	=head3 The special problem of SIGPIPE	1792	=head3 The special problem of SIGPIPE
1684		1793
1685	While not really specific to libev, it is easy to forget about C<SIGPIPE>:	1794	While not really specific to libev, it is easy to forget about C<SIGPIPE>:
1686	when writing to a pipe whose other end has been closed, your program gets	1795	when writing to a pipe whose other end has been closed, your program gets
…		…
1737	=item ev_io_init (ev_io *, callback, int fd, int events)	1846	=item ev_io_init (ev_io *, callback, int fd, int events)
1738		1847
1739	=item ev_io_set (ev_io *, int fd, int events)	1848	=item ev_io_set (ev_io *, int fd, int events)
1740		1849
1741	Configures an C<ev_io> watcher. The C<fd> is the file descriptor to	1850	Configures an C<ev_io> watcher. The C<fd> is the file descriptor to
1742	receive events for and C<events> is either C<EV_READ>, C<EV_WRITE> or	1851	receive events for and C<events> is either C<EV_READ>, C<EV_WRITE>, both
1743	C<EV_READ | EV_WRITE>, to express the desire to receive the given events.	1852	C<EV_READ | EV_WRITE> or C<0>, to express the desire to receive the given
		1853	events.
1744		1854
1745	=item int fd [read-only]	1855	Note that setting the C<events> to C<0> and starting the watcher is
		1856	supported, but not specially optimized - if your program sometimes happens
		1857	to generate this combination this is fine, but if it is easy to avoid
		1858	starting an io watcher watching for no events you should do so.
1746		1859
1747	The file descriptor being watched.	1860	=item ev_io_modify (ev_io *, int events)
1748		1861
		1862	Similar to C<ev_io_set>, but only changes the requested events. Using this
		1863	might be faster with some backends, as libev can assume that the C<fd>
		1864	still refers to the same underlying file description, something it cannot
		1865	do when using C<ev_io_set>.
		1866
		1867	=item int fd [no-modify]
		1868
		1869	The file descriptor being watched. While it can be read at any time, you
		1870	must not modify this member even when the watcher is stopped - always use
		1871	C<ev_io_set> for that.
		1872
1749	=item int events [read-only]	1873	=item int events [no-modify]
1750		1874
1751	The events being watched.	1875	The set of events the fd is being watched for, among other flags. Remember
		1876	that this is a bit set - to test for C<EV_READ>, use C<< w->events &
		1877	EV_READ >>, and similarly for C<EV_WRITE>.
		1878
		1879	As with C<fd>, you must not modify this member even when the watcher is
		1880	stopped, always use C<ev_io_set> or C<ev_io_modify> for that.
1752		1881
1753	=back	1882	=back
1754		1883
1755	=head3 Examples	1884	=head3 Examples
1756		1885
…		…
1784	detecting time jumps is hard, and some inaccuracies are unavoidable (the	1913	detecting time jumps is hard, and some inaccuracies are unavoidable (the
1785	monotonic clock option helps a lot here).	1914	monotonic clock option helps a lot here).
1786		1915
1787	The callback is guaranteed to be invoked only I<after> its timeout has	1916	The callback is guaranteed to be invoked only I<after> its timeout has
1788	passed (not I<at>, so on systems with very low-resolution clocks this	1917	passed (not I<at>, so on systems with very low-resolution clocks this
1789	might introduce a small delay). If multiple timers become ready during the	1918	might introduce a small delay, see "the special problem of being too
		1919	early", below). If multiple timers become ready during the same loop
1790	same loop iteration then the ones with earlier time-out values are invoked	1920	iteration then the ones with earlier time-out values are invoked before
1791	before ones of the same priority with later time-out values (but this is	1921	ones of the same priority with later time-out values (but this is no
1792	no longer true when a callback calls C<ev_run> recursively).	1922	longer true when a callback calls C<ev_run> recursively).
1793		1923
1794	=head3 Be smart about timeouts	1924	=head3 Be smart about timeouts
1795		1925
1796	Many real-world problems involve some kind of timeout, usually for error	1926	Many real-world problems involve some kind of timeout, usually for error
1797	recovery. A typical example is an HTTP request - if the other side hangs,	1927	recovery. A typical example is an HTTP request - if the other side hangs,
…		…
1872		2002
1873	In this case, it would be more efficient to leave the C<ev_timer> alone,	2003	In this case, it would be more efficient to leave the C<ev_timer> alone,
1874	but remember the time of last activity, and check for a real timeout only	2004	but remember the time of last activity, and check for a real timeout only
1875	within the callback:	2005	within the callback:
1876		2006
		2007	ev_tstamp timeout = 60.;
1877	ev_tstamp last_activity; // time of last activity	2008	ev_tstamp last_activity; // time of last activity
		2009	ev_timer timer;
1878		2010
1879	static void	2011	static void
1880	callback (EV_P_ ev_timer *w, int revents)	2012	callback (EV_P_ ev_timer *w, int revents)
1881	{	2013	{
1882	ev_tstamp now = ev_now (EV_A);	2014	// calculate when the timeout would happen
1883	ev_tstamp timeout = last_activity + 60.;	2015	ev_tstamp after = last_activity - ev_now (EV_A) + timeout;
1884		2016
1885	// if last_activity + 60. is older than now, we did time out	2017	// if negative, it means we the timeout already occurred
1886	if (timeout < now)	2018	if (after < 0.)
1887	{	2019	{
1888	// timeout occurred, take action	2020	// timeout occurred, take action
1889	}	2021	}
1890	else	2022	else
1891	{	2023	{
1892	// callback was invoked, but there was some activity, re-arm	2024	// callback was invoked, but there was some recent
1893	// the watcher to fire in last_activity + 60, which is	2025	// activity. simply restart the timer to time out
1894	// guaranteed to be in the future, so "again" is positive:	2026	// after "after" seconds, which is the earliest time
1895	w->repeat = timeout - now;	2027	// the timeout can occur.
		2028	ev_timer_set (w, after, 0.);
1896	ev_timer_again (EV_A_ w);	2029	ev_timer_start (EV_A_ w);
1897	}	2030	}
1898	}	2031	}
1899		2032
1900	To summarise the callback: first calculate the real timeout (defined	2033	To summarise the callback: first calculate in how many seconds the
1901	as "60 seconds after the last activity"), then check if that time has	2034	timeout will occur (by calculating the absolute time when it would occur,
1902	been reached, which means something I<did>, in fact, time out. Otherwise	2035	C<last_activity + timeout>, and subtracting the current time, C<ev_now
1903	the callback was invoked too early (C<timeout> is in the future), so	2036	(EV_A)> from that).
1904	re-schedule the timer to fire at that future time, to see if maybe we have
1905	a timeout then.
1906		2037
1907	Note how C<ev_timer_again> is used, taking advantage of the	2038	If this value is negative, then we are already past the timeout, i.e. we
1908	C<ev_timer_again> optimisation when the timer is already running.	2039	timed out, and need to do whatever is needed in this case.
		2040
		2041	Otherwise, we now the earliest time at which the timeout would trigger,
		2042	and simply start the timer with this timeout value.
		2043
		2044	In other words, each time the callback is invoked it will check whether
		2045	the timeout occurred. If not, it will simply reschedule itself to check
		2046	again at the earliest time it could time out. Rinse. Repeat.
1909		2047
1910	This scheme causes more callback invocations (about one every 60 seconds	2048	This scheme causes more callback invocations (about one every 60 seconds
1911	minus half the average time between activity), but virtually no calls to	2049	minus half the average time between activity), but virtually no calls to
1912	libev to change the timeout.	2050	libev to change the timeout.
1913		2051
1914	To start the timer, simply initialise the watcher and set C<last_activity>	2052	To start the machinery, simply initialise the watcher and set
1915	to the current time (meaning we just have some activity :), then call the	2053	C<last_activity> to the current time (meaning there was some activity just
1916	callback, which will "do the right thing" and start the timer:	2054	now), then call the callback, which will "do the right thing" and start
		2055	the timer:
1917		2056
		2057	last_activity = ev_now (EV_A);
1918	ev_init (timer, callback);	2058	ev_init (&timer, callback);
1919	last_activity = ev_now (loop);	2059	callback (EV_A_ &timer, 0);
1920	callback (loop, timer, EV_TIMER);
1921		2060
1922	And when there is some activity, simply store the current time in	2061	When there is some activity, simply store the current time in
1923	C<last_activity>, no libev calls at all:	2062	C<last_activity>, no libev calls at all:
1924		2063
		2064	if (activity detected)
1925	last_activity = ev_now (loop);	2065	last_activity = ev_now (EV_A);
		2066
		2067	When your timeout value changes, then the timeout can be changed by simply
		2068	providing a new value, stopping the timer and calling the callback, which
		2069	will again do the right thing (for example, time out immediately :).
		2070
		2071	timeout = new_value;
		2072	ev_timer_stop (EV_A_ &timer);
		2073	callback (EV_A_ &timer, 0);
1926		2074
1927	This technique is slightly more complex, but in most cases where the	2075	This technique is slightly more complex, but in most cases where the
1928	time-out is unlikely to be triggered, much more efficient.	2076	time-out is unlikely to be triggered, much more efficient.
1929
1930	Changing the timeout is trivial as well (if it isn't hard-coded in the
1931	callback :) - just change the timeout and invoke the callback, which will
1932	fix things for you.
1933		2077
1934	=item 4. Wee, just use a double-linked list for your timeouts.	2078	=item 4. Wee, just use a double-linked list for your timeouts.
1935		2079
1936	If there is not one request, but many thousands (millions...), all	2080	If there is not one request, but many thousands (millions...), all
1937	employing some kind of timeout with the same timeout value, then one can	2081	employing some kind of timeout with the same timeout value, then one can
…		…
1964	Method #1 is almost always a bad idea, and buys you nothing. Method #4 is	2108	Method #1 is almost always a bad idea, and buys you nothing. Method #4 is
1965	rather complicated, but extremely efficient, something that really pays	2109	rather complicated, but extremely efficient, something that really pays
1966	off after the first million or so of active timers, i.e. it's usually	2110	off after the first million or so of active timers, i.e. it's usually
1967	overkill :)	2111	overkill :)
1968		2112
		2113	=head3 The special problem of being too early
		2114
		2115	If you ask a timer to call your callback after three seconds, then
		2116	you expect it to be invoked after three seconds - but of course, this
		2117	cannot be guaranteed to infinite precision. Less obviously, it cannot be
		2118	guaranteed to any precision by libev - imagine somebody suspending the
		2119	process with a STOP signal for a few hours for example.
		2120
		2121	So, libev tries to invoke your callback as soon as possible I<after> the
		2122	delay has occurred, but cannot guarantee this.
		2123
		2124	A less obvious failure mode is calling your callback too early: many event
		2125	loops compare timestamps with a "elapsed delay >= requested delay", but
		2126	this can cause your callback to be invoked much earlier than you would
		2127	expect.
		2128
		2129	To see why, imagine a system with a clock that only offers full second
		2130	resolution (think windows if you can't come up with a broken enough OS
		2131	yourself). If you schedule a one-second timer at the time 500.9, then the
		2132	event loop will schedule your timeout to elapse at a system time of 500
		2133	(500.9 truncated to the resolution) + 1, or 501.
		2134
		2135	If an event library looks at the timeout 0.1s later, it will see "501 >=
		2136	501" and invoke the callback 0.1s after it was started, even though a
		2137	one-second delay was requested - this is being "too early", despite best
		2138	intentions.
		2139
		2140	This is the reason why libev will never invoke the callback if the elapsed
		2141	delay equals the requested delay, but only when the elapsed delay is
		2142	larger than the requested delay. In the example above, libev would only invoke
		2143	the callback at system time 502, or 1.1s after the timer was started.
		2144
		2145	So, while libev cannot guarantee that your callback will be invoked
		2146	exactly when requested, it I<can> and I<does> guarantee that the requested
		2147	delay has actually elapsed, or in other words, it always errs on the "too
		2148	late" side of things.
		2149
1969	=head3 The special problem of time updates	2150	=head3 The special problem of time updates
1970		2151
1971	Establishing the current time is a costly operation (it usually takes at	2152	Establishing the current time is a costly operation (it usually takes
1972	least two system calls): EV therefore updates its idea of the current	2153	at least one system call): EV therefore updates its idea of the current
1973	time only before and after C<ev_run> collects new events, which causes a	2154	time only before and after C<ev_run> collects new events, which causes a
1974	growing difference between C<ev_now ()> and C<ev_time ()> when handling	2155	growing difference between C<ev_now ()> and C<ev_time ()> when handling
1975	lots of events in one iteration.	2156	lots of events in one iteration.
1976		2157
1977	The relative timeouts are calculated relative to the C<ev_now ()>	2158	The relative timeouts are calculated relative to the C<ev_now ()>
1978	time. This is usually the right thing as this timestamp refers to the time	2159	time. This is usually the right thing as this timestamp refers to the time
1979	of the event triggering whatever timeout you are modifying/starting. If	2160	of the event triggering whatever timeout you are modifying/starting. If
1980	you suspect event processing to be delayed and you I<need> to base the	2161	you suspect event processing to be delayed and you I<need> to base the
1981	timeout on the current time, use something like this to adjust for this:	2162	timeout on the current time, use something like the following to adjust
		2163	for it:
1982		2164
1983	ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);	2165	ev_timer_set (&timer, after + (ev_time () - ev_now ()), 0.);
1984		2166
1985	If the event loop is suspended for a long time, you can also force an	2167	If the event loop is suspended for a long time, you can also force an
1986	update of the time returned by C<ev_now ()> by calling C<ev_now_update	2168	update of the time returned by C<ev_now ()> by calling C<ev_now_update
1987	()>.	2169	()>, although that will push the event time of all outstanding events
		2170	further into the future.
		2171
		2172	=head3 The special problem of unsynchronised clocks
		2173
		2174	Modern systems have a variety of clocks - libev itself uses the normal
		2175	"wall clock" clock and, if available, the monotonic clock (to avoid time
		2176	jumps).
		2177
		2178	Neither of these clocks is synchronised with each other or any other clock
		2179	on the system, so C<ev_time ()> might return a considerably different time
		2180	than C<gettimeofday ()> or C<time ()>. On a GNU/Linux system, for example,
		2181	a call to C<gettimeofday> might return a second count that is one higher
		2182	than a directly following call to C<time>.
		2183
		2184	The moral of this is to only compare libev-related timestamps with
		2185	C<ev_time ()> and C<ev_now ()>, at least if you want better precision than
		2186	a second or so.
		2187
		2188	One more problem arises due to this lack of synchronisation: if libev uses
		2189	the system monotonic clock and you compare timestamps from C<ev_time>
		2190	or C<ev_now> from when you started your timer and when your callback is
		2191	invoked, you will find that sometimes the callback is a bit "early".
		2192
		2193	This is because C<ev_timer>s work in real time, not wall clock time, so
		2194	libev makes sure your callback is not invoked before the delay happened,
		2195	I<measured according to the real time>, not the system clock.
		2196
		2197	If your timeouts are based on a physical timescale (e.g. "time out this
		2198	connection after 100 seconds") then this shouldn't bother you as it is
		2199	exactly the right behaviour.
		2200
		2201	If you want to compare wall clock/system timestamps to your timers, then
		2202	you need to use C<ev_periodic>s, as these are based on the wall clock
		2203	time, where your comparisons will always generate correct results.
1988		2204
1989	=head3 The special problems of suspended animation	2205	=head3 The special problems of suspended animation
1990		2206
1991	When you leave the server world it is quite customary to hit machines that	2207	When you leave the server world it is quite customary to hit machines that
1992	can suspend/hibernate - what happens to the clocks during such a suspend?	2208	can suspend/hibernate - what happens to the clocks during such a suspend?
…		…
2022		2238
2023	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)	2239	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)
2024		2240
2025	=item ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)	2241	=item ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)
2026		2242
2027	Configure the timer to trigger after C<after> seconds. If C<repeat>	2243	Configure the timer to trigger after C<after> seconds (fractional and
2028	is C<0.>, then it will automatically be stopped once the timeout is	2244	negative values are supported). If C<repeat> is C<0.>, then it will
2029	reached. If it is positive, then the timer will automatically be	2245	automatically be stopped once the timeout is reached. If it is positive,
2030	configured to trigger again C<repeat> seconds later, again, and again,	2246	then the timer will automatically be configured to trigger again C<repeat>
2031	until stopped manually.	2247	seconds later, again, and again, until stopped manually.
2032		2248
2033	The timer itself will do a best-effort at avoiding drift, that is, if	2249	The timer itself will do a best-effort at avoiding drift, that is, if
2034	you configure a timer to trigger every 10 seconds, then it will normally	2250	you configure a timer to trigger every 10 seconds, then it will normally
2035	trigger at exactly 10 second intervals. If, however, your program cannot	2251	trigger at exactly 10 second intervals. If, however, your program cannot
2036	keep up with the timer (because it takes longer than those 10 seconds to	2252	keep up with the timer (because it takes longer than those 10 seconds to
2037	do stuff) the timer will not fire more than once per event loop iteration.	2253	do stuff) the timer will not fire more than once per event loop iteration.
2038		2254
2039	=item ev_timer_again (loop, ev_timer *)	2255	=item ev_timer_again (loop, ev_timer *)
2040		2256
2041	This will act as if the timer timed out and restart it again if it is	2257	This will act as if the timer timed out, and restarts it again if it is
2042	repeating. The exact semantics are:	2258	repeating. It basically works like calling C<ev_timer_stop>, updating the
		2259	timeout to the C<repeat> value and calling C<ev_timer_start>.
2043		2260
		2261	The exact semantics are as in the following rules, all of which will be
		2262	applied to the watcher:
		2263
		2264	=over 4
		2265
2044	If the timer is pending, its pending status is cleared.	2266	=item If the timer is pending, the pending status is always cleared.
2045		2267
2046	If the timer is started but non-repeating, stop it (as if it timed out).	2268	=item If the timer is started but non-repeating, stop it (as if it timed
		2269	out, without invoking it).
2047		2270
2048	If the timer is repeating, either start it if necessary (with the	2271	=item If the timer is repeating, make the C<repeat> value the new timeout
2049	C<repeat> value), or reset the running timer to the C<repeat> value.	2272	and start the timer, if necessary.
2050		2273
		2274	=back
		2275
2051	This sounds a bit complicated, see L<Be smart about timeouts>, above, for a	2276	This sounds a bit complicated, see L</Be smart about timeouts>, above, for a
2052	usage example.	2277	usage example.
2053		2278
2054	=item ev_tstamp ev_timer_remaining (loop, ev_timer *)	2279	=item ev_tstamp ev_timer_remaining (loop, ev_timer *)
2055		2280
2056	Returns the remaining time until a timer fires. If the timer is active,	2281	Returns the remaining time until a timer fires. If the timer is active,
…		…
2109	Periodic watchers are also timers of a kind, but they are very versatile	2334	Periodic watchers are also timers of a kind, but they are very versatile
2110	(and unfortunately a bit complex).	2335	(and unfortunately a bit complex).
2111		2336
2112	Unlike C<ev_timer>, periodic watchers are not based on real time (or	2337	Unlike C<ev_timer>, periodic watchers are not based on real time (or
2113	relative time, the physical time that passes) but on wall clock time	2338	relative time, the physical time that passes) but on wall clock time
2114	(absolute time, the thing you can read on your calender or clock). The	2339	(absolute time, the thing you can read on your calendar or clock). The
2115	difference is that wall clock time can run faster or slower than real	2340	difference is that wall clock time can run faster or slower than real
2116	time, and time jumps are not uncommon (e.g. when you adjust your	2341	time, and time jumps are not uncommon (e.g. when you adjust your
2117	wrist-watch).	2342	wrist-watch).
2118		2343
2119	You can tell a periodic watcher to trigger after some specific point	2344	You can tell a periodic watcher to trigger after some specific point
…		…
2124	C<ev_timer>, which would still trigger roughly 10 seconds after starting	2349	C<ev_timer>, which would still trigger roughly 10 seconds after starting
2125	it, as it uses a relative timeout).	2350	it, as it uses a relative timeout).
2126		2351
2127	C<ev_periodic> watchers can also be used to implement vastly more complex	2352	C<ev_periodic> watchers can also be used to implement vastly more complex
2128	timers, such as triggering an event on each "midnight, local time", or	2353	timers, such as triggering an event on each "midnight, local time", or
2129	other complicated rules. This cannot be done with C<ev_timer> watchers, as	2354	other complicated rules. This cannot easily be done with C<ev_timer>
2130	those cannot react to time jumps.	2355	watchers, as those cannot react to time jumps.
2131		2356
2132	As with timers, the callback is guaranteed to be invoked only when the	2357	As with timers, the callback is guaranteed to be invoked only when the
2133	point in time where it is supposed to trigger has passed. If multiple	2358	point in time where it is supposed to trigger has passed. If multiple
2134	timers become ready during the same loop iteration then the ones with	2359	timers become ready during the same loop iteration then the ones with
2135	earlier time-out values are invoked before ones with later time-out values	2360	earlier time-out values are invoked before ones with later time-out values
…		…
2176		2401
2177	Another way to think about it (for the mathematically inclined) is that	2402	Another way to think about it (for the mathematically inclined) is that
2178	C<ev_periodic> will try to run the callback in this mode at the next possible	2403	C<ev_periodic> will try to run the callback in this mode at the next possible
2179	time where C<time = offset (mod interval)>, regardless of any time jumps.	2404	time where C<time = offset (mod interval)>, regardless of any time jumps.
2180		2405
2181	For numerical stability it is preferable that the C<offset> value is near	2406	The C<interval> I<MUST> be positive, and for numerical stability, the
2182	C<ev_now ()> (the current time), but there is no range requirement for	2407	interval value should be higher than C<1/8192> (which is around 100
2183	this value, and in fact is often specified as zero.	2408	microseconds) and C<offset> should be higher than C<0> and should have
		2409	at most a similar magnitude as the current time (say, within a factor of
		2410	ten). Typical values for offset are, in fact, C<0> or something between
		2411	C<0> and C<interval>, which is also the recommended range.
2184		2412
2185	Note also that there is an upper limit to how often a timer can fire (CPU	2413	Note also that there is an upper limit to how often a timer can fire (CPU
2186	speed for example), so if C<interval> is very small then timing stability	2414	speed for example), so if C<interval> is very small then timing stability
2187	will of course deteriorate. Libev itself tries to be exact to be about one	2415	will of course deteriorate. Libev itself tries to be exact to be about one
2188	millisecond (if the OS supports it and the machine is fast enough).	2416	millisecond (if the OS supports it and the machine is fast enough).
…		…
2218		2446
2219	NOTE: I<< This callback must always return a time that is higher than or	2447	NOTE: I<< This callback must always return a time that is higher than or
2220	equal to the passed C<now> value >>.	2448	equal to the passed C<now> value >>.
2221		2449
2222	This can be used to create very complex timers, such as a timer that	2450	This can be used to create very complex timers, such as a timer that
2223	triggers on "next midnight, local time". To do this, you would calculate the	2451	triggers on "next midnight, local time". To do this, you would calculate
2224	next midnight after C<now> and return the timestamp value for this. How	2452	the next midnight after C<now> and return the timestamp value for
2225	you do this is, again, up to you (but it is not trivial, which is the main	2453	this. Here is a (completely untested, no error checking) example on how to
2226	reason I omitted it as an example).	2454	do this:
		2455
		2456	#include <time.h>
		2457
		2458	static ev_tstamp
		2459	my_rescheduler (ev_periodic *w, ev_tstamp now)
		2460	{
		2461	time_t tnow = (time_t)now;
		2462	struct tm tm;
		2463	localtime_r (&tnow, &tm);
		2464
		2465	tm.tm_sec = tm.tm_min = tm.tm_hour = 0; // midnight current day
		2466	++tm.tm_mday; // midnight next day
		2467
		2468	return mktime (&tm);
		2469	}
		2470
		2471	Note: this code might run into trouble on days that have more then two
		2472	midnights (beginning and end).
2227		2473
2228	=back	2474	=back
2229		2475
2230	=item ev_periodic_again (loop, ev_periodic *)	2476	=item ev_periodic_again (loop, ev_periodic *)
2231		2477
…		…
2296		2542
2297	ev_periodic hourly_tick;	2543	ev_periodic hourly_tick;
2298	ev_periodic_init (&hourly_tick, clock_cb,	2544	ev_periodic_init (&hourly_tick, clock_cb,
2299	fmod (ev_now (loop), 3600.), 3600., 0);	2545	fmod (ev_now (loop), 3600.), 3600., 0);
2300	ev_periodic_start (loop, &hourly_tick);	2546	ev_periodic_start (loop, &hourly_tick);
2301		2547
2302		2548
2303	=head2 C<ev_signal> - signal me when a signal gets signalled!	2549	=head2 C<ev_signal> - signal me when a signal gets signalled!
2304		2550
2305	Signal watchers will trigger an event when the process receives a specific	2551	Signal watchers will trigger an event when the process receives a specific
2306	signal one or more times. Even though signals are very asynchronous, libev	2552	signal one or more times. Even though signals are very asynchronous, libev
…		…
2316	only within the same loop, i.e. you can watch for C<SIGINT> in your	2562	only within the same loop, i.e. you can watch for C<SIGINT> in your
2317	default loop and for C<SIGIO> in another loop, but you cannot watch for	2563	default loop and for C<SIGIO> in another loop, but you cannot watch for
2318	C<SIGINT> in both the default loop and another loop at the same time. At	2564	C<SIGINT> in both the default loop and another loop at the same time. At
2319	the moment, C<SIGCHLD> is permanently tied to the default loop.	2565	the moment, C<SIGCHLD> is permanently tied to the default loop.
2320		2566
2321	When the first watcher gets started will libev actually register something	2567	Only after the first watcher for a signal is started will libev actually
2322	with the kernel (thus it coexists with your own signal handlers as long as	2568	register something with the kernel. It thus coexists with your own signal
2323	you don't register any with libev for the same signal).	2569	handlers as long as you don't register any with libev for the same signal.
2324		2570
2325	If possible and supported, libev will install its handlers with	2571	If possible and supported, libev will install its handlers with
2326	C<SA_RESTART> (or equivalent) behaviour enabled, so system calls should	2572	C<SA_RESTART> (or equivalent) behaviour enabled, so system calls should
2327	not be unduly interrupted. If you have a problem with system calls getting	2573	not be unduly interrupted. If you have a problem with system calls getting
2328	interrupted by signals you can block all signals in an C<ev_check> watcher	2574	interrupted by signals you can block all signals in an C<ev_check> watcher
…		…
2331	=head3 The special problem of inheritance over fork/execve/pthread_create	2577	=head3 The special problem of inheritance over fork/execve/pthread_create
2332		2578
2333	Both the signal mask (C<sigprocmask>) and the signal disposition	2579	Both the signal mask (C<sigprocmask>) and the signal disposition
2334	(C<sigaction>) are unspecified after starting a signal watcher (and after	2580	(C<sigaction>) are unspecified after starting a signal watcher (and after
2335	stopping it again), that is, libev might or might not block the signal,	2581	stopping it again), that is, libev might or might not block the signal,
2336	and might or might not set or restore the installed signal handler.	2582	and might or might not set or restore the installed signal handler (but
		2583	see C<EVFLAG_NOSIGMASK>).
2337		2584
2338	While this does not matter for the signal disposition (libev never	2585	While this does not matter for the signal disposition (libev never
2339	sets signals to C<SIG_IGN>, so handlers will be reset to C<SIG_DFL> on	2586	sets signals to C<SIG_IGN>, so handlers will be reset to C<SIG_DFL> on
2340	C<execve>), this matters for the signal mask: many programs do not expect	2587	C<execve>), this matters for the signal mask: many programs do not expect
2341	certain signals to be blocked.	2588	certain signals to be blocked.
…		…
2512		2759
2513	=head2 C<ev_stat> - did the file attributes just change?	2760	=head2 C<ev_stat> - did the file attributes just change?
2514		2761
2515	This watches a file system path for attribute changes. That is, it calls	2762	This watches a file system path for attribute changes. That is, it calls
2516	C<stat> on that path in regular intervals (or when the OS says it changed)	2763	C<stat> on that path in regular intervals (or when the OS says it changed)
2517	and sees if it changed compared to the last time, invoking the callback if	2764	and sees if it changed compared to the last time, invoking the callback
2518	it did.	2765	if it did. Starting the watcher C<stat>'s the file, so only changes that
		2766	happen after the watcher has been started will be reported.
2519		2767
2520	The path does not need to exist: changing from "path exists" to "path does	2768	The path does not need to exist: changing from "path exists" to "path does
2521	not exist" is a status change like any other. The condition "path does not	2769	not exist" is a status change like any other. The condition "path does not
2522	exist" (or more correctly "path cannot be stat'ed") is signified by the	2770	exist" (or more correctly "path cannot be stat'ed") is signified by the
2523	C<st_nlink> field being zero (which is otherwise always forced to be at	2771	C<st_nlink> field being zero (which is otherwise always forced to be at
…		…
2753	Apart from keeping your process non-blocking (which is a useful	3001	Apart from keeping your process non-blocking (which is a useful
2754	effect on its own sometimes), idle watchers are a good place to do	3002	effect on its own sometimes), idle watchers are a good place to do
2755	"pseudo-background processing", or delay processing stuff to after the	3003	"pseudo-background processing", or delay processing stuff to after the
2756	event loop has handled all outstanding events.	3004	event loop has handled all outstanding events.
2757		3005
		3006	=head3 Abusing an C<ev_idle> watcher for its side-effect
		3007
		3008	As long as there is at least one active idle watcher, libev will never
		3009	sleep unnecessarily. Or in other words, it will loop as fast as possible.
		3010	For this to work, the idle watcher doesn't need to be invoked at all - the
		3011	lowest priority will do.
		3012
		3013	This mode of operation can be useful together with an C<ev_check> watcher,
		3014	to do something on each event loop iteration - for example to balance load
		3015	between different connections.
		3016
		3017	See L</Abusing an ev_check watcher for its side-effect> for a longer
		3018	example.
		3019
2758	=head3 Watcher-Specific Functions and Data Members	3020	=head3 Watcher-Specific Functions and Data Members
2759		3021
2760	=over 4	3022	=over 4
2761		3023
2762	=item ev_idle_init (ev_idle *, callback)	3024	=item ev_idle_init (ev_idle *, callback)
…		…
2773	callback, free it. Also, use no error checking, as usual.	3035	callback, free it. Also, use no error checking, as usual.
2774		3036
2775	static void	3037	static void
2776	idle_cb (struct ev_loop *loop, ev_idle *w, int revents)	3038	idle_cb (struct ev_loop *loop, ev_idle *w, int revents)
2777	{	3039	{
		3040	// stop the watcher
		3041	ev_idle_stop (loop, w);
		3042
		3043	// now we can free it
2778	free (w);	3044	free (w);
		3045
2779	// now do something you wanted to do when the program has	3046	// now do something you wanted to do when the program has
2780	// no longer anything immediate to do.	3047	// no longer anything immediate to do.
2781	}	3048	}
2782		3049
2783	ev_idle *idle_watcher = malloc (sizeof (ev_idle));	3050	ev_idle *idle_watcher = malloc (sizeof (ev_idle));
…		…
2785	ev_idle_start (loop, idle_watcher);	3052	ev_idle_start (loop, idle_watcher);
2786		3053
2787		3054
2788	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!	3055	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!
2789		3056
2790	Prepare and check watchers are usually (but not always) used in pairs:	3057	Prepare and check watchers are often (but not always) used in pairs:
2791	prepare watchers get invoked before the process blocks and check watchers	3058	prepare watchers get invoked before the process blocks and check watchers
2792	afterwards.	3059	afterwards.
2793		3060
2794	You I<must not> call C<ev_run> or similar functions that enter	3061	You I<must not> call C<ev_run> (or similar functions that enter the
2795	the current event loop from either C<ev_prepare> or C<ev_check>	3062	current event loop) or C<ev_loop_fork> from either C<ev_prepare> or
2796	watchers. Other loops than the current one are fine, however. The	3063	C<ev_check> watchers. Other loops than the current one are fine,
2797	rationale behind this is that you do not need to check for recursion in	3064	however. The rationale behind this is that you do not need to check
2798	those watchers, i.e. the sequence will always be C<ev_prepare>, blocking,	3065	for recursion in those watchers, i.e. the sequence will always be
2799	C<ev_check> so if you have one watcher of each kind they will always be	3066	C<ev_prepare>, blocking, C<ev_check> so if you have one watcher of each
2800	called in pairs bracketing the blocking call.	3067	kind they will always be called in pairs bracketing the blocking call.
2801		3068
2802	Their main purpose is to integrate other event mechanisms into libev and	3069	Their main purpose is to integrate other event mechanisms into libev and
2803	their use is somewhat advanced. They could be used, for example, to track	3070	their use is somewhat advanced. They could be used, for example, to track
2804	variable changes, implement your own watchers, integrate net-snmp or a	3071	variable changes, implement your own watchers, integrate net-snmp or a
2805	coroutine library and lots more. They are also occasionally useful if	3072	coroutine library and lots more. They are also occasionally useful if
…		…
2823	with priority higher than or equal to the event loop and one coroutine	3090	with priority higher than or equal to the event loop and one coroutine
2824	of lower priority, but only once, using idle watchers to keep the event	3091	of lower priority, but only once, using idle watchers to keep the event
2825	loop from blocking if lower-priority coroutines are active, thus mapping	3092	loop from blocking if lower-priority coroutines are active, thus mapping
2826	low-priority coroutines to idle/background tasks).	3093	low-priority coroutines to idle/background tasks).
2827		3094
2828	It is recommended to give C<ev_check> watchers highest (C<EV_MAXPRI>)	3095	When used for this purpose, it is recommended to give C<ev_check> watchers
2829	priority, to ensure that they are being run before any other watchers	3096	highest (C<EV_MAXPRI>) priority, to ensure that they are being run before
2830	after the poll (this doesn't matter for C<ev_prepare> watchers).	3097	any other watchers after the poll (this doesn't matter for C<ev_prepare>
		3098	watchers).
2831		3099
2832	Also, C<ev_check> watchers (and C<ev_prepare> watchers, too) should not	3100	Also, C<ev_check> watchers (and C<ev_prepare> watchers, too) should not
2833	activate ("feed") events into libev. While libev fully supports this, they	3101	activate ("feed") events into libev. While libev fully supports this, they
2834	might get executed before other C<ev_check> watchers did their job. As	3102	might get executed before other C<ev_check> watchers did their job. As
2835	C<ev_check> watchers are often used to embed other (non-libev) event	3103	C<ev_check> watchers are often used to embed other (non-libev) event
2836	loops those other event loops might be in an unusable state until their	3104	loops those other event loops might be in an unusable state until their
2837	C<ev_check> watcher ran (always remind yourself to coexist peacefully with	3105	C<ev_check> watcher ran (always remind yourself to coexist peacefully with
2838	others).	3106	others).
		3107
		3108	=head3 Abusing an C<ev_check> watcher for its side-effect
		3109
		3110	C<ev_check> (and less often also C<ev_prepare>) watchers can also be
		3111	useful because they are called once per event loop iteration. For
		3112	example, if you want to handle a large number of connections fairly, you
		3113	normally only do a bit of work for each active connection, and if there
		3114	is more work to do, you wait for the next event loop iteration, so other
		3115	connections have a chance of making progress.
		3116
		3117	Using an C<ev_check> watcher is almost enough: it will be called on the
		3118	next event loop iteration. However, that isn't as soon as possible -
		3119	without external events, your C<ev_check> watcher will not be invoked.
		3120
		3121	This is where C<ev_idle> watchers come in handy - all you need is a
		3122	single global idle watcher that is active as long as you have one active
		3123	C<ev_check> watcher. The C<ev_idle> watcher makes sure the event loop
		3124	will not sleep, and the C<ev_check> watcher makes sure a callback gets
		3125	invoked. Neither watcher alone can do that.
2839		3126
2840	=head3 Watcher-Specific Functions and Data Members	3127	=head3 Watcher-Specific Functions and Data Members
2841		3128
2842	=over 4	3129	=over 4
2843		3130
…		…
3044		3331
3045	=over 4	3332	=over 4
3046		3333
3047	=item ev_embed_init (ev_embed *, callback, struct ev_loop *embedded_loop)	3334	=item ev_embed_init (ev_embed *, callback, struct ev_loop *embedded_loop)
3048		3335
3049	=item ev_embed_set (ev_embed *, callback, struct ev_loop *embedded_loop)	3336	=item ev_embed_set (ev_embed *, struct ev_loop *embedded_loop)
3050		3337
3051	Configures the watcher to embed the given loop, which must be	3338	Configures the watcher to embed the given loop, which must be
3052	embeddable. If the callback is C<0>, then C<ev_embed_sweep> will be	3339	embeddable. If the callback is C<0>, then C<ev_embed_sweep> will be
3053	invoked automatically, otherwise it is the responsibility of the callback	3340	invoked automatically, otherwise it is the responsibility of the callback
3054	to invoke it (it will continue to be called until the sweep has been done,	3341	to invoke it (it will continue to be called until the sweep has been done,
…		…
3075	used).	3362	used).
3076		3363
3077	struct ev_loop *loop_hi = ev_default_init (0);	3364	struct ev_loop *loop_hi = ev_default_init (0);
3078	struct ev_loop *loop_lo = 0;	3365	struct ev_loop *loop_lo = 0;
3079	ev_embed embed;	3366	ev_embed embed;
3080		3367
3081	// see if there is a chance of getting one that works	3368	// see if there is a chance of getting one that works
3082	// (remember that a flags value of 0 means autodetection)	3369	// (remember that a flags value of 0 means autodetection)
3083	loop_lo = ev_embeddable_backends () & ev_recommended_backends ()	3370	loop_lo = ev_embeddable_backends () & ev_recommended_backends ()
3084	? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())	3371	? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())
3085	: 0;	3372	: 0;
…		…
3099	C<loop_socket>. (One might optionally use C<EVFLAG_NOENV>, too).	3386	C<loop_socket>. (One might optionally use C<EVFLAG_NOENV>, too).
3100		3387
3101	struct ev_loop *loop = ev_default_init (0);	3388	struct ev_loop *loop = ev_default_init (0);
3102	struct ev_loop *loop_socket = 0;	3389	struct ev_loop *loop_socket = 0;
3103	ev_embed embed;	3390	ev_embed embed;
3104		3391
3105	if (ev_supported_backends () & ~ev_recommended_backends () & EVBACKEND_KQUEUE)	3392	if (ev_supported_backends () & ~ev_recommended_backends () & EVBACKEND_KQUEUE)
3106	if ((loop_socket = ev_loop_new (EVBACKEND_KQUEUE))	3393	if ((loop_socket = ev_loop_new (EVBACKEND_KQUEUE))
3107	{	3394	{
3108	ev_embed_init (&embed, 0, loop_socket);	3395	ev_embed_init (&embed, 0, loop_socket);
3109	ev_embed_start (loop, &embed);	3396	ev_embed_start (loop, &embed);
…		…
3117		3404
3118	=head2 C<ev_fork> - the audacity to resume the event loop after a fork	3405	=head2 C<ev_fork> - the audacity to resume the event loop after a fork
3119		3406
3120	Fork watchers are called when a C<fork ()> was detected (usually because	3407	Fork watchers are called when a C<fork ()> was detected (usually because
3121	whoever is a good citizen cared to tell libev about it by calling	3408	whoever is a good citizen cared to tell libev about it by calling
3122	C<ev_default_fork> or C<ev_loop_fork>). The invocation is done before the	3409	C<ev_loop_fork>). The invocation is done before the event loop blocks next
3123	event loop blocks next and before C<ev_check> watchers are being called,	3410	and before C<ev_check> watchers are being called, and only in the child
3124	and only in the child after the fork. If whoever good citizen calling	3411	after the fork. If whoever good citizen calling C<ev_default_fork> cheats
3125	C<ev_default_fork> cheats and calls it in the wrong process, the fork	3412	and calls it in the wrong process, the fork handlers will be invoked, too,
3126	handlers will be invoked, too, of course.	3413	of course.
3127		3414
3128	=head3 The special problem of life after fork - how is it possible?	3415	=head3 The special problem of life after fork - how is it possible?
3129		3416
3130	Most uses of C<fork()> consist of forking, then some simple calls to set	3417	Most uses of C<fork ()> consist of forking, then some simple calls to set
3131	up/change the process environment, followed by a call to C<exec()>. This	3418	up/change the process environment, followed by a call to C<exec()>. This
3132	sequence should be handled by libev without any problems.	3419	sequence should be handled by libev without any problems.
3133		3420
3134	This changes when the application actually wants to do event handling	3421	This changes when the application actually wants to do event handling
3135	in the child, or both parent in child, in effect "continuing" after the	3422	in the child, or both parent in child, in effect "continuing" after the
…		…
3212	atexit (program_exits);	3499	atexit (program_exits);
3213		3500
3214		3501
3215	=head2 C<ev_async> - how to wake up an event loop	3502	=head2 C<ev_async> - how to wake up an event loop
3216		3503
3217	In general, you cannot use an C<ev_run> from multiple threads or other	3504	In general, you cannot use an C<ev_loop> from multiple threads or other
3218	asynchronous sources such as signal handlers (as opposed to multiple event	3505	asynchronous sources such as signal handlers (as opposed to multiple event
3219	loops - those are of course safe to use in different threads).	3506	loops - those are of course safe to use in different threads).
3220		3507
3221	Sometimes, however, you need to wake up an event loop you do not control,	3508	Sometimes, however, you need to wake up an event loop you do not control,
3222	for example because it belongs to another thread. This is what C<ev_async>	3509	for example because it belongs to another thread. This is what C<ev_async>
…		…
3224	it by calling C<ev_async_send>, which is thread- and signal safe.	3511	it by calling C<ev_async_send>, which is thread- and signal safe.
3225		3512
3226	This functionality is very similar to C<ev_signal> watchers, as signals,	3513	This functionality is very similar to C<ev_signal> watchers, as signals,
3227	too, are asynchronous in nature, and signals, too, will be compressed	3514	too, are asynchronous in nature, and signals, too, will be compressed
3228	(i.e. the number of callback invocations may be less than the number of	3515	(i.e. the number of callback invocations may be less than the number of
3229	C<ev_async_sent> calls). In fact, you could use signal watchers as a kind	3516	C<ev_async_send> calls). In fact, you could use signal watchers as a kind
3230	of "global async watchers" by using a watcher on an otherwise unused	3517	of "global async watchers" by using a watcher on an otherwise unused
3231	signal, and C<ev_feed_signal> to signal this watcher from another thread,	3518	signal, and C<ev_feed_signal> to signal this watcher from another thread,
3232	even without knowing which loop owns the signal.	3519	even without knowing which loop owns the signal.
3233
3234	Unlike C<ev_signal> watchers, C<ev_async> works with any event loop, not
3235	just the default loop.
3236		3520
3237	=head3 Queueing	3521	=head3 Queueing
3238		3522
3239	C<ev_async> does not support queueing of data in any way. The reason	3523	C<ev_async> does not support queueing of data in any way. The reason
3240	is that the author does not know of a simple (or any) algorithm for a	3524	is that the author does not know of a simple (or any) algorithm for a
…		…
3332	trust me.	3616	trust me.
3333		3617
3334	=item ev_async_send (loop, ev_async *)	3618	=item ev_async_send (loop, ev_async *)
3335		3619
3336	Sends/signals/activates the given C<ev_async> watcher, that is, feeds	3620	Sends/signals/activates the given C<ev_async> watcher, that is, feeds
3337	an C<EV_ASYNC> event on the watcher into the event loop. Unlike	3621	an C<EV_ASYNC> event on the watcher into the event loop, and instantly
		3622	returns.
		3623
3338	C<ev_feed_event>, this call is safe to do from other threads, signal or	3624	Unlike C<ev_feed_event>, this call is safe to do from other threads,
3339	similar contexts (see the discussion of C<EV_ATOMIC_T> in the embedding	3625	signal or similar contexts (see the discussion of C<EV_ATOMIC_T> in the
3340	section below on what exactly this means).	3626	embedding section below on what exactly this means).
3341		3627
3342	Note that, as with other watchers in libev, multiple events might get	3628	Note that, as with other watchers in libev, multiple events might get
3343	compressed into a single callback invocation (another way to look at this	3629	compressed into a single callback invocation (another way to look at
3344	is that C<ev_async> watchers are level-triggered, set on C<ev_async_send>,	3630	this is that C<ev_async> watchers are level-triggered: they are set on
3345	reset when the event loop detects that).	3631	C<ev_async_send>, reset when the event loop detects that).
3346		3632
3347	This call incurs the overhead of a system call only once per event loop	3633	This call incurs the overhead of at most one extra system call per event
3348	iteration, so while the overhead might be noticeable, it doesn't apply to	3634	loop iteration, if the event loop is blocked, and no syscall at all if
3349	repeated calls to C<ev_async_send> for the same event loop.	3635	the event loop (or your program) is processing events. That means that
		3636	repeated calls are basically free (there is no need to avoid calls for
		3637	performance reasons) and that the overhead becomes smaller (typically
		3638	zero) under load.
3350		3639
3351	=item bool = ev_async_pending (ev_async *)	3640	=item bool = ev_async_pending (ev_async *)
3352		3641
3353	Returns a non-zero value when C<ev_async_send> has been called on the	3642	Returns a non-zero value when C<ev_async_send> has been called on the
3354	watcher but the event has not yet been processed (or even noted) by the	3643	watcher but the event has not yet been processed (or even noted) by the
…		…
3371		3660
3372	There are some other functions of possible interest. Described. Here. Now.	3661	There are some other functions of possible interest. Described. Here. Now.
3373		3662
3374	=over 4	3663	=over 4
3375		3664
3376	=item ev_once (loop, int fd, int events, ev_tstamp timeout, callback)	3665	=item ev_once (loop, int fd, int events, ev_tstamp timeout, callback, arg)
3377		3666
3378	This function combines a simple timer and an I/O watcher, calls your	3667	This function combines a simple timer and an I/O watcher, calls your
3379	callback on whichever event happens first and automatically stops both	3668	callback on whichever event happens first and automatically stops both
3380	watchers. This is useful if you want to wait for a single event on an fd	3669	watchers. This is useful if you want to wait for a single event on an fd
3381	or timeout without having to allocate/configure/start/stop/free one or	3670	or timeout without having to allocate/configure/start/stop/free one or
…		…
3409	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);	3698	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);
3410		3699
3411	=item ev_feed_fd_event (loop, int fd, int revents)	3700	=item ev_feed_fd_event (loop, int fd, int revents)
3412		3701
3413	Feed an event on the given fd, as if a file descriptor backend detected	3702	Feed an event on the given fd, as if a file descriptor backend detected
3414	the given events it.	3703	the given events.
3415		3704
3416	=item ev_feed_signal_event (loop, int signum)	3705	=item ev_feed_signal_event (loop, int signum)
3417		3706
3418	Feed an event as if the given signal occurred. See also C<ev_feed_signal>,	3707	Feed an event as if the given signal occurred. See also C<ev_feed_signal>,
3419	which is async-safe.	3708	which is async-safe.
…		…
3425		3714
3426	This section explains some common idioms that are not immediately	3715	This section explains some common idioms that are not immediately
3427	obvious. Note that examples are sprinkled over the whole manual, and this	3716	obvious. Note that examples are sprinkled over the whole manual, and this
3428	section only contains stuff that wouldn't fit anywhere else.	3717	section only contains stuff that wouldn't fit anywhere else.
3429		3718
3430	=over 4	3719	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER
3431		3720
3432	=item Model/nested event loop invocations and exit conditions.	3721	Each watcher has, by default, a C<void *data> member that you can read
		3722	or modify at any time: libev will completely ignore it. This can be used
		3723	to associate arbitrary data with your watcher. If you need more data and
		3724	don't want to allocate memory separately and store a pointer to it in that
		3725	data member, you can also "subclass" the watcher type and provide your own
		3726	data:
		3727
		3728	struct my_io
		3729	{
		3730	ev_io io;
		3731	int otherfd;
		3732	void *somedata;
		3733	struct whatever *mostinteresting;
		3734	};
		3735
		3736	...
		3737	struct my_io w;
		3738	ev_io_init (&w.io, my_cb, fd, EV_READ);
		3739
		3740	And since your callback will be called with a pointer to the watcher, you
		3741	can cast it back to your own type:
		3742
		3743	static void my_cb (struct ev_loop *loop, ev_io *w_, int revents)
		3744	{
		3745	struct my_io *w = (struct my_io *)w_;
		3746	...
		3747	}
		3748
		3749	More interesting and less C-conformant ways of casting your callback
		3750	function type instead have been omitted.
		3751
		3752	=head2 BUILDING YOUR OWN COMPOSITE WATCHERS
		3753
		3754	Another common scenario is to use some data structure with multiple
		3755	embedded watchers, in effect creating your own watcher that combines
		3756	multiple libev event sources into one "super-watcher":
		3757
		3758	struct my_biggy
		3759	{
		3760	int some_data;
		3761	ev_timer t1;
		3762	ev_timer t2;
		3763	}
		3764
		3765	In this case getting the pointer to C<my_biggy> is a bit more
		3766	complicated: Either you store the address of your C<my_biggy> struct in
		3767	the C<data> member of the watcher (for woozies or C++ coders), or you need
		3768	to use some pointer arithmetic using C<offsetof> inside your watchers (for
		3769	real programmers):
		3770
		3771	#include <stddef.h>
		3772
		3773	static void
		3774	t1_cb (EV_P_ ev_timer *w, int revents)
		3775	{
		3776	struct my_biggy big = (struct my_biggy *)
		3777	(((char *)w) - offsetof (struct my_biggy, t1));
		3778	}
		3779
		3780	static void
		3781	t2_cb (EV_P_ ev_timer *w, int revents)
		3782	{
		3783	struct my_biggy big = (struct my_biggy *)
		3784	(((char *)w) - offsetof (struct my_biggy, t2));
		3785	}
		3786
		3787	=head2 AVOIDING FINISHING BEFORE RETURNING
		3788
		3789	Often you have structures like this in event-based programs:
		3790
		3791	callback ()
		3792	{
		3793	free (request);
		3794	}
		3795
		3796	request = start_new_request (..., callback);
		3797
		3798	The intent is to start some "lengthy" operation. The C<request> could be
		3799	used to cancel the operation, or do other things with it.
		3800
		3801	It's not uncommon to have code paths in C<start_new_request> that
		3802	immediately invoke the callback, for example, to report errors. Or you add
		3803	some caching layer that finds that it can skip the lengthy aspects of the
		3804	operation and simply invoke the callback with the result.
		3805
		3806	The problem here is that this will happen I<before> C<start_new_request>
		3807	has returned, so C<request> is not set.
		3808
		3809	Even if you pass the request by some safer means to the callback, you
		3810	might want to do something to the request after starting it, such as
		3811	canceling it, which probably isn't working so well when the callback has
		3812	already been invoked.
		3813
		3814	A common way around all these issues is to make sure that
		3815	C<start_new_request> I<always> returns before the callback is invoked. If
		3816	C<start_new_request> immediately knows the result, it can artificially
		3817	delay invoking the callback by using a C<prepare> or C<idle> watcher for
		3818	example, or more sneakily, by reusing an existing (stopped) watcher and
		3819	pushing it into the pending queue:
		3820
		3821	ev_set_cb (watcher, callback);
		3822	ev_feed_event (EV_A_ watcher, 0);
		3823
		3824	This way, C<start_new_request> can safely return before the callback is
		3825	invoked, while not delaying callback invocation too much.
		3826
		3827	=head2 MODEL/NESTED EVENT LOOP INVOCATIONS AND EXIT CONDITIONS
3433		3828
3434	Often (especially in GUI toolkits) there are places where you have	3829	Often (especially in GUI toolkits) there are places where you have
3435	I<modal> interaction, which is most easily implemented by recursively	3830	I<modal> interaction, which is most easily implemented by recursively
3436	invoking C<ev_run>.	3831	invoking C<ev_run>.
3437		3832
3438	This brings the problem of exiting - a callback might want to finish the	3833	This brings the problem of exiting - a callback might want to finish the
3439	main C<ev_run> call, but not the nested one (e.g. user clicked "Quit", but	3834	main C<ev_run> call, but not the nested one (e.g. user clicked "Quit", but
3440	a modal "Are you sure?" dialog is still waiting), or just the nested one	3835	a modal "Are you sure?" dialog is still waiting), or just the nested one
3441	and not the main one (e.g. user clocked "Ok" in a modal dialog), or some	3836	and not the main one (e.g. user clocked "Ok" in a modal dialog), or some
3442	other combination: In these cases, C<ev_break> will not work alone.	3837	other combination: In these cases, a simple C<ev_break> will not work.
3443		3838
3444	The solution is to maintain "break this loop" variable for each C<ev_run>	3839	The solution is to maintain "break this loop" variable for each C<ev_run>
3445	invocation, and use a loop around C<ev_run> until the condition is	3840	invocation, and use a loop around C<ev_run> until the condition is
3446	triggered, using C<EVRUN_ONCE>:	3841	triggered, using C<EVRUN_ONCE>:
3447		3842
…		…
3449	int exit_main_loop = 0;	3844	int exit_main_loop = 0;
3450		3845
3451	while (!exit_main_loop)	3846	while (!exit_main_loop)
3452	ev_run (EV_DEFAULT_ EVRUN_ONCE);	3847	ev_run (EV_DEFAULT_ EVRUN_ONCE);
3453		3848
3454	// in a model watcher	3849	// in a modal watcher
3455	int exit_nested_loop = 0;	3850	int exit_nested_loop = 0;
3456		3851
3457	while (!exit_nested_loop)	3852	while (!exit_nested_loop)
3458	ev_run (EV_A_ EVRUN_ONCE);	3853	ev_run (EV_A_ EVRUN_ONCE);
3459		3854
…		…
3466	exit_main_loop = 1;	3861	exit_main_loop = 1;
3467		3862
3468	// exit both	3863	// exit both
3469	exit_main_loop = exit_nested_loop = 1;	3864	exit_main_loop = exit_nested_loop = 1;
3470		3865
3471	=back	3866	=head2 THREAD LOCKING EXAMPLE
		3867
		3868	Here is a fictitious example of how to run an event loop in a different
		3869	thread from where callbacks are being invoked and watchers are
		3870	created/added/removed.
		3871
		3872	For a real-world example, see the C<EV::Loop::Async> perl module,
		3873	which uses exactly this technique (which is suited for many high-level
		3874	languages).
		3875
		3876	The example uses a pthread mutex to protect the loop data, a condition
		3877	variable to wait for callback invocations, an async watcher to notify the
		3878	event loop thread and an unspecified mechanism to wake up the main thread.
		3879
		3880	First, you need to associate some data with the event loop:
		3881
		3882	typedef struct {
		3883	pthread_mutex_t lock; /* global loop lock */
		3884	pthread_t tid;
		3885	pthread_cond_t invoke_cv;
		3886	ev_async async_w;
		3887	} userdata;
		3888
		3889	void prepare_loop (EV_P)
		3890	{
		3891	// for simplicity, we use a static userdata struct.
		3892	static userdata u;
		3893
		3894	ev_async_init (&u.async_w, async_cb);
		3895	ev_async_start (EV_A_ &u.async_w);
		3896
		3897	pthread_mutex_init (&u.lock, 0);
		3898	pthread_cond_init (&u.invoke_cv, 0);
		3899
		3900	// now associate this with the loop
		3901	ev_set_userdata (EV_A_ &u);
		3902	ev_set_invoke_pending_cb (EV_A_ l_invoke);
		3903	ev_set_loop_release_cb (EV_A_ l_release, l_acquire);
		3904
		3905	// then create the thread running ev_run
		3906	pthread_create (&u.tid, 0, l_run, EV_A);
		3907	}
		3908
		3909	The callback for the C<ev_async> watcher does nothing: the watcher is used
		3910	solely to wake up the event loop so it takes notice of any new watchers
		3911	that might have been added:
		3912
		3913	static void
		3914	async_cb (EV_P_ ev_async *w, int revents)
		3915	{
		3916	// just used for the side effects
		3917	}
		3918
		3919	The C<l_release> and C<l_acquire> callbacks simply unlock/lock the mutex
		3920	protecting the loop data, respectively.
		3921
		3922	static void
		3923	l_release (EV_P)
		3924	{
		3925	userdata *u = ev_userdata (EV_A);
		3926	pthread_mutex_unlock (&u->lock);
		3927	}
		3928
		3929	static void
		3930	l_acquire (EV_P)
		3931	{
		3932	userdata *u = ev_userdata (EV_A);
		3933	pthread_mutex_lock (&u->lock);
		3934	}
		3935
		3936	The event loop thread first acquires the mutex, and then jumps straight
		3937	into C<ev_run>:
		3938
		3939	void *
		3940	l_run (void *thr_arg)
		3941	{
		3942	struct ev_loop *loop = (struct ev_loop *)thr_arg;
		3943
		3944	l_acquire (EV_A);
		3945	pthread_setcanceltype (PTHREAD_CANCEL_ASYNCHRONOUS, 0);
		3946	ev_run (EV_A_ 0);
		3947	l_release (EV_A);
		3948
		3949	return 0;
		3950	}
		3951
		3952	Instead of invoking all pending watchers, the C<l_invoke> callback will
		3953	signal the main thread via some unspecified mechanism (signals? pipe
		3954	writes? C<Async::Interrupt>?) and then waits until all pending watchers
		3955	have been called (in a while loop because a) spurious wakeups are possible
		3956	and b) skipping inter-thread-communication when there are no pending
		3957	watchers is very beneficial):
		3958
		3959	static void
		3960	l_invoke (EV_P)
		3961	{
		3962	userdata *u = ev_userdata (EV_A);
		3963
		3964	while (ev_pending_count (EV_A))
		3965	{
		3966	wake_up_other_thread_in_some_magic_or_not_so_magic_way ();
		3967	pthread_cond_wait (&u->invoke_cv, &u->lock);
		3968	}
		3969	}
		3970
		3971	Now, whenever the main thread gets told to invoke pending watchers, it
		3972	will grab the lock, call C<ev_invoke_pending> and then signal the loop
		3973	thread to continue:
		3974
		3975	static void
		3976	real_invoke_pending (EV_P)
		3977	{
		3978	userdata *u = ev_userdata (EV_A);
		3979
		3980	pthread_mutex_lock (&u->lock);
		3981	ev_invoke_pending (EV_A);
		3982	pthread_cond_signal (&u->invoke_cv);
		3983	pthread_mutex_unlock (&u->lock);
		3984	}
		3985
		3986	Whenever you want to start/stop a watcher or do other modifications to an
		3987	event loop, you will now have to lock:
		3988
		3989	ev_timer timeout_watcher;
		3990	userdata *u = ev_userdata (EV_A);
		3991
		3992	ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);
		3993
		3994	pthread_mutex_lock (&u->lock);
		3995	ev_timer_start (EV_A_ &timeout_watcher);
		3996	ev_async_send (EV_A_ &u->async_w);
		3997	pthread_mutex_unlock (&u->lock);
		3998
		3999	Note that sending the C<ev_async> watcher is required because otherwise
		4000	an event loop currently blocking in the kernel will have no knowledge
		4001	about the newly added timer. By waking up the loop it will pick up any new
		4002	watchers in the next event loop iteration.
		4003
		4004	=head2 THREADS, COROUTINES, CONTINUATIONS, QUEUES... INSTEAD OF CALLBACKS
		4005
		4006	While the overhead of a callback that e.g. schedules a thread is small, it
		4007	is still an overhead. If you embed libev, and your main usage is with some
		4008	kind of threads or coroutines, you might want to customise libev so that
		4009	doesn't need callbacks anymore.
		4010
		4011	Imagine you have coroutines that you can switch to using a function
		4012	C<switch_to (coro)>, that libev runs in a coroutine called C<libev_coro>
		4013	and that due to some magic, the currently active coroutine is stored in a
		4014	global called C<current_coro>. Then you can build your own "wait for libev
		4015	event" primitive by changing C<EV_CB_DECLARE> and C<EV_CB_INVOKE> (note
		4016	the differing C<;> conventions):
		4017
		4018	#define EV_CB_DECLARE(type) struct my_coro *cb;
		4019	#define EV_CB_INVOKE(watcher) switch_to ((watcher)->cb)
		4020
		4021	That means instead of having a C callback function, you store the
		4022	coroutine to switch to in each watcher, and instead of having libev call
		4023	your callback, you instead have it switch to that coroutine.
		4024
		4025	A coroutine might now wait for an event with a function called
		4026	C<wait_for_event>. (the watcher needs to be started, as always, but it doesn't
		4027	matter when, or whether the watcher is active or not when this function is
		4028	called):
		4029
		4030	void
		4031	wait_for_event (ev_watcher *w)
		4032	{
		4033	ev_set_cb (w, current_coro);
		4034	switch_to (libev_coro);
		4035	}
		4036
		4037	That basically suspends the coroutine inside C<wait_for_event> and
		4038	continues the libev coroutine, which, when appropriate, switches back to
		4039	this or any other coroutine.
		4040
		4041	You can do similar tricks if you have, say, threads with an event queue -
		4042	instead of storing a coroutine, you store the queue object and instead of
		4043	switching to a coroutine, you push the watcher onto the queue and notify
		4044	any waiters.
		4045
		4046	To embed libev, see L</EMBEDDING>, but in short, it's easiest to create two
		4047	files, F<my_ev.h> and F<my_ev.c> that include the respective libev files:
		4048
		4049	// my_ev.h
		4050	#define EV_CB_DECLARE(type) struct my_coro *cb;
		4051	#define EV_CB_INVOKE(watcher) switch_to ((watcher)->cb)
		4052	#include "../libev/ev.h"
		4053
		4054	// my_ev.c
		4055	#define EV_H "my_ev.h"
		4056	#include "../libev/ev.c"
		4057
		4058	And then use F<my_ev.h> when you would normally use F<ev.h>, and compile
		4059	F<my_ev.c> into your project. When properly specifying include paths, you
		4060	can even use F<ev.h> as header file name directly.
3472		4061
3473		4062
3474	=head1 LIBEVENT EMULATION	4063	=head1 LIBEVENT EMULATION
3475		4064
3476	Libev offers a compatibility emulation layer for libevent. It cannot	4065	Libev offers a compatibility emulation layer for libevent. It cannot
…		…
3506		4095
3507	=back	4096	=back
3508		4097
3509	=head1 C++ SUPPORT	4098	=head1 C++ SUPPORT
3510		4099
		4100	=head2 C API
		4101
		4102	The normal C API should work fine when used from C++: both ev.h and the
		4103	libev sources can be compiled as C++. Therefore, code that uses the C API
		4104	will work fine.
		4105
		4106	Proper exception specifications might have to be added to callbacks passed
		4107	to libev: exceptions may be thrown only from watcher callbacks, all other
		4108	callbacks (allocator, syserr, loop acquire/release and periodic reschedule
		4109	callbacks) must not throw exceptions, and might need a C<noexcept>
		4110	specification. If you have code that needs to be compiled as both C and
		4111	C++ you can use the C<EV_NOEXCEPT> macro for this:
		4112
		4113	static void
		4114	fatal_error (const char *msg) EV_NOEXCEPT
		4115	{
		4116	perror (msg);
		4117	abort ();
		4118	}
		4119
		4120	...
		4121	ev_set_syserr_cb (fatal_error);
		4122
		4123	The only API functions that can currently throw exceptions are C<ev_run>,
		4124	C<ev_invoke>, C<ev_invoke_pending> and C<ev_loop_destroy> (the latter
		4125	because it runs cleanup watchers).
		4126
		4127	Throwing exceptions in watcher callbacks is only supported if libev itself
		4128	is compiled with a C++ compiler or your C and C++ environments allow
		4129	throwing exceptions through C libraries (most do).
		4130
		4131	=head2 C++ API
		4132
3511	Libev comes with some simplistic wrapper classes for C++ that mainly allow	4133	Libev comes with some simplistic wrapper classes for C++ that mainly allow
3512	you to use some convenience methods to start/stop watchers and also change	4134	you to use some convenience methods to start/stop watchers and also change
3513	the callback model to a model using method callbacks on objects.	4135	the callback model to a model using method callbacks on objects.
3514		4136
3515	To use it,	4137	To use it,
3516		4138
3517	#include <ev++.h>	4139	#include <ev++.h>
3518		4140
3519	This automatically includes F<ev.h> and puts all of its definitions (many	4141	This automatically includes F<ev.h> and puts all of its definitions (many
3520	of them macros) into the global namespace. All C++ specific things are	4142	of them macros) into the global namespace. All C++ specific things are
3521	put into the C<ev> namespace. It should support all the same embedding	4143	put into the C<ev> namespace. It should support all the same embedding
…		…
3530	with C<operator ()> can be used as callbacks. Other types should be easy	4152	with C<operator ()> can be used as callbacks. Other types should be easy
3531	to add as long as they only need one additional pointer for context. If	4153	to add as long as they only need one additional pointer for context. If
3532	you need support for other types of functors please contact the author	4154	you need support for other types of functors please contact the author
3533	(preferably after implementing it).	4155	(preferably after implementing it).
3534		4156
		4157	For all this to work, your C++ compiler either has to use the same calling
		4158	conventions as your C compiler (for static member functions), or you have
		4159	to embed libev and compile libev itself as C++.
		4160
3535	Here is a list of things available in the C<ev> namespace:	4161	Here is a list of things available in the C<ev> namespace:
3536		4162
3537	=over 4	4163	=over 4
3538		4164
3539	=item C<ev::READ>, C<ev::WRITE> etc.	4165	=item C<ev::READ>, C<ev::WRITE> etc.
…		…
3548	=item C<ev::io>, C<ev::timer>, C<ev::periodic>, C<ev::idle>, C<ev::sig> etc.	4174	=item C<ev::io>, C<ev::timer>, C<ev::periodic>, C<ev::idle>, C<ev::sig> etc.
3549		4175
3550	For each C<ev_TYPE> watcher in F<ev.h> there is a corresponding class of	4176	For each C<ev_TYPE> watcher in F<ev.h> there is a corresponding class of
3551	the same name in the C<ev> namespace, with the exception of C<ev_signal>	4177	the same name in the C<ev> namespace, with the exception of C<ev_signal>
3552	which is called C<ev::sig> to avoid clashes with the C<signal> macro	4178	which is called C<ev::sig> to avoid clashes with the C<signal> macro
3553	defines by many implementations.	4179	defined by many implementations.
3554		4180
3555	All of those classes have these methods:	4181	All of those classes have these methods:
3556		4182
3557	=over 4	4183	=over 4
3558		4184
…		…
3620	void operator() (ev::io &w, int revents)	4246	void operator() (ev::io &w, int revents)
3621	{	4247	{
3622	...	4248	...
3623	}	4249	}
3624	}	4250	}
3625		4251
3626	myfunctor f;	4252	myfunctor f;
3627		4253
3628	ev::io w;	4254	ev::io w;
3629	w.set (&f);	4255	w.set (&f);
3630		4256
…		…
3648	Associates a different C<struct ev_loop> with this watcher. You can only	4274	Associates a different C<struct ev_loop> with this watcher. You can only
3649	do this when the watcher is inactive (and not pending either).	4275	do this when the watcher is inactive (and not pending either).
3650		4276
3651	=item w->set ([arguments])	4277	=item w->set ([arguments])
3652		4278
3653	Basically the same as C<ev_TYPE_set>, with the same arguments. Either this	4279	Basically the same as C<ev_TYPE_set> (except for C<ev::embed> watchers>),
3654	method or a suitable start method must be called at least once. Unlike the	4280	with the same arguments. Either this method or a suitable start method
3655	C counterpart, an active watcher gets automatically stopped and restarted	4281	must be called at least once. Unlike the C counterpart, an active watcher
3656	when reconfiguring it with this method.	4282	gets automatically stopped and restarted when reconfiguring it with this
		4283	method.
		4284
		4285	For C<ev::embed> watchers this method is called C<set_embed>, to avoid
		4286	clashing with the C<set (loop)> method.
		4287
		4288	For C<ev::io> watchers there is an additional C<set> method that acepts a
		4289	new event mask only, and internally calls C<ev_io_modify>.
3657		4290
3658	=item w->start ()	4291	=item w->start ()
3659		4292
3660	Starts the watcher. Note that there is no C<loop> argument, as the	4293	Starts the watcher. Note that there is no C<loop> argument, as the
3661	constructor already stores the event loop.	4294	constructor already stores the event loop.
…		…
3691	watchers in the constructor.	4324	watchers in the constructor.
3692		4325
3693	class myclass	4326	class myclass
3694	{	4327	{
3695	ev::io io ; void io_cb (ev::io &w, int revents);	4328	ev::io io ; void io_cb (ev::io &w, int revents);
3696	ev::io2 io2 ; void io2_cb (ev::io &w, int revents);	4329	ev::io io2 ; void io2_cb (ev::io &w, int revents);
3697	ev::idle idle; void idle_cb (ev::idle &w, int revents);	4330	ev::idle idle; void idle_cb (ev::idle &w, int revents);
3698		4331
3699	myclass (int fd)	4332	myclass (int fd)
3700	{	4333	{
3701	io .set <myclass, &myclass::io_cb > (this);	4334	io .set <myclass, &myclass::io_cb > (this);
…		…
3752	L<http://hackage.haskell.org/cgi-bin/hackage-scripts/package/hlibev>.	4385	L<http://hackage.haskell.org/cgi-bin/hackage-scripts/package/hlibev>.
3753		4386
3754	=item D	4387	=item D
3755		4388
3756	Leandro Lucarella has written a D language binding (F<ev.d>) for libev, to	4389	Leandro Lucarella has written a D language binding (F<ev.d>) for libev, to
3757	be found at L<http://proj.llucax.com.ar/wiki/evd>.	4390	be found at L<http://www.llucax.com.ar/proj/ev.d/index.html>.
3758		4391
3759	=item Ocaml	4392	=item Ocaml
3760		4393
3761	Erkki Seppala has written Ocaml bindings for libev, to be found at	4394	Erkki Seppala has written Ocaml bindings for libev, to be found at
3762	L<http://modeemi.cs.tut.fi/~flux/software/ocaml-ev/>.	4395	L<http://modeemi.cs.tut.fi/~flux/software/ocaml-ev/>.
…		…
3765		4398
3766	Brian Maher has written a partial interface to libev for lua (at the	4399	Brian Maher has written a partial interface to libev for lua (at the
3767	time of this writing, only C<ev_io> and C<ev_timer>), to be found at	4400	time of this writing, only C<ev_io> and C<ev_timer>), to be found at
3768	L<http://github.com/brimworks/lua-ev>.	4401	L<http://github.com/brimworks/lua-ev>.
3769		4402
		4403	=item Javascript
		4404
		4405	Node.js (L<http://nodejs.org>) uses libev as the underlying event library.
		4406
		4407	=item Others
		4408
		4409	There are others, and I stopped counting.
		4410
3770	=back	4411	=back
3771		4412
3772		4413
3773	=head1 MACRO MAGIC	4414	=head1 MACRO MAGIC
3774		4415
…		…
3810	suitable for use with C<EV_A>.	4451	suitable for use with C<EV_A>.
3811		4452
3812	=item C<EV_DEFAULT>, C<EV_DEFAULT_>	4453	=item C<EV_DEFAULT>, C<EV_DEFAULT_>
3813		4454
3814	Similar to the other two macros, this gives you the value of the default	4455	Similar to the other two macros, this gives you the value of the default
3815	loop, if multiple loops are supported ("ev loop default").	4456	loop, if multiple loops are supported ("ev loop default"). The default loop
		4457	will be initialised if it isn't already initialised.
		4458
		4459	For non-multiplicity builds, these macros do nothing, so you always have
		4460	to initialise the loop somewhere.
3816		4461
3817	=item C<EV_DEFAULT_UC>, C<EV_DEFAULT_UC_>	4462	=item C<EV_DEFAULT_UC>, C<EV_DEFAULT_UC_>
3818		4463
3819	Usage identical to C<EV_DEFAULT> and C<EV_DEFAULT_>, but requires that the	4464	Usage identical to C<EV_DEFAULT> and C<EV_DEFAULT_>, but requires that the
3820	default loop has been initialised (C<UC> == unchecked). Their behaviour	4465	default loop has been initialised (C<UC> == unchecked). Their behaviour
…		…
3887	ev_vars.h	4532	ev_vars.h
3888	ev_wrap.h	4533	ev_wrap.h
3889		4534
3890	ev_win32.c required on win32 platforms only	4535	ev_win32.c required on win32 platforms only
3891		4536
3892	ev_select.c only when select backend is enabled (which is enabled by default)	4537	ev_select.c only when select backend is enabled
3893	ev_poll.c only when poll backend is enabled (disabled by default)	4538	ev_poll.c only when poll backend is enabled
3894	ev_epoll.c only when the epoll backend is enabled (disabled by default)	4539	ev_epoll.c only when the epoll backend is enabled
		4540	ev_linuxaio.c only when the linux aio backend is enabled
		4541	ev_iouring.c only when the linux io_uring backend is enabled
3895	ev_kqueue.c only when the kqueue backend is enabled (disabled by default)	4542	ev_kqueue.c only when the kqueue backend is enabled
3896	ev_port.c only when the solaris port backend is enabled (disabled by default)	4543	ev_port.c only when the solaris port backend is enabled
3897		4544
3898	F<ev.c> includes the backend files directly when enabled, so you only need	4545	F<ev.c> includes the backend files directly when enabled, so you only need
3899	to compile this single file.	4546	to compile this single file.
3900		4547
3901	=head3 LIBEVENT COMPATIBILITY API	4548	=head3 LIBEVENT COMPATIBILITY API
…		…
3965	supported). It will also not define any of the structs usually found in	4612	supported). It will also not define any of the structs usually found in
3966	F<event.h> that are not directly supported by the libev core alone.	4613	F<event.h> that are not directly supported by the libev core alone.
3967		4614
3968	In standalone mode, libev will still try to automatically deduce the	4615	In standalone mode, libev will still try to automatically deduce the
3969	configuration, but has to be more conservative.	4616	configuration, but has to be more conservative.
		4617
		4618	=item EV_USE_FLOOR
		4619
		4620	If defined to be C<1>, libev will use the C<floor ()> function for its
		4621	periodic reschedule calculations, otherwise libev will fall back on a
		4622	portable (slower) implementation. If you enable this, you usually have to
		4623	link against libm or something equivalent. Enabling this when the C<floor>
		4624	function is not available will fail, so the safe default is to not enable
		4625	this.
3970		4626
3971	=item EV_USE_MONOTONIC	4627	=item EV_USE_MONOTONIC
3972		4628
3973	If defined to be C<1>, libev will try to detect the availability of the	4629	If defined to be C<1>, libev will try to detect the availability of the
3974	monotonic clock option at both compile time and runtime. Otherwise no	4630	monotonic clock option at both compile time and runtime. Otherwise no
…		…
4011	available and will probe for kernel support at runtime. This will improve	4667	available and will probe for kernel support at runtime. This will improve
4012	C<ev_signal> and C<ev_async> performance and reduce resource consumption.	4668	C<ev_signal> and C<ev_async> performance and reduce resource consumption.
4013	If undefined, it will be enabled if the headers indicate GNU/Linux + Glibc	4669	If undefined, it will be enabled if the headers indicate GNU/Linux + Glibc
4014	2.7 or newer, otherwise disabled.	4670	2.7 or newer, otherwise disabled.
4015		4671
		4672	=item EV_USE_SIGNALFD
		4673
		4674	If defined to be C<1>, then libev will assume that C<signalfd ()> is
		4675	available and will probe for kernel support at runtime. This enables
		4676	the use of EVFLAG_SIGNALFD for faster and simpler signal handling. If
		4677	undefined, it will be enabled if the headers indicate GNU/Linux + Glibc
		4678	2.7 or newer, otherwise disabled.
		4679
		4680	=item EV_USE_TIMERFD
		4681
		4682	If defined to be C<1>, then libev will assume that C<timerfd ()> is
		4683	available and will probe for kernel support at runtime. This allows
		4684	libev to detect time jumps accurately. If undefined, it will be enabled
		4685	if the headers indicate GNU/Linux + Glibc 2.8 or newer and define
		4686	C<TFD_TIMER_CANCEL_ON_SET>, otherwise disabled.
		4687
		4688	=item EV_USE_EVENTFD
		4689
		4690	If defined to be C<1>, then libev will assume that C<eventfd ()> is
		4691	available and will probe for kernel support at runtime. This will improve
		4692	C<ev_signal> and C<ev_async> performance and reduce resource consumption.
		4693	If undefined, it will be enabled if the headers indicate GNU/Linux + Glibc
		4694	2.7 or newer, otherwise disabled.
		4695
4016	=item EV_USE_SELECT	4696	=item EV_USE_SELECT
4017		4697
4018	If undefined or defined to be C<1>, libev will compile in support for the	4698	If undefined or defined to be C<1>, libev will compile in support for the
4019	C<select>(2) backend. No attempt at auto-detection will be done: if no	4699	C<select>(2) backend. No attempt at auto-detection will be done: if no
4020	other method takes over, select will be it. Otherwise the select backend	4700	other method takes over, select will be it. Otherwise the select backend
…		…
4060	If programs implement their own fd to handle mapping on win32, then this	4740	If programs implement their own fd to handle mapping on win32, then this
4061	macro can be used to override the C<close> function, useful to unregister	4741	macro can be used to override the C<close> function, useful to unregister
4062	file descriptors again. Note that the replacement function has to close	4742	file descriptors again. Note that the replacement function has to close
4063	the underlying OS handle.	4743	the underlying OS handle.
4064		4744
		4745	=item EV_USE_WSASOCKET
		4746
		4747	If defined to be C<1>, libev will use C<WSASocket> to create its internal
		4748	communication socket, which works better in some environments. Otherwise,
		4749	the normal C<socket> function will be used, which works better in other
		4750	environments.
		4751
4065	=item EV_USE_POLL	4752	=item EV_USE_POLL
4066		4753
4067	If defined to be C<1>, libev will compile in support for the C<poll>(2)	4754	If defined to be C<1>, libev will compile in support for the C<poll>(2)
4068	backend. Otherwise it will be enabled on non-win32 platforms. It	4755	backend. Otherwise it will be enabled on non-win32 platforms. It
4069	takes precedence over select.	4756	takes precedence over select.
…		…
4073	If defined to be C<1>, libev will compile in support for the Linux	4760	If defined to be C<1>, libev will compile in support for the Linux
4074	C<epoll>(7) backend. Its availability will be detected at runtime,	4761	C<epoll>(7) backend. Its availability will be detected at runtime,
4075	otherwise another method will be used as fallback. This is the preferred	4762	otherwise another method will be used as fallback. This is the preferred
4076	backend for GNU/Linux systems. If undefined, it will be enabled if the	4763	backend for GNU/Linux systems. If undefined, it will be enabled if the
4077	headers indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.	4764	headers indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.
		4765
		4766	=item EV_USE_LINUXAIO
		4767
		4768	If defined to be C<1>, libev will compile in support for the Linux aio
		4769	backend (C<EV_USE_EPOLL> must also be enabled). If undefined, it will be
		4770	enabled on linux, otherwise disabled.
		4771
		4772	=item EV_USE_IOURING
		4773
		4774	If defined to be C<1>, libev will compile in support for the Linux
		4775	io_uring backend (C<EV_USE_EPOLL> must also be enabled). Due to it's
		4776	current limitations it has to be requested explicitly. If undefined, it
		4777	will be enabled on linux, otherwise disabled.
4078		4778
4079	=item EV_USE_KQUEUE	4779	=item EV_USE_KQUEUE
4080		4780
4081	If defined to be C<1>, libev will compile in support for the BSD style	4781	If defined to be C<1>, libev will compile in support for the BSD style
4082	C<kqueue>(2) backend. Its actual availability will be detected at runtime,	4782	C<kqueue>(2) backend. Its actual availability will be detected at runtime,
…		…
4104	If defined to be C<1>, libev will compile in support for the Linux inotify	4804	If defined to be C<1>, libev will compile in support for the Linux inotify
4105	interface to speed up C<ev_stat> watchers. Its actual availability will	4805	interface to speed up C<ev_stat> watchers. Its actual availability will
4106	be detected at runtime. If undefined, it will be enabled if the headers	4806	be detected at runtime. If undefined, it will be enabled if the headers
4107	indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.	4807	indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.
4108		4808
		4809	=item EV_NO_SMP
		4810
		4811	If defined to be C<1>, libev will assume that memory is always coherent
		4812	between threads, that is, threads can be used, but threads never run on
		4813	different cpus (or different cpu cores). This reduces dependencies
		4814	and makes libev faster.
		4815
		4816	=item EV_NO_THREADS
		4817
		4818	If defined to be C<1>, libev will assume that it will never be called from
		4819	different threads (that includes signal handlers), which is a stronger
		4820	assumption than C<EV_NO_SMP>, above. This reduces dependencies and makes
		4821	libev faster.
		4822
4109	=item EV_ATOMIC_T	4823	=item EV_ATOMIC_T
4110		4824
4111	Libev requires an integer type (suitable for storing C<0> or C<1>) whose	4825	Libev requires an integer type (suitable for storing C<0> or C<1>) whose
4112	access is atomic with respect to other threads or signal contexts. No such	4826	access is atomic with respect to other threads or signal contexts. No
4113	type is easily found in the C language, so you can provide your own type	4827	such type is easily found in the C language, so you can provide your own
4114	that you know is safe for your purposes. It is used both for signal handler "locking"	4828	type that you know is safe for your purposes. It is used both for signal
4115	as well as for signal and thread safety in C<ev_async> watchers.	4829	handler "locking" as well as for signal and thread safety in C<ev_async>
		4830	watchers.
4116		4831
4117	In the absence of this define, libev will use C<sig_atomic_t volatile>	4832	In the absence of this define, libev will use C<sig_atomic_t volatile>
4118	(from F<signal.h>), which is usually good enough on most platforms.	4833	(from F<signal.h>), which is usually good enough on most platforms.
4119		4834
4120	=item EV_H (h)	4835	=item EV_H (h)
…		…
4147	will have the C<struct ev_loop *> as first argument, and you can create	4862	will have the C<struct ev_loop *> as first argument, and you can create
4148	additional independent event loops. Otherwise there will be no support	4863	additional independent event loops. Otherwise there will be no support
4149	for multiple event loops and there is no first event loop pointer	4864	for multiple event loops and there is no first event loop pointer
4150	argument. Instead, all functions act on the single default loop.	4865	argument. Instead, all functions act on the single default loop.
4151		4866
		4867	Note that C<EV_DEFAULT> and C<EV_DEFAULT_> will no longer provide a
		4868	default loop when multiplicity is switched off - you always have to
		4869	initialise the loop manually in this case.
		4870
4152	=item EV_MINPRI	4871	=item EV_MINPRI
4153		4872
4154	=item EV_MAXPRI	4873	=item EV_MAXPRI
4155		4874
4156	The range of allowed priorities. C<EV_MINPRI> must be smaller or equal to	4875	The range of allowed priorities. C<EV_MINPRI> must be smaller or equal to
…		…
4192	#define EV_USE_POLL 1	4911	#define EV_USE_POLL 1
4193	#define EV_CHILD_ENABLE 1	4912	#define EV_CHILD_ENABLE 1
4194	#define EV_ASYNC_ENABLE 1	4913	#define EV_ASYNC_ENABLE 1
4195		4914
4196	The actual value is a bitset, it can be a combination of the following	4915	The actual value is a bitset, it can be a combination of the following
4197	values:	4916	values (by default, all of these are enabled):
4198		4917
4199	=over 4	4918	=over 4
4200		4919
4201	=item C<1> - faster/larger code	4920	=item C<1> - faster/larger code
4202		4921
…		…
4206	code size by roughly 30% on amd64).	4925	code size by roughly 30% on amd64).
4207		4926
4208	When optimising for size, use of compiler flags such as C<-Os> with	4927	When optimising for size, use of compiler flags such as C<-Os> with
4209	gcc is recommended, as well as C<-DNDEBUG>, as libev contains a number of	4928	gcc is recommended, as well as C<-DNDEBUG>, as libev contains a number of
4210	assertions.	4929	assertions.
		4930
		4931	The default is off when C<__OPTIMIZE_SIZE__> is defined by your compiler
		4932	(e.g. gcc with C<-Os>).
4211		4933
4212	=item C<2> - faster/larger data structures	4934	=item C<2> - faster/larger data structures
4213		4935
4214	Replaces the small 2-heap for timer management by a faster 4-heap, larger	4936	Replaces the small 2-heap for timer management by a faster 4-heap, larger
4215	hash table sizes and so on. This will usually further increase code size	4937	hash table sizes and so on. This will usually further increase code size
4216	and can additionally have an effect on the size of data structures at	4938	and can additionally have an effect on the size of data structures at
4217	runtime.	4939	runtime.
4218		4940
		4941	The default is off when C<__OPTIMIZE_SIZE__> is defined by your compiler
		4942	(e.g. gcc with C<-Os>).
		4943
4219	=item C<4> - full API configuration	4944	=item C<4> - full API configuration
4220		4945
4221	This enables priorities (sets C<EV_MAXPRI>=2 and C<EV_MINPRI>=-2), and	4946	This enables priorities (sets C<EV_MAXPRI>=2 and C<EV_MINPRI>=-2), and
4222	enables multiplicity (C<EV_MULTIPLICITY>=1).	4947	enables multiplicity (C<EV_MULTIPLICITY>=1).
4223		4948
…		…
4253		4978
4254	With an intelligent-enough linker (gcc+binutils are intelligent enough	4979	With an intelligent-enough linker (gcc+binutils are intelligent enough
4255	when you use C<-Wl,--gc-sections -ffunction-sections>) functions unused by	4980	when you use C<-Wl,--gc-sections -ffunction-sections>) functions unused by
4256	your program might be left out as well - a binary starting a timer and an	4981	your program might be left out as well - a binary starting a timer and an
4257	I/O watcher then might come out at only 5Kb.	4982	I/O watcher then might come out at only 5Kb.
		4983
		4984	=item EV_API_STATIC
		4985
		4986	If this symbol is defined (by default it is not), then all identifiers
		4987	will have static linkage. This means that libev will not export any
		4988	identifiers, and you cannot link against libev anymore. This can be useful
		4989	when you embed libev, only want to use libev functions in a single file,
		4990	and do not want its identifiers to be visible.
		4991
		4992	To use this, define C<EV_API_STATIC> and include F<ev.c> in the file that
		4993	wants to use libev.
		4994
		4995	This option only works when libev is compiled with a C compiler, as C++
		4996	doesn't support the required declaration syntax.
4258		4997
4259	=item EV_AVOID_STDIO	4998	=item EV_AVOID_STDIO
4260		4999
4261	If this is set to C<1> at compiletime, then libev will avoid using stdio	5000	If this is set to C<1> at compiletime, then libev will avoid using stdio
4262	functions (printf, scanf, perror etc.). This will increase the code size	5001	functions (printf, scanf, perror etc.). This will increase the code size
…		…
4320	in. If set to C<1>, then verification code will be compiled in, but not	5059	in. If set to C<1>, then verification code will be compiled in, but not
4321	called. If set to C<2>, then the internal verification code will be	5060	called. If set to C<2>, then the internal verification code will be
4322	called once per loop, which can slow down libev. If set to C<3>, then the	5061	called once per loop, which can slow down libev. If set to C<3>, then the
4323	verification code will be called very frequently, which will slow down	5062	verification code will be called very frequently, which will slow down
4324	libev considerably.	5063	libev considerably.
		5064
		5065	Verification errors are reported via C's C<assert> mechanism, so if you
		5066	disable that (e.g. by defining C<NDEBUG>) then no errors will be reported.
4325		5067
4326	The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it	5068	The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it
4327	will be C<0>.	5069	will be C<0>.
4328		5070
4329	=item EV_COMMON	5071	=item EV_COMMON
…		…
4406	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:	5148	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:
4407		5149
4408	#include "ev_cpp.h"	5150	#include "ev_cpp.h"
4409	#include "ev.c"	5151	#include "ev.c"
4410		5152
4411	=head1 INTERACTION WITH OTHER PROGRAMS OR LIBRARIES	5153	=head1 INTERACTION WITH OTHER PROGRAMS, LIBRARIES OR THE ENVIRONMENT
4412		5154
4413	=head2 THREADS AND COROUTINES	5155	=head2 THREADS AND COROUTINES
4414		5156
4415	=head3 THREADS	5157	=head3 THREADS
4416		5158
…		…
4467	default loop and triggering an C<ev_async> watcher from the default loop	5209	default loop and triggering an C<ev_async> watcher from the default loop
4468	watcher callback into the event loop interested in the signal.	5210	watcher callback into the event loop interested in the signal.
4469		5211
4470	=back	5212	=back
4471		5213
4472	=head4 THREAD LOCKING EXAMPLE	5214	See also L</THREAD LOCKING EXAMPLE>.
4473
4474	Here is a fictitious example of how to run an event loop in a different
4475	thread than where callbacks are being invoked and watchers are
4476	created/added/removed.
4477
4478	For a real-world example, see the C<EV::Loop::Async> perl module,
4479	which uses exactly this technique (which is suited for many high-level
4480	languages).
4481
4482	The example uses a pthread mutex to protect the loop data, a condition
4483	variable to wait for callback invocations, an async watcher to notify the
4484	event loop thread and an unspecified mechanism to wake up the main thread.
4485
4486	First, you need to associate some data with the event loop:
4487
4488	typedef struct {
4489	mutex_t lock; /* global loop lock */
4490	ev_async async_w;
4491	thread_t tid;
4492	cond_t invoke_cv;
4493	} userdata;
4494
4495	void prepare_loop (EV_P)
4496	{
4497	// for simplicity, we use a static userdata struct.
4498	static userdata u;
4499
4500	ev_async_init (&u->async_w, async_cb);
4501	ev_async_start (EV_A_ &u->async_w);
4502
4503	pthread_mutex_init (&u->lock, 0);
4504	pthread_cond_init (&u->invoke_cv, 0);
4505
4506	// now associate this with the loop
4507	ev_set_userdata (EV_A_ u);
4508	ev_set_invoke_pending_cb (EV_A_ l_invoke);
4509	ev_set_loop_release_cb (EV_A_ l_release, l_acquire);
4510
4511	// then create the thread running ev_loop
4512	pthread_create (&u->tid, 0, l_run, EV_A);
4513	}
4514
4515	The callback for the C<ev_async> watcher does nothing: the watcher is used
4516	solely to wake up the event loop so it takes notice of any new watchers
4517	that might have been added:
4518
4519	static void
4520	async_cb (EV_P_ ev_async *w, int revents)
4521	{
4522	// just used for the side effects
4523	}
4524
4525	The C<l_release> and C<l_acquire> callbacks simply unlock/lock the mutex
4526	protecting the loop data, respectively.
4527
4528	static void
4529	l_release (EV_P)
4530	{
4531	userdata *u = ev_userdata (EV_A);
4532	pthread_mutex_unlock (&u->lock);
4533	}
4534
4535	static void
4536	l_acquire (EV_P)
4537	{
4538	userdata *u = ev_userdata (EV_A);
4539	pthread_mutex_lock (&u->lock);
4540	}
4541
4542	The event loop thread first acquires the mutex, and then jumps straight
4543	into C<ev_run>:
4544
4545	void *
4546	l_run (void *thr_arg)
4547	{
4548	struct ev_loop *loop = (struct ev_loop *)thr_arg;
4549
4550	l_acquire (EV_A);
4551	pthread_setcanceltype (PTHREAD_CANCEL_ASYNCHRONOUS, 0);
4552	ev_run (EV_A_ 0);
4553	l_release (EV_A);
4554
4555	return 0;
4556	}
4557
4558	Instead of invoking all pending watchers, the C<l_invoke> callback will
4559	signal the main thread via some unspecified mechanism (signals? pipe
4560	writes? C<Async::Interrupt>?) and then waits until all pending watchers
4561	have been called (in a while loop because a) spurious wakeups are possible
4562	and b) skipping inter-thread-communication when there are no pending
4563	watchers is very beneficial):
4564
4565	static void
4566	l_invoke (EV_P)
4567	{
4568	userdata *u = ev_userdata (EV_A);
4569
4570	while (ev_pending_count (EV_A))
4571	{
4572	wake_up_other_thread_in_some_magic_or_not_so_magic_way ();
4573	pthread_cond_wait (&u->invoke_cv, &u->lock);
4574	}
4575	}
4576
4577	Now, whenever the main thread gets told to invoke pending watchers, it
4578	will grab the lock, call C<ev_invoke_pending> and then signal the loop
4579	thread to continue:
4580
4581	static void
4582	real_invoke_pending (EV_P)
4583	{
4584	userdata *u = ev_userdata (EV_A);
4585
4586	pthread_mutex_lock (&u->lock);
4587	ev_invoke_pending (EV_A);
4588	pthread_cond_signal (&u->invoke_cv);
4589	pthread_mutex_unlock (&u->lock);
4590	}
4591
4592	Whenever you want to start/stop a watcher or do other modifications to an
4593	event loop, you will now have to lock:
4594
4595	ev_timer timeout_watcher;
4596	userdata *u = ev_userdata (EV_A);
4597
4598	ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);
4599
4600	pthread_mutex_lock (&u->lock);
4601	ev_timer_start (EV_A_ &timeout_watcher);
4602	ev_async_send (EV_A_ &u->async_w);
4603	pthread_mutex_unlock (&u->lock);
4604
4605	Note that sending the C<ev_async> watcher is required because otherwise
4606	an event loop currently blocking in the kernel will have no knowledge
4607	about the newly added timer. By waking up the loop it will pick up any new
4608	watchers in the next event loop iteration.
4609		5215
4610	=head3 COROUTINES	5216	=head3 COROUTINES
4611		5217
4612	Libev is very accommodating to coroutines ("cooperative threads"):	5218	Libev is very accommodating to coroutines ("cooperative threads"):
4613	libev fully supports nesting calls to its functions from different	5219	libev fully supports nesting calls to its functions from different
…		…
4778	requires, and its I/O model is fundamentally incompatible with the POSIX	5384	requires, and its I/O model is fundamentally incompatible with the POSIX
4779	model. Libev still offers limited functionality on this platform in	5385	model. Libev still offers limited functionality on this platform in
4780	the form of the C<EVBACKEND_SELECT> backend, and only supports socket	5386	the form of the C<EVBACKEND_SELECT> backend, and only supports socket
4781	descriptors. This only applies when using Win32 natively, not when using	5387	descriptors. This only applies when using Win32 natively, not when using
4782	e.g. cygwin. Actually, it only applies to the microsofts own compilers,	5388	e.g. cygwin. Actually, it only applies to the microsofts own compilers,
4783	as every compielr comes with a slightly differently broken/incompatible	5389	as every compiler comes with a slightly differently broken/incompatible
4784	environment.	5390	environment.
4785		5391
4786	Lifting these limitations would basically require the full	5392	Lifting these limitations would basically require the full
4787	re-implementation of the I/O system. If you are into this kind of thing,	5393	re-implementation of the I/O system. If you are into this kind of thing,
4788	then note that glib does exactly that for you in a very portable way (note	5394	then note that glib does exactly that for you in a very portable way (note
…		…
4882	structure (guaranteed by POSIX but not by ISO C for example), but it also	5488	structure (guaranteed by POSIX but not by ISO C for example), but it also
4883	assumes that the same (machine) code can be used to call any watcher	5489	assumes that the same (machine) code can be used to call any watcher
4884	callback: The watcher callbacks have different type signatures, but libev	5490	callback: The watcher callbacks have different type signatures, but libev
4885	calls them using an C<ev_watcher *> internally.	5491	calls them using an C<ev_watcher *> internally.
4886		5492
		5493	=item null pointers and integer zero are represented by 0 bytes
		5494
		5495	Libev uses C<memset> to initialise structs and arrays to C<0> bytes, and
		5496	relies on this setting pointers and integers to null.
		5497
4887	=item pointer accesses must be thread-atomic	5498	=item pointer accesses must be thread-atomic
4888		5499
4889	Accessing a pointer value must be atomic, it must both be readable and	5500	Accessing a pointer value must be atomic, it must both be readable and
4890	writable in one piece - this is the case on all current architectures.	5501	writable in one piece - this is the case on all current architectures.
4891		5502
…		…
4904	thread" or will block signals process-wide, both behaviours would	5515	thread" or will block signals process-wide, both behaviours would
4905	be compatible with libev. Interaction between C<sigprocmask> and	5516	be compatible with libev. Interaction between C<sigprocmask> and
4906	C<pthread_sigmask> could complicate things, however.	5517	C<pthread_sigmask> could complicate things, however.
4907		5518
4908	The most portable way to handle signals is to block signals in all threads	5519	The most portable way to handle signals is to block signals in all threads
4909	except the initial one, and run the default loop in the initial thread as	5520	except the initial one, and run the signal handling loop in the initial
4910	well.	5521	thread as well.
4911		5522
4912	=item C<long> must be large enough for common memory allocation sizes	5523	=item C<long> must be large enough for common memory allocation sizes
4913		5524
4914	To improve portability and simplify its API, libev uses C<long> internally	5525	To improve portability and simplify its API, libev uses C<long> internally
4915	instead of C<size_t> when allocating its data structures. On non-POSIX	5526	instead of C<size_t> when allocating its data structures. On non-POSIX
…		…
4921		5532
4922	The type C<double> is used to represent timestamps. It is required to	5533	The type C<double> is used to represent timestamps. It is required to
4923	have at least 51 bits of mantissa (and 9 bits of exponent), which is	5534	have at least 51 bits of mantissa (and 9 bits of exponent), which is
4924	good enough for at least into the year 4000 with millisecond accuracy	5535	good enough for at least into the year 4000 with millisecond accuracy
4925	(the design goal for libev). This requirement is overfulfilled by	5536	(the design goal for libev). This requirement is overfulfilled by
4926	implementations using IEEE 754, which is basically all existing ones. With	5537	implementations using IEEE 754, which is basically all existing ones.
		5538
4927	IEEE 754 doubles, you get microsecond accuracy until at least 2200.	5539	With IEEE 754 doubles, you get microsecond accuracy until at least the
		5540	year 2255 (and millisecond accuracy till the year 287396 - by then, libev
		5541	is either obsolete or somebody patched it to use C<long double> or
		5542	something like that, just kidding).
4928		5543
4929	=back	5544	=back
4930		5545
4931	If you know of other additional requirements drop me a note.	5546	If you know of other additional requirements drop me a note.
4932		5547
…		…
4994	=item Processing ev_async_send: O(number_of_async_watchers)	5609	=item Processing ev_async_send: O(number_of_async_watchers)
4995		5610
4996	=item Processing signals: O(max_signal_number)	5611	=item Processing signals: O(max_signal_number)
4997		5612
4998	Sending involves a system call I<iff> there were no other C<ev_async_send>	5613	Sending involves a system call I<iff> there were no other C<ev_async_send>
4999	calls in the current loop iteration. Checking for async and signal events	5614	calls in the current loop iteration and the loop is currently
		5615	blocked. Checking for async and signal events involves iterating over all
5000	involves iterating over all running async watchers or all signal numbers.	5616	running async watchers or all signal numbers.
5001		5617
5002	=back	5618	=back
5003		5619
5004		5620
5005	=head1 PORTING FROM LIBEV 3.X TO 4.X	5621	=head1 PORTING FROM LIBEV 3.X TO 4.X
…		…
5014	=over 4	5630	=over 4
5015		5631
5016	=item C<EV_COMPAT3> backwards compatibility mechanism	5632	=item C<EV_COMPAT3> backwards compatibility mechanism
5017		5633
5018	The backward compatibility mechanism can be controlled by	5634	The backward compatibility mechanism can be controlled by
5019	C<EV_COMPAT3>. See L<PREPROCESSOR SYMBOLS/MACROS> in the L<EMBEDDING>	5635	C<EV_COMPAT3>. See L</"PREPROCESSOR SYMBOLS/MACROS"> in the L</EMBEDDING>
5020	section.	5636	section.
5021		5637
5022	=item C<ev_default_destroy> and C<ev_default_fork> have been removed	5638	=item C<ev_default_destroy> and C<ev_default_fork> have been removed
5023		5639
5024	These calls can be replaced easily by their C<ev_loop_xxx> counterparts:	5640	These calls can be replaced easily by their C<ev_loop_xxx> counterparts:
…		…
5067	=over 4	5683	=over 4
5068		5684
5069	=item active	5685	=item active
5070		5686
5071	A watcher is active as long as it has been started and not yet stopped.	5687	A watcher is active as long as it has been started and not yet stopped.
5072	See L<WATCHER STATES> for details.	5688	See L</WATCHER STATES> for details.
5073		5689
5074	=item application	5690	=item application
5075		5691
5076	In this document, an application is whatever is using libev.	5692	In this document, an application is whatever is using libev.
5077		5693
…		…
5113	watchers and events.	5729	watchers and events.
5114		5730
5115	=item pending	5731	=item pending
5116		5732
5117	A watcher is pending as soon as the corresponding event has been	5733	A watcher is pending as soon as the corresponding event has been
5118	detected. See L<WATCHER STATES> for details.	5734	detected. See L</WATCHER STATES> for details.
5119		5735
5120	=item real time	5736	=item real time
5121		5737
5122	The physical time that is observed. It is apparently strictly monotonic :)	5738	The physical time that is observed. It is apparently strictly monotonic :)
5123		5739
5124	=item wall-clock time	5740	=item wall-clock time
5125		5741
5126	The time and date as shown on clocks. Unlike real time, it can actually	5742	The time and date as shown on clocks. Unlike real time, it can actually
5127	be wrong and jump forwards and backwards, e.g. when the you adjust your	5743	be wrong and jump forwards and backwards, e.g. when you adjust your
5128	clock.	5744	clock.
5129		5745
5130	=item watcher	5746	=item watcher
5131		5747
5132	A data structure that describes interest in certain events. Watchers need	5748	A data structure that describes interest in certain events. Watchers need
…		…
5135	=back	5751	=back
5136		5752
5137	=head1 AUTHOR	5753	=head1 AUTHOR
5138		5754
5139	Marc Lehmann <libev@schmorp.de>, with repeated corrections by Mikael	5755	Marc Lehmann <libev@schmorp.de>, with repeated corrections by Mikael
5140	Magnusson and Emanuele Giaquinta.	5756	Magnusson and Emanuele Giaquinta, and minor corrections by many others.
5141		5757

Diff Legend

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