ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/libev/ev.pod

Comparing libev/ev.pod (file contents):
Revision 1.178 by root, Sat Sep 13 18:25:50 2008 UTC vs.
Revision 1.282 by root, Wed Mar 10 08:19:39 2010 UTC

…		…
9	=head2 EXAMPLE PROGRAM	9	=head2 EXAMPLE PROGRAM
10		10
11	// a single header file is required	11	// a single header file is required
12	#include <ev.h>	12	#include <ev.h>
13		13
		14	#include <stdio.h> // for puts
		15
14	// every watcher type has its own typedef'd struct	16	// every watcher type has its own typedef'd struct
15	// with the name ev_<type>	17	// with the name ev_TYPE
16	ev_io stdin_watcher;	18	ev_io stdin_watcher;
17	ev_timer timeout_watcher;	19	ev_timer timeout_watcher;
18		20
19	// all watcher callbacks have a similar signature	21	// all watcher callbacks have a similar signature
20	// this callback is called when data is readable on stdin	22	// this callback is called when data is readable on stdin
21	static void	23	static void
22	stdin_cb (EV_P_ struct ev_io *w, int revents)	24	stdin_cb (EV_P_ ev_io *w, int revents)
23	{	25	{
24	puts ("stdin ready");	26	puts ("stdin ready");
25	// for one-shot events, one must manually stop the watcher	27	// for one-shot events, one must manually stop the watcher
26	// with its corresponding stop function.	28	// with its corresponding stop function.
27	ev_io_stop (EV_A_ w);	29	ev_io_stop (EV_A_ w);
…		…
30	ev_unloop (EV_A_ EVUNLOOP_ALL);	32	ev_unloop (EV_A_ EVUNLOOP_ALL);
31	}	33	}
32		34
33	// another callback, this time for a time-out	35	// another callback, this time for a time-out
34	static void	36	static void
35	timeout_cb (EV_P_ struct ev_timer *w, int revents)	37	timeout_cb (EV_P_ ev_timer *w, int revents)
36	{	38	{
37	puts ("timeout");	39	puts ("timeout");
38	// this causes the innermost ev_loop to stop iterating	40	// this causes the innermost ev_loop to stop iterating
39	ev_unloop (EV_A_ EVUNLOOP_ONE);	41	ev_unloop (EV_A_ EVUNLOOP_ONE);
40	}	42	}
…		…
60		62
61	// unloop was called, so exit	63	// unloop was called, so exit
62	return 0;	64	return 0;
63	}	65	}
64		66
65	=head1 DESCRIPTION	67	=head1 ABOUT THIS DOCUMENT
		68
		69	This document documents the libev software package.
66		70
67	The newest version of this document is also available as an html-formatted	71	The newest version of this document is also available as an html-formatted
68	web page you might find easier to navigate when reading it for the first	72	web page you might find easier to navigate when reading it for the first
69	time: L<http://pod.tst.eu/http://cvs.schmorp.de/libev/ev.pod>.	73	time: L<http://pod.tst.eu/http://cvs.schmorp.de/libev/ev.pod>.
		74
		75	While this document tries to be as complete as possible in documenting
		76	libev, its usage and the rationale behind its design, it is not a tutorial
		77	on event-based programming, nor will it introduce event-based programming
		78	with libev.
		79
		80	Familarity with event based programming techniques in general is assumed
		81	throughout this document.
		82
		83	=head1 ABOUT LIBEV
70		84
71	Libev is an event loop: you register interest in certain events (such as a	85	Libev is an event loop: you register interest in certain events (such as a
72	file descriptor being readable or a timeout occurring), and it will manage	86	file descriptor being readable or a timeout occurring), and it will manage
73	these event sources and provide your program with events.	87	these event sources and provide your program with events.
74		88
…		…
84	=head2 FEATURES	98	=head2 FEATURES
85		99
86	Libev supports C<select>, C<poll>, the Linux-specific C<epoll>, the	100	Libev supports C<select>, C<poll>, the Linux-specific C<epoll>, the
87	BSD-specific C<kqueue> and the Solaris-specific event port mechanisms	101	BSD-specific C<kqueue> and the Solaris-specific event port mechanisms
88	for file descriptor events (C<ev_io>), the Linux C<inotify> interface	102	for file descriptor events (C<ev_io>), the Linux C<inotify> interface
89	(for C<ev_stat>), relative timers (C<ev_timer>), absolute timers	103	(for C<ev_stat>), Linux eventfd/signalfd (for faster and cleaner
90	with customised rescheduling (C<ev_periodic>), synchronous signals	104	inter-thread wakeup (C<ev_async>)/signal handling (C<ev_signal>)) relative
91	(C<ev_signal>), process status change events (C<ev_child>), and event	105	timers (C<ev_timer>), absolute timers with customised rescheduling
92	watchers dealing with the event loop mechanism itself (C<ev_idle>,	106	(C<ev_periodic>), synchronous signals (C<ev_signal>), process status
93	C<ev_embed>, C<ev_prepare> and C<ev_check> watchers) as well as	107	change events (C<ev_child>), and event watchers dealing with the event
94	file watchers (C<ev_stat>) and even limited support for fork events	108	loop mechanism itself (C<ev_idle>, C<ev_embed>, C<ev_prepare> and
95	(C<ev_fork>).	109	C<ev_check> watchers) as well as file watchers (C<ev_stat>) and even
		110	limited support for fork events (C<ev_fork>).
96		111
97	It also is quite fast (see this	112	It also is quite fast (see this
98	L<benchmark|http://libev.schmorp.de/bench.html> comparing it to libevent	113	L<benchmark|http://libev.schmorp.de/bench.html> comparing it to libevent
99	for example).	114	for example).
100		115
…		…
108	name C<loop> (which is always of type C<struct ev_loop *>) will not have	123	name C<loop> (which is always of type C<struct ev_loop *>) will not have
109	this argument.	124	this argument.
110		125
111	=head2 TIME REPRESENTATION	126	=head2 TIME REPRESENTATION
112		127
113	Libev represents time as a single floating point number, representing the	128	Libev represents time as a single floating point number, representing
114	(fractional) number of seconds since the (POSIX) epoch (somewhere near	129	the (fractional) number of seconds since the (POSIX) epoch (somewhere
115	the beginning of 1970, details are complicated, don't ask). This type is	130	near the beginning of 1970, details are complicated, don't ask). This
116	called C<ev_tstamp>, which is what you should use too. It usually aliases	131	type is called C<ev_tstamp>, which is what you should use too. It usually
117	to the C<double> type in C, and when you need to do any calculations on	132	aliases to the C<double> type in C. When you need to do any calculations
118	it, you should treat it as some floating point value. Unlike the name	133	on it, you should treat it as some floating point value. Unlike the name
119	component C<stamp> might indicate, it is also used for time differences	134	component C<stamp> might indicate, it is also used for time differences
120	throughout libev.	135	throughout libev.
121		136
122	=head1 ERROR HANDLING	137	=head1 ERROR HANDLING
123		138
…		…
214	C<ev_embeddable_backends () & ev_supported_backends ()>, likewise for	229	C<ev_embeddable_backends () & ev_supported_backends ()>, likewise for
215	recommended ones.	230	recommended ones.
216		231
217	See the description of C<ev_embed> watchers for more info.	232	See the description of C<ev_embed> watchers for more info.
218		233
219	=item ev_set_allocator (void *(*cb)(void *ptr, long size))	234	=item ev_set_allocator (void *(*cb)(void *ptr, long size)) [NOT REENTRANT]
220		235
221	Sets the allocation function to use (the prototype is similar - the	236	Sets the allocation function to use (the prototype is similar - the
222	semantics are identical to the C<realloc> C89/SuS/POSIX function). It is	237	semantics are identical to the C<realloc> C89/SuS/POSIX function). It is
223	used to allocate and free memory (no surprises here). If it returns zero	238	used to allocate and free memory (no surprises here). If it returns zero
224	when memory needs to be allocated (C<size != 0>), the library might abort	239	when memory needs to be allocated (C<size != 0>), the library might abort
…		…
250	}	265	}
251		266
252	...	267	...
253	ev_set_allocator (persistent_realloc);	268	ev_set_allocator (persistent_realloc);
254		269
255	=item ev_set_syserr_cb (void (*cb)(const char *msg));	270	=item ev_set_syserr_cb (void (*cb)(const char *msg)); [NOT REENTRANT]
256		271
257	Set the callback function to call on a retryable system call error (such	272	Set the callback function to call on a retryable system call error (such
258	as failed select, poll, epoll_wait). The message is a printable string	273	as failed select, poll, epoll_wait). The message is a printable string
259	indicating the system call or subsystem causing the problem. If this	274	indicating the system call or subsystem causing the problem. If this
260	callback is set, then libev will expect it to remedy the situation, no	275	callback is set, then libev will expect it to remedy the situation, no
…		…
276		291
277	=back	292	=back
278		293
279	=head1 FUNCTIONS CONTROLLING THE EVENT LOOP	294	=head1 FUNCTIONS CONTROLLING THE EVENT LOOP
280		295
281	An event loop is described by a C<struct ev_loop *>. The library knows two	296	An event loop is described by a C<struct ev_loop *> (the C<struct>
282	types of such loops, the I<default> loop, which supports signals and child	297	is I<not> optional in this case, as there is also an C<ev_loop>
283	events, and dynamically created loops which do not.	298	I<function>).
		299
		300	The library knows two types of such loops, the I<default> loop, which
		301	supports signals and child events, and dynamically created loops which do
		302	not.
284		303
285	=over 4	304	=over 4
286		305
287	=item struct ev_loop *ev_default_loop (unsigned int flags)	306	=item struct ev_loop *ev_default_loop (unsigned int flags)
288		307
…		…
294	If you don't know what event loop to use, use the one returned from this	313	If you don't know what event loop to use, use the one returned from this
295	function.	314	function.
296		315
297	Note that this function is I<not> thread-safe, so if you want to use it	316	Note that this function is I<not> thread-safe, so if you want to use it
298	from multiple threads, you have to lock (note also that this is unlikely,	317	from multiple threads, you have to lock (note also that this is unlikely,
299	as loops cannot bes hared easily between threads anyway).	318	as loops cannot be shared easily between threads anyway).
300		319
301	The default loop is the only loop that can handle C<ev_signal> and	320	The default loop is the only loop that can handle C<ev_signal> and
302	C<ev_child> watchers, and to do this, it always registers a handler	321	C<ev_child> watchers, and to do this, it always registers a handler
303	for C<SIGCHLD>. If this is a problem for your application you can either	322	for C<SIGCHLD>. If this is a problem for your application you can either
304	create a dynamic loop with C<ev_loop_new> that doesn't do that, or you	323	create a dynamic loop with C<ev_loop_new> that doesn't do that, or you
…		…
344	flag.	363	flag.
345		364
346	This flag setting cannot be overridden or specified in the C<LIBEV_FLAGS>	365	This flag setting cannot be overridden or specified in the C<LIBEV_FLAGS>
347	environment variable.	366	environment variable.
348		367
		368	=item C<EVFLAG_NOINOTIFY>
		369
		370	When this flag is specified, then libev will not attempt to use the
		371	I<inotify> API for it's C<ev_stat> watchers. Apart from debugging and
		372	testing, this flag can be useful to conserve inotify file descriptors, as
		373	otherwise each loop using C<ev_stat> watchers consumes one inotify handle.
		374
		375	=item C<EVFLAG_SIGNALFD>
		376
		377	When this flag is specified, then libev will attempt to use the
		378	I<signalfd> API for it's C<ev_signal> (and C<ev_child>) watchers. This API
		379	delivers signals synchronously, which makes it both faster and might make
		380	it possible to get the queued signal data. It can also simplify signal
		381	handling with threads, as long as you properly block signals in your
		382	threads that are not interested in handling them.
		383
		384	Signalfd will not be used by default as this changes your signal mask, and
		385	there are a lot of shoddy libraries and programs (glib's threadpool for
		386	example) that can't properly initialise their signal masks.
		387
349	=item C<EVBACKEND_SELECT> (value 1, portable select backend)	388	=item C<EVBACKEND_SELECT> (value 1, portable select backend)
350		389
351	This is your standard select(2) backend. Not I<completely> standard, as	390	This is your standard select(2) backend. Not I<completely> standard, as
352	libev tries to roll its own fd_set with no limits on the number of fds,	391	libev tries to roll its own fd_set with no limits on the number of fds,
353	but if that fails, expect a fairly low limit on the number of fds when	392	but if that fails, expect a fairly low limit on the number of fds when
…		…
359	writing a server, you should C<accept ()> in a loop to accept as many	398	writing a server, you should C<accept ()> in a loop to accept as many
360	connections as possible during one iteration. You might also want to have	399	connections as possible during one iteration. You might also want to have
361	a look at C<ev_set_io_collect_interval ()> to increase the amount of	400	a look at C<ev_set_io_collect_interval ()> to increase the amount of
362	readiness notifications you get per iteration.	401	readiness notifications you get per iteration.
363		402
		403	This backend maps C<EV_READ> to the C<readfds> set and C<EV_WRITE> to the
		404	C<writefds> set (and to work around Microsoft Windows bugs, also onto the
		405	C<exceptfds> set on that platform).
		406
364	=item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows)	407	=item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows)
365		408
366	And this is your standard poll(2) backend. It's more complicated	409	And this is your standard poll(2) backend. It's more complicated
367	than select, but handles sparse fds better and has no artificial	410	than select, but handles sparse fds better and has no artificial
368	limit on the number of fds you can use (except it will slow down	411	limit on the number of fds you can use (except it will slow down
369	considerably with a lot of inactive fds). It scales similarly to select,	412	considerably with a lot of inactive fds). It scales similarly to select,
370	i.e. O(total_fds). See the entry for C<EVBACKEND_SELECT>, above, for	413	i.e. O(total_fds). See the entry for C<EVBACKEND_SELECT>, above, for
371	performance tips.	414	performance tips.
372		415
		416	This backend maps C<EV_READ> to C<POLLIN | POLLERR | POLLHUP>, and
		417	C<EV_WRITE> to C<POLLOUT | POLLERR | POLLHUP>.
		418
373	=item C<EVBACKEND_EPOLL> (value 4, Linux)	419	=item C<EVBACKEND_EPOLL> (value 4, Linux)
		420
		421	Use the linux-specific epoll(7) interface (for both pre- and post-2.6.9
		422	kernels).
374		423
375	For few fds, this backend is a bit little slower than poll and select,	424	For few fds, this backend is a bit little slower than poll and select,
376	but it scales phenomenally better. While poll and select usually scale	425	but it scales phenomenally better. While poll and select usually scale
377	like O(total_fds) where n is the total number of fds (or the highest fd),	426	like O(total_fds) where n is the total number of fds (or the highest fd),
378	epoll scales either O(1) or O(active_fds). The epoll design has a number	427	epoll scales either O(1) or O(active_fds).
379	of shortcomings, such as silently dropping events in some hard-to-detect	428
380	cases and requiring a system call per fd change, no fork support and bad	429	The epoll mechanism deserves honorable mention as the most misdesigned
381	support for dup.	430	of the more advanced event mechanisms: mere annoyances include silently
		431	dropping file descriptors, requiring a system call per change per file
		432	descriptor (and unnecessary guessing of parameters), problems with dup and
		433	so on. The biggest issue is fork races, however - if a program forks then
		434	I<both> parent and child process have to recreate the epoll set, which can
		435	take considerable time (one syscall per file descriptor) and is of course
		436	hard to detect.
		437
		438	Epoll is also notoriously buggy - embedding epoll fds I<should> work, but
		439	of course I<doesn't>, and epoll just loves to report events for totally
		440	I<different> file descriptors (even already closed ones, so one cannot
		441	even remove them from the set) than registered in the set (especially
		442	on SMP systems). Libev tries to counter these spurious notifications by
		443	employing an additional generation counter and comparing that against the
		444	events to filter out spurious ones, recreating the set when required.
382		445
383	While stopping, setting and starting an I/O watcher in the same iteration	446	While stopping, setting and starting an I/O watcher in the same iteration
384	will result in some caching, there is still a system call per such incident	447	will result in some caching, there is still a system call per such
385	(because the fd could point to a different file description now), so its	448	incident (because the same I<file descriptor> could point to a different
386	best to avoid that. Also, C<dup ()>'ed file descriptors might not work	449	I<file description> now), so its best to avoid that. Also, C<dup ()>'ed
387	very well if you register events for both fds.	450	file descriptors might not work very well if you register events for both
388		451	file descriptors.
389	Please note that epoll sometimes generates spurious notifications, so you
390	need to use non-blocking I/O or other means to avoid blocking when no data
391	(or space) is available.
392		452
393	Best performance from this backend is achieved by not unregistering all	453	Best performance from this backend is achieved by not unregistering all
394	watchers for a file descriptor until it has been closed, if possible, i.e.	454	watchers for a file descriptor until it has been closed, if possible,
395	keep at least one watcher active per fd at all times.	455	i.e. keep at least one watcher active per fd at all times. Stopping and
		456	starting a watcher (without re-setting it) also usually doesn't cause
		457	extra overhead. A fork can both result in spurious notifications as well
		458	as in libev having to destroy and recreate the epoll object, which can
		459	take considerable time and thus should be avoided.
		460
		461	All this means that, in practice, C<EVBACKEND_SELECT> can be as fast or
		462	faster than epoll for maybe up to a hundred file descriptors, depending on
		463	the usage. So sad.
396		464
397	While nominally embeddable in other event loops, this feature is broken in	465	While nominally embeddable in other event loops, this feature is broken in
398	all kernel versions tested so far.	466	all kernel versions tested so far.
		467
		468	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
		469	C<EVBACKEND_POLL>.
399		470
400	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)	471	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)
401		472
402	Kqueue deserves special mention, as at the time of this writing, it	473	Kqueue deserves special mention, as at the time of this writing, it
403	was broken on all BSDs except NetBSD (usually it doesn't work reliably	474	was broken on all BSDs except NetBSD (usually it doesn't work reliably
404	with anything but sockets and pipes, except on Darwin, where of course	475	with anything but sockets and pipes, except on Darwin, where of course
405	it's completely useless). For this reason it's not being "auto-detected"	476	it's completely useless). Unlike epoll, however, whose brokenness
		477	is by design, these kqueue bugs can (and eventually will) be fixed
		478	without API changes to existing programs. For this reason it's not being
406	unless you explicitly specify it explicitly in the flags (i.e. using	479	"auto-detected" unless you explicitly specify it in the flags (i.e. using
407	C<EVBACKEND_KQUEUE>) or libev was compiled on a known-to-be-good (-enough)	480	C<EVBACKEND_KQUEUE>) or libev was compiled on a known-to-be-good (-enough)
408	system like NetBSD.	481	system like NetBSD.
409		482
410	You still can embed kqueue into a normal poll or select backend and use it	483	You still can embed kqueue into a normal poll or select backend and use it
411	only for sockets (after having made sure that sockets work with kqueue on	484	only for sockets (after having made sure that sockets work with kqueue on
…		…
413		486
414	It scales in the same way as the epoll backend, but the interface to the	487	It scales in the same way as the epoll backend, but the interface to the
415	kernel is more efficient (which says nothing about its actual speed, of	488	kernel is more efficient (which says nothing about its actual speed, of
416	course). While stopping, setting and starting an I/O watcher does never	489	course). While stopping, setting and starting an I/O watcher does never
417	cause an extra system call as with C<EVBACKEND_EPOLL>, it still adds up to	490	cause an extra system call as with C<EVBACKEND_EPOLL>, it still adds up to
418	two event changes per incident, support for C<fork ()> is very bad and it	491	two event changes per incident. Support for C<fork ()> is very bad (but
419	drops fds silently in similarly hard-to-detect cases.	492	sane, unlike epoll) and it drops fds silently in similarly hard-to-detect
		493	cases
420		494
421	This backend usually performs well under most conditions.	495	This backend usually performs well under most conditions.
422		496
423	While nominally embeddable in other event loops, this doesn't work	497	While nominally embeddable in other event loops, this doesn't work
424	everywhere, so you might need to test for this. And since it is broken	498	everywhere, so you might need to test for this. And since it is broken
425	almost everywhere, you should only use it when you have a lot of sockets	499	almost everywhere, you should only use it when you have a lot of sockets
426	(for which it usually works), by embedding it into another event loop	500	(for which it usually works), by embedding it into another event loop
427	(e.g. C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>) and using it only for	501	(e.g. C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> (but C<poll> is of course
428	sockets.	502	also broken on OS X)) and, did I mention it, using it only for sockets.
		503
		504	This backend maps C<EV_READ> into an C<EVFILT_READ> kevent with
		505	C<NOTE_EOF>, and C<EV_WRITE> into an C<EVFILT_WRITE> kevent with
		506	C<NOTE_EOF>.
429		507
430	=item C<EVBACKEND_DEVPOLL> (value 16, Solaris 8)	508	=item C<EVBACKEND_DEVPOLL> (value 16, Solaris 8)
431		509
432	This is not implemented yet (and might never be, unless you send me an	510	This is not implemented yet (and might never be, unless you send me an
433	implementation). According to reports, C</dev/poll> only supports sockets	511	implementation). According to reports, C</dev/poll> only supports sockets
…		…
446	While this backend scales well, it requires one system call per active	524	While this backend scales well, it requires one system call per active
447	file descriptor per loop iteration. For small and medium numbers of file	525	file descriptor per loop iteration. For small and medium numbers of file
448	descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend	526	descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend
449	might perform better.	527	might perform better.
450		528
451	On the positive side, ignoring the spurious readiness notifications, this	529	On the positive side, with the exception of the spurious readiness
452	backend actually performed to specification in all tests and is fully	530	notifications, this backend actually performed fully to specification
453	embeddable, which is a rare feat among the OS-specific backends.	531	in all tests and is fully embeddable, which is a rare feat among the
		532	OS-specific backends (I vastly prefer correctness over speed hacks).
		533
		534	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
		535	C<EVBACKEND_POLL>.
454		536
455	=item C<EVBACKEND_ALL>	537	=item C<EVBACKEND_ALL>
456		538
457	Try all backends (even potentially broken ones that wouldn't be tried	539	Try all backends (even potentially broken ones that wouldn't be tried
458	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as	540	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as
…		…
460		542
461	It is definitely not recommended to use this flag.	543	It is definitely not recommended to use this flag.
462		544
463	=back	545	=back
464		546
465	If one or more of these are or'ed into the flags value, then only these	547	If one or more of the backend flags are or'ed into the flags value,
466	backends will be tried (in the reverse order as listed here). If none are	548	then only these backends will be tried (in the reverse order as listed
467	specified, all backends in C<ev_recommended_backends ()> will be tried.	549	here). If none are specified, all backends in C<ev_recommended_backends
		550	()> will be tried.
468		551
469	The most typical usage is like this:	552	Example: This is the most typical usage.
470		553
471	if (!ev_default_loop (0))	554	if (!ev_default_loop (0))
472	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");	555	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
473		556
474	Restrict libev to the select and poll backends, and do not allow	557	Example: Restrict libev to the select and poll backends, and do not allow
475	environment settings to be taken into account:	558	environment settings to be taken into account:
476		559
477	ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);	560	ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);
478		561
479	Use whatever libev has to offer, but make sure that kqueue is used if	562	Example: Use whatever libev has to offer, but make sure that kqueue is
480	available (warning, breaks stuff, best use only with your own private	563	used if available (warning, breaks stuff, best use only with your own
481	event loop and only if you know the OS supports your types of fds):	564	private event loop and only if you know the OS supports your types of
		565	fds):
482		566
483	ev_default_loop (ev_recommended_backends () | EVBACKEND_KQUEUE);	567	ev_default_loop (ev_recommended_backends () | EVBACKEND_KQUEUE);
484		568
485	=item struct ev_loop *ev_loop_new (unsigned int flags)	569	=item struct ev_loop *ev_loop_new (unsigned int flags)
486		570
…		…
507	responsibility to either stop all watchers cleanly yourself I<before>	591	responsibility to either stop all watchers cleanly yourself I<before>
508	calling this function, or cope with the fact afterwards (which is usually	592	calling this function, or cope with the fact afterwards (which is usually
509	the easiest thing, you can just ignore the watchers and/or C<free ()> them	593	the easiest thing, you can just ignore the watchers and/or C<free ()> them
510	for example).	594	for example).
511		595
512	Note that certain global state, such as signal state, will not be freed by	596	Note that certain global state, such as signal state (and installed signal
513	this function, and related watchers (such as signal and child watchers)	597	handlers), will not be freed by this function, and related watchers (such
514	would need to be stopped manually.	598	as signal and child watchers) would need to be stopped manually.
515		599
516	In general it is not advisable to call this function except in the	600	In general it is not advisable to call this function except in the
517	rare occasion where you really need to free e.g. the signal handling	601	rare occasion where you really need to free e.g. the signal handling
518	pipe fds. If you need dynamically allocated loops it is better to use	602	pipe fds. If you need dynamically allocated loops it is better to use
519	C<ev_loop_new> and C<ev_loop_destroy>).	603	C<ev_loop_new> and C<ev_loop_destroy>.
520		604
521	=item ev_loop_destroy (loop)	605	=item ev_loop_destroy (loop)
522		606
523	Like C<ev_default_destroy>, but destroys an event loop created by an	607	Like C<ev_default_destroy>, but destroys an event loop created by an
524	earlier call to C<ev_loop_new>.	608	earlier call to C<ev_loop_new>.
…		…
544		628
545	=item ev_loop_fork (loop)	629	=item ev_loop_fork (loop)
546		630
547	Like C<ev_default_fork>, but acts on an event loop created by	631	Like C<ev_default_fork>, but acts on an event loop created by
548	C<ev_loop_new>. Yes, you have to call this on every allocated event loop	632	C<ev_loop_new>. Yes, you have to call this on every allocated event loop
549	after fork, and how you do this is entirely your own problem.	633	after fork that you want to re-use in the child, and how you do this is
		634	entirely your own problem.
550		635
551	=item int ev_is_default_loop (loop)	636	=item int ev_is_default_loop (loop)
552		637
553	Returns true when the given loop actually is the default loop, false otherwise.	638	Returns true when the given loop is, in fact, the default loop, and false
		639	otherwise.
554		640
555	=item unsigned int ev_loop_count (loop)	641	=item unsigned int ev_loop_count (loop)
556		642
557	Returns the count of loop iterations for the loop, which is identical to	643	Returns the count of loop iterations for the loop, which is identical to
558	the number of times libev did poll for new events. It starts at C<0> and	644	the number of times libev did poll for new events. It starts at C<0> and
559	happily wraps around with enough iterations.	645	happily wraps around with enough iterations.
560		646
561	This value can sometimes be useful as a generation counter of sorts (it	647	This value can sometimes be useful as a generation counter of sorts (it
562	"ticks" the number of loop iterations), as it roughly corresponds with	648	"ticks" the number of loop iterations), as it roughly corresponds with
563	C<ev_prepare> and C<ev_check> calls.	649	C<ev_prepare> and C<ev_check> calls.
		650
		651	=item unsigned int ev_loop_depth (loop)
		652
		653	Returns the number of times C<ev_loop> was entered minus the number of
		654	times C<ev_loop> was exited, in other words, the recursion depth.
		655
		656	Outside C<ev_loop>, this number is zero. In a callback, this number is
		657	C<1>, unless C<ev_loop> was invoked recursively (or from another thread),
		658	in which case it is higher.
		659
		660	Leaving C<ev_loop> abnormally (setjmp/longjmp, cancelling the thread
		661	etc.), doesn't count as exit.
564		662
565	=item unsigned int ev_backend (loop)	663	=item unsigned int ev_backend (loop)
566		664
567	Returns one of the C<EVBACKEND_*> flags indicating the event backend in	665	Returns one of the C<EVBACKEND_*> flags indicating the event backend in
568	use.	666	use.
…		…
583		681
584	This function is rarely useful, but when some event callback runs for a	682	This function is rarely useful, but when some event callback runs for a
585	very long time without entering the event loop, updating libev's idea of	683	very long time without entering the event loop, updating libev's idea of
586	the current time is a good idea.	684	the current time is a good idea.
587		685
588	See also "The special problem of time updates" in the C<ev_timer> section.	686	See also L<The special problem of time updates> in the C<ev_timer> section.
		687
		688	=item ev_suspend (loop)
		689
		690	=item ev_resume (loop)
		691
		692	These two functions suspend and resume a loop, for use when the loop is
		693	not used for a while and timeouts should not be processed.
		694
		695	A typical use case would be an interactive program such as a game: When
		696	the user presses C<^Z> to suspend the game and resumes it an hour later it
		697	would be best to handle timeouts as if no time had actually passed while
		698	the program was suspended. This can be achieved by calling C<ev_suspend>
		699	in your C<SIGTSTP> handler, sending yourself a C<SIGSTOP> and calling
		700	C<ev_resume> directly afterwards to resume timer processing.
		701
		702	Effectively, all C<ev_timer> watchers will be delayed by the time spend
		703	between C<ev_suspend> and C<ev_resume>, and all C<ev_periodic> watchers
		704	will be rescheduled (that is, they will lose any events that would have
		705	occured while suspended).
		706
		707	After calling C<ev_suspend> you B<must not> call I<any> function on the
		708	given loop other than C<ev_resume>, and you B<must not> call C<ev_resume>
		709	without a previous call to C<ev_suspend>.
		710
		711	Calling C<ev_suspend>/C<ev_resume> has the side effect of updating the
		712	event loop time (see C<ev_now_update>).
589		713
590	=item ev_loop (loop, int flags)	714	=item ev_loop (loop, int flags)
591		715
592	Finally, this is it, the event handler. This function usually is called	716	Finally, this is it, the event handler. This function usually is called
593	after you initialised all your watchers and you want to start handling	717	after you have initialised all your watchers and you want to start
594	events.	718	handling events.
595		719
596	If the flags argument is specified as C<0>, it will not return until	720	If the flags argument is specified as C<0>, it will not return until
597	either no event watchers are active anymore or C<ev_unloop> was called.	721	either no event watchers are active anymore or C<ev_unloop> was called.
598		722
599	Please note that an explicit C<ev_unloop> is usually better than	723	Please note that an explicit C<ev_unloop> is usually better than
600	relying on all watchers to be stopped when deciding when a program has	724	relying on all watchers to be stopped when deciding when a program has
601	finished (especially in interactive programs), but having a program that	725	finished (especially in interactive programs), but having a program
602	automatically loops as long as it has to and no longer by virtue of	726	that automatically loops as long as it has to and no longer by virtue
603	relying on its watchers stopping correctly is a thing of beauty.	727	of relying on its watchers stopping correctly, that is truly a thing of
		728	beauty.
604		729
605	A flags value of C<EVLOOP_NONBLOCK> will look for new events, will handle	730	A flags value of C<EVLOOP_NONBLOCK> will look for new events, will handle
606	those events and any outstanding ones, but will not block your process in	731	those events and any already outstanding ones, but will not block your
607	case there are no events and will return after one iteration of the loop.	732	process in case there are no events and will return after one iteration of
		733	the loop.
608		734
609	A flags value of C<EVLOOP_ONESHOT> will look for new events (waiting if	735	A flags value of C<EVLOOP_ONESHOT> will look for new events (waiting if
610	necessary) and will handle those and any outstanding ones. It will block	736	necessary) and will handle those and any already outstanding ones. It
611	your process until at least one new event arrives, and will return after	737	will block your process until at least one new event arrives (which could
612	one iteration of the loop. This is useful if you are waiting for some	738	be an event internal to libev itself, so there is no guarantee that a
613	external event in conjunction with something not expressible using other	739	user-registered callback will be called), and will return after one
		740	iteration of the loop.
		741
		742	This is useful if you are waiting for some external event in conjunction
		743	with something not expressible using other libev watchers (i.e. "roll your
614	libev watchers. However, a pair of C<ev_prepare>/C<ev_check> watchers is	744	own C<ev_loop>"). However, a pair of C<ev_prepare>/C<ev_check> watchers is
615	usually a better approach for this kind of thing.	745	usually a better approach for this kind of thing.
616		746
617	Here are the gory details of what C<ev_loop> does:	747	Here are the gory details of what C<ev_loop> does:
618		748
619	- Before the first iteration, call any pending watchers.	749	- Before the first iteration, call any pending watchers.
…		…
629	any active watchers at all will result in not sleeping).	759	any active watchers at all will result in not sleeping).
630	- Sleep if the I/O and timer collect interval say so.	760	- Sleep if the I/O and timer collect interval say so.
631	- Block the process, waiting for any events.	761	- Block the process, waiting for any events.
632	- Queue all outstanding I/O (fd) events.	762	- Queue all outstanding I/O (fd) events.
633	- Update the "event loop time" (ev_now ()), and do time jump adjustments.	763	- Update the "event loop time" (ev_now ()), and do time jump adjustments.
634	- Queue all outstanding timers.	764	- Queue all expired timers.
635	- Queue all outstanding periodics.	765	- Queue all expired periodics.
636	- Unless any events are pending now, queue all idle watchers.	766	- Unless any events are pending now, queue all idle watchers.
637	- Queue all check watchers.	767	- Queue all check watchers.
638	- Call all queued watchers in reverse order (i.e. check watchers first).	768	- Call all queued watchers in reverse order (i.e. check watchers first).
639	Signals and child watchers are implemented as I/O watchers, and will	769	Signals and child watchers are implemented as I/O watchers, and will
640	be handled here by queueing them when their watcher gets executed.	770	be handled here by queueing them when their watcher gets executed.
…		…
657	C<EVUNLOOP_ONE>, which will make the innermost C<ev_loop> call return, or	787	C<EVUNLOOP_ONE>, which will make the innermost C<ev_loop> call return, or
658	C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.	788	C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.
659		789
660	This "unloop state" will be cleared when entering C<ev_loop> again.	790	This "unloop state" will be cleared when entering C<ev_loop> again.
661		791
		792	It is safe to call C<ev_unloop> from otuside any C<ev_loop> calls.
		793
662	=item ev_ref (loop)	794	=item ev_ref (loop)
663		795
664	=item ev_unref (loop)	796	=item ev_unref (loop)
665		797
666	Ref/unref can be used to add or remove a reference count on the event	798	Ref/unref can be used to add or remove a reference count on the event
667	loop: Every watcher keeps one reference, and as long as the reference	799	loop: Every watcher keeps one reference, and as long as the reference
668	count is nonzero, C<ev_loop> will not return on its own. If you have	800	count is nonzero, C<ev_loop> will not return on its own.
669	a watcher you never unregister that should not keep C<ev_loop> from	801
670	returning, ev_unref() after starting, and ev_ref() before stopping it. For	802	This is useful when you have a watcher that you never intend to
		803	unregister, but that nevertheless should not keep C<ev_loop> from
		804	returning. In such a case, call C<ev_unref> after starting, and C<ev_ref>
		805	before stopping it.
		806
671	example, libev itself uses this for its internal signal pipe: It is not	807	As an example, libev itself uses this for its internal signal pipe: It
672	visible to the libev user and should not keep C<ev_loop> from exiting if	808	is not visible to the libev user and should not keep C<ev_loop> from
673	no event watchers registered by it are active. It is also an excellent	809	exiting if no event watchers registered by it are active. It is also an
674	way to do this for generic recurring timers or from within third-party	810	excellent way to do this for generic recurring timers or from within
675	libraries. Just remember to I<unref after start> and I<ref before stop>	811	third-party libraries. Just remember to I<unref after start> and I<ref
676	(but only if the watcher wasn't active before, or was active before,	812	before stop> (but only if the watcher wasn't active before, or was active
677	respectively).	813	before, respectively. Note also that libev might stop watchers itself
		814	(e.g. non-repeating timers) in which case you have to C<ev_ref>
		815	in the callback).
678		816
679	Example: Create a signal watcher, but keep it from keeping C<ev_loop>	817	Example: Create a signal watcher, but keep it from keeping C<ev_loop>
680	running when nothing else is active.	818	running when nothing else is active.
681		819
682	struct ev_signal exitsig;	820	ev_signal exitsig;
683	ev_signal_init (&exitsig, sig_cb, SIGINT);	821	ev_signal_init (&exitsig, sig_cb, SIGINT);
684	ev_signal_start (loop, &exitsig);	822	ev_signal_start (loop, &exitsig);
685	evf_unref (loop);	823	evf_unref (loop);
686		824
687	Example: For some weird reason, unregister the above signal handler again.	825	Example: For some weird reason, unregister the above signal handler again.
…		…
701	Setting these to a higher value (the C<interval> I<must> be >= C<0>)	839	Setting these to a higher value (the C<interval> I<must> be >= C<0>)
702	allows libev to delay invocation of I/O and timer/periodic callbacks	840	allows libev to delay invocation of I/O and timer/periodic callbacks
703	to increase efficiency of loop iterations (or to increase power-saving	841	to increase efficiency of loop iterations (or to increase power-saving
704	opportunities).	842	opportunities).
705		843
706	The background is that sometimes your program runs just fast enough to	844	The idea is that sometimes your program runs just fast enough to handle
707	handle one (or very few) event(s) per loop iteration. While this makes	845	one (or very few) event(s) per loop iteration. While this makes the
708	the program responsive, it also wastes a lot of CPU time to poll for new	846	program responsive, it also wastes a lot of CPU time to poll for new
709	events, especially with backends like C<select ()> which have a high	847	events, especially with backends like C<select ()> which have a high
710	overhead for the actual polling but can deliver many events at once.	848	overhead for the actual polling but can deliver many events at once.
711		849
712	By setting a higher I<io collect interval> you allow libev to spend more	850	By setting a higher I<io collect interval> you allow libev to spend more
713	time collecting I/O events, so you can handle more events per iteration,	851	time collecting I/O events, so you can handle more events per iteration,
714	at the cost of increasing latency. Timeouts (both C<ev_periodic> and	852	at the cost of increasing latency. Timeouts (both C<ev_periodic> and
715	C<ev_timer>) will be not affected. Setting this to a non-null value will	853	C<ev_timer>) will be not affected. Setting this to a non-null value will
716	introduce an additional C<ev_sleep ()> call into most loop iterations.	854	introduce an additional C<ev_sleep ()> call into most loop iterations. The
		855	sleep time ensures that libev will not poll for I/O events more often then
		856	once per this interval, on average.
717		857
718	Likewise, by setting a higher I<timeout collect interval> you allow libev	858	Likewise, by setting a higher I<timeout collect interval> you allow libev
719	to spend more time collecting timeouts, at the expense of increased	859	to spend more time collecting timeouts, at the expense of increased
720	latency (the watcher callback will be called later). C<ev_io> watchers	860	latency/jitter/inexactness (the watcher callback will be called
721	will not be affected. Setting this to a non-null value will not introduce	861	later). C<ev_io> watchers will not be affected. Setting this to a non-null
722	any overhead in libev.	862	value will not introduce any overhead in libev.
723		863
724	Many (busy) programs can usually benefit by setting the I/O collect	864	Many (busy) programs can usually benefit by setting the I/O collect
725	interval to a value near C<0.1> or so, which is often enough for	865	interval to a value near C<0.1> or so, which is often enough for
726	interactive servers (of course not for games), likewise for timeouts. It	866	interactive servers (of course not for games), likewise for timeouts. It
727	usually doesn't make much sense to set it to a lower value than C<0.01>,	867	usually doesn't make much sense to set it to a lower value than C<0.01>,
728	as this approaches the timing granularity of most systems.	868	as this approaches the timing granularity of most systems. Note that if
		869	you do transactions with the outside world and you can't increase the
		870	parallelity, then this setting will limit your transaction rate (if you
		871	need to poll once per transaction and the I/O collect interval is 0.01,
		872	then you can't do more than 100 transations per second).
729		873
730	Setting the I<timeout collect interval> can improve the opportunity for	874	Setting the I<timeout collect interval> can improve the opportunity for
731	saving power, as the program will "bundle" timer callback invocations that	875	saving power, as the program will "bundle" timer callback invocations that
732	are "near" in time together, by delaying some, thus reducing the number of	876	are "near" in time together, by delaying some, thus reducing the number of
733	times the process sleeps and wakes up again. Another useful technique to	877	times the process sleeps and wakes up again. Another useful technique to
734	reduce iterations/wake-ups is to use C<ev_periodic> watchers and make sure	878	reduce iterations/wake-ups is to use C<ev_periodic> watchers and make sure
735	they fire on, say, one-second boundaries only.	879	they fire on, say, one-second boundaries only.
736		880
		881	Example: we only need 0.1s timeout granularity, and we wish not to poll
		882	more often than 100 times per second:
		883
		884	ev_set_timeout_collect_interval (EV_DEFAULT_UC_ 0.1);
		885	ev_set_io_collect_interval (EV_DEFAULT_UC_ 0.01);
		886
		887	=item ev_invoke_pending (loop)
		888
		889	This call will simply invoke all pending watchers while resetting their
		890	pending state. Normally, C<ev_loop> does this automatically when required,
		891	but when overriding the invoke callback this call comes handy.
		892
		893	=item int ev_pending_count (loop)
		894
		895	Returns the number of pending watchers - zero indicates that no watchers
		896	are pending.
		897
		898	=item ev_set_invoke_pending_cb (loop, void (*invoke_pending_cb)(EV_P))
		899
		900	This overrides the invoke pending functionality of the loop: Instead of
		901	invoking all pending watchers when there are any, C<ev_loop> will call
		902	this callback instead. This is useful, for example, when you want to
		903	invoke the actual watchers inside another context (another thread etc.).
		904
		905	If you want to reset the callback, use C<ev_invoke_pending> as new
		906	callback.
		907
		908	=item ev_set_loop_release_cb (loop, void (*release)(EV_P), void (*acquire)(EV_P))
		909
		910	Sometimes you want to share the same loop between multiple threads. This
		911	can be done relatively simply by putting mutex_lock/unlock calls around
		912	each call to a libev function.
		913
		914	However, C<ev_loop> can run an indefinite time, so it is not feasible to
		915	wait for it to return. One way around this is to wake up the loop via
		916	C<ev_unloop> and C<av_async_send>, another way is to set these I<release>
		917	and I<acquire> callbacks on the loop.
		918
		919	When set, then C<release> will be called just before the thread is
		920	suspended waiting for new events, and C<acquire> is called just
		921	afterwards.
		922
		923	Ideally, C<release> will just call your mutex_unlock function, and
		924	C<acquire> will just call the mutex_lock function again.
		925
		926	While event loop modifications are allowed between invocations of
		927	C<release> and C<acquire> (that's their only purpose after all), no
		928	modifications done will affect the event loop, i.e. adding watchers will
		929	have no effect on the set of file descriptors being watched, or the time
		930	waited. Use an C<ev_async> watcher to wake up C<ev_loop> when you want it
		931	to take note of any changes you made.
		932
		933	In theory, threads executing C<ev_loop> will be async-cancel safe between
		934	invocations of C<release> and C<acquire>.
		935
		936	See also the locking example in the C<THREADS> section later in this
		937	document.
		938
		939	=item ev_set_userdata (loop, void *data)
		940
		941	=item ev_userdata (loop)
		942
		943	Set and retrieve a single C<void *> associated with a loop. When
		944	C<ev_set_userdata> has never been called, then C<ev_userdata> returns
		945	C<0.>
		946
		947	These two functions can be used to associate arbitrary data with a loop,
		948	and are intended solely for the C<invoke_pending_cb>, C<release> and
		949	C<acquire> callbacks described above, but of course can be (ab-)used for
		950	any other purpose as well.
		951
737	=item ev_loop_verify (loop)	952	=item ev_loop_verify (loop)
738		953
739	This function only does something when C<EV_VERIFY> support has been	954	This function only does something when C<EV_VERIFY> support has been
740	compiled in. It tries to go through all internal structures and checks	955	compiled in, which is the default for non-minimal builds. It tries to go
741	them for validity. If anything is found to be inconsistent, it will print	956	through all internal structures and checks them for validity. If anything
742	an error message to standard error and call C<abort ()>.	957	is found to be inconsistent, it will print an error message to standard
		958	error and call C<abort ()>.
743		959
744	This can be used to catch bugs inside libev itself: under normal	960	This can be used to catch bugs inside libev itself: under normal
745	circumstances, this function will never abort as of course libev keeps its	961	circumstances, this function will never abort as of course libev keeps its
746	data structures consistent.	962	data structures consistent.
747		963
748	=back	964	=back
749		965
750		966
751	=head1 ANATOMY OF A WATCHER	967	=head1 ANATOMY OF A WATCHER
752		968
		969	In the following description, uppercase C<TYPE> in names stands for the
		970	watcher type, e.g. C<ev_TYPE_start> can mean C<ev_timer_start> for timer
		971	watchers and C<ev_io_start> for I/O watchers.
		972
753	A watcher is a structure that you create and register to record your	973	A watcher is a structure that you create and register to record your
754	interest in some event. For instance, if you want to wait for STDIN to	974	interest in some event. For instance, if you want to wait for STDIN to
755	become readable, you would create an C<ev_io> watcher for that:	975	become readable, you would create an C<ev_io> watcher for that:
756		976
757	static void my_cb (struct ev_loop *loop, struct ev_io *w, int revents)	977	static void my_cb (struct ev_loop *loop, ev_io *w, int revents)
758	{	978	{
759	ev_io_stop (w);	979	ev_io_stop (w);
760	ev_unloop (loop, EVUNLOOP_ALL);	980	ev_unloop (loop, EVUNLOOP_ALL);
761	}	981	}
762		982
763	struct ev_loop *loop = ev_default_loop (0);	983	struct ev_loop *loop = ev_default_loop (0);
		984
764	struct ev_io stdin_watcher;	985	ev_io stdin_watcher;
		986
765	ev_init (&stdin_watcher, my_cb);	987	ev_init (&stdin_watcher, my_cb);
766	ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ);	988	ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ);
767	ev_io_start (loop, &stdin_watcher);	989	ev_io_start (loop, &stdin_watcher);
		990
768	ev_loop (loop, 0);	991	ev_loop (loop, 0);
769		992
770	As you can see, you are responsible for allocating the memory for your	993	As you can see, you are responsible for allocating the memory for your
771	watcher structures (and it is usually a bad idea to do this on the stack,	994	watcher structures (and it is I<usually> a bad idea to do this on the
772	although this can sometimes be quite valid).	995	stack).
		996
		997	Each watcher has an associated watcher structure (called C<struct ev_TYPE>
		998	or simply C<ev_TYPE>, as typedefs are provided for all watcher structs).
773		999
774	Each watcher structure must be initialised by a call to C<ev_init	1000	Each watcher structure must be initialised by a call to C<ev_init
775	(watcher *, callback)>, which expects a callback to be provided. This	1001	(watcher *, callback)>, which expects a callback to be provided. This
776	callback gets invoked each time the event occurs (or, in the case of I/O	1002	callback gets invoked each time the event occurs (or, in the case of I/O
777	watchers, each time the event loop detects that the file descriptor given	1003	watchers, each time the event loop detects that the file descriptor given
778	is readable and/or writable).	1004	is readable and/or writable).
779		1005
780	Each watcher type has its own C<< ev_<type>_set (watcher *, ...) >> macro	1006	Each watcher type further has its own C<< ev_TYPE_set (watcher *, ...) >>
781	with arguments specific to this watcher type. There is also a macro	1007	macro to configure it, with arguments specific to the watcher type. There
782	to combine initialisation and setting in one call: C<< ev_<type>_init	1008	is also a macro to combine initialisation and setting in one call: C<<
783	(watcher *, callback, ...) >>.	1009	ev_TYPE_init (watcher *, callback, ...) >>.
784		1010
785	To make the watcher actually watch out for events, you have to start it	1011	To make the watcher actually watch out for events, you have to start it
786	with a watcher-specific start function (C<< ev_<type>_start (loop, watcher	1012	with a watcher-specific start function (C<< ev_TYPE_start (loop, watcher
787	*) >>), and you can stop watching for events at any time by calling the	1013	*) >>), and you can stop watching for events at any time by calling the
788	corresponding stop function (C<< ev_<type>_stop (loop, watcher *) >>.	1014	corresponding stop function (C<< ev_TYPE_stop (loop, watcher *) >>.
789		1015
790	As long as your watcher is active (has been started but not stopped) you	1016	As long as your watcher is active (has been started but not stopped) you
791	must not touch the values stored in it. Most specifically you must never	1017	must not touch the values stored in it. Most specifically you must never
792	reinitialise it or call its C<set> macro.	1018	reinitialise it or call its C<ev_TYPE_set> macro.
793		1019
794	Each and every callback receives the event loop pointer as first, the	1020	Each and every callback receives the event loop pointer as first, the
795	registered watcher structure as second, and a bitset of received events as	1021	registered watcher structure as second, and a bitset of received events as
796	third argument.	1022	third argument.
797		1023
…		…
855		1081
856	=item C<EV_ASYNC>	1082	=item C<EV_ASYNC>
857		1083
858	The given async watcher has been asynchronously notified (see C<ev_async>).	1084	The given async watcher has been asynchronously notified (see C<ev_async>).
859		1085
		1086	=item C<EV_CUSTOM>
		1087
		1088	Not ever sent (or otherwise used) by libev itself, but can be freely used
		1089	by libev users to signal watchers (e.g. via C<ev_feed_event>).
		1090
860	=item C<EV_ERROR>	1091	=item C<EV_ERROR>
861		1092
862	An unspecified error has occurred, the watcher has been stopped. This might	1093	An unspecified error has occurred, the watcher has been stopped. This might
863	happen because the watcher could not be properly started because libev	1094	happen because the watcher could not be properly started because libev
864	ran out of memory, a file descriptor was found to be closed or any other	1095	ran out of memory, a file descriptor was found to be closed or any other
		1096	problem. Libev considers these application bugs.
		1097
865	problem. You best act on it by reporting the problem and somehow coping	1098	You best act on it by reporting the problem and somehow coping with the
866	with the watcher being stopped.	1099	watcher being stopped. Note that well-written programs should not receive
		1100	an error ever, so when your watcher receives it, this usually indicates a
		1101	bug in your program.
867		1102
868	Libev will usually signal a few "dummy" events together with an error,	1103	Libev will usually signal a few "dummy" events together with an error, for
869	for example it might indicate that a fd is readable or writable, and if	1104	example it might indicate that a fd is readable or writable, and if your
870	your callbacks is well-written it can just attempt the operation and cope	1105	callbacks is well-written it can just attempt the operation and cope with
871	with the error from read() or write(). This will not work in multi-threaded	1106	the error from read() or write(). This will not work in multi-threaded
872	programs, though, so beware.	1107	programs, though, as the fd could already be closed and reused for another
		1108	thing, so beware.
873		1109
874	=back	1110	=back
875		1111
876	=head2 GENERIC WATCHER FUNCTIONS	1112	=head2 GENERIC WATCHER FUNCTIONS
877
878	In the following description, C<TYPE> stands for the watcher type,
879	e.g. C<timer> for C<ev_timer> watchers and C<io> for C<ev_io> watchers.
880		1113
881	=over 4	1114	=over 4
882		1115
883	=item C<ev_init> (ev_TYPE *watcher, callback)	1116	=item C<ev_init> (ev_TYPE *watcher, callback)
884		1117
…		…
890	which rolls both calls into one.	1123	which rolls both calls into one.
891		1124
892	You can reinitialise a watcher at any time as long as it has been stopped	1125	You can reinitialise a watcher at any time as long as it has been stopped
893	(or never started) and there are no pending events outstanding.	1126	(or never started) and there are no pending events outstanding.
894		1127
895	The callback is always of type C<void (*)(ev_loop *loop, ev_TYPE *watcher,	1128	The callback is always of type C<void (*)(struct ev_loop *loop, ev_TYPE *watcher,
896	int revents)>.	1129	int revents)>.
897		1130
		1131	Example: Initialise an C<ev_io> watcher in two steps.
		1132
		1133	ev_io w;
		1134	ev_init (&w, my_cb);
		1135	ev_io_set (&w, STDIN_FILENO, EV_READ);
		1136
898	=item C<ev_TYPE_set> (ev_TYPE *, [args])	1137	=item C<ev_TYPE_set> (ev_TYPE *watcher, [args])
899		1138
900	This macro initialises the type-specific parts of a watcher. You need to	1139	This macro initialises the type-specific parts of a watcher. You need to
901	call C<ev_init> at least once before you call this macro, but you can	1140	call C<ev_init> at least once before you call this macro, but you can
902	call C<ev_TYPE_set> any number of times. You must not, however, call this	1141	call C<ev_TYPE_set> any number of times. You must not, however, call this
903	macro on a watcher that is active (it can be pending, however, which is a	1142	macro on a watcher that is active (it can be pending, however, which is a
904	difference to the C<ev_init> macro).	1143	difference to the C<ev_init> macro).
905		1144
906	Although some watcher types do not have type-specific arguments	1145	Although some watcher types do not have type-specific arguments
907	(e.g. C<ev_prepare>) you still need to call its C<set> macro.	1146	(e.g. C<ev_prepare>) you still need to call its C<set> macro.
908		1147
		1148	See C<ev_init>, above, for an example.
		1149
909	=item C<ev_TYPE_init> (ev_TYPE *watcher, callback, [args])	1150	=item C<ev_TYPE_init> (ev_TYPE *watcher, callback, [args])
910		1151
911	This convenience macro rolls both C<ev_init> and C<ev_TYPE_set> macro	1152	This convenience macro rolls both C<ev_init> and C<ev_TYPE_set> macro
912	calls into a single call. This is the most convenient method to initialise	1153	calls into a single call. This is the most convenient method to initialise
913	a watcher. The same limitations apply, of course.	1154	a watcher. The same limitations apply, of course.
914		1155
		1156	Example: Initialise and set an C<ev_io> watcher in one step.
		1157
		1158	ev_io_init (&w, my_cb, STDIN_FILENO, EV_READ);
		1159
915	=item C<ev_TYPE_start> (loop *, ev_TYPE *watcher)	1160	=item C<ev_TYPE_start> (loop, ev_TYPE *watcher)
916		1161
917	Starts (activates) the given watcher. Only active watchers will receive	1162	Starts (activates) the given watcher. Only active watchers will receive
918	events. If the watcher is already active nothing will happen.	1163	events. If the watcher is already active nothing will happen.
919		1164
		1165	Example: Start the C<ev_io> watcher that is being abused as example in this
		1166	whole section.
		1167
		1168	ev_io_start (EV_DEFAULT_UC, &w);
		1169
920	=item C<ev_TYPE_stop> (loop *, ev_TYPE *watcher)	1170	=item C<ev_TYPE_stop> (loop, ev_TYPE *watcher)
921		1171
922	Stops the given watcher again (if active) and clears the pending	1172	Stops the given watcher if active, and clears the pending status (whether
		1173	the watcher was active or not).
		1174
923	status. It is possible that stopped watchers are pending (for example,	1175	It is possible that stopped watchers are pending - for example,
924	non-repeating timers are being stopped when they become pending), but	1176	non-repeating timers are being stopped when they become pending - but
925	C<ev_TYPE_stop> ensures that the watcher is neither active nor pending. If	1177	calling C<ev_TYPE_stop> ensures that the watcher is neither active nor
926	you want to free or reuse the memory used by the watcher it is therefore a	1178	pending. If you want to free or reuse the memory used by the watcher it is
927	good idea to always call its C<ev_TYPE_stop> function.	1179	therefore a good idea to always call its C<ev_TYPE_stop> function.
928		1180
929	=item bool ev_is_active (ev_TYPE *watcher)	1181	=item bool ev_is_active (ev_TYPE *watcher)
930		1182
931	Returns a true value iff the watcher is active (i.e. it has been started	1183	Returns a true value iff the watcher is active (i.e. it has been started
932	and not yet been stopped). As long as a watcher is active you must not modify	1184	and not yet been stopped). As long as a watcher is active you must not modify
…		…
948	=item ev_cb_set (ev_TYPE *watcher, callback)	1200	=item ev_cb_set (ev_TYPE *watcher, callback)
949		1201
950	Change the callback. You can change the callback at virtually any time	1202	Change the callback. You can change the callback at virtually any time
951	(modulo threads).	1203	(modulo threads).
952		1204
953	=item ev_set_priority (ev_TYPE *watcher, priority)	1205	=item ev_set_priority (ev_TYPE *watcher, int priority)
954		1206
955	=item int ev_priority (ev_TYPE *watcher)	1207	=item int ev_priority (ev_TYPE *watcher)
956		1208
957	Set and query the priority of the watcher. The priority is a small	1209	Set and query the priority of the watcher. The priority is a small
958	integer between C<EV_MAXPRI> (default: C<2>) and C<EV_MINPRI>	1210	integer between C<EV_MAXPRI> (default: C<2>) and C<EV_MINPRI>
959	(default: C<-2>). Pending watchers with higher priority will be invoked	1211	(default: C<-2>). Pending watchers with higher priority will be invoked
960	before watchers with lower priority, but priority will not keep watchers	1212	before watchers with lower priority, but priority will not keep watchers
961	from being executed (except for C<ev_idle> watchers).	1213	from being executed (except for C<ev_idle> watchers).
962		1214
963	This means that priorities are I<only> used for ordering callback
964	invocation after new events have been received. This is useful, for
965	example, to reduce latency after idling, or more often, to bind two
966	watchers on the same event and make sure one is called first.
967
968	If you need to suppress invocation when higher priority events are pending	1215	If you need to suppress invocation when higher priority events are pending
969	you need to look at C<ev_idle> watchers, which provide this functionality.	1216	you need to look at C<ev_idle> watchers, which provide this functionality.
970		1217
971	You I<must not> change the priority of a watcher as long as it is active or	1218	You I<must not> change the priority of a watcher as long as it is active or
972	pending.	1219	pending.
973		1220
		1221	Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is
		1222	fine, as long as you do not mind that the priority value you query might
		1223	or might not have been clamped to the valid range.
		1224
974	The default priority used by watchers when no priority has been set is	1225	The default priority used by watchers when no priority has been set is
975	always C<0>, which is supposed to not be too high and not be too low :).	1226	always C<0>, which is supposed to not be too high and not be too low :).
976		1227
977	Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is	1228	See L<WATCHER PRIORITY MODELS>, below, for a more thorough treatment of
978	fine, as long as you do not mind that the priority value you query might	1229	priorities.
979	or might not have been adjusted to be within valid range.
980		1230
981	=item ev_invoke (loop, ev_TYPE *watcher, int revents)	1231	=item ev_invoke (loop, ev_TYPE *watcher, int revents)
982		1232
983	Invoke the C<watcher> with the given C<loop> and C<revents>. Neither	1233	Invoke the C<watcher> with the given C<loop> and C<revents>. Neither
984	C<loop> nor C<revents> need to be valid as long as the watcher callback	1234	C<loop> nor C<revents> need to be valid as long as the watcher callback
985	can deal with that fact.	1235	can deal with that fact, as both are simply passed through to the
		1236	callback.
986		1237
987	=item int ev_clear_pending (loop, ev_TYPE *watcher)	1238	=item int ev_clear_pending (loop, ev_TYPE *watcher)
988		1239
989	If the watcher is pending, this function returns clears its pending status	1240	If the watcher is pending, this function clears its pending status and
990	and returns its C<revents> bitset (as if its callback was invoked). If the	1241	returns its C<revents> bitset (as if its callback was invoked). If the
991	watcher isn't pending it does nothing and returns C<0>.	1242	watcher isn't pending it does nothing and returns C<0>.
992		1243
		1244	Sometimes it can be useful to "poll" a watcher instead of waiting for its
		1245	callback to be invoked, which can be accomplished with this function.
		1246
		1247	=item ev_feed_event (loop, ev_TYPE *watcher, int revents)
		1248
		1249	Feeds the given event set into the event loop, as if the specified event
		1250	had happened for the specified watcher (which must be a pointer to an
		1251	initialised but not necessarily started event watcher). Obviously you must
		1252	not free the watcher as long as it has pending events.
		1253
		1254	Stopping the watcher, letting libev invoke it, or calling
		1255	C<ev_clear_pending> will clear the pending event, even if the watcher was
		1256	not started in the first place.
		1257
		1258	See also C<ev_feed_fd_event> and C<ev_feed_signal_event> for related
		1259	functions that do not need a watcher.
		1260
993	=back	1261	=back
994		1262
995		1263
996	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER	1264	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER
997		1265
998	Each watcher has, by default, a member C<void *data> that you can change	1266	Each watcher has, by default, a member C<void *data> that you can change
999	and read at any time, libev will completely ignore it. This can be used	1267	and read at any time: libev will completely ignore it. This can be used
1000	to associate arbitrary data with your watcher. If you need more data and	1268	to associate arbitrary data with your watcher. If you need more data and
1001	don't want to allocate memory and store a pointer to it in that data	1269	don't want to allocate memory and store a pointer to it in that data
1002	member, you can also "subclass" the watcher type and provide your own	1270	member, you can also "subclass" the watcher type and provide your own
1003	data:	1271	data:
1004		1272
1005	struct my_io	1273	struct my_io
1006	{	1274	{
1007	struct ev_io io;	1275	ev_io io;
1008	int otherfd;	1276	int otherfd;
1009	void *somedata;	1277	void *somedata;
1010	struct whatever *mostinteresting;	1278	struct whatever *mostinteresting;
1011	};	1279	};
1012		1280
…		…
1015	ev_io_init (&w.io, my_cb, fd, EV_READ);	1283	ev_io_init (&w.io, my_cb, fd, EV_READ);
1016		1284
1017	And since your callback will be called with a pointer to the watcher, you	1285	And since your callback will be called with a pointer to the watcher, you
1018	can cast it back to your own type:	1286	can cast it back to your own type:
1019		1287
1020	static void my_cb (struct ev_loop *loop, struct ev_io *w_, int revents)	1288	static void my_cb (struct ev_loop *loop, ev_io *w_, int revents)
1021	{	1289	{
1022	struct my_io *w = (struct my_io *)w_;	1290	struct my_io *w = (struct my_io *)w_;
1023	...	1291	...
1024	}	1292	}
1025		1293
…		…
1036	ev_timer t2;	1304	ev_timer t2;
1037	}	1305	}
1038		1306
1039	In this case getting the pointer to C<my_biggy> is a bit more	1307	In this case getting the pointer to C<my_biggy> is a bit more
1040	complicated: Either you store the address of your C<my_biggy> struct	1308	complicated: Either you store the address of your C<my_biggy> struct
1041	in the C<data> member of the watcher, or you need to use some pointer	1309	in the C<data> member of the watcher (for woozies), or you need to use
1042	arithmetic using C<offsetof> inside your watchers:	1310	some pointer arithmetic using C<offsetof> inside your watchers (for real
		1311	programmers):
1043		1312
1044	#include <stddef.h>	1313	#include <stddef.h>
1045		1314
1046	static void	1315	static void
1047	t1_cb (EV_P_ struct ev_timer *w, int revents)	1316	t1_cb (EV_P_ ev_timer *w, int revents)
1048	{	1317	{
1049	struct my_biggy big = (struct my_biggy *	1318	struct my_biggy big = (struct my_biggy *)
1050	(((char *)w) - offsetof (struct my_biggy, t1));	1319	(((char *)w) - offsetof (struct my_biggy, t1));
1051	}	1320	}
1052		1321
1053	static void	1322	static void
1054	t2_cb (EV_P_ struct ev_timer *w, int revents)	1323	t2_cb (EV_P_ ev_timer *w, int revents)
1055	{	1324	{
1056	struct my_biggy big = (struct my_biggy *	1325	struct my_biggy big = (struct my_biggy *)
1057	(((char *)w) - offsetof (struct my_biggy, t2));	1326	(((char *)w) - offsetof (struct my_biggy, t2));
1058	}	1327	}
		1328
		1329	=head2 WATCHER PRIORITY MODELS
		1330
		1331	Many event loops support I<watcher priorities>, which are usually small
		1332	integers that influence the ordering of event callback invocation
		1333	between watchers in some way, all else being equal.
		1334
		1335	In libev, Watcher priorities can be set using C<ev_set_priority>. See its
		1336	description for the more technical details such as the actual priority
		1337	range.
		1338
		1339	There are two common ways how these these priorities are being interpreted
		1340	by event loops:
		1341
		1342	In the more common lock-out model, higher priorities "lock out" invocation
		1343	of lower priority watchers, which means as long as higher priority
		1344	watchers receive events, lower priority watchers are not being invoked.
		1345
		1346	The less common only-for-ordering model uses priorities solely to order
		1347	callback invocation within a single event loop iteration: Higher priority
		1348	watchers are invoked before lower priority ones, but they all get invoked
		1349	before polling for new events.
		1350
		1351	Libev uses the second (only-for-ordering) model for all its watchers
		1352	except for idle watchers (which use the lock-out model).
		1353
		1354	The rationale behind this is that implementing the lock-out model for
		1355	watchers is not well supported by most kernel interfaces, and most event
		1356	libraries will just poll for the same events again and again as long as
		1357	their callbacks have not been executed, which is very inefficient in the
		1358	common case of one high-priority watcher locking out a mass of lower
		1359	priority ones.
		1360
		1361	Static (ordering) priorities are most useful when you have two or more
		1362	watchers handling the same resource: a typical usage example is having an
		1363	C<ev_io> watcher to receive data, and an associated C<ev_timer> to handle
		1364	timeouts. Under load, data might be received while the program handles
		1365	other jobs, but since timers normally get invoked first, the timeout
		1366	handler will be executed before checking for data. In that case, giving
		1367	the timer a lower priority than the I/O watcher ensures that I/O will be
		1368	handled first even under adverse conditions (which is usually, but not
		1369	always, what you want).
		1370
		1371	Since idle watchers use the "lock-out" model, meaning that idle watchers
		1372	will only be executed when no same or higher priority watchers have
		1373	received events, they can be used to implement the "lock-out" model when
		1374	required.
		1375
		1376	For example, to emulate how many other event libraries handle priorities,
		1377	you can associate an C<ev_idle> watcher to each such watcher, and in
		1378	the normal watcher callback, you just start the idle watcher. The real
		1379	processing is done in the idle watcher callback. This causes libev to
		1380	continously poll and process kernel event data for the watcher, but when
		1381	the lock-out case is known to be rare (which in turn is rare :), this is
		1382	workable.
		1383
		1384	Usually, however, the lock-out model implemented that way will perform
		1385	miserably under the type of load it was designed to handle. In that case,
		1386	it might be preferable to stop the real watcher before starting the
		1387	idle watcher, so the kernel will not have to process the event in case
		1388	the actual processing will be delayed for considerable time.
		1389
		1390	Here is an example of an I/O watcher that should run at a strictly lower
		1391	priority than the default, and which should only process data when no
		1392	other events are pending:
		1393
		1394	ev_idle idle; // actual processing watcher
		1395	ev_io io; // actual event watcher
		1396
		1397	static void
		1398	io_cb (EV_P_ ev_io *w, int revents)
		1399	{
		1400	// stop the I/O watcher, we received the event, but
		1401	// are not yet ready to handle it.
		1402	ev_io_stop (EV_A_ w);
		1403
		1404	// start the idle watcher to ahndle the actual event.
		1405	// it will not be executed as long as other watchers
		1406	// with the default priority are receiving events.
		1407	ev_idle_start (EV_A_ &idle);
		1408	}
		1409
		1410	static void
		1411	idle_cb (EV_P_ ev_idle *w, int revents)
		1412	{
		1413	// actual processing
		1414	read (STDIN_FILENO, ...);
		1415
		1416	// have to start the I/O watcher again, as
		1417	// we have handled the event
		1418	ev_io_start (EV_P_ &io);
		1419	}
		1420
		1421	// initialisation
		1422	ev_idle_init (&idle, idle_cb);
		1423	ev_io_init (&io, io_cb, STDIN_FILENO, EV_READ);
		1424	ev_io_start (EV_DEFAULT_ &io);
		1425
		1426	In the "real" world, it might also be beneficial to start a timer, so that
		1427	low-priority connections can not be locked out forever under load. This
		1428	enables your program to keep a lower latency for important connections
		1429	during short periods of high load, while not completely locking out less
		1430	important ones.
1059		1431
1060		1432
1061	=head1 WATCHER TYPES	1433	=head1 WATCHER TYPES
1062		1434
1063	This section describes each watcher in detail, but will not repeat	1435	This section describes each watcher in detail, but will not repeat
…		…
1087	In general you can register as many read and/or write event watchers per	1459	In general you can register as many read and/or write event watchers per
1088	fd as you want (as long as you don't confuse yourself). Setting all file	1460	fd as you want (as long as you don't confuse yourself). Setting all file
1089	descriptors to non-blocking mode is also usually a good idea (but not	1461	descriptors to non-blocking mode is also usually a good idea (but not
1090	required if you know what you are doing).	1462	required if you know what you are doing).
1091		1463
1092	If you must do this, then force the use of a known-to-be-good backend	1464	If you cannot use non-blocking mode, then force the use of a
1093	(at the time of this writing, this includes only C<EVBACKEND_SELECT> and	1465	known-to-be-good backend (at the time of this writing, this includes only
1094	C<EVBACKEND_POLL>).	1466	C<EVBACKEND_SELECT> and C<EVBACKEND_POLL>). The same applies to file
		1467	descriptors for which non-blocking operation makes no sense (such as
		1468	files) - libev doesn't guarentee any specific behaviour in that case.
1095		1469
1096	Another thing you have to watch out for is that it is quite easy to	1470	Another thing you have to watch out for is that it is quite easy to
1097	receive "spurious" readiness notifications, that is your callback might	1471	receive "spurious" readiness notifications, that is your callback might
1098	be called with C<EV_READ> but a subsequent C<read>(2) will actually block	1472	be called with C<EV_READ> but a subsequent C<read>(2) will actually block
1099	because there is no data. Not only are some backends known to create a	1473	because there is no data. Not only are some backends known to create a
1100	lot of those (for example Solaris ports), it is very easy to get into	1474	lot of those (for example Solaris ports), it is very easy to get into
1101	this situation even with a relatively standard program structure. Thus	1475	this situation even with a relatively standard program structure. Thus
1102	it is best to always use non-blocking I/O: An extra C<read>(2) returning	1476	it is best to always use non-blocking I/O: An extra C<read>(2) returning
1103	C<EAGAIN> is far preferable to a program hanging until some data arrives.	1477	C<EAGAIN> is far preferable to a program hanging until some data arrives.
1104		1478
1105	If you cannot run the fd in non-blocking mode (for example you should not	1479	If you cannot run the fd in non-blocking mode (for example you should
1106	play around with an Xlib connection), then you have to separately re-test	1480	not play around with an Xlib connection), then you have to separately
1107	whether a file descriptor is really ready with a known-to-be good interface	1481	re-test whether a file descriptor is really ready with a known-to-be good
1108	such as poll (fortunately in our Xlib example, Xlib already does this on	1482	interface such as poll (fortunately in our Xlib example, Xlib already
1109	its own, so its quite safe to use).	1483	does this on its own, so its quite safe to use). Some people additionally
		1484	use C<SIGALRM> and an interval timer, just to be sure you won't block
		1485	indefinitely.
		1486
		1487	But really, best use non-blocking mode.
1110		1488
1111	=head3 The special problem of disappearing file descriptors	1489	=head3 The special problem of disappearing file descriptors
1112		1490
1113	Some backends (e.g. kqueue, epoll) need to be told about closing a file	1491	Some backends (e.g. kqueue, epoll) need to be told about closing a file
1114	descriptor (either by calling C<close> explicitly or by any other means,	1492	descriptor (either due to calling C<close> explicitly or any other means,
1115	such as C<dup>). The reason is that you register interest in some file	1493	such as C<dup2>). The reason is that you register interest in some file
1116	descriptor, but when it goes away, the operating system will silently drop	1494	descriptor, but when it goes away, the operating system will silently drop
1117	this interest. If another file descriptor with the same number then is	1495	this interest. If another file descriptor with the same number then is
1118	registered with libev, there is no efficient way to see that this is, in	1496	registered with libev, there is no efficient way to see that this is, in
1119	fact, a different file descriptor.	1497	fact, a different file descriptor.
1120		1498
…		…
1151	enable C<EVFLAG_FORKCHECK>, or resort to C<EVBACKEND_SELECT> or	1529	enable C<EVFLAG_FORKCHECK>, or resort to C<EVBACKEND_SELECT> or
1152	C<EVBACKEND_POLL>.	1530	C<EVBACKEND_POLL>.
1153		1531
1154	=head3 The special problem of SIGPIPE	1532	=head3 The special problem of SIGPIPE
1155		1533
1156	While not really specific to libev, it is easy to forget about SIGPIPE:	1534	While not really specific to libev, it is easy to forget about C<SIGPIPE>:
1157	when writing to a pipe whose other end has been closed, your program gets	1535	when writing to a pipe whose other end has been closed, your program gets
1158	send a SIGPIPE, which, by default, aborts your program. For most programs	1536	sent a SIGPIPE, which, by default, aborts your program. For most programs
1159	this is sensible behaviour, for daemons, this is usually undesirable.	1537	this is sensible behaviour, for daemons, this is usually undesirable.
1160		1538
1161	So when you encounter spurious, unexplained daemon exits, make sure you	1539	So when you encounter spurious, unexplained daemon exits, make sure you
1162	ignore SIGPIPE (and maybe make sure you log the exit status of your daemon	1540	ignore SIGPIPE (and maybe make sure you log the exit status of your daemon
1163	somewhere, as that would have given you a big clue).	1541	somewhere, as that would have given you a big clue).
…		…
1170	=item ev_io_init (ev_io *, callback, int fd, int events)	1548	=item ev_io_init (ev_io *, callback, int fd, int events)
1171		1549
1172	=item ev_io_set (ev_io *, int fd, int events)	1550	=item ev_io_set (ev_io *, int fd, int events)
1173		1551
1174	Configures an C<ev_io> watcher. The C<fd> is the file descriptor to	1552	Configures an C<ev_io> watcher. The C<fd> is the file descriptor to
1175	receive events for and events is either C<EV_READ>, C<EV_WRITE> or	1553	receive events for and C<events> is either C<EV_READ>, C<EV_WRITE> or
1176	C<EV_READ | EV_WRITE> to receive the given events.	1554	C<EV_READ | EV_WRITE>, to express the desire to receive the given events.
1177		1555
1178	=item int fd [read-only]	1556	=item int fd [read-only]
1179		1557
1180	The file descriptor being watched.	1558	The file descriptor being watched.
1181		1559
…		…
1190	Example: Call C<stdin_readable_cb> when STDIN_FILENO has become, well	1568	Example: Call C<stdin_readable_cb> when STDIN_FILENO has become, well
1191	readable, but only once. Since it is likely line-buffered, you could	1569	readable, but only once. Since it is likely line-buffered, you could
1192	attempt to read a whole line in the callback.	1570	attempt to read a whole line in the callback.
1193		1571
1194	static void	1572	static void
1195	stdin_readable_cb (struct ev_loop *loop, struct ev_io *w, int revents)	1573	stdin_readable_cb (struct ev_loop *loop, ev_io *w, int revents)
1196	{	1574	{
1197	ev_io_stop (loop, w);	1575	ev_io_stop (loop, w);
1198	.. read from stdin here (or from w->fd) and haqndle any I/O errors	1576	.. read from stdin here (or from w->fd) and handle any I/O errors
1199	}	1577	}
1200		1578
1201	...	1579	...
1202	struct ev_loop *loop = ev_default_init (0);	1580	struct ev_loop *loop = ev_default_init (0);
1203	struct ev_io stdin_readable;	1581	ev_io stdin_readable;
1204	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);	1582	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);
1205	ev_io_start (loop, &stdin_readable);	1583	ev_io_start (loop, &stdin_readable);
1206	ev_loop (loop, 0);	1584	ev_loop (loop, 0);
1207		1585
1208		1586
…		…
1211	Timer watchers are simple relative timers that generate an event after a	1589	Timer watchers are simple relative timers that generate an event after a
1212	given time, and optionally repeating in regular intervals after that.	1590	given time, and optionally repeating in regular intervals after that.
1213		1591
1214	The timers are based on real time, that is, if you register an event that	1592	The timers are based on real time, that is, if you register an event that
1215	times out after an hour and you reset your system clock to January last	1593	times out after an hour and you reset your system clock to January last
1216	year, it will still time out after (roughly) and hour. "Roughly" because	1594	year, it will still time out after (roughly) one hour. "Roughly" because
1217	detecting time jumps is hard, and some inaccuracies are unavoidable (the	1595	detecting time jumps is hard, and some inaccuracies are unavoidable (the
1218	monotonic clock option helps a lot here).	1596	monotonic clock option helps a lot here).
1219		1597
1220	The callback is guaranteed to be invoked only after its timeout has passed,	1598	The callback is guaranteed to be invoked only I<after> its timeout has
1221	but if multiple timers become ready during the same loop iteration then	1599	passed (not I<at>, so on systems with very low-resolution clocks this
1222	order of execution is undefined.	1600	might introduce a small delay). If multiple timers become ready during the
		1601	same loop iteration then the ones with earlier time-out values are invoked
		1602	before ones of the same priority with later time-out values (but this is
		1603	no longer true when a callback calls C<ev_loop> recursively).
		1604
		1605	=head3 Be smart about timeouts
		1606
		1607	Many real-world problems involve some kind of timeout, usually for error
		1608	recovery. A typical example is an HTTP request - if the other side hangs,
		1609	you want to raise some error after a while.
		1610
		1611	What follows are some ways to handle this problem, from obvious and
		1612	inefficient to smart and efficient.
		1613
		1614	In the following, a 60 second activity timeout is assumed - a timeout that
		1615	gets reset to 60 seconds each time there is activity (e.g. each time some
		1616	data or other life sign was received).
		1617
		1618	=over 4
		1619
		1620	=item 1. Use a timer and stop, reinitialise and start it on activity.
		1621
		1622	This is the most obvious, but not the most simple way: In the beginning,
		1623	start the watcher:
		1624
		1625	ev_timer_init (timer, callback, 60., 0.);
		1626	ev_timer_start (loop, timer);
		1627
		1628	Then, each time there is some activity, C<ev_timer_stop> it, initialise it
		1629	and start it again:
		1630
		1631	ev_timer_stop (loop, timer);
		1632	ev_timer_set (timer, 60., 0.);
		1633	ev_timer_start (loop, timer);
		1634
		1635	This is relatively simple to implement, but means that each time there is
		1636	some activity, libev will first have to remove the timer from its internal
		1637	data structure and then add it again. Libev tries to be fast, but it's
		1638	still not a constant-time operation.
		1639
		1640	=item 2. Use a timer and re-start it with C<ev_timer_again> inactivity.
		1641
		1642	This is the easiest way, and involves using C<ev_timer_again> instead of
		1643	C<ev_timer_start>.
		1644
		1645	To implement this, configure an C<ev_timer> with a C<repeat> value
		1646	of C<60> and then call C<ev_timer_again> at start and each time you
		1647	successfully read or write some data. If you go into an idle state where
		1648	you do not expect data to travel on the socket, you can C<ev_timer_stop>
		1649	the timer, and C<ev_timer_again> will automatically restart it if need be.
		1650
		1651	That means you can ignore both the C<ev_timer_start> function and the
		1652	C<after> argument to C<ev_timer_set>, and only ever use the C<repeat>
		1653	member and C<ev_timer_again>.
		1654
		1655	At start:
		1656
		1657	ev_init (timer, callback);
		1658	timer->repeat = 60.;
		1659	ev_timer_again (loop, timer);
		1660
		1661	Each time there is some activity:
		1662
		1663	ev_timer_again (loop, timer);
		1664
		1665	It is even possible to change the time-out on the fly, regardless of
		1666	whether the watcher is active or not:
		1667
		1668	timer->repeat = 30.;
		1669	ev_timer_again (loop, timer);
		1670
		1671	This is slightly more efficient then stopping/starting the timer each time
		1672	you want to modify its timeout value, as libev does not have to completely
		1673	remove and re-insert the timer from/into its internal data structure.
		1674
		1675	It is, however, even simpler than the "obvious" way to do it.
		1676
		1677	=item 3. Let the timer time out, but then re-arm it as required.
		1678
		1679	This method is more tricky, but usually most efficient: Most timeouts are
		1680	relatively long compared to the intervals between other activity - in
		1681	our example, within 60 seconds, there are usually many I/O events with
		1682	associated activity resets.
		1683
		1684	In this case, it would be more efficient to leave the C<ev_timer> alone,
		1685	but remember the time of last activity, and check for a real timeout only
		1686	within the callback:
		1687
		1688	ev_tstamp last_activity; // time of last activity
		1689
		1690	static void
		1691	callback (EV_P_ ev_timer *w, int revents)
		1692	{
		1693	ev_tstamp now = ev_now (EV_A);
		1694	ev_tstamp timeout = last_activity + 60.;
		1695
		1696	// if last_activity + 60. is older than now, we did time out
		1697	if (timeout < now)
		1698	{
		1699	// timeout occured, take action
		1700	}
		1701	else
		1702	{
		1703	// callback was invoked, but there was some activity, re-arm
		1704	// the watcher to fire in last_activity + 60, which is
		1705	// guaranteed to be in the future, so "again" is positive:
		1706	w->repeat = timeout - now;
		1707	ev_timer_again (EV_A_ w);
		1708	}
		1709	}
		1710
		1711	To summarise the callback: first calculate the real timeout (defined
		1712	as "60 seconds after the last activity"), then check if that time has
		1713	been reached, which means something I<did>, in fact, time out. Otherwise
		1714	the callback was invoked too early (C<timeout> is in the future), so
		1715	re-schedule the timer to fire at that future time, to see if maybe we have
		1716	a timeout then.
		1717
		1718	Note how C<ev_timer_again> is used, taking advantage of the
		1719	C<ev_timer_again> optimisation when the timer is already running.
		1720
		1721	This scheme causes more callback invocations (about one every 60 seconds
		1722	minus half the average time between activity), but virtually no calls to
		1723	libev to change the timeout.
		1724
		1725	To start the timer, simply initialise the watcher and set C<last_activity>
		1726	to the current time (meaning we just have some activity :), then call the
		1727	callback, which will "do the right thing" and start the timer:
		1728
		1729	ev_init (timer, callback);
		1730	last_activity = ev_now (loop);
		1731	callback (loop, timer, EV_TIMEOUT);
		1732
		1733	And when there is some activity, simply store the current time in
		1734	C<last_activity>, no libev calls at all:
		1735
		1736	last_actiivty = ev_now (loop);
		1737
		1738	This technique is slightly more complex, but in most cases where the
		1739	time-out is unlikely to be triggered, much more efficient.
		1740
		1741	Changing the timeout is trivial as well (if it isn't hard-coded in the
		1742	callback :) - just change the timeout and invoke the callback, which will
		1743	fix things for you.
		1744
		1745	=item 4. Wee, just use a double-linked list for your timeouts.
		1746
		1747	If there is not one request, but many thousands (millions...), all
		1748	employing some kind of timeout with the same timeout value, then one can
		1749	do even better:
		1750
		1751	When starting the timeout, calculate the timeout value and put the timeout
		1752	at the I<end> of the list.
		1753
		1754	Then use an C<ev_timer> to fire when the timeout at the I<beginning> of
		1755	the list is expected to fire (for example, using the technique #3).
		1756
		1757	When there is some activity, remove the timer from the list, recalculate
		1758	the timeout, append it to the end of the list again, and make sure to
		1759	update the C<ev_timer> if it was taken from the beginning of the list.
		1760
		1761	This way, one can manage an unlimited number of timeouts in O(1) time for
		1762	starting, stopping and updating the timers, at the expense of a major
		1763	complication, and having to use a constant timeout. The constant timeout
		1764	ensures that the list stays sorted.
		1765
		1766	=back
		1767
		1768	So which method the best?
		1769
		1770	Method #2 is a simple no-brain-required solution that is adequate in most
		1771	situations. Method #3 requires a bit more thinking, but handles many cases
		1772	better, and isn't very complicated either. In most case, choosing either
		1773	one is fine, with #3 being better in typical situations.
		1774
		1775	Method #1 is almost always a bad idea, and buys you nothing. Method #4 is
		1776	rather complicated, but extremely efficient, something that really pays
		1777	off after the first million or so of active timers, i.e. it's usually
		1778	overkill :)
1223		1779
1224	=head3 The special problem of time updates	1780	=head3 The special problem of time updates
1225		1781
1226	Establishing the current time is a costly operation (it usually takes at	1782	Establishing the current time is a costly operation (it usually takes at
1227	least two system calls): EV therefore updates its idea of the current	1783	least two system calls): EV therefore updates its idea of the current
1228	time only before and after C<ev_loop> polls for new events, which causes	1784	time only before and after C<ev_loop> collects new events, which causes a
1229	a growing difference between C<ev_now ()> and C<ev_time ()> when handling	1785	growing difference between C<ev_now ()> and C<ev_time ()> when handling
1230	lots of events.	1786	lots of events in one iteration.
1231		1787
1232	The relative timeouts are calculated relative to the C<ev_now ()>	1788	The relative timeouts are calculated relative to the C<ev_now ()>
1233	time. This is usually the right thing as this timestamp refers to the time	1789	time. This is usually the right thing as this timestamp refers to the time
1234	of the event triggering whatever timeout you are modifying/starting. If	1790	of the event triggering whatever timeout you are modifying/starting. If
1235	you suspect event processing to be delayed and you I<need> to base the	1791	you suspect event processing to be delayed and you I<need> to base the
…		…
1239		1795
1240	If the event loop is suspended for a long time, you can also force an	1796	If the event loop is suspended for a long time, you can also force an
1241	update of the time returned by C<ev_now ()> by calling C<ev_now_update	1797	update of the time returned by C<ev_now ()> by calling C<ev_now_update
1242	()>.	1798	()>.
1243		1799
		1800	=head3 The special problems of suspended animation
		1801
		1802	When you leave the server world it is quite customary to hit machines that
		1803	can suspend/hibernate - what happens to the clocks during such a suspend?
		1804
		1805	Some quick tests made with a Linux 2.6.28 indicate that a suspend freezes
		1806	all processes, while the clocks (C<times>, C<CLOCK_MONOTONIC>) continue
		1807	to run until the system is suspended, but they will not advance while the
		1808	system is suspended. That means, on resume, it will be as if the program
		1809	was frozen for a few seconds, but the suspend time will not be counted
		1810	towards C<ev_timer> when a monotonic clock source is used. The real time
		1811	clock advanced as expected, but if it is used as sole clocksource, then a
		1812	long suspend would be detected as a time jump by libev, and timers would
		1813	be adjusted accordingly.
		1814
		1815	I would not be surprised to see different behaviour in different between
		1816	operating systems, OS versions or even different hardware.
		1817
		1818	The other form of suspend (job control, or sending a SIGSTOP) will see a
		1819	time jump in the monotonic clocks and the realtime clock. If the program
		1820	is suspended for a very long time, and monotonic clock sources are in use,
		1821	then you can expect C<ev_timer>s to expire as the full suspension time
		1822	will be counted towards the timers. When no monotonic clock source is in
		1823	use, then libev will again assume a timejump and adjust accordingly.
		1824
		1825	It might be beneficial for this latter case to call C<ev_suspend>
		1826	and C<ev_resume> in code that handles C<SIGTSTP>, to at least get
		1827	deterministic behaviour in this case (you can do nothing against
		1828	C<SIGSTOP>).
		1829
1244	=head3 Watcher-Specific Functions and Data Members	1830	=head3 Watcher-Specific Functions and Data Members
1245		1831
1246	=over 4	1832	=over 4
1247		1833
1248	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)	1834	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)
…		…
1271	If the timer is started but non-repeating, stop it (as if it timed out).	1857	If the timer is started but non-repeating, stop it (as if it timed out).
1272		1858
1273	If the timer is repeating, either start it if necessary (with the	1859	If the timer is repeating, either start it if necessary (with the
1274	C<repeat> value), or reset the running timer to the C<repeat> value.	1860	C<repeat> value), or reset the running timer to the C<repeat> value.
1275		1861
1276	This sounds a bit complicated, but here is a useful and typical	1862	This sounds a bit complicated, see L<Be smart about timeouts>, above, for a
1277	example: Imagine you have a TCP connection and you want a so-called idle	1863	usage example.
1278	timeout, that is, you want to be called when there have been, say, 60
1279	seconds of inactivity on the socket. The easiest way to do this is to
1280	configure an C<ev_timer> with a C<repeat> value of C<60> and then call
1281	C<ev_timer_again> each time you successfully read or write some data. If
1282	you go into an idle state where you do not expect data to travel on the
1283	socket, you can C<ev_timer_stop> the timer, and C<ev_timer_again> will
1284	automatically restart it if need be.
1285		1864
1286	That means you can ignore the C<after> value and C<ev_timer_start>	1865	=item ev_tstamp ev_timer_remaining (loop, ev_timer *)
1287	altogether and only ever use the C<repeat> value and C<ev_timer_again>:
1288		1866
1289	ev_timer_init (timer, callback, 0., 5.);	1867	Returns the remaining time until a timer fires. If the timer is active,
1290	ev_timer_again (loop, timer);	1868	then this time is relative to the current event loop time, otherwise it's
1291	...	1869	the timeout value currently configured.
1292	timer->again = 17.;
1293	ev_timer_again (loop, timer);
1294	...
1295	timer->again = 10.;
1296	ev_timer_again (loop, timer);
1297		1870
1298	This is more slightly efficient then stopping/starting the timer each time	1871	That is, after an C<ev_timer_set (w, 5, 7)>, C<ev_timer_remaining> returns
1299	you want to modify its timeout value.	1872	C<5>. When the timer is started and one second passes, C<ev_timer_remaining>
		1873	will return C<4>. When the timer expires and is restarted, it will return
		1874	roughly C<7> (likely slightly less as callback invocation takes some time,
		1875	too), and so on.
1300		1876
1301	=item ev_tstamp repeat [read-write]	1877	=item ev_tstamp repeat [read-write]
1302		1878
1303	The current C<repeat> value. Will be used each time the watcher times out	1879	The current C<repeat> value. Will be used each time the watcher times out
1304	or C<ev_timer_again> is called and determines the next timeout (if any),	1880	or C<ev_timer_again> is called, and determines the next timeout (if any),
1305	which is also when any modifications are taken into account.	1881	which is also when any modifications are taken into account.
1306		1882
1307	=back	1883	=back
1308		1884
1309	=head3 Examples	1885	=head3 Examples
1310		1886
1311	Example: Create a timer that fires after 60 seconds.	1887	Example: Create a timer that fires after 60 seconds.
1312		1888
1313	static void	1889	static void
1314	one_minute_cb (struct ev_loop *loop, struct ev_timer *w, int revents)	1890	one_minute_cb (struct ev_loop *loop, ev_timer *w, int revents)
1315	{	1891	{
1316	.. one minute over, w is actually stopped right here	1892	.. one minute over, w is actually stopped right here
1317	}	1893	}
1318		1894
1319	struct ev_timer mytimer;	1895	ev_timer mytimer;
1320	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);	1896	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);
1321	ev_timer_start (loop, &mytimer);	1897	ev_timer_start (loop, &mytimer);
1322		1898
1323	Example: Create a timeout timer that times out after 10 seconds of	1899	Example: Create a timeout timer that times out after 10 seconds of
1324	inactivity.	1900	inactivity.
1325		1901
1326	static void	1902	static void
1327	timeout_cb (struct ev_loop *loop, struct ev_timer *w, int revents)	1903	timeout_cb (struct ev_loop *loop, ev_timer *w, int revents)
1328	{	1904	{
1329	.. ten seconds without any activity	1905	.. ten seconds without any activity
1330	}	1906	}
1331		1907
1332	struct ev_timer mytimer;	1908	ev_timer mytimer;
1333	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */	1909	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */
1334	ev_timer_again (&mytimer); /* start timer */	1910	ev_timer_again (&mytimer); /* start timer */
1335	ev_loop (loop, 0);	1911	ev_loop (loop, 0);
1336		1912
1337	// and in some piece of code that gets executed on any "activity":	1913	// and in some piece of code that gets executed on any "activity":
…		…
1342	=head2 C<ev_periodic> - to cron or not to cron?	1918	=head2 C<ev_periodic> - to cron or not to cron?
1343		1919
1344	Periodic watchers are also timers of a kind, but they are very versatile	1920	Periodic watchers are also timers of a kind, but they are very versatile
1345	(and unfortunately a bit complex).	1921	(and unfortunately a bit complex).
1346		1922
1347	Unlike C<ev_timer>'s, they are not based on real time (or relative time)	1923	Unlike C<ev_timer>, periodic watchers are not based on real time (or
1348	but on wall clock time (absolute time). You can tell a periodic watcher	1924	relative time, the physical time that passes) but on wall clock time
1349	to trigger after some specific point in time. For example, if you tell a	1925	(absolute time, the thing you can read on your calender or clock). The
1350	periodic watcher to trigger in 10 seconds (by specifying e.g. C<ev_now ()	1926	difference is that wall clock time can run faster or slower than real
1351	+ 10.>, that is, an absolute time not a delay) and then reset your system	1927	time, and time jumps are not uncommon (e.g. when you adjust your
1352	clock to January of the previous year, then it will take more than year	1928	wrist-watch).
1353	to trigger the event (unlike an C<ev_timer>, which would still trigger
1354	roughly 10 seconds later as it uses a relative timeout).
1355		1929
		1930	You can tell a periodic watcher to trigger after some specific point
		1931	in time: for example, if you tell a periodic watcher to trigger "in 10
		1932	seconds" (by specifying e.g. C<ev_now () + 10.>, that is, an absolute time
		1933	not a delay) and then reset your system clock to January of the previous
		1934	year, then it will take a year or more to trigger the event (unlike an
		1935	C<ev_timer>, which would still trigger roughly 10 seconds after starting
		1936	it, as it uses a relative timeout).
		1937
1356	C<ev_periodic>s can also be used to implement vastly more complex timers,	1938	C<ev_periodic> watchers can also be used to implement vastly more complex
1357	such as triggering an event on each "midnight, local time", or other	1939	timers, such as triggering an event on each "midnight, local time", or
1358	complicated, rules.	1940	other complicated rules. This cannot be done with C<ev_timer> watchers, as
		1941	those cannot react to time jumps.
1359		1942
1360	As with timers, the callback is guaranteed to be invoked only when the	1943	As with timers, the callback is guaranteed to be invoked only when the
1361	time (C<at>) has passed, but if multiple periodic timers become ready	1944	point in time where it is supposed to trigger has passed. If multiple
1362	during the same loop iteration then order of execution is undefined.	1945	timers become ready during the same loop iteration then the ones with
		1946	earlier time-out values are invoked before ones with later time-out values
		1947	(but this is no longer true when a callback calls C<ev_loop> recursively).
1363		1948
1364	=head3 Watcher-Specific Functions and Data Members	1949	=head3 Watcher-Specific Functions and Data Members
1365		1950
1366	=over 4	1951	=over 4
1367		1952
1368	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)	1953	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp offset, ev_tstamp interval, reschedule_cb)
1369		1954
1370	=item ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)	1955	=item ev_periodic_set (ev_periodic *, ev_tstamp offset, ev_tstamp interval, reschedule_cb)
1371		1956
1372	Lots of arguments, lets sort it out... There are basically three modes of	1957	Lots of arguments, let's sort it out... There are basically three modes of
1373	operation, and we will explain them from simplest to complex:	1958	operation, and we will explain them from simplest to most complex:
1374		1959
1375	=over 4	1960	=over 4
1376		1961
1377	=item * absolute timer (at = time, interval = reschedule_cb = 0)	1962	=item * absolute timer (offset = absolute time, interval = 0, reschedule_cb = 0)
1378		1963
1379	In this configuration the watcher triggers an event after the wall clock	1964	In this configuration the watcher triggers an event after the wall clock
1380	time C<at> has passed and doesn't repeat. It will not adjust when a time	1965	time C<offset> has passed. It will not repeat and will not adjust when a
1381	jump occurs, that is, if it is to be run at January 1st 2011 then it will	1966	time jump occurs, that is, if it is to be run at January 1st 2011 then it
1382	run when the system time reaches or surpasses this time.	1967	will be stopped and invoked when the system clock reaches or surpasses
		1968	this point in time.
1383		1969
1384	=item * repeating interval timer (at = offset, interval > 0, reschedule_cb = 0)	1970	=item * repeating interval timer (offset = offset within interval, interval > 0, reschedule_cb = 0)
1385		1971
1386	In this mode the watcher will always be scheduled to time out at the next	1972	In this mode the watcher will always be scheduled to time out at the next
1387	C<at + N * interval> time (for some integer N, which can also be negative)	1973	C<offset + N * interval> time (for some integer N, which can also be
1388	and then repeat, regardless of any time jumps.	1974	negative) and then repeat, regardless of any time jumps. The C<offset>
		1975	argument is merely an offset into the C<interval> periods.
1389		1976
1390	This can be used to create timers that do not drift with respect to system	1977	This can be used to create timers that do not drift with respect to the
1391	time, for example, here is a C<ev_periodic> that triggers each hour, on	1978	system clock, for example, here is an C<ev_periodic> that triggers each
1392	the hour:	1979	hour, on the hour (with respect to UTC):
1393		1980
1394	ev_periodic_set (&periodic, 0., 3600., 0);	1981	ev_periodic_set (&periodic, 0., 3600., 0);
1395		1982
1396	This doesn't mean there will always be 3600 seconds in between triggers,	1983	This doesn't mean there will always be 3600 seconds in between triggers,
1397	but only that the callback will be called when the system time shows a	1984	but only that the callback will be called when the system time shows a
1398	full hour (UTC), or more correctly, when the system time is evenly divisible	1985	full hour (UTC), or more correctly, when the system time is evenly divisible
1399	by 3600.	1986	by 3600.
1400		1987
1401	Another way to think about it (for the mathematically inclined) is that	1988	Another way to think about it (for the mathematically inclined) is that
1402	C<ev_periodic> will try to run the callback in this mode at the next possible	1989	C<ev_periodic> will try to run the callback in this mode at the next possible
1403	time where C<time = at (mod interval)>, regardless of any time jumps.	1990	time where C<time = offset (mod interval)>, regardless of any time jumps.
1404		1991
1405	For numerical stability it is preferable that the C<at> value is near	1992	For numerical stability it is preferable that the C<offset> value is near
1406	C<ev_now ()> (the current time), but there is no range requirement for	1993	C<ev_now ()> (the current time), but there is no range requirement for
1407	this value, and in fact is often specified as zero.	1994	this value, and in fact is often specified as zero.
1408		1995
1409	Note also that there is an upper limit to how often a timer can fire (CPU	1996	Note also that there is an upper limit to how often a timer can fire (CPU
1410	speed for example), so if C<interval> is very small then timing stability	1997	speed for example), so if C<interval> is very small then timing stability
1411	will of course deteriorate. Libev itself tries to be exact to be about one	1998	will of course deteriorate. Libev itself tries to be exact to be about one
1412	millisecond (if the OS supports it and the machine is fast enough).	1999	millisecond (if the OS supports it and the machine is fast enough).
1413		2000
1414	=item * manual reschedule mode (at and interval ignored, reschedule_cb = callback)	2001	=item * manual reschedule mode (offset ignored, interval ignored, reschedule_cb = callback)
1415		2002
1416	In this mode the values for C<interval> and C<at> are both being	2003	In this mode the values for C<interval> and C<offset> are both being
1417	ignored. Instead, each time the periodic watcher gets scheduled, the	2004	ignored. Instead, each time the periodic watcher gets scheduled, the
1418	reschedule callback will be called with the watcher as first, and the	2005	reschedule callback will be called with the watcher as first, and the
1419	current time as second argument.	2006	current time as second argument.
1420		2007
1421	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher,	2008	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher, ever,
1422	ever, or make ANY event loop modifications whatsoever>.	2009	or make ANY other event loop modifications whatsoever, unless explicitly
		2010	allowed by documentation here>.
1423		2011
1424	If you need to stop it, return C<now + 1e30> (or so, fudge fudge) and stop	2012	If you need to stop it, return C<now + 1e30> (or so, fudge fudge) and stop
1425	it afterwards (e.g. by starting an C<ev_prepare> watcher, which is the	2013	it afterwards (e.g. by starting an C<ev_prepare> watcher, which is the
1426	only event loop modification you are allowed to do).	2014	only event loop modification you are allowed to do).
1427		2015
1428	The callback prototype is C<ev_tstamp (*reschedule_cb)(struct ev_periodic	2016	The callback prototype is C<ev_tstamp (*reschedule_cb)(ev_periodic
1429	*w, ev_tstamp now)>, e.g.:	2017	*w, ev_tstamp now)>, e.g.:
1430		2018
		2019	static ev_tstamp
1431	static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now)	2020	my_rescheduler (ev_periodic *w, ev_tstamp now)
1432	{	2021	{
1433	return now + 60.;	2022	return now + 60.;
1434	}	2023	}
1435		2024
1436	It must return the next time to trigger, based on the passed time value	2025	It must return the next time to trigger, based on the passed time value
…		…
1456	a different time than the last time it was called (e.g. in a crond like	2045	a different time than the last time it was called (e.g. in a crond like
1457	program when the crontabs have changed).	2046	program when the crontabs have changed).
1458		2047
1459	=item ev_tstamp ev_periodic_at (ev_periodic *)	2048	=item ev_tstamp ev_periodic_at (ev_periodic *)
1460		2049
1461	When active, returns the absolute time that the watcher is supposed to	2050	When active, returns the absolute time that the watcher is supposed
1462	trigger next.	2051	to trigger next. This is not the same as the C<offset> argument to
		2052	C<ev_periodic_set>, but indeed works even in interval and manual
		2053	rescheduling modes.
1463		2054
1464	=item ev_tstamp offset [read-write]	2055	=item ev_tstamp offset [read-write]
1465		2056
1466	When repeating, this contains the offset value, otherwise this is the	2057	When repeating, this contains the offset value, otherwise this is the
1467	absolute point in time (the C<at> value passed to C<ev_periodic_set>).	2058	absolute point in time (the C<offset> value passed to C<ev_periodic_set>,
		2059	although libev might modify this value for better numerical stability).
1468		2060
1469	Can be modified any time, but changes only take effect when the periodic	2061	Can be modified any time, but changes only take effect when the periodic
1470	timer fires or C<ev_periodic_again> is being called.	2062	timer fires or C<ev_periodic_again> is being called.
1471		2063
1472	=item ev_tstamp interval [read-write]	2064	=item ev_tstamp interval [read-write]
1473		2065
1474	The current interval value. Can be modified any time, but changes only	2066	The current interval value. Can be modified any time, but changes only
1475	take effect when the periodic timer fires or C<ev_periodic_again> is being	2067	take effect when the periodic timer fires or C<ev_periodic_again> is being
1476	called.	2068	called.
1477		2069
1478	=item ev_tstamp (*reschedule_cb)(struct ev_periodic *w, ev_tstamp now) [read-write]	2070	=item ev_tstamp (*reschedule_cb)(ev_periodic *w, ev_tstamp now) [read-write]
1479		2071
1480	The current reschedule callback, or C<0>, if this functionality is	2072	The current reschedule callback, or C<0>, if this functionality is
1481	switched off. Can be changed any time, but changes only take effect when	2073	switched off. Can be changed any time, but changes only take effect when
1482	the periodic timer fires or C<ev_periodic_again> is being called.	2074	the periodic timer fires or C<ev_periodic_again> is being called.
1483		2075
1484	=back	2076	=back
1485		2077
1486	=head3 Examples	2078	=head3 Examples
1487		2079
1488	Example: Call a callback every hour, or, more precisely, whenever the	2080	Example: Call a callback every hour, or, more precisely, whenever the
1489	system clock is divisible by 3600. The callback invocation times have	2081	system time is divisible by 3600. The callback invocation times have
1490	potentially a lot of jitter, but good long-term stability.	2082	potentially a lot of jitter, but good long-term stability.
1491		2083
1492	static void	2084	static void
1493	clock_cb (struct ev_loop *loop, struct ev_io *w, int revents)	2085	clock_cb (struct ev_loop *loop, ev_io *w, int revents)
1494	{	2086	{
1495	... its now a full hour (UTC, or TAI or whatever your clock follows)	2087	... its now a full hour (UTC, or TAI or whatever your clock follows)
1496	}	2088	}
1497		2089
1498	struct ev_periodic hourly_tick;	2090	ev_periodic hourly_tick;
1499	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);	2091	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);
1500	ev_periodic_start (loop, &hourly_tick);	2092	ev_periodic_start (loop, &hourly_tick);
1501		2093
1502	Example: The same as above, but use a reschedule callback to do it:	2094	Example: The same as above, but use a reschedule callback to do it:
1503		2095
1504	#include <math.h>	2096	#include <math.h>
1505		2097
1506	static ev_tstamp	2098	static ev_tstamp
1507	my_scheduler_cb (struct ev_periodic *w, ev_tstamp now)	2099	my_scheduler_cb (ev_periodic *w, ev_tstamp now)
1508	{	2100	{
1509	return fmod (now, 3600.) + 3600.;	2101	return now + (3600. - fmod (now, 3600.));
1510	}	2102	}
1511		2103
1512	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);	2104	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);
1513		2105
1514	Example: Call a callback every hour, starting now:	2106	Example: Call a callback every hour, starting now:
1515		2107
1516	struct ev_periodic hourly_tick;	2108	ev_periodic hourly_tick;
1517	ev_periodic_init (&hourly_tick, clock_cb,	2109	ev_periodic_init (&hourly_tick, clock_cb,
1518	fmod (ev_now (loop), 3600.), 3600., 0);	2110	fmod (ev_now (loop), 3600.), 3600., 0);
1519	ev_periodic_start (loop, &hourly_tick);	2111	ev_periodic_start (loop, &hourly_tick);
1520		2112
1521		2113
…		…
1524	Signal watchers will trigger an event when the process receives a specific	2116	Signal watchers will trigger an event when the process receives a specific
1525	signal one or more times. Even though signals are very asynchronous, libev	2117	signal one or more times. Even though signals are very asynchronous, libev
1526	will try it's best to deliver signals synchronously, i.e. as part of the	2118	will try it's best to deliver signals synchronously, i.e. as part of the
1527	normal event processing, like any other event.	2119	normal event processing, like any other event.
1528		2120
		2121	If you want signals to be delivered truly asynchronously, just use
		2122	C<sigaction> as you would do without libev and forget about sharing
		2123	the signal. You can even use C<ev_async> from a signal handler to
		2124	synchronously wake up an event loop.
		2125
1529	You can configure as many watchers as you like per signal. Only when the	2126	You can configure as many watchers as you like for the same signal, but
		2127	only within the same loop, i.e. you can watch for C<SIGINT> in your
		2128	default loop and for C<SIGIO> in another loop, but you cannot watch for
		2129	C<SIGINT> in both the default loop and another loop at the same time. At
		2130	the moment, C<SIGCHLD> is permanently tied to the default loop.
		2131
1530	first watcher gets started will libev actually register a signal watcher	2132	When the first watcher gets started will libev actually register something
1531	with the kernel (thus it coexists with your own signal handlers as long	2133	with the kernel (thus it coexists with your own signal handlers as long as
1532	as you don't register any with libev). Similarly, when the last signal	2134	you don't register any with libev for the same signal).
1533	watcher for a signal is stopped libev will reset the signal handler to
1534	SIG_DFL (regardless of what it was set to before).
1535		2135
1536	If possible and supported, libev will install its handlers with	2136	If possible and supported, libev will install its handlers with
1537	C<SA_RESTART> behaviour enabled, so system calls should not be unduly	2137	C<SA_RESTART> (or equivalent) behaviour enabled, so system calls should
1538	interrupted. If you have a problem with system calls getting interrupted by	2138	not be unduly interrupted. If you have a problem with system calls getting
1539	signals you can block all signals in an C<ev_check> watcher and unblock	2139	interrupted by signals you can block all signals in an C<ev_check> watcher
1540	them in an C<ev_prepare> watcher.	2140	and unblock them in an C<ev_prepare> watcher.
		2141
		2142	=head3 The special problem of inheritance over fork/execve/pthread_create
		2143
		2144	Both the signal mask (C<sigprocmask>) and the signal disposition
		2145	(C<sigaction>) are unspecified after starting a signal watcher (and after
		2146	stopping it again), that is, libev might or might not block the signal,
		2147	and might or might not set or restore the installed signal handler.
		2148
		2149	While this does not matter for the signal disposition (libev never
		2150	sets signals to C<SIG_IGN>, so handlers will be reset to C<SIG_DFL> on
		2151	C<execve>), this matters for the signal mask: many programs do not expect
		2152	certain signals to be blocked.
		2153
		2154	This means that before calling C<exec> (from the child) you should reset
		2155	the signal mask to whatever "default" you expect (all clear is a good
		2156	choice usually).
		2157
		2158	The simplest way to ensure that the signal mask is reset in the child is
		2159	to install a fork handler with C<pthread_atfork> that resets it. That will
		2160	catch fork calls done by libraries (such as the libc) as well.
		2161
		2162	In current versions of libev, the signal will not be blocked indefinitely
		2163	unless you use the C<signalfd> API (C<EV_SIGNALFD>). While this reduces
		2164	the window of opportunity for problems, it will not go away, as libev
		2165	I<has> to modify the signal mask, at least temporarily.
		2166
		2167	So I can't stress this enough: I<If you do not reset your signal mask when
		2168	you expect it to be empty, you have a race condition in your code>. This
		2169	is not a libev-specific thing, this is true for most event libraries.
1541		2170
1542	=head3 Watcher-Specific Functions and Data Members	2171	=head3 Watcher-Specific Functions and Data Members
1543		2172
1544	=over 4	2173	=over 4
1545		2174
…		…
1556		2185
1557	=back	2186	=back
1558		2187
1559	=head3 Examples	2188	=head3 Examples
1560		2189
1561	Example: Try to exit cleanly on SIGINT and SIGTERM.	2190	Example: Try to exit cleanly on SIGINT.
1562		2191
1563	static void	2192	static void
1564	sigint_cb (struct ev_loop *loop, struct ev_signal *w, int revents)	2193	sigint_cb (struct ev_loop *loop, ev_signal *w, int revents)
1565	{	2194	{
1566	ev_unloop (loop, EVUNLOOP_ALL);	2195	ev_unloop (loop, EVUNLOOP_ALL);
1567	}	2196	}
1568		2197
1569	struct ev_signal signal_watcher;	2198	ev_signal signal_watcher;
1570	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);	2199	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);
1571	ev_signal_start (loop, &sigint_cb);	2200	ev_signal_start (loop, &signal_watcher);
1572		2201
1573		2202
1574	=head2 C<ev_child> - watch out for process status changes	2203	=head2 C<ev_child> - watch out for process status changes
1575		2204
1576	Child watchers trigger when your process receives a SIGCHLD in response to	2205	Child watchers trigger when your process receives a SIGCHLD in response to
1577	some child status changes (most typically when a child of yours dies). It	2206	some child status changes (most typically when a child of yours dies or
1578	is permissible to install a child watcher I<after> the child has been	2207	exits). It is permissible to install a child watcher I<after> the child
1579	forked (which implies it might have already exited), as long as the event	2208	has been forked (which implies it might have already exited), as long
1580	loop isn't entered (or is continued from a watcher).	2209	as the event loop isn't entered (or is continued from a watcher), i.e.,
		2210	forking and then immediately registering a watcher for the child is fine,
		2211	but forking and registering a watcher a few event loop iterations later or
		2212	in the next callback invocation is not.
1581		2213
1582	Only the default event loop is capable of handling signals, and therefore	2214	Only the default event loop is capable of handling signals, and therefore
1583	you can only register child watchers in the default event loop.	2215	you can only register child watchers in the default event loop.
1584		2216
		2217	Due to some design glitches inside libev, child watchers will always be
		2218	handled at maximum priority (their priority is set to C<EV_MAXPRI> by
		2219	libev)
		2220
1585	=head3 Process Interaction	2221	=head3 Process Interaction
1586		2222
1587	Libev grabs C<SIGCHLD> as soon as the default event loop is	2223	Libev grabs C<SIGCHLD> as soon as the default event loop is
1588	initialised. This is necessary to guarantee proper behaviour even if	2224	initialised. This is necessary to guarantee proper behaviour even if the
1589	the first child watcher is started after the child exits. The occurrence	2225	first child watcher is started after the child exits. The occurrence
1590	of C<SIGCHLD> is recorded asynchronously, but child reaping is done	2226	of C<SIGCHLD> is recorded asynchronously, but child reaping is done
1591	synchronously as part of the event loop processing. Libev always reaps all	2227	synchronously as part of the event loop processing. Libev always reaps all
1592	children, even ones not watched.	2228	children, even ones not watched.
1593		2229
1594	=head3 Overriding the Built-In Processing	2230	=head3 Overriding the Built-In Processing
…		…
1604	=head3 Stopping the Child Watcher	2240	=head3 Stopping the Child Watcher
1605		2241
1606	Currently, the child watcher never gets stopped, even when the	2242	Currently, the child watcher never gets stopped, even when the
1607	child terminates, so normally one needs to stop the watcher in the	2243	child terminates, so normally one needs to stop the watcher in the
1608	callback. Future versions of libev might stop the watcher automatically	2244	callback. Future versions of libev might stop the watcher automatically
1609	when a child exit is detected.	2245	when a child exit is detected (calling C<ev_child_stop> twice is not a
		2246	problem).
1610		2247
1611	=head3 Watcher-Specific Functions and Data Members	2248	=head3 Watcher-Specific Functions and Data Members
1612		2249
1613	=over 4	2250	=over 4
1614		2251
…		…
1646	its completion.	2283	its completion.
1647		2284
1648	ev_child cw;	2285	ev_child cw;
1649		2286
1650	static void	2287	static void
1651	child_cb (EV_P_ struct ev_child *w, int revents)	2288	child_cb (EV_P_ ev_child *w, int revents)
1652	{	2289	{
1653	ev_child_stop (EV_A_ w);	2290	ev_child_stop (EV_A_ w);
1654	printf ("process %d exited with status %x\n", w->rpid, w->rstatus);	2291	printf ("process %d exited with status %x\n", w->rpid, w->rstatus);
1655	}	2292	}
1656		2293
…		…
1671		2308
1672		2309
1673	=head2 C<ev_stat> - did the file attributes just change?	2310	=head2 C<ev_stat> - did the file attributes just change?
1674		2311
1675	This watches a file system path for attribute changes. That is, it calls	2312	This watches a file system path for attribute changes. That is, it calls
1676	C<stat> regularly (or when the OS says it changed) and sees if it changed	2313	C<stat> on that path in regular intervals (or when the OS says it changed)
1677	compared to the last time, invoking the callback if it did.	2314	and sees if it changed compared to the last time, invoking the callback if
		2315	it did.
1678		2316
1679	The path does not need to exist: changing from "path exists" to "path does	2317	The path does not need to exist: changing from "path exists" to "path does
1680	not exist" is a status change like any other. The condition "path does	2318	not exist" is a status change like any other. The condition "path does not
1681	not exist" is signified by the C<st_nlink> field being zero (which is	2319	exist" (or more correctly "path cannot be stat'ed") is signified by the
1682	otherwise always forced to be at least one) and all the other fields of	2320	C<st_nlink> field being zero (which is otherwise always forced to be at
1683	the stat buffer having unspecified contents.	2321	least one) and all the other fields of the stat buffer having unspecified
		2322	contents.
1684		2323
1685	The path I<should> be absolute and I<must not> end in a slash. If it is	2324	The path I<must not> end in a slash or contain special components such as
		2325	C<.> or C<..>. The path I<should> be absolute: If it is relative and
1686	relative and your working directory changes, the behaviour is undefined.	2326	your working directory changes, then the behaviour is undefined.
1687		2327
1688	Since there is no standard to do this, the portable implementation simply	2328	Since there is no portable change notification interface available, the
1689	calls C<stat (2)> regularly on the path to see if it changed somehow. You	2329	portable implementation simply calls C<stat(2)> regularly on the path
1690	can specify a recommended polling interval for this case. If you specify	2330	to see if it changed somehow. You can specify a recommended polling
1691	a polling interval of C<0> (highly recommended!) then a I<suitable,	2331	interval for this case. If you specify a polling interval of C<0> (highly
1692	unspecified default> value will be used (which you can expect to be around	2332	recommended!) then a I<suitable, unspecified default> value will be used
1693	five seconds, although this might change dynamically). Libev will also	2333	(which you can expect to be around five seconds, although this might
1694	impose a minimum interval which is currently around C<0.1>, but thats	2334	change dynamically). Libev will also impose a minimum interval which is
1695	usually overkill.	2335	currently around C<0.1>, but that's usually overkill.
1696		2336
1697	This watcher type is not meant for massive numbers of stat watchers,	2337	This watcher type is not meant for massive numbers of stat watchers,
1698	as even with OS-supported change notifications, this can be	2338	as even with OS-supported change notifications, this can be
1699	resource-intensive.	2339	resource-intensive.
1700		2340
1701	At the time of this writing, only the Linux inotify interface is	2341	At the time of this writing, the only OS-specific interface implemented
1702	implemented (implementing kqueue support is left as an exercise for the	2342	is the Linux inotify interface (implementing kqueue support is left as an
1703	reader, note, however, that the author sees no way of implementing ev_stat	2343	exercise for the reader. Note, however, that the author sees no way of
1704	semantics with kqueue). Inotify will be used to give hints only and should	2344	implementing C<ev_stat> semantics with kqueue, except as a hint).
1705	not change the semantics of C<ev_stat> watchers, which means that libev
1706	sometimes needs to fall back to regular polling again even with inotify,
1707	but changes are usually detected immediately, and if the file exists there
1708	will be no polling.
1709		2345
1710	=head3 ABI Issues (Largefile Support)	2346	=head3 ABI Issues (Largefile Support)
1711		2347
1712	Libev by default (unless the user overrides this) uses the default	2348	Libev by default (unless the user overrides this) uses the default
1713	compilation environment, which means that on systems with large file	2349	compilation environment, which means that on systems with large file
1714	support disabled by default, you get the 32 bit version of the stat	2350	support disabled by default, you get the 32 bit version of the stat
1715	structure. When using the library from programs that change the ABI to	2351	structure. When using the library from programs that change the ABI to
1716	use 64 bit file offsets the programs will fail. In that case you have to	2352	use 64 bit file offsets the programs will fail. In that case you have to
1717	compile libev with the same flags to get binary compatibility. This is	2353	compile libev with the same flags to get binary compatibility. This is
1718	obviously the case with any flags that change the ABI, but the problem is	2354	obviously the case with any flags that change the ABI, but the problem is
1719	most noticeably disabled with ev_stat and large file support.	2355	most noticeably displayed with ev_stat and large file support.
1720		2356
1721	The solution for this is to lobby your distribution maker to make large	2357	The solution for this is to lobby your distribution maker to make large
1722	file interfaces available by default (as e.g. FreeBSD does) and not	2358	file interfaces available by default (as e.g. FreeBSD does) and not
1723	optional. Libev cannot simply switch on large file support because it has	2359	optional. Libev cannot simply switch on large file support because it has
1724	to exchange stat structures with application programs compiled using the	2360	to exchange stat structures with application programs compiled using the
1725	default compilation environment.	2361	default compilation environment.
1726		2362
1727	=head3 Inotify	2363	=head3 Inotify and Kqueue
1728		2364
1729	When C<inotify (7)> support has been compiled into libev (generally only	2365	When C<inotify (7)> support has been compiled into libev and present at
1730	available on Linux) and present at runtime, it will be used to speed up	2366	runtime, it will be used to speed up change detection where possible. The
1731	change detection where possible. The inotify descriptor will be created lazily	2367	inotify descriptor will be created lazily when the first C<ev_stat>
1732	when the first C<ev_stat> watcher is being started.	2368	watcher is being started.
1733		2369
1734	Inotify presence does not change the semantics of C<ev_stat> watchers	2370	Inotify presence does not change the semantics of C<ev_stat> watchers
1735	except that changes might be detected earlier, and in some cases, to avoid	2371	except that changes might be detected earlier, and in some cases, to avoid
1736	making regular C<stat> calls. Even in the presence of inotify support	2372	making regular C<stat> calls. Even in the presence of inotify support
1737	there are many cases where libev has to resort to regular C<stat> polling.	2373	there are many cases where libev has to resort to regular C<stat> polling,
		2374	but as long as kernel 2.6.25 or newer is used (2.6.24 and older have too
		2375	many bugs), the path exists (i.e. stat succeeds), and the path resides on
		2376	a local filesystem (libev currently assumes only ext2/3, jfs, reiserfs and
		2377	xfs are fully working) libev usually gets away without polling.
1738		2378
1739	(There is no support for kqueue, as apparently it cannot be used to	2379	There is no support for kqueue, as apparently it cannot be used to
1740	implement this functionality, due to the requirement of having a file	2380	implement this functionality, due to the requirement of having a file
1741	descriptor open on the object at all times).	2381	descriptor open on the object at all times, and detecting renames, unlinks
		2382	etc. is difficult.
		2383
		2384	=head3 C<stat ()> is a synchronous operation
		2385
		2386	Libev doesn't normally do any kind of I/O itself, and so is not blocking
		2387	the process. The exception are C<ev_stat> watchers - those call C<stat
		2388	()>, which is a synchronous operation.
		2389
		2390	For local paths, this usually doesn't matter: unless the system is very
		2391	busy or the intervals between stat's are large, a stat call will be fast,
		2392	as the path data is usually in memory already (except when starting the
		2393	watcher).
		2394
		2395	For networked file systems, calling C<stat ()> can block an indefinite
		2396	time due to network issues, and even under good conditions, a stat call
		2397	often takes multiple milliseconds.
		2398
		2399	Therefore, it is best to avoid using C<ev_stat> watchers on networked
		2400	paths, although this is fully supported by libev.
1742		2401
1743	=head3 The special problem of stat time resolution	2402	=head3 The special problem of stat time resolution
1744		2403
1745	The C<stat ()> system call only supports full-second resolution portably, and	2404	The C<stat ()> system call only supports full-second resolution portably,
1746	even on systems where the resolution is higher, many file systems still	2405	and even on systems where the resolution is higher, most file systems
1747	only support whole seconds.	2406	still only support whole seconds.
1748		2407
1749	That means that, if the time is the only thing that changes, you can	2408	That means that, if the time is the only thing that changes, you can
1750	easily miss updates: on the first update, C<ev_stat> detects a change and	2409	easily miss updates: on the first update, C<ev_stat> detects a change and
1751	calls your callback, which does something. When there is another update	2410	calls your callback, which does something. When there is another update
1752	within the same second, C<ev_stat> will be unable to detect it as the stat	2411	within the same second, C<ev_stat> will be unable to detect unless the
1753	data does not change.	2412	stat data does change in other ways (e.g. file size).
1754		2413
1755	The solution to this is to delay acting on a change for slightly more	2414	The solution to this is to delay acting on a change for slightly more
1756	than a second (or till slightly after the next full second boundary), using	2415	than a second (or till slightly after the next full second boundary), using
1757	a roughly one-second-delay C<ev_timer> (e.g. C<ev_timer_set (w, 0., 1.02);	2416	a roughly one-second-delay C<ev_timer> (e.g. C<ev_timer_set (w, 0., 1.02);
1758	ev_timer_again (loop, w)>).	2417	ev_timer_again (loop, w)>).
…		…
1778	C<path>. The C<interval> is a hint on how quickly a change is expected to	2437	C<path>. The C<interval> is a hint on how quickly a change is expected to
1779	be detected and should normally be specified as C<0> to let libev choose	2438	be detected and should normally be specified as C<0> to let libev choose
1780	a suitable value. The memory pointed to by C<path> must point to the same	2439	a suitable value. The memory pointed to by C<path> must point to the same
1781	path for as long as the watcher is active.	2440	path for as long as the watcher is active.
1782		2441
1783	The callback will receive C<EV_STAT> when a change was detected, relative	2442	The callback will receive an C<EV_STAT> event when a change was detected,
1784	to the attributes at the time the watcher was started (or the last change	2443	relative to the attributes at the time the watcher was started (or the
1785	was detected).	2444	last change was detected).
1786		2445
1787	=item ev_stat_stat (loop, ev_stat *)	2446	=item ev_stat_stat (loop, ev_stat *)
1788		2447
1789	Updates the stat buffer immediately with new values. If you change the	2448	Updates the stat buffer immediately with new values. If you change the
1790	watched path in your callback, you could call this function to avoid	2449	watched path in your callback, you could call this function to avoid
…		…
1873		2532
1874		2533
1875	=head2 C<ev_idle> - when you've got nothing better to do...	2534	=head2 C<ev_idle> - when you've got nothing better to do...
1876		2535
1877	Idle watchers trigger events when no other events of the same or higher	2536	Idle watchers trigger events when no other events of the same or higher
1878	priority are pending (prepare, check and other idle watchers do not	2537	priority are pending (prepare, check and other idle watchers do not count
1879	count).	2538	as receiving "events").
1880		2539
1881	That is, as long as your process is busy handling sockets or timeouts	2540	That is, as long as your process is busy handling sockets or timeouts
1882	(or even signals, imagine) of the same or higher priority it will not be	2541	(or even signals, imagine) of the same or higher priority it will not be
1883	triggered. But when your process is idle (or only lower-priority watchers	2542	triggered. But when your process is idle (or only lower-priority watchers
1884	are pending), the idle watchers are being called once per event loop	2543	are pending), the idle watchers are being called once per event loop
…		…
1895		2554
1896	=head3 Watcher-Specific Functions and Data Members	2555	=head3 Watcher-Specific Functions and Data Members
1897		2556
1898	=over 4	2557	=over 4
1899		2558
1900	=item ev_idle_init (ev_signal *, callback)	2559	=item ev_idle_init (ev_idle *, callback)
1901		2560
1902	Initialises and configures the idle watcher - it has no parameters of any	2561	Initialises and configures the idle watcher - it has no parameters of any
1903	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,	2562	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,
1904	believe me.	2563	believe me.
1905		2564
…		…
1909		2568
1910	Example: Dynamically allocate an C<ev_idle> watcher, start it, and in the	2569	Example: Dynamically allocate an C<ev_idle> watcher, start it, and in the
1911	callback, free it. Also, use no error checking, as usual.	2570	callback, free it. Also, use no error checking, as usual.
1912		2571
1913	static void	2572	static void
1914	idle_cb (struct ev_loop *loop, struct ev_idle *w, int revents)	2573	idle_cb (struct ev_loop *loop, ev_idle *w, int revents)
1915	{	2574	{
1916	free (w);	2575	free (w);
1917	// now do something you wanted to do when the program has	2576	// now do something you wanted to do when the program has
1918	// no longer anything immediate to do.	2577	// no longer anything immediate to do.
1919	}	2578	}
1920		2579
1921	struct ev_idle *idle_watcher = malloc (sizeof (struct ev_idle));	2580	ev_idle *idle_watcher = malloc (sizeof (ev_idle));
1922	ev_idle_init (idle_watcher, idle_cb);	2581	ev_idle_init (idle_watcher, idle_cb);
1923	ev_idle_start (loop, idle_cb);	2582	ev_idle_start (loop, idle_watcher);
1924		2583
1925		2584
1926	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!	2585	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!
1927		2586
1928	Prepare and check watchers are usually (but not always) used in tandem:	2587	Prepare and check watchers are usually (but not always) used in pairs:
1929	prepare watchers get invoked before the process blocks and check watchers	2588	prepare watchers get invoked before the process blocks and check watchers
1930	afterwards.	2589	afterwards.
1931		2590
1932	You I<must not> call C<ev_loop> or similar functions that enter	2591	You I<must not> call C<ev_loop> or similar functions that enter
1933	the current event loop from either C<ev_prepare> or C<ev_check>	2592	the current event loop from either C<ev_prepare> or C<ev_check>
…		…
1936	those watchers, i.e. the sequence will always be C<ev_prepare>, blocking,	2595	those watchers, i.e. the sequence will always be C<ev_prepare>, blocking,
1937	C<ev_check> so if you have one watcher of each kind they will always be	2596	C<ev_check> so if you have one watcher of each kind they will always be
1938	called in pairs bracketing the blocking call.	2597	called in pairs bracketing the blocking call.
1939		2598
1940	Their main purpose is to integrate other event mechanisms into libev and	2599	Their main purpose is to integrate other event mechanisms into libev and
1941	their use is somewhat advanced. This could be used, for example, to track	2600	their use is somewhat advanced. They could be used, for example, to track
1942	variable changes, implement your own watchers, integrate net-snmp or a	2601	variable changes, implement your own watchers, integrate net-snmp or a
1943	coroutine library and lots more. They are also occasionally useful if	2602	coroutine library and lots more. They are also occasionally useful if
1944	you cache some data and want to flush it before blocking (for example,	2603	you cache some data and want to flush it before blocking (for example,
1945	in X programs you might want to do an C<XFlush ()> in an C<ev_prepare>	2604	in X programs you might want to do an C<XFlush ()> in an C<ev_prepare>
1946	watcher).	2605	watcher).
1947		2606
1948	This is done by examining in each prepare call which file descriptors need	2607	This is done by examining in each prepare call which file descriptors
1949	to be watched by the other library, registering C<ev_io> watchers for	2608	need to be watched by the other library, registering C<ev_io> watchers
1950	them and starting an C<ev_timer> watcher for any timeouts (many libraries	2609	for them and starting an C<ev_timer> watcher for any timeouts (many
1951	provide just this functionality). Then, in the check watcher you check for	2610	libraries provide exactly this functionality). Then, in the check watcher,
1952	any events that occurred (by checking the pending status of all watchers	2611	you check for any events that occurred (by checking the pending status
1953	and stopping them) and call back into the library. The I/O and timer	2612	of all watchers and stopping them) and call back into the library. The
1954	callbacks will never actually be called (but must be valid nevertheless,	2613	I/O and timer callbacks will never actually be called (but must be valid
1955	because you never know, you know?).	2614	nevertheless, because you never know, you know?).
1956		2615
1957	As another example, the Perl Coro module uses these hooks to integrate	2616	As another example, the Perl Coro module uses these hooks to integrate
1958	coroutines into libev programs, by yielding to other active coroutines	2617	coroutines into libev programs, by yielding to other active coroutines
1959	during each prepare and only letting the process block if no coroutines	2618	during each prepare and only letting the process block if no coroutines
1960	are ready to run (it's actually more complicated: it only runs coroutines	2619	are ready to run (it's actually more complicated: it only runs coroutines
…		…
1963	loop from blocking if lower-priority coroutines are active, thus mapping	2622	loop from blocking if lower-priority coroutines are active, thus mapping
1964	low-priority coroutines to idle/background tasks).	2623	low-priority coroutines to idle/background tasks).
1965		2624
1966	It is recommended to give C<ev_check> watchers highest (C<EV_MAXPRI>)	2625	It is recommended to give C<ev_check> watchers highest (C<EV_MAXPRI>)
1967	priority, to ensure that they are being run before any other watchers	2626	priority, to ensure that they are being run before any other watchers
		2627	after the poll (this doesn't matter for C<ev_prepare> watchers).
		2628
1968	after the poll. Also, C<ev_check> watchers (and C<ev_prepare> watchers,	2629	Also, C<ev_check> watchers (and C<ev_prepare> watchers, too) should not
1969	too) should not activate ("feed") events into libev. While libev fully	2630	activate ("feed") events into libev. While libev fully supports this, they
1970	supports this, they might get executed before other C<ev_check> watchers	2631	might get executed before other C<ev_check> watchers did their job. As
1971	did their job. As C<ev_check> watchers are often used to embed other	2632	C<ev_check> watchers are often used to embed other (non-libev) event
1972	(non-libev) event loops those other event loops might be in an unusable	2633	loops those other event loops might be in an unusable state until their
1973	state until their C<ev_check> watcher ran (always remind yourself to	2634	C<ev_check> watcher ran (always remind yourself to coexist peacefully with
1974	coexist peacefully with others).	2635	others).
1975		2636
1976	=head3 Watcher-Specific Functions and Data Members	2637	=head3 Watcher-Specific Functions and Data Members
1977		2638
1978	=over 4	2639	=over 4
1979		2640
…		…
1981		2642
1982	=item ev_check_init (ev_check *, callback)	2643	=item ev_check_init (ev_check *, callback)
1983		2644
1984	Initialises and configures the prepare or check watcher - they have no	2645	Initialises and configures the prepare or check watcher - they have no
1985	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>	2646	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>
1986	macros, but using them is utterly, utterly and completely pointless.	2647	macros, but using them is utterly, utterly, utterly and completely
		2648	pointless.
1987		2649
1988	=back	2650	=back
1989		2651
1990	=head3 Examples	2652	=head3 Examples
1991		2653
…		…
2004		2666
2005	static ev_io iow [nfd];	2667	static ev_io iow [nfd];
2006	static ev_timer tw;	2668	static ev_timer tw;
2007		2669
2008	static void	2670	static void
2009	io_cb (ev_loop *loop, ev_io *w, int revents)	2671	io_cb (struct ev_loop *loop, ev_io *w, int revents)
2010	{	2672	{
2011	}	2673	}
2012		2674
2013	// create io watchers for each fd and a timer before blocking	2675	// create io watchers for each fd and a timer before blocking
2014	static void	2676	static void
2015	adns_prepare_cb (ev_loop *loop, ev_prepare *w, int revents)	2677	adns_prepare_cb (struct ev_loop *loop, ev_prepare *w, int revents)
2016	{	2678	{
2017	int timeout = 3600000;	2679	int timeout = 3600000;
2018	struct pollfd fds [nfd];	2680	struct pollfd fds [nfd];
2019	// actual code will need to loop here and realloc etc.	2681	// actual code will need to loop here and realloc etc.
2020	adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));	2682	adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));
2021		2683
2022	/* the callback is illegal, but won't be called as we stop during check */	2684	/* the callback is illegal, but won't be called as we stop during check */
2023	ev_timer_init (&tw, 0, timeout * 1e-3);	2685	ev_timer_init (&tw, 0, timeout * 1e-3, 0.);
2024	ev_timer_start (loop, &tw);	2686	ev_timer_start (loop, &tw);
2025		2687
2026	// create one ev_io per pollfd	2688	// create one ev_io per pollfd
2027	for (int i = 0; i < nfd; ++i)	2689	for (int i = 0; i < nfd; ++i)
2028	{	2690	{
…		…
2035	}	2697	}
2036	}	2698	}
2037		2699
2038	// stop all watchers after blocking	2700	// stop all watchers after blocking
2039	static void	2701	static void
2040	adns_check_cb (ev_loop *loop, ev_check *w, int revents)	2702	adns_check_cb (struct ev_loop *loop, ev_check *w, int revents)
2041	{	2703	{
2042	ev_timer_stop (loop, &tw);	2704	ev_timer_stop (loop, &tw);
2043		2705
2044	for (int i = 0; i < nfd; ++i)	2706	for (int i = 0; i < nfd; ++i)
2045	{	2707	{
…		…
2084	}	2746	}
2085		2747
2086	// do not ever call adns_afterpoll	2748	// do not ever call adns_afterpoll
2087		2749
2088	Method 4: Do not use a prepare or check watcher because the module you	2750	Method 4: Do not use a prepare or check watcher because the module you
2089	want to embed is too inflexible to support it. Instead, you can override	2751	want to embed is not flexible enough to support it. Instead, you can
2090	their poll function. The drawback with this solution is that the main	2752	override their poll function. The drawback with this solution is that the
2091	loop is now no longer controllable by EV. The C<Glib::EV> module does	2753	main loop is now no longer controllable by EV. The C<Glib::EV> module uses
2092	this.	2754	this approach, effectively embedding EV as a client into the horrible
		2755	libglib event loop.
2093		2756
2094	static gint	2757	static gint
2095	event_poll_func (GPollFD *fds, guint nfds, gint timeout)	2758	event_poll_func (GPollFD *fds, guint nfds, gint timeout)
2096	{	2759	{
2097	int got_events = 0;	2760	int got_events = 0;
…		…
2128	prioritise I/O.	2791	prioritise I/O.
2129		2792
2130	As an example for a bug workaround, the kqueue backend might only support	2793	As an example for a bug workaround, the kqueue backend might only support
2131	sockets on some platform, so it is unusable as generic backend, but you	2794	sockets on some platform, so it is unusable as generic backend, but you
2132	still want to make use of it because you have many sockets and it scales	2795	still want to make use of it because you have many sockets and it scales
2133	so nicely. In this case, you would create a kqueue-based loop and embed it	2796	so nicely. In this case, you would create a kqueue-based loop and embed
2134	into your default loop (which might use e.g. poll). Overall operation will	2797	it into your default loop (which might use e.g. poll). Overall operation
2135	be a bit slower because first libev has to poll and then call kevent, but	2798	will be a bit slower because first libev has to call C<poll> and then
2136	at least you can use both at what they are best.	2799	C<kevent>, but at least you can use both mechanisms for what they are
		2800	best: C<kqueue> for scalable sockets and C<poll> if you want it to work :)
2137		2801
2138	As for prioritising I/O: rarely you have the case where some fds have	2802	As for prioritising I/O: under rare circumstances you have the case where
2139	to be watched and handled very quickly (with low latency), and even	2803	some fds have to be watched and handled very quickly (with low latency),
2140	priorities and idle watchers might have too much overhead. In this case	2804	and even priorities and idle watchers might have too much overhead. In
2141	you would put all the high priority stuff in one loop and all the rest in	2805	this case you would put all the high priority stuff in one loop and all
2142	a second one, and embed the second one in the first.	2806	the rest in a second one, and embed the second one in the first.
2143		2807
2144	As long as the watcher is active, the callback will be invoked every time	2808	As long as the watcher is active, the callback will be invoked every
2145	there might be events pending in the embedded loop. The callback must then	2809	time there might be events pending in the embedded loop. The callback
2146	call C<ev_embed_sweep (mainloop, watcher)> to make a single sweep and invoke	2810	must then call C<ev_embed_sweep (mainloop, watcher)> to make a single
2147	their callbacks (you could also start an idle watcher to give the embedded	2811	sweep and invoke their callbacks (the callback doesn't need to invoke the
2148	loop strictly lower priority for example). You can also set the callback	2812	C<ev_embed_sweep> function directly, it could also start an idle watcher
2149	to C<0>, in which case the embed watcher will automatically execute the	2813	to give the embedded loop strictly lower priority for example).
2150	embedded loop sweep.
2151		2814
2152	As long as the watcher is started it will automatically handle events. The	2815	You can also set the callback to C<0>, in which case the embed watcher
2153	callback will be invoked whenever some events have been handled. You can	2816	will automatically execute the embedded loop sweep whenever necessary.
2154	set the callback to C<0> to avoid having to specify one if you are not
2155	interested in that.
2156		2817
2157	Also, there have not currently been made special provisions for forking:	2818	Fork detection will be handled transparently while the C<ev_embed> watcher
2158	when you fork, you not only have to call C<ev_loop_fork> on both loops,	2819	is active, i.e., the embedded loop will automatically be forked when the
2159	but you will also have to stop and restart any C<ev_embed> watchers	2820	embedding loop forks. In other cases, the user is responsible for calling
2160	yourself.	2821	C<ev_loop_fork> on the embedded loop.
2161		2822
2162	Unfortunately, not all backends are embeddable, only the ones returned by	2823	Unfortunately, not all backends are embeddable: only the ones returned by
2163	C<ev_embeddable_backends> are, which, unfortunately, does not include any	2824	C<ev_embeddable_backends> are, which, unfortunately, does not include any
2164	portable one.	2825	portable one.
2165		2826
2166	So when you want to use this feature you will always have to be prepared	2827	So when you want to use this feature you will always have to be prepared
2167	that you cannot get an embeddable loop. The recommended way to get around	2828	that you cannot get an embeddable loop. The recommended way to get around
2168	this is to have a separate variables for your embeddable loop, try to	2829	this is to have a separate variables for your embeddable loop, try to
2169	create it, and if that fails, use the normal loop for everything.	2830	create it, and if that fails, use the normal loop for everything.
		2831
		2832	=head3 C<ev_embed> and fork
		2833
		2834	While the C<ev_embed> watcher is running, forks in the embedding loop will
		2835	automatically be applied to the embedded loop as well, so no special
		2836	fork handling is required in that case. When the watcher is not running,
		2837	however, it is still the task of the libev user to call C<ev_loop_fork ()>
		2838	as applicable.
2170		2839
2171	=head3 Watcher-Specific Functions and Data Members	2840	=head3 Watcher-Specific Functions and Data Members
2172		2841
2173	=over 4	2842	=over 4
2174		2843
…		…
2202	C<loop_lo> (which is C<loop_hi> in the case no embeddable loop can be	2871	C<loop_lo> (which is C<loop_hi> in the case no embeddable loop can be
2203	used).	2872	used).
2204		2873
2205	struct ev_loop *loop_hi = ev_default_init (0);	2874	struct ev_loop *loop_hi = ev_default_init (0);
2206	struct ev_loop *loop_lo = 0;	2875	struct ev_loop *loop_lo = 0;
2207	struct ev_embed embed;	2876	ev_embed embed;
2208		2877
2209	// see if there is a chance of getting one that works	2878	// see if there is a chance of getting one that works
2210	// (remember that a flags value of 0 means autodetection)	2879	// (remember that a flags value of 0 means autodetection)
2211	loop_lo = ev_embeddable_backends () & ev_recommended_backends ()	2880	loop_lo = ev_embeddable_backends () & ev_recommended_backends ()
2212	? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())	2881	? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())
…		…
2226	kqueue implementation). Store the kqueue/socket-only event loop in	2895	kqueue implementation). Store the kqueue/socket-only event loop in
2227	C<loop_socket>. (One might optionally use C<EVFLAG_NOENV>, too).	2896	C<loop_socket>. (One might optionally use C<EVFLAG_NOENV>, too).
2228		2897
2229	struct ev_loop *loop = ev_default_init (0);	2898	struct ev_loop *loop = ev_default_init (0);
2230	struct ev_loop *loop_socket = 0;	2899	struct ev_loop *loop_socket = 0;
2231	struct ev_embed embed;	2900	ev_embed embed;
2232		2901
2233	if (ev_supported_backends () & ~ev_recommended_backends () & EVBACKEND_KQUEUE)	2902	if (ev_supported_backends () & ~ev_recommended_backends () & EVBACKEND_KQUEUE)
2234	if ((loop_socket = ev_loop_new (EVBACKEND_KQUEUE))	2903	if ((loop_socket = ev_loop_new (EVBACKEND_KQUEUE))
2235	{	2904	{
2236	ev_embed_init (&embed, 0, loop_socket);	2905	ev_embed_init (&embed, 0, loop_socket);
…		…
2251	event loop blocks next and before C<ev_check> watchers are being called,	2920	event loop blocks next and before C<ev_check> watchers are being called,
2252	and only in the child after the fork. If whoever good citizen calling	2921	and only in the child after the fork. If whoever good citizen calling
2253	C<ev_default_fork> cheats and calls it in the wrong process, the fork	2922	C<ev_default_fork> cheats and calls it in the wrong process, the fork
2254	handlers will be invoked, too, of course.	2923	handlers will be invoked, too, of course.
2255		2924
		2925	=head3 The special problem of life after fork - how is it possible?
		2926
		2927	Most uses of C<fork()> consist of forking, then some simple calls to ste
		2928	up/change the process environment, followed by a call to C<exec()>. This
		2929	sequence should be handled by libev without any problems.
		2930
		2931	This changes when the application actually wants to do event handling
		2932	in the child, or both parent in child, in effect "continuing" after the
		2933	fork.
		2934
		2935	The default mode of operation (for libev, with application help to detect
		2936	forks) is to duplicate all the state in the child, as would be expected
		2937	when I<either> the parent I<or> the child process continues.
		2938
		2939	When both processes want to continue using libev, then this is usually the
		2940	wrong result. In that case, usually one process (typically the parent) is
		2941	supposed to continue with all watchers in place as before, while the other
		2942	process typically wants to start fresh, i.e. without any active watchers.
		2943
		2944	The cleanest and most efficient way to achieve that with libev is to
		2945	simply create a new event loop, which of course will be "empty", and
		2946	use that for new watchers. This has the advantage of not touching more
		2947	memory than necessary, and thus avoiding the copy-on-write, and the
		2948	disadvantage of having to use multiple event loops (which do not support
		2949	signal watchers).
		2950
		2951	When this is not possible, or you want to use the default loop for
		2952	other reasons, then in the process that wants to start "fresh", call
		2953	C<ev_default_destroy ()> followed by C<ev_default_loop (...)>. Destroying
		2954	the default loop will "orphan" (not stop) all registered watchers, so you
		2955	have to be careful not to execute code that modifies those watchers. Note
		2956	also that in that case, you have to re-register any signal watchers.
		2957
2256	=head3 Watcher-Specific Functions and Data Members	2958	=head3 Watcher-Specific Functions and Data Members
2257		2959
2258	=over 4	2960	=over 4
2259		2961
2260	=item ev_fork_init (ev_signal *, callback)	2962	=item ev_fork_init (ev_signal *, callback)
…		…
2289	=head3 Queueing	2991	=head3 Queueing
2290		2992
2291	C<ev_async> does not support queueing of data in any way. The reason	2993	C<ev_async> does not support queueing of data in any way. The reason
2292	is that the author does not know of a simple (or any) algorithm for a	2994	is that the author does not know of a simple (or any) algorithm for a
2293	multiple-writer-single-reader queue that works in all cases and doesn't	2995	multiple-writer-single-reader queue that works in all cases and doesn't
2294	need elaborate support such as pthreads.	2996	need elaborate support such as pthreads or unportable memory access
		2997	semantics.
2295		2998
2296	That means that if you want to queue data, you have to provide your own	2999	That means that if you want to queue data, you have to provide your own
2297	queue. But at least I can tell you would implement locking around your	3000	queue. But at least I can tell you how to implement locking around your
2298	queue:	3001	queue:
2299		3002
2300	=over 4	3003	=over 4
2301		3004
2302	=item queueing from a signal handler context	3005	=item queueing from a signal handler context
2303		3006
2304	To implement race-free queueing, you simply add to the queue in the signal	3007	To implement race-free queueing, you simply add to the queue in the signal
2305	handler but you block the signal handler in the watcher callback. Here is an example that does that for	3008	handler but you block the signal handler in the watcher callback. Here is
2306	some fictitious SIGUSR1 handler:	3009	an example that does that for some fictitious SIGUSR1 handler:
2307		3010
2308	static ev_async mysig;	3011	static ev_async mysig;
2309		3012
2310	static void	3013	static void
2311	sigusr1_handler (void)	3014	sigusr1_handler (void)
…		…
2377	=over 4	3080	=over 4
2378		3081
2379	=item ev_async_init (ev_async *, callback)	3082	=item ev_async_init (ev_async *, callback)
2380		3083
2381	Initialises and configures the async watcher - it has no parameters of any	3084	Initialises and configures the async watcher - it has no parameters of any
2382	kind. There is a C<ev_asynd_set> macro, but using it is utterly pointless,	3085	kind. There is a C<ev_async_set> macro, but using it is utterly pointless,
2383	believe me.	3086	trust me.
2384		3087
2385	=item ev_async_send (loop, ev_async *)	3088	=item ev_async_send (loop, ev_async *)
2386		3089
2387	Sends/signals/activates the given C<ev_async> watcher, that is, feeds	3090	Sends/signals/activates the given C<ev_async> watcher, that is, feeds
2388	an C<EV_ASYNC> event on the watcher into the event loop. Unlike	3091	an C<EV_ASYNC> event on the watcher into the event loop. Unlike
2389	C<ev_feed_event>, this call is safe to do in other threads, signal or	3092	C<ev_feed_event>, this call is safe to do from other threads, signal or
2390	similar contexts (see the discussion of C<EV_ATOMIC_T> in the embedding	3093	similar contexts (see the discussion of C<EV_ATOMIC_T> in the embedding
2391	section below on what exactly this means).	3094	section below on what exactly this means).
2392		3095
		3096	Note that, as with other watchers in libev, multiple events might get
		3097	compressed into a single callback invocation (another way to look at this
		3098	is that C<ev_async> watchers are level-triggered, set on C<ev_async_send>,
		3099	reset when the event loop detects that).
		3100
2393	This call incurs the overhead of a system call only once per loop iteration,	3101	This call incurs the overhead of a system call only once per event loop
2394	so while the overhead might be noticeable, it doesn't apply to repeated	3102	iteration, so while the overhead might be noticeable, it doesn't apply to
2395	calls to C<ev_async_send>.	3103	repeated calls to C<ev_async_send> for the same event loop.
2396		3104
2397	=item bool = ev_async_pending (ev_async *)	3105	=item bool = ev_async_pending (ev_async *)
2398		3106
2399	Returns a non-zero value when C<ev_async_send> has been called on the	3107	Returns a non-zero value when C<ev_async_send> has been called on the
2400	watcher but the event has not yet been processed (or even noted) by the	3108	watcher but the event has not yet been processed (or even noted) by the
…		…
2403	C<ev_async_send> sets a flag in the watcher and wakes up the loop. When	3111	C<ev_async_send> sets a flag in the watcher and wakes up the loop. When
2404	the loop iterates next and checks for the watcher to have become active,	3112	the loop iterates next and checks for the watcher to have become active,
2405	it will reset the flag again. C<ev_async_pending> can be used to very	3113	it will reset the flag again. C<ev_async_pending> can be used to very
2406	quickly check whether invoking the loop might be a good idea.	3114	quickly check whether invoking the loop might be a good idea.
2407		3115
2408	Not that this does I<not> check whether the watcher itself is pending, only	3116	Not that this does I<not> check whether the watcher itself is pending,
2409	whether it has been requested to make this watcher pending.	3117	only whether it has been requested to make this watcher pending: there
		3118	is a time window between the event loop checking and resetting the async
		3119	notification, and the callback being invoked.
2410		3120
2411	=back	3121	=back
2412		3122
2413		3123
2414	=head1 OTHER FUNCTIONS	3124	=head1 OTHER FUNCTIONS
…		…
2418	=over 4	3128	=over 4
2419		3129
2420	=item ev_once (loop, int fd, int events, ev_tstamp timeout, callback)	3130	=item ev_once (loop, int fd, int events, ev_tstamp timeout, callback)
2421		3131
2422	This function combines a simple timer and an I/O watcher, calls your	3132	This function combines a simple timer and an I/O watcher, calls your
2423	callback on whichever event happens first and automatically stop both	3133	callback on whichever event happens first and automatically stops both
2424	watchers. This is useful if you want to wait for a single event on an fd	3134	watchers. This is useful if you want to wait for a single event on an fd
2425	or timeout without having to allocate/configure/start/stop/free one or	3135	or timeout without having to allocate/configure/start/stop/free one or
2426	more watchers yourself.	3136	more watchers yourself.
2427		3137
2428	If C<fd> is less than 0, then no I/O watcher will be started and events	3138	If C<fd> is less than 0, then no I/O watcher will be started and the
2429	is being ignored. Otherwise, an C<ev_io> watcher for the given C<fd> and	3139	C<events> argument is being ignored. Otherwise, an C<ev_io> watcher for
2430	C<events> set will be created and started.	3140	the given C<fd> and C<events> set will be created and started.
2431		3141
2432	If C<timeout> is less than 0, then no timeout watcher will be	3142	If C<timeout> is less than 0, then no timeout watcher will be
2433	started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and	3143	started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and
2434	repeat = 0) will be started. While C<0> is a valid timeout, it is of	3144	repeat = 0) will be started. C<0> is a valid timeout.
2435	dubious value.
2436		3145
2437	The callback has the type C<void (*cb)(int revents, void *arg)> and gets	3146	The callback has the type C<void (*cb)(int revents, void *arg)> and gets
2438	passed an C<revents> set like normal event callbacks (a combination of	3147	passed an C<revents> set like normal event callbacks (a combination of
2439	C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMEOUT>) and the C<arg>	3148	C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMEOUT>) and the C<arg>
2440	value passed to C<ev_once>:	3149	value passed to C<ev_once>. Note that it is possible to receive I<both>
		3150	a timeout and an io event at the same time - you probably should give io
		3151	events precedence.
		3152
		3153	Example: wait up to ten seconds for data to appear on STDIN_FILENO.
2441		3154
2442	static void stdin_ready (int revents, void *arg)	3155	static void stdin_ready (int revents, void *arg)
2443	{	3156	{
		3157	if (revents & EV_READ)
		3158	/* stdin might have data for us, joy! */;
2444	if (revents & EV_TIMEOUT)	3159	else if (revents & EV_TIMEOUT)
2445	/* doh, nothing entered */;	3160	/* doh, nothing entered */;
2446	else if (revents & EV_READ)
2447	/* stdin might have data for us, joy! */;
2448	}	3161	}
2449		3162
2450	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);	3163	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);
2451		3164
2452	=item ev_feed_event (ev_loop *, watcher *, int revents)
2453
2454	Feeds the given event set into the event loop, as if the specified event
2455	had happened for the specified watcher (which must be a pointer to an
2456	initialised but not necessarily started event watcher).
2457
2458	=item ev_feed_fd_event (ev_loop *, int fd, int revents)	3165	=item ev_feed_fd_event (loop, int fd, int revents)
2459		3166
2460	Feed an event on the given fd, as if a file descriptor backend detected	3167	Feed an event on the given fd, as if a file descriptor backend detected
2461	the given events it.	3168	the given events it.
2462		3169
2463	=item ev_feed_signal_event (ev_loop *loop, int signum)	3170	=item ev_feed_signal_event (loop, int signum)
2464		3171
2465	Feed an event as if the given signal occurred (C<loop> must be the default	3172	Feed an event as if the given signal occurred (C<loop> must be the default
2466	loop!).	3173	loop!).
2467		3174
2468	=back	3175	=back
…		…
2548		3255
2549	=over 4	3256	=over 4
2550		3257
2551	=item ev::TYPE::TYPE ()	3258	=item ev::TYPE::TYPE ()
2552		3259
2553	=item ev::TYPE::TYPE (struct ev_loop *)	3260	=item ev::TYPE::TYPE (loop)
2554		3261
2555	=item ev::TYPE::~TYPE	3262	=item ev::TYPE::~TYPE
2556		3263
2557	The constructor (optionally) takes an event loop to associate the watcher	3264	The constructor (optionally) takes an event loop to associate the watcher
2558	with. If it is omitted, it will use C<EV_DEFAULT>.	3265	with. If it is omitted, it will use C<EV_DEFAULT>.
…		…
2590		3297
2591	myclass obj;	3298	myclass obj;
2592	ev::io iow;	3299	ev::io iow;
2593	iow.set <myclass, &myclass::io_cb> (&obj);	3300	iow.set <myclass, &myclass::io_cb> (&obj);
2594		3301
		3302	=item w->set (object *)
		3303
		3304	This is an B<experimental> feature that might go away in a future version.
		3305
		3306	This is a variation of a method callback - leaving out the method to call
		3307	will default the method to C<operator ()>, which makes it possible to use
		3308	functor objects without having to manually specify the C<operator ()> all
		3309	the time. Incidentally, you can then also leave out the template argument
		3310	list.
		3311
		3312	The C<operator ()> method prototype must be C<void operator ()(watcher &w,
		3313	int revents)>.
		3314
		3315	See the method-C<set> above for more details.
		3316
		3317	Example: use a functor object as callback.
		3318
		3319	struct myfunctor
		3320	{
		3321	void operator() (ev::io &w, int revents)
		3322	{
		3323	...
		3324	}
		3325	}
		3326
		3327	myfunctor f;
		3328
		3329	ev::io w;
		3330	w.set (&f);
		3331
2595	=item w->set<function> (void *data = 0)	3332	=item w->set<function> (void *data = 0)
2596		3333
2597	Also sets a callback, but uses a static method or plain function as	3334	Also sets a callback, but uses a static method or plain function as
2598	callback. The optional C<data> argument will be stored in the watcher's	3335	callback. The optional C<data> argument will be stored in the watcher's
2599	C<data> member and is free for you to use.	3336	C<data> member and is free for you to use.
2600		3337
2601	The prototype of the C<function> must be C<void (*)(ev::TYPE &w, int)>.	3338	The prototype of the C<function> must be C<void (*)(ev::TYPE &w, int)>.
2602		3339
2603	See the method-C<set> above for more details.	3340	See the method-C<set> above for more details.
2604		3341
2605	Example:	3342	Example: Use a plain function as callback.
2606		3343
2607	static void io_cb (ev::io &w, int revents) { }	3344	static void io_cb (ev::io &w, int revents) { }
2608	iow.set <io_cb> ();	3345	iow.set <io_cb> ();
2609		3346
2610	=item w->set (struct ev_loop *)	3347	=item w->set (loop)
2611		3348
2612	Associates a different C<struct ev_loop> with this watcher. You can only	3349	Associates a different C<struct ev_loop> with this watcher. You can only
2613	do this when the watcher is inactive (and not pending either).	3350	do this when the watcher is inactive (and not pending either).
2614		3351
2615	=item w->set ([arguments])	3352	=item w->set ([arguments])
…		…
2648	Example: Define a class with an IO and idle watcher, start one of them in	3385	Example: Define a class with an IO and idle watcher, start one of them in
2649	the constructor.	3386	the constructor.
2650		3387
2651	class myclass	3388	class myclass
2652	{	3389	{
2653	ev::io io; void io_cb (ev::io &w, int revents);	3390	ev::io io ; void io_cb (ev::io &w, int revents);
2654	ev:idle idle void idle_cb (ev::idle &w, int revents);	3391	ev::idle idle; void idle_cb (ev::idle &w, int revents);
2655		3392
2656	myclass (int fd)	3393	myclass (int fd)
2657	{	3394	{
2658	io .set <myclass, &myclass::io_cb > (this);	3395	io .set <myclass, &myclass::io_cb > (this);
2659	idle.set <myclass, &myclass::idle_cb> (this);	3396	idle.set <myclass, &myclass::idle_cb> (this);
…		…
2675	=item Perl	3412	=item Perl
2676		3413
2677	The EV module implements the full libev API and is actually used to test	3414	The EV module implements the full libev API and is actually used to test
2678	libev. EV is developed together with libev. Apart from the EV core module,	3415	libev. EV is developed together with libev. Apart from the EV core module,
2679	there are additional modules that implement libev-compatible interfaces	3416	there are additional modules that implement libev-compatible interfaces
2680	to C<libadns> (C<EV::ADNS>), C<Net::SNMP> (C<Net::SNMP::EV>) and the	3417	to C<libadns> (C<EV::ADNS>, but C<AnyEvent::DNS> is preferred nowadays),
2681	C<libglib> event core (C<Glib::EV> and C<EV::Glib>).	3418	C<Net::SNMP> (C<Net::SNMP::EV>) and the C<libglib> event core (C<Glib::EV>
		3419	and C<EV::Glib>).
2682		3420
2683	It can be found and installed via CPAN, its homepage is at	3421	It can be found and installed via CPAN, its homepage is at
2684	L<http://software.schmorp.de/pkg/EV>.	3422	L<http://software.schmorp.de/pkg/EV>.
2685		3423
2686	=item Python	3424	=item Python
2687		3425
2688	Python bindings can be found at L<http://code.google.com/p/pyev/>. It	3426	Python bindings can be found at L<http://code.google.com/p/pyev/>. It
2689	seems to be quite complete and well-documented. Note, however, that the	3427	seems to be quite complete and well-documented.
2690	patch they require for libev is outright dangerous as it breaks the ABI
2691	for everybody else, and therefore, should never be applied in an installed
2692	libev (if python requires an incompatible ABI then it needs to embed
2693	libev).
2694		3428
2695	=item Ruby	3429	=item Ruby
2696		3430
2697	Tony Arcieri has written a ruby extension that offers access to a subset	3431	Tony Arcieri has written a ruby extension that offers access to a subset
2698	of the libev API and adds file handle abstractions, asynchronous DNS and	3432	of the libev API and adds file handle abstractions, asynchronous DNS and
2699	more on top of it. It can be found via gem servers. Its homepage is at	3433	more on top of it. It can be found via gem servers. Its homepage is at
2700	L<http://rev.rubyforge.org/>.	3434	L<http://rev.rubyforge.org/>.
2701		3435
		3436	Roger Pack reports that using the link order C<-lws2_32 -lmsvcrt-ruby-190>
		3437	makes rev work even on mingw.
		3438
		3439	=item Haskell
		3440
		3441	A haskell binding to libev is available at
		3442	L<http://hackage.haskell.org/cgi-bin/hackage-scripts/package/hlibev>.
		3443
2702	=item D	3444	=item D
2703		3445
2704	Leandro Lucarella has written a D language binding (F<ev.d>) for libev, to	3446	Leandro Lucarella has written a D language binding (F<ev.d>) for libev, to
2705	be found at L<http://proj.llucax.com.ar/wiki/evd>.	3447	be found at L<http://proj.llucax.com.ar/wiki/evd>.
		3448
		3449	=item Ocaml
		3450
		3451	Erkki Seppala has written Ocaml bindings for libev, to be found at
		3452	L<http://modeemi.cs.tut.fi/~flux/software/ocaml-ev/>.
		3453
		3454	=item Lua
		3455
		3456	Brian Maher has written a partial interface to libev for lua (at the
		3457	time of this writing, only C<ev_io> and C<ev_timer>), to be found at
		3458	L<http://github.com/brimworks/lua-ev>.
2706		3459
2707	=back	3460	=back
2708		3461
2709		3462
2710	=head1 MACRO MAGIC	3463	=head1 MACRO MAGIC
…		…
2811		3564
2812	#define EV_STANDALONE 1	3565	#define EV_STANDALONE 1
2813	#include "ev.h"	3566	#include "ev.h"
2814		3567
2815	Both header files and implementation files can be compiled with a C++	3568	Both header files and implementation files can be compiled with a C++
2816	compiler (at least, thats a stated goal, and breakage will be treated	3569	compiler (at least, that's a stated goal, and breakage will be treated
2817	as a bug).	3570	as a bug).
2818		3571
2819	You need the following files in your source tree, or in a directory	3572	You need the following files in your source tree, or in a directory
2820	in your include path (e.g. in libev/ when using -Ilibev):	3573	in your include path (e.g. in libev/ when using -Ilibev):
2821		3574
…		…
2864	libev.m4	3617	libev.m4
2865		3618
2866	=head2 PREPROCESSOR SYMBOLS/MACROS	3619	=head2 PREPROCESSOR SYMBOLS/MACROS
2867		3620
2868	Libev can be configured via a variety of preprocessor symbols you have to	3621	Libev can be configured via a variety of preprocessor symbols you have to
2869	define before including any of its files. The default in the absence of	3622	define before including (or compiling) any of its files. The default in
2870	autoconf is noted for every option.	3623	the absence of autoconf is documented for every option.
		3624
		3625	Symbols marked with "(h)" do not change the ABI, and can have different
		3626	values when compiling libev vs. including F<ev.h>, so it is permissible
		3627	to redefine them before including F<ev.h> without breakign compatibility
		3628	to a compiled library. All other symbols change the ABI, which means all
		3629	users of libev and the libev code itself must be compiled with compatible
		3630	settings.
2871		3631
2872	=over 4	3632	=over 4
2873		3633
2874	=item EV_STANDALONE	3634	=item EV_STANDALONE (h)
2875		3635
2876	Must always be C<1> if you do not use autoconf configuration, which	3636	Must always be C<1> if you do not use autoconf configuration, which
2877	keeps libev from including F<config.h>, and it also defines dummy	3637	keeps libev from including F<config.h>, and it also defines dummy
2878	implementations for some libevent functions (such as logging, which is not	3638	implementations for some libevent functions (such as logging, which is not
2879	supported). It will also not define any of the structs usually found in	3639	supported). It will also not define any of the structs usually found in
2880	F<event.h> that are not directly supported by the libev core alone.	3640	F<event.h> that are not directly supported by the libev core alone.
2881		3641
		3642	In standalone mode, libev will still try to automatically deduce the
		3643	configuration, but has to be more conservative.
		3644
2882	=item EV_USE_MONOTONIC	3645	=item EV_USE_MONOTONIC
2883		3646
2884	If defined to be C<1>, libev will try to detect the availability of the	3647	If defined to be C<1>, libev will try to detect the availability of the
2885	monotonic clock option at both compile time and runtime. Otherwise no use	3648	monotonic clock option at both compile time and runtime. Otherwise no
2886	of the monotonic clock option will be attempted. If you enable this, you	3649	use of the monotonic clock option will be attempted. If you enable this,
2887	usually have to link against librt or something similar. Enabling it when	3650	you usually have to link against librt or something similar. Enabling it
2888	the functionality isn't available is safe, though, although you have	3651	when the functionality isn't available is safe, though, although you have
2889	to make sure you link against any libraries where the C<clock_gettime>	3652	to make sure you link against any libraries where the C<clock_gettime>
2890	function is hiding in (often F<-lrt>).	3653	function is hiding in (often F<-lrt>). See also C<EV_USE_CLOCK_SYSCALL>.
2891		3654
2892	=item EV_USE_REALTIME	3655	=item EV_USE_REALTIME
2893		3656
2894	If defined to be C<1>, libev will try to detect the availability of the	3657	If defined to be C<1>, libev will try to detect the availability of the
2895	real-time clock option at compile time (and assume its availability at	3658	real-time clock option at compile time (and assume its availability
2896	runtime if successful). Otherwise no use of the real-time clock option will	3659	at runtime if successful). Otherwise no use of the real-time clock
2897	be attempted. This effectively replaces C<gettimeofday> by C<clock_get	3660	option will be attempted. This effectively replaces C<gettimeofday>
2898	(CLOCK_REALTIME, ...)> and will not normally affect correctness. See the	3661	by C<clock_get (CLOCK_REALTIME, ...)> and will not normally affect
2899	note about libraries in the description of C<EV_USE_MONOTONIC>, though.	3662	correctness. See the note about libraries in the description of
		3663	C<EV_USE_MONOTONIC>, though. Defaults to the opposite value of
		3664	C<EV_USE_CLOCK_SYSCALL>.
		3665
		3666	=item EV_USE_CLOCK_SYSCALL
		3667
		3668	If defined to be C<1>, libev will try to use a direct syscall instead
		3669	of calling the system-provided C<clock_gettime> function. This option
		3670	exists because on GNU/Linux, C<clock_gettime> is in C<librt>, but C<librt>
		3671	unconditionally pulls in C<libpthread>, slowing down single-threaded
		3672	programs needlessly. Using a direct syscall is slightly slower (in
		3673	theory), because no optimised vdso implementation can be used, but avoids
		3674	the pthread dependency. Defaults to C<1> on GNU/Linux with glibc 2.x or
		3675	higher, as it simplifies linking (no need for C<-lrt>).
2900		3676
2901	=item EV_USE_NANOSLEEP	3677	=item EV_USE_NANOSLEEP
2902		3678
2903	If defined to be C<1>, libev will assume that C<nanosleep ()> is available	3679	If defined to be C<1>, libev will assume that C<nanosleep ()> is available
2904	and will use it for delays. Otherwise it will use C<select ()>.	3680	and will use it for delays. Otherwise it will use C<select ()>.
…		…
2920		3696
2921	=item EV_SELECT_USE_FD_SET	3697	=item EV_SELECT_USE_FD_SET
2922		3698
2923	If defined to C<1>, then the select backend will use the system C<fd_set>	3699	If defined to C<1>, then the select backend will use the system C<fd_set>
2924	structure. This is useful if libev doesn't compile due to a missing	3700	structure. This is useful if libev doesn't compile due to a missing
2925	C<NFDBITS> or C<fd_mask> definition or it mis-guesses the bitset layout on	3701	C<NFDBITS> or C<fd_mask> definition or it mis-guesses the bitset layout
2926	exotic systems. This usually limits the range of file descriptors to some	3702	on exotic systems. This usually limits the range of file descriptors to
2927	low limit such as 1024 or might have other limitations (winsocket only	3703	some low limit such as 1024 or might have other limitations (winsocket
2928	allows 64 sockets). The C<FD_SETSIZE> macro, set before compilation, might	3704	only allows 64 sockets). The C<FD_SETSIZE> macro, set before compilation,
2929	influence the size of the C<fd_set> used.	3705	configures the maximum size of the C<fd_set>.
2930		3706
2931	=item EV_SELECT_IS_WINSOCKET	3707	=item EV_SELECT_IS_WINSOCKET
2932		3708
2933	When defined to C<1>, the select backend will assume that	3709	When defined to C<1>, the select backend will assume that
2934	select/socket/connect etc. don't understand file descriptors but	3710	select/socket/connect etc. don't understand file descriptors but
…		…
2936	be used is the winsock select). This means that it will call	3712	be used is the winsock select). This means that it will call
2937	C<_get_osfhandle> on the fd to convert it to an OS handle. Otherwise,	3713	C<_get_osfhandle> on the fd to convert it to an OS handle. Otherwise,
2938	it is assumed that all these functions actually work on fds, even	3714	it is assumed that all these functions actually work on fds, even
2939	on win32. Should not be defined on non-win32 platforms.	3715	on win32. Should not be defined on non-win32 platforms.
2940		3716
2941	=item EV_FD_TO_WIN32_HANDLE	3717	=item EV_FD_TO_WIN32_HANDLE(fd)
2942		3718
2943	If C<EV_SELECT_IS_WINSOCKET> is enabled, then libev needs a way to map	3719	If C<EV_SELECT_IS_WINSOCKET> is enabled, then libev needs a way to map
2944	file descriptors to socket handles. When not defining this symbol (the	3720	file descriptors to socket handles. When not defining this symbol (the
2945	default), then libev will call C<_get_osfhandle>, which is usually	3721	default), then libev will call C<_get_osfhandle>, which is usually
2946	correct. In some cases, programs use their own file descriptor management,	3722	correct. In some cases, programs use their own file descriptor management,
2947	in which case they can provide this function to map fds to socket handles.	3723	in which case they can provide this function to map fds to socket handles.
		3724
		3725	=item EV_WIN32_HANDLE_TO_FD(handle)
		3726
		3727	If C<EV_SELECT_IS_WINSOCKET> then libev maps handles to file descriptors
		3728	using the standard C<_open_osfhandle> function. For programs implementing
		3729	their own fd to handle mapping, overwriting this function makes it easier
		3730	to do so. This can be done by defining this macro to an appropriate value.
		3731
		3732	=item EV_WIN32_CLOSE_FD(fd)
		3733
		3734	If programs implement their own fd to handle mapping on win32, then this
		3735	macro can be used to override the C<close> function, useful to unregister
		3736	file descriptors again. Note that the replacement function has to close
		3737	the underlying OS handle.
2948		3738
2949	=item EV_USE_POLL	3739	=item EV_USE_POLL
2950		3740
2951	If defined to be C<1>, libev will compile in support for the C<poll>(2)	3741	If defined to be C<1>, libev will compile in support for the C<poll>(2)
2952	backend. Otherwise it will be enabled on non-win32 platforms. It	3742	backend. Otherwise it will be enabled on non-win32 platforms. It
…		…
2999	as well as for signal and thread safety in C<ev_async> watchers.	3789	as well as for signal and thread safety in C<ev_async> watchers.
3000		3790
3001	In the absence of this define, libev will use C<sig_atomic_t volatile>	3791	In the absence of this define, libev will use C<sig_atomic_t volatile>
3002	(from F<signal.h>), which is usually good enough on most platforms.	3792	(from F<signal.h>), which is usually good enough on most platforms.
3003		3793
3004	=item EV_H	3794	=item EV_H (h)
3005		3795
3006	The name of the F<ev.h> header file used to include it. The default if	3796	The name of the F<ev.h> header file used to include it. The default if
3007	undefined is C<"ev.h"> in F<event.h>, F<ev.c> and F<ev++.h>. This can be	3797	undefined is C<"ev.h"> in F<event.h>, F<ev.c> and F<ev++.h>. This can be
3008	used to virtually rename the F<ev.h> header file in case of conflicts.	3798	used to virtually rename the F<ev.h> header file in case of conflicts.
3009		3799
3010	=item EV_CONFIG_H	3800	=item EV_CONFIG_H (h)
3011		3801
3012	If C<EV_STANDALONE> isn't C<1>, this variable can be used to override	3802	If C<EV_STANDALONE> isn't C<1>, this variable can be used to override
3013	F<ev.c>'s idea of where to find the F<config.h> file, similarly to	3803	F<ev.c>'s idea of where to find the F<config.h> file, similarly to
3014	C<EV_H>, above.	3804	C<EV_H>, above.
3015		3805
3016	=item EV_EVENT_H	3806	=item EV_EVENT_H (h)
3017		3807
3018	Similarly to C<EV_H>, this macro can be used to override F<event.c>'s idea	3808	Similarly to C<EV_H>, this macro can be used to override F<event.c>'s idea
3019	of how the F<event.h> header can be found, the default is C<"event.h">.	3809	of how the F<event.h> header can be found, the default is C<"event.h">.
3020		3810
3021	=item EV_PROTOTYPES	3811	=item EV_PROTOTYPES (h)
3022		3812
3023	If defined to be C<0>, then F<ev.h> will not define any function	3813	If defined to be C<0>, then F<ev.h> will not define any function
3024	prototypes, but still define all the structs and other symbols. This is	3814	prototypes, but still define all the structs and other symbols. This is
3025	occasionally useful if you want to provide your own wrapper functions	3815	occasionally useful if you want to provide your own wrapper functions
3026	around libev functions.	3816	around libev functions.
…		…
3045	When doing priority-based operations, libev usually has to linearly search	3835	When doing priority-based operations, libev usually has to linearly search
3046	all the priorities, so having many of them (hundreds) uses a lot of space	3836	all the priorities, so having many of them (hundreds) uses a lot of space
3047	and time, so using the defaults of five priorities (-2 .. +2) is usually	3837	and time, so using the defaults of five priorities (-2 .. +2) is usually
3048	fine.	3838	fine.
3049		3839
3050	If your embedding application does not need any priorities, defining these both to	3840	If your embedding application does not need any priorities, defining these
3051	C<0> will save some memory and CPU.	3841	both to C<0> will save some memory and CPU.
3052		3842
3053	=item EV_PERIODIC_ENABLE	3843	=item EV_PERIODIC_ENABLE
3054		3844
3055	If undefined or defined to be C<1>, then periodic timers are supported. If	3845	If undefined or defined to be C<1>, then periodic timers are supported. If
3056	defined to be C<0>, then they are not. Disabling them saves a few kB of	3846	defined to be C<0>, then they are not. Disabling them saves a few kB of
…		…
3063	code.	3853	code.
3064		3854
3065	=item EV_EMBED_ENABLE	3855	=item EV_EMBED_ENABLE
3066		3856
3067	If undefined or defined to be C<1>, then embed watchers are supported. If	3857	If undefined or defined to be C<1>, then embed watchers are supported. If
3068	defined to be C<0>, then they are not.	3858	defined to be C<0>, then they are not. Embed watchers rely on most other
		3859	watcher types, which therefore must not be disabled.
3069		3860
3070	=item EV_STAT_ENABLE	3861	=item EV_STAT_ENABLE
3071		3862
3072	If undefined or defined to be C<1>, then stat watchers are supported. If	3863	If undefined or defined to be C<1>, then stat watchers are supported. If
3073	defined to be C<0>, then they are not.	3864	defined to be C<0>, then they are not.
…		…
3075	=item EV_FORK_ENABLE	3866	=item EV_FORK_ENABLE
3076		3867
3077	If undefined or defined to be C<1>, then fork watchers are supported. If	3868	If undefined or defined to be C<1>, then fork watchers are supported. If
3078	defined to be C<0>, then they are not.	3869	defined to be C<0>, then they are not.
3079		3870
		3871	=item EV_SIGNAL_ENABLE
		3872
		3873	If undefined or defined to be C<1>, then signal watchers are supported. If
		3874	defined to be C<0>, then they are not.
		3875
3080	=item EV_ASYNC_ENABLE	3876	=item EV_ASYNC_ENABLE
3081		3877
3082	If undefined or defined to be C<1>, then async watchers are supported. If	3878	If undefined or defined to be C<1>, then async watchers are supported. If
3083	defined to be C<0>, then they are not.	3879	defined to be C<0>, then they are not.
3084		3880
		3881	=item EV_CHILD_ENABLE
		3882
		3883	If undefined or defined to be C<1> (and C<_WIN32> is not defined), then
		3884	child watchers are supported. If defined to be C<0>, then they are not.
		3885
3085	=item EV_MINIMAL	3886	=item EV_MINIMAL
3086		3887
3087	If you need to shave off some kilobytes of code at the expense of some	3888	If you need to shave off some kilobytes of code at the expense of some
3088	speed, define this symbol to C<1>. Currently this is used to override some	3889	speed (but with the full API), define this symbol to C<1>. Currently this
3089	inlining decisions, saves roughly 30% code size on amd64. It also selects a	3890	is used to override some inlining decisions, saves roughly 30% code size
3090	much smaller 2-heap for timer management over the default 4-heap.	3891	on amd64. It also selects a much smaller 2-heap for timer management over
		3892	the default 4-heap.
		3893
		3894	You can save even more by disabling watcher types you do not need
		3895	and setting C<EV_MAXPRI> == C<EV_MINPRI>. Also, disabling C<assert>
		3896	(C<-DNDEBUG>) will usually reduce code size a lot. Disabling inotify,
		3897	eventfd and signalfd will further help, and disabling backends one doesn't
		3898	need (e.g. poll, epoll, kqueue, ports) will help further.
		3899
		3900	Defining C<EV_MINIMAL> to C<2> will additionally reduce the core API to
		3901	provide a bare-bones event library. See C<ev.h> for details on what parts
		3902	of the API are still available, and do not complain if this subset changes
		3903	over time.
		3904
		3905	This example set of settings reduces the compiled size of libev from 24Kb
		3906	to 8Kb on my GNU/Linux amd64 system (and leaves little in - there is also
		3907	an effect on the amount of memory used). With an intelligent-enough linker
		3908	further unused functions might be left out as well automatically.
		3909
		3910	// tuning and API changes
		3911	#define EV_MINIMAL 2
		3912	#define EV_MULTIPLICITY 0
		3913	#define EV_MINPRI 0
		3914	#define EV_MAXPRI 0
		3915
		3916	// OS-specific backends
		3917	#define EV_USE_INOTIFY 0
		3918	#define EV_USE_EVENTFD 0
		3919	#define EV_USE_SIGNALFD 0
		3920	#define EV_USE_REALTIME 0
		3921	#define EV_USE_MONOTONIC 0
		3922	#define EV_USE_CLOCK_SYSCALL 0
		3923
		3924	// disable all backends except select
		3925	#define EV_USE_POLL 0
		3926	#define EV_USE_PORT 0
		3927	#define EV_USE_KQUEUE 0
		3928	#define EV_USE_EPOLL 0
		3929
		3930	// disable all watcher types that cna be disabled
		3931	#define EV_STAT_ENABLE 0
		3932	#define EV_PERIODIC_ENABLE 0
		3933	#define EV_IDLE_ENABLE 0
		3934	#define EV_FORK_ENABLE 0
		3935	#define EV_SIGNAL_ENABLE 0
		3936	#define EV_CHILD_ENABLE 0
		3937	#define EV_ASYNC_ENABLE 0
		3938	#define EV_EMBED_ENABLE 0
		3939
		3940	=item EV_AVOID_STDIO
		3941
		3942	If this is set to C<1> at compiletime, then libev will avoid using stdio
		3943	functions (printf, scanf, perror etc.). This will increase the codesize
		3944	somewhat, but if your program doesn't otherwise depend on stdio and your
		3945	libc allows it, this avoids linking in the stdio library which is quite
		3946	big.
		3947
		3948	Note that error messages might become less precise when this option is
		3949	enabled.
		3950
		3951	=item EV_NSIG
		3952
		3953	The highest supported signal number, +1 (or, the number of
		3954	signals): Normally, libev tries to deduce the maximum number of signals
		3955	automatically, but sometimes this fails, in which case it can be
		3956	specified. Also, using a lower number than detected (C<32> should be
		3957	good for about any system in existance) can save some memory, as libev
		3958	statically allocates some 12-24 bytes per signal number.
3091		3959
3092	=item EV_PID_HASHSIZE	3960	=item EV_PID_HASHSIZE
3093		3961
3094	C<ev_child> watchers use a small hash table to distribute workload by	3962	C<ev_child> watchers use a small hash table to distribute workload by
3095	pid. The default size is C<16> (or C<1> with C<EV_MINIMAL>), usually more	3963	pid. The default size is C<16> (or C<1> with C<EV_MINIMAL>), usually more
…		…
3105	two).	3973	two).
3106		3974
3107	=item EV_USE_4HEAP	3975	=item EV_USE_4HEAP
3108		3976
3109	Heaps are not very cache-efficient. To improve the cache-efficiency of the	3977	Heaps are not very cache-efficient. To improve the cache-efficiency of the
3110	timer and periodics heap, libev uses a 4-heap when this symbol is defined	3978	timer and periodics heaps, libev uses a 4-heap when this symbol is defined
3111	to C<1>. The 4-heap uses more complicated (longer) code but has	3979	to C<1>. The 4-heap uses more complicated (longer) code but has noticeably
3112	noticeably faster performance with many (thousands) of watchers.	3980	faster performance with many (thousands) of watchers.
3113		3981
3114	The default is C<1> unless C<EV_MINIMAL> is set in which case it is C<0>	3982	The default is C<1> unless C<EV_MINIMAL> is set in which case it is C<0>
3115	(disabled).	3983	(disabled).
3116		3984
3117	=item EV_HEAP_CACHE_AT	3985	=item EV_HEAP_CACHE_AT
3118		3986
3119	Heaps are not very cache-efficient. To improve the cache-efficiency of the	3987	Heaps are not very cache-efficient. To improve the cache-efficiency of the
3120	timer and periodics heap, libev can cache the timestamp (I<at>) within	3988	timer and periodics heaps, libev can cache the timestamp (I<at>) within
3121	the heap structure (selected by defining C<EV_HEAP_CACHE_AT> to C<1>),	3989	the heap structure (selected by defining C<EV_HEAP_CACHE_AT> to C<1>),
3122	which uses 8-12 bytes more per watcher and a few hundred bytes more code,	3990	which uses 8-12 bytes more per watcher and a few hundred bytes more code,
3123	but avoids random read accesses on heap changes. This improves performance	3991	but avoids random read accesses on heap changes. This improves performance
3124	noticeably with with many (hundreds) of watchers.	3992	noticeably with many (hundreds) of watchers.
3125		3993
3126	The default is C<1> unless C<EV_MINIMAL> is set in which case it is C<0>	3994	The default is C<1> unless C<EV_MINIMAL> is set in which case it is C<0>
3127	(disabled).	3995	(disabled).
3128		3996
3129	=item EV_VERIFY	3997	=item EV_VERIFY
…		…
3135	called once per loop, which can slow down libev. If set to C<3>, then the	4003	called once per loop, which can slow down libev. If set to C<3>, then the
3136	verification code will be called very frequently, which will slow down	4004	verification code will be called very frequently, which will slow down
3137	libev considerably.	4005	libev considerably.
3138		4006
3139	The default is C<1>, unless C<EV_MINIMAL> is set, in which case it will be	4007	The default is C<1>, unless C<EV_MINIMAL> is set, in which case it will be
3140	C<0.>	4008	C<0>.
3141		4009
3142	=item EV_COMMON	4010	=item EV_COMMON
3143		4011
3144	By default, all watchers have a C<void *data> member. By redefining	4012	By default, all watchers have a C<void *data> member. By redefining
3145	this macro to a something else you can include more and other types of	4013	this macro to a something else you can include more and other types of
…		…
3162	and the way callbacks are invoked and set. Must expand to a struct member	4030	and the way callbacks are invoked and set. Must expand to a struct member
3163	definition and a statement, respectively. See the F<ev.h> header file for	4031	definition and a statement, respectively. See the F<ev.h> header file for
3164	their default definitions. One possible use for overriding these is to	4032	their default definitions. One possible use for overriding these is to
3165	avoid the C<struct ev_loop *> as first argument in all cases, or to use	4033	avoid the C<struct ev_loop *> as first argument in all cases, or to use
3166	method calls instead of plain function calls in C++.	4034	method calls instead of plain function calls in C++.
		4035
		4036	=back
3167		4037
3168	=head2 EXPORTED API SYMBOLS	4038	=head2 EXPORTED API SYMBOLS
3169		4039
3170	If you need to re-export the API (e.g. via a DLL) and you need a list of	4040	If you need to re-export the API (e.g. via a DLL) and you need a list of
3171	exported symbols, you can use the provided F<Symbol.*> files which list	4041	exported symbols, you can use the provided F<Symbol.*> files which list
…		…
3218	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:	4088	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:
3219		4089
3220	#include "ev_cpp.h"	4090	#include "ev_cpp.h"
3221	#include "ev.c"	4091	#include "ev.c"
3222		4092
		4093	=head1 INTERACTION WITH OTHER PROGRAMS OR LIBRARIES
3223		4094
3224	=head1 THREADS AND COROUTINES	4095	=head2 THREADS AND COROUTINES
3225		4096
3226	=head2 THREADS	4097	=head3 THREADS
3227		4098
3228	Libev itself is completely thread-safe, but it uses no locking. This	4099	All libev functions are reentrant and thread-safe unless explicitly
		4100	documented otherwise, but libev implements no locking itself. This means
3229	means that you can use as many loops as you want in parallel, as long as	4101	that you can use as many loops as you want in parallel, as long as there
3230	only one thread ever calls into one libev function with the same loop	4102	are no concurrent calls into any libev function with the same loop
3231	parameter.	4103	parameter (C<ev_default_*> calls have an implicit default loop parameter,
		4104	of course): libev guarantees that different event loops share no data
		4105	structures that need any locking.
3232		4106
3233	Or put differently: calls with different loop parameters can be done in	4107	Or to put it differently: calls with different loop parameters can be done
3234	parallel from multiple threads, calls with the same loop parameter must be	4108	concurrently from multiple threads, calls with the same loop parameter
3235	done serially (but can be done from different threads, as long as only one	4109	must be done serially (but can be done from different threads, as long as
3236	thread ever is inside a call at any point in time, e.g. by using a mutex	4110	only one thread ever is inside a call at any point in time, e.g. by using
3237	per loop).	4111	a mutex per loop).
		4112
		4113	Specifically to support threads (and signal handlers), libev implements
		4114	so-called C<ev_async> watchers, which allow some limited form of
		4115	concurrency on the same event loop, namely waking it up "from the
		4116	outside".
3238		4117
3239	If you want to know which design (one loop, locking, or multiple loops	4118	If you want to know which design (one loop, locking, or multiple loops
3240	without or something else still) is best for your problem, then I cannot	4119	without or something else still) is best for your problem, then I cannot
3241	help you. I can give some generic advice however:	4120	help you, but here is some generic advice:
3242		4121
3243	=over 4	4122	=over 4
3244		4123
3245	=item * most applications have a main thread: use the default libev loop	4124	=item * most applications have a main thread: use the default libev loop
3246	in that thread, or create a separate thread running only the default loop.	4125	in that thread, or create a separate thread running only the default loop.
…		…
3258		4137
3259	Choosing a model is hard - look around, learn, know that usually you can do	4138	Choosing a model is hard - look around, learn, know that usually you can do
3260	better than you currently do :-)	4139	better than you currently do :-)
3261		4140
3262	=item * often you need to talk to some other thread which blocks in the	4141	=item * often you need to talk to some other thread which blocks in the
		4142	event loop.
		4143
3263	event loop - C<ev_async> watchers can be used to wake them up from other	4144	C<ev_async> watchers can be used to wake them up from other threads safely
3264	threads safely (or from signal contexts...).	4145	(or from signal contexts...).
		4146
		4147	An example use would be to communicate signals or other events that only
		4148	work in the default loop by registering the signal watcher with the
		4149	default loop and triggering an C<ev_async> watcher from the default loop
		4150	watcher callback into the event loop interested in the signal.
3265		4151
3266	=back	4152	=back
3267		4153
		4154	=head4 THREAD LOCKING EXAMPLE
		4155
		4156	Here is a fictitious example of how to run an event loop in a different
		4157	thread than where callbacks are being invoked and watchers are
		4158	created/added/removed.
		4159
		4160	For a real-world example, see the C<EV::Loop::Async> perl module,
		4161	which uses exactly this technique (which is suited for many high-level
		4162	languages).
		4163
		4164	The example uses a pthread mutex to protect the loop data, a condition
		4165	variable to wait for callback invocations, an async watcher to notify the
		4166	event loop thread and an unspecified mechanism to wake up the main thread.
		4167
		4168	First, you need to associate some data with the event loop:
		4169
		4170	typedef struct {
		4171	mutex_t lock; /* global loop lock */
		4172	ev_async async_w;
		4173	thread_t tid;
		4174	cond_t invoke_cv;
		4175	} userdata;
		4176
		4177	void prepare_loop (EV_P)
		4178	{
		4179	// for simplicity, we use a static userdata struct.
		4180	static userdata u;
		4181
		4182	ev_async_init (&u->async_w, async_cb);
		4183	ev_async_start (EV_A_ &u->async_w);
		4184
		4185	pthread_mutex_init (&u->lock, 0);
		4186	pthread_cond_init (&u->invoke_cv, 0);
		4187
		4188	// now associate this with the loop
		4189	ev_set_userdata (EV_A_ u);
		4190	ev_set_invoke_pending_cb (EV_A_ l_invoke);
		4191	ev_set_loop_release_cb (EV_A_ l_release, l_acquire);
		4192
		4193	// then create the thread running ev_loop
		4194	pthread_create (&u->tid, 0, l_run, EV_A);
		4195	}
		4196
		4197	The callback for the C<ev_async> watcher does nothing: the watcher is used
		4198	solely to wake up the event loop so it takes notice of any new watchers
		4199	that might have been added:
		4200
		4201	static void
		4202	async_cb (EV_P_ ev_async *w, int revents)
		4203	{
		4204	// just used for the side effects
		4205	}
		4206
		4207	The C<l_release> and C<l_acquire> callbacks simply unlock/lock the mutex
		4208	protecting the loop data, respectively.
		4209
		4210	static void
		4211	l_release (EV_P)
		4212	{
		4213	userdata *u = ev_userdata (EV_A);
		4214	pthread_mutex_unlock (&u->lock);
		4215	}
		4216
		4217	static void
		4218	l_acquire (EV_P)
		4219	{
		4220	userdata *u = ev_userdata (EV_A);
		4221	pthread_mutex_lock (&u->lock);
		4222	}
		4223
		4224	The event loop thread first acquires the mutex, and then jumps straight
		4225	into C<ev_loop>:
		4226
		4227	void *
		4228	l_run (void *thr_arg)
		4229	{
		4230	struct ev_loop *loop = (struct ev_loop *)thr_arg;
		4231
		4232	l_acquire (EV_A);
		4233	pthread_setcanceltype (PTHREAD_CANCEL_ASYNCHRONOUS, 0);
		4234	ev_loop (EV_A_ 0);
		4235	l_release (EV_A);
		4236
		4237	return 0;
		4238	}
		4239
		4240	Instead of invoking all pending watchers, the C<l_invoke> callback will
		4241	signal the main thread via some unspecified mechanism (signals? pipe
		4242	writes? C<Async::Interrupt>?) and then waits until all pending watchers
		4243	have been called (in a while loop because a) spurious wakeups are possible
		4244	and b) skipping inter-thread-communication when there are no pending
		4245	watchers is very beneficial):
		4246
		4247	static void
		4248	l_invoke (EV_P)
		4249	{
		4250	userdata *u = ev_userdata (EV_A);
		4251
		4252	while (ev_pending_count (EV_A))
		4253	{
		4254	wake_up_other_thread_in_some_magic_or_not_so_magic_way ();
		4255	pthread_cond_wait (&u->invoke_cv, &u->lock);
		4256	}
		4257	}
		4258
		4259	Now, whenever the main thread gets told to invoke pending watchers, it
		4260	will grab the lock, call C<ev_invoke_pending> and then signal the loop
		4261	thread to continue:
		4262
		4263	static void
		4264	real_invoke_pending (EV_P)
		4265	{
		4266	userdata *u = ev_userdata (EV_A);
		4267
		4268	pthread_mutex_lock (&u->lock);
		4269	ev_invoke_pending (EV_A);
		4270	pthread_cond_signal (&u->invoke_cv);
		4271	pthread_mutex_unlock (&u->lock);
		4272	}
		4273
		4274	Whenever you want to start/stop a watcher or do other modifications to an
		4275	event loop, you will now have to lock:
		4276
		4277	ev_timer timeout_watcher;
		4278	userdata *u = ev_userdata (EV_A);
		4279
		4280	ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);
		4281
		4282	pthread_mutex_lock (&u->lock);
		4283	ev_timer_start (EV_A_ &timeout_watcher);
		4284	ev_async_send (EV_A_ &u->async_w);
		4285	pthread_mutex_unlock (&u->lock);
		4286
		4287	Note that sending the C<ev_async> watcher is required because otherwise
		4288	an event loop currently blocking in the kernel will have no knowledge
		4289	about the newly added timer. By waking up the loop it will pick up any new
		4290	watchers in the next event loop iteration.
		4291
3268	=head2 COROUTINES	4292	=head3 COROUTINES
3269		4293
3270	Libev is much more accommodating to coroutines ("cooperative threads"):	4294	Libev is very accommodating to coroutines ("cooperative threads"):
3271	libev fully supports nesting calls to it's functions from different	4295	libev fully supports nesting calls to its functions from different
3272	coroutines (e.g. you can call C<ev_loop> on the same loop from two	4296	coroutines (e.g. you can call C<ev_loop> on the same loop from two
3273	different coroutines and switch freely between both coroutines running the	4297	different coroutines, and switch freely between both coroutines running
3274	loop, as long as you don't confuse yourself). The only exception is that	4298	the loop, as long as you don't confuse yourself). The only exception is
3275	you must not do this from C<ev_periodic> reschedule callbacks.	4299	that you must not do this from C<ev_periodic> reschedule callbacks.
3276		4300
3277	Care has been invested into making sure that libev does not keep local	4301	Care has been taken to ensure that libev does not keep local state inside
3278	state inside C<ev_loop>, and other calls do not usually allow coroutine	4302	C<ev_loop>, and other calls do not usually allow for coroutine switches as
3279	switches.	4303	they do not call any callbacks.
3280		4304
		4305	=head2 COMPILER WARNINGS
3281		4306
3282	=head1 COMPLEXITIES	4307	Depending on your compiler and compiler settings, you might get no or a
		4308	lot of warnings when compiling libev code. Some people are apparently
		4309	scared by this.
3283		4310
3284	In this section the complexities of (many of) the algorithms used inside	4311	However, these are unavoidable for many reasons. For one, each compiler
3285	libev will be explained. For complexity discussions about backends see the	4312	has different warnings, and each user has different tastes regarding
3286	documentation for C<ev_default_init>.	4313	warning options. "Warn-free" code therefore cannot be a goal except when
		4314	targeting a specific compiler and compiler-version.
3287		4315
3288	All of the following are about amortised time: If an array needs to be	4316	Another reason is that some compiler warnings require elaborate
3289	extended, libev needs to realloc and move the whole array, but this	4317	workarounds, or other changes to the code that make it less clear and less
3290	happens asymptotically never with higher number of elements, so O(1) might	4318	maintainable.
3291	mean it might do a lengthy realloc operation in rare cases, but on average
3292	it is much faster and asymptotically approaches constant time.
3293		4319
3294	=over 4	4320	And of course, some compiler warnings are just plain stupid, or simply
		4321	wrong (because they don't actually warn about the condition their message
		4322	seems to warn about). For example, certain older gcc versions had some
		4323	warnings that resulted an extreme number of false positives. These have
		4324	been fixed, but some people still insist on making code warn-free with
		4325	such buggy versions.
3295		4326
3296	=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers)	4327	While libev is written to generate as few warnings as possible,
		4328	"warn-free" code is not a goal, and it is recommended not to build libev
		4329	with any compiler warnings enabled unless you are prepared to cope with
		4330	them (e.g. by ignoring them). Remember that warnings are just that:
		4331	warnings, not errors, or proof of bugs.
3297		4332
3298	This means that, when you have a watcher that triggers in one hour and
3299	there are 100 watchers that would trigger before that then inserting will
3300	have to skip roughly seven (C<ld 100>) of these watchers.
3301		4333
3302	=item Changing timer/periodic watchers (by autorepeat or calling again): O(log skipped_other_timers)	4334	=head2 VALGRIND
3303		4335
3304	That means that changing a timer costs less than removing/adding them	4336	Valgrind has a special section here because it is a popular tool that is
3305	as only the relative motion in the event queue has to be paid for.	4337	highly useful. Unfortunately, valgrind reports are very hard to interpret.
3306		4338
3307	=item Starting io/check/prepare/idle/signal/child/fork/async watchers: O(1)	4339	If you think you found a bug (memory leak, uninitialised data access etc.)
		4340	in libev, then check twice: If valgrind reports something like:
3308		4341
3309	These just add the watcher into an array or at the head of a list.	4342	==2274== definitely lost: 0 bytes in 0 blocks.
		4343	==2274== possibly lost: 0 bytes in 0 blocks.
		4344	==2274== still reachable: 256 bytes in 1 blocks.
3310		4345
3311	=item Stopping check/prepare/idle/fork/async watchers: O(1)	4346	Then there is no memory leak, just as memory accounted to global variables
		4347	is not a memleak - the memory is still being referenced, and didn't leak.
3312		4348
3313	=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % EV_PID_HASHSIZE))	4349	Similarly, under some circumstances, valgrind might report kernel bugs
		4350	as if it were a bug in libev (e.g. in realloc or in the poll backend,
		4351	although an acceptable workaround has been found here), or it might be
		4352	confused.
3314		4353
3315	These watchers are stored in lists then need to be walked to find the	4354	Keep in mind that valgrind is a very good tool, but only a tool. Don't
3316	correct watcher to remove. The lists are usually short (you don't usually	4355	make it into some kind of religion.
3317	have many watchers waiting for the same fd or signal).
3318		4356
3319	=item Finding the next timer in each loop iteration: O(1)	4357	If you are unsure about something, feel free to contact the mailing list
		4358	with the full valgrind report and an explanation on why you think this
		4359	is a bug in libev (best check the archives, too :). However, don't be
		4360	annoyed when you get a brisk "this is no bug" answer and take the chance
		4361	of learning how to interpret valgrind properly.
3320		4362
3321	By virtue of using a binary or 4-heap, the next timer is always found at a	4363	If you need, for some reason, empty reports from valgrind for your project
3322	fixed position in the storage array.	4364	I suggest using suppression lists.
3323		4365
3324	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)
3325		4366
3326	A change means an I/O watcher gets started or stopped, which requires	4367	=head1 PORTABILITY NOTES
3327	libev to recalculate its status (and possibly tell the kernel, depending
3328	on backend and whether C<ev_io_set> was used).
3329		4368
3330	=item Activating one watcher (putting it into the pending state): O(1)
3331
3332	=item Priority handling: O(number_of_priorities)
3333
3334	Priorities are implemented by allocating some space for each
3335	priority. When doing priority-based operations, libev usually has to
3336	linearly search all the priorities, but starting/stopping and activating
3337	watchers becomes O(1) w.r.t. priority handling.
3338
3339	=item Sending an ev_async: O(1)
3340
3341	=item Processing ev_async_send: O(number_of_async_watchers)
3342
3343	=item Processing signals: O(max_signal_number)
3344
3345	Sending involves a system call I<iff> there were no other C<ev_async_send>
3346	calls in the current loop iteration. Checking for async and signal events
3347	involves iterating over all running async watchers or all signal numbers.
3348
3349	=back
3350
3351
3352	=head1 WIN32 PLATFORM LIMITATIONS AND WORKAROUNDS	4369	=head2 WIN32 PLATFORM LIMITATIONS AND WORKAROUNDS
3353		4370
3354	Win32 doesn't support any of the standards (e.g. POSIX) that libev	4371	Win32 doesn't support any of the standards (e.g. POSIX) that libev
3355	requires, and its I/O model is fundamentally incompatible with the POSIX	4372	requires, and its I/O model is fundamentally incompatible with the POSIX
3356	model. Libev still offers limited functionality on this platform in	4373	model. Libev still offers limited functionality on this platform in
3357	the form of the C<EVBACKEND_SELECT> backend, and only supports socket	4374	the form of the C<EVBACKEND_SELECT> backend, and only supports socket
…		…
3364	way (note also that glib is the slowest event library known to man).	4381	way (note also that glib is the slowest event library known to man).
3365		4382
3366	There is no supported compilation method available on windows except	4383	There is no supported compilation method available on windows except
3367	embedding it into other applications.	4384	embedding it into other applications.
3368		4385
		4386	Sensible signal handling is officially unsupported by Microsoft - libev
		4387	tries its best, but under most conditions, signals will simply not work.
		4388
3369	Not a libev limitation but worth mentioning: windows apparently doesn't	4389	Not a libev limitation but worth mentioning: windows apparently doesn't
3370	accept large writes: instead of resulting in a partial write, windows will	4390	accept large writes: instead of resulting in a partial write, windows will
3371	either accept everything or return C<ENOBUFS> if the buffer is too large,	4391	either accept everything or return C<ENOBUFS> if the buffer is too large,
3372	so make sure you only write small amounts into your sockets (less than a	4392	so make sure you only write small amounts into your sockets (less than a
3373	megabyte seems safe, but thsi apparently depends on the amount of memory	4393	megabyte seems safe, but this apparently depends on the amount of memory
3374	available).	4394	available).
3375		4395
3376	Due to the many, low, and arbitrary limits on the win32 platform and	4396	Due to the many, low, and arbitrary limits on the win32 platform and
3377	the abysmal performance of winsockets, using a large number of sockets	4397	the abysmal performance of winsockets, using a large number of sockets
3378	is not recommended (and not reasonable). If your program needs to use	4398	is not recommended (and not reasonable). If your program needs to use
3379	more than a hundred or so sockets, then likely it needs to use a totally	4399	more than a hundred or so sockets, then likely it needs to use a totally
3380	different implementation for windows, as libev offers the POSIX readiness	4400	different implementation for windows, as libev offers the POSIX readiness
3381	notification model, which cannot be implemented efficiently on windows	4401	notification model, which cannot be implemented efficiently on windows
3382	(Microsoft monopoly games).	4402	(due to Microsoft monopoly games).
3383		4403
3384	A typical way to use libev under windows is to embed it (see the embedding	4404	A typical way to use libev under windows is to embed it (see the embedding
3385	section for details) and use the following F<evwrap.h> header file instead	4405	section for details) and use the following F<evwrap.h> header file instead
3386	of F<ev.h>:	4406	of F<ev.h>:
3387		4407
…		…
3389	#define EV_SELECT_IS_WINSOCKET 1 /* configure libev for windows select */	4409	#define EV_SELECT_IS_WINSOCKET 1 /* configure libev for windows select */
3390		4410
3391	#include "ev.h"	4411	#include "ev.h"
3392		4412
3393	And compile the following F<evwrap.c> file into your project (make sure	4413	And compile the following F<evwrap.c> file into your project (make sure
3394	you do I<not> compile the F<ev.c> or any other embedded soruce files!):	4414	you do I<not> compile the F<ev.c> or any other embedded source files!):
3395		4415
3396	#include "evwrap.h"	4416	#include "evwrap.h"
3397	#include "ev.c"	4417	#include "ev.c"
3398		4418
3399	=over 4	4419	=over 4
…		…
3423		4443
3424	Early versions of winsocket's select only supported waiting for a maximum	4444	Early versions of winsocket's select only supported waiting for a maximum
3425	of C<64> handles (probably owning to the fact that all windows kernels	4445	of C<64> handles (probably owning to the fact that all windows kernels
3426	can only wait for C<64> things at the same time internally; Microsoft	4446	can only wait for C<64> things at the same time internally; Microsoft
3427	recommends spawning a chain of threads and wait for 63 handles and the	4447	recommends spawning a chain of threads and wait for 63 handles and the
3428	previous thread in each. Great).	4448	previous thread in each. Sounds great!).
3429		4449
3430	Newer versions support more handles, but you need to define C<FD_SETSIZE>	4450	Newer versions support more handles, but you need to define C<FD_SETSIZE>
3431	to some high number (e.g. C<2048>) before compiling the winsocket select	4451	to some high number (e.g. C<2048>) before compiling the winsocket select
3432	call (which might be in libev or elsewhere, for example, perl does its own	4452	call (which might be in libev or elsewhere, for example, perl and many
3433	select emulation on windows).	4453	other interpreters do their own select emulation on windows).
3434		4454
3435	Another limit is the number of file descriptors in the Microsoft runtime	4455	Another limit is the number of file descriptors in the Microsoft runtime
3436	libraries, which by default is C<64> (there must be a hidden I<64> fetish	4456	libraries, which by default is C<64> (there must be a hidden I<64>
3437	or something like this inside Microsoft). You can increase this by calling	4457	fetish or something like this inside Microsoft). You can increase this
3438	C<_setmaxstdio>, which can increase this limit to C<2048> (another	4458	by calling C<_setmaxstdio>, which can increase this limit to C<2048>
3439	arbitrary limit), but is broken in many versions of the Microsoft runtime	4459	(another arbitrary limit), but is broken in many versions of the Microsoft
3440	libraries.
3441
3442	This might get you to about C<512> or C<2048> sockets (depending on	4460	runtime libraries. This might get you to about C<512> or C<2048> sockets
3443	windows version and/or the phase of the moon). To get more, you need to	4461	(depending on windows version and/or the phase of the moon). To get more,
3444	wrap all I/O functions and provide your own fd management, but the cost of	4462	you need to wrap all I/O functions and provide your own fd management, but
3445	calling select (O(n²)) will likely make this unworkable.	4463	the cost of calling select (O(n²)) will likely make this unworkable.
3446		4464
3447	=back	4465	=back
3448		4466
3449
3450	=head1 PORTABILITY REQUIREMENTS	4467	=head2 PORTABILITY REQUIREMENTS
3451		4468
3452	In addition to a working ISO-C implementation, libev relies on a few	4469	In addition to a working ISO-C implementation and of course the
3453	additional extensions:	4470	backend-specific APIs, libev relies on a few additional extensions:
3454		4471
3455	=over 4	4472	=over 4
3456		4473
3457	=item C<void (*)(ev_watcher_type *, int revents)> must have compatible	4474	=item C<void (*)(ev_watcher_type *, int revents)> must have compatible
3458	calling conventions regardless of C<ev_watcher_type *>.	4475	calling conventions regardless of C<ev_watcher_type *>.
…		…
3464	calls them using an C<ev_watcher *> internally.	4481	calls them using an C<ev_watcher *> internally.
3465		4482
3466	=item C<sig_atomic_t volatile> must be thread-atomic as well	4483	=item C<sig_atomic_t volatile> must be thread-atomic as well
3467		4484
3468	The type C<sig_atomic_t volatile> (or whatever is defined as	4485	The type C<sig_atomic_t volatile> (or whatever is defined as
3469	C<EV_ATOMIC_T>) must be atomic w.r.t. accesses from different	4486	C<EV_ATOMIC_T>) must be atomic with respect to accesses from different
3470	threads. This is not part of the specification for C<sig_atomic_t>, but is	4487	threads. This is not part of the specification for C<sig_atomic_t>, but is
3471	believed to be sufficiently portable.	4488	believed to be sufficiently portable.
3472		4489
3473	=item C<sigprocmask> must work in a threaded environment	4490	=item C<sigprocmask> must work in a threaded environment
3474		4491
…		…
3483	except the initial one, and run the default loop in the initial thread as	4500	except the initial one, and run the default loop in the initial thread as
3484	well.	4501	well.
3485		4502
3486	=item C<long> must be large enough for common memory allocation sizes	4503	=item C<long> must be large enough for common memory allocation sizes
3487		4504
3488	To improve portability and simplify using libev, libev uses C<long>	4505	To improve portability and simplify its API, libev uses C<long> internally
3489	internally instead of C<size_t> when allocating its data structures. On	4506	instead of C<size_t> when allocating its data structures. On non-POSIX
3490	non-POSIX systems (Microsoft...) this might be unexpectedly low, but	4507	systems (Microsoft...) this might be unexpectedly low, but is still at
3491	is still at least 31 bits everywhere, which is enough for hundreds of	4508	least 31 bits everywhere, which is enough for hundreds of millions of
3492	millions of watchers.	4509	watchers.
3493		4510
3494	=item C<double> must hold a time value in seconds with enough accuracy	4511	=item C<double> must hold a time value in seconds with enough accuracy
3495		4512
3496	The type C<double> is used to represent timestamps. It is required to	4513	The type C<double> is used to represent timestamps. It is required to
3497	have at least 51 bits of mantissa (and 9 bits of exponent), which is good	4514	have at least 51 bits of mantissa (and 9 bits of exponent), which is good
3498	enough for at least into the year 4000. This requirement is fulfilled by	4515	enough for at least into the year 4000. This requirement is fulfilled by
3499	implementations implementing IEEE 754 (basically all existing ones).	4516	implementations implementing IEEE 754, which is basically all existing
		4517	ones. With IEEE 754 doubles, you get microsecond accuracy until at least
		4518	2200.
3500		4519
3501	=back	4520	=back
3502		4521
3503	If you know of other additional requirements drop me a note.	4522	If you know of other additional requirements drop me a note.
3504		4523
3505		4524
3506	=head1 COMPILER WARNINGS	4525	=head1 ALGORITHMIC COMPLEXITIES
3507		4526
3508	Depending on your compiler and compiler settings, you might get no or a	4527	In this section the complexities of (many of) the algorithms used inside
3509	lot of warnings when compiling libev code. Some people are apparently	4528	libev will be documented. For complexity discussions about backends see
3510	scared by this.	4529	the documentation for C<ev_default_init>.
3511		4530
3512	However, these are unavoidable for many reasons. For one, each compiler	4531	All of the following are about amortised time: If an array needs to be
3513	has different warnings, and each user has different tastes regarding	4532	extended, libev needs to realloc and move the whole array, but this
3514	warning options. "Warn-free" code therefore cannot be a goal except when	4533	happens asymptotically rarer with higher number of elements, so O(1) might
3515	targeting a specific compiler and compiler-version.	4534	mean that libev does a lengthy realloc operation in rare cases, but on
		4535	average it is much faster and asymptotically approaches constant time.
3516		4536
3517	Another reason is that some compiler warnings require elaborate	4537	=over 4
3518	workarounds, or other changes to the code that make it less clear and less
3519	maintainable.
3520		4538
3521	And of course, some compiler warnings are just plain stupid, or simply	4539	=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers)
3522	wrong (because they don't actually warn about the condition their message
3523	seems to warn about).
3524		4540
3525	While libev is written to generate as few warnings as possible,	4541	This means that, when you have a watcher that triggers in one hour and
3526	"warn-free" code is not a goal, and it is recommended not to build libev	4542	there are 100 watchers that would trigger before that, then inserting will
3527	with any compiler warnings enabled unless you are prepared to cope with	4543	have to skip roughly seven (C<ld 100>) of these watchers.
3528	them (e.g. by ignoring them). Remember that warnings are just that:
3529	warnings, not errors, or proof of bugs.
3530		4544
		4545	=item Changing timer/periodic watchers (by autorepeat or calling again): O(log skipped_other_timers)
3531		4546
3532	=head1 VALGRIND	4547	That means that changing a timer costs less than removing/adding them,
		4548	as only the relative motion in the event queue has to be paid for.
3533		4549
3534	Valgrind has a special section here because it is a popular tool that is	4550	=item Starting io/check/prepare/idle/signal/child/fork/async watchers: O(1)
3535	highly useful, but valgrind reports are very hard to interpret.
3536		4551
3537	If you think you found a bug (memory leak, uninitialised data access etc.)	4552	These just add the watcher into an array or at the head of a list.
3538	in libev, then check twice: If valgrind reports something like:
3539		4553
3540	==2274== definitely lost: 0 bytes in 0 blocks.	4554	=item Stopping check/prepare/idle/fork/async watchers: O(1)
3541	==2274== possibly lost: 0 bytes in 0 blocks.
3542	==2274== still reachable: 256 bytes in 1 blocks.
3543		4555
3544	Then there is no memory leak. Similarly, under some circumstances,	4556	=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % EV_PID_HASHSIZE))
3545	valgrind might report kernel bugs as if it were a bug in libev, or it
3546	might be confused (it is a very good tool, but only a tool).
3547		4557
3548	If you are unsure about something, feel free to contact the mailing list	4558	These watchers are stored in lists, so they need to be walked to find the
3549	with the full valgrind report and an explanation on why you think this is	4559	correct watcher to remove. The lists are usually short (you don't usually
3550	a bug in libev. However, don't be annoyed when you get a brisk "this is	4560	have many watchers waiting for the same fd or signal: one is typical, two
3551	no bug" answer and take the chance of learning how to interpret valgrind	4561	is rare).
3552	properly.
3553		4562
3554	If you need, for some reason, empty reports from valgrind for your project	4563	=item Finding the next timer in each loop iteration: O(1)
3555	I suggest using suppression lists.
3556		4564
		4565	By virtue of using a binary or 4-heap, the next timer is always found at a
		4566	fixed position in the storage array.
		4567
		4568	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)
		4569
		4570	A change means an I/O watcher gets started or stopped, which requires
		4571	libev to recalculate its status (and possibly tell the kernel, depending
		4572	on backend and whether C<ev_io_set> was used).
		4573
		4574	=item Activating one watcher (putting it into the pending state): O(1)
		4575
		4576	=item Priority handling: O(number_of_priorities)
		4577
		4578	Priorities are implemented by allocating some space for each
		4579	priority. When doing priority-based operations, libev usually has to
		4580	linearly search all the priorities, but starting/stopping and activating
		4581	watchers becomes O(1) with respect to priority handling.
		4582
		4583	=item Sending an ev_async: O(1)
		4584
		4585	=item Processing ev_async_send: O(number_of_async_watchers)
		4586
		4587	=item Processing signals: O(max_signal_number)
		4588
		4589	Sending involves a system call I<iff> there were no other C<ev_async_send>
		4590	calls in the current loop iteration. Checking for async and signal events
		4591	involves iterating over all running async watchers or all signal numbers.
		4592
		4593	=back
		4594
		4595
		4596	=head1 GLOSSARY
		4597
		4598	=over 4
		4599
		4600	=item active
		4601
		4602	A watcher is active as long as it has been started (has been attached to
		4603	an event loop) but not yet stopped (disassociated from the event loop).
		4604
		4605	=item application
		4606
		4607	In this document, an application is whatever is using libev.
		4608
		4609	=item callback
		4610
		4611	The address of a function that is called when some event has been
		4612	detected. Callbacks are being passed the event loop, the watcher that
		4613	received the event, and the actual event bitset.
		4614
		4615	=item callback invocation
		4616
		4617	The act of calling the callback associated with a watcher.
		4618
		4619	=item event
		4620
		4621	A change of state of some external event, such as data now being available
		4622	for reading on a file descriptor, time having passed or simply not having
		4623	any other events happening anymore.
		4624
		4625	In libev, events are represented as single bits (such as C<EV_READ> or
		4626	C<EV_TIMEOUT>).
		4627
		4628	=item event library
		4629
		4630	A software package implementing an event model and loop.
		4631
		4632	=item event loop
		4633
		4634	An entity that handles and processes external events and converts them
		4635	into callback invocations.
		4636
		4637	=item event model
		4638
		4639	The model used to describe how an event loop handles and processes
		4640	watchers and events.
		4641
		4642	=item pending
		4643
		4644	A watcher is pending as soon as the corresponding event has been detected,
		4645	and stops being pending as soon as the watcher will be invoked or its
		4646	pending status is explicitly cleared by the application.
		4647
		4648	A watcher can be pending, but not active. Stopping a watcher also clears
		4649	its pending status.
		4650
		4651	=item real time
		4652
		4653	The physical time that is observed. It is apparently strictly monotonic :)
		4654
		4655	=item wall-clock time
		4656
		4657	The time and date as shown on clocks. Unlike real time, it can actually
		4658	be wrong and jump forwards and backwards, e.g. when the you adjust your
		4659	clock.
		4660
		4661	=item watcher
		4662
		4663	A data structure that describes interest in certain events. Watchers need
		4664	to be started (attached to an event loop) before they can receive events.
		4665
		4666	=item watcher invocation
		4667
		4668	The act of calling the callback associated with a watcher.
		4669
		4670	=back
3557		4671
3558	=head1 AUTHOR	4672	=head1 AUTHOR
3559		4673
3560	Marc Lehmann <libev@schmorp.de>.	4674	Marc Lehmann <libev@schmorp.de>, with repeated corrections by Mikael Magnusson.
3561		4675

Diff Legend

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