ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/libev/ev.pod

Comparing libev/ev.pod (file contents):
Revision 1.175 by root, Mon Sep 8 16:36:14 2008 UTC vs.
Revision 1.277 by root, Thu Dec 31 06:50:17 2009 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 is both faster and might make
		380	it possible to get the queued signal data.
		381
		382	Signalfd will not be used by default as this changes your signal mask, and
		383	there are a lot of shoddy libraries and programs (glib's threadpool for
		384	example) that can't properly initialise their signal masks.
		385
349	=item C<EVBACKEND_SELECT> (value 1, portable select backend)	386	=item C<EVBACKEND_SELECT> (value 1, portable select backend)
350		387
351	This is your standard select(2) backend. Not I<completely> standard, as	388	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,	389	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	390	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	396	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	397	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	398	a look at C<ev_set_io_collect_interval ()> to increase the amount of
362	readiness notifications you get per iteration.	399	readiness notifications you get per iteration.
363		400
		401	This backend maps C<EV_READ> to the C<readfds> set and C<EV_WRITE> to the
		402	C<writefds> set (and to work around Microsoft Windows bugs, also onto the
		403	C<exceptfds> set on that platform).
		404
364	=item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows)	405	=item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows)
365		406
366	And this is your standard poll(2) backend. It's more complicated	407	And this is your standard poll(2) backend. It's more complicated
367	than select, but handles sparse fds better and has no artificial	408	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	409	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,	410	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	411	i.e. O(total_fds). See the entry for C<EVBACKEND_SELECT>, above, for
371	performance tips.	412	performance tips.
372		413
		414	This backend maps C<EV_READ> to C<POLLIN | POLLERR | POLLHUP>, and
		415	C<EV_WRITE> to C<POLLOUT | POLLERR | POLLHUP>.
		416
373	=item C<EVBACKEND_EPOLL> (value 4, Linux)	417	=item C<EVBACKEND_EPOLL> (value 4, Linux)
		418
		419	Use the linux-specific epoll(7) interface (for both pre- and post-2.6.9
		420	kernels).
374		421
375	For few fds, this backend is a bit little slower than poll and select,	422	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	423	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),	424	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	425	epoll scales either O(1) or O(active_fds).
379	of shortcomings, such as silently dropping events in some hard-to-detect	426
380	cases and requiring a system call per fd change, no fork support and bad	427	The epoll mechanism deserves honorable mention as the most misdesigned
381	support for dup.	428	of the more advanced event mechanisms: mere annoyances include silently
		429	dropping file descriptors, requiring a system call per change per file
		430	descriptor (and unnecessary guessing of parameters), problems with dup and
		431	so on. The biggest issue is fork races, however - if a program forks then
		432	I<both> parent and child process have to recreate the epoll set, which can
		433	take considerable time (one syscall per file descriptor) and is of course
		434	hard to detect.
		435
		436	Epoll is also notoriously buggy - embedding epoll fds I<should> work, but
		437	of course I<doesn't>, and epoll just loves to report events for totally
		438	I<different> file descriptors (even already closed ones, so one cannot
		439	even remove them from the set) than registered in the set (especially
		440	on SMP systems). Libev tries to counter these spurious notifications by
		441	employing an additional generation counter and comparing that against the
		442	events to filter out spurious ones, recreating the set when required.
382		443
383	While stopping, setting and starting an I/O watcher in the same iteration	444	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	445	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	446	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	447	I<file description> now), so its best to avoid that. Also, C<dup ()>'ed
387	very well if you register events for both fds.	448	file descriptors might not work very well if you register events for both
388		449	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		450
393	Best performance from this backend is achieved by not unregistering all	451	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.	452	watchers for a file descriptor until it has been closed, if possible,
395	keep at least one watcher active per fd at all times.	453	i.e. keep at least one watcher active per fd at all times. Stopping and
		454	starting a watcher (without re-setting it) also usually doesn't cause
		455	extra overhead. A fork can both result in spurious notifications as well
		456	as in libev having to destroy and recreate the epoll object, which can
		457	take considerable time and thus should be avoided.
		458
		459	All this means that, in practice, C<EVBACKEND_SELECT> can be as fast or
		460	faster than epoll for maybe up to a hundred file descriptors, depending on
		461	the usage. So sad.
396		462
397	While nominally embeddable in other event loops, this feature is broken in	463	While nominally embeddable in other event loops, this feature is broken in
398	all kernel versions tested so far.	464	all kernel versions tested so far.
		465
		466	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
		467	C<EVBACKEND_POLL>.
399		468
400	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)	469	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)
401		470
402	Kqueue deserves special mention, as at the time of this writing, it	471	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	472	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	473	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"	474	it's completely useless). Unlike epoll, however, whose brokenness
		475	is by design, these kqueue bugs can (and eventually will) be fixed
		476	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	477	"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)	478	C<EVBACKEND_KQUEUE>) or libev was compiled on a known-to-be-good (-enough)
408	system like NetBSD.	479	system like NetBSD.
409		480
410	You still can embed kqueue into a normal poll or select backend and use it	481	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	482	only for sockets (after having made sure that sockets work with kqueue on
…		…
413		484
414	It scales in the same way as the epoll backend, but the interface to the	485	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	486	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	487	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	488	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	489	two event changes per incident. Support for C<fork ()> is very bad (but
419	drops fds silently in similarly hard-to-detect cases.	490	sane, unlike epoll) and it drops fds silently in similarly hard-to-detect
		491	cases
420		492
421	This backend usually performs well under most conditions.	493	This backend usually performs well under most conditions.
422		494
423	While nominally embeddable in other event loops, this doesn't work	495	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	496	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	497	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	498	(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	499	(e.g. C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> (but C<poll> is of course
428	sockets.	500	also broken on OS X)) and, did I mention it, using it only for sockets.
		501
		502	This backend maps C<EV_READ> into an C<EVFILT_READ> kevent with
		503	C<NOTE_EOF>, and C<EV_WRITE> into an C<EVFILT_WRITE> kevent with
		504	C<NOTE_EOF>.
429		505
430	=item C<EVBACKEND_DEVPOLL> (value 16, Solaris 8)	506	=item C<EVBACKEND_DEVPOLL> (value 16, Solaris 8)
431		507
432	This is not implemented yet (and might never be, unless you send me an	508	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	509	implementation). According to reports, C</dev/poll> only supports sockets
…		…
446	While this backend scales well, it requires one system call per active	522	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	523	file descriptor per loop iteration. For small and medium numbers of file
448	descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend	524	descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend
449	might perform better.	525	might perform better.
450		526
451	On the positive side, ignoring the spurious readiness notifications, this	527	On the positive side, with the exception of the spurious readiness
452	backend actually performed to specification in all tests and is fully	528	notifications, this backend actually performed fully to specification
453	embeddable, which is a rare feat among the OS-specific backends.	529	in all tests and is fully embeddable, which is a rare feat among the
		530	OS-specific backends (I vastly prefer correctness over speed hacks).
		531
		532	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
		533	C<EVBACKEND_POLL>.
454		534
455	=item C<EVBACKEND_ALL>	535	=item C<EVBACKEND_ALL>
456		536
457	Try all backends (even potentially broken ones that wouldn't be tried	537	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	538	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as
…		…
460		540
461	It is definitely not recommended to use this flag.	541	It is definitely not recommended to use this flag.
462		542
463	=back	543	=back
464		544
465	If one or more of these are or'ed into the flags value, then only these	545	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	546	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.	547	here). If none are specified, all backends in C<ev_recommended_backends
		548	()> will be tried.
468		549
469	The most typical usage is like this:	550	Example: This is the most typical usage.
470		551
471	if (!ev_default_loop (0))	552	if (!ev_default_loop (0))
472	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");	553	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
473		554
474	Restrict libev to the select and poll backends, and do not allow	555	Example: Restrict libev to the select and poll backends, and do not allow
475	environment settings to be taken into account:	556	environment settings to be taken into account:
476		557
477	ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);	558	ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);
478		559
479	Use whatever libev has to offer, but make sure that kqueue is used if	560	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	561	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):	562	private event loop and only if you know the OS supports your types of
		563	fds):
482		564
483	ev_default_loop (ev_recommended_backends () | EVBACKEND_KQUEUE);	565	ev_default_loop (ev_recommended_backends () | EVBACKEND_KQUEUE);
484		566
485	=item struct ev_loop *ev_loop_new (unsigned int flags)	567	=item struct ev_loop *ev_loop_new (unsigned int flags)
486		568
…		…
507	responsibility to either stop all watchers cleanly yourself I<before>	589	responsibility to either stop all watchers cleanly yourself I<before>
508	calling this function, or cope with the fact afterwards (which is usually	590	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	591	the easiest thing, you can just ignore the watchers and/or C<free ()> them
510	for example).	592	for example).
511		593
512	Note that certain global state, such as signal state, will not be freed by	594	Note that certain global state, such as signal state (and installed signal
513	this function, and related watchers (such as signal and child watchers)	595	handlers), will not be freed by this function, and related watchers (such
514	would need to be stopped manually.	596	as signal and child watchers) would need to be stopped manually.
515		597
516	In general it is not advisable to call this function except in the	598	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	599	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	600	pipe fds. If you need dynamically allocated loops it is better to use
519	C<ev_loop_new> and C<ev_loop_destroy>).	601	C<ev_loop_new> and C<ev_loop_destroy>.
520		602
521	=item ev_loop_destroy (loop)	603	=item ev_loop_destroy (loop)
522		604
523	Like C<ev_default_destroy>, but destroys an event loop created by an	605	Like C<ev_default_destroy>, but destroys an event loop created by an
524	earlier call to C<ev_loop_new>.	606	earlier call to C<ev_loop_new>.
…		…
544		626
545	=item ev_loop_fork (loop)	627	=item ev_loop_fork (loop)
546		628
547	Like C<ev_default_fork>, but acts on an event loop created by	629	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	630	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.	631	after fork that you want to re-use in the child, and how you do this is
		632	entirely your own problem.
550		633
551	=item int ev_is_default_loop (loop)	634	=item int ev_is_default_loop (loop)
552		635
553	Returns true when the given loop actually is the default loop, false otherwise.	636	Returns true when the given loop is, in fact, the default loop, and false
		637	otherwise.
554		638
555	=item unsigned int ev_loop_count (loop)	639	=item unsigned int ev_loop_count (loop)
556		640
557	Returns the count of loop iterations for the loop, which is identical to	641	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	642	the number of times libev did poll for new events. It starts at C<0> and
559	happily wraps around with enough iterations.	643	happily wraps around with enough iterations.
560		644
561	This value can sometimes be useful as a generation counter of sorts (it	645	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	646	"ticks" the number of loop iterations), as it roughly corresponds with
563	C<ev_prepare> and C<ev_check> calls.	647	C<ev_prepare> and C<ev_check> calls.
		648
		649	=item unsigned int ev_loop_depth (loop)
		650
		651	Returns the number of times C<ev_loop> was entered minus the number of
		652	times C<ev_loop> was exited, in other words, the recursion depth.
		653
		654	Outside C<ev_loop>, this number is zero. In a callback, this number is
		655	C<1>, unless C<ev_loop> was invoked recursively (or from another thread),
		656	in which case it is higher.
		657
		658	Leaving C<ev_loop> abnormally (setjmp/longjmp, cancelling the thread
		659	etc.), doesn't count as exit.
564		660
565	=item unsigned int ev_backend (loop)	661	=item unsigned int ev_backend (loop)
566		662
567	Returns one of the C<EVBACKEND_*> flags indicating the event backend in	663	Returns one of the C<EVBACKEND_*> flags indicating the event backend in
568	use.	664	use.
…		…
573	received events and started processing them. This timestamp does not	669	received events and started processing them. This timestamp does not
574	change as long as callbacks are being processed, and this is also the base	670	change as long as callbacks are being processed, and this is also the base
575	time used for relative timers. You can treat it as the timestamp of the	671	time used for relative timers. You can treat it as the timestamp of the
576	event occurring (or more correctly, libev finding out about it).	672	event occurring (or more correctly, libev finding out about it).
577		673
		674	=item ev_now_update (loop)
		675
		676	Establishes the current time by querying the kernel, updating the time
		677	returned by C<ev_now ()> in the progress. This is a costly operation and
		678	is usually done automatically within C<ev_loop ()>.
		679
		680	This function is rarely useful, but when some event callback runs for a
		681	very long time without entering the event loop, updating libev's idea of
		682	the current time is a good idea.
		683
		684	See also L<The special problem of time updates> in the C<ev_timer> section.
		685
		686	=item ev_suspend (loop)
		687
		688	=item ev_resume (loop)
		689
		690	These two functions suspend and resume a loop, for use when the loop is
		691	not used for a while and timeouts should not be processed.
		692
		693	A typical use case would be an interactive program such as a game: When
		694	the user presses C<^Z> to suspend the game and resumes it an hour later it
		695	would be best to handle timeouts as if no time had actually passed while
		696	the program was suspended. This can be achieved by calling C<ev_suspend>
		697	in your C<SIGTSTP> handler, sending yourself a C<SIGSTOP> and calling
		698	C<ev_resume> directly afterwards to resume timer processing.
		699
		700	Effectively, all C<ev_timer> watchers will be delayed by the time spend
		701	between C<ev_suspend> and C<ev_resume>, and all C<ev_periodic> watchers
		702	will be rescheduled (that is, they will lose any events that would have
		703	occured while suspended).
		704
		705	After calling C<ev_suspend> you B<must not> call I<any> function on the
		706	given loop other than C<ev_resume>, and you B<must not> call C<ev_resume>
		707	without a previous call to C<ev_suspend>.
		708
		709	Calling C<ev_suspend>/C<ev_resume> has the side effect of updating the
		710	event loop time (see C<ev_now_update>).
		711
578	=item ev_loop (loop, int flags)	712	=item ev_loop (loop, int flags)
579		713
580	Finally, this is it, the event handler. This function usually is called	714	Finally, this is it, the event handler. This function usually is called
581	after you initialised all your watchers and you want to start handling	715	after you have initialised all your watchers and you want to start
582	events.	716	handling events.
583		717
584	If the flags argument is specified as C<0>, it will not return until	718	If the flags argument is specified as C<0>, it will not return until
585	either no event watchers are active anymore or C<ev_unloop> was called.	719	either no event watchers are active anymore or C<ev_unloop> was called.
586		720
587	Please note that an explicit C<ev_unloop> is usually better than	721	Please note that an explicit C<ev_unloop> is usually better than
588	relying on all watchers to be stopped when deciding when a program has	722	relying on all watchers to be stopped when deciding when a program has
589	finished (especially in interactive programs), but having a program that	723	finished (especially in interactive programs), but having a program
590	automatically loops as long as it has to and no longer by virtue of	724	that automatically loops as long as it has to and no longer by virtue
591	relying on its watchers stopping correctly is a thing of beauty.	725	of relying on its watchers stopping correctly, that is truly a thing of
		726	beauty.
592		727
593	A flags value of C<EVLOOP_NONBLOCK> will look for new events, will handle	728	A flags value of C<EVLOOP_NONBLOCK> will look for new events, will handle
594	those events and any outstanding ones, but will not block your process in	729	those events and any already outstanding ones, but will not block your
595	case there are no events and will return after one iteration of the loop.	730	process in case there are no events and will return after one iteration of
		731	the loop.
596		732
597	A flags value of C<EVLOOP_ONESHOT> will look for new events (waiting if	733	A flags value of C<EVLOOP_ONESHOT> will look for new events (waiting if
598	necessary) and will handle those and any outstanding ones. It will block	734	necessary) and will handle those and any already outstanding ones. It
599	your process until at least one new event arrives, and will return after	735	will block your process until at least one new event arrives (which could
600	one iteration of the loop. This is useful if you are waiting for some	736	be an event internal to libev itself, so there is no guarantee that a
601	external event in conjunction with something not expressible using other	737	user-registered callback will be called), and will return after one
		738	iteration of the loop.
		739
		740	This is useful if you are waiting for some external event in conjunction
		741	with something not expressible using other libev watchers (i.e. "roll your
602	libev watchers. However, a pair of C<ev_prepare>/C<ev_check> watchers is	742	own C<ev_loop>"). However, a pair of C<ev_prepare>/C<ev_check> watchers is
603	usually a better approach for this kind of thing.	743	usually a better approach for this kind of thing.
604		744
605	Here are the gory details of what C<ev_loop> does:	745	Here are the gory details of what C<ev_loop> does:
606		746
607	- Before the first iteration, call any pending watchers.	747	- Before the first iteration, call any pending watchers.
…		…
617	any active watchers at all will result in not sleeping).	757	any active watchers at all will result in not sleeping).
618	- Sleep if the I/O and timer collect interval say so.	758	- Sleep if the I/O and timer collect interval say so.
619	- Block the process, waiting for any events.	759	- Block the process, waiting for any events.
620	- Queue all outstanding I/O (fd) events.	760	- Queue all outstanding I/O (fd) events.
621	- Update the "event loop time" (ev_now ()), and do time jump adjustments.	761	- Update the "event loop time" (ev_now ()), and do time jump adjustments.
622	- Queue all outstanding timers.	762	- Queue all expired timers.
623	- Queue all outstanding periodics.	763	- Queue all expired periodics.
624	- Unless any events are pending now, queue all idle watchers.	764	- Unless any events are pending now, queue all idle watchers.
625	- Queue all check watchers.	765	- Queue all check watchers.
626	- Call all queued watchers in reverse order (i.e. check watchers first).	766	- Call all queued watchers in reverse order (i.e. check watchers first).
627	Signals and child watchers are implemented as I/O watchers, and will	767	Signals and child watchers are implemented as I/O watchers, and will
628	be handled here by queueing them when their watcher gets executed.	768	be handled here by queueing them when their watcher gets executed.
…		…
645	C<EVUNLOOP_ONE>, which will make the innermost C<ev_loop> call return, or	785	C<EVUNLOOP_ONE>, which will make the innermost C<ev_loop> call return, or
646	C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.	786	C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.
647		787
648	This "unloop state" will be cleared when entering C<ev_loop> again.	788	This "unloop state" will be cleared when entering C<ev_loop> again.
649		789
		790	It is safe to call C<ev_unloop> from otuside any C<ev_loop> calls.
		791
650	=item ev_ref (loop)	792	=item ev_ref (loop)
651		793
652	=item ev_unref (loop)	794	=item ev_unref (loop)
653		795
654	Ref/unref can be used to add or remove a reference count on the event	796	Ref/unref can be used to add or remove a reference count on the event
655	loop: Every watcher keeps one reference, and as long as the reference	797	loop: Every watcher keeps one reference, and as long as the reference
656	count is nonzero, C<ev_loop> will not return on its own. If you have	798	count is nonzero, C<ev_loop> will not return on its own.
657	a watcher you never unregister that should not keep C<ev_loop> from	799
658	returning, ev_unref() after starting, and ev_ref() before stopping it. For	800	This is useful when you have a watcher that you never intend to
		801	unregister, but that nevertheless should not keep C<ev_loop> from
		802	returning. In such a case, call C<ev_unref> after starting, and C<ev_ref>
		803	before stopping it.
		804
659	example, libev itself uses this for its internal signal pipe: It is not	805	As an example, libev itself uses this for its internal signal pipe: It
660	visible to the libev user and should not keep C<ev_loop> from exiting if	806	is not visible to the libev user and should not keep C<ev_loop> from
661	no event watchers registered by it are active. It is also an excellent	807	exiting if no event watchers registered by it are active. It is also an
662	way to do this for generic recurring timers or from within third-party	808	excellent way to do this for generic recurring timers or from within
663	libraries. Just remember to I<unref after start> and I<ref before stop>	809	third-party libraries. Just remember to I<unref after start> and I<ref
664	(but only if the watcher wasn't active before, or was active before,	810	before stop> (but only if the watcher wasn't active before, or was active
665	respectively).	811	before, respectively. Note also that libev might stop watchers itself
		812	(e.g. non-repeating timers) in which case you have to C<ev_ref>
		813	in the callback).
666		814
667	Example: Create a signal watcher, but keep it from keeping C<ev_loop>	815	Example: Create a signal watcher, but keep it from keeping C<ev_loop>
668	running when nothing else is active.	816	running when nothing else is active.
669		817
670	struct ev_signal exitsig;	818	ev_signal exitsig;
671	ev_signal_init (&exitsig, sig_cb, SIGINT);	819	ev_signal_init (&exitsig, sig_cb, SIGINT);
672	ev_signal_start (loop, &exitsig);	820	ev_signal_start (loop, &exitsig);
673	evf_unref (loop);	821	evf_unref (loop);
674		822
675	Example: For some weird reason, unregister the above signal handler again.	823	Example: For some weird reason, unregister the above signal handler again.
…		…
689	Setting these to a higher value (the C<interval> I<must> be >= C<0>)	837	Setting these to a higher value (the C<interval> I<must> be >= C<0>)
690	allows libev to delay invocation of I/O and timer/periodic callbacks	838	allows libev to delay invocation of I/O and timer/periodic callbacks
691	to increase efficiency of loop iterations (or to increase power-saving	839	to increase efficiency of loop iterations (or to increase power-saving
692	opportunities).	840	opportunities).
693		841
694	The background is that sometimes your program runs just fast enough to	842	The idea is that sometimes your program runs just fast enough to handle
695	handle one (or very few) event(s) per loop iteration. While this makes	843	one (or very few) event(s) per loop iteration. While this makes the
696	the program responsive, it also wastes a lot of CPU time to poll for new	844	program responsive, it also wastes a lot of CPU time to poll for new
697	events, especially with backends like C<select ()> which have a high	845	events, especially with backends like C<select ()> which have a high
698	overhead for the actual polling but can deliver many events at once.	846	overhead for the actual polling but can deliver many events at once.
699		847
700	By setting a higher I<io collect interval> you allow libev to spend more	848	By setting a higher I<io collect interval> you allow libev to spend more
701	time collecting I/O events, so you can handle more events per iteration,	849	time collecting I/O events, so you can handle more events per iteration,
702	at the cost of increasing latency. Timeouts (both C<ev_periodic> and	850	at the cost of increasing latency. Timeouts (both C<ev_periodic> and
703	C<ev_timer>) will be not affected. Setting this to a non-null value will	851	C<ev_timer>) will be not affected. Setting this to a non-null value will
704	introduce an additional C<ev_sleep ()> call into most loop iterations.	852	introduce an additional C<ev_sleep ()> call into most loop iterations. The
		853	sleep time ensures that libev will not poll for I/O events more often then
		854	once per this interval, on average.
705		855
706	Likewise, by setting a higher I<timeout collect interval> you allow libev	856	Likewise, by setting a higher I<timeout collect interval> you allow libev
707	to spend more time collecting timeouts, at the expense of increased	857	to spend more time collecting timeouts, at the expense of increased
708	latency (the watcher callback will be called later). C<ev_io> watchers	858	latency/jitter/inexactness (the watcher callback will be called
709	will not be affected. Setting this to a non-null value will not introduce	859	later). C<ev_io> watchers will not be affected. Setting this to a non-null
710	any overhead in libev.	860	value will not introduce any overhead in libev.
711		861
712	Many (busy) programs can usually benefit by setting the I/O collect	862	Many (busy) programs can usually benefit by setting the I/O collect
713	interval to a value near C<0.1> or so, which is often enough for	863	interval to a value near C<0.1> or so, which is often enough for
714	interactive servers (of course not for games), likewise for timeouts. It	864	interactive servers (of course not for games), likewise for timeouts. It
715	usually doesn't make much sense to set it to a lower value than C<0.01>,	865	usually doesn't make much sense to set it to a lower value than C<0.01>,
716	as this approaches the timing granularity of most systems.	866	as this approaches the timing granularity of most systems. Note that if
		867	you do transactions with the outside world and you can't increase the
		868	parallelity, then this setting will limit your transaction rate (if you
		869	need to poll once per transaction and the I/O collect interval is 0.01,
		870	then you can't do more than 100 transations per second).
717		871
718	Setting the I<timeout collect interval> can improve the opportunity for	872	Setting the I<timeout collect interval> can improve the opportunity for
719	saving power, as the program will "bundle" timer callback invocations that	873	saving power, as the program will "bundle" timer callback invocations that
720	are "near" in time together, by delaying some, thus reducing the number of	874	are "near" in time together, by delaying some, thus reducing the number of
721	times the process sleeps and wakes up again. Another useful technique to	875	times the process sleeps and wakes up again. Another useful technique to
722	reduce iterations/wake-ups is to use C<ev_periodic> watchers and make sure	876	reduce iterations/wake-ups is to use C<ev_periodic> watchers and make sure
723	they fire on, say, one-second boundaries only.	877	they fire on, say, one-second boundaries only.
724		878
		879	Example: we only need 0.1s timeout granularity, and we wish not to poll
		880	more often than 100 times per second:
		881
		882	ev_set_timeout_collect_interval (EV_DEFAULT_UC_ 0.1);
		883	ev_set_io_collect_interval (EV_DEFAULT_UC_ 0.01);
		884
		885	=item ev_invoke_pending (loop)
		886
		887	This call will simply invoke all pending watchers while resetting their
		888	pending state. Normally, C<ev_loop> does this automatically when required,
		889	but when overriding the invoke callback this call comes handy.
		890
		891	=item int ev_pending_count (loop)
		892
		893	Returns the number of pending watchers - zero indicates that no watchers
		894	are pending.
		895
		896	=item ev_set_invoke_pending_cb (loop, void (*invoke_pending_cb)(EV_P))
		897
		898	This overrides the invoke pending functionality of the loop: Instead of
		899	invoking all pending watchers when there are any, C<ev_loop> will call
		900	this callback instead. This is useful, for example, when you want to
		901	invoke the actual watchers inside another context (another thread etc.).
		902
		903	If you want to reset the callback, use C<ev_invoke_pending> as new
		904	callback.
		905
		906	=item ev_set_loop_release_cb (loop, void (*release)(EV_P), void (*acquire)(EV_P))
		907
		908	Sometimes you want to share the same loop between multiple threads. This
		909	can be done relatively simply by putting mutex_lock/unlock calls around
		910	each call to a libev function.
		911
		912	However, C<ev_loop> can run an indefinite time, so it is not feasible to
		913	wait for it to return. One way around this is to wake up the loop via
		914	C<ev_unloop> and C<av_async_send>, another way is to set these I<release>
		915	and I<acquire> callbacks on the loop.
		916
		917	When set, then C<release> will be called just before the thread is
		918	suspended waiting for new events, and C<acquire> is called just
		919	afterwards.
		920
		921	Ideally, C<release> will just call your mutex_unlock function, and
		922	C<acquire> will just call the mutex_lock function again.
		923
		924	While event loop modifications are allowed between invocations of
		925	C<release> and C<acquire> (that's their only purpose after all), no
		926	modifications done will affect the event loop, i.e. adding watchers will
		927	have no effect on the set of file descriptors being watched, or the time
		928	waited. Use an C<ev_async> watcher to wake up C<ev_loop> when you want it
		929	to take note of any changes you made.
		930
		931	In theory, threads executing C<ev_loop> will be async-cancel safe between
		932	invocations of C<release> and C<acquire>.
		933
		934	See also the locking example in the C<THREADS> section later in this
		935	document.
		936
		937	=item ev_set_userdata (loop, void *data)
		938
		939	=item ev_userdata (loop)
		940
		941	Set and retrieve a single C<void *> associated with a loop. When
		942	C<ev_set_userdata> has never been called, then C<ev_userdata> returns
		943	C<0.>
		944
		945	These two functions can be used to associate arbitrary data with a loop,
		946	and are intended solely for the C<invoke_pending_cb>, C<release> and
		947	C<acquire> callbacks described above, but of course can be (ab-)used for
		948	any other purpose as well.
		949
725	=item ev_loop_verify (loop)	950	=item ev_loop_verify (loop)
726		951
727	This function only does something when C<EV_VERIFY> support has been	952	This function only does something when C<EV_VERIFY> support has been
728	compiled in. It tries to go through all internal structures and checks	953	compiled in, which is the default for non-minimal builds. It tries to go
729	them for validity. If anything is found to be inconsistent, it will print	954	through all internal structures and checks them for validity. If anything
730	an error message to standard error and call C<abort ()>.	955	is found to be inconsistent, it will print an error message to standard
		956	error and call C<abort ()>.
731		957
732	This can be used to catch bugs inside libev itself: under normal	958	This can be used to catch bugs inside libev itself: under normal
733	circumstances, this function will never abort as of course libev keeps its	959	circumstances, this function will never abort as of course libev keeps its
734	data structures consistent.	960	data structures consistent.
735		961
736	=back	962	=back
737		963
738		964
739	=head1 ANATOMY OF A WATCHER	965	=head1 ANATOMY OF A WATCHER
740		966
		967	In the following description, uppercase C<TYPE> in names stands for the
		968	watcher type, e.g. C<ev_TYPE_start> can mean C<ev_timer_start> for timer
		969	watchers and C<ev_io_start> for I/O watchers.
		970
741	A watcher is a structure that you create and register to record your	971	A watcher is a structure that you create and register to record your
742	interest in some event. For instance, if you want to wait for STDIN to	972	interest in some event. For instance, if you want to wait for STDIN to
743	become readable, you would create an C<ev_io> watcher for that:	973	become readable, you would create an C<ev_io> watcher for that:
744		974
745	static void my_cb (struct ev_loop *loop, struct ev_io *w, int revents)	975	static void my_cb (struct ev_loop *loop, ev_io *w, int revents)
746	{	976	{
747	ev_io_stop (w);	977	ev_io_stop (w);
748	ev_unloop (loop, EVUNLOOP_ALL);	978	ev_unloop (loop, EVUNLOOP_ALL);
749	}	979	}
750		980
751	struct ev_loop *loop = ev_default_loop (0);	981	struct ev_loop *loop = ev_default_loop (0);
		982
752	struct ev_io stdin_watcher;	983	ev_io stdin_watcher;
		984
753	ev_init (&stdin_watcher, my_cb);	985	ev_init (&stdin_watcher, my_cb);
754	ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ);	986	ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ);
755	ev_io_start (loop, &stdin_watcher);	987	ev_io_start (loop, &stdin_watcher);
		988
756	ev_loop (loop, 0);	989	ev_loop (loop, 0);
757		990
758	As you can see, you are responsible for allocating the memory for your	991	As you can see, you are responsible for allocating the memory for your
759	watcher structures (and it is usually a bad idea to do this on the stack,	992	watcher structures (and it is I<usually> a bad idea to do this on the
760	although this can sometimes be quite valid).	993	stack).
		994
		995	Each watcher has an associated watcher structure (called C<struct ev_TYPE>
		996	or simply C<ev_TYPE>, as typedefs are provided for all watcher structs).
761		997
762	Each watcher structure must be initialised by a call to C<ev_init	998	Each watcher structure must be initialised by a call to C<ev_init
763	(watcher *, callback)>, which expects a callback to be provided. This	999	(watcher *, callback)>, which expects a callback to be provided. This
764	callback gets invoked each time the event occurs (or, in the case of I/O	1000	callback gets invoked each time the event occurs (or, in the case of I/O
765	watchers, each time the event loop detects that the file descriptor given	1001	watchers, each time the event loop detects that the file descriptor given
766	is readable and/or writable).	1002	is readable and/or writable).
767		1003
768	Each watcher type has its own C<< ev_<type>_set (watcher *, ...) >> macro	1004	Each watcher type further has its own C<< ev_TYPE_set (watcher *, ...) >>
769	with arguments specific to this watcher type. There is also a macro	1005	macro to configure it, with arguments specific to the watcher type. There
770	to combine initialisation and setting in one call: C<< ev_<type>_init	1006	is also a macro to combine initialisation and setting in one call: C<<
771	(watcher *, callback, ...) >>.	1007	ev_TYPE_init (watcher *, callback, ...) >>.
772		1008
773	To make the watcher actually watch out for events, you have to start it	1009	To make the watcher actually watch out for events, you have to start it
774	with a watcher-specific start function (C<< ev_<type>_start (loop, watcher	1010	with a watcher-specific start function (C<< ev_TYPE_start (loop, watcher
775	*) >>), and you can stop watching for events at any time by calling the	1011	*) >>), and you can stop watching for events at any time by calling the
776	corresponding stop function (C<< ev_<type>_stop (loop, watcher *) >>.	1012	corresponding stop function (C<< ev_TYPE_stop (loop, watcher *) >>.
777		1013
778	As long as your watcher is active (has been started but not stopped) you	1014	As long as your watcher is active (has been started but not stopped) you
779	must not touch the values stored in it. Most specifically you must never	1015	must not touch the values stored in it. Most specifically you must never
780	reinitialise it or call its C<set> macro.	1016	reinitialise it or call its C<ev_TYPE_set> macro.
781		1017
782	Each and every callback receives the event loop pointer as first, the	1018	Each and every callback receives the event loop pointer as first, the
783	registered watcher structure as second, and a bitset of received events as	1019	registered watcher structure as second, and a bitset of received events as
784	third argument.	1020	third argument.
785		1021
…		…
843		1079
844	=item C<EV_ASYNC>	1080	=item C<EV_ASYNC>
845		1081
846	The given async watcher has been asynchronously notified (see C<ev_async>).	1082	The given async watcher has been asynchronously notified (see C<ev_async>).
847		1083
		1084	=item C<EV_CUSTOM>
		1085
		1086	Not ever sent (or otherwise used) by libev itself, but can be freely used
		1087	by libev users to signal watchers (e.g. via C<ev_feed_event>).
		1088
848	=item C<EV_ERROR>	1089	=item C<EV_ERROR>
849		1090
850	An unspecified error has occurred, the watcher has been stopped. This might	1091	An unspecified error has occurred, the watcher has been stopped. This might
851	happen because the watcher could not be properly started because libev	1092	happen because the watcher could not be properly started because libev
852	ran out of memory, a file descriptor was found to be closed or any other	1093	ran out of memory, a file descriptor was found to be closed or any other
		1094	problem. Libev considers these application bugs.
		1095
853	problem. You best act on it by reporting the problem and somehow coping	1096	You best act on it by reporting the problem and somehow coping with the
854	with the watcher being stopped.	1097	watcher being stopped. Note that well-written programs should not receive
		1098	an error ever, so when your watcher receives it, this usually indicates a
		1099	bug in your program.
855		1100
856	Libev will usually signal a few "dummy" events together with an error,	1101	Libev will usually signal a few "dummy" events together with an error, for
857	for example it might indicate that a fd is readable or writable, and if	1102	example it might indicate that a fd is readable or writable, and if your
858	your callbacks is well-written it can just attempt the operation and cope	1103	callbacks is well-written it can just attempt the operation and cope with
859	with the error from read() or write(). This will not work in multi-threaded	1104	the error from read() or write(). This will not work in multi-threaded
860	programs, though, so beware.	1105	programs, though, as the fd could already be closed and reused for another
		1106	thing, so beware.
861		1107
862	=back	1108	=back
863		1109
864	=head2 GENERIC WATCHER FUNCTIONS	1110	=head2 GENERIC WATCHER FUNCTIONS
865
866	In the following description, C<TYPE> stands for the watcher type,
867	e.g. C<timer> for C<ev_timer> watchers and C<io> for C<ev_io> watchers.
868		1111
869	=over 4	1112	=over 4
870		1113
871	=item C<ev_init> (ev_TYPE *watcher, callback)	1114	=item C<ev_init> (ev_TYPE *watcher, callback)
872		1115
…		…
878	which rolls both calls into one.	1121	which rolls both calls into one.
879		1122
880	You can reinitialise a watcher at any time as long as it has been stopped	1123	You can reinitialise a watcher at any time as long as it has been stopped
881	(or never started) and there are no pending events outstanding.	1124	(or never started) and there are no pending events outstanding.
882		1125
883	The callback is always of type C<void (*)(ev_loop *loop, ev_TYPE *watcher,	1126	The callback is always of type C<void (*)(struct ev_loop *loop, ev_TYPE *watcher,
884	int revents)>.	1127	int revents)>.
885		1128
		1129	Example: Initialise an C<ev_io> watcher in two steps.
		1130
		1131	ev_io w;
		1132	ev_init (&w, my_cb);
		1133	ev_io_set (&w, STDIN_FILENO, EV_READ);
		1134
886	=item C<ev_TYPE_set> (ev_TYPE *, [args])	1135	=item C<ev_TYPE_set> (ev_TYPE *watcher, [args])
887		1136
888	This macro initialises the type-specific parts of a watcher. You need to	1137	This macro initialises the type-specific parts of a watcher. You need to
889	call C<ev_init> at least once before you call this macro, but you can	1138	call C<ev_init> at least once before you call this macro, but you can
890	call C<ev_TYPE_set> any number of times. You must not, however, call this	1139	call C<ev_TYPE_set> any number of times. You must not, however, call this
891	macro on a watcher that is active (it can be pending, however, which is a	1140	macro on a watcher that is active (it can be pending, however, which is a
892	difference to the C<ev_init> macro).	1141	difference to the C<ev_init> macro).
893		1142
894	Although some watcher types do not have type-specific arguments	1143	Although some watcher types do not have type-specific arguments
895	(e.g. C<ev_prepare>) you still need to call its C<set> macro.	1144	(e.g. C<ev_prepare>) you still need to call its C<set> macro.
896		1145
		1146	See C<ev_init>, above, for an example.
		1147
897	=item C<ev_TYPE_init> (ev_TYPE *watcher, callback, [args])	1148	=item C<ev_TYPE_init> (ev_TYPE *watcher, callback, [args])
898		1149
899	This convenience macro rolls both C<ev_init> and C<ev_TYPE_set> macro	1150	This convenience macro rolls both C<ev_init> and C<ev_TYPE_set> macro
900	calls into a single call. This is the most convenient method to initialise	1151	calls into a single call. This is the most convenient method to initialise
901	a watcher. The same limitations apply, of course.	1152	a watcher. The same limitations apply, of course.
902		1153
		1154	Example: Initialise and set an C<ev_io> watcher in one step.
		1155
		1156	ev_io_init (&w, my_cb, STDIN_FILENO, EV_READ);
		1157
903	=item C<ev_TYPE_start> (loop *, ev_TYPE *watcher)	1158	=item C<ev_TYPE_start> (loop, ev_TYPE *watcher)
904		1159
905	Starts (activates) the given watcher. Only active watchers will receive	1160	Starts (activates) the given watcher. Only active watchers will receive
906	events. If the watcher is already active nothing will happen.	1161	events. If the watcher is already active nothing will happen.
907		1162
		1163	Example: Start the C<ev_io> watcher that is being abused as example in this
		1164	whole section.
		1165
		1166	ev_io_start (EV_DEFAULT_UC, &w);
		1167
908	=item C<ev_TYPE_stop> (loop *, ev_TYPE *watcher)	1168	=item C<ev_TYPE_stop> (loop, ev_TYPE *watcher)
909		1169
910	Stops the given watcher again (if active) and clears the pending	1170	Stops the given watcher if active, and clears the pending status (whether
		1171	the watcher was active or not).
		1172
911	status. It is possible that stopped watchers are pending (for example,	1173	It is possible that stopped watchers are pending - for example,
912	non-repeating timers are being stopped when they become pending), but	1174	non-repeating timers are being stopped when they become pending - but
913	C<ev_TYPE_stop> ensures that the watcher is neither active nor pending. If	1175	calling C<ev_TYPE_stop> ensures that the watcher is neither active nor
914	you want to free or reuse the memory used by the watcher it is therefore a	1176	pending. If you want to free or reuse the memory used by the watcher it is
915	good idea to always call its C<ev_TYPE_stop> function.	1177	therefore a good idea to always call its C<ev_TYPE_stop> function.
916		1178
917	=item bool ev_is_active (ev_TYPE *watcher)	1179	=item bool ev_is_active (ev_TYPE *watcher)
918		1180
919	Returns a true value iff the watcher is active (i.e. it has been started	1181	Returns a true value iff the watcher is active (i.e. it has been started
920	and not yet been stopped). As long as a watcher is active you must not modify	1182	and not yet been stopped). As long as a watcher is active you must not modify
…		…
936	=item ev_cb_set (ev_TYPE *watcher, callback)	1198	=item ev_cb_set (ev_TYPE *watcher, callback)
937		1199
938	Change the callback. You can change the callback at virtually any time	1200	Change the callback. You can change the callback at virtually any time
939	(modulo threads).	1201	(modulo threads).
940		1202
941	=item ev_set_priority (ev_TYPE *watcher, priority)	1203	=item ev_set_priority (ev_TYPE *watcher, int priority)
942		1204
943	=item int ev_priority (ev_TYPE *watcher)	1205	=item int ev_priority (ev_TYPE *watcher)
944		1206
945	Set and query the priority of the watcher. The priority is a small	1207	Set and query the priority of the watcher. The priority is a small
946	integer between C<EV_MAXPRI> (default: C<2>) and C<EV_MINPRI>	1208	integer between C<EV_MAXPRI> (default: C<2>) and C<EV_MINPRI>
947	(default: C<-2>). Pending watchers with higher priority will be invoked	1209	(default: C<-2>). Pending watchers with higher priority will be invoked
948	before watchers with lower priority, but priority will not keep watchers	1210	before watchers with lower priority, but priority will not keep watchers
949	from being executed (except for C<ev_idle> watchers).	1211	from being executed (except for C<ev_idle> watchers).
950		1212
951	This means that priorities are I<only> used for ordering callback
952	invocation after new events have been received. This is useful, for
953	example, to reduce latency after idling, or more often, to bind two
954	watchers on the same event and make sure one is called first.
955
956	If you need to suppress invocation when higher priority events are pending	1213	If you need to suppress invocation when higher priority events are pending
957	you need to look at C<ev_idle> watchers, which provide this functionality.	1214	you need to look at C<ev_idle> watchers, which provide this functionality.
958		1215
959	You I<must not> change the priority of a watcher as long as it is active or	1216	You I<must not> change the priority of a watcher as long as it is active or
960	pending.	1217	pending.
961		1218
		1219	Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is
		1220	fine, as long as you do not mind that the priority value you query might
		1221	or might not have been clamped to the valid range.
		1222
962	The default priority used by watchers when no priority has been set is	1223	The default priority used by watchers when no priority has been set is
963	always C<0>, which is supposed to not be too high and not be too low :).	1224	always C<0>, which is supposed to not be too high and not be too low :).
964		1225
965	Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is	1226	See L<WATCHER PRIORITY MODELS>, below, for a more thorough treatment of
966	fine, as long as you do not mind that the priority value you query might	1227	priorities.
967	or might not have been adjusted to be within valid range.
968		1228
969	=item ev_invoke (loop, ev_TYPE *watcher, int revents)	1229	=item ev_invoke (loop, ev_TYPE *watcher, int revents)
970		1230
971	Invoke the C<watcher> with the given C<loop> and C<revents>. Neither	1231	Invoke the C<watcher> with the given C<loop> and C<revents>. Neither
972	C<loop> nor C<revents> need to be valid as long as the watcher callback	1232	C<loop> nor C<revents> need to be valid as long as the watcher callback
973	can deal with that fact.	1233	can deal with that fact, as both are simply passed through to the
		1234	callback.
974		1235
975	=item int ev_clear_pending (loop, ev_TYPE *watcher)	1236	=item int ev_clear_pending (loop, ev_TYPE *watcher)
976		1237
977	If the watcher is pending, this function returns clears its pending status	1238	If the watcher is pending, this function clears its pending status and
978	and returns its C<revents> bitset (as if its callback was invoked). If the	1239	returns its C<revents> bitset (as if its callback was invoked). If the
979	watcher isn't pending it does nothing and returns C<0>.	1240	watcher isn't pending it does nothing and returns C<0>.
980		1241
		1242	Sometimes it can be useful to "poll" a watcher instead of waiting for its
		1243	callback to be invoked, which can be accomplished with this function.
		1244
		1245	=item ev_feed_event (loop, ev_TYPE *watcher, int revents)
		1246
		1247	Feeds the given event set into the event loop, as if the specified event
		1248	had happened for the specified watcher (which must be a pointer to an
		1249	initialised but not necessarily started event watcher). Obviously you must
		1250	not free the watcher as long as it has pending events.
		1251
		1252	Stopping the watcher, letting libev invoke it, or calling
		1253	C<ev_clear_pending> will clear the pending event, even if the watcher was
		1254	not started in the first place.
		1255
		1256	See also C<ev_feed_fd_event> and C<ev_feed_signal_event> for related
		1257	functions that do not need a watcher.
		1258
981	=back	1259	=back
982		1260
983		1261
984	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER	1262	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER
985		1263
986	Each watcher has, by default, a member C<void *data> that you can change	1264	Each watcher has, by default, a member C<void *data> that you can change
987	and read at any time, libev will completely ignore it. This can be used	1265	and read at any time: libev will completely ignore it. This can be used
988	to associate arbitrary data with your watcher. If you need more data and	1266	to associate arbitrary data with your watcher. If you need more data and
989	don't want to allocate memory and store a pointer to it in that data	1267	don't want to allocate memory and store a pointer to it in that data
990	member, you can also "subclass" the watcher type and provide your own	1268	member, you can also "subclass" the watcher type and provide your own
991	data:	1269	data:
992		1270
993	struct my_io	1271	struct my_io
994	{	1272	{
995	struct ev_io io;	1273	ev_io io;
996	int otherfd;	1274	int otherfd;
997	void *somedata;	1275	void *somedata;
998	struct whatever *mostinteresting;	1276	struct whatever *mostinteresting;
999	}	1277	};
		1278
		1279	...
		1280	struct my_io w;
		1281	ev_io_init (&w.io, my_cb, fd, EV_READ);
1000		1282
1001	And since your callback will be called with a pointer to the watcher, you	1283	And since your callback will be called with a pointer to the watcher, you
1002	can cast it back to your own type:	1284	can cast it back to your own type:
1003		1285
1004	static void my_cb (struct ev_loop *loop, struct ev_io *w_, int revents)	1286	static void my_cb (struct ev_loop *loop, ev_io *w_, int revents)
1005	{	1287	{
1006	struct my_io *w = (struct my_io *)w_;	1288	struct my_io *w = (struct my_io *)w_;
1007	...	1289	...
1008	}	1290	}
1009		1291
1010	More interesting and less C-conformant ways of casting your callback type	1292	More interesting and less C-conformant ways of casting your callback type
1011	instead have been omitted.	1293	instead have been omitted.
1012		1294
1013	Another common scenario is having some data structure with multiple	1295	Another common scenario is to use some data structure with multiple
1014	watchers:	1296	embedded watchers:
1015		1297
1016	struct my_biggy	1298	struct my_biggy
1017	{	1299	{
1018	int some_data;	1300	int some_data;
1019	ev_timer t1;	1301	ev_timer t1;
1020	ev_timer t2;	1302	ev_timer t2;
1021	}	1303	}
1022		1304
1023	In this case getting the pointer to C<my_biggy> is a bit more complicated,	1305	In this case getting the pointer to C<my_biggy> is a bit more
1024	you need to use C<offsetof>:	1306	complicated: Either you store the address of your C<my_biggy> struct
		1307	in the C<data> member of the watcher (for woozies), or you need to use
		1308	some pointer arithmetic using C<offsetof> inside your watchers (for real
		1309	programmers):
1025		1310
1026	#include <stddef.h>	1311	#include <stddef.h>
1027		1312
1028	static void	1313	static void
1029	t1_cb (EV_P_ struct ev_timer *w, int revents)	1314	t1_cb (EV_P_ ev_timer *w, int revents)
1030	{	1315	{
1031	struct my_biggy big = (struct my_biggy *	1316	struct my_biggy big = (struct my_biggy *)
1032	(((char *)w) - offsetof (struct my_biggy, t1));	1317	(((char *)w) - offsetof (struct my_biggy, t1));
1033	}	1318	}
1034		1319
1035	static void	1320	static void
1036	t2_cb (EV_P_ struct ev_timer *w, int revents)	1321	t2_cb (EV_P_ ev_timer *w, int revents)
1037	{	1322	{
1038	struct my_biggy big = (struct my_biggy *	1323	struct my_biggy big = (struct my_biggy *)
1039	(((char *)w) - offsetof (struct my_biggy, t2));	1324	(((char *)w) - offsetof (struct my_biggy, t2));
1040	}	1325	}
		1326
		1327	=head2 WATCHER PRIORITY MODELS
		1328
		1329	Many event loops support I<watcher priorities>, which are usually small
		1330	integers that influence the ordering of event callback invocation
		1331	between watchers in some way, all else being equal.
		1332
		1333	In libev, Watcher priorities can be set using C<ev_set_priority>. See its
		1334	description for the more technical details such as the actual priority
		1335	range.
		1336
		1337	There are two common ways how these these priorities are being interpreted
		1338	by event loops:
		1339
		1340	In the more common lock-out model, higher priorities "lock out" invocation
		1341	of lower priority watchers, which means as long as higher priority
		1342	watchers receive events, lower priority watchers are not being invoked.
		1343
		1344	The less common only-for-ordering model uses priorities solely to order
		1345	callback invocation within a single event loop iteration: Higher priority
		1346	watchers are invoked before lower priority ones, but they all get invoked
		1347	before polling for new events.
		1348
		1349	Libev uses the second (only-for-ordering) model for all its watchers
		1350	except for idle watchers (which use the lock-out model).
		1351
		1352	The rationale behind this is that implementing the lock-out model for
		1353	watchers is not well supported by most kernel interfaces, and most event
		1354	libraries will just poll for the same events again and again as long as
		1355	their callbacks have not been executed, which is very inefficient in the
		1356	common case of one high-priority watcher locking out a mass of lower
		1357	priority ones.
		1358
		1359	Static (ordering) priorities are most useful when you have two or more
		1360	watchers handling the same resource: a typical usage example is having an
		1361	C<ev_io> watcher to receive data, and an associated C<ev_timer> to handle
		1362	timeouts. Under load, data might be received while the program handles
		1363	other jobs, but since timers normally get invoked first, the timeout
		1364	handler will be executed before checking for data. In that case, giving
		1365	the timer a lower priority than the I/O watcher ensures that I/O will be
		1366	handled first even under adverse conditions (which is usually, but not
		1367	always, what you want).
		1368
		1369	Since idle watchers use the "lock-out" model, meaning that idle watchers
		1370	will only be executed when no same or higher priority watchers have
		1371	received events, they can be used to implement the "lock-out" model when
		1372	required.
		1373
		1374	For example, to emulate how many other event libraries handle priorities,
		1375	you can associate an C<ev_idle> watcher to each such watcher, and in
		1376	the normal watcher callback, you just start the idle watcher. The real
		1377	processing is done in the idle watcher callback. This causes libev to
		1378	continously poll and process kernel event data for the watcher, but when
		1379	the lock-out case is known to be rare (which in turn is rare :), this is
		1380	workable.
		1381
		1382	Usually, however, the lock-out model implemented that way will perform
		1383	miserably under the type of load it was designed to handle. In that case,
		1384	it might be preferable to stop the real watcher before starting the
		1385	idle watcher, so the kernel will not have to process the event in case
		1386	the actual processing will be delayed for considerable time.
		1387
		1388	Here is an example of an I/O watcher that should run at a strictly lower
		1389	priority than the default, and which should only process data when no
		1390	other events are pending:
		1391
		1392	ev_idle idle; // actual processing watcher
		1393	ev_io io; // actual event watcher
		1394
		1395	static void
		1396	io_cb (EV_P_ ev_io *w, int revents)
		1397	{
		1398	// stop the I/O watcher, we received the event, but
		1399	// are not yet ready to handle it.
		1400	ev_io_stop (EV_A_ w);
		1401
		1402	// start the idle watcher to ahndle the actual event.
		1403	// it will not be executed as long as other watchers
		1404	// with the default priority are receiving events.
		1405	ev_idle_start (EV_A_ &idle);
		1406	}
		1407
		1408	static void
		1409	idle_cb (EV_P_ ev_idle *w, int revents)
		1410	{
		1411	// actual processing
		1412	read (STDIN_FILENO, ...);
		1413
		1414	// have to start the I/O watcher again, as
		1415	// we have handled the event
		1416	ev_io_start (EV_P_ &io);
		1417	}
		1418
		1419	// initialisation
		1420	ev_idle_init (&idle, idle_cb);
		1421	ev_io_init (&io, io_cb, STDIN_FILENO, EV_READ);
		1422	ev_io_start (EV_DEFAULT_ &io);
		1423
		1424	In the "real" world, it might also be beneficial to start a timer, so that
		1425	low-priority connections can not be locked out forever under load. This
		1426	enables your program to keep a lower latency for important connections
		1427	during short periods of high load, while not completely locking out less
		1428	important ones.
1041		1429
1042		1430
1043	=head1 WATCHER TYPES	1431	=head1 WATCHER TYPES
1044		1432
1045	This section describes each watcher in detail, but will not repeat	1433	This section describes each watcher in detail, but will not repeat
…		…
1069	In general you can register as many read and/or write event watchers per	1457	In general you can register as many read and/or write event watchers per
1070	fd as you want (as long as you don't confuse yourself). Setting all file	1458	fd as you want (as long as you don't confuse yourself). Setting all file
1071	descriptors to non-blocking mode is also usually a good idea (but not	1459	descriptors to non-blocking mode is also usually a good idea (but not
1072	required if you know what you are doing).	1460	required if you know what you are doing).
1073		1461
1074	If you must do this, then force the use of a known-to-be-good backend	1462	If you cannot use non-blocking mode, then force the use of a
1075	(at the time of this writing, this includes only C<EVBACKEND_SELECT> and	1463	known-to-be-good backend (at the time of this writing, this includes only
1076	C<EVBACKEND_POLL>).	1464	C<EVBACKEND_SELECT> and C<EVBACKEND_POLL>). The same applies to file
		1465	descriptors for which non-blocking operation makes no sense (such as
		1466	files) - libev doesn't guarentee any specific behaviour in that case.
1077		1467
1078	Another thing you have to watch out for is that it is quite easy to	1468	Another thing you have to watch out for is that it is quite easy to
1079	receive "spurious" readiness notifications, that is your callback might	1469	receive "spurious" readiness notifications, that is your callback might
1080	be called with C<EV_READ> but a subsequent C<read>(2) will actually block	1470	be called with C<EV_READ> but a subsequent C<read>(2) will actually block
1081	because there is no data. Not only are some backends known to create a	1471	because there is no data. Not only are some backends known to create a
1082	lot of those (for example Solaris ports), it is very easy to get into	1472	lot of those (for example Solaris ports), it is very easy to get into
1083	this situation even with a relatively standard program structure. Thus	1473	this situation even with a relatively standard program structure. Thus
1084	it is best to always use non-blocking I/O: An extra C<read>(2) returning	1474	it is best to always use non-blocking I/O: An extra C<read>(2) returning
1085	C<EAGAIN> is far preferable to a program hanging until some data arrives.	1475	C<EAGAIN> is far preferable to a program hanging until some data arrives.
1086		1476
1087	If you cannot run the fd in non-blocking mode (for example you should not	1477	If you cannot run the fd in non-blocking mode (for example you should
1088	play around with an Xlib connection), then you have to separately re-test	1478	not play around with an Xlib connection), then you have to separately
1089	whether a file descriptor is really ready with a known-to-be good interface	1479	re-test whether a file descriptor is really ready with a known-to-be good
1090	such as poll (fortunately in our Xlib example, Xlib already does this on	1480	interface such as poll (fortunately in our Xlib example, Xlib already
1091	its own, so its quite safe to use).	1481	does this on its own, so its quite safe to use). Some people additionally
		1482	use C<SIGALRM> and an interval timer, just to be sure you won't block
		1483	indefinitely.
		1484
		1485	But really, best use non-blocking mode.
1092		1486
1093	=head3 The special problem of disappearing file descriptors	1487	=head3 The special problem of disappearing file descriptors
1094		1488
1095	Some backends (e.g. kqueue, epoll) need to be told about closing a file	1489	Some backends (e.g. kqueue, epoll) need to be told about closing a file
1096	descriptor (either by calling C<close> explicitly or by any other means,	1490	descriptor (either due to calling C<close> explicitly or any other means,
1097	such as C<dup>). The reason is that you register interest in some file	1491	such as C<dup2>). The reason is that you register interest in some file
1098	descriptor, but when it goes away, the operating system will silently drop	1492	descriptor, but when it goes away, the operating system will silently drop
1099	this interest. If another file descriptor with the same number then is	1493	this interest. If another file descriptor with the same number then is
1100	registered with libev, there is no efficient way to see that this is, in	1494	registered with libev, there is no efficient way to see that this is, in
1101	fact, a different file descriptor.	1495	fact, a different file descriptor.
1102		1496
…		…
1133	enable C<EVFLAG_FORKCHECK>, or resort to C<EVBACKEND_SELECT> or	1527	enable C<EVFLAG_FORKCHECK>, or resort to C<EVBACKEND_SELECT> or
1134	C<EVBACKEND_POLL>.	1528	C<EVBACKEND_POLL>.
1135		1529
1136	=head3 The special problem of SIGPIPE	1530	=head3 The special problem of SIGPIPE
1137		1531
1138	While not really specific to libev, it is easy to forget about SIGPIPE:	1532	While not really specific to libev, it is easy to forget about C<SIGPIPE>:
1139	when writing to a pipe whose other end has been closed, your program gets	1533	when writing to a pipe whose other end has been closed, your program gets
1140	send a SIGPIPE, which, by default, aborts your program. For most programs	1534	sent a SIGPIPE, which, by default, aborts your program. For most programs
1141	this is sensible behaviour, for daemons, this is usually undesirable.	1535	this is sensible behaviour, for daemons, this is usually undesirable.
1142		1536
1143	So when you encounter spurious, unexplained daemon exits, make sure you	1537	So when you encounter spurious, unexplained daemon exits, make sure you
1144	ignore SIGPIPE (and maybe make sure you log the exit status of your daemon	1538	ignore SIGPIPE (and maybe make sure you log the exit status of your daemon
1145	somewhere, as that would have given you a big clue).	1539	somewhere, as that would have given you a big clue).
…		…
1152	=item ev_io_init (ev_io *, callback, int fd, int events)	1546	=item ev_io_init (ev_io *, callback, int fd, int events)
1153		1547
1154	=item ev_io_set (ev_io *, int fd, int events)	1548	=item ev_io_set (ev_io *, int fd, int events)
1155		1549
1156	Configures an C<ev_io> watcher. The C<fd> is the file descriptor to	1550	Configures an C<ev_io> watcher. The C<fd> is the file descriptor to
1157	receive events for and events is either C<EV_READ>, C<EV_WRITE> or	1551	receive events for and C<events> is either C<EV_READ>, C<EV_WRITE> or
1158	C<EV_READ | EV_WRITE> to receive the given events.	1552	C<EV_READ | EV_WRITE>, to express the desire to receive the given events.
1159		1553
1160	=item int fd [read-only]	1554	=item int fd [read-only]
1161		1555
1162	The file descriptor being watched.	1556	The file descriptor being watched.
1163		1557
…		…
1172	Example: Call C<stdin_readable_cb> when STDIN_FILENO has become, well	1566	Example: Call C<stdin_readable_cb> when STDIN_FILENO has become, well
1173	readable, but only once. Since it is likely line-buffered, you could	1567	readable, but only once. Since it is likely line-buffered, you could
1174	attempt to read a whole line in the callback.	1568	attempt to read a whole line in the callback.
1175		1569
1176	static void	1570	static void
1177	stdin_readable_cb (struct ev_loop *loop, struct ev_io *w, int revents)	1571	stdin_readable_cb (struct ev_loop *loop, ev_io *w, int revents)
1178	{	1572	{
1179	ev_io_stop (loop, w);	1573	ev_io_stop (loop, w);
1180	.. read from stdin here (or from w->fd) and haqndle any I/O errors	1574	.. read from stdin here (or from w->fd) and handle any I/O errors
1181	}	1575	}
1182		1576
1183	...	1577	...
1184	struct ev_loop *loop = ev_default_init (0);	1578	struct ev_loop *loop = ev_default_init (0);
1185	struct ev_io stdin_readable;	1579	ev_io stdin_readable;
1186	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);	1580	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);
1187	ev_io_start (loop, &stdin_readable);	1581	ev_io_start (loop, &stdin_readable);
1188	ev_loop (loop, 0);	1582	ev_loop (loop, 0);
1189		1583
1190		1584
…		…
1193	Timer watchers are simple relative timers that generate an event after a	1587	Timer watchers are simple relative timers that generate an event after a
1194	given time, and optionally repeating in regular intervals after that.	1588	given time, and optionally repeating in regular intervals after that.
1195		1589
1196	The timers are based on real time, that is, if you register an event that	1590	The timers are based on real time, that is, if you register an event that
1197	times out after an hour and you reset your system clock to January last	1591	times out after an hour and you reset your system clock to January last
1198	year, it will still time out after (roughly) and hour. "Roughly" because	1592	year, it will still time out after (roughly) one hour. "Roughly" because
1199	detecting time jumps is hard, and some inaccuracies are unavoidable (the	1593	detecting time jumps is hard, and some inaccuracies are unavoidable (the
1200	monotonic clock option helps a lot here).	1594	monotonic clock option helps a lot here).
1201		1595
1202	The callback is guaranteed to be invoked only after its timeout has passed,	1596	The callback is guaranteed to be invoked only I<after> its timeout has
1203	but if multiple timers become ready during the same loop iteration then	1597	passed (not I<at>, so on systems with very low-resolution clocks this
1204	order of execution is undefined.	1598	might introduce a small delay). If multiple timers become ready during the
		1599	same loop iteration then the ones with earlier time-out values are invoked
		1600	before ones of the same priority with later time-out values (but this is
		1601	no longer true when a callback calls C<ev_loop> recursively).
		1602
		1603	=head3 Be smart about timeouts
		1604
		1605	Many real-world problems involve some kind of timeout, usually for error
		1606	recovery. A typical example is an HTTP request - if the other side hangs,
		1607	you want to raise some error after a while.
		1608
		1609	What follows are some ways to handle this problem, from obvious and
		1610	inefficient to smart and efficient.
		1611
		1612	In the following, a 60 second activity timeout is assumed - a timeout that
		1613	gets reset to 60 seconds each time there is activity (e.g. each time some
		1614	data or other life sign was received).
		1615
		1616	=over 4
		1617
		1618	=item 1. Use a timer and stop, reinitialise and start it on activity.
		1619
		1620	This is the most obvious, but not the most simple way: In the beginning,
		1621	start the watcher:
		1622
		1623	ev_timer_init (timer, callback, 60., 0.);
		1624	ev_timer_start (loop, timer);
		1625
		1626	Then, each time there is some activity, C<ev_timer_stop> it, initialise it
		1627	and start it again:
		1628
		1629	ev_timer_stop (loop, timer);
		1630	ev_timer_set (timer, 60., 0.);
		1631	ev_timer_start (loop, timer);
		1632
		1633	This is relatively simple to implement, but means that each time there is
		1634	some activity, libev will first have to remove the timer from its internal
		1635	data structure and then add it again. Libev tries to be fast, but it's
		1636	still not a constant-time operation.
		1637
		1638	=item 2. Use a timer and re-start it with C<ev_timer_again> inactivity.
		1639
		1640	This is the easiest way, and involves using C<ev_timer_again> instead of
		1641	C<ev_timer_start>.
		1642
		1643	To implement this, configure an C<ev_timer> with a C<repeat> value
		1644	of C<60> and then call C<ev_timer_again> at start and each time you
		1645	successfully read or write some data. If you go into an idle state where
		1646	you do not expect data to travel on the socket, you can C<ev_timer_stop>
		1647	the timer, and C<ev_timer_again> will automatically restart it if need be.
		1648
		1649	That means you can ignore both the C<ev_timer_start> function and the
		1650	C<after> argument to C<ev_timer_set>, and only ever use the C<repeat>
		1651	member and C<ev_timer_again>.
		1652
		1653	At start:
		1654
		1655	ev_init (timer, callback);
		1656	timer->repeat = 60.;
		1657	ev_timer_again (loop, timer);
		1658
		1659	Each time there is some activity:
		1660
		1661	ev_timer_again (loop, timer);
		1662
		1663	It is even possible to change the time-out on the fly, regardless of
		1664	whether the watcher is active or not:
		1665
		1666	timer->repeat = 30.;
		1667	ev_timer_again (loop, timer);
		1668
		1669	This is slightly more efficient then stopping/starting the timer each time
		1670	you want to modify its timeout value, as libev does not have to completely
		1671	remove and re-insert the timer from/into its internal data structure.
		1672
		1673	It is, however, even simpler than the "obvious" way to do it.
		1674
		1675	=item 3. Let the timer time out, but then re-arm it as required.
		1676
		1677	This method is more tricky, but usually most efficient: Most timeouts are
		1678	relatively long compared to the intervals between other activity - in
		1679	our example, within 60 seconds, there are usually many I/O events with
		1680	associated activity resets.
		1681
		1682	In this case, it would be more efficient to leave the C<ev_timer> alone,
		1683	but remember the time of last activity, and check for a real timeout only
		1684	within the callback:
		1685
		1686	ev_tstamp last_activity; // time of last activity
		1687
		1688	static void
		1689	callback (EV_P_ ev_timer *w, int revents)
		1690	{
		1691	ev_tstamp now = ev_now (EV_A);
		1692	ev_tstamp timeout = last_activity + 60.;
		1693
		1694	// if last_activity + 60. is older than now, we did time out
		1695	if (timeout < now)
		1696	{
		1697	// timeout occured, take action
		1698	}
		1699	else
		1700	{
		1701	// callback was invoked, but there was some activity, re-arm
		1702	// the watcher to fire in last_activity + 60, which is
		1703	// guaranteed to be in the future, so "again" is positive:
		1704	w->repeat = timeout - now;
		1705	ev_timer_again (EV_A_ w);
		1706	}
		1707	}
		1708
		1709	To summarise the callback: first calculate the real timeout (defined
		1710	as "60 seconds after the last activity"), then check if that time has
		1711	been reached, which means something I<did>, in fact, time out. Otherwise
		1712	the callback was invoked too early (C<timeout> is in the future), so
		1713	re-schedule the timer to fire at that future time, to see if maybe we have
		1714	a timeout then.
		1715
		1716	Note how C<ev_timer_again> is used, taking advantage of the
		1717	C<ev_timer_again> optimisation when the timer is already running.
		1718
		1719	This scheme causes more callback invocations (about one every 60 seconds
		1720	minus half the average time between activity), but virtually no calls to
		1721	libev to change the timeout.
		1722
		1723	To start the timer, simply initialise the watcher and set C<last_activity>
		1724	to the current time (meaning we just have some activity :), then call the
		1725	callback, which will "do the right thing" and start the timer:
		1726
		1727	ev_init (timer, callback);
		1728	last_activity = ev_now (loop);
		1729	callback (loop, timer, EV_TIMEOUT);
		1730
		1731	And when there is some activity, simply store the current time in
		1732	C<last_activity>, no libev calls at all:
		1733
		1734	last_actiivty = ev_now (loop);
		1735
		1736	This technique is slightly more complex, but in most cases where the
		1737	time-out is unlikely to be triggered, much more efficient.
		1738
		1739	Changing the timeout is trivial as well (if it isn't hard-coded in the
		1740	callback :) - just change the timeout and invoke the callback, which will
		1741	fix things for you.
		1742
		1743	=item 4. Wee, just use a double-linked list for your timeouts.
		1744
		1745	If there is not one request, but many thousands (millions...), all
		1746	employing some kind of timeout with the same timeout value, then one can
		1747	do even better:
		1748
		1749	When starting the timeout, calculate the timeout value and put the timeout
		1750	at the I<end> of the list.
		1751
		1752	Then use an C<ev_timer> to fire when the timeout at the I<beginning> of
		1753	the list is expected to fire (for example, using the technique #3).
		1754
		1755	When there is some activity, remove the timer from the list, recalculate
		1756	the timeout, append it to the end of the list again, and make sure to
		1757	update the C<ev_timer> if it was taken from the beginning of the list.
		1758
		1759	This way, one can manage an unlimited number of timeouts in O(1) time for
		1760	starting, stopping and updating the timers, at the expense of a major
		1761	complication, and having to use a constant timeout. The constant timeout
		1762	ensures that the list stays sorted.
		1763
		1764	=back
		1765
		1766	So which method the best?
		1767
		1768	Method #2 is a simple no-brain-required solution that is adequate in most
		1769	situations. Method #3 requires a bit more thinking, but handles many cases
		1770	better, and isn't very complicated either. In most case, choosing either
		1771	one is fine, with #3 being better in typical situations.
		1772
		1773	Method #1 is almost always a bad idea, and buys you nothing. Method #4 is
		1774	rather complicated, but extremely efficient, something that really pays
		1775	off after the first million or so of active timers, i.e. it's usually
		1776	overkill :)
1205		1777
1206	=head3 The special problem of time updates	1778	=head3 The special problem of time updates
1207		1779
1208	Requesting the current time is a costly operation (it usually takes at	1780	Establishing the current time is a costly operation (it usually takes at
1209	least two syscalls): EV therefore updates it's idea of the current time	1781	least two system calls): EV therefore updates its idea of the current
1210	only before and after C<ev_loop> polls for new events, which causes the	1782	time only before and after C<ev_loop> collects new events, which causes a
1211	difference between C<ev_now ()> and C<ev_time ()>.	1783	growing difference between C<ev_now ()> and C<ev_time ()> when handling
		1784	lots of events in one iteration.
1212		1785
1213	The relative timeouts are calculated relative to the C<ev_now ()>	1786	The relative timeouts are calculated relative to the C<ev_now ()>
1214	time. This is usually the right thing as this timestamp refers to the time	1787	time. This is usually the right thing as this timestamp refers to the time
1215	of the event triggering whatever timeout you are modifying/starting. If	1788	of the event triggering whatever timeout you are modifying/starting. If
1216	you suspect event processing to be delayed and you I<need> to base the	1789	you suspect event processing to be delayed and you I<need> to base the
1217	timeout on the current time, use something like this to adjust for this:	1790	timeout on the current time, use something like this to adjust for this:
1218		1791
1219	ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);	1792	ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);
1220		1793
		1794	If the event loop is suspended for a long time, you can also force an
		1795	update of the time returned by C<ev_now ()> by calling C<ev_now_update
		1796	()>.
		1797
		1798	=head3 The special problems of suspended animation
		1799
		1800	When you leave the server world it is quite customary to hit machines that
		1801	can suspend/hibernate - what happens to the clocks during such a suspend?
		1802
		1803	Some quick tests made with a Linux 2.6.28 indicate that a suspend freezes
		1804	all processes, while the clocks (C<times>, C<CLOCK_MONOTONIC>) continue
		1805	to run until the system is suspended, but they will not advance while the
		1806	system is suspended. That means, on resume, it will be as if the program
		1807	was frozen for a few seconds, but the suspend time will not be counted
		1808	towards C<ev_timer> when a monotonic clock source is used. The real time
		1809	clock advanced as expected, but if it is used as sole clocksource, then a
		1810	long suspend would be detected as a time jump by libev, and timers would
		1811	be adjusted accordingly.
		1812
		1813	I would not be surprised to see different behaviour in different between
		1814	operating systems, OS versions or even different hardware.
		1815
		1816	The other form of suspend (job control, or sending a SIGSTOP) will see a
		1817	time jump in the monotonic clocks and the realtime clock. If the program
		1818	is suspended for a very long time, and monotonic clock sources are in use,
		1819	then you can expect C<ev_timer>s to expire as the full suspension time
		1820	will be counted towards the timers. When no monotonic clock source is in
		1821	use, then libev will again assume a timejump and adjust accordingly.
		1822
		1823	It might be beneficial for this latter case to call C<ev_suspend>
		1824	and C<ev_resume> in code that handles C<SIGTSTP>, to at least get
		1825	deterministic behaviour in this case (you can do nothing against
		1826	C<SIGSTOP>).
		1827
1221	=head3 Watcher-Specific Functions and Data Members	1828	=head3 Watcher-Specific Functions and Data Members
1222		1829
1223	=over 4	1830	=over 4
1224		1831
1225	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)	1832	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)
…		…
1248	If the timer is started but non-repeating, stop it (as if it timed out).	1855	If the timer is started but non-repeating, stop it (as if it timed out).
1249		1856
1250	If the timer is repeating, either start it if necessary (with the	1857	If the timer is repeating, either start it if necessary (with the
1251	C<repeat> value), or reset the running timer to the C<repeat> value.	1858	C<repeat> value), or reset the running timer to the C<repeat> value.
1252		1859
1253	This sounds a bit complicated, but here is a useful and typical	1860	This sounds a bit complicated, see L<Be smart about timeouts>, above, for a
1254	example: Imagine you have a TCP connection and you want a so-called idle	1861	usage example.
1255	timeout, that is, you want to be called when there have been, say, 60
1256	seconds of inactivity on the socket. The easiest way to do this is to
1257	configure an C<ev_timer> with a C<repeat> value of C<60> and then call
1258	C<ev_timer_again> each time you successfully read or write some data. If
1259	you go into an idle state where you do not expect data to travel on the
1260	socket, you can C<ev_timer_stop> the timer, and C<ev_timer_again> will
1261	automatically restart it if need be.
1262		1862
1263	That means you can ignore the C<after> value and C<ev_timer_start>	1863	=item ev_tstamp ev_timer_remaining (loop, ev_timer *)
1264	altogether and only ever use the C<repeat> value and C<ev_timer_again>:
1265		1864
1266	ev_timer_init (timer, callback, 0., 5.);	1865	Returns the remaining time until a timer fires. If the timer is active,
1267	ev_timer_again (loop, timer);	1866	then this time is relative to the current event loop time, otherwise it's
1268	...	1867	the timeout value currently configured.
1269	timer->again = 17.;
1270	ev_timer_again (loop, timer);
1271	...
1272	timer->again = 10.;
1273	ev_timer_again (loop, timer);
1274		1868
1275	This is more slightly efficient then stopping/starting the timer each time	1869	That is, after an C<ev_timer_set (w, 5, 7)>, C<ev_timer_remaining> returns
1276	you want to modify its timeout value.	1870	C<5>. When the timer is started and one second passes, C<ev_timer_remain>
		1871	will return C<4>. When the timer expires and is restarted, it will return
		1872	roughly C<7> (likely slightly less as callback invocation takes some time,
		1873	too), and so on.
1277		1874
1278	=item ev_tstamp repeat [read-write]	1875	=item ev_tstamp repeat [read-write]
1279		1876
1280	The current C<repeat> value. Will be used each time the watcher times out	1877	The current C<repeat> value. Will be used each time the watcher times out
1281	or C<ev_timer_again> is called and determines the next timeout (if any),	1878	or C<ev_timer_again> is called, and determines the next timeout (if any),
1282	which is also when any modifications are taken into account.	1879	which is also when any modifications are taken into account.
1283		1880
1284	=back	1881	=back
1285		1882
1286	=head3 Examples	1883	=head3 Examples
1287		1884
1288	Example: Create a timer that fires after 60 seconds.	1885	Example: Create a timer that fires after 60 seconds.
1289		1886
1290	static void	1887	static void
1291	one_minute_cb (struct ev_loop *loop, struct ev_timer *w, int revents)	1888	one_minute_cb (struct ev_loop *loop, ev_timer *w, int revents)
1292	{	1889	{
1293	.. one minute over, w is actually stopped right here	1890	.. one minute over, w is actually stopped right here
1294	}	1891	}
1295		1892
1296	struct ev_timer mytimer;	1893	ev_timer mytimer;
1297	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);	1894	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);
1298	ev_timer_start (loop, &mytimer);	1895	ev_timer_start (loop, &mytimer);
1299		1896
1300	Example: Create a timeout timer that times out after 10 seconds of	1897	Example: Create a timeout timer that times out after 10 seconds of
1301	inactivity.	1898	inactivity.
1302		1899
1303	static void	1900	static void
1304	timeout_cb (struct ev_loop *loop, struct ev_timer *w, int revents)	1901	timeout_cb (struct ev_loop *loop, ev_timer *w, int revents)
1305	{	1902	{
1306	.. ten seconds without any activity	1903	.. ten seconds without any activity
1307	}	1904	}
1308		1905
1309	struct ev_timer mytimer;	1906	ev_timer mytimer;
1310	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */	1907	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */
1311	ev_timer_again (&mytimer); /* start timer */	1908	ev_timer_again (&mytimer); /* start timer */
1312	ev_loop (loop, 0);	1909	ev_loop (loop, 0);
1313		1910
1314	// and in some piece of code that gets executed on any "activity":	1911	// and in some piece of code that gets executed on any "activity":
…		…
1319	=head2 C<ev_periodic> - to cron or not to cron?	1916	=head2 C<ev_periodic> - to cron or not to cron?
1320		1917
1321	Periodic watchers are also timers of a kind, but they are very versatile	1918	Periodic watchers are also timers of a kind, but they are very versatile
1322	(and unfortunately a bit complex).	1919	(and unfortunately a bit complex).
1323		1920
1324	Unlike C<ev_timer>'s, they are not based on real time (or relative time)	1921	Unlike C<ev_timer>, periodic watchers are not based on real time (or
1325	but on wall clock time (absolute time). You can tell a periodic watcher	1922	relative time, the physical time that passes) but on wall clock time
1326	to trigger after some specific point in time. For example, if you tell a	1923	(absolute time, the thing you can read on your calender or clock). The
1327	periodic watcher to trigger in 10 seconds (by specifying e.g. C<ev_now ()	1924	difference is that wall clock time can run faster or slower than real
1328	+ 10.>, that is, an absolute time not a delay) and then reset your system	1925	time, and time jumps are not uncommon (e.g. when you adjust your
1329	clock to January of the previous year, then it will take more than year	1926	wrist-watch).
1330	to trigger the event (unlike an C<ev_timer>, which would still trigger
1331	roughly 10 seconds later as it uses a relative timeout).
1332		1927
		1928	You can tell a periodic watcher to trigger after some specific point
		1929	in time: for example, if you tell a periodic watcher to trigger "in 10
		1930	seconds" (by specifying e.g. C<ev_now () + 10.>, that is, an absolute time
		1931	not a delay) and then reset your system clock to January of the previous
		1932	year, then it will take a year or more to trigger the event (unlike an
		1933	C<ev_timer>, which would still trigger roughly 10 seconds after starting
		1934	it, as it uses a relative timeout).
		1935
1333	C<ev_periodic>s can also be used to implement vastly more complex timers,	1936	C<ev_periodic> watchers can also be used to implement vastly more complex
1334	such as triggering an event on each "midnight, local time", or other	1937	timers, such as triggering an event on each "midnight, local time", or
1335	complicated, rules.	1938	other complicated rules. This cannot be done with C<ev_timer> watchers, as
		1939	those cannot react to time jumps.
1336		1940
1337	As with timers, the callback is guaranteed to be invoked only when the	1941	As with timers, the callback is guaranteed to be invoked only when the
1338	time (C<at>) has passed, but if multiple periodic timers become ready	1942	point in time where it is supposed to trigger has passed. If multiple
1339	during the same loop iteration then order of execution is undefined.	1943	timers become ready during the same loop iteration then the ones with
		1944	earlier time-out values are invoked before ones with later time-out values
		1945	(but this is no longer true when a callback calls C<ev_loop> recursively).
1340		1946
1341	=head3 Watcher-Specific Functions and Data Members	1947	=head3 Watcher-Specific Functions and Data Members
1342		1948
1343	=over 4	1949	=over 4
1344		1950
1345	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)	1951	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp offset, ev_tstamp interval, reschedule_cb)
1346		1952
1347	=item ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)	1953	=item ev_periodic_set (ev_periodic *, ev_tstamp offset, ev_tstamp interval, reschedule_cb)
1348		1954
1349	Lots of arguments, lets sort it out... There are basically three modes of	1955	Lots of arguments, let's sort it out... There are basically three modes of
1350	operation, and we will explain them from simplest to complex:	1956	operation, and we will explain them from simplest to most complex:
1351		1957
1352	=over 4	1958	=over 4
1353		1959
1354	=item * absolute timer (at = time, interval = reschedule_cb = 0)	1960	=item * absolute timer (offset = absolute time, interval = 0, reschedule_cb = 0)
1355		1961
1356	In this configuration the watcher triggers an event after the wall clock	1962	In this configuration the watcher triggers an event after the wall clock
1357	time C<at> has passed and doesn't repeat. It will not adjust when a time	1963	time C<offset> has passed. It will not repeat and will not adjust when a
1358	jump occurs, that is, if it is to be run at January 1st 2011 then it will	1964	time jump occurs, that is, if it is to be run at January 1st 2011 then it
1359	run when the system time reaches or surpasses this time.	1965	will be stopped and invoked when the system clock reaches or surpasses
		1966	this point in time.
1360		1967
1361	=item * repeating interval timer (at = offset, interval > 0, reschedule_cb = 0)	1968	=item * repeating interval timer (offset = offset within interval, interval > 0, reschedule_cb = 0)
1362		1969
1363	In this mode the watcher will always be scheduled to time out at the next	1970	In this mode the watcher will always be scheduled to time out at the next
1364	C<at + N * interval> time (for some integer N, which can also be negative)	1971	C<offset + N * interval> time (for some integer N, which can also be
1365	and then repeat, regardless of any time jumps.	1972	negative) and then repeat, regardless of any time jumps. The C<offset>
		1973	argument is merely an offset into the C<interval> periods.
1366		1974
1367	This can be used to create timers that do not drift with respect to system	1975	This can be used to create timers that do not drift with respect to the
1368	time, for example, here is a C<ev_periodic> that triggers each hour, on	1976	system clock, for example, here is an C<ev_periodic> that triggers each
1369	the hour:	1977	hour, on the hour (with respect to UTC):
1370		1978
1371	ev_periodic_set (&periodic, 0., 3600., 0);	1979	ev_periodic_set (&periodic, 0., 3600., 0);
1372		1980
1373	This doesn't mean there will always be 3600 seconds in between triggers,	1981	This doesn't mean there will always be 3600 seconds in between triggers,
1374	but only that the callback will be called when the system time shows a	1982	but only that the callback will be called when the system time shows a
1375	full hour (UTC), or more correctly, when the system time is evenly divisible	1983	full hour (UTC), or more correctly, when the system time is evenly divisible
1376	by 3600.	1984	by 3600.
1377		1985
1378	Another way to think about it (for the mathematically inclined) is that	1986	Another way to think about it (for the mathematically inclined) is that
1379	C<ev_periodic> will try to run the callback in this mode at the next possible	1987	C<ev_periodic> will try to run the callback in this mode at the next possible
1380	time where C<time = at (mod interval)>, regardless of any time jumps.	1988	time where C<time = offset (mod interval)>, regardless of any time jumps.
1381		1989
1382	For numerical stability it is preferable that the C<at> value is near	1990	For numerical stability it is preferable that the C<offset> value is near
1383	C<ev_now ()> (the current time), but there is no range requirement for	1991	C<ev_now ()> (the current time), but there is no range requirement for
1384	this value, and in fact is often specified as zero.	1992	this value, and in fact is often specified as zero.
1385		1993
1386	Note also that there is an upper limit to how often a timer can fire (CPU	1994	Note also that there is an upper limit to how often a timer can fire (CPU
1387	speed for example), so if C<interval> is very small then timing stability	1995	speed for example), so if C<interval> is very small then timing stability
1388	will of course deteriorate. Libev itself tries to be exact to be about one	1996	will of course deteriorate. Libev itself tries to be exact to be about one
1389	millisecond (if the OS supports it and the machine is fast enough).	1997	millisecond (if the OS supports it and the machine is fast enough).
1390		1998
1391	=item * manual reschedule mode (at and interval ignored, reschedule_cb = callback)	1999	=item * manual reschedule mode (offset ignored, interval ignored, reschedule_cb = callback)
1392		2000
1393	In this mode the values for C<interval> and C<at> are both being	2001	In this mode the values for C<interval> and C<offset> are both being
1394	ignored. Instead, each time the periodic watcher gets scheduled, the	2002	ignored. Instead, each time the periodic watcher gets scheduled, the
1395	reschedule callback will be called with the watcher as first, and the	2003	reschedule callback will be called with the watcher as first, and the
1396	current time as second argument.	2004	current time as second argument.
1397		2005
1398	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher,	2006	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher, ever,
1399	ever, or make ANY event loop modifications whatsoever>.	2007	or make ANY other event loop modifications whatsoever, unless explicitly
		2008	allowed by documentation here>.
1400		2009
1401	If you need to stop it, return C<now + 1e30> (or so, fudge fudge) and stop	2010	If you need to stop it, return C<now + 1e30> (or so, fudge fudge) and stop
1402	it afterwards (e.g. by starting an C<ev_prepare> watcher, which is the	2011	it afterwards (e.g. by starting an C<ev_prepare> watcher, which is the
1403	only event loop modification you are allowed to do).	2012	only event loop modification you are allowed to do).
1404		2013
1405	The callback prototype is C<ev_tstamp (*reschedule_cb)(struct ev_periodic	2014	The callback prototype is C<ev_tstamp (*reschedule_cb)(ev_periodic
1406	*w, ev_tstamp now)>, e.g.:	2015	*w, ev_tstamp now)>, e.g.:
1407		2016
		2017	static ev_tstamp
1408	static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now)	2018	my_rescheduler (ev_periodic *w, ev_tstamp now)
1409	{	2019	{
1410	return now + 60.;	2020	return now + 60.;
1411	}	2021	}
1412		2022
1413	It must return the next time to trigger, based on the passed time value	2023	It must return the next time to trigger, based on the passed time value
…		…
1433	a different time than the last time it was called (e.g. in a crond like	2043	a different time than the last time it was called (e.g. in a crond like
1434	program when the crontabs have changed).	2044	program when the crontabs have changed).
1435		2045
1436	=item ev_tstamp ev_periodic_at (ev_periodic *)	2046	=item ev_tstamp ev_periodic_at (ev_periodic *)
1437		2047
1438	When active, returns the absolute time that the watcher is supposed to	2048	When active, returns the absolute time that the watcher is supposed
1439	trigger next.	2049	to trigger next. This is not the same as the C<offset> argument to
		2050	C<ev_periodic_set>, but indeed works even in interval and manual
		2051	rescheduling modes.
1440		2052
1441	=item ev_tstamp offset [read-write]	2053	=item ev_tstamp offset [read-write]
1442		2054
1443	When repeating, this contains the offset value, otherwise this is the	2055	When repeating, this contains the offset value, otherwise this is the
1444	absolute point in time (the C<at> value passed to C<ev_periodic_set>).	2056	absolute point in time (the C<offset> value passed to C<ev_periodic_set>,
		2057	although libev might modify this value for better numerical stability).
1445		2058
1446	Can be modified any time, but changes only take effect when the periodic	2059	Can be modified any time, but changes only take effect when the periodic
1447	timer fires or C<ev_periodic_again> is being called.	2060	timer fires or C<ev_periodic_again> is being called.
1448		2061
1449	=item ev_tstamp interval [read-write]	2062	=item ev_tstamp interval [read-write]
1450		2063
1451	The current interval value. Can be modified any time, but changes only	2064	The current interval value. Can be modified any time, but changes only
1452	take effect when the periodic timer fires or C<ev_periodic_again> is being	2065	take effect when the periodic timer fires or C<ev_periodic_again> is being
1453	called.	2066	called.
1454		2067
1455	=item ev_tstamp (*reschedule_cb)(struct ev_periodic *w, ev_tstamp now) [read-write]	2068	=item ev_tstamp (*reschedule_cb)(ev_periodic *w, ev_tstamp now) [read-write]
1456		2069
1457	The current reschedule callback, or C<0>, if this functionality is	2070	The current reschedule callback, or C<0>, if this functionality is
1458	switched off. Can be changed any time, but changes only take effect when	2071	switched off. Can be changed any time, but changes only take effect when
1459	the periodic timer fires or C<ev_periodic_again> is being called.	2072	the periodic timer fires or C<ev_periodic_again> is being called.
1460		2073
1461	=back	2074	=back
1462		2075
1463	=head3 Examples	2076	=head3 Examples
1464		2077
1465	Example: Call a callback every hour, or, more precisely, whenever the	2078	Example: Call a callback every hour, or, more precisely, whenever the
1466	system clock is divisible by 3600. The callback invocation times have	2079	system time is divisible by 3600. The callback invocation times have
1467	potentially a lot of jitter, but good long-term stability.	2080	potentially a lot of jitter, but good long-term stability.
1468		2081
1469	static void	2082	static void
1470	clock_cb (struct ev_loop *loop, struct ev_io *w, int revents)	2083	clock_cb (struct ev_loop *loop, ev_io *w, int revents)
1471	{	2084	{
1472	... its now a full hour (UTC, or TAI or whatever your clock follows)	2085	... its now a full hour (UTC, or TAI or whatever your clock follows)
1473	}	2086	}
1474		2087
1475	struct ev_periodic hourly_tick;	2088	ev_periodic hourly_tick;
1476	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);	2089	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);
1477	ev_periodic_start (loop, &hourly_tick);	2090	ev_periodic_start (loop, &hourly_tick);
1478		2091
1479	Example: The same as above, but use a reschedule callback to do it:	2092	Example: The same as above, but use a reschedule callback to do it:
1480		2093
1481	#include <math.h>	2094	#include <math.h>
1482		2095
1483	static ev_tstamp	2096	static ev_tstamp
1484	my_scheduler_cb (struct ev_periodic *w, ev_tstamp now)	2097	my_scheduler_cb (ev_periodic *w, ev_tstamp now)
1485	{	2098	{
1486	return fmod (now, 3600.) + 3600.;	2099	return now + (3600. - fmod (now, 3600.));
1487	}	2100	}
1488		2101
1489	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);	2102	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);
1490		2103
1491	Example: Call a callback every hour, starting now:	2104	Example: Call a callback every hour, starting now:
1492		2105
1493	struct ev_periodic hourly_tick;	2106	ev_periodic hourly_tick;
1494	ev_periodic_init (&hourly_tick, clock_cb,	2107	ev_periodic_init (&hourly_tick, clock_cb,
1495	fmod (ev_now (loop), 3600.), 3600., 0);	2108	fmod (ev_now (loop), 3600.), 3600., 0);
1496	ev_periodic_start (loop, &hourly_tick);	2109	ev_periodic_start (loop, &hourly_tick);
1497		2110
1498		2111
…		…
1501	Signal watchers will trigger an event when the process receives a specific	2114	Signal watchers will trigger an event when the process receives a specific
1502	signal one or more times. Even though signals are very asynchronous, libev	2115	signal one or more times. Even though signals are very asynchronous, libev
1503	will try it's best to deliver signals synchronously, i.e. as part of the	2116	will try it's best to deliver signals synchronously, i.e. as part of the
1504	normal event processing, like any other event.	2117	normal event processing, like any other event.
1505		2118
		2119	If you want signals to be delivered truly asynchronously, just use
		2120	C<sigaction> as you would do without libev and forget about sharing
		2121	the signal. You can even use C<ev_async> from a signal handler to
		2122	synchronously wake up an event loop.
		2123
1506	You can configure as many watchers as you like per signal. Only when the	2124	You can configure as many watchers as you like for the same signal, but
		2125	only within the same loop, i.e. you can watch for C<SIGINT> in your
		2126	default loop and for C<SIGIO> in another loop, but you cannot watch for
		2127	C<SIGINT> in both the default loop and another loop at the same time. At
		2128	the moment, C<SIGCHLD> is permanently tied to the default loop.
		2129
1507	first watcher gets started will libev actually register a signal watcher	2130	When the first watcher gets started will libev actually register something
1508	with the kernel (thus it coexists with your own signal handlers as long	2131	with the kernel (thus it coexists with your own signal handlers as long as
1509	as you don't register any with libev). Similarly, when the last signal	2132	you don't register any with libev for the same signal).
1510	watcher for a signal is stopped libev will reset the signal handler to
1511	SIG_DFL (regardless of what it was set to before).
1512		2133
1513	If possible and supported, libev will install its handlers with	2134	If possible and supported, libev will install its handlers with
1514	C<SA_RESTART> behaviour enabled, so system calls should not be unduly	2135	C<SA_RESTART> (or equivalent) behaviour enabled, so system calls should
1515	interrupted. If you have a problem with system calls getting interrupted by	2136	not be unduly interrupted. If you have a problem with system calls getting
1516	signals you can block all signals in an C<ev_check> watcher and unblock	2137	interrupted by signals you can block all signals in an C<ev_check> watcher
1517	them in an C<ev_prepare> watcher.	2138	and unblock them in an C<ev_prepare> watcher.
		2139
		2140	=head3 The special problem of inheritance over fork/execve/pthread_create
		2141
		2142	Both the signal mask (C<sigprocmask>) and the signal disposition
		2143	(C<sigaction>) are unspecified after starting a signal watcher (and after
		2144	stopping it again), that is, libev might or might not block the signal,
		2145	and might or might not set or restore the installed signal handler.
		2146
		2147	While this does not matter for the signal disposition (libev never
		2148	sets signals to C<SIG_IGN>, so handlers will be reset to C<SIG_DFL> on
		2149	C<execve>), this matters for the signal mask: many programs do not expect
		2150	certain signals to be blocked.
		2151
		2152	This means that before calling C<exec> (from the child) you should reset
		2153	the signal mask to whatever "default" you expect (all clear is a good
		2154	choice usually).
		2155
		2156	The simplest way to ensure that the signal mask is reset in the child is
		2157	to install a fork handler with C<pthread_atfork> that resets it. That will
		2158	catch fork calls done by libraries (such as the libc) as well.
		2159
		2160	In current versions of libev, the signal will not be blocked indefinitely
		2161	unless you use the C<signalfd> API (C<EV_SIGNALFD>). While this reduces
		2162	the window of opportunity for problems, it will not go away, as libev
		2163	I<has> to modify the signal mask, at least temporarily.
		2164
		2165	So I can't stress this enough I<if you do not reset your signal mask
		2166	when you expect it to be empty, you have a race condition in your
		2167	program>. This is not a libev-specific thing, this is true for most event
		2168	libraries.
1518		2169
1519	=head3 Watcher-Specific Functions and Data Members	2170	=head3 Watcher-Specific Functions and Data Members
1520		2171
1521	=over 4	2172	=over 4
1522		2173
…		…
1533		2184
1534	=back	2185	=back
1535		2186
1536	=head3 Examples	2187	=head3 Examples
1537		2188
1538	Example: Try to exit cleanly on SIGINT and SIGTERM.	2189	Example: Try to exit cleanly on SIGINT.
1539		2190
1540	static void	2191	static void
1541	sigint_cb (struct ev_loop *loop, struct ev_signal *w, int revents)	2192	sigint_cb (struct ev_loop *loop, ev_signal *w, int revents)
1542	{	2193	{
1543	ev_unloop (loop, EVUNLOOP_ALL);	2194	ev_unloop (loop, EVUNLOOP_ALL);
1544	}	2195	}
1545		2196
1546	struct ev_signal signal_watcher;	2197	ev_signal signal_watcher;
1547	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);	2198	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);
1548	ev_signal_start (loop, &sigint_cb);	2199	ev_signal_start (loop, &signal_watcher);
1549		2200
1550		2201
1551	=head2 C<ev_child> - watch out for process status changes	2202	=head2 C<ev_child> - watch out for process status changes
1552		2203
1553	Child watchers trigger when your process receives a SIGCHLD in response to	2204	Child watchers trigger when your process receives a SIGCHLD in response to
1554	some child status changes (most typically when a child of yours dies). It	2205	some child status changes (most typically when a child of yours dies or
1555	is permissible to install a child watcher I<after> the child has been	2206	exits). It is permissible to install a child watcher I<after> the child
1556	forked (which implies it might have already exited), as long as the event	2207	has been forked (which implies it might have already exited), as long
1557	loop isn't entered (or is continued from a watcher).	2208	as the event loop isn't entered (or is continued from a watcher), i.e.,
		2209	forking and then immediately registering a watcher for the child is fine,
		2210	but forking and registering a watcher a few event loop iterations later or
		2211	in the next callback invocation is not.
1558		2212
1559	Only the default event loop is capable of handling signals, and therefore	2213	Only the default event loop is capable of handling signals, and therefore
1560	you can only register child watchers in the default event loop.	2214	you can only register child watchers in the default event loop.
1561		2215
		2216	Due to some design glitches inside libev, child watchers will always be
		2217	handled at maximum priority (their priority is set to C<EV_MAXPRI> by
		2218	libev)
		2219
1562	=head3 Process Interaction	2220	=head3 Process Interaction
1563		2221
1564	Libev grabs C<SIGCHLD> as soon as the default event loop is	2222	Libev grabs C<SIGCHLD> as soon as the default event loop is
1565	initialised. This is necessary to guarantee proper behaviour even if	2223	initialised. This is necessary to guarantee proper behaviour even if the
1566	the first child watcher is started after the child exits. The occurrence	2224	first child watcher is started after the child exits. The occurrence
1567	of C<SIGCHLD> is recorded asynchronously, but child reaping is done	2225	of C<SIGCHLD> is recorded asynchronously, but child reaping is done
1568	synchronously as part of the event loop processing. Libev always reaps all	2226	synchronously as part of the event loop processing. Libev always reaps all
1569	children, even ones not watched.	2227	children, even ones not watched.
1570		2228
1571	=head3 Overriding the Built-In Processing	2229	=head3 Overriding the Built-In Processing
…		…
1581	=head3 Stopping the Child Watcher	2239	=head3 Stopping the Child Watcher
1582		2240
1583	Currently, the child watcher never gets stopped, even when the	2241	Currently, the child watcher never gets stopped, even when the
1584	child terminates, so normally one needs to stop the watcher in the	2242	child terminates, so normally one needs to stop the watcher in the
1585	callback. Future versions of libev might stop the watcher automatically	2243	callback. Future versions of libev might stop the watcher automatically
1586	when a child exit is detected.	2244	when a child exit is detected (calling C<ev_child_stop> twice is not a
		2245	problem).
1587		2246
1588	=head3 Watcher-Specific Functions and Data Members	2247	=head3 Watcher-Specific Functions and Data Members
1589		2248
1590	=over 4	2249	=over 4
1591		2250
…		…
1623	its completion.	2282	its completion.
1624		2283
1625	ev_child cw;	2284	ev_child cw;
1626		2285
1627	static void	2286	static void
1628	child_cb (EV_P_ struct ev_child *w, int revents)	2287	child_cb (EV_P_ ev_child *w, int revents)
1629	{	2288	{
1630	ev_child_stop (EV_A_ w);	2289	ev_child_stop (EV_A_ w);
1631	printf ("process %d exited with status %x\n", w->rpid, w->rstatus);	2290	printf ("process %d exited with status %x\n", w->rpid, w->rstatus);
1632	}	2291	}
1633		2292
…		…
1648		2307
1649		2308
1650	=head2 C<ev_stat> - did the file attributes just change?	2309	=head2 C<ev_stat> - did the file attributes just change?
1651		2310
1652	This watches a file system path for attribute changes. That is, it calls	2311	This watches a file system path for attribute changes. That is, it calls
1653	C<stat> regularly (or when the OS says it changed) and sees if it changed	2312	C<stat> on that path in regular intervals (or when the OS says it changed)
1654	compared to the last time, invoking the callback if it did.	2313	and sees if it changed compared to the last time, invoking the callback if
		2314	it did.
1655		2315
1656	The path does not need to exist: changing from "path exists" to "path does	2316	The path does not need to exist: changing from "path exists" to "path does
1657	not exist" is a status change like any other. The condition "path does	2317	not exist" is a status change like any other. The condition "path does not
1658	not exist" is signified by the C<st_nlink> field being zero (which is	2318	exist" (or more correctly "path cannot be stat'ed") is signified by the
1659	otherwise always forced to be at least one) and all the other fields of	2319	C<st_nlink> field being zero (which is otherwise always forced to be at
1660	the stat buffer having unspecified contents.	2320	least one) and all the other fields of the stat buffer having unspecified
		2321	contents.
1661		2322
1662	The path I<should> be absolute and I<must not> end in a slash. If it is	2323	The path I<must not> end in a slash or contain special components such as
		2324	C<.> or C<..>. The path I<should> be absolute: If it is relative and
1663	relative and your working directory changes, the behaviour is undefined.	2325	your working directory changes, then the behaviour is undefined.
1664		2326
1665	Since there is no standard to do this, the portable implementation simply	2327	Since there is no portable change notification interface available, the
1666	calls C<stat (2)> regularly on the path to see if it changed somehow. You	2328	portable implementation simply calls C<stat(2)> regularly on the path
1667	can specify a recommended polling interval for this case. If you specify	2329	to see if it changed somehow. You can specify a recommended polling
1668	a polling interval of C<0> (highly recommended!) then a I<suitable,	2330	interval for this case. If you specify a polling interval of C<0> (highly
1669	unspecified default> value will be used (which you can expect to be around	2331	recommended!) then a I<suitable, unspecified default> value will be used
1670	five seconds, although this might change dynamically). Libev will also	2332	(which you can expect to be around five seconds, although this might
1671	impose a minimum interval which is currently around C<0.1>, but thats	2333	change dynamically). Libev will also impose a minimum interval which is
1672	usually overkill.	2334	currently around C<0.1>, but that's usually overkill.
1673		2335
1674	This watcher type is not meant for massive numbers of stat watchers,	2336	This watcher type is not meant for massive numbers of stat watchers,
1675	as even with OS-supported change notifications, this can be	2337	as even with OS-supported change notifications, this can be
1676	resource-intensive.	2338	resource-intensive.
1677		2339
1678	At the time of this writing, only the Linux inotify interface is	2340	At the time of this writing, the only OS-specific interface implemented
1679	implemented (implementing kqueue support is left as an exercise for the	2341	is the Linux inotify interface (implementing kqueue support is left as an
1680	reader, note, however, that the author sees no way of implementing ev_stat	2342	exercise for the reader. Note, however, that the author sees no way of
1681	semantics with kqueue). Inotify will be used to give hints only and should	2343	implementing C<ev_stat> semantics with kqueue, except as a hint).
1682	not change the semantics of C<ev_stat> watchers, which means that libev
1683	sometimes needs to fall back to regular polling again even with inotify,
1684	but changes are usually detected immediately, and if the file exists there
1685	will be no polling.
1686		2344
1687	=head3 ABI Issues (Largefile Support)	2345	=head3 ABI Issues (Largefile Support)
1688		2346
1689	Libev by default (unless the user overrides this) uses the default	2347	Libev by default (unless the user overrides this) uses the default
1690	compilation environment, which means that on systems with large file	2348	compilation environment, which means that on systems with large file
1691	support disabled by default, you get the 32 bit version of the stat	2349	support disabled by default, you get the 32 bit version of the stat
1692	structure. When using the library from programs that change the ABI to	2350	structure. When using the library from programs that change the ABI to
1693	use 64 bit file offsets the programs will fail. In that case you have to	2351	use 64 bit file offsets the programs will fail. In that case you have to
1694	compile libev with the same flags to get binary compatibility. This is	2352	compile libev with the same flags to get binary compatibility. This is
1695	obviously the case with any flags that change the ABI, but the problem is	2353	obviously the case with any flags that change the ABI, but the problem is
1696	most noticeably disabled with ev_stat and large file support.	2354	most noticeably displayed with ev_stat and large file support.
1697		2355
1698	The solution for this is to lobby your distribution maker to make large	2356	The solution for this is to lobby your distribution maker to make large
1699	file interfaces available by default (as e.g. FreeBSD does) and not	2357	file interfaces available by default (as e.g. FreeBSD does) and not
1700	optional. Libev cannot simply switch on large file support because it has	2358	optional. Libev cannot simply switch on large file support because it has
1701	to exchange stat structures with application programs compiled using the	2359	to exchange stat structures with application programs compiled using the
1702	default compilation environment.	2360	default compilation environment.
1703		2361
1704	=head3 Inotify	2362	=head3 Inotify and Kqueue
1705		2363
1706	When C<inotify (7)> support has been compiled into libev (generally only	2364	When C<inotify (7)> support has been compiled into libev and present at
1707	available on Linux) and present at runtime, it will be used to speed up	2365	runtime, it will be used to speed up change detection where possible. The
1708	change detection where possible. The inotify descriptor will be created lazily	2366	inotify descriptor will be created lazily when the first C<ev_stat>
1709	when the first C<ev_stat> watcher is being started.	2367	watcher is being started.
1710		2368
1711	Inotify presence does not change the semantics of C<ev_stat> watchers	2369	Inotify presence does not change the semantics of C<ev_stat> watchers
1712	except that changes might be detected earlier, and in some cases, to avoid	2370	except that changes might be detected earlier, and in some cases, to avoid
1713	making regular C<stat> calls. Even in the presence of inotify support	2371	making regular C<stat> calls. Even in the presence of inotify support
1714	there are many cases where libev has to resort to regular C<stat> polling.	2372	there are many cases where libev has to resort to regular C<stat> polling,
		2373	but as long as kernel 2.6.25 or newer is used (2.6.24 and older have too
		2374	many bugs), the path exists (i.e. stat succeeds), and the path resides on
		2375	a local filesystem (libev currently assumes only ext2/3, jfs, reiserfs and
		2376	xfs are fully working) libev usually gets away without polling.
1715		2377
1716	(There is no support for kqueue, as apparently it cannot be used to	2378	There is no support for kqueue, as apparently it cannot be used to
1717	implement this functionality, due to the requirement of having a file	2379	implement this functionality, due to the requirement of having a file
1718	descriptor open on the object at all times).	2380	descriptor open on the object at all times, and detecting renames, unlinks
		2381	etc. is difficult.
		2382
		2383	=head3 C<stat ()> is a synchronous operation
		2384
		2385	Libev doesn't normally do any kind of I/O itself, and so is not blocking
		2386	the process. The exception are C<ev_stat> watchers - those call C<stat
		2387	()>, which is a synchronous operation.
		2388
		2389	For local paths, this usually doesn't matter: unless the system is very
		2390	busy or the intervals between stat's are large, a stat call will be fast,
		2391	as the path data is usually in memory already (except when starting the
		2392	watcher).
		2393
		2394	For networked file systems, calling C<stat ()> can block an indefinite
		2395	time due to network issues, and even under good conditions, a stat call
		2396	often takes multiple milliseconds.
		2397
		2398	Therefore, it is best to avoid using C<ev_stat> watchers on networked
		2399	paths, although this is fully supported by libev.
1719		2400
1720	=head3 The special problem of stat time resolution	2401	=head3 The special problem of stat time resolution
1721		2402
1722	The C<stat ()> system call only supports full-second resolution portably, and	2403	The C<stat ()> system call only supports full-second resolution portably,
1723	even on systems where the resolution is higher, many file systems still	2404	and even on systems where the resolution is higher, most file systems
1724	only support whole seconds.	2405	still only support whole seconds.
1725		2406
1726	That means that, if the time is the only thing that changes, you can	2407	That means that, if the time is the only thing that changes, you can
1727	easily miss updates: on the first update, C<ev_stat> detects a change and	2408	easily miss updates: on the first update, C<ev_stat> detects a change and
1728	calls your callback, which does something. When there is another update	2409	calls your callback, which does something. When there is another update
1729	within the same second, C<ev_stat> will be unable to detect it as the stat	2410	within the same second, C<ev_stat> will be unable to detect unless the
1730	data does not change.	2411	stat data does change in other ways (e.g. file size).
1731		2412
1732	The solution to this is to delay acting on a change for slightly more	2413	The solution to this is to delay acting on a change for slightly more
1733	than a second (or till slightly after the next full second boundary), using	2414	than a second (or till slightly after the next full second boundary), using
1734	a roughly one-second-delay C<ev_timer> (e.g. C<ev_timer_set (w, 0., 1.02);	2415	a roughly one-second-delay C<ev_timer> (e.g. C<ev_timer_set (w, 0., 1.02);
1735	ev_timer_again (loop, w)>).	2416	ev_timer_again (loop, w)>).
…		…
1755	C<path>. The C<interval> is a hint on how quickly a change is expected to	2436	C<path>. The C<interval> is a hint on how quickly a change is expected to
1756	be detected and should normally be specified as C<0> to let libev choose	2437	be detected and should normally be specified as C<0> to let libev choose
1757	a suitable value. The memory pointed to by C<path> must point to the same	2438	a suitable value. The memory pointed to by C<path> must point to the same
1758	path for as long as the watcher is active.	2439	path for as long as the watcher is active.
1759		2440
1760	The callback will receive C<EV_STAT> when a change was detected, relative	2441	The callback will receive an C<EV_STAT> event when a change was detected,
1761	to the attributes at the time the watcher was started (or the last change	2442	relative to the attributes at the time the watcher was started (or the
1762	was detected).	2443	last change was detected).
1763		2444
1764	=item ev_stat_stat (loop, ev_stat *)	2445	=item ev_stat_stat (loop, ev_stat *)
1765		2446
1766	Updates the stat buffer immediately with new values. If you change the	2447	Updates the stat buffer immediately with new values. If you change the
1767	watched path in your callback, you could call this function to avoid	2448	watched path in your callback, you could call this function to avoid
…		…
1850		2531
1851		2532
1852	=head2 C<ev_idle> - when you've got nothing better to do...	2533	=head2 C<ev_idle> - when you've got nothing better to do...
1853		2534
1854	Idle watchers trigger events when no other events of the same or higher	2535	Idle watchers trigger events when no other events of the same or higher
1855	priority are pending (prepare, check and other idle watchers do not	2536	priority are pending (prepare, check and other idle watchers do not count
1856	count).	2537	as receiving "events").
1857		2538
1858	That is, as long as your process is busy handling sockets or timeouts	2539	That is, as long as your process is busy handling sockets or timeouts
1859	(or even signals, imagine) of the same or higher priority it will not be	2540	(or even signals, imagine) of the same or higher priority it will not be
1860	triggered. But when your process is idle (or only lower-priority watchers	2541	triggered. But when your process is idle (or only lower-priority watchers
1861	are pending), the idle watchers are being called once per event loop	2542	are pending), the idle watchers are being called once per event loop
…		…
1872		2553
1873	=head3 Watcher-Specific Functions and Data Members	2554	=head3 Watcher-Specific Functions and Data Members
1874		2555
1875	=over 4	2556	=over 4
1876		2557
1877	=item ev_idle_init (ev_signal *, callback)	2558	=item ev_idle_init (ev_idle *, callback)
1878		2559
1879	Initialises and configures the idle watcher - it has no parameters of any	2560	Initialises and configures the idle watcher - it has no parameters of any
1880	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,	2561	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,
1881	believe me.	2562	believe me.
1882		2563
…		…
1886		2567
1887	Example: Dynamically allocate an C<ev_idle> watcher, start it, and in the	2568	Example: Dynamically allocate an C<ev_idle> watcher, start it, and in the
1888	callback, free it. Also, use no error checking, as usual.	2569	callback, free it. Also, use no error checking, as usual.
1889		2570
1890	static void	2571	static void
1891	idle_cb (struct ev_loop *loop, struct ev_idle *w, int revents)	2572	idle_cb (struct ev_loop *loop, ev_idle *w, int revents)
1892	{	2573	{
1893	free (w);	2574	free (w);
1894	// now do something you wanted to do when the program has	2575	// now do something you wanted to do when the program has
1895	// no longer anything immediate to do.	2576	// no longer anything immediate to do.
1896	}	2577	}
1897		2578
1898	struct ev_idle *idle_watcher = malloc (sizeof (struct ev_idle));	2579	ev_idle *idle_watcher = malloc (sizeof (ev_idle));
1899	ev_idle_init (idle_watcher, idle_cb);	2580	ev_idle_init (idle_watcher, idle_cb);
1900	ev_idle_start (loop, idle_cb);	2581	ev_idle_start (loop, idle_watcher);
1901		2582
1902		2583
1903	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!	2584	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!
1904		2585
1905	Prepare and check watchers are usually (but not always) used in tandem:	2586	Prepare and check watchers are usually (but not always) used in pairs:
1906	prepare watchers get invoked before the process blocks and check watchers	2587	prepare watchers get invoked before the process blocks and check watchers
1907	afterwards.	2588	afterwards.
1908		2589
1909	You I<must not> call C<ev_loop> or similar functions that enter	2590	You I<must not> call C<ev_loop> or similar functions that enter
1910	the current event loop from either C<ev_prepare> or C<ev_check>	2591	the current event loop from either C<ev_prepare> or C<ev_check>
…		…
1913	those watchers, i.e. the sequence will always be C<ev_prepare>, blocking,	2594	those watchers, i.e. the sequence will always be C<ev_prepare>, blocking,
1914	C<ev_check> so if you have one watcher of each kind they will always be	2595	C<ev_check> so if you have one watcher of each kind they will always be
1915	called in pairs bracketing the blocking call.	2596	called in pairs bracketing the blocking call.
1916		2597
1917	Their main purpose is to integrate other event mechanisms into libev and	2598	Their main purpose is to integrate other event mechanisms into libev and
1918	their use is somewhat advanced. This could be used, for example, to track	2599	their use is somewhat advanced. They could be used, for example, to track
1919	variable changes, implement your own watchers, integrate net-snmp or a	2600	variable changes, implement your own watchers, integrate net-snmp or a
1920	coroutine library and lots more. They are also occasionally useful if	2601	coroutine library and lots more. They are also occasionally useful if
1921	you cache some data and want to flush it before blocking (for example,	2602	you cache some data and want to flush it before blocking (for example,
1922	in X programs you might want to do an C<XFlush ()> in an C<ev_prepare>	2603	in X programs you might want to do an C<XFlush ()> in an C<ev_prepare>
1923	watcher).	2604	watcher).
1924		2605
1925	This is done by examining in each prepare call which file descriptors need	2606	This is done by examining in each prepare call which file descriptors
1926	to be watched by the other library, registering C<ev_io> watchers for	2607	need to be watched by the other library, registering C<ev_io> watchers
1927	them and starting an C<ev_timer> watcher for any timeouts (many libraries	2608	for them and starting an C<ev_timer> watcher for any timeouts (many
1928	provide just this functionality). Then, in the check watcher you check for	2609	libraries provide exactly this functionality). Then, in the check watcher,
1929	any events that occurred (by checking the pending status of all watchers	2610	you check for any events that occurred (by checking the pending status
1930	and stopping them) and call back into the library. The I/O and timer	2611	of all watchers and stopping them) and call back into the library. The
1931	callbacks will never actually be called (but must be valid nevertheless,	2612	I/O and timer callbacks will never actually be called (but must be valid
1932	because you never know, you know?).	2613	nevertheless, because you never know, you know?).
1933		2614
1934	As another example, the Perl Coro module uses these hooks to integrate	2615	As another example, the Perl Coro module uses these hooks to integrate
1935	coroutines into libev programs, by yielding to other active coroutines	2616	coroutines into libev programs, by yielding to other active coroutines
1936	during each prepare and only letting the process block if no coroutines	2617	during each prepare and only letting the process block if no coroutines
1937	are ready to run (it's actually more complicated: it only runs coroutines	2618	are ready to run (it's actually more complicated: it only runs coroutines
…		…
1940	loop from blocking if lower-priority coroutines are active, thus mapping	2621	loop from blocking if lower-priority coroutines are active, thus mapping
1941	low-priority coroutines to idle/background tasks).	2622	low-priority coroutines to idle/background tasks).
1942		2623
1943	It is recommended to give C<ev_check> watchers highest (C<EV_MAXPRI>)	2624	It is recommended to give C<ev_check> watchers highest (C<EV_MAXPRI>)
1944	priority, to ensure that they are being run before any other watchers	2625	priority, to ensure that they are being run before any other watchers
		2626	after the poll (this doesn't matter for C<ev_prepare> watchers).
		2627
1945	after the poll. Also, C<ev_check> watchers (and C<ev_prepare> watchers,	2628	Also, C<ev_check> watchers (and C<ev_prepare> watchers, too) should not
1946	too) should not activate ("feed") events into libev. While libev fully	2629	activate ("feed") events into libev. While libev fully supports this, they
1947	supports this, they might get executed before other C<ev_check> watchers	2630	might get executed before other C<ev_check> watchers did their job. As
1948	did their job. As C<ev_check> watchers are often used to embed other	2631	C<ev_check> watchers are often used to embed other (non-libev) event
1949	(non-libev) event loops those other event loops might be in an unusable	2632	loops those other event loops might be in an unusable state until their
1950	state until their C<ev_check> watcher ran (always remind yourself to	2633	C<ev_check> watcher ran (always remind yourself to coexist peacefully with
1951	coexist peacefully with others).	2634	others).
1952		2635
1953	=head3 Watcher-Specific Functions and Data Members	2636	=head3 Watcher-Specific Functions and Data Members
1954		2637
1955	=over 4	2638	=over 4
1956		2639
…		…
1958		2641
1959	=item ev_check_init (ev_check *, callback)	2642	=item ev_check_init (ev_check *, callback)
1960		2643
1961	Initialises and configures the prepare or check watcher - they have no	2644	Initialises and configures the prepare or check watcher - they have no
1962	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>	2645	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>
1963	macros, but using them is utterly, utterly and completely pointless.	2646	macros, but using them is utterly, utterly, utterly and completely
		2647	pointless.
1964		2648
1965	=back	2649	=back
1966		2650
1967	=head3 Examples	2651	=head3 Examples
1968		2652
…		…
1981		2665
1982	static ev_io iow [nfd];	2666	static ev_io iow [nfd];
1983	static ev_timer tw;	2667	static ev_timer tw;
1984		2668
1985	static void	2669	static void
1986	io_cb (ev_loop *loop, ev_io *w, int revents)	2670	io_cb (struct ev_loop *loop, ev_io *w, int revents)
1987	{	2671	{
1988	}	2672	}
1989		2673
1990	// create io watchers for each fd and a timer before blocking	2674	// create io watchers for each fd and a timer before blocking
1991	static void	2675	static void
1992	adns_prepare_cb (ev_loop *loop, ev_prepare *w, int revents)	2676	adns_prepare_cb (struct ev_loop *loop, ev_prepare *w, int revents)
1993	{	2677	{
1994	int timeout = 3600000;	2678	int timeout = 3600000;
1995	struct pollfd fds [nfd];	2679	struct pollfd fds [nfd];
1996	// actual code will need to loop here and realloc etc.	2680	// actual code will need to loop here and realloc etc.
1997	adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));	2681	adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));
1998		2682
1999	/* the callback is illegal, but won't be called as we stop during check */	2683	/* the callback is illegal, but won't be called as we stop during check */
2000	ev_timer_init (&tw, 0, timeout * 1e-3);	2684	ev_timer_init (&tw, 0, timeout * 1e-3, 0.);
2001	ev_timer_start (loop, &tw);	2685	ev_timer_start (loop, &tw);
2002		2686
2003	// create one ev_io per pollfd	2687	// create one ev_io per pollfd
2004	for (int i = 0; i < nfd; ++i)	2688	for (int i = 0; i < nfd; ++i)
2005	{	2689	{
…		…
2012	}	2696	}
2013	}	2697	}
2014		2698
2015	// stop all watchers after blocking	2699	// stop all watchers after blocking
2016	static void	2700	static void
2017	adns_check_cb (ev_loop *loop, ev_check *w, int revents)	2701	adns_check_cb (struct ev_loop *loop, ev_check *w, int revents)
2018	{	2702	{
2019	ev_timer_stop (loop, &tw);	2703	ev_timer_stop (loop, &tw);
2020		2704
2021	for (int i = 0; i < nfd; ++i)	2705	for (int i = 0; i < nfd; ++i)
2022	{	2706	{
…		…
2061	}	2745	}
2062		2746
2063	// do not ever call adns_afterpoll	2747	// do not ever call adns_afterpoll
2064		2748
2065	Method 4: Do not use a prepare or check watcher because the module you	2749	Method 4: Do not use a prepare or check watcher because the module you
2066	want to embed is too inflexible to support it. Instead, you can override	2750	want to embed is not flexible enough to support it. Instead, you can
2067	their poll function. The drawback with this solution is that the main	2751	override their poll function. The drawback with this solution is that the
2068	loop is now no longer controllable by EV. The C<Glib::EV> module does	2752	main loop is now no longer controllable by EV. The C<Glib::EV> module uses
2069	this.	2753	this approach, effectively embedding EV as a client into the horrible
		2754	libglib event loop.
2070		2755
2071	static gint	2756	static gint
2072	event_poll_func (GPollFD *fds, guint nfds, gint timeout)	2757	event_poll_func (GPollFD *fds, guint nfds, gint timeout)
2073	{	2758	{
2074	int got_events = 0;	2759	int got_events = 0;
…		…
2105	prioritise I/O.	2790	prioritise I/O.
2106		2791
2107	As an example for a bug workaround, the kqueue backend might only support	2792	As an example for a bug workaround, the kqueue backend might only support
2108	sockets on some platform, so it is unusable as generic backend, but you	2793	sockets on some platform, so it is unusable as generic backend, but you
2109	still want to make use of it because you have many sockets and it scales	2794	still want to make use of it because you have many sockets and it scales
2110	so nicely. In this case, you would create a kqueue-based loop and embed it	2795	so nicely. In this case, you would create a kqueue-based loop and embed
2111	into your default loop (which might use e.g. poll). Overall operation will	2796	it into your default loop (which might use e.g. poll). Overall operation
2112	be a bit slower because first libev has to poll and then call kevent, but	2797	will be a bit slower because first libev has to call C<poll> and then
2113	at least you can use both at what they are best.	2798	C<kevent>, but at least you can use both mechanisms for what they are
		2799	best: C<kqueue> for scalable sockets and C<poll> if you want it to work :)
2114		2800
2115	As for prioritising I/O: rarely you have the case where some fds have	2801	As for prioritising I/O: under rare circumstances you have the case where
2116	to be watched and handled very quickly (with low latency), and even	2802	some fds have to be watched and handled very quickly (with low latency),
2117	priorities and idle watchers might have too much overhead. In this case	2803	and even priorities and idle watchers might have too much overhead. In
2118	you would put all the high priority stuff in one loop and all the rest in	2804	this case you would put all the high priority stuff in one loop and all
2119	a second one, and embed the second one in the first.	2805	the rest in a second one, and embed the second one in the first.
2120		2806
2121	As long as the watcher is active, the callback will be invoked every time	2807	As long as the watcher is active, the callback will be invoked every
2122	there might be events pending in the embedded loop. The callback must then	2808	time there might be events pending in the embedded loop. The callback
2123	call C<ev_embed_sweep (mainloop, watcher)> to make a single sweep and invoke	2809	must then call C<ev_embed_sweep (mainloop, watcher)> to make a single
2124	their callbacks (you could also start an idle watcher to give the embedded	2810	sweep and invoke their callbacks (the callback doesn't need to invoke the
2125	loop strictly lower priority for example). You can also set the callback	2811	C<ev_embed_sweep> function directly, it could also start an idle watcher
2126	to C<0>, in which case the embed watcher will automatically execute the	2812	to give the embedded loop strictly lower priority for example).
2127	embedded loop sweep.
2128		2813
2129	As long as the watcher is started it will automatically handle events. The	2814	You can also set the callback to C<0>, in which case the embed watcher
2130	callback will be invoked whenever some events have been handled. You can	2815	will automatically execute the embedded loop sweep whenever necessary.
2131	set the callback to C<0> to avoid having to specify one if you are not
2132	interested in that.
2133		2816
2134	Also, there have not currently been made special provisions for forking:	2817	Fork detection will be handled transparently while the C<ev_embed> watcher
2135	when you fork, you not only have to call C<ev_loop_fork> on both loops,	2818	is active, i.e., the embedded loop will automatically be forked when the
2136	but you will also have to stop and restart any C<ev_embed> watchers	2819	embedding loop forks. In other cases, the user is responsible for calling
2137	yourself.	2820	C<ev_loop_fork> on the embedded loop.
2138		2821
2139	Unfortunately, not all backends are embeddable, only the ones returned by	2822	Unfortunately, not all backends are embeddable: only the ones returned by
2140	C<ev_embeddable_backends> are, which, unfortunately, does not include any	2823	C<ev_embeddable_backends> are, which, unfortunately, does not include any
2141	portable one.	2824	portable one.
2142		2825
2143	So when you want to use this feature you will always have to be prepared	2826	So when you want to use this feature you will always have to be prepared
2144	that you cannot get an embeddable loop. The recommended way to get around	2827	that you cannot get an embeddable loop. The recommended way to get around
2145	this is to have a separate variables for your embeddable loop, try to	2828	this is to have a separate variables for your embeddable loop, try to
2146	create it, and if that fails, use the normal loop for everything.	2829	create it, and if that fails, use the normal loop for everything.
		2830
		2831	=head3 C<ev_embed> and fork
		2832
		2833	While the C<ev_embed> watcher is running, forks in the embedding loop will
		2834	automatically be applied to the embedded loop as well, so no special
		2835	fork handling is required in that case. When the watcher is not running,
		2836	however, it is still the task of the libev user to call C<ev_loop_fork ()>
		2837	as applicable.
2147		2838
2148	=head3 Watcher-Specific Functions and Data Members	2839	=head3 Watcher-Specific Functions and Data Members
2149		2840
2150	=over 4	2841	=over 4
2151		2842
…		…
2179	C<loop_lo> (which is C<loop_hi> in the case no embeddable loop can be	2870	C<loop_lo> (which is C<loop_hi> in the case no embeddable loop can be
2180	used).	2871	used).
2181		2872
2182	struct ev_loop *loop_hi = ev_default_init (0);	2873	struct ev_loop *loop_hi = ev_default_init (0);
2183	struct ev_loop *loop_lo = 0;	2874	struct ev_loop *loop_lo = 0;
2184	struct ev_embed embed;	2875	ev_embed embed;
2185		2876
2186	// see if there is a chance of getting one that works	2877	// see if there is a chance of getting one that works
2187	// (remember that a flags value of 0 means autodetection)	2878	// (remember that a flags value of 0 means autodetection)
2188	loop_lo = ev_embeddable_backends () & ev_recommended_backends ()	2879	loop_lo = ev_embeddable_backends () & ev_recommended_backends ()
2189	? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())	2880	? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())
…		…
2203	kqueue implementation). Store the kqueue/socket-only event loop in	2894	kqueue implementation). Store the kqueue/socket-only event loop in
2204	C<loop_socket>. (One might optionally use C<EVFLAG_NOENV>, too).	2895	C<loop_socket>. (One might optionally use C<EVFLAG_NOENV>, too).
2205		2896
2206	struct ev_loop *loop = ev_default_init (0);	2897	struct ev_loop *loop = ev_default_init (0);
2207	struct ev_loop *loop_socket = 0;	2898	struct ev_loop *loop_socket = 0;
2208	struct ev_embed embed;	2899	ev_embed embed;
2209		2900
2210	if (ev_supported_backends () & ~ev_recommended_backends () & EVBACKEND_KQUEUE)	2901	if (ev_supported_backends () & ~ev_recommended_backends () & EVBACKEND_KQUEUE)
2211	if ((loop_socket = ev_loop_new (EVBACKEND_KQUEUE))	2902	if ((loop_socket = ev_loop_new (EVBACKEND_KQUEUE))
2212	{	2903	{
2213	ev_embed_init (&embed, 0, loop_socket);	2904	ev_embed_init (&embed, 0, loop_socket);
…		…
2228	event loop blocks next and before C<ev_check> watchers are being called,	2919	event loop blocks next and before C<ev_check> watchers are being called,
2229	and only in the child after the fork. If whoever good citizen calling	2920	and only in the child after the fork. If whoever good citizen calling
2230	C<ev_default_fork> cheats and calls it in the wrong process, the fork	2921	C<ev_default_fork> cheats and calls it in the wrong process, the fork
2231	handlers will be invoked, too, of course.	2922	handlers will be invoked, too, of course.
2232		2923
		2924	=head3 The special problem of life after fork - how is it possible?
		2925
		2926	Most uses of C<fork()> consist of forking, then some simple calls to ste
		2927	up/change the process environment, followed by a call to C<exec()>. This
		2928	sequence should be handled by libev without any problems.
		2929
		2930	This changes when the application actually wants to do event handling
		2931	in the child, or both parent in child, in effect "continuing" after the
		2932	fork.
		2933
		2934	The default mode of operation (for libev, with application help to detect
		2935	forks) is to duplicate all the state in the child, as would be expected
		2936	when I<either> the parent I<or> the child process continues.
		2937
		2938	When both processes want to continue using libev, then this is usually the
		2939	wrong result. In that case, usually one process (typically the parent) is
		2940	supposed to continue with all watchers in place as before, while the other
		2941	process typically wants to start fresh, i.e. without any active watchers.
		2942
		2943	The cleanest and most efficient way to achieve that with libev is to
		2944	simply create a new event loop, which of course will be "empty", and
		2945	use that for new watchers. This has the advantage of not touching more
		2946	memory than necessary, and thus avoiding the copy-on-write, and the
		2947	disadvantage of having to use multiple event loops (which do not support
		2948	signal watchers).
		2949
		2950	When this is not possible, or you want to use the default loop for
		2951	other reasons, then in the process that wants to start "fresh", call
		2952	C<ev_default_destroy ()> followed by C<ev_default_loop (...)>. Destroying
		2953	the default loop will "orphan" (not stop) all registered watchers, so you
		2954	have to be careful not to execute code that modifies those watchers. Note
		2955	also that in that case, you have to re-register any signal watchers.
		2956
2233	=head3 Watcher-Specific Functions and Data Members	2957	=head3 Watcher-Specific Functions and Data Members
2234		2958
2235	=over 4	2959	=over 4
2236		2960
2237	=item ev_fork_init (ev_signal *, callback)	2961	=item ev_fork_init (ev_signal *, callback)
…		…
2266	=head3 Queueing	2990	=head3 Queueing
2267		2991
2268	C<ev_async> does not support queueing of data in any way. The reason	2992	C<ev_async> does not support queueing of data in any way. The reason
2269	is that the author does not know of a simple (or any) algorithm for a	2993	is that the author does not know of a simple (or any) algorithm for a
2270	multiple-writer-single-reader queue that works in all cases and doesn't	2994	multiple-writer-single-reader queue that works in all cases and doesn't
2271	need elaborate support such as pthreads.	2995	need elaborate support such as pthreads or unportable memory access
		2996	semantics.
2272		2997
2273	That means that if you want to queue data, you have to provide your own	2998	That means that if you want to queue data, you have to provide your own
2274	queue. But at least I can tell you would implement locking around your	2999	queue. But at least I can tell you how to implement locking around your
2275	queue:	3000	queue:
2276		3001
2277	=over 4	3002	=over 4
2278		3003
2279	=item queueing from a signal handler context	3004	=item queueing from a signal handler context
2280		3005
2281	To implement race-free queueing, you simply add to the queue in the signal	3006	To implement race-free queueing, you simply add to the queue in the signal
2282	handler but you block the signal handler in the watcher callback. Here is an example that does that for	3007	handler but you block the signal handler in the watcher callback. Here is
2283	some fictitious SIGUSR1 handler:	3008	an example that does that for some fictitious SIGUSR1 handler:
2284		3009
2285	static ev_async mysig;	3010	static ev_async mysig;
2286		3011
2287	static void	3012	static void
2288	sigusr1_handler (void)	3013	sigusr1_handler (void)
…		…
2354	=over 4	3079	=over 4
2355		3080
2356	=item ev_async_init (ev_async *, callback)	3081	=item ev_async_init (ev_async *, callback)
2357		3082
2358	Initialises and configures the async watcher - it has no parameters of any	3083	Initialises and configures the async watcher - it has no parameters of any
2359	kind. There is a C<ev_asynd_set> macro, but using it is utterly pointless,	3084	kind. There is a C<ev_async_set> macro, but using it is utterly pointless,
2360	believe me.	3085	trust me.
2361		3086
2362	=item ev_async_send (loop, ev_async *)	3087	=item ev_async_send (loop, ev_async *)
2363		3088
2364	Sends/signals/activates the given C<ev_async> watcher, that is, feeds	3089	Sends/signals/activates the given C<ev_async> watcher, that is, feeds
2365	an C<EV_ASYNC> event on the watcher into the event loop. Unlike	3090	an C<EV_ASYNC> event on the watcher into the event loop. Unlike
2366	C<ev_feed_event>, this call is safe to do in other threads, signal or	3091	C<ev_feed_event>, this call is safe to do from other threads, signal or
2367	similar contexts (see the discussion of C<EV_ATOMIC_T> in the embedding	3092	similar contexts (see the discussion of C<EV_ATOMIC_T> in the embedding
2368	section below on what exactly this means).	3093	section below on what exactly this means).
2369		3094
		3095	Note that, as with other watchers in libev, multiple events might get
		3096	compressed into a single callback invocation (another way to look at this
		3097	is that C<ev_async> watchers are level-triggered, set on C<ev_async_send>,
		3098	reset when the event loop detects that).
		3099
2370	This call incurs the overhead of a system call only once per loop iteration,	3100	This call incurs the overhead of a system call only once per event loop
2371	so while the overhead might be noticeable, it doesn't apply to repeated	3101	iteration, so while the overhead might be noticeable, it doesn't apply to
2372	calls to C<ev_async_send>.	3102	repeated calls to C<ev_async_send> for the same event loop.
2373		3103
2374	=item bool = ev_async_pending (ev_async *)	3104	=item bool = ev_async_pending (ev_async *)
2375		3105
2376	Returns a non-zero value when C<ev_async_send> has been called on the	3106	Returns a non-zero value when C<ev_async_send> has been called on the
2377	watcher but the event has not yet been processed (or even noted) by the	3107	watcher but the event has not yet been processed (or even noted) by the
…		…
2380	C<ev_async_send> sets a flag in the watcher and wakes up the loop. When	3110	C<ev_async_send> sets a flag in the watcher and wakes up the loop. When
2381	the loop iterates next and checks for the watcher to have become active,	3111	the loop iterates next and checks for the watcher to have become active,
2382	it will reset the flag again. C<ev_async_pending> can be used to very	3112	it will reset the flag again. C<ev_async_pending> can be used to very
2383	quickly check whether invoking the loop might be a good idea.	3113	quickly check whether invoking the loop might be a good idea.
2384		3114
2385	Not that this does I<not> check whether the watcher itself is pending, only	3115	Not that this does I<not> check whether the watcher itself is pending,
2386	whether it has been requested to make this watcher pending.	3116	only whether it has been requested to make this watcher pending: there
		3117	is a time window between the event loop checking and resetting the async
		3118	notification, and the callback being invoked.
2387		3119
2388	=back	3120	=back
2389		3121
2390		3122
2391	=head1 OTHER FUNCTIONS	3123	=head1 OTHER FUNCTIONS
…		…
2395	=over 4	3127	=over 4
2396		3128
2397	=item ev_once (loop, int fd, int events, ev_tstamp timeout, callback)	3129	=item ev_once (loop, int fd, int events, ev_tstamp timeout, callback)
2398		3130
2399	This function combines a simple timer and an I/O watcher, calls your	3131	This function combines a simple timer and an I/O watcher, calls your
2400	callback on whichever event happens first and automatically stop both	3132	callback on whichever event happens first and automatically stops both
2401	watchers. This is useful if you want to wait for a single event on an fd	3133	watchers. This is useful if you want to wait for a single event on an fd
2402	or timeout without having to allocate/configure/start/stop/free one or	3134	or timeout without having to allocate/configure/start/stop/free one or
2403	more watchers yourself.	3135	more watchers yourself.
2404		3136
2405	If C<fd> is less than 0, then no I/O watcher will be started and events	3137	If C<fd> is less than 0, then no I/O watcher will be started and the
2406	is being ignored. Otherwise, an C<ev_io> watcher for the given C<fd> and	3138	C<events> argument is being ignored. Otherwise, an C<ev_io> watcher for
2407	C<events> set will be created and started.	3139	the given C<fd> and C<events> set will be created and started.
2408		3140
2409	If C<timeout> is less than 0, then no timeout watcher will be	3141	If C<timeout> is less than 0, then no timeout watcher will be
2410	started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and	3142	started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and
2411	repeat = 0) will be started. While C<0> is a valid timeout, it is of	3143	repeat = 0) will be started. C<0> is a valid timeout.
2412	dubious value.
2413		3144
2414	The callback has the type C<void (*cb)(int revents, void *arg)> and gets	3145	The callback has the type C<void (*cb)(int revents, void *arg)> and gets
2415	passed an C<revents> set like normal event callbacks (a combination of	3146	passed an C<revents> set like normal event callbacks (a combination of
2416	C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMEOUT>) and the C<arg>	3147	C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMEOUT>) and the C<arg>
2417	value passed to C<ev_once>:	3148	value passed to C<ev_once>. Note that it is possible to receive I<both>
		3149	a timeout and an io event at the same time - you probably should give io
		3150	events precedence.
		3151
		3152	Example: wait up to ten seconds for data to appear on STDIN_FILENO.
2418		3153
2419	static void stdin_ready (int revents, void *arg)	3154	static void stdin_ready (int revents, void *arg)
2420	{	3155	{
		3156	if (revents & EV_READ)
		3157	/* stdin might have data for us, joy! */;
2421	if (revents & EV_TIMEOUT)	3158	else if (revents & EV_TIMEOUT)
2422	/* doh, nothing entered */;	3159	/* doh, nothing entered */;
2423	else if (revents & EV_READ)
2424	/* stdin might have data for us, joy! */;
2425	}	3160	}
2426		3161
2427	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);	3162	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);
2428		3163
2429	=item ev_feed_event (ev_loop *, watcher *, int revents)
2430
2431	Feeds the given event set into the event loop, as if the specified event
2432	had happened for the specified watcher (which must be a pointer to an
2433	initialised but not necessarily started event watcher).
2434
2435	=item ev_feed_fd_event (ev_loop *, int fd, int revents)	3164	=item ev_feed_fd_event (loop, int fd, int revents)
2436		3165
2437	Feed an event on the given fd, as if a file descriptor backend detected	3166	Feed an event on the given fd, as if a file descriptor backend detected
2438	the given events it.	3167	the given events it.
2439		3168
2440	=item ev_feed_signal_event (ev_loop *loop, int signum)	3169	=item ev_feed_signal_event (loop, int signum)
2441		3170
2442	Feed an event as if the given signal occurred (C<loop> must be the default	3171	Feed an event as if the given signal occurred (C<loop> must be the default
2443	loop!).	3172	loop!).
2444		3173
2445	=back	3174	=back
…		…
2525		3254
2526	=over 4	3255	=over 4
2527		3256
2528	=item ev::TYPE::TYPE ()	3257	=item ev::TYPE::TYPE ()
2529		3258
2530	=item ev::TYPE::TYPE (struct ev_loop *)	3259	=item ev::TYPE::TYPE (loop)
2531		3260
2532	=item ev::TYPE::~TYPE	3261	=item ev::TYPE::~TYPE
2533		3262
2534	The constructor (optionally) takes an event loop to associate the watcher	3263	The constructor (optionally) takes an event loop to associate the watcher
2535	with. If it is omitted, it will use C<EV_DEFAULT>.	3264	with. If it is omitted, it will use C<EV_DEFAULT>.
…		…
2567		3296
2568	myclass obj;	3297	myclass obj;
2569	ev::io iow;	3298	ev::io iow;
2570	iow.set <myclass, &myclass::io_cb> (&obj);	3299	iow.set <myclass, &myclass::io_cb> (&obj);
2571		3300
		3301	=item w->set (object *)
		3302
		3303	This is an B<experimental> feature that might go away in a future version.
		3304
		3305	This is a variation of a method callback - leaving out the method to call
		3306	will default the method to C<operator ()>, which makes it possible to use
		3307	functor objects without having to manually specify the C<operator ()> all
		3308	the time. Incidentally, you can then also leave out the template argument
		3309	list.
		3310
		3311	The C<operator ()> method prototype must be C<void operator ()(watcher &w,
		3312	int revents)>.
		3313
		3314	See the method-C<set> above for more details.
		3315
		3316	Example: use a functor object as callback.
		3317
		3318	struct myfunctor
		3319	{
		3320	void operator() (ev::io &w, int revents)
		3321	{
		3322	...
		3323	}
		3324	}
		3325
		3326	myfunctor f;
		3327
		3328	ev::io w;
		3329	w.set (&f);
		3330
2572	=item w->set<function> (void *data = 0)	3331	=item w->set<function> (void *data = 0)
2573		3332
2574	Also sets a callback, but uses a static method or plain function as	3333	Also sets a callback, but uses a static method or plain function as
2575	callback. The optional C<data> argument will be stored in the watcher's	3334	callback. The optional C<data> argument will be stored in the watcher's
2576	C<data> member and is free for you to use.	3335	C<data> member and is free for you to use.
2577		3336
2578	The prototype of the C<function> must be C<void (*)(ev::TYPE &w, int)>.	3337	The prototype of the C<function> must be C<void (*)(ev::TYPE &w, int)>.
2579		3338
2580	See the method-C<set> above for more details.	3339	See the method-C<set> above for more details.
2581		3340
2582	Example:	3341	Example: Use a plain function as callback.
2583		3342
2584	static void io_cb (ev::io &w, int revents) { }	3343	static void io_cb (ev::io &w, int revents) { }
2585	iow.set <io_cb> ();	3344	iow.set <io_cb> ();
2586		3345
2587	=item w->set (struct ev_loop *)	3346	=item w->set (loop)
2588		3347
2589	Associates a different C<struct ev_loop> with this watcher. You can only	3348	Associates a different C<struct ev_loop> with this watcher. You can only
2590	do this when the watcher is inactive (and not pending either).	3349	do this when the watcher is inactive (and not pending either).
2591		3350
2592	=item w->set ([arguments])	3351	=item w->set ([arguments])
…		…
2625	Example: Define a class with an IO and idle watcher, start one of them in	3384	Example: Define a class with an IO and idle watcher, start one of them in
2626	the constructor.	3385	the constructor.
2627		3386
2628	class myclass	3387	class myclass
2629	{	3388	{
2630	ev::io io; void io_cb (ev::io &w, int revents);	3389	ev::io io ; void io_cb (ev::io &w, int revents);
2631	ev:idle idle void idle_cb (ev::idle &w, int revents);	3390	ev::idle idle; void idle_cb (ev::idle &w, int revents);
2632		3391
2633	myclass (int fd)	3392	myclass (int fd)
2634	{	3393	{
2635	io .set <myclass, &myclass::io_cb > (this);	3394	io .set <myclass, &myclass::io_cb > (this);
2636	idle.set <myclass, &myclass::idle_cb> (this);	3395	idle.set <myclass, &myclass::idle_cb> (this);
…		…
2652	=item Perl	3411	=item Perl
2653		3412
2654	The EV module implements the full libev API and is actually used to test	3413	The EV module implements the full libev API and is actually used to test
2655	libev. EV is developed together with libev. Apart from the EV core module,	3414	libev. EV is developed together with libev. Apart from the EV core module,
2656	there are additional modules that implement libev-compatible interfaces	3415	there are additional modules that implement libev-compatible interfaces
2657	to C<libadns> (C<EV::ADNS>), C<Net::SNMP> (C<Net::SNMP::EV>) and the	3416	to C<libadns> (C<EV::ADNS>, but C<AnyEvent::DNS> is preferred nowadays),
2658	C<libglib> event core (C<Glib::EV> and C<EV::Glib>).	3417	C<Net::SNMP> (C<Net::SNMP::EV>) and the C<libglib> event core (C<Glib::EV>
		3418	and C<EV::Glib>).
2659		3419
2660	It can be found and installed via CPAN, its homepage is at	3420	It can be found and installed via CPAN, its homepage is at
2661	L<http://software.schmorp.de/pkg/EV>.	3421	L<http://software.schmorp.de/pkg/EV>.
2662		3422
2663	=item Python	3423	=item Python
2664		3424
2665	Python bindings can be found at L<http://code.google.com/p/pyev/>. It	3425	Python bindings can be found at L<http://code.google.com/p/pyev/>. It
2666	seems to be quite complete and well-documented. Note, however, that the	3426	seems to be quite complete and well-documented.
2667	patch they require for libev is outright dangerous as it breaks the ABI
2668	for everybody else, and therefore, should never be applied in an installed
2669	libev (if python requires an incompatible ABI then it needs to embed
2670	libev).
2671		3427
2672	=item Ruby	3428	=item Ruby
2673		3429
2674	Tony Arcieri has written a ruby extension that offers access to a subset	3430	Tony Arcieri has written a ruby extension that offers access to a subset
2675	of the libev API and adds file handle abstractions, asynchronous DNS and	3431	of the libev API and adds file handle abstractions, asynchronous DNS and
2676	more on top of it. It can be found via gem servers. Its homepage is at	3432	more on top of it. It can be found via gem servers. Its homepage is at
2677	L<http://rev.rubyforge.org/>.	3433	L<http://rev.rubyforge.org/>.
2678		3434
		3435	Roger Pack reports that using the link order C<-lws2_32 -lmsvcrt-ruby-190>
		3436	makes rev work even on mingw.
		3437
		3438	=item Haskell
		3439
		3440	A haskell binding to libev is available at
		3441	L<http://hackage.haskell.org/cgi-bin/hackage-scripts/package/hlibev>.
		3442
2679	=item D	3443	=item D
2680		3444
2681	Leandro Lucarella has written a D language binding (F<ev.d>) for libev, to	3445	Leandro Lucarella has written a D language binding (F<ev.d>) for libev, to
2682	be found at L<http://proj.llucax.com.ar/wiki/evd>.	3446	be found at L<http://proj.llucax.com.ar/wiki/evd>.
		3447
		3448	=item Ocaml
		3449
		3450	Erkki Seppala has written Ocaml bindings for libev, to be found at
		3451	L<http://modeemi.cs.tut.fi/~flux/software/ocaml-ev/>.
		3452
		3453	=item Lua
		3454
		3455	Brian Maher has written a partial interface to libev
		3456	for lua (only C<ev_io> and C<ev_timer>), to be found at
		3457	L<http://github.com/brimworks/lua-ev>.
2683		3458
2684	=back	3459	=back
2685		3460
2686		3461
2687	=head1 MACRO MAGIC	3462	=head1 MACRO MAGIC
…		…
2788		3563
2789	#define EV_STANDALONE 1	3564	#define EV_STANDALONE 1
2790	#include "ev.h"	3565	#include "ev.h"
2791		3566
2792	Both header files and implementation files can be compiled with a C++	3567	Both header files and implementation files can be compiled with a C++
2793	compiler (at least, thats a stated goal, and breakage will be treated	3568	compiler (at least, that's a stated goal, and breakage will be treated
2794	as a bug).	3569	as a bug).
2795		3570
2796	You need the following files in your source tree, or in a directory	3571	You need the following files in your source tree, or in a directory
2797	in your include path (e.g. in libev/ when using -Ilibev):	3572	in your include path (e.g. in libev/ when using -Ilibev):
2798		3573
…		…
2842		3617
2843	=head2 PREPROCESSOR SYMBOLS/MACROS	3618	=head2 PREPROCESSOR SYMBOLS/MACROS
2844		3619
2845	Libev can be configured via a variety of preprocessor symbols you have to	3620	Libev can be configured via a variety of preprocessor symbols you have to
2846	define before including any of its files. The default in the absence of	3621	define before including any of its files. The default in the absence of
2847	autoconf is noted for every option.	3622	autoconf is documented for every option.
2848		3623
2849	=over 4	3624	=over 4
2850		3625
2851	=item EV_STANDALONE	3626	=item EV_STANDALONE
2852		3627
…		…
2854	keeps libev from including F<config.h>, and it also defines dummy	3629	keeps libev from including F<config.h>, and it also defines dummy
2855	implementations for some libevent functions (such as logging, which is not	3630	implementations for some libevent functions (such as logging, which is not
2856	supported). It will also not define any of the structs usually found in	3631	supported). It will also not define any of the structs usually found in
2857	F<event.h> that are not directly supported by the libev core alone.	3632	F<event.h> that are not directly supported by the libev core alone.
2858		3633
		3634	In standalone mode, libev will still try to automatically deduce the
		3635	configuration, but has to be more conservative.
		3636
2859	=item EV_USE_MONOTONIC	3637	=item EV_USE_MONOTONIC
2860		3638
2861	If defined to be C<1>, libev will try to detect the availability of the	3639	If defined to be C<1>, libev will try to detect the availability of the
2862	monotonic clock option at both compile time and runtime. Otherwise no use	3640	monotonic clock option at both compile time and runtime. Otherwise no
2863	of the monotonic clock option will be attempted. If you enable this, you	3641	use of the monotonic clock option will be attempted. If you enable this,
2864	usually have to link against librt or something similar. Enabling it when	3642	you usually have to link against librt or something similar. Enabling it
2865	the functionality isn't available is safe, though, although you have	3643	when the functionality isn't available is safe, though, although you have
2866	to make sure you link against any libraries where the C<clock_gettime>	3644	to make sure you link against any libraries where the C<clock_gettime>
2867	function is hiding in (often F<-lrt>).	3645	function is hiding in (often F<-lrt>). See also C<EV_USE_CLOCK_SYSCALL>.
2868		3646
2869	=item EV_USE_REALTIME	3647	=item EV_USE_REALTIME
2870		3648
2871	If defined to be C<1>, libev will try to detect the availability of the	3649	If defined to be C<1>, libev will try to detect the availability of the
2872	real-time clock option at compile time (and assume its availability at	3650	real-time clock option at compile time (and assume its availability
2873	runtime if successful). Otherwise no use of the real-time clock option will	3651	at runtime if successful). Otherwise no use of the real-time clock
2874	be attempted. This effectively replaces C<gettimeofday> by C<clock_get	3652	option will be attempted. This effectively replaces C<gettimeofday>
2875	(CLOCK_REALTIME, ...)> and will not normally affect correctness. See the	3653	by C<clock_get (CLOCK_REALTIME, ...)> and will not normally affect
2876	note about libraries in the description of C<EV_USE_MONOTONIC>, though.	3654	correctness. See the note about libraries in the description of
		3655	C<EV_USE_MONOTONIC>, though. Defaults to the opposite value of
		3656	C<EV_USE_CLOCK_SYSCALL>.
		3657
		3658	=item EV_USE_CLOCK_SYSCALL
		3659
		3660	If defined to be C<1>, libev will try to use a direct syscall instead
		3661	of calling the system-provided C<clock_gettime> function. This option
		3662	exists because on GNU/Linux, C<clock_gettime> is in C<librt>, but C<librt>
		3663	unconditionally pulls in C<libpthread>, slowing down single-threaded
		3664	programs needlessly. Using a direct syscall is slightly slower (in
		3665	theory), because no optimised vdso implementation can be used, but avoids
		3666	the pthread dependency. Defaults to C<1> on GNU/Linux with glibc 2.x or
		3667	higher, as it simplifies linking (no need for C<-lrt>).
2877		3668
2878	=item EV_USE_NANOSLEEP	3669	=item EV_USE_NANOSLEEP
2879		3670
2880	If defined to be C<1>, libev will assume that C<nanosleep ()> is available	3671	If defined to be C<1>, libev will assume that C<nanosleep ()> is available
2881	and will use it for delays. Otherwise it will use C<select ()>.	3672	and will use it for delays. Otherwise it will use C<select ()>.
…		…
2897		3688
2898	=item EV_SELECT_USE_FD_SET	3689	=item EV_SELECT_USE_FD_SET
2899		3690
2900	If defined to C<1>, then the select backend will use the system C<fd_set>	3691	If defined to C<1>, then the select backend will use the system C<fd_set>
2901	structure. This is useful if libev doesn't compile due to a missing	3692	structure. This is useful if libev doesn't compile due to a missing
2902	C<NFDBITS> or C<fd_mask> definition or it mis-guesses the bitset layout on	3693	C<NFDBITS> or C<fd_mask> definition or it mis-guesses the bitset layout
2903	exotic systems. This usually limits the range of file descriptors to some	3694	on exotic systems. This usually limits the range of file descriptors to
2904	low limit such as 1024 or might have other limitations (winsocket only	3695	some low limit such as 1024 or might have other limitations (winsocket
2905	allows 64 sockets). The C<FD_SETSIZE> macro, set before compilation, might	3696	only allows 64 sockets). The C<FD_SETSIZE> macro, set before compilation,
2906	influence the size of the C<fd_set> used.	3697	configures the maximum size of the C<fd_set>.
2907		3698
2908	=item EV_SELECT_IS_WINSOCKET	3699	=item EV_SELECT_IS_WINSOCKET
2909		3700
2910	When defined to C<1>, the select backend will assume that	3701	When defined to C<1>, the select backend will assume that
2911	select/socket/connect etc. don't understand file descriptors but	3702	select/socket/connect etc. don't understand file descriptors but
…		…
2913	be used is the winsock select). This means that it will call	3704	be used is the winsock select). This means that it will call
2914	C<_get_osfhandle> on the fd to convert it to an OS handle. Otherwise,	3705	C<_get_osfhandle> on the fd to convert it to an OS handle. Otherwise,
2915	it is assumed that all these functions actually work on fds, even	3706	it is assumed that all these functions actually work on fds, even
2916	on win32. Should not be defined on non-win32 platforms.	3707	on win32. Should not be defined on non-win32 platforms.
2917		3708
2918	=item EV_FD_TO_WIN32_HANDLE	3709	=item EV_FD_TO_WIN32_HANDLE(fd)
2919		3710
2920	If C<EV_SELECT_IS_WINSOCKET> is enabled, then libev needs a way to map	3711	If C<EV_SELECT_IS_WINSOCKET> is enabled, then libev needs a way to map
2921	file descriptors to socket handles. When not defining this symbol (the	3712	file descriptors to socket handles. When not defining this symbol (the
2922	default), then libev will call C<_get_osfhandle>, which is usually	3713	default), then libev will call C<_get_osfhandle>, which is usually
2923	correct. In some cases, programs use their own file descriptor management,	3714	correct. In some cases, programs use their own file descriptor management,
2924	in which case they can provide this function to map fds to socket handles.	3715	in which case they can provide this function to map fds to socket handles.
		3716
		3717	=item EV_WIN32_HANDLE_TO_FD(handle)
		3718
		3719	If C<EV_SELECT_IS_WINSOCKET> then libev maps handles to file descriptors
		3720	using the standard C<_open_osfhandle> function. For programs implementing
		3721	their own fd to handle mapping, overwriting this function makes it easier
		3722	to do so. This can be done by defining this macro to an appropriate value.
		3723
		3724	=item EV_WIN32_CLOSE_FD(fd)
		3725
		3726	If programs implement their own fd to handle mapping on win32, then this
		3727	macro can be used to override the C<close> function, useful to unregister
		3728	file descriptors again. Note that the replacement function has to close
		3729	the underlying OS handle.
2925		3730
2926	=item EV_USE_POLL	3731	=item EV_USE_POLL
2927		3732
2928	If defined to be C<1>, libev will compile in support for the C<poll>(2)	3733	If defined to be C<1>, libev will compile in support for the C<poll>(2)
2929	backend. Otherwise it will be enabled on non-win32 platforms. It	3734	backend. Otherwise it will be enabled on non-win32 platforms. It
…		…
3022	When doing priority-based operations, libev usually has to linearly search	3827	When doing priority-based operations, libev usually has to linearly search
3023	all the priorities, so having many of them (hundreds) uses a lot of space	3828	all the priorities, so having many of them (hundreds) uses a lot of space
3024	and time, so using the defaults of five priorities (-2 .. +2) is usually	3829	and time, so using the defaults of five priorities (-2 .. +2) is usually
3025	fine.	3830	fine.
3026		3831
3027	If your embedding application does not need any priorities, defining these both to	3832	If your embedding application does not need any priorities, defining these
3028	C<0> will save some memory and CPU.	3833	both to C<0> will save some memory and CPU.
3029		3834
3030	=item EV_PERIODIC_ENABLE	3835	=item EV_PERIODIC_ENABLE
3031		3836
3032	If undefined or defined to be C<1>, then periodic timers are supported. If	3837	If undefined or defined to be C<1>, then periodic timers are supported. If
3033	defined to be C<0>, then they are not. Disabling them saves a few kB of	3838	defined to be C<0>, then they are not. Disabling them saves a few kB of
…		…
3040	code.	3845	code.
3041		3846
3042	=item EV_EMBED_ENABLE	3847	=item EV_EMBED_ENABLE
3043		3848
3044	If undefined or defined to be C<1>, then embed watchers are supported. If	3849	If undefined or defined to be C<1>, then embed watchers are supported. If
3045	defined to be C<0>, then they are not.	3850	defined to be C<0>, then they are not. Embed watchers rely on most other
		3851	watcher types, which therefore must not be disabled.
3046		3852
3047	=item EV_STAT_ENABLE	3853	=item EV_STAT_ENABLE
3048		3854
3049	If undefined or defined to be C<1>, then stat watchers are supported. If	3855	If undefined or defined to be C<1>, then stat watchers are supported. If
3050	defined to be C<0>, then they are not.	3856	defined to be C<0>, then they are not.
…		…
3060	defined to be C<0>, then they are not.	3866	defined to be C<0>, then they are not.
3061		3867
3062	=item EV_MINIMAL	3868	=item EV_MINIMAL
3063		3869
3064	If you need to shave off some kilobytes of code at the expense of some	3870	If you need to shave off some kilobytes of code at the expense of some
3065	speed, define this symbol to C<1>. Currently this is used to override some	3871	speed (but with the full API), define this symbol to C<1>. Currently this
3066	inlining decisions, saves roughly 30% code size on amd64. It also selects a	3872	is used to override some inlining decisions, saves roughly 30% code size
3067	much smaller 2-heap for timer management over the default 4-heap.	3873	on amd64. It also selects a much smaller 2-heap for timer management over
		3874	the default 4-heap.
		3875
		3876	You can save even more by disabling watcher types you do not need
		3877	and setting C<EV_MAXPRI> == C<EV_MINPRI>. Also, disabling C<assert>
		3878	(C<-DNDEBUG>) will usually reduce code size a lot.
		3879
		3880	Defining C<EV_MINIMAL> to C<2> will additionally reduce the core API to
		3881	provide a bare-bones event library. See C<ev.h> for details on what parts
		3882	of the API are still available, and do not complain if this subset changes
		3883	over time.
		3884
		3885	=item EV_NSIG
		3886
		3887	The highest supported signal number, +1 (or, the number of
		3888	signals): Normally, libev tries to deduce the maximum number of signals
		3889	automatically, but sometimes this fails, in which case it can be
		3890	specified. Also, using a lower number than detected (C<32> should be
		3891	good for about any system in existance) can save some memory, as libev
		3892	statically allocates some 12-24 bytes per signal number.
3068		3893
3069	=item EV_PID_HASHSIZE	3894	=item EV_PID_HASHSIZE
3070		3895
3071	C<ev_child> watchers use a small hash table to distribute workload by	3896	C<ev_child> watchers use a small hash table to distribute workload by
3072	pid. The default size is C<16> (or C<1> with C<EV_MINIMAL>), usually more	3897	pid. The default size is C<16> (or C<1> with C<EV_MINIMAL>), usually more
…		…
3082	two).	3907	two).
3083		3908
3084	=item EV_USE_4HEAP	3909	=item EV_USE_4HEAP
3085		3910
3086	Heaps are not very cache-efficient. To improve the cache-efficiency of the	3911	Heaps are not very cache-efficient. To improve the cache-efficiency of the
3087	timer and periodics heap, libev uses a 4-heap when this symbol is defined	3912	timer and periodics heaps, libev uses a 4-heap when this symbol is defined
3088	to C<1>. The 4-heap uses more complicated (longer) code but has	3913	to C<1>. The 4-heap uses more complicated (longer) code but has noticeably
3089	noticeably faster performance with many (thousands) of watchers.	3914	faster performance with many (thousands) of watchers.
3090		3915
3091	The default is C<1> unless C<EV_MINIMAL> is set in which case it is C<0>	3916	The default is C<1> unless C<EV_MINIMAL> is set in which case it is C<0>
3092	(disabled).	3917	(disabled).
3093		3918
3094	=item EV_HEAP_CACHE_AT	3919	=item EV_HEAP_CACHE_AT
3095		3920
3096	Heaps are not very cache-efficient. To improve the cache-efficiency of the	3921	Heaps are not very cache-efficient. To improve the cache-efficiency of the
3097	timer and periodics heap, libev can cache the timestamp (I<at>) within	3922	timer and periodics heaps, libev can cache the timestamp (I<at>) within
3098	the heap structure (selected by defining C<EV_HEAP_CACHE_AT> to C<1>),	3923	the heap structure (selected by defining C<EV_HEAP_CACHE_AT> to C<1>),
3099	which uses 8-12 bytes more per watcher and a few hundred bytes more code,	3924	which uses 8-12 bytes more per watcher and a few hundred bytes more code,
3100	but avoids random read accesses on heap changes. This improves performance	3925	but avoids random read accesses on heap changes. This improves performance
3101	noticeably with with many (hundreds) of watchers.	3926	noticeably with many (hundreds) of watchers.
3102		3927
3103	The default is C<1> unless C<EV_MINIMAL> is set in which case it is C<0>	3928	The default is C<1> unless C<EV_MINIMAL> is set in which case it is C<0>
3104	(disabled).	3929	(disabled).
3105		3930
3106	=item EV_VERIFY	3931	=item EV_VERIFY
…		…
3112	called once per loop, which can slow down libev. If set to C<3>, then the	3937	called once per loop, which can slow down libev. If set to C<3>, then the
3113	verification code will be called very frequently, which will slow down	3938	verification code will be called very frequently, which will slow down
3114	libev considerably.	3939	libev considerably.
3115		3940
3116	The default is C<1>, unless C<EV_MINIMAL> is set, in which case it will be	3941	The default is C<1>, unless C<EV_MINIMAL> is set, in which case it will be
3117	C<0.>	3942	C<0>.
3118		3943
3119	=item EV_COMMON	3944	=item EV_COMMON
3120		3945
3121	By default, all watchers have a C<void *data> member. By redefining	3946	By default, all watchers have a C<void *data> member. By redefining
3122	this macro to a something else you can include more and other types of	3947	this macro to a something else you can include more and other types of
…		…
3139	and the way callbacks are invoked and set. Must expand to a struct member	3964	and the way callbacks are invoked and set. Must expand to a struct member
3140	definition and a statement, respectively. See the F<ev.h> header file for	3965	definition and a statement, respectively. See the F<ev.h> header file for
3141	their default definitions. One possible use for overriding these is to	3966	their default definitions. One possible use for overriding these is to
3142	avoid the C<struct ev_loop *> as first argument in all cases, or to use	3967	avoid the C<struct ev_loop *> as first argument in all cases, or to use
3143	method calls instead of plain function calls in C++.	3968	method calls instead of plain function calls in C++.
		3969
		3970	=back
3144		3971
3145	=head2 EXPORTED API SYMBOLS	3972	=head2 EXPORTED API SYMBOLS
3146		3973
3147	If you need to re-export the API (e.g. via a DLL) and you need a list of	3974	If you need to re-export the API (e.g. via a DLL) and you need a list of
3148	exported symbols, you can use the provided F<Symbol.*> files which list	3975	exported symbols, you can use the provided F<Symbol.*> files which list
…		…
3195	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:	4022	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:
3196		4023
3197	#include "ev_cpp.h"	4024	#include "ev_cpp.h"
3198	#include "ev.c"	4025	#include "ev.c"
3199		4026
		4027	=head1 INTERACTION WITH OTHER PROGRAMS OR LIBRARIES
3200		4028
3201	=head1 THREADS AND COROUTINES	4029	=head2 THREADS AND COROUTINES
3202		4030
3203	=head2 THREADS	4031	=head3 THREADS
3204		4032
3205	Libev itself is completely thread-safe, but it uses no locking. This	4033	All libev functions are reentrant and thread-safe unless explicitly
		4034	documented otherwise, but libev implements no locking itself. This means
3206	means that you can use as many loops as you want in parallel, as long as	4035	that you can use as many loops as you want in parallel, as long as there
3207	only one thread ever calls into one libev function with the same loop	4036	are no concurrent calls into any libev function with the same loop
3208	parameter.	4037	parameter (C<ev_default_*> calls have an implicit default loop parameter,
		4038	of course): libev guarantees that different event loops share no data
		4039	structures that need any locking.
3209		4040
3210	Or put differently: calls with different loop parameters can be done in	4041	Or to put it differently: calls with different loop parameters can be done
3211	parallel from multiple threads, calls with the same loop parameter must be	4042	concurrently from multiple threads, calls with the same loop parameter
3212	done serially (but can be done from different threads, as long as only one	4043	must be done serially (but can be done from different threads, as long as
3213	thread ever is inside a call at any point in time, e.g. by using a mutex	4044	only one thread ever is inside a call at any point in time, e.g. by using
3214	per loop).	4045	a mutex per loop).
		4046
		4047	Specifically to support threads (and signal handlers), libev implements
		4048	so-called C<ev_async> watchers, which allow some limited form of
		4049	concurrency on the same event loop, namely waking it up "from the
		4050	outside".
3215		4051
3216	If you want to know which design (one loop, locking, or multiple loops	4052	If you want to know which design (one loop, locking, or multiple loops
3217	without or something else still) is best for your problem, then I cannot	4053	without or something else still) is best for your problem, then I cannot
3218	help you. I can give some generic advice however:	4054	help you, but here is some generic advice:
3219		4055
3220	=over 4	4056	=over 4
3221		4057
3222	=item * most applications have a main thread: use the default libev loop	4058	=item * most applications have a main thread: use the default libev loop
3223	in that thread, or create a separate thread running only the default loop.	4059	in that thread, or create a separate thread running only the default loop.
…		…
3235		4071
3236	Choosing a model is hard - look around, learn, know that usually you can do	4072	Choosing a model is hard - look around, learn, know that usually you can do
3237	better than you currently do :-)	4073	better than you currently do :-)
3238		4074
3239	=item * often you need to talk to some other thread which blocks in the	4075	=item * often you need to talk to some other thread which blocks in the
		4076	event loop.
		4077
3240	event loop - C<ev_async> watchers can be used to wake them up from other	4078	C<ev_async> watchers can be used to wake them up from other threads safely
3241	threads safely (or from signal contexts...).	4079	(or from signal contexts...).
		4080
		4081	An example use would be to communicate signals or other events that only
		4082	work in the default loop by registering the signal watcher with the
		4083	default loop and triggering an C<ev_async> watcher from the default loop
		4084	watcher callback into the event loop interested in the signal.
3242		4085
3243	=back	4086	=back
3244		4087
		4088	=head4 THREAD LOCKING EXAMPLE
		4089
		4090	Here is a fictitious example of how to run an event loop in a different
		4091	thread than where callbacks are being invoked and watchers are
		4092	created/added/removed.
		4093
		4094	For a real-world example, see the C<EV::Loop::Async> perl module,
		4095	which uses exactly this technique (which is suited for many high-level
		4096	languages).
		4097
		4098	The example uses a pthread mutex to protect the loop data, a condition
		4099	variable to wait for callback invocations, an async watcher to notify the
		4100	event loop thread and an unspecified mechanism to wake up the main thread.
		4101
		4102	First, you need to associate some data with the event loop:
		4103
		4104	typedef struct {
		4105	mutex_t lock; /* global loop lock */
		4106	ev_async async_w;
		4107	thread_t tid;
		4108	cond_t invoke_cv;
		4109	} userdata;
		4110
		4111	void prepare_loop (EV_P)
		4112	{
		4113	// for simplicity, we use a static userdata struct.
		4114	static userdata u;
		4115
		4116	ev_async_init (&u->async_w, async_cb);
		4117	ev_async_start (EV_A_ &u->async_w);
		4118
		4119	pthread_mutex_init (&u->lock, 0);
		4120	pthread_cond_init (&u->invoke_cv, 0);
		4121
		4122	// now associate this with the loop
		4123	ev_set_userdata (EV_A_ u);
		4124	ev_set_invoke_pending_cb (EV_A_ l_invoke);
		4125	ev_set_loop_release_cb (EV_A_ l_release, l_acquire);
		4126
		4127	// then create the thread running ev_loop
		4128	pthread_create (&u->tid, 0, l_run, EV_A);
		4129	}
		4130
		4131	The callback for the C<ev_async> watcher does nothing: the watcher is used
		4132	solely to wake up the event loop so it takes notice of any new watchers
		4133	that might have been added:
		4134
		4135	static void
		4136	async_cb (EV_P_ ev_async *w, int revents)
		4137	{
		4138	// just used for the side effects
		4139	}
		4140
		4141	The C<l_release> and C<l_acquire> callbacks simply unlock/lock the mutex
		4142	protecting the loop data, respectively.
		4143
		4144	static void
		4145	l_release (EV_P)
		4146	{
		4147	userdata *u = ev_userdata (EV_A);
		4148	pthread_mutex_unlock (&u->lock);
		4149	}
		4150
		4151	static void
		4152	l_acquire (EV_P)
		4153	{
		4154	userdata *u = ev_userdata (EV_A);
		4155	pthread_mutex_lock (&u->lock);
		4156	}
		4157
		4158	The event loop thread first acquires the mutex, and then jumps straight
		4159	into C<ev_loop>:
		4160
		4161	void *
		4162	l_run (void *thr_arg)
		4163	{
		4164	struct ev_loop *loop = (struct ev_loop *)thr_arg;
		4165
		4166	l_acquire (EV_A);
		4167	pthread_setcanceltype (PTHREAD_CANCEL_ASYNCHRONOUS, 0);
		4168	ev_loop (EV_A_ 0);
		4169	l_release (EV_A);
		4170
		4171	return 0;
		4172	}
		4173
		4174	Instead of invoking all pending watchers, the C<l_invoke> callback will
		4175	signal the main thread via some unspecified mechanism (signals? pipe
		4176	writes? C<Async::Interrupt>?) and then waits until all pending watchers
		4177	have been called (in a while loop because a) spurious wakeups are possible
		4178	and b) skipping inter-thread-communication when there are no pending
		4179	watchers is very beneficial):
		4180
		4181	static void
		4182	l_invoke (EV_P)
		4183	{
		4184	userdata *u = ev_userdata (EV_A);
		4185
		4186	while (ev_pending_count (EV_A))
		4187	{
		4188	wake_up_other_thread_in_some_magic_or_not_so_magic_way ();
		4189	pthread_cond_wait (&u->invoke_cv, &u->lock);
		4190	}
		4191	}
		4192
		4193	Now, whenever the main thread gets told to invoke pending watchers, it
		4194	will grab the lock, call C<ev_invoke_pending> and then signal the loop
		4195	thread to continue:
		4196
		4197	static void
		4198	real_invoke_pending (EV_P)
		4199	{
		4200	userdata *u = ev_userdata (EV_A);
		4201
		4202	pthread_mutex_lock (&u->lock);
		4203	ev_invoke_pending (EV_A);
		4204	pthread_cond_signal (&u->invoke_cv);
		4205	pthread_mutex_unlock (&u->lock);
		4206	}
		4207
		4208	Whenever you want to start/stop a watcher or do other modifications to an
		4209	event loop, you will now have to lock:
		4210
		4211	ev_timer timeout_watcher;
		4212	userdata *u = ev_userdata (EV_A);
		4213
		4214	ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);
		4215
		4216	pthread_mutex_lock (&u->lock);
		4217	ev_timer_start (EV_A_ &timeout_watcher);
		4218	ev_async_send (EV_A_ &u->async_w);
		4219	pthread_mutex_unlock (&u->lock);
		4220
		4221	Note that sending the C<ev_async> watcher is required because otherwise
		4222	an event loop currently blocking in the kernel will have no knowledge
		4223	about the newly added timer. By waking up the loop it will pick up any new
		4224	watchers in the next event loop iteration.
		4225
3245	=head2 COROUTINES	4226	=head3 COROUTINES
3246		4227
3247	Libev is much more accommodating to coroutines ("cooperative threads"):	4228	Libev is very accommodating to coroutines ("cooperative threads"):
3248	libev fully supports nesting calls to it's functions from different	4229	libev fully supports nesting calls to its functions from different
3249	coroutines (e.g. you can call C<ev_loop> on the same loop from two	4230	coroutines (e.g. you can call C<ev_loop> on the same loop from two
3250	different coroutines and switch freely between both coroutines running the	4231	different coroutines, and switch freely between both coroutines running
3251	loop, as long as you don't confuse yourself). The only exception is that	4232	the loop, as long as you don't confuse yourself). The only exception is
3252	you must not do this from C<ev_periodic> reschedule callbacks.	4233	that you must not do this from C<ev_periodic> reschedule callbacks.
3253		4234
3254	Care has been invested into making sure that libev does not keep local	4235	Care has been taken to ensure that libev does not keep local state inside
3255	state inside C<ev_loop>, and other calls do not usually allow coroutine	4236	C<ev_loop>, and other calls do not usually allow for coroutine switches as
3256	switches.	4237	they do not call any callbacks.
3257		4238
		4239	=head2 COMPILER WARNINGS
3258		4240
3259	=head1 COMPLEXITIES	4241	Depending on your compiler and compiler settings, you might get no or a
		4242	lot of warnings when compiling libev code. Some people are apparently
		4243	scared by this.
3260		4244
3261	In this section the complexities of (many of) the algorithms used inside	4245	However, these are unavoidable for many reasons. For one, each compiler
3262	libev will be explained. For complexity discussions about backends see the	4246	has different warnings, and each user has different tastes regarding
3263	documentation for C<ev_default_init>.	4247	warning options. "Warn-free" code therefore cannot be a goal except when
		4248	targeting a specific compiler and compiler-version.
3264		4249
3265	All of the following are about amortised time: If an array needs to be	4250	Another reason is that some compiler warnings require elaborate
3266	extended, libev needs to realloc and move the whole array, but this	4251	workarounds, or other changes to the code that make it less clear and less
3267	happens asymptotically never with higher number of elements, so O(1) might	4252	maintainable.
3268	mean it might do a lengthy realloc operation in rare cases, but on average
3269	it is much faster and asymptotically approaches constant time.
3270		4253
3271	=over 4	4254	And of course, some compiler warnings are just plain stupid, or simply
		4255	wrong (because they don't actually warn about the condition their message
		4256	seems to warn about). For example, certain older gcc versions had some
		4257	warnings that resulted an extreme number of false positives. These have
		4258	been fixed, but some people still insist on making code warn-free with
		4259	such buggy versions.
3272		4260
3273	=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers)	4261	While libev is written to generate as few warnings as possible,
		4262	"warn-free" code is not a goal, and it is recommended not to build libev
		4263	with any compiler warnings enabled unless you are prepared to cope with
		4264	them (e.g. by ignoring them). Remember that warnings are just that:
		4265	warnings, not errors, or proof of bugs.
3274		4266
3275	This means that, when you have a watcher that triggers in one hour and
3276	there are 100 watchers that would trigger before that then inserting will
3277	have to skip roughly seven (C<ld 100>) of these watchers.
3278		4267
3279	=item Changing timer/periodic watchers (by autorepeat or calling again): O(log skipped_other_timers)	4268	=head2 VALGRIND
3280		4269
3281	That means that changing a timer costs less than removing/adding them	4270	Valgrind has a special section here because it is a popular tool that is
3282	as only the relative motion in the event queue has to be paid for.	4271	highly useful. Unfortunately, valgrind reports are very hard to interpret.
3283		4272
3284	=item Starting io/check/prepare/idle/signal/child/fork/async watchers: O(1)	4273	If you think you found a bug (memory leak, uninitialised data access etc.)
		4274	in libev, then check twice: If valgrind reports something like:
3285		4275
3286	These just add the watcher into an array or at the head of a list.	4276	==2274== definitely lost: 0 bytes in 0 blocks.
		4277	==2274== possibly lost: 0 bytes in 0 blocks.
		4278	==2274== still reachable: 256 bytes in 1 blocks.
3287		4279
3288	=item Stopping check/prepare/idle/fork/async watchers: O(1)	4280	Then there is no memory leak, just as memory accounted to global variables
		4281	is not a memleak - the memory is still being referenced, and didn't leak.
3289		4282
3290	=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % EV_PID_HASHSIZE))	4283	Similarly, under some circumstances, valgrind might report kernel bugs
		4284	as if it were a bug in libev (e.g. in realloc or in the poll backend,
		4285	although an acceptable workaround has been found here), or it might be
		4286	confused.
3291		4287
3292	These watchers are stored in lists then need to be walked to find the	4288	Keep in mind that valgrind is a very good tool, but only a tool. Don't
3293	correct watcher to remove. The lists are usually short (you don't usually	4289	make it into some kind of religion.
3294	have many watchers waiting for the same fd or signal).
3295		4290
3296	=item Finding the next timer in each loop iteration: O(1)	4291	If you are unsure about something, feel free to contact the mailing list
		4292	with the full valgrind report and an explanation on why you think this
		4293	is a bug in libev (best check the archives, too :). However, don't be
		4294	annoyed when you get a brisk "this is no bug" answer and take the chance
		4295	of learning how to interpret valgrind properly.
3297		4296
3298	By virtue of using a binary or 4-heap, the next timer is always found at a	4297	If you need, for some reason, empty reports from valgrind for your project
3299	fixed position in the storage array.	4298	I suggest using suppression lists.
3300		4299
3301	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)
3302		4300
3303	A change means an I/O watcher gets started or stopped, which requires	4301	=head1 PORTABILITY NOTES
3304	libev to recalculate its status (and possibly tell the kernel, depending
3305	on backend and whether C<ev_io_set> was used).
3306		4302
3307	=item Activating one watcher (putting it into the pending state): O(1)
3308
3309	=item Priority handling: O(number_of_priorities)
3310
3311	Priorities are implemented by allocating some space for each
3312	priority. When doing priority-based operations, libev usually has to
3313	linearly search all the priorities, but starting/stopping and activating
3314	watchers becomes O(1) w.r.t. priority handling.
3315
3316	=item Sending an ev_async: O(1)
3317
3318	=item Processing ev_async_send: O(number_of_async_watchers)
3319
3320	=item Processing signals: O(max_signal_number)
3321
3322	Sending involves a system call I<iff> there were no other C<ev_async_send>
3323	calls in the current loop iteration. Checking for async and signal events
3324	involves iterating over all running async watchers or all signal numbers.
3325
3326	=back
3327
3328
3329	=head1 WIN32 PLATFORM LIMITATIONS AND WORKAROUNDS	4303	=head2 WIN32 PLATFORM LIMITATIONS AND WORKAROUNDS
3330		4304
3331	Win32 doesn't support any of the standards (e.g. POSIX) that libev	4305	Win32 doesn't support any of the standards (e.g. POSIX) that libev
3332	requires, and its I/O model is fundamentally incompatible with the POSIX	4306	requires, and its I/O model is fundamentally incompatible with the POSIX
3333	model. Libev still offers limited functionality on this platform in	4307	model. Libev still offers limited functionality on this platform in
3334	the form of the C<EVBACKEND_SELECT> backend, and only supports socket	4308	the form of the C<EVBACKEND_SELECT> backend, and only supports socket
…		…
3341	way (note also that glib is the slowest event library known to man).	4315	way (note also that glib is the slowest event library known to man).
3342		4316
3343	There is no supported compilation method available on windows except	4317	There is no supported compilation method available on windows except
3344	embedding it into other applications.	4318	embedding it into other applications.
3345		4319
		4320	Sensible signal handling is officially unsupported by Microsoft - libev
		4321	tries its best, but under most conditions, signals will simply not work.
		4322
3346	Not a libev limitation but worth mentioning: windows apparently doesn't	4323	Not a libev limitation but worth mentioning: windows apparently doesn't
3347	accept large writes: instead of resulting in a partial write, windows will	4324	accept large writes: instead of resulting in a partial write, windows will
3348	either accept everything or return C<ENOBUFS> if the buffer is too large,	4325	either accept everything or return C<ENOBUFS> if the buffer is too large,
3349	so make sure you only write small amounts into your sockets (less than a	4326	so make sure you only write small amounts into your sockets (less than a
3350	megabyte seems safe, but thsi apparently depends on the amount of memory	4327	megabyte seems safe, but this apparently depends on the amount of memory
3351	available).	4328	available).
3352		4329
3353	Due to the many, low, and arbitrary limits on the win32 platform and	4330	Due to the many, low, and arbitrary limits on the win32 platform and
3354	the abysmal performance of winsockets, using a large number of sockets	4331	the abysmal performance of winsockets, using a large number of sockets
3355	is not recommended (and not reasonable). If your program needs to use	4332	is not recommended (and not reasonable). If your program needs to use
3356	more than a hundred or so sockets, then likely it needs to use a totally	4333	more than a hundred or so sockets, then likely it needs to use a totally
3357	different implementation for windows, as libev offers the POSIX readiness	4334	different implementation for windows, as libev offers the POSIX readiness
3358	notification model, which cannot be implemented efficiently on windows	4335	notification model, which cannot be implemented efficiently on windows
3359	(Microsoft monopoly games).	4336	(due to Microsoft monopoly games).
3360		4337
3361	A typical way to use libev under windows is to embed it (see the embedding	4338	A typical way to use libev under windows is to embed it (see the embedding
3362	section for details) and use the following F<evwrap.h> header file instead	4339	section for details) and use the following F<evwrap.h> header file instead
3363	of F<ev.h>:	4340	of F<ev.h>:
3364		4341
…		…
3366	#define EV_SELECT_IS_WINSOCKET 1 /* configure libev for windows select */	4343	#define EV_SELECT_IS_WINSOCKET 1 /* configure libev for windows select */
3367		4344
3368	#include "ev.h"	4345	#include "ev.h"
3369		4346
3370	And compile the following F<evwrap.c> file into your project (make sure	4347	And compile the following F<evwrap.c> file into your project (make sure
3371	you do I<not> compile the F<ev.c> or any other embedded soruce files!):	4348	you do I<not> compile the F<ev.c> or any other embedded source files!):
3372		4349
3373	#include "evwrap.h"	4350	#include "evwrap.h"
3374	#include "ev.c"	4351	#include "ev.c"
3375		4352
3376	=over 4	4353	=over 4
…		…
3400		4377
3401	Early versions of winsocket's select only supported waiting for a maximum	4378	Early versions of winsocket's select only supported waiting for a maximum
3402	of C<64> handles (probably owning to the fact that all windows kernels	4379	of C<64> handles (probably owning to the fact that all windows kernels
3403	can only wait for C<64> things at the same time internally; Microsoft	4380	can only wait for C<64> things at the same time internally; Microsoft
3404	recommends spawning a chain of threads and wait for 63 handles and the	4381	recommends spawning a chain of threads and wait for 63 handles and the
3405	previous thread in each. Great).	4382	previous thread in each. Sounds great!).
3406		4383
3407	Newer versions support more handles, but you need to define C<FD_SETSIZE>	4384	Newer versions support more handles, but you need to define C<FD_SETSIZE>
3408	to some high number (e.g. C<2048>) before compiling the winsocket select	4385	to some high number (e.g. C<2048>) before compiling the winsocket select
3409	call (which might be in libev or elsewhere, for example, perl does its own	4386	call (which might be in libev or elsewhere, for example, perl and many
3410	select emulation on windows).	4387	other interpreters do their own select emulation on windows).
3411		4388
3412	Another limit is the number of file descriptors in the Microsoft runtime	4389	Another limit is the number of file descriptors in the Microsoft runtime
3413	libraries, which by default is C<64> (there must be a hidden I<64> fetish	4390	libraries, which by default is C<64> (there must be a hidden I<64>
3414	or something like this inside Microsoft). You can increase this by calling	4391	fetish or something like this inside Microsoft). You can increase this
3415	C<_setmaxstdio>, which can increase this limit to C<2048> (another	4392	by calling C<_setmaxstdio>, which can increase this limit to C<2048>
3416	arbitrary limit), but is broken in many versions of the Microsoft runtime	4393	(another arbitrary limit), but is broken in many versions of the Microsoft
3417	libraries.
3418
3419	This might get you to about C<512> or C<2048> sockets (depending on	4394	runtime libraries. This might get you to about C<512> or C<2048> sockets
3420	windows version and/or the phase of the moon). To get more, you need to	4395	(depending on windows version and/or the phase of the moon). To get more,
3421	wrap all I/O functions and provide your own fd management, but the cost of	4396	you need to wrap all I/O functions and provide your own fd management, but
3422	calling select (O(n²)) will likely make this unworkable.	4397	the cost of calling select (O(n²)) will likely make this unworkable.
3423		4398
3424	=back	4399	=back
3425		4400
3426
3427	=head1 PORTABILITY REQUIREMENTS	4401	=head2 PORTABILITY REQUIREMENTS
3428		4402
3429	In addition to a working ISO-C implementation, libev relies on a few	4403	In addition to a working ISO-C implementation and of course the
3430	additional extensions:	4404	backend-specific APIs, libev relies on a few additional extensions:
3431		4405
3432	=over 4	4406	=over 4
3433		4407
3434	=item C<void (*)(ev_watcher_type *, int revents)> must have compatible	4408	=item C<void (*)(ev_watcher_type *, int revents)> must have compatible
3435	calling conventions regardless of C<ev_watcher_type *>.	4409	calling conventions regardless of C<ev_watcher_type *>.
…		…
3441	calls them using an C<ev_watcher *> internally.	4415	calls them using an C<ev_watcher *> internally.
3442		4416
3443	=item C<sig_atomic_t volatile> must be thread-atomic as well	4417	=item C<sig_atomic_t volatile> must be thread-atomic as well
3444		4418
3445	The type C<sig_atomic_t volatile> (or whatever is defined as	4419	The type C<sig_atomic_t volatile> (or whatever is defined as
3446	C<EV_ATOMIC_T>) must be atomic w.r.t. accesses from different	4420	C<EV_ATOMIC_T>) must be atomic with respect to accesses from different
3447	threads. This is not part of the specification for C<sig_atomic_t>, but is	4421	threads. This is not part of the specification for C<sig_atomic_t>, but is
3448	believed to be sufficiently portable.	4422	believed to be sufficiently portable.
3449		4423
3450	=item C<sigprocmask> must work in a threaded environment	4424	=item C<sigprocmask> must work in a threaded environment
3451		4425
…		…
3460	except the initial one, and run the default loop in the initial thread as	4434	except the initial one, and run the default loop in the initial thread as
3461	well.	4435	well.
3462		4436
3463	=item C<long> must be large enough for common memory allocation sizes	4437	=item C<long> must be large enough for common memory allocation sizes
3464		4438
3465	To improve portability and simplify using libev, libev uses C<long>	4439	To improve portability and simplify its API, libev uses C<long> internally
3466	internally instead of C<size_t> when allocating its data structures. On	4440	instead of C<size_t> when allocating its data structures. On non-POSIX
3467	non-POSIX systems (Microsoft...) this might be unexpectedly low, but	4441	systems (Microsoft...) this might be unexpectedly low, but is still at
3468	is still at least 31 bits everywhere, which is enough for hundreds of	4442	least 31 bits everywhere, which is enough for hundreds of millions of
3469	millions of watchers.	4443	watchers.
3470		4444
3471	=item C<double> must hold a time value in seconds with enough accuracy	4445	=item C<double> must hold a time value in seconds with enough accuracy
3472		4446
3473	The type C<double> is used to represent timestamps. It is required to	4447	The type C<double> is used to represent timestamps. It is required to
3474	have at least 51 bits of mantissa (and 9 bits of exponent), which is good	4448	have at least 51 bits of mantissa (and 9 bits of exponent), which is good
3475	enough for at least into the year 4000. This requirement is fulfilled by	4449	enough for at least into the year 4000. This requirement is fulfilled by
3476	implementations implementing IEEE 754 (basically all existing ones).	4450	implementations implementing IEEE 754, which is basically all existing
		4451	ones. With IEEE 754 doubles, you get microsecond accuracy until at least
		4452	2200.
3477		4453
3478	=back	4454	=back
3479		4455
3480	If you know of other additional requirements drop me a note.	4456	If you know of other additional requirements drop me a note.
3481		4457
3482		4458
3483	=head1 COMPILER WARNINGS	4459	=head1 ALGORITHMIC COMPLEXITIES
3484		4460
3485	Depending on your compiler and compiler settings, you might get no or a	4461	In this section the complexities of (many of) the algorithms used inside
3486	lot of warnings when compiling libev code. Some people are apparently	4462	libev will be documented. For complexity discussions about backends see
3487	scared by this.	4463	the documentation for C<ev_default_init>.
3488		4464
3489	However, these are unavoidable for many reasons. For one, each compiler	4465	All of the following are about amortised time: If an array needs to be
3490	has different warnings, and each user has different tastes regarding	4466	extended, libev needs to realloc and move the whole array, but this
3491	warning options. "Warn-free" code therefore cannot be a goal except when	4467	happens asymptotically rarer with higher number of elements, so O(1) might
3492	targeting a specific compiler and compiler-version.	4468	mean that libev does a lengthy realloc operation in rare cases, but on
		4469	average it is much faster and asymptotically approaches constant time.
3493		4470
3494	Another reason is that some compiler warnings require elaborate	4471	=over 4
3495	workarounds, or other changes to the code that make it less clear and less
3496	maintainable.
3497		4472
3498	And of course, some compiler warnings are just plain stupid, or simply	4473	=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers)
3499	wrong (because they don't actually warn about the condition their message
3500	seems to warn about).
3501		4474
3502	While libev is written to generate as few warnings as possible,	4475	This means that, when you have a watcher that triggers in one hour and
3503	"warn-free" code is not a goal, and it is recommended not to build libev	4476	there are 100 watchers that would trigger before that, then inserting will
3504	with any compiler warnings enabled unless you are prepared to cope with	4477	have to skip roughly seven (C<ld 100>) of these watchers.
3505	them (e.g. by ignoring them). Remember that warnings are just that:
3506	warnings, not errors, or proof of bugs.
3507		4478
		4479	=item Changing timer/periodic watchers (by autorepeat or calling again): O(log skipped_other_timers)
3508		4480
3509	=head1 VALGRIND	4481	That means that changing a timer costs less than removing/adding them,
		4482	as only the relative motion in the event queue has to be paid for.
3510		4483
3511	Valgrind has a special section here because it is a popular tool that is	4484	=item Starting io/check/prepare/idle/signal/child/fork/async watchers: O(1)
3512	highly useful, but valgrind reports are very hard to interpret.
3513		4485
3514	If you think you found a bug (memory leak, uninitialised data access etc.)	4486	These just add the watcher into an array or at the head of a list.
3515	in libev, then check twice: If valgrind reports something like:
3516		4487
3517	==2274== definitely lost: 0 bytes in 0 blocks.	4488	=item Stopping check/prepare/idle/fork/async watchers: O(1)
3518	==2274== possibly lost: 0 bytes in 0 blocks.
3519	==2274== still reachable: 256 bytes in 1 blocks.
3520		4489
3521	Then there is no memory leak. Similarly, under some circumstances,	4490	=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % EV_PID_HASHSIZE))
3522	valgrind might report kernel bugs as if it were a bug in libev, or it
3523	might be confused (it is a very good tool, but only a tool).
3524		4491
3525	If you are unsure about something, feel free to contact the mailing list	4492	These watchers are stored in lists, so they need to be walked to find the
3526	with the full valgrind report and an explanation on why you think this is	4493	correct watcher to remove. The lists are usually short (you don't usually
3527	a bug in libev. However, don't be annoyed when you get a brisk "this is	4494	have many watchers waiting for the same fd or signal: one is typical, two
3528	no bug" answer and take the chance of learning how to interpret valgrind	4495	is rare).
3529	properly.
3530		4496
3531	If you need, for some reason, empty reports from valgrind for your project	4497	=item Finding the next timer in each loop iteration: O(1)
3532	I suggest using suppression lists.
3533		4498
		4499	By virtue of using a binary or 4-heap, the next timer is always found at a
		4500	fixed position in the storage array.
		4501
		4502	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)
		4503
		4504	A change means an I/O watcher gets started or stopped, which requires
		4505	libev to recalculate its status (and possibly tell the kernel, depending
		4506	on backend and whether C<ev_io_set> was used).
		4507
		4508	=item Activating one watcher (putting it into the pending state): O(1)
		4509
		4510	=item Priority handling: O(number_of_priorities)
		4511
		4512	Priorities are implemented by allocating some space for each
		4513	priority. When doing priority-based operations, libev usually has to
		4514	linearly search all the priorities, but starting/stopping and activating
		4515	watchers becomes O(1) with respect to priority handling.
		4516
		4517	=item Sending an ev_async: O(1)
		4518
		4519	=item Processing ev_async_send: O(number_of_async_watchers)
		4520
		4521	=item Processing signals: O(max_signal_number)
		4522
		4523	Sending involves a system call I<iff> there were no other C<ev_async_send>
		4524	calls in the current loop iteration. Checking for async and signal events
		4525	involves iterating over all running async watchers or all signal numbers.
		4526
		4527	=back
		4528
		4529
		4530	=head1 GLOSSARY
		4531
		4532	=over 4
		4533
		4534	=item active
		4535
		4536	A watcher is active as long as it has been started (has been attached to
		4537	an event loop) but not yet stopped (disassociated from the event loop).
		4538
		4539	=item application
		4540
		4541	In this document, an application is whatever is using libev.
		4542
		4543	=item callback
		4544
		4545	The address of a function that is called when some event has been
		4546	detected. Callbacks are being passed the event loop, the watcher that
		4547	received the event, and the actual event bitset.
		4548
		4549	=item callback invocation
		4550
		4551	The act of calling the callback associated with a watcher.
		4552
		4553	=item event
		4554
		4555	A change of state of some external event, such as data now being available
		4556	for reading on a file descriptor, time having passed or simply not having
		4557	any other events happening anymore.
		4558
		4559	In libev, events are represented as single bits (such as C<EV_READ> or
		4560	C<EV_TIMEOUT>).
		4561
		4562	=item event library
		4563
		4564	A software package implementing an event model and loop.
		4565
		4566	=item event loop
		4567
		4568	An entity that handles and processes external events and converts them
		4569	into callback invocations.
		4570
		4571	=item event model
		4572
		4573	The model used to describe how an event loop handles and processes
		4574	watchers and events.
		4575
		4576	=item pending
		4577
		4578	A watcher is pending as soon as the corresponding event has been detected,
		4579	and stops being pending as soon as the watcher will be invoked or its
		4580	pending status is explicitly cleared by the application.
		4581
		4582	A watcher can be pending, but not active. Stopping a watcher also clears
		4583	its pending status.
		4584
		4585	=item real time
		4586
		4587	The physical time that is observed. It is apparently strictly monotonic :)
		4588
		4589	=item wall-clock time
		4590
		4591	The time and date as shown on clocks. Unlike real time, it can actually
		4592	be wrong and jump forwards and backwards, e.g. when the you adjust your
		4593	clock.
		4594
		4595	=item watcher
		4596
		4597	A data structure that describes interest in certain events. Watchers need
		4598	to be started (attached to an event loop) before they can receive events.
		4599
		4600	=item watcher invocation
		4601
		4602	The act of calling the callback associated with a watcher.
		4603
		4604	=back
3534		4605
3535	=head1 AUTHOR	4606	=head1 AUTHOR
3536		4607
3537	Marc Lehmann <libev@schmorp.de>.	4608	Marc Lehmann <libev@schmorp.de>, with repeated corrections by Mikael Magnusson.
3538		4609

Diff Legend

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