ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/libev/ev.pod

Comparing libev/ev.pod (file contents):
Revision 1.188 by root, Tue Sep 30 18:35:47 2008 UTC vs.
Revision 1.256 by root, Tue Jul 14 20:31:21 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
…		…
103	Libev is very configurable. In this manual the default (and most common)	117	Libev is very configurable. In this manual the default (and most common)
104	configuration will be described, which supports multiple event loops. For	118	configuration will be described, which supports multiple event loops. For
105	more info about various configuration options please have a look at	119	more info about various configuration options please have a look at
106	B<EMBED> section in this manual. If libev was configured without support	120	B<EMBED> section in this manual. If libev was configured without support
107	for multiple event loops, then all functions taking an initial argument of	121	for multiple event loops, then all functions taking an initial argument of
108	name C<loop> (which is always of type C<struct ev_loop *>) will not have	122	name C<loop> (which is always of type C<ev_loop *>) will not have
109	this argument.	123	this argument.
110		124
111	=head2 TIME REPRESENTATION	125	=head2 TIME REPRESENTATION
112		126
113	Libev represents time as a single floating point number, representing the	127	Libev represents time as a single floating point number, representing
114	(fractional) number of seconds since the (POSIX) epoch (somewhere near	128	the (fractional) number of seconds since the (POSIX) epoch (somewhere
115	the beginning of 1970, details are complicated, don't ask). This type is	129	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	130	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	131	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	132	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	133	component C<stamp> might indicate, it is also used for time differences
120	throughout libev.	134	throughout libev.
121		135
122	=head1 ERROR HANDLING	136	=head1 ERROR HANDLING
123		137
…		…
276		290
277	=back	291	=back
278		292
279	=head1 FUNCTIONS CONTROLLING THE EVENT LOOP	293	=head1 FUNCTIONS CONTROLLING THE EVENT LOOP
280		294
281	An event loop is described by a C<struct ev_loop *>. The library knows two	295	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	296	is I<not> optional in this case, as there is also an C<ev_loop>
283	events, and dynamically created loops which do not.	297	I<function>).
		298
		299	The library knows two types of such loops, the I<default> loop, which
		300	supports signals and child events, and dynamically created loops which do
		301	not.
284		302
285	=over 4	303	=over 4
286		304
287	=item struct ev_loop *ev_default_loop (unsigned int flags)	305	=item struct ev_loop *ev_default_loop (unsigned int flags)
288		306
…		…
294	If you don't know what event loop to use, use the one returned from this	312	If you don't know what event loop to use, use the one returned from this
295	function.	313	function.
296		314
297	Note that this function is I<not> thread-safe, so if you want to use it	315	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,	316	from multiple threads, you have to lock (note also that this is unlikely,
299	as loops cannot bes hared easily between threads anyway).	317	as loops cannot be shared easily between threads anyway).
300		318
301	The default loop is the only loop that can handle C<ev_signal> and	319	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	320	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	321	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	322	create a dynamic loop with C<ev_loop_new> that doesn't do that, or you
…		…
380	=item C<EVBACKEND_EPOLL> (value 4, Linux)	398	=item C<EVBACKEND_EPOLL> (value 4, Linux)
381		399
382	For few fds, this backend is a bit little slower than poll and select,	400	For few fds, this backend is a bit little slower than poll and select,
383	but it scales phenomenally better. While poll and select usually scale	401	but it scales phenomenally better. While poll and select usually scale
384	like O(total_fds) where n is the total number of fds (or the highest fd),	402	like O(total_fds) where n is the total number of fds (or the highest fd),
385	epoll scales either O(1) or O(active_fds). The epoll design has a number	403	epoll scales either O(1) or O(active_fds).
386	of shortcomings, such as silently dropping events in some hard-to-detect	404
387	cases and requiring a system call per fd change, no fork support and bad	405	The epoll mechanism deserves honorable mention as the most misdesigned
388	support for dup.	406	of the more advanced event mechanisms: mere annoyances include silently
		407	dropping file descriptors, requiring a system call per change per file
		408	descriptor (and unnecessary guessing of parameters), problems with dup and
		409	so on. The biggest issue is fork races, however - if a program forks then
		410	I<both> parent and child process have to recreate the epoll set, which can
		411	take considerable time (one syscall per file descriptor) and is of course
		412	hard to detect.
		413
		414	Epoll is also notoriously buggy - embedding epoll fds I<should> work, but
		415	of course I<doesn't>, and epoll just loves to report events for totally
		416	I<different> file descriptors (even already closed ones, so one cannot
		417	even remove them from the set) than registered in the set (especially
		418	on SMP systems). Libev tries to counter these spurious notifications by
		419	employing an additional generation counter and comparing that against the
		420	events to filter out spurious ones, recreating the set when required.
389		421
390	While stopping, setting and starting an I/O watcher in the same iteration	422	While stopping, setting and starting an I/O watcher in the same iteration
391	will result in some caching, there is still a system call per such incident	423	will result in some caching, there is still a system call per such
392	(because the fd could point to a different file description now), so its	424	incident (because the same I<file descriptor> could point to a different
393	best to avoid that. Also, C<dup ()>'ed file descriptors might not work	425	I<file description> now), so its best to avoid that. Also, C<dup ()>'ed
394	very well if you register events for both fds.	426	file descriptors might not work very well if you register events for both
395		427	file descriptors.
396	Please note that epoll sometimes generates spurious notifications, so you
397	need to use non-blocking I/O or other means to avoid blocking when no data
398	(or space) is available.
399		428
400	Best performance from this backend is achieved by not unregistering all	429	Best performance from this backend is achieved by not unregistering all
401	watchers for a file descriptor until it has been closed, if possible,	430	watchers for a file descriptor until it has been closed, if possible,
402	i.e. keep at least one watcher active per fd at all times. Stopping and	431	i.e. keep at least one watcher active per fd at all times. Stopping and
403	starting a watcher (without re-setting it) also usually doesn't cause	432	starting a watcher (without re-setting it) also usually doesn't cause
404	extra overhead.	433	extra overhead. A fork can both result in spurious notifications as well
		434	as in libev having to destroy and recreate the epoll object, which can
		435	take considerable time and thus should be avoided.
		436
		437	All this means that, in practice, C<EVBACKEND_SELECT> can be as fast or
		438	faster than epoll for maybe up to a hundred file descriptors, depending on
		439	the usage. So sad.
405		440
406	While nominally embeddable in other event loops, this feature is broken in	441	While nominally embeddable in other event loops, this feature is broken in
407	all kernel versions tested so far.	442	all kernel versions tested so far.
408		443
409	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as	444	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
410	C<EVBACKEND_POLL>.	445	C<EVBACKEND_POLL>.
411		446
412	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)	447	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)
413		448
414	Kqueue deserves special mention, as at the time of this writing, it was	449	Kqueue deserves special mention, as at the time of this writing, it
415	broken on all BSDs except NetBSD (usually it doesn't work reliably with	450	was broken on all BSDs except NetBSD (usually it doesn't work reliably
416	anything but sockets and pipes, except on Darwin, where of course it's	451	with anything but sockets and pipes, except on Darwin, where of course
417	completely useless). For this reason it's not being "auto-detected" unless	452	it's completely useless). Unlike epoll, however, whose brokenness
418	you explicitly specify it in the flags (i.e. using C<EVBACKEND_KQUEUE>) or	453	is by design, these kqueue bugs can (and eventually will) be fixed
419	libev was compiled on a known-to-be-good (-enough) system like NetBSD.	454	without API changes to existing programs. For this reason it's not being
		455	"auto-detected" unless you explicitly specify it in the flags (i.e. using
		456	C<EVBACKEND_KQUEUE>) or libev was compiled on a known-to-be-good (-enough)
		457	system like NetBSD.
420		458
421	You still can embed kqueue into a normal poll or select backend and use it	459	You still can embed kqueue into a normal poll or select backend and use it
422	only for sockets (after having made sure that sockets work with kqueue on	460	only for sockets (after having made sure that sockets work with kqueue on
423	the target platform). See C<ev_embed> watchers for more info.	461	the target platform). See C<ev_embed> watchers for more info.
424		462
425	It scales in the same way as the epoll backend, but the interface to the	463	It scales in the same way as the epoll backend, but the interface to the
426	kernel is more efficient (which says nothing about its actual speed, of	464	kernel is more efficient (which says nothing about its actual speed, of
427	course). While stopping, setting and starting an I/O watcher does never	465	course). While stopping, setting and starting an I/O watcher does never
428	cause an extra system call as with C<EVBACKEND_EPOLL>, it still adds up to	466	cause an extra system call as with C<EVBACKEND_EPOLL>, it still adds up to
429	two event changes per incident. Support for C<fork ()> is very bad and it	467	two event changes per incident. Support for C<fork ()> is very bad (but
430	drops fds silently in similarly hard-to-detect cases.	468	sane, unlike epoll) and it drops fds silently in similarly hard-to-detect
		469	cases
431		470
432	This backend usually performs well under most conditions.	471	This backend usually performs well under most conditions.
433		472
434	While nominally embeddable in other event loops, this doesn't work	473	While nominally embeddable in other event loops, this doesn't work
435	everywhere, so you might need to test for this. And since it is broken	474	everywhere, so you might need to test for this. And since it is broken
436	almost everywhere, you should only use it when you have a lot of sockets	475	almost everywhere, you should only use it when you have a lot of sockets
437	(for which it usually works), by embedding it into another event loop	476	(for which it usually works), by embedding it into another event loop
438	(e.g. C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>) and, did I mention it,	477	(e.g. C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> (but C<poll> is of course
439	using it only for sockets.	478	also broken on OS X)) and, did I mention it, using it only for sockets.
440		479
441	This backend maps C<EV_READ> into an C<EVFILT_READ> kevent with	480	This backend maps C<EV_READ> into an C<EVFILT_READ> kevent with
442	C<NOTE_EOF>, and C<EV_WRITE> into an C<EVFILT_WRITE> kevent with	481	C<NOTE_EOF>, and C<EV_WRITE> into an C<EVFILT_WRITE> kevent with
443	C<NOTE_EOF>.	482	C<NOTE_EOF>.
444		483
…		…
464	might perform better.	503	might perform better.
465		504
466	On the positive side, with the exception of the spurious readiness	505	On the positive side, with the exception of the spurious readiness
467	notifications, this backend actually performed fully to specification	506	notifications, this backend actually performed fully to specification
468	in all tests and is fully embeddable, which is a rare feat among the	507	in all tests and is fully embeddable, which is a rare feat among the
469	OS-specific backends.	508	OS-specific backends (I vastly prefer correctness over speed hacks).
470		509
471	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as	510	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
472	C<EVBACKEND_POLL>.	511	C<EVBACKEND_POLL>.
473		512
474	=item C<EVBACKEND_ALL>	513	=item C<EVBACKEND_ALL>
…		…
527	responsibility to either stop all watchers cleanly yourself I<before>	566	responsibility to either stop all watchers cleanly yourself I<before>
528	calling this function, or cope with the fact afterwards (which is usually	567	calling this function, or cope with the fact afterwards (which is usually
529	the easiest thing, you can just ignore the watchers and/or C<free ()> them	568	the easiest thing, you can just ignore the watchers and/or C<free ()> them
530	for example).	569	for example).
531		570
532	Note that certain global state, such as signal state, will not be freed by	571	Note that certain global state, such as signal state (and installed signal
533	this function, and related watchers (such as signal and child watchers)	572	handlers), will not be freed by this function, and related watchers (such
534	would need to be stopped manually.	573	as signal and child watchers) would need to be stopped manually.
535		574
536	In general it is not advisable to call this function except in the	575	In general it is not advisable to call this function except in the
537	rare occasion where you really need to free e.g. the signal handling	576	rare occasion where you really need to free e.g. the signal handling
538	pipe fds. If you need dynamically allocated loops it is better to use	577	pipe fds. If you need dynamically allocated loops it is better to use
539	C<ev_loop_new> and C<ev_loop_destroy>).	578	C<ev_loop_new> and C<ev_loop_destroy>).
…		…
582		621
583	This value can sometimes be useful as a generation counter of sorts (it	622	This value can sometimes be useful as a generation counter of sorts (it
584	"ticks" the number of loop iterations), as it roughly corresponds with	623	"ticks" the number of loop iterations), as it roughly corresponds with
585	C<ev_prepare> and C<ev_check> calls.	624	C<ev_prepare> and C<ev_check> calls.
586		625
		626	=item unsigned int ev_loop_depth (loop)
		627
		628	Returns the number of times C<ev_loop> was entered minus the number of
		629	times C<ev_loop> was exited, in other words, the recursion depth.
		630
		631	Outside C<ev_loop>, this number is zero. In a callback, this number is
		632	C<1>, unless C<ev_loop> was invoked recursively (or from another thread),
		633	in which case it is higher.
		634
		635	Leaving C<ev_loop> abnormally (setjmp/longjmp, cancelling the thread
		636	etc.), doesn't count as exit.
		637
587	=item unsigned int ev_backend (loop)	638	=item unsigned int ev_backend (loop)
588		639
589	Returns one of the C<EVBACKEND_*> flags indicating the event backend in	640	Returns one of the C<EVBACKEND_*> flags indicating the event backend in
590	use.	641	use.
591		642
…		…
605		656
606	This function is rarely useful, but when some event callback runs for a	657	This function is rarely useful, but when some event callback runs for a
607	very long time without entering the event loop, updating libev's idea of	658	very long time without entering the event loop, updating libev's idea of
608	the current time is a good idea.	659	the current time is a good idea.
609		660
610	See also "The special problem of time updates" in the C<ev_timer> section.	661	See also L<The special problem of time updates> in the C<ev_timer> section.
		662
		663	=item ev_suspend (loop)
		664
		665	=item ev_resume (loop)
		666
		667	These two functions suspend and resume a loop, for use when the loop is
		668	not used for a while and timeouts should not be processed.
		669
		670	A typical use case would be an interactive program such as a game: When
		671	the user presses C<^Z> to suspend the game and resumes it an hour later it
		672	would be best to handle timeouts as if no time had actually passed while
		673	the program was suspended. This can be achieved by calling C<ev_suspend>
		674	in your C<SIGTSTP> handler, sending yourself a C<SIGSTOP> and calling
		675	C<ev_resume> directly afterwards to resume timer processing.
		676
		677	Effectively, all C<ev_timer> watchers will be delayed by the time spend
		678	between C<ev_suspend> and C<ev_resume>, and all C<ev_periodic> watchers
		679	will be rescheduled (that is, they will lose any events that would have
		680	occured while suspended).
		681
		682	After calling C<ev_suspend> you B<must not> call I<any> function on the
		683	given loop other than C<ev_resume>, and you B<must not> call C<ev_resume>
		684	without a previous call to C<ev_suspend>.
		685
		686	Calling C<ev_suspend>/C<ev_resume> has the side effect of updating the
		687	event loop time (see C<ev_now_update>).
611		688
612	=item ev_loop (loop, int flags)	689	=item ev_loop (loop, int flags)
613		690
614	Finally, this is it, the event handler. This function usually is called	691	Finally, this is it, the event handler. This function usually is called
615	after you initialised all your watchers and you want to start handling	692	after you initialised all your watchers and you want to start handling
…		…
631	the loop.	708	the loop.
632		709
633	A flags value of C<EVLOOP_ONESHOT> will look for new events (waiting if	710	A flags value of C<EVLOOP_ONESHOT> will look for new events (waiting if
634	necessary) and will handle those and any already outstanding ones. It	711	necessary) and will handle those and any already outstanding ones. It
635	will block your process until at least one new event arrives (which could	712	will block your process until at least one new event arrives (which could
636	be an event internal to libev itself, so there is no guarentee that a	713	be an event internal to libev itself, so there is no guarantee that a
637	user-registered callback will be called), and will return after one	714	user-registered callback will be called), and will return after one
638	iteration of the loop.	715	iteration of the loop.
639		716
640	This is useful if you are waiting for some external event in conjunction	717	This is useful if you are waiting for some external event in conjunction
641	with something not expressible using other libev watchers (i.e. "roll your	718	with something not expressible using other libev watchers (i.e. "roll your
…		…
685	C<EVUNLOOP_ONE>, which will make the innermost C<ev_loop> call return, or	762	C<EVUNLOOP_ONE>, which will make the innermost C<ev_loop> call return, or
686	C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.	763	C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.
687		764
688	This "unloop state" will be cleared when entering C<ev_loop> again.	765	This "unloop state" will be cleared when entering C<ev_loop> again.
689		766
		767	It is safe to call C<ev_unloop> from otuside any C<ev_loop> calls.
		768
690	=item ev_ref (loop)	769	=item ev_ref (loop)
691		770
692	=item ev_unref (loop)	771	=item ev_unref (loop)
693		772
694	Ref/unref can be used to add or remove a reference count on the event	773	Ref/unref can be used to add or remove a reference count on the event
…		…
697		776
698	If you have a watcher you never unregister that should not keep C<ev_loop>	777	If you have a watcher you never unregister that should not keep C<ev_loop>
699	from returning, call ev_unref() after starting, and ev_ref() before	778	from returning, call ev_unref() after starting, and ev_ref() before
700	stopping it.	779	stopping it.
701		780
702	As an example, libev itself uses this for its internal signal pipe: It is	781	As an example, libev itself uses this for its internal signal pipe: It
703	not visible to the libev user and should not keep C<ev_loop> from exiting	782	is not visible to the libev user and should not keep C<ev_loop> from
704	if no event watchers registered by it are active. It is also an excellent	783	exiting if no event watchers registered by it are active. It is also an
705	way to do this for generic recurring timers or from within third-party	784	excellent way to do this for generic recurring timers or from within
706	libraries. Just remember to I<unref after start> and I<ref before stop>	785	third-party libraries. Just remember to I<unref after start> and I<ref
707	(but only if the watcher wasn't active before, or was active before,	786	before stop> (but only if the watcher wasn't active before, or was active
708	respectively).	787	before, respectively. Note also that libev might stop watchers itself
		788	(e.g. non-repeating timers) in which case you have to C<ev_ref>
		789	in the callback).
709		790
710	Example: Create a signal watcher, but keep it from keeping C<ev_loop>	791	Example: Create a signal watcher, but keep it from keeping C<ev_loop>
711	running when nothing else is active.	792	running when nothing else is active.
712		793
713	struct ev_signal exitsig;	794	ev_signal exitsig;
714	ev_signal_init (&exitsig, sig_cb, SIGINT);	795	ev_signal_init (&exitsig, sig_cb, SIGINT);
715	ev_signal_start (loop, &exitsig);	796	ev_signal_start (loop, &exitsig);
716	evf_unref (loop);	797	evf_unref (loop);
717		798
718	Example: For some weird reason, unregister the above signal handler again.	799	Example: For some weird reason, unregister the above signal handler again.
…		…
742		823
743	By setting a higher I<io collect interval> you allow libev to spend more	824	By setting a higher I<io collect interval> you allow libev to spend more
744	time collecting I/O events, so you can handle more events per iteration,	825	time collecting I/O events, so you can handle more events per iteration,
745	at the cost of increasing latency. Timeouts (both C<ev_periodic> and	826	at the cost of increasing latency. Timeouts (both C<ev_periodic> and
746	C<ev_timer>) will be not affected. Setting this to a non-null value will	827	C<ev_timer>) will be not affected. Setting this to a non-null value will
747	introduce an additional C<ev_sleep ()> call into most loop iterations.	828	introduce an additional C<ev_sleep ()> call into most loop iterations. The
		829	sleep time ensures that libev will not poll for I/O events more often then
		830	once per this interval, on average.
748		831
749	Likewise, by setting a higher I<timeout collect interval> you allow libev	832	Likewise, by setting a higher I<timeout collect interval> you allow libev
750	to spend more time collecting timeouts, at the expense of increased	833	to spend more time collecting timeouts, at the expense of increased
751	latency/jitter/inexactness (the watcher callback will be called	834	latency/jitter/inexactness (the watcher callback will be called
752	later). C<ev_io> watchers will not be affected. Setting this to a non-null	835	later). C<ev_io> watchers will not be affected. Setting this to a non-null
…		…
754		837
755	Many (busy) programs can usually benefit by setting the I/O collect	838	Many (busy) programs can usually benefit by setting the I/O collect
756	interval to a value near C<0.1> or so, which is often enough for	839	interval to a value near C<0.1> or so, which is often enough for
757	interactive servers (of course not for games), likewise for timeouts. It	840	interactive servers (of course not for games), likewise for timeouts. It
758	usually doesn't make much sense to set it to a lower value than C<0.01>,	841	usually doesn't make much sense to set it to a lower value than C<0.01>,
759	as this approaches the timing granularity of most systems.	842	as this approaches the timing granularity of most systems. Note that if
		843	you do transactions with the outside world and you can't increase the
		844	parallelity, then this setting will limit your transaction rate (if you
		845	need to poll once per transaction and the I/O collect interval is 0.01,
		846	then you can't do more than 100 transations per second).
760		847
761	Setting the I<timeout collect interval> can improve the opportunity for	848	Setting the I<timeout collect interval> can improve the opportunity for
762	saving power, as the program will "bundle" timer callback invocations that	849	saving power, as the program will "bundle" timer callback invocations that
763	are "near" in time together, by delaying some, thus reducing the number of	850	are "near" in time together, by delaying some, thus reducing the number of
764	times the process sleeps and wakes up again. Another useful technique to	851	times the process sleeps and wakes up again. Another useful technique to
765	reduce iterations/wake-ups is to use C<ev_periodic> watchers and make sure	852	reduce iterations/wake-ups is to use C<ev_periodic> watchers and make sure
766	they fire on, say, one-second boundaries only.	853	they fire on, say, one-second boundaries only.
767		854
		855	Example: we only need 0.1s timeout granularity, and we wish not to poll
		856	more often than 100 times per second:
		857
		858	ev_set_timeout_collect_interval (EV_DEFAULT_UC_ 0.1);
		859	ev_set_io_collect_interval (EV_DEFAULT_UC_ 0.01);
		860
		861	=item ev_invoke_pending (loop)
		862
		863	This call will simply invoke all pending watchers while resetting their
		864	pending state. Normally, C<ev_loop> does this automatically when required,
		865	but when overriding the invoke callback this call comes handy.
		866
		867	=item int ev_pending_count (loop)
		868
		869	Returns the number of pending watchers - zero indicates that no watchers
		870	are pending.
		871
		872	=item ev_set_invoke_pending_cb (loop, void (*invoke_pending_cb)(EV_P))
		873
		874	This overrides the invoke pending functionality of the loop: Instead of
		875	invoking all pending watchers when there are any, C<ev_loop> will call
		876	this callback instead. This is useful, for example, when you want to
		877	invoke the actual watchers inside another context (another thread etc.).
		878
		879	If you want to reset the callback, use C<ev_invoke_pending> as new
		880	callback.
		881
		882	=item ev_set_loop_release_cb (loop, void (*release)(EV_P), void (*acquire)(EV_P))
		883
		884	Sometimes you want to share the same loop between multiple threads. This
		885	can be done relatively simply by putting mutex_lock/unlock calls around
		886	each call to a libev function.
		887
		888	However, C<ev_loop> can run an indefinite time, so it is not feasible to
		889	wait for it to return. One way around this is to wake up the loop via
		890	C<ev_unloop> and C<av_async_send>, another way is to set these I<release>
		891	and I<acquire> callbacks on the loop.
		892
		893	When set, then C<release> will be called just before the thread is
		894	suspended waiting for new events, and C<acquire> is called just
		895	afterwards.
		896
		897	Ideally, C<release> will just call your mutex_unlock function, and
		898	C<acquire> will just call the mutex_lock function again.
		899
		900	While event loop modifications are allowed between invocations of
		901	C<release> and C<acquire> (that's their only purpose after all), no
		902	modifications done will affect the event loop, i.e. adding watchers will
		903	have no effect on the set of file descriptors being watched, or the time
		904	waited. USe an C<ev_async> watcher to wake up C<ev_loop> when you want it
		905	to take note of any changes you made.
		906
		907	In theory, threads executing C<ev_loop> will be async-cancel safe between
		908	invocations of C<release> and C<acquire>.
		909
		910	See also the locking example in the C<THREADS> section later in this
		911	document.
		912
		913	=item ev_set_userdata (loop, void *data)
		914
		915	=item ev_userdata (loop)
		916
		917	Set and retrieve a single C<void *> associated with a loop. When
		918	C<ev_set_userdata> has never been called, then C<ev_userdata> returns
		919	C<0.>
		920
		921	These two functions can be used to associate arbitrary data with a loop,
		922	and are intended solely for the C<invoke_pending_cb>, C<release> and
		923	C<acquire> callbacks described above, but of course can be (ab-)used for
		924	any other purpose as well.
		925
768	=item ev_loop_verify (loop)	926	=item ev_loop_verify (loop)
769		927
770	This function only does something when C<EV_VERIFY> support has been	928	This function only does something when C<EV_VERIFY> support has been
771	compiled in. which is the default for non-minimal builds. It tries to go	929	compiled in, which is the default for non-minimal builds. It tries to go
772	through all internal structures and checks them for validity. If anything	930	through all internal structures and checks them for validity. If anything
773	is found to be inconsistent, it will print an error message to standard	931	is found to be inconsistent, it will print an error message to standard
774	error and call C<abort ()>.	932	error and call C<abort ()>.
775		933
776	This can be used to catch bugs inside libev itself: under normal	934	This can be used to catch bugs inside libev itself: under normal
…		…
780	=back	938	=back
781		939
782		940
783	=head1 ANATOMY OF A WATCHER	941	=head1 ANATOMY OF A WATCHER
784		942
		943	In the following description, uppercase C<TYPE> in names stands for the
		944	watcher type, e.g. C<ev_TYPE_start> can mean C<ev_timer_start> for timer
		945	watchers and C<ev_io_start> for I/O watchers.
		946
785	A watcher is a structure that you create and register to record your	947	A watcher is a structure that you create and register to record your
786	interest in some event. For instance, if you want to wait for STDIN to	948	interest in some event. For instance, if you want to wait for STDIN to
787	become readable, you would create an C<ev_io> watcher for that:	949	become readable, you would create an C<ev_io> watcher for that:
788		950
789	static void my_cb (struct ev_loop *loop, struct ev_io *w, int revents)	951	static void my_cb (struct ev_loop *loop, ev_io *w, int revents)
790	{	952	{
791	ev_io_stop (w);	953	ev_io_stop (w);
792	ev_unloop (loop, EVUNLOOP_ALL);	954	ev_unloop (loop, EVUNLOOP_ALL);
793	}	955	}
794		956
795	struct ev_loop *loop = ev_default_loop (0);	957	struct ev_loop *loop = ev_default_loop (0);
		958
796	struct ev_io stdin_watcher;	959	ev_io stdin_watcher;
		960
797	ev_init (&stdin_watcher, my_cb);	961	ev_init (&stdin_watcher, my_cb);
798	ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ);	962	ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ);
799	ev_io_start (loop, &stdin_watcher);	963	ev_io_start (loop, &stdin_watcher);
		964
800	ev_loop (loop, 0);	965	ev_loop (loop, 0);
801		966
802	As you can see, you are responsible for allocating the memory for your	967	As you can see, you are responsible for allocating the memory for your
803	watcher structures (and it is usually a bad idea to do this on the stack,	968	watcher structures (and it is I<usually> a bad idea to do this on the
804	although this can sometimes be quite valid).	969	stack).
		970
		971	Each watcher has an associated watcher structure (called C<struct ev_TYPE>
		972	or simply C<ev_TYPE>, as typedefs are provided for all watcher structs).
805		973
806	Each watcher structure must be initialised by a call to C<ev_init	974	Each watcher structure must be initialised by a call to C<ev_init
807	(watcher *, callback)>, which expects a callback to be provided. This	975	(watcher *, callback)>, which expects a callback to be provided. This
808	callback gets invoked each time the event occurs (or, in the case of I/O	976	callback gets invoked each time the event occurs (or, in the case of I/O
809	watchers, each time the event loop detects that the file descriptor given	977	watchers, each time the event loop detects that the file descriptor given
810	is readable and/or writable).	978	is readable and/or writable).
811		979
812	Each watcher type has its own C<< ev_<type>_set (watcher *, ...) >> macro	980	Each watcher type further has its own C<< ev_TYPE_set (watcher *, ...) >>
813	with arguments specific to this watcher type. There is also a macro	981	macro to configure it, with arguments specific to the watcher type. There
814	to combine initialisation and setting in one call: C<< ev_<type>_init	982	is also a macro to combine initialisation and setting in one call: C<<
815	(watcher *, callback, ...) >>.	983	ev_TYPE_init (watcher *, callback, ...) >>.
816		984
817	To make the watcher actually watch out for events, you have to start it	985	To make the watcher actually watch out for events, you have to start it
818	with a watcher-specific start function (C<< ev_<type>_start (loop, watcher	986	with a watcher-specific start function (C<< ev_TYPE_start (loop, watcher
819	*) >>), and you can stop watching for events at any time by calling the	987	*) >>), and you can stop watching for events at any time by calling the
820	corresponding stop function (C<< ev_<type>_stop (loop, watcher *) >>.	988	corresponding stop function (C<< ev_TYPE_stop (loop, watcher *) >>.
821		989
822	As long as your watcher is active (has been started but not stopped) you	990	As long as your watcher is active (has been started but not stopped) you
823	must not touch the values stored in it. Most specifically you must never	991	must not touch the values stored in it. Most specifically you must never
824	reinitialise it or call its C<set> macro.	992	reinitialise it or call its C<ev_TYPE_set> macro.
825		993
826	Each and every callback receives the event loop pointer as first, the	994	Each and every callback receives the event loop pointer as first, the
827	registered watcher structure as second, and a bitset of received events as	995	registered watcher structure as second, and a bitset of received events as
828	third argument.	996	third argument.
829		997
…		…
887		1055
888	=item C<EV_ASYNC>	1056	=item C<EV_ASYNC>
889		1057
890	The given async watcher has been asynchronously notified (see C<ev_async>).	1058	The given async watcher has been asynchronously notified (see C<ev_async>).
891		1059
		1060	=item C<EV_CUSTOM>
		1061
		1062	Not ever sent (or otherwise used) by libev itself, but can be freely used
		1063	by libev users to signal watchers (e.g. via C<ev_feed_event>).
		1064
892	=item C<EV_ERROR>	1065	=item C<EV_ERROR>
893		1066
894	An unspecified error has occurred, the watcher has been stopped. This might	1067	An unspecified error has occurred, the watcher has been stopped. This might
895	happen because the watcher could not be properly started because libev	1068	happen because the watcher could not be properly started because libev
896	ran out of memory, a file descriptor was found to be closed or any other	1069	ran out of memory, a file descriptor was found to be closed or any other
		1070	problem. Libev considers these application bugs.
		1071
897	problem. You best act on it by reporting the problem and somehow coping	1072	You best act on it by reporting the problem and somehow coping with the
898	with the watcher being stopped.	1073	watcher being stopped. Note that well-written programs should not receive
		1074	an error ever, so when your watcher receives it, this usually indicates a
		1075	bug in your program.
899		1076
900	Libev will usually signal a few "dummy" events together with an error, for	1077	Libev will usually signal a few "dummy" events together with an error, for
901	example it might indicate that a fd is readable or writable, and if your	1078	example it might indicate that a fd is readable or writable, and if your
902	callbacks is well-written it can just attempt the operation and cope with	1079	callbacks is well-written it can just attempt the operation and cope with
903	the error from read() or write(). This will not work in multi-threaded	1080	the error from read() or write(). This will not work in multi-threaded
…		…
906		1083
907	=back	1084	=back
908		1085
909	=head2 GENERIC WATCHER FUNCTIONS	1086	=head2 GENERIC WATCHER FUNCTIONS
910		1087
911	In the following description, C<TYPE> stands for the watcher type,
912	e.g. C<timer> for C<ev_timer> watchers and C<io> for C<ev_io> watchers.
913
914	=over 4	1088	=over 4
915		1089
916	=item C<ev_init> (ev_TYPE *watcher, callback)	1090	=item C<ev_init> (ev_TYPE *watcher, callback)
917		1091
918	This macro initialises the generic portion of a watcher. The contents	1092	This macro initialises the generic portion of a watcher. The contents
…		…
923	which rolls both calls into one.	1097	which rolls both calls into one.
924		1098
925	You can reinitialise a watcher at any time as long as it has been stopped	1099	You can reinitialise a watcher at any time as long as it has been stopped
926	(or never started) and there are no pending events outstanding.	1100	(or never started) and there are no pending events outstanding.
927		1101
928	The callback is always of type C<void (*)(ev_loop *loop, ev_TYPE *watcher,	1102	The callback is always of type C<void (*)(struct ev_loop *loop, ev_TYPE *watcher,
929	int revents)>.	1103	int revents)>.
930		1104
931	Example: Initialise an C<ev_io> watcher in two steps.	1105	Example: Initialise an C<ev_io> watcher in two steps.
932		1106
933	ev_io w;	1107	ev_io w;
…		…
967		1141
968	ev_io_start (EV_DEFAULT_UC, &w);	1142	ev_io_start (EV_DEFAULT_UC, &w);
969		1143
970	=item C<ev_TYPE_stop> (loop *, ev_TYPE *watcher)	1144	=item C<ev_TYPE_stop> (loop *, ev_TYPE *watcher)
971		1145
972	Stops the given watcher again (if active) and clears the pending	1146	Stops the given watcher if active, and clears the pending status (whether
		1147	the watcher was active or not).
		1148
973	status. It is possible that stopped watchers are pending (for example,	1149	It is possible that stopped watchers are pending - for example,
974	non-repeating timers are being stopped when they become pending), but	1150	non-repeating timers are being stopped when they become pending - but
975	C<ev_TYPE_stop> ensures that the watcher is neither active nor pending. If	1151	calling C<ev_TYPE_stop> ensures that the watcher is neither active nor
976	you want to free or reuse the memory used by the watcher it is therefore a	1152	pending. If you want to free or reuse the memory used by the watcher it is
977	good idea to always call its C<ev_TYPE_stop> function.	1153	therefore a good idea to always call its C<ev_TYPE_stop> function.
978		1154
979	=item bool ev_is_active (ev_TYPE *watcher)	1155	=item bool ev_is_active (ev_TYPE *watcher)
980		1156
981	Returns a true value iff the watcher is active (i.e. it has been started	1157	Returns a true value iff the watcher is active (i.e. it has been started
982	and not yet been stopped). As long as a watcher is active you must not modify	1158	and not yet been stopped). As long as a watcher is active you must not modify
…		…
1008	integer between C<EV_MAXPRI> (default: C<2>) and C<EV_MINPRI>	1184	integer between C<EV_MAXPRI> (default: C<2>) and C<EV_MINPRI>
1009	(default: C<-2>). Pending watchers with higher priority will be invoked	1185	(default: C<-2>). Pending watchers with higher priority will be invoked
1010	before watchers with lower priority, but priority will not keep watchers	1186	before watchers with lower priority, but priority will not keep watchers
1011	from being executed (except for C<ev_idle> watchers).	1187	from being executed (except for C<ev_idle> watchers).
1012		1188
1013	This means that priorities are I<only> used for ordering callback
1014	invocation after new events have been received. This is useful, for
1015	example, to reduce latency after idling, or more often, to bind two
1016	watchers on the same event and make sure one is called first.
1017
1018	If you need to suppress invocation when higher priority events are pending	1189	If you need to suppress invocation when higher priority events are pending
1019	you need to look at C<ev_idle> watchers, which provide this functionality.	1190	you need to look at C<ev_idle> watchers, which provide this functionality.
1020		1191
1021	You I<must not> change the priority of a watcher as long as it is active or	1192	You I<must not> change the priority of a watcher as long as it is active or
1022	pending.	1193	pending.
1023		1194
		1195	Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is
		1196	fine, as long as you do not mind that the priority value you query might
		1197	or might not have been clamped to the valid range.
		1198
1024	The default priority used by watchers when no priority has been set is	1199	The default priority used by watchers when no priority has been set is
1025	always C<0>, which is supposed to not be too high and not be too low :).	1200	always C<0>, which is supposed to not be too high and not be too low :).
1026		1201
1027	Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is	1202	See L<WATCHER PRIORITY MODELS>, below, for a more thorough treatment of
1028	fine, as long as you do not mind that the priority value you query might	1203	priorities.
1029	or might not have been adjusted to be within valid range.
1030		1204
1031	=item ev_invoke (loop, ev_TYPE *watcher, int revents)	1205	=item ev_invoke (loop, ev_TYPE *watcher, int revents)
1032		1206
1033	Invoke the C<watcher> with the given C<loop> and C<revents>. Neither	1207	Invoke the C<watcher> with the given C<loop> and C<revents>. Neither
1034	C<loop> nor C<revents> need to be valid as long as the watcher callback	1208	C<loop> nor C<revents> need to be valid as long as the watcher callback
…		…
1056	member, you can also "subclass" the watcher type and provide your own	1230	member, you can also "subclass" the watcher type and provide your own
1057	data:	1231	data:
1058		1232
1059	struct my_io	1233	struct my_io
1060	{	1234	{
1061	struct ev_io io;	1235	ev_io io;
1062	int otherfd;	1236	int otherfd;
1063	void *somedata;	1237	void *somedata;
1064	struct whatever *mostinteresting;	1238	struct whatever *mostinteresting;
1065	};	1239	};
1066		1240
…		…
1069	ev_io_init (&w.io, my_cb, fd, EV_READ);	1243	ev_io_init (&w.io, my_cb, fd, EV_READ);
1070		1244
1071	And since your callback will be called with a pointer to the watcher, you	1245	And since your callback will be called with a pointer to the watcher, you
1072	can cast it back to your own type:	1246	can cast it back to your own type:
1073		1247
1074	static void my_cb (struct ev_loop *loop, struct ev_io *w_, int revents)	1248	static void my_cb (struct ev_loop *loop, ev_io *w_, int revents)
1075	{	1249	{
1076	struct my_io *w = (struct my_io *)w_;	1250	struct my_io *w = (struct my_io *)w_;
1077	...	1251	...
1078	}	1252	}
1079		1253
…		…
1097	programmers):	1271	programmers):
1098		1272
1099	#include <stddef.h>	1273	#include <stddef.h>
1100		1274
1101	static void	1275	static void
1102	t1_cb (EV_P_ struct ev_timer *w, int revents)	1276	t1_cb (EV_P_ ev_timer *w, int revents)
1103	{	1277	{
1104	struct my_biggy big = (struct my_biggy *	1278	struct my_biggy big = (struct my_biggy *)
1105	(((char *)w) - offsetof (struct my_biggy, t1));	1279	(((char *)w) - offsetof (struct my_biggy, t1));
1106	}	1280	}
1107		1281
1108	static void	1282	static void
1109	t2_cb (EV_P_ struct ev_timer *w, int revents)	1283	t2_cb (EV_P_ ev_timer *w, int revents)
1110	{	1284	{
1111	struct my_biggy big = (struct my_biggy *	1285	struct my_biggy big = (struct my_biggy *)
1112	(((char *)w) - offsetof (struct my_biggy, t2));	1286	(((char *)w) - offsetof (struct my_biggy, t2));
1113	}	1287	}
		1288
		1289	=head2 WATCHER PRIORITY MODELS
		1290
		1291	Many event loops support I<watcher priorities>, which are usually small
		1292	integers that influence the ordering of event callback invocation
		1293	between watchers in some way, all else being equal.
		1294
		1295	In libev, Watcher priorities can be set using C<ev_set_priority>. See its
		1296	description for the more technical details such as the actual priority
		1297	range.
		1298
		1299	There are two common ways how these these priorities are being interpreted
		1300	by event loops:
		1301
		1302	In the more common lock-out model, higher priorities "lock out" invocation
		1303	of lower priority watchers, which means as long as higher priority
		1304	watchers receive events, lower priority watchers are not being invoked.
		1305
		1306	The less common only-for-ordering model uses priorities solely to order
		1307	callback invocation within a single event loop iteration: Higher priority
		1308	watchers are invoked before lower priority ones, but they all get invoked
		1309	before polling for new events.
		1310
		1311	Libev uses the second (only-for-ordering) model for all its watchers
		1312	except for idle watchers (which use the lock-out model).
		1313
		1314	The rationale behind this is that implementing the lock-out model for
		1315	watchers is not well supported by most kernel interfaces, and most event
		1316	libraries will just poll for the same events again and again as long as
		1317	their callbacks have not been executed, which is very inefficient in the
		1318	common case of one high-priority watcher locking out a mass of lower
		1319	priority ones.
		1320
		1321	Static (ordering) priorities are most useful when you have two or more
		1322	watchers handling the same resource: a typical usage example is having an
		1323	C<ev_io> watcher to receive data, and an associated C<ev_timer> to handle
		1324	timeouts. Under load, data might be received while the program handles
		1325	other jobs, but since timers normally get invoked first, the timeout
		1326	handler will be executed before checking for data. In that case, giving
		1327	the timer a lower priority than the I/O watcher ensures that I/O will be
		1328	handled first even under adverse conditions (which is usually, but not
		1329	always, what you want).
		1330
		1331	Since idle watchers use the "lock-out" model, meaning that idle watchers
		1332	will only be executed when no same or higher priority watchers have
		1333	received events, they can be used to implement the "lock-out" model when
		1334	required.
		1335
		1336	For example, to emulate how many other event libraries handle priorities,
		1337	you can associate an C<ev_idle> watcher to each such watcher, and in
		1338	the normal watcher callback, you just start the idle watcher. The real
		1339	processing is done in the idle watcher callback. This causes libev to
		1340	continously poll and process kernel event data for the watcher, but when
		1341	the lock-out case is known to be rare (which in turn is rare :), this is
		1342	workable.
		1343
		1344	Usually, however, the lock-out model implemented that way will perform
		1345	miserably under the type of load it was designed to handle. In that case,
		1346	it might be preferable to stop the real watcher before starting the
		1347	idle watcher, so the kernel will not have to process the event in case
		1348	the actual processing will be delayed for considerable time.
		1349
		1350	Here is an example of an I/O watcher that should run at a strictly lower
		1351	priority than the default, and which should only process data when no
		1352	other events are pending:
		1353
		1354	ev_idle idle; // actual processing watcher
		1355	ev_io io; // actual event watcher
		1356
		1357	static void
		1358	io_cb (EV_P_ ev_io *w, int revents)
		1359	{
		1360	// stop the I/O watcher, we received the event, but
		1361	// are not yet ready to handle it.
		1362	ev_io_stop (EV_A_ w);
		1363
		1364	// start the idle watcher to ahndle the actual event.
		1365	// it will not be executed as long as other watchers
		1366	// with the default priority are receiving events.
		1367	ev_idle_start (EV_A_ &idle);
		1368	}
		1369
		1370	static void
		1371	idle_cb (EV_P_ ev_idle *w, int revents)
		1372	{
		1373	// actual processing
		1374	read (STDIN_FILENO, ...);
		1375
		1376	// have to start the I/O watcher again, as
		1377	// we have handled the event
		1378	ev_io_start (EV_P_ &io);
		1379	}
		1380
		1381	// initialisation
		1382	ev_idle_init (&idle, idle_cb);
		1383	ev_io_init (&io, io_cb, STDIN_FILENO, EV_READ);
		1384	ev_io_start (EV_DEFAULT_ &io);
		1385
		1386	In the "real" world, it might also be beneficial to start a timer, so that
		1387	low-priority connections can not be locked out forever under load. This
		1388	enables your program to keep a lower latency for important connections
		1389	during short periods of high load, while not completely locking out less
		1390	important ones.
1114		1391
1115		1392
1116	=head1 WATCHER TYPES	1393	=head1 WATCHER TYPES
1117		1394
1118	This section describes each watcher in detail, but will not repeat	1395	This section describes each watcher in detail, but will not repeat
…		…
1144	descriptors to non-blocking mode is also usually a good idea (but not	1421	descriptors to non-blocking mode is also usually a good idea (but not
1145	required if you know what you are doing).	1422	required if you know what you are doing).
1146		1423
1147	If you cannot use non-blocking mode, then force the use of a	1424	If you cannot use non-blocking mode, then force the use of a
1148	known-to-be-good backend (at the time of this writing, this includes only	1425	known-to-be-good backend (at the time of this writing, this includes only
1149	C<EVBACKEND_SELECT> and C<EVBACKEND_POLL>).	1426	C<EVBACKEND_SELECT> and C<EVBACKEND_POLL>). The same applies to file
		1427	descriptors for which non-blocking operation makes no sense (such as
		1428	files) - libev doesn't guarentee any specific behaviour in that case.
1150		1429
1151	Another thing you have to watch out for is that it is quite easy to	1430	Another thing you have to watch out for is that it is quite easy to
1152	receive "spurious" readiness notifications, that is your callback might	1431	receive "spurious" readiness notifications, that is your callback might
1153	be called with C<EV_READ> but a subsequent C<read>(2) will actually block	1432	be called with C<EV_READ> but a subsequent C<read>(2) will actually block
1154	because there is no data. Not only are some backends known to create a	1433	because there is no data. Not only are some backends known to create a
…		…
1249	Example: Call C<stdin_readable_cb> when STDIN_FILENO has become, well	1528	Example: Call C<stdin_readable_cb> when STDIN_FILENO has become, well
1250	readable, but only once. Since it is likely line-buffered, you could	1529	readable, but only once. Since it is likely line-buffered, you could
1251	attempt to read a whole line in the callback.	1530	attempt to read a whole line in the callback.
1252		1531
1253	static void	1532	static void
1254	stdin_readable_cb (struct ev_loop *loop, struct ev_io *w, int revents)	1533	stdin_readable_cb (struct ev_loop *loop, ev_io *w, int revents)
1255	{	1534	{
1256	ev_io_stop (loop, w);	1535	ev_io_stop (loop, w);
1257	.. read from stdin here (or from w->fd) and handle any I/O errors	1536	.. read from stdin here (or from w->fd) and handle any I/O errors
1258	}	1537	}
1259		1538
1260	...	1539	...
1261	struct ev_loop *loop = ev_default_init (0);	1540	struct ev_loop *loop = ev_default_init (0);
1262	struct ev_io stdin_readable;	1541	ev_io stdin_readable;
1263	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);	1542	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);
1264	ev_io_start (loop, &stdin_readable);	1543	ev_io_start (loop, &stdin_readable);
1265	ev_loop (loop, 0);	1544	ev_loop (loop, 0);
1266		1545
1267		1546
…		…
1275	year, it will still time out after (roughly) one hour. "Roughly" because	1554	year, it will still time out after (roughly) one hour. "Roughly" because
1276	detecting time jumps is hard, and some inaccuracies are unavoidable (the	1555	detecting time jumps is hard, and some inaccuracies are unavoidable (the
1277	monotonic clock option helps a lot here).	1556	monotonic clock option helps a lot here).
1278		1557
1279	The callback is guaranteed to be invoked only I<after> its timeout has	1558	The callback is guaranteed to be invoked only I<after> its timeout has
1280	passed, but if multiple timers become ready during the same loop iteration	1559	passed (not I<at>, so on systems with very low-resolution clocks this
1281	then order of execution is undefined.	1560	might introduce a small delay). If multiple timers become ready during the
		1561	same loop iteration then the ones with earlier time-out values are invoked
		1562	before ones of the same priority with later time-out values (but this is
		1563	no longer true when a callback calls C<ev_loop> recursively).
		1564
		1565	=head3 Be smart about timeouts
		1566
		1567	Many real-world problems involve some kind of timeout, usually for error
		1568	recovery. A typical example is an HTTP request - if the other side hangs,
		1569	you want to raise some error after a while.
		1570
		1571	What follows are some ways to handle this problem, from obvious and
		1572	inefficient to smart and efficient.
		1573
		1574	In the following, a 60 second activity timeout is assumed - a timeout that
		1575	gets reset to 60 seconds each time there is activity (e.g. each time some
		1576	data or other life sign was received).
		1577
		1578	=over 4
		1579
		1580	=item 1. Use a timer and stop, reinitialise and start it on activity.
		1581
		1582	This is the most obvious, but not the most simple way: In the beginning,
		1583	start the watcher:
		1584
		1585	ev_timer_init (timer, callback, 60., 0.);
		1586	ev_timer_start (loop, timer);
		1587
		1588	Then, each time there is some activity, C<ev_timer_stop> it, initialise it
		1589	and start it again:
		1590
		1591	ev_timer_stop (loop, timer);
		1592	ev_timer_set (timer, 60., 0.);
		1593	ev_timer_start (loop, timer);
		1594
		1595	This is relatively simple to implement, but means that each time there is
		1596	some activity, libev will first have to remove the timer from its internal
		1597	data structure and then add it again. Libev tries to be fast, but it's
		1598	still not a constant-time operation.
		1599
		1600	=item 2. Use a timer and re-start it with C<ev_timer_again> inactivity.
		1601
		1602	This is the easiest way, and involves using C<ev_timer_again> instead of
		1603	C<ev_timer_start>.
		1604
		1605	To implement this, configure an C<ev_timer> with a C<repeat> value
		1606	of C<60> and then call C<ev_timer_again> at start and each time you
		1607	successfully read or write some data. If you go into an idle state where
		1608	you do not expect data to travel on the socket, you can C<ev_timer_stop>
		1609	the timer, and C<ev_timer_again> will automatically restart it if need be.
		1610
		1611	That means you can ignore both the C<ev_timer_start> function and the
		1612	C<after> argument to C<ev_timer_set>, and only ever use the C<repeat>
		1613	member and C<ev_timer_again>.
		1614
		1615	At start:
		1616
		1617	ev_init (timer, callback);
		1618	timer->repeat = 60.;
		1619	ev_timer_again (loop, timer);
		1620
		1621	Each time there is some activity:
		1622
		1623	ev_timer_again (loop, timer);
		1624
		1625	It is even possible to change the time-out on the fly, regardless of
		1626	whether the watcher is active or not:
		1627
		1628	timer->repeat = 30.;
		1629	ev_timer_again (loop, timer);
		1630
		1631	This is slightly more efficient then stopping/starting the timer each time
		1632	you want to modify its timeout value, as libev does not have to completely
		1633	remove and re-insert the timer from/into its internal data structure.
		1634
		1635	It is, however, even simpler than the "obvious" way to do it.
		1636
		1637	=item 3. Let the timer time out, but then re-arm it as required.
		1638
		1639	This method is more tricky, but usually most efficient: Most timeouts are
		1640	relatively long compared to the intervals between other activity - in
		1641	our example, within 60 seconds, there are usually many I/O events with
		1642	associated activity resets.
		1643
		1644	In this case, it would be more efficient to leave the C<ev_timer> alone,
		1645	but remember the time of last activity, and check for a real timeout only
		1646	within the callback:
		1647
		1648	ev_tstamp last_activity; // time of last activity
		1649
		1650	static void
		1651	callback (EV_P_ ev_timer *w, int revents)
		1652	{
		1653	ev_tstamp now = ev_now (EV_A);
		1654	ev_tstamp timeout = last_activity + 60.;
		1655
		1656	// if last_activity + 60. is older than now, we did time out
		1657	if (timeout < now)
		1658	{
		1659	// timeout occured, take action
		1660	}
		1661	else
		1662	{
		1663	// callback was invoked, but there was some activity, re-arm
		1664	// the watcher to fire in last_activity + 60, which is
		1665	// guaranteed to be in the future, so "again" is positive:
		1666	w->repeat = timeout - now;
		1667	ev_timer_again (EV_A_ w);
		1668	}
		1669	}
		1670
		1671	To summarise the callback: first calculate the real timeout (defined
		1672	as "60 seconds after the last activity"), then check if that time has
		1673	been reached, which means something I<did>, in fact, time out. Otherwise
		1674	the callback was invoked too early (C<timeout> is in the future), so
		1675	re-schedule the timer to fire at that future time, to see if maybe we have
		1676	a timeout then.
		1677
		1678	Note how C<ev_timer_again> is used, taking advantage of the
		1679	C<ev_timer_again> optimisation when the timer is already running.
		1680
		1681	This scheme causes more callback invocations (about one every 60 seconds
		1682	minus half the average time between activity), but virtually no calls to
		1683	libev to change the timeout.
		1684
		1685	To start the timer, simply initialise the watcher and set C<last_activity>
		1686	to the current time (meaning we just have some activity :), then call the
		1687	callback, which will "do the right thing" and start the timer:
		1688
		1689	ev_init (timer, callback);
		1690	last_activity = ev_now (loop);
		1691	callback (loop, timer, EV_TIMEOUT);
		1692
		1693	And when there is some activity, simply store the current time in
		1694	C<last_activity>, no libev calls at all:
		1695
		1696	last_actiivty = ev_now (loop);
		1697
		1698	This technique is slightly more complex, but in most cases where the
		1699	time-out is unlikely to be triggered, much more efficient.
		1700
		1701	Changing the timeout is trivial as well (if it isn't hard-coded in the
		1702	callback :) - just change the timeout and invoke the callback, which will
		1703	fix things for you.
		1704
		1705	=item 4. Wee, just use a double-linked list for your timeouts.
		1706
		1707	If there is not one request, but many thousands (millions...), all
		1708	employing some kind of timeout with the same timeout value, then one can
		1709	do even better:
		1710
		1711	When starting the timeout, calculate the timeout value and put the timeout
		1712	at the I<end> of the list.
		1713
		1714	Then use an C<ev_timer> to fire when the timeout at the I<beginning> of
		1715	the list is expected to fire (for example, using the technique #3).
		1716
		1717	When there is some activity, remove the timer from the list, recalculate
		1718	the timeout, append it to the end of the list again, and make sure to
		1719	update the C<ev_timer> if it was taken from the beginning of the list.
		1720
		1721	This way, one can manage an unlimited number of timeouts in O(1) time for
		1722	starting, stopping and updating the timers, at the expense of a major
		1723	complication, and having to use a constant timeout. The constant timeout
		1724	ensures that the list stays sorted.
		1725
		1726	=back
		1727
		1728	So which method the best?
		1729
		1730	Method #2 is a simple no-brain-required solution that is adequate in most
		1731	situations. Method #3 requires a bit more thinking, but handles many cases
		1732	better, and isn't very complicated either. In most case, choosing either
		1733	one is fine, with #3 being better in typical situations.
		1734
		1735	Method #1 is almost always a bad idea, and buys you nothing. Method #4 is
		1736	rather complicated, but extremely efficient, something that really pays
		1737	off after the first million or so of active timers, i.e. it's usually
		1738	overkill :)
1282		1739
1283	=head3 The special problem of time updates	1740	=head3 The special problem of time updates
1284		1741
1285	Establishing the current time is a costly operation (it usually takes at	1742	Establishing the current time is a costly operation (it usually takes at
1286	least two system calls): EV therefore updates its idea of the current	1743	least two system calls): EV therefore updates its idea of the current
…		…
1330	If the timer is started but non-repeating, stop it (as if it timed out).	1787	If the timer is started but non-repeating, stop it (as if it timed out).
1331		1788
1332	If the timer is repeating, either start it if necessary (with the	1789	If the timer is repeating, either start it if necessary (with the
1333	C<repeat> value), or reset the running timer to the C<repeat> value.	1790	C<repeat> value), or reset the running timer to the C<repeat> value.
1334		1791
1335	This sounds a bit complicated, but here is a useful and typical	1792	This sounds a bit complicated, see L<Be smart about timeouts>, above, for a
1336	example: Imagine you have a TCP connection and you want a so-called idle	1793	usage example.
1337	timeout, that is, you want to be called when there have been, say, 60
1338	seconds of inactivity on the socket. The easiest way to do this is to
1339	configure an C<ev_timer> with a C<repeat> value of C<60> and then call
1340	C<ev_timer_again> each time you successfully read or write some data. If
1341	you go into an idle state where you do not expect data to travel on the
1342	socket, you can C<ev_timer_stop> the timer, and C<ev_timer_again> will
1343	automatically restart it if need be.
1344
1345	That means you can ignore the C<after> value and C<ev_timer_start>
1346	altogether and only ever use the C<repeat> value and C<ev_timer_again>:
1347
1348	ev_timer_init (timer, callback, 0., 5.);
1349	ev_timer_again (loop, timer);
1350	...
1351	timer->again = 17.;
1352	ev_timer_again (loop, timer);
1353	...
1354	timer->again = 10.;
1355	ev_timer_again (loop, timer);
1356
1357	This is more slightly efficient then stopping/starting the timer each time
1358	you want to modify its timeout value.
1359
1360	Note, however, that it is often even more efficient to remember the
1361	time of the last activity and let the timer time-out naturally. In the
1362	callback, you then check whether the time-out is real, or, if there was
1363	some activity, you reschedule the watcher to time-out in "last_activity +
1364	timeout - ev_now ()" seconds.
1365		1794
1366	=item ev_tstamp repeat [read-write]	1795	=item ev_tstamp repeat [read-write]
1367		1796
1368	The current C<repeat> value. Will be used each time the watcher times out	1797	The current C<repeat> value. Will be used each time the watcher times out
1369	or C<ev_timer_again> is called, and determines the next timeout (if any),	1798	or C<ev_timer_again> is called, and determines the next timeout (if any),
…		…
1374	=head3 Examples	1803	=head3 Examples
1375		1804
1376	Example: Create a timer that fires after 60 seconds.	1805	Example: Create a timer that fires after 60 seconds.
1377		1806
1378	static void	1807	static void
1379	one_minute_cb (struct ev_loop *loop, struct ev_timer *w, int revents)	1808	one_minute_cb (struct ev_loop *loop, ev_timer *w, int revents)
1380	{	1809	{
1381	.. one minute over, w is actually stopped right here	1810	.. one minute over, w is actually stopped right here
1382	}	1811	}
1383		1812
1384	struct ev_timer mytimer;	1813	ev_timer mytimer;
1385	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);	1814	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);
1386	ev_timer_start (loop, &mytimer);	1815	ev_timer_start (loop, &mytimer);
1387		1816
1388	Example: Create a timeout timer that times out after 10 seconds of	1817	Example: Create a timeout timer that times out after 10 seconds of
1389	inactivity.	1818	inactivity.
1390		1819
1391	static void	1820	static void
1392	timeout_cb (struct ev_loop *loop, struct ev_timer *w, int revents)	1821	timeout_cb (struct ev_loop *loop, ev_timer *w, int revents)
1393	{	1822	{
1394	.. ten seconds without any activity	1823	.. ten seconds without any activity
1395	}	1824	}
1396		1825
1397	struct ev_timer mytimer;	1826	ev_timer mytimer;
1398	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */	1827	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */
1399	ev_timer_again (&mytimer); /* start timer */	1828	ev_timer_again (&mytimer); /* start timer */
1400	ev_loop (loop, 0);	1829	ev_loop (loop, 0);
1401		1830
1402	// and in some piece of code that gets executed on any "activity":	1831	// and in some piece of code that gets executed on any "activity":
…		…
1407	=head2 C<ev_periodic> - to cron or not to cron?	1836	=head2 C<ev_periodic> - to cron or not to cron?
1408		1837
1409	Periodic watchers are also timers of a kind, but they are very versatile	1838	Periodic watchers are also timers of a kind, but they are very versatile
1410	(and unfortunately a bit complex).	1839	(and unfortunately a bit complex).
1411		1840
1412	Unlike C<ev_timer>'s, they are not based on real time (or relative time)	1841	Unlike C<ev_timer>, periodic watchers are not based on real time (or
1413	but on wall clock time (absolute time). You can tell a periodic watcher	1842	relative time, the physical time that passes) but on wall clock time
1414	to trigger after some specific point in time. For example, if you tell a	1843	(absolute time, the thing you can read on your calender or clock). The
1415	periodic watcher to trigger in 10 seconds (by specifying e.g. C<ev_now ()	1844	difference is that wall clock time can run faster or slower than real
1416	+ 10.>, that is, an absolute time not a delay) and then reset your system	1845	time, and time jumps are not uncommon (e.g. when you adjust your
1417	clock to January of the previous year, then it will take more than year	1846	wrist-watch).
1418	to trigger the event (unlike an C<ev_timer>, which would still trigger
1419	roughly 10 seconds later as it uses a relative timeout).
1420		1847
		1848	You can tell a periodic watcher to trigger after some specific point
		1849	in time: for example, if you tell a periodic watcher to trigger "in 10
		1850	seconds" (by specifying e.g. C<ev_now () + 10.>, that is, an absolute time
		1851	not a delay) and then reset your system clock to January of the previous
		1852	year, then it will take a year or more to trigger the event (unlike an
		1853	C<ev_timer>, which would still trigger roughly 10 seconds after starting
		1854	it, as it uses a relative timeout).
		1855
1421	C<ev_periodic>s can also be used to implement vastly more complex timers,	1856	C<ev_periodic> watchers can also be used to implement vastly more complex
1422	such as triggering an event on each "midnight, local time", or other	1857	timers, such as triggering an event on each "midnight, local time", or
1423	complicated rules.	1858	other complicated rules. This cannot be done with C<ev_timer> watchers, as
		1859	those cannot react to time jumps.
1424		1860
1425	As with timers, the callback is guaranteed to be invoked only when the	1861	As with timers, the callback is guaranteed to be invoked only when the
1426	time (C<at>) has passed, but if multiple periodic timers become ready	1862	point in time where it is supposed to trigger has passed. If multiple
1427	during the same loop iteration, then order of execution is undefined.	1863	timers become ready during the same loop iteration then the ones with
		1864	earlier time-out values are invoked before ones with later time-out values
		1865	(but this is no longer true when a callback calls C<ev_loop> recursively).
1428		1866
1429	=head3 Watcher-Specific Functions and Data Members	1867	=head3 Watcher-Specific Functions and Data Members
1430		1868
1431	=over 4	1869	=over 4
1432		1870
1433	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)	1871	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp offset, ev_tstamp interval, reschedule_cb)
1434		1872
1435	=item ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)	1873	=item ev_periodic_set (ev_periodic *, ev_tstamp offset, ev_tstamp interval, reschedule_cb)
1436		1874
1437	Lots of arguments, lets sort it out... There are basically three modes of	1875	Lots of arguments, let's sort it out... There are basically three modes of
1438	operation, and we will explain them from simplest to most complex:	1876	operation, and we will explain them from simplest to most complex:
1439		1877
1440	=over 4	1878	=over 4
1441		1879
1442	=item * absolute timer (at = time, interval = reschedule_cb = 0)	1880	=item * absolute timer (offset = absolute time, interval = 0, reschedule_cb = 0)
1443		1881
1444	In this configuration the watcher triggers an event after the wall clock	1882	In this configuration the watcher triggers an event after the wall clock
1445	time C<at> has passed. It will not repeat and will not adjust when a time	1883	time C<offset> has passed. It will not repeat and will not adjust when a
1446	jump occurs, that is, if it is to be run at January 1st 2011 then it will	1884	time jump occurs, that is, if it is to be run at January 1st 2011 then it
1447	only run when the system clock reaches or surpasses this time.	1885	will be stopped and invoked when the system clock reaches or surpasses
		1886	this point in time.
1448		1887
1449	=item * repeating interval timer (at = offset, interval > 0, reschedule_cb = 0)	1888	=item * repeating interval timer (offset = offset within interval, interval > 0, reschedule_cb = 0)
1450		1889
1451	In this mode the watcher will always be scheduled to time out at the next	1890	In this mode the watcher will always be scheduled to time out at the next
1452	C<at + N * interval> time (for some integer N, which can also be negative)	1891	C<offset + N * interval> time (for some integer N, which can also be
1453	and then repeat, regardless of any time jumps.	1892	negative) and then repeat, regardless of any time jumps. The C<offset>
		1893	argument is merely an offset into the C<interval> periods.
1454		1894
1455	This can be used to create timers that do not drift with respect to the	1895	This can be used to create timers that do not drift with respect to the
1456	system clock, for example, here is a C<ev_periodic> that triggers each	1896	system clock, for example, here is an C<ev_periodic> that triggers each
1457	hour, on the hour:	1897	hour, on the hour (with respect to UTC):
1458		1898
1459	ev_periodic_set (&periodic, 0., 3600., 0);	1899	ev_periodic_set (&periodic, 0., 3600., 0);
1460		1900
1461	This doesn't mean there will always be 3600 seconds in between triggers,	1901	This doesn't mean there will always be 3600 seconds in between triggers,
1462	but only that the callback will be called when the system time shows a	1902	but only that the callback will be called when the system time shows a
1463	full hour (UTC), or more correctly, when the system time is evenly divisible	1903	full hour (UTC), or more correctly, when the system time is evenly divisible
1464	by 3600.	1904	by 3600.
1465		1905
1466	Another way to think about it (for the mathematically inclined) is that	1906	Another way to think about it (for the mathematically inclined) is that
1467	C<ev_periodic> will try to run the callback in this mode at the next possible	1907	C<ev_periodic> will try to run the callback in this mode at the next possible
1468	time where C<time = at (mod interval)>, regardless of any time jumps.	1908	time where C<time = offset (mod interval)>, regardless of any time jumps.
1469		1909
1470	For numerical stability it is preferable that the C<at> value is near	1910	For numerical stability it is preferable that the C<offset> value is near
1471	C<ev_now ()> (the current time), but there is no range requirement for	1911	C<ev_now ()> (the current time), but there is no range requirement for
1472	this value, and in fact is often specified as zero.	1912	this value, and in fact is often specified as zero.
1473		1913
1474	Note also that there is an upper limit to how often a timer can fire (CPU	1914	Note also that there is an upper limit to how often a timer can fire (CPU
1475	speed for example), so if C<interval> is very small then timing stability	1915	speed for example), so if C<interval> is very small then timing stability
1476	will of course deteriorate. Libev itself tries to be exact to be about one	1916	will of course deteriorate. Libev itself tries to be exact to be about one
1477	millisecond (if the OS supports it and the machine is fast enough).	1917	millisecond (if the OS supports it and the machine is fast enough).
1478		1918
1479	=item * manual reschedule mode (at and interval ignored, reschedule_cb = callback)	1919	=item * manual reschedule mode (offset ignored, interval ignored, reschedule_cb = callback)
1480		1920
1481	In this mode the values for C<interval> and C<at> are both being	1921	In this mode the values for C<interval> and C<offset> are both being
1482	ignored. Instead, each time the periodic watcher gets scheduled, the	1922	ignored. Instead, each time the periodic watcher gets scheduled, the
1483	reschedule callback will be called with the watcher as first, and the	1923	reschedule callback will be called with the watcher as first, and the
1484	current time as second argument.	1924	current time as second argument.
1485		1925
1486	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher,	1926	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher, ever,
1487	ever, or make ANY event loop modifications whatsoever>.	1927	or make ANY other event loop modifications whatsoever, unless explicitly
		1928	allowed by documentation here>.
1488		1929
1489	If you need to stop it, return C<now + 1e30> (or so, fudge fudge) and stop	1930	If you need to stop it, return C<now + 1e30> (or so, fudge fudge) and stop
1490	it afterwards (e.g. by starting an C<ev_prepare> watcher, which is the	1931	it afterwards (e.g. by starting an C<ev_prepare> watcher, which is the
1491	only event loop modification you are allowed to do).	1932	only event loop modification you are allowed to do).
1492		1933
1493	The callback prototype is C<ev_tstamp (*reschedule_cb)(struct ev_periodic	1934	The callback prototype is C<ev_tstamp (*reschedule_cb)(ev_periodic
1494	*w, ev_tstamp now)>, e.g.:	1935	*w, ev_tstamp now)>, e.g.:
1495		1936
		1937	static ev_tstamp
1496	static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now)	1938	my_rescheduler (ev_periodic *w, ev_tstamp now)
1497	{	1939	{
1498	return now + 60.;	1940	return now + 60.;
1499	}	1941	}
1500		1942
1501	It must return the next time to trigger, based on the passed time value	1943	It must return the next time to trigger, based on the passed time value
…		…
1521	a different time than the last time it was called (e.g. in a crond like	1963	a different time than the last time it was called (e.g. in a crond like
1522	program when the crontabs have changed).	1964	program when the crontabs have changed).
1523		1965
1524	=item ev_tstamp ev_periodic_at (ev_periodic *)	1966	=item ev_tstamp ev_periodic_at (ev_periodic *)
1525		1967
1526	When active, returns the absolute time that the watcher is supposed to	1968	When active, returns the absolute time that the watcher is supposed
1527	trigger next.	1969	to trigger next. This is not the same as the C<offset> argument to
		1970	C<ev_periodic_set>, but indeed works even in interval and manual
		1971	rescheduling modes.
1528		1972
1529	=item ev_tstamp offset [read-write]	1973	=item ev_tstamp offset [read-write]
1530		1974
1531	When repeating, this contains the offset value, otherwise this is the	1975	When repeating, this contains the offset value, otherwise this is the
1532	absolute point in time (the C<at> value passed to C<ev_periodic_set>).	1976	absolute point in time (the C<offset> value passed to C<ev_periodic_set>,
		1977	although libev might modify this value for better numerical stability).
1533		1978
1534	Can be modified any time, but changes only take effect when the periodic	1979	Can be modified any time, but changes only take effect when the periodic
1535	timer fires or C<ev_periodic_again> is being called.	1980	timer fires or C<ev_periodic_again> is being called.
1536		1981
1537	=item ev_tstamp interval [read-write]	1982	=item ev_tstamp interval [read-write]
1538		1983
1539	The current interval value. Can be modified any time, but changes only	1984	The current interval value. Can be modified any time, but changes only
1540	take effect when the periodic timer fires or C<ev_periodic_again> is being	1985	take effect when the periodic timer fires or C<ev_periodic_again> is being
1541	called.	1986	called.
1542		1987
1543	=item ev_tstamp (*reschedule_cb)(struct ev_periodic *w, ev_tstamp now) [read-write]	1988	=item ev_tstamp (*reschedule_cb)(ev_periodic *w, ev_tstamp now) [read-write]
1544		1989
1545	The current reschedule callback, or C<0>, if this functionality is	1990	The current reschedule callback, or C<0>, if this functionality is
1546	switched off. Can be changed any time, but changes only take effect when	1991	switched off. Can be changed any time, but changes only take effect when
1547	the periodic timer fires or C<ev_periodic_again> is being called.	1992	the periodic timer fires or C<ev_periodic_again> is being called.
1548		1993
…		…
1553	Example: Call a callback every hour, or, more precisely, whenever the	1998	Example: Call a callback every hour, or, more precisely, whenever the
1554	system time is divisible by 3600. The callback invocation times have	1999	system time is divisible by 3600. The callback invocation times have
1555	potentially a lot of jitter, but good long-term stability.	2000	potentially a lot of jitter, but good long-term stability.
1556		2001
1557	static void	2002	static void
1558	clock_cb (struct ev_loop *loop, struct ev_io *w, int revents)	2003	clock_cb (struct ev_loop *loop, ev_io *w, int revents)
1559	{	2004	{
1560	... its now a full hour (UTC, or TAI or whatever your clock follows)	2005	... its now a full hour (UTC, or TAI or whatever your clock follows)
1561	}	2006	}
1562		2007
1563	struct ev_periodic hourly_tick;	2008	ev_periodic hourly_tick;
1564	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);	2009	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);
1565	ev_periodic_start (loop, &hourly_tick);	2010	ev_periodic_start (loop, &hourly_tick);
1566		2011
1567	Example: The same as above, but use a reschedule callback to do it:	2012	Example: The same as above, but use a reschedule callback to do it:
1568		2013
1569	#include <math.h>	2014	#include <math.h>
1570		2015
1571	static ev_tstamp	2016	static ev_tstamp
1572	my_scheduler_cb (struct ev_periodic *w, ev_tstamp now)	2017	my_scheduler_cb (ev_periodic *w, ev_tstamp now)
1573	{	2018	{
1574	return now + (3600. - fmod (now, 3600.));	2019	return now + (3600. - fmod (now, 3600.));
1575	}	2020	}
1576		2021
1577	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);	2022	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);
1578		2023
1579	Example: Call a callback every hour, starting now:	2024	Example: Call a callback every hour, starting now:
1580		2025
1581	struct ev_periodic hourly_tick;	2026	ev_periodic hourly_tick;
1582	ev_periodic_init (&hourly_tick, clock_cb,	2027	ev_periodic_init (&hourly_tick, clock_cb,
1583	fmod (ev_now (loop), 3600.), 3600., 0);	2028	fmod (ev_now (loop), 3600.), 3600., 0);
1584	ev_periodic_start (loop, &hourly_tick);	2029	ev_periodic_start (loop, &hourly_tick);
1585		2030
1586		2031
…		…
1628	=head3 Examples	2073	=head3 Examples
1629		2074
1630	Example: Try to exit cleanly on SIGINT.	2075	Example: Try to exit cleanly on SIGINT.
1631		2076
1632	static void	2077	static void
1633	sigint_cb (struct ev_loop *loop, struct ev_signal *w, int revents)	2078	sigint_cb (struct ev_loop *loop, ev_signal *w, int revents)
1634	{	2079	{
1635	ev_unloop (loop, EVUNLOOP_ALL);	2080	ev_unloop (loop, EVUNLOOP_ALL);
1636	}	2081	}
1637		2082
1638	struct ev_signal signal_watcher;	2083	ev_signal signal_watcher;
1639	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);	2084	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);
1640	ev_signal_start (loop, &signal_watcher);	2085	ev_signal_start (loop, &signal_watcher);
1641		2086
1642		2087
1643	=head2 C<ev_child> - watch out for process status changes	2088	=head2 C<ev_child> - watch out for process status changes
…		…
1646	some child status changes (most typically when a child of yours dies or	2091	some child status changes (most typically when a child of yours dies or
1647	exits). It is permissible to install a child watcher I<after> the child	2092	exits). It is permissible to install a child watcher I<after> the child
1648	has been forked (which implies it might have already exited), as long	2093	has been forked (which implies it might have already exited), as long
1649	as the event loop isn't entered (or is continued from a watcher), i.e.,	2094	as the event loop isn't entered (or is continued from a watcher), i.e.,
1650	forking and then immediately registering a watcher for the child is fine,	2095	forking and then immediately registering a watcher for the child is fine,
1651	but forking and registering a watcher a few event loop iterations later is	2096	but forking and registering a watcher a few event loop iterations later or
1652	not.	2097	in the next callback invocation is not.
1653		2098
1654	Only the default event loop is capable of handling signals, and therefore	2099	Only the default event loop is capable of handling signals, and therefore
1655	you can only register child watchers in the default event loop.	2100	you can only register child watchers in the default event loop.
		2101
		2102	Due to some design glitches inside libev, child watchers will always be
		2103	handled at maximum priority (their priority is set to C<EV_MAXPRI> by
		2104	libev)
1656		2105
1657	=head3 Process Interaction	2106	=head3 Process Interaction
1658		2107
1659	Libev grabs C<SIGCHLD> as soon as the default event loop is	2108	Libev grabs C<SIGCHLD> as soon as the default event loop is
1660	initialised. This is necessary to guarantee proper behaviour even if	2109	initialised. This is necessary to guarantee proper behaviour even if
…		…
1718	its completion.	2167	its completion.
1719		2168
1720	ev_child cw;	2169	ev_child cw;
1721		2170
1722	static void	2171	static void
1723	child_cb (EV_P_ struct ev_child *w, int revents)	2172	child_cb (EV_P_ ev_child *w, int revents)
1724	{	2173	{
1725	ev_child_stop (EV_A_ w);	2174	ev_child_stop (EV_A_ w);
1726	printf ("process %d exited with status %x\n", w->rpid, w->rstatus);	2175	printf ("process %d exited with status %x\n", w->rpid, w->rstatus);
1727	}	2176	}
1728		2177
…		…
1743		2192
1744		2193
1745	=head2 C<ev_stat> - did the file attributes just change?	2194	=head2 C<ev_stat> - did the file attributes just change?
1746		2195
1747	This watches a file system path for attribute changes. That is, it calls	2196	This watches a file system path for attribute changes. That is, it calls
1748	C<stat> regularly (or when the OS says it changed) and sees if it changed	2197	C<stat> on that path in regular intervals (or when the OS says it changed)
1749	compared to the last time, invoking the callback if it did.	2198	and sees if it changed compared to the last time, invoking the callback if
		2199	it did.
1750		2200
1751	The path does not need to exist: changing from "path exists" to "path does	2201	The path does not need to exist: changing from "path exists" to "path does
1752	not exist" is a status change like any other. The condition "path does	2202	not exist" is a status change like any other. The condition "path does not
1753	not exist" is signified by the C<st_nlink> field being zero (which is	2203	exist" (or more correctly "path cannot be stat'ed") is signified by the
1754	otherwise always forced to be at least one) and all the other fields of	2204	C<st_nlink> field being zero (which is otherwise always forced to be at
1755	the stat buffer having unspecified contents.	2205	least one) and all the other fields of the stat buffer having unspecified
		2206	contents.
1756		2207
1757	The path I<should> be absolute and I<must not> end in a slash. If it is	2208	The path I<must not> end in a slash or contain special components such as
		2209	C<.> or C<..>. The path I<should> be absolute: If it is relative and
1758	relative and your working directory changes, the behaviour is undefined.	2210	your working directory changes, then the behaviour is undefined.
1759		2211
1760	Since there is no standard kernel interface to do this, the portable	2212	Since there is no portable change notification interface available, the
1761	implementation simply calls C<stat (2)> regularly on the path to see if	2213	portable implementation simply calls C<stat(2)> regularly on the path
1762	it changed somehow. You can specify a recommended polling interval for	2214	to see if it changed somehow. You can specify a recommended polling
1763	this case. If you specify a polling interval of C<0> (highly recommended!)	2215	interval for this case. If you specify a polling interval of C<0> (highly
1764	then a I<suitable, unspecified default> value will be used (which	2216	recommended!) then a I<suitable, unspecified default> value will be used
1765	you can expect to be around five seconds, although this might change	2217	(which you can expect to be around five seconds, although this might
1766	dynamically). Libev will also impose a minimum interval which is currently	2218	change dynamically). Libev will also impose a minimum interval which is
1767	around C<0.1>, but thats usually overkill.	2219	currently around C<0.1>, but that's usually overkill.
1768		2220
1769	This watcher type is not meant for massive numbers of stat watchers,	2221	This watcher type is not meant for massive numbers of stat watchers,
1770	as even with OS-supported change notifications, this can be	2222	as even with OS-supported change notifications, this can be
1771	resource-intensive.	2223	resource-intensive.
1772		2224
1773	At the time of this writing, the only OS-specific interface implemented	2225	At the time of this writing, the only OS-specific interface implemented
1774	is the Linux inotify interface (implementing kqueue support is left as	2226	is the Linux inotify interface (implementing kqueue support is left as an
1775	an exercise for the reader. Note, however, that the author sees no way	2227	exercise for the reader. Note, however, that the author sees no way of
1776	of implementing C<ev_stat> semantics with kqueue).	2228	implementing C<ev_stat> semantics with kqueue, except as a hint).
1777		2229
1778	=head3 ABI Issues (Largefile Support)	2230	=head3 ABI Issues (Largefile Support)
1779		2231
1780	Libev by default (unless the user overrides this) uses the default	2232	Libev by default (unless the user overrides this) uses the default
1781	compilation environment, which means that on systems with large file	2233	compilation environment, which means that on systems with large file
1782	support disabled by default, you get the 32 bit version of the stat	2234	support disabled by default, you get the 32 bit version of the stat
1783	structure. When using the library from programs that change the ABI to	2235	structure. When using the library from programs that change the ABI to
1784	use 64 bit file offsets the programs will fail. In that case you have to	2236	use 64 bit file offsets the programs will fail. In that case you have to
1785	compile libev with the same flags to get binary compatibility. This is	2237	compile libev with the same flags to get binary compatibility. This is
1786	obviously the case with any flags that change the ABI, but the problem is	2238	obviously the case with any flags that change the ABI, but the problem is
1787	most noticeably disabled with ev_stat and large file support.	2239	most noticeably displayed with ev_stat and large file support.
1788		2240
1789	The solution for this is to lobby your distribution maker to make large	2241	The solution for this is to lobby your distribution maker to make large
1790	file interfaces available by default (as e.g. FreeBSD does) and not	2242	file interfaces available by default (as e.g. FreeBSD does) and not
1791	optional. Libev cannot simply switch on large file support because it has	2243	optional. Libev cannot simply switch on large file support because it has
1792	to exchange stat structures with application programs compiled using the	2244	to exchange stat structures with application programs compiled using the
1793	default compilation environment.	2245	default compilation environment.
1794		2246
1795	=head3 Inotify and Kqueue	2247	=head3 Inotify and Kqueue
1796		2248
1797	When C<inotify (7)> support has been compiled into libev (generally only	2249	When C<inotify (7)> support has been compiled into libev and present at
1798	available with Linux) and present at runtime, it will be used to speed up	2250	runtime, it will be used to speed up change detection where possible. The
1799	change detection where possible. The inotify descriptor will be created lazily	2251	inotify descriptor will be created lazily when the first C<ev_stat>
1800	when the first C<ev_stat> watcher is being started.	2252	watcher is being started.
1801		2253
1802	Inotify presence does not change the semantics of C<ev_stat> watchers	2254	Inotify presence does not change the semantics of C<ev_stat> watchers
1803	except that changes might be detected earlier, and in some cases, to avoid	2255	except that changes might be detected earlier, and in some cases, to avoid
1804	making regular C<stat> calls. Even in the presence of inotify support	2256	making regular C<stat> calls. Even in the presence of inotify support
1805	there are many cases where libev has to resort to regular C<stat> polling,	2257	there are many cases where libev has to resort to regular C<stat> polling,
1806	but as long as the path exists, libev usually gets away without polling.	2258	but as long as kernel 2.6.25 or newer is used (2.6.24 and older have too
		2259	many bugs), the path exists (i.e. stat succeeds), and the path resides on
		2260	a local filesystem (libev currently assumes only ext2/3, jfs, reiserfs and
		2261	xfs are fully working) libev usually gets away without polling.
1807		2262
1808	There is no support for kqueue, as apparently it cannot be used to	2263	There is no support for kqueue, as apparently it cannot be used to
1809	implement this functionality, due to the requirement of having a file	2264	implement this functionality, due to the requirement of having a file
1810	descriptor open on the object at all times, and detecting renames, unlinks	2265	descriptor open on the object at all times, and detecting renames, unlinks
1811	etc. is difficult.	2266	etc. is difficult.
1812		2267
		2268	=head3 C<stat ()> is a synchronous operation
		2269
		2270	Libev doesn't normally do any kind of I/O itself, and so is not blocking
		2271	the process. The exception are C<ev_stat> watchers - those call C<stat
		2272	()>, which is a synchronous operation.
		2273
		2274	For local paths, this usually doesn't matter: unless the system is very
		2275	busy or the intervals between stat's are large, a stat call will be fast,
		2276	as the path data is usually in memory already (except when starting the
		2277	watcher).
		2278
		2279	For networked file systems, calling C<stat ()> can block an indefinite
		2280	time due to network issues, and even under good conditions, a stat call
		2281	often takes multiple milliseconds.
		2282
		2283	Therefore, it is best to avoid using C<ev_stat> watchers on networked
		2284	paths, although this is fully supported by libev.
		2285
1813	=head3 The special problem of stat time resolution	2286	=head3 The special problem of stat time resolution
1814		2287
1815	The C<stat ()> system call only supports full-second resolution portably, and	2288	The C<stat ()> system call only supports full-second resolution portably,
1816	even on systems where the resolution is higher, most file systems still	2289	and even on systems where the resolution is higher, most file systems
1817	only support whole seconds.	2290	still only support whole seconds.
1818		2291
1819	That means that, if the time is the only thing that changes, you can	2292	That means that, if the time is the only thing that changes, you can
1820	easily miss updates: on the first update, C<ev_stat> detects a change and	2293	easily miss updates: on the first update, C<ev_stat> detects a change and
1821	calls your callback, which does something. When there is another update	2294	calls your callback, which does something. When there is another update
1822	within the same second, C<ev_stat> will be unable to detect unless the	2295	within the same second, C<ev_stat> will be unable to detect unless the
…		…
1965		2438
1966	=head3 Watcher-Specific Functions and Data Members	2439	=head3 Watcher-Specific Functions and Data Members
1967		2440
1968	=over 4	2441	=over 4
1969		2442
1970	=item ev_idle_init (ev_signal *, callback)	2443	=item ev_idle_init (ev_idle *, callback)
1971		2444
1972	Initialises and configures the idle watcher - it has no parameters of any	2445	Initialises and configures the idle watcher - it has no parameters of any
1973	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,	2446	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,
1974	believe me.	2447	believe me.
1975		2448
…		…
1979		2452
1980	Example: Dynamically allocate an C<ev_idle> watcher, start it, and in the	2453	Example: Dynamically allocate an C<ev_idle> watcher, start it, and in the
1981	callback, free it. Also, use no error checking, as usual.	2454	callback, free it. Also, use no error checking, as usual.
1982		2455
1983	static void	2456	static void
1984	idle_cb (struct ev_loop *loop, struct ev_idle *w, int revents)	2457	idle_cb (struct ev_loop *loop, ev_idle *w, int revents)
1985	{	2458	{
1986	free (w);	2459	free (w);
1987	// now do something you wanted to do when the program has	2460	// now do something you wanted to do when the program has
1988	// no longer anything immediate to do.	2461	// no longer anything immediate to do.
1989	}	2462	}
1990		2463
1991	struct ev_idle *idle_watcher = malloc (sizeof (struct ev_idle));	2464	ev_idle *idle_watcher = malloc (sizeof (ev_idle));
1992	ev_idle_init (idle_watcher, idle_cb);	2465	ev_idle_init (idle_watcher, idle_cb);
1993	ev_idle_start (loop, idle_cb);	2466	ev_idle_start (loop, idle_watcher);
1994		2467
1995		2468
1996	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!	2469	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!
1997		2470
1998	Prepare and check watchers are usually (but not always) used in pairs:	2471	Prepare and check watchers are usually (but not always) used in pairs:
…		…
2077		2550
2078	static ev_io iow [nfd];	2551	static ev_io iow [nfd];
2079	static ev_timer tw;	2552	static ev_timer tw;
2080		2553
2081	static void	2554	static void
2082	io_cb (ev_loop *loop, ev_io *w, int revents)	2555	io_cb (struct ev_loop *loop, ev_io *w, int revents)
2083	{	2556	{
2084	}	2557	}
2085		2558
2086	// create io watchers for each fd and a timer before blocking	2559	// create io watchers for each fd and a timer before blocking
2087	static void	2560	static void
2088	adns_prepare_cb (ev_loop *loop, ev_prepare *w, int revents)	2561	adns_prepare_cb (struct ev_loop *loop, ev_prepare *w, int revents)
2089	{	2562	{
2090	int timeout = 3600000;	2563	int timeout = 3600000;
2091	struct pollfd fds [nfd];	2564	struct pollfd fds [nfd];
2092	// actual code will need to loop here and realloc etc.	2565	// actual code will need to loop here and realloc etc.
2093	adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));	2566	adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));
2094		2567
2095	/* the callback is illegal, but won't be called as we stop during check */	2568	/* the callback is illegal, but won't be called as we stop during check */
2096	ev_timer_init (&tw, 0, timeout * 1e-3);	2569	ev_timer_init (&tw, 0, timeout * 1e-3, 0.);
2097	ev_timer_start (loop, &tw);	2570	ev_timer_start (loop, &tw);
2098		2571
2099	// create one ev_io per pollfd	2572	// create one ev_io per pollfd
2100	for (int i = 0; i < nfd; ++i)	2573	for (int i = 0; i < nfd; ++i)
2101	{	2574	{
…		…
2108	}	2581	}
2109	}	2582	}
2110		2583
2111	// stop all watchers after blocking	2584	// stop all watchers after blocking
2112	static void	2585	static void
2113	adns_check_cb (ev_loop *loop, ev_check *w, int revents)	2586	adns_check_cb (struct ev_loop *loop, ev_check *w, int revents)
2114	{	2587	{
2115	ev_timer_stop (loop, &tw);	2588	ev_timer_stop (loop, &tw);
2116		2589
2117	for (int i = 0; i < nfd; ++i)	2590	for (int i = 0; i < nfd; ++i)
2118	{	2591	{
…		…
2214	some fds have to be watched and handled very quickly (with low latency),	2687	some fds have to be watched and handled very quickly (with low latency),
2215	and even priorities and idle watchers might have too much overhead. In	2688	and even priorities and idle watchers might have too much overhead. In
2216	this case you would put all the high priority stuff in one loop and all	2689	this case you would put all the high priority stuff in one loop and all
2217	the rest in a second one, and embed the second one in the first.	2690	the rest in a second one, and embed the second one in the first.
2218		2691
2219	As long as the watcher is active, the callback will be invoked every time	2692	As long as the watcher is active, the callback will be invoked every
2220	there might be events pending in the embedded loop. The callback must then	2693	time there might be events pending in the embedded loop. The callback
2221	call C<ev_embed_sweep (mainloop, watcher)> to make a single sweep and invoke	2694	must then call C<ev_embed_sweep (mainloop, watcher)> to make a single
2222	their callbacks (you could also start an idle watcher to give the embedded	2695	sweep and invoke their callbacks (the callback doesn't need to invoke the
2223	loop strictly lower priority for example). You can also set the callback	2696	C<ev_embed_sweep> function directly, it could also start an idle watcher
2224	to C<0>, in which case the embed watcher will automatically execute the	2697	to give the embedded loop strictly lower priority for example).
2225	embedded loop sweep.
2226		2698
2227	As long as the watcher is started it will automatically handle events. The	2699	You can also set the callback to C<0>, in which case the embed watcher
2228	callback will be invoked whenever some events have been handled. You can	2700	will automatically execute the embedded loop sweep whenever necessary.
2229	set the callback to C<0> to avoid having to specify one if you are not
2230	interested in that.
2231		2701
2232	Also, there have not currently been made special provisions for forking:	2702	Fork detection will be handled transparently while the C<ev_embed> watcher
2233	when you fork, you not only have to call C<ev_loop_fork> on both loops,	2703	is active, i.e., the embedded loop will automatically be forked when the
2234	but you will also have to stop and restart any C<ev_embed> watchers	2704	embedding loop forks. In other cases, the user is responsible for calling
2235	yourself - but you can use a fork watcher to handle this automatically,	2705	C<ev_loop_fork> on the embedded loop.
2236	and future versions of libev might do just that.
2237		2706
2238	Unfortunately, not all backends are embeddable: only the ones returned by	2707	Unfortunately, not all backends are embeddable: only the ones returned by
2239	C<ev_embeddable_backends> are, which, unfortunately, does not include any	2708	C<ev_embeddable_backends> are, which, unfortunately, does not include any
2240	portable one.	2709	portable one.
2241		2710
…		…
2286	C<loop_lo> (which is C<loop_hi> in the case no embeddable loop can be	2755	C<loop_lo> (which is C<loop_hi> in the case no embeddable loop can be
2287	used).	2756	used).
2288		2757
2289	struct ev_loop *loop_hi = ev_default_init (0);	2758	struct ev_loop *loop_hi = ev_default_init (0);
2290	struct ev_loop *loop_lo = 0;	2759	struct ev_loop *loop_lo = 0;
2291	struct ev_embed embed;	2760	ev_embed embed;
2292		2761
2293	// see if there is a chance of getting one that works	2762	// see if there is a chance of getting one that works
2294	// (remember that a flags value of 0 means autodetection)	2763	// (remember that a flags value of 0 means autodetection)
2295	loop_lo = ev_embeddable_backends () & ev_recommended_backends ()	2764	loop_lo = ev_embeddable_backends () & ev_recommended_backends ()
2296	? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())	2765	? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())
…		…
2310	kqueue implementation). Store the kqueue/socket-only event loop in	2779	kqueue implementation). Store the kqueue/socket-only event loop in
2311	C<loop_socket>. (One might optionally use C<EVFLAG_NOENV>, too).	2780	C<loop_socket>. (One might optionally use C<EVFLAG_NOENV>, too).
2312		2781
2313	struct ev_loop *loop = ev_default_init (0);	2782	struct ev_loop *loop = ev_default_init (0);
2314	struct ev_loop *loop_socket = 0;	2783	struct ev_loop *loop_socket = 0;
2315	struct ev_embed embed;	2784	ev_embed embed;
2316		2785
2317	if (ev_supported_backends () & ~ev_recommended_backends () & EVBACKEND_KQUEUE)	2786	if (ev_supported_backends () & ~ev_recommended_backends () & EVBACKEND_KQUEUE)
2318	if ((loop_socket = ev_loop_new (EVBACKEND_KQUEUE))	2787	if ((loop_socket = ev_loop_new (EVBACKEND_KQUEUE))
2319	{	2788	{
2320	ev_embed_init (&embed, 0, loop_socket);	2789	ev_embed_init (&embed, 0, loop_socket);
…		…
2335	event loop blocks next and before C<ev_check> watchers are being called,	2804	event loop blocks next and before C<ev_check> watchers are being called,
2336	and only in the child after the fork. If whoever good citizen calling	2805	and only in the child after the fork. If whoever good citizen calling
2337	C<ev_default_fork> cheats and calls it in the wrong process, the fork	2806	C<ev_default_fork> cheats and calls it in the wrong process, the fork
2338	handlers will be invoked, too, of course.	2807	handlers will be invoked, too, of course.
2339		2808
		2809	=head3 The special problem of life after fork - how is it possible?
		2810
		2811	Most uses of C<fork()> consist of forking, then some simple calls to ste
		2812	up/change the process environment, followed by a call to C<exec()>. This
		2813	sequence should be handled by libev without any problems.
		2814
		2815	This changes when the application actually wants to do event handling
		2816	in the child, or both parent in child, in effect "continuing" after the
		2817	fork.
		2818
		2819	The default mode of operation (for libev, with application help to detect
		2820	forks) is to duplicate all the state in the child, as would be expected
		2821	when I<either> the parent I<or> the child process continues.
		2822
		2823	When both processes want to continue using libev, then this is usually the
		2824	wrong result. In that case, usually one process (typically the parent) is
		2825	supposed to continue with all watchers in place as before, while the other
		2826	process typically wants to start fresh, i.e. without any active watchers.
		2827
		2828	The cleanest and most efficient way to achieve that with libev is to
		2829	simply create a new event loop, which of course will be "empty", and
		2830	use that for new watchers. This has the advantage of not touching more
		2831	memory than necessary, and thus avoiding the copy-on-write, and the
		2832	disadvantage of having to use multiple event loops (which do not support
		2833	signal watchers).
		2834
		2835	When this is not possible, or you want to use the default loop for
		2836	other reasons, then in the process that wants to start "fresh", call
		2837	C<ev_default_destroy ()> followed by C<ev_default_loop (...)>. Destroying
		2838	the default loop will "orphan" (not stop) all registered watchers, so you
		2839	have to be careful not to execute code that modifies those watchers. Note
		2840	also that in that case, you have to re-register any signal watchers.
		2841
2340	=head3 Watcher-Specific Functions and Data Members	2842	=head3 Watcher-Specific Functions and Data Members
2341		2843
2342	=over 4	2844	=over 4
2343		2845
2344	=item ev_fork_init (ev_signal *, callback)	2846	=item ev_fork_init (ev_signal *, callback)
…		…
2384	=over 4	2886	=over 4
2385		2887
2386	=item queueing from a signal handler context	2888	=item queueing from a signal handler context
2387		2889
2388	To implement race-free queueing, you simply add to the queue in the signal	2890	To implement race-free queueing, you simply add to the queue in the signal
2389	handler but you block the signal handler in the watcher callback. Here is an example that does that for	2891	handler but you block the signal handler in the watcher callback. Here is
2390	some fictitious SIGUSR1 handler:	2892	an example that does that for some fictitious SIGUSR1 handler:
2391		2893
2392	static ev_async mysig;	2894	static ev_async mysig;
2393		2895
2394	static void	2896	static void
2395	sigusr1_handler (void)	2897	sigusr1_handler (void)
…		…
2461	=over 4	2963	=over 4
2462		2964
2463	=item ev_async_init (ev_async *, callback)	2965	=item ev_async_init (ev_async *, callback)
2464		2966
2465	Initialises and configures the async watcher - it has no parameters of any	2967	Initialises and configures the async watcher - it has no parameters of any
2466	kind. There is a C<ev_asynd_set> macro, but using it is utterly pointless,	2968	kind. There is a C<ev_async_set> macro, but using it is utterly pointless,
2467	trust me.	2969	trust me.
2468		2970
2469	=item ev_async_send (loop, ev_async *)	2971	=item ev_async_send (loop, ev_async *)
2470		2972
2471	Sends/signals/activates the given C<ev_async> watcher, that is, feeds	2973	Sends/signals/activates the given C<ev_async> watcher, that is, feeds
2472	an C<EV_ASYNC> event on the watcher into the event loop. Unlike	2974	an C<EV_ASYNC> event on the watcher into the event loop. Unlike
2473	C<ev_feed_event>, this call is safe to do from other threads, signal or	2975	C<ev_feed_event>, this call is safe to do from other threads, signal or
2474	similar contexts (see the discussion of C<EV_ATOMIC_T> in the embedding	2976	similar contexts (see the discussion of C<EV_ATOMIC_T> in the embedding
2475	section below on what exactly this means).	2977	section below on what exactly this means).
2476		2978
		2979	Note that, as with other watchers in libev, multiple events might get
		2980	compressed into a single callback invocation (another way to look at this
		2981	is that C<ev_async> watchers are level-triggered, set on C<ev_async_send>,
		2982	reset when the event loop detects that).
		2983
2477	This call incurs the overhead of a system call only once per loop iteration,	2984	This call incurs the overhead of a system call only once per event loop
2478	so while the overhead might be noticeable, it doesn't apply to repeated	2985	iteration, so while the overhead might be noticeable, it doesn't apply to
2479	calls to C<ev_async_send>.	2986	repeated calls to C<ev_async_send> for the same event loop.
2480		2987
2481	=item bool = ev_async_pending (ev_async *)	2988	=item bool = ev_async_pending (ev_async *)
2482		2989
2483	Returns a non-zero value when C<ev_async_send> has been called on the	2990	Returns a non-zero value when C<ev_async_send> has been called on the
2484	watcher but the event has not yet been processed (or even noted) by the	2991	watcher but the event has not yet been processed (or even noted) by the
…		…
2487	C<ev_async_send> sets a flag in the watcher and wakes up the loop. When	2994	C<ev_async_send> sets a flag in the watcher and wakes up the loop. When
2488	the loop iterates next and checks for the watcher to have become active,	2995	the loop iterates next and checks for the watcher to have become active,
2489	it will reset the flag again. C<ev_async_pending> can be used to very	2996	it will reset the flag again. C<ev_async_pending> can be used to very
2490	quickly check whether invoking the loop might be a good idea.	2997	quickly check whether invoking the loop might be a good idea.
2491		2998
2492	Not that this does I<not> check whether the watcher itself is pending, only	2999	Not that this does I<not> check whether the watcher itself is pending,
2493	whether it has been requested to make this watcher pending.	3000	only whether it has been requested to make this watcher pending: there
		3001	is a time window between the event loop checking and resetting the async
		3002	notification, and the callback being invoked.
2494		3003
2495	=back	3004	=back
2496		3005
2497		3006
2498	=head1 OTHER FUNCTIONS	3007	=head1 OTHER FUNCTIONS
…		…
2502	=over 4	3011	=over 4
2503		3012
2504	=item ev_once (loop, int fd, int events, ev_tstamp timeout, callback)	3013	=item ev_once (loop, int fd, int events, ev_tstamp timeout, callback)
2505		3014
2506	This function combines a simple timer and an I/O watcher, calls your	3015	This function combines a simple timer and an I/O watcher, calls your
2507	callback on whichever event happens first and automatically stop both	3016	callback on whichever event happens first and automatically stops both
2508	watchers. This is useful if you want to wait for a single event on an fd	3017	watchers. This is useful if you want to wait for a single event on an fd
2509	or timeout without having to allocate/configure/start/stop/free one or	3018	or timeout without having to allocate/configure/start/stop/free one or
2510	more watchers yourself.	3019	more watchers yourself.
2511		3020
2512	If C<fd> is less than 0, then no I/O watcher will be started and events	3021	If C<fd> is less than 0, then no I/O watcher will be started and the
2513	is being ignored. Otherwise, an C<ev_io> watcher for the given C<fd> and	3022	C<events> argument is being ignored. Otherwise, an C<ev_io> watcher for
2514	C<events> set will be created and started.	3023	the given C<fd> and C<events> set will be created and started.
2515		3024
2516	If C<timeout> is less than 0, then no timeout watcher will be	3025	If C<timeout> is less than 0, then no timeout watcher will be
2517	started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and	3026	started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and
2518	repeat = 0) will be started. While C<0> is a valid timeout, it is of	3027	repeat = 0) will be started. C<0> is a valid timeout.
2519	dubious value.
2520		3028
2521	The callback has the type C<void (*cb)(int revents, void *arg)> and gets	3029	The callback has the type C<void (*cb)(int revents, void *arg)> and gets
2522	passed an C<revents> set like normal event callbacks (a combination of	3030	passed an C<revents> set like normal event callbacks (a combination of
2523	C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMEOUT>) and the C<arg>	3031	C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMEOUT>) and the C<arg>
2524	value passed to C<ev_once>:	3032	value passed to C<ev_once>. Note that it is possible to receive I<both>
		3033	a timeout and an io event at the same time - you probably should give io
		3034	events precedence.
		3035
		3036	Example: wait up to ten seconds for data to appear on STDIN_FILENO.
2525		3037
2526	static void stdin_ready (int revents, void *arg)	3038	static void stdin_ready (int revents, void *arg)
2527	{	3039	{
		3040	if (revents & EV_READ)
		3041	/* stdin might have data for us, joy! */;
2528	if (revents & EV_TIMEOUT)	3042	else if (revents & EV_TIMEOUT)
2529	/* doh, nothing entered */;	3043	/* doh, nothing entered */;
2530	else if (revents & EV_READ)
2531	/* stdin might have data for us, joy! */;
2532	}	3044	}
2533		3045
2534	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);	3046	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);
2535		3047
2536	=item ev_feed_event (ev_loop *, watcher *, int revents)	3048	=item ev_feed_event (struct ev_loop *, watcher *, int revents)
2537		3049
2538	Feeds the given event set into the event loop, as if the specified event	3050	Feeds the given event set into the event loop, as if the specified event
2539	had happened for the specified watcher (which must be a pointer to an	3051	had happened for the specified watcher (which must be a pointer to an
2540	initialised but not necessarily started event watcher).	3052	initialised but not necessarily started event watcher).
2541		3053
2542	=item ev_feed_fd_event (ev_loop *, int fd, int revents)	3054	=item ev_feed_fd_event (struct ev_loop *, int fd, int revents)
2543		3055
2544	Feed an event on the given fd, as if a file descriptor backend detected	3056	Feed an event on the given fd, as if a file descriptor backend detected
2545	the given events it.	3057	the given events it.
2546		3058
2547	=item ev_feed_signal_event (ev_loop *loop, int signum)	3059	=item ev_feed_signal_event (struct ev_loop *loop, int signum)
2548		3060
2549	Feed an event as if the given signal occurred (C<loop> must be the default	3061	Feed an event as if the given signal occurred (C<loop> must be the default
2550	loop!).	3062	loop!).
2551		3063
2552	=back	3064	=back
…		…
2673	}	3185	}
2674		3186
2675	myclass obj;	3187	myclass obj;
2676	ev::io iow;	3188	ev::io iow;
2677	iow.set <myclass, &myclass::io_cb> (&obj);	3189	iow.set <myclass, &myclass::io_cb> (&obj);
		3190
		3191	=item w->set (object *)
		3192
		3193	This is an B<experimental> feature that might go away in a future version.
		3194
		3195	This is a variation of a method callback - leaving out the method to call
		3196	will default the method to C<operator ()>, which makes it possible to use
		3197	functor objects without having to manually specify the C<operator ()> all
		3198	the time. Incidentally, you can then also leave out the template argument
		3199	list.
		3200
		3201	The C<operator ()> method prototype must be C<void operator ()(watcher &w,
		3202	int revents)>.
		3203
		3204	See the method-C<set> above for more details.
		3205
		3206	Example: use a functor object as callback.
		3207
		3208	struct myfunctor
		3209	{
		3210	void operator() (ev::io &w, int revents)
		3211	{
		3212	...
		3213	}
		3214	}
		3215
		3216	myfunctor f;
		3217
		3218	ev::io w;
		3219	w.set (&f);
2678		3220
2679	=item w->set<function> (void *data = 0)	3221	=item w->set<function> (void *data = 0)
2680		3222
2681	Also sets a callback, but uses a static method or plain function as	3223	Also sets a callback, but uses a static method or plain function as
2682	callback. The optional C<data> argument will be stored in the watcher's	3224	callback. The optional C<data> argument will be stored in the watcher's
…		…
2769	L<http://software.schmorp.de/pkg/EV>.	3311	L<http://software.schmorp.de/pkg/EV>.
2770		3312
2771	=item Python	3313	=item Python
2772		3314
2773	Python bindings can be found at L<http://code.google.com/p/pyev/>. It	3315	Python bindings can be found at L<http://code.google.com/p/pyev/>. It
2774	seems to be quite complete and well-documented. Note, however, that the	3316	seems to be quite complete and well-documented.
2775	patch they require for libev is outright dangerous as it breaks the ABI
2776	for everybody else, and therefore, should never be applied in an installed
2777	libev (if python requires an incompatible ABI then it needs to embed
2778	libev).
2779		3317
2780	=item Ruby	3318	=item Ruby
2781		3319
2782	Tony Arcieri has written a ruby extension that offers access to a subset	3320	Tony Arcieri has written a ruby extension that offers access to a subset
2783	of the libev API and adds file handle abstractions, asynchronous DNS and	3321	of the libev API and adds file handle abstractions, asynchronous DNS and
2784	more on top of it. It can be found via gem servers. Its homepage is at	3322	more on top of it. It can be found via gem servers. Its homepage is at
2785	L<http://rev.rubyforge.org/>.	3323	L<http://rev.rubyforge.org/>.
2786		3324
		3325	Roger Pack reports that using the link order C<-lws2_32 -lmsvcrt-ruby-190>
		3326	makes rev work even on mingw.
		3327
		3328	=item Haskell
		3329
		3330	A haskell binding to libev is available at
		3331	L<http://hackage.haskell.org/cgi-bin/hackage-scripts/package/hlibev>.
		3332
2787	=item D	3333	=item D
2788		3334
2789	Leandro Lucarella has written a D language binding (F<ev.d>) for libev, to	3335	Leandro Lucarella has written a D language binding (F<ev.d>) for libev, to
2790	be found at L<http://proj.llucax.com.ar/wiki/evd>.	3336	be found at L<http://proj.llucax.com.ar/wiki/evd>.
		3337
		3338	=item Ocaml
		3339
		3340	Erkki Seppala has written Ocaml bindings for libev, to be found at
		3341	L<http://modeemi.cs.tut.fi/~flux/software/ocaml-ev/>.
2791		3342
2792	=back	3343	=back
2793		3344
2794		3345
2795	=head1 MACRO MAGIC	3346	=head1 MACRO MAGIC
…		…
2896		3447
2897	#define EV_STANDALONE 1	3448	#define EV_STANDALONE 1
2898	#include "ev.h"	3449	#include "ev.h"
2899		3450
2900	Both header files and implementation files can be compiled with a C++	3451	Both header files and implementation files can be compiled with a C++
2901	compiler (at least, thats a stated goal, and breakage will be treated	3452	compiler (at least, that's a stated goal, and breakage will be treated
2902	as a bug).	3453	as a bug).
2903		3454
2904	You need the following files in your source tree, or in a directory	3455	You need the following files in your source tree, or in a directory
2905	in your include path (e.g. in libev/ when using -Ilibev):	3456	in your include path (e.g. in libev/ when using -Ilibev):
2906		3457
…		…
2962	keeps libev from including F<config.h>, and it also defines dummy	3513	keeps libev from including F<config.h>, and it also defines dummy
2963	implementations for some libevent functions (such as logging, which is not	3514	implementations for some libevent functions (such as logging, which is not
2964	supported). It will also not define any of the structs usually found in	3515	supported). It will also not define any of the structs usually found in
2965	F<event.h> that are not directly supported by the libev core alone.	3516	F<event.h> that are not directly supported by the libev core alone.
2966		3517
		3518	In stanbdalone mode, libev will still try to automatically deduce the
		3519	configuration, but has to be more conservative.
		3520
2967	=item EV_USE_MONOTONIC	3521	=item EV_USE_MONOTONIC
2968		3522
2969	If defined to be C<1>, libev will try to detect the availability of the	3523	If defined to be C<1>, libev will try to detect the availability of the
2970	monotonic clock option at both compile time and runtime. Otherwise no use	3524	monotonic clock option at both compile time and runtime. Otherwise no
2971	of the monotonic clock option will be attempted. If you enable this, you	3525	use of the monotonic clock option will be attempted. If you enable this,
2972	usually have to link against librt or something similar. Enabling it when	3526	you usually have to link against librt or something similar. Enabling it
2973	the functionality isn't available is safe, though, although you have	3527	when the functionality isn't available is safe, though, although you have
2974	to make sure you link against any libraries where the C<clock_gettime>	3528	to make sure you link against any libraries where the C<clock_gettime>
2975	function is hiding in (often F<-lrt>).	3529	function is hiding in (often F<-lrt>). See also C<EV_USE_CLOCK_SYSCALL>.
2976		3530
2977	=item EV_USE_REALTIME	3531	=item EV_USE_REALTIME
2978		3532
2979	If defined to be C<1>, libev will try to detect the availability of the	3533	If defined to be C<1>, libev will try to detect the availability of the
2980	real-time clock option at compile time (and assume its availability at	3534	real-time clock option at compile time (and assume its availability
2981	runtime if successful). Otherwise no use of the real-time clock option will	3535	at runtime if successful). Otherwise no use of the real-time clock
2982	be attempted. This effectively replaces C<gettimeofday> by C<clock_get	3536	option will be attempted. This effectively replaces C<gettimeofday>
2983	(CLOCK_REALTIME, ...)> and will not normally affect correctness. See the	3537	by C<clock_get (CLOCK_REALTIME, ...)> and will not normally affect
2984	note about libraries in the description of C<EV_USE_MONOTONIC>, though.	3538	correctness. See the note about libraries in the description of
		3539	C<EV_USE_MONOTONIC>, though. Defaults to the opposite value of
		3540	C<EV_USE_CLOCK_SYSCALL>.
		3541
		3542	=item EV_USE_CLOCK_SYSCALL
		3543
		3544	If defined to be C<1>, libev will try to use a direct syscall instead
		3545	of calling the system-provided C<clock_gettime> function. This option
		3546	exists because on GNU/Linux, C<clock_gettime> is in C<librt>, but C<librt>
		3547	unconditionally pulls in C<libpthread>, slowing down single-threaded
		3548	programs needlessly. Using a direct syscall is slightly slower (in
		3549	theory), because no optimised vdso implementation can be used, but avoids
		3550	the pthread dependency. Defaults to C<1> on GNU/Linux with glibc 2.x or
		3551	higher, as it simplifies linking (no need for C<-lrt>).
2985		3552
2986	=item EV_USE_NANOSLEEP	3553	=item EV_USE_NANOSLEEP
2987		3554
2988	If defined to be C<1>, libev will assume that C<nanosleep ()> is available	3555	If defined to be C<1>, libev will assume that C<nanosleep ()> is available
2989	and will use it for delays. Otherwise it will use C<select ()>.	3556	and will use it for delays. Otherwise it will use C<select ()>.
…		…
3005		3572
3006	=item EV_SELECT_USE_FD_SET	3573	=item EV_SELECT_USE_FD_SET
3007		3574
3008	If defined to C<1>, then the select backend will use the system C<fd_set>	3575	If defined to C<1>, then the select backend will use the system C<fd_set>
3009	structure. This is useful if libev doesn't compile due to a missing	3576	structure. This is useful if libev doesn't compile due to a missing
3010	C<NFDBITS> or C<fd_mask> definition or it mis-guesses the bitset layout on	3577	C<NFDBITS> or C<fd_mask> definition or it mis-guesses the bitset layout
3011	exotic systems. This usually limits the range of file descriptors to some	3578	on exotic systems. This usually limits the range of file descriptors to
3012	low limit such as 1024 or might have other limitations (winsocket only	3579	some low limit such as 1024 or might have other limitations (winsocket
3013	allows 64 sockets). The C<FD_SETSIZE> macro, set before compilation, might	3580	only allows 64 sockets). The C<FD_SETSIZE> macro, set before compilation,
3014	influence the size of the C<fd_set> used.	3581	configures the maximum size of the C<fd_set>.
3015		3582
3016	=item EV_SELECT_IS_WINSOCKET	3583	=item EV_SELECT_IS_WINSOCKET
3017		3584
3018	When defined to C<1>, the select backend will assume that	3585	When defined to C<1>, the select backend will assume that
3019	select/socket/connect etc. don't understand file descriptors but	3586	select/socket/connect etc. don't understand file descriptors but
…		…
3169	defined to be C<0>, then they are not.	3736	defined to be C<0>, then they are not.
3170		3737
3171	=item EV_MINIMAL	3738	=item EV_MINIMAL
3172		3739
3173	If you need to shave off some kilobytes of code at the expense of some	3740	If you need to shave off some kilobytes of code at the expense of some
3174	speed, define this symbol to C<1>. Currently this is used to override some	3741	speed (but with the full API), define this symbol to C<1>. Currently this
3175	inlining decisions, saves roughly 30% code size on amd64. It also selects a	3742	is used to override some inlining decisions, saves roughly 30% code size
3176	much smaller 2-heap for timer management over the default 4-heap.	3743	on amd64. It also selects a much smaller 2-heap for timer management over
		3744	the default 4-heap.
		3745
		3746	You can save even more by disabling watcher types you do not need
		3747	and setting C<EV_MAXPRI> == C<EV_MINPRI>. Also, disabling C<assert>
		3748	(C<-DNDEBUG>) will usually reduce code size a lot.
		3749
		3750	Defining C<EV_MINIMAL> to C<2> will additionally reduce the core API to
		3751	provide a bare-bones event library. See C<ev.h> for details on what parts
		3752	of the API are still available, and do not complain if this subset changes
		3753	over time.
3177		3754
3178	=item EV_PID_HASHSIZE	3755	=item EV_PID_HASHSIZE
3179		3756
3180	C<ev_child> watchers use a small hash table to distribute workload by	3757	C<ev_child> watchers use a small hash table to distribute workload by
3181	pid. The default size is C<16> (or C<1> with C<EV_MINIMAL>), usually more	3758	pid. The default size is C<16> (or C<1> with C<EV_MINIMAL>), usually more
…		…
3306	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:	3883	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:
3307		3884
3308	#include "ev_cpp.h"	3885	#include "ev_cpp.h"
3309	#include "ev.c"	3886	#include "ev.c"
3310		3887
		3888	=head1 INTERACTION WITH OTHER PROGRAMS OR LIBRARIES
3311		3889
3312	=head1 THREADS AND COROUTINES	3890	=head2 THREADS AND COROUTINES
3313		3891
3314	=head2 THREADS	3892	=head3 THREADS
3315		3893
3316	All libev functions are reentrant and thread-safe unless explicitly	3894	All libev functions are reentrant and thread-safe unless explicitly
3317	documented otherwise, but it uses no locking itself. This means that you	3895	documented otherwise, but libev implements no locking itself. This means
3318	can use as many loops as you want in parallel, as long as there are no	3896	that you can use as many loops as you want in parallel, as long as there
3319	concurrent calls into any libev function with the same loop parameter	3897	are no concurrent calls into any libev function with the same loop
3320	(C<ev_default_*> calls have an implicit default loop parameter, of	3898	parameter (C<ev_default_*> calls have an implicit default loop parameter,
3321	course): libev guarantees that different event loops share no data	3899	of course): libev guarantees that different event loops share no data
3322	structures that need any locking.	3900	structures that need any locking.
3323		3901
3324	Or to put it differently: calls with different loop parameters can be done	3902	Or to put it differently: calls with different loop parameters can be done
3325	concurrently from multiple threads, calls with the same loop parameter	3903	concurrently from multiple threads, calls with the same loop parameter
3326	must be done serially (but can be done from different threads, as long as	3904	must be done serially (but can be done from different threads, as long as
…		…
3366	default loop and triggering an C<ev_async> watcher from the default loop	3944	default loop and triggering an C<ev_async> watcher from the default loop
3367	watcher callback into the event loop interested in the signal.	3945	watcher callback into the event loop interested in the signal.
3368		3946
3369	=back	3947	=back
3370		3948
		3949	=head4 THREAD LOCKING EXAMPLE
		3950
		3951	Here is a fictitious example of how to run an event loop in a different
		3952	thread than where callbacks are being invoked and watchers are
		3953	created/added/removed.
		3954
		3955	For a real-world example, see the C<EV::Loop::Async> perl module,
		3956	which uses exactly this technique (which is suited for many high-level
		3957	languages).
		3958
		3959	The example uses a pthread mutex to protect the loop data, a condition
		3960	variable to wait for callback invocations, an async watcher to notify the
		3961	event loop thread and an unspecified mechanism to wake up the main thread.
		3962
		3963	First, you need to associate some data with the event loop:
		3964
		3965	typedef struct {
		3966	mutex_t lock; /* global loop lock */
		3967	ev_async async_w;
		3968	thread_t tid;
		3969	cond_t invoke_cv;
		3970	} userdata;
		3971
		3972	void prepare_loop (EV_P)
		3973	{
		3974	// for simplicity, we use a static userdata struct.
		3975	static userdata u;
		3976
		3977	ev_async_init (&u->async_w, async_cb);
		3978	ev_async_start (EV_A_ &u->async_w);
		3979
		3980	pthread_mutex_init (&u->lock, 0);
		3981	pthread_cond_init (&u->invoke_cv, 0);
		3982
		3983	// now associate this with the loop
		3984	ev_set_userdata (EV_A_ u);
		3985	ev_set_invoke_pending_cb (EV_A_ l_invoke);
		3986	ev_set_loop_release_cb (EV_A_ l_release, l_acquire);
		3987
		3988	// then create the thread running ev_loop
		3989	pthread_create (&u->tid, 0, l_run, EV_A);
		3990	}
		3991
		3992	The callback for the C<ev_async> watcher does nothing: the watcher is used
		3993	solely to wake up the event loop so it takes notice of any new watchers
		3994	that might have been added:
		3995
		3996	static void
		3997	async_cb (EV_P_ ev_async *w, int revents)
		3998	{
		3999	// just used for the side effects
		4000	}
		4001
		4002	The C<l_release> and C<l_acquire> callbacks simply unlock/lock the mutex
		4003	protecting the loop data, respectively.
		4004
		4005	static void
		4006	l_release (EV_P)
		4007	{
		4008	userdata *u = ev_userdata (EV_A);
		4009	pthread_mutex_unlock (&u->lock);
		4010	}
		4011
		4012	static void
		4013	l_acquire (EV_P)
		4014	{
		4015	userdata *u = ev_userdata (EV_A);
		4016	pthread_mutex_lock (&u->lock);
		4017	}
		4018
		4019	The event loop thread first acquires the mutex, and then jumps straight
		4020	into C<ev_loop>:
		4021
		4022	void *
		4023	l_run (void *thr_arg)
		4024	{
		4025	struct ev_loop *loop = (struct ev_loop *)thr_arg;
		4026
		4027	l_acquire (EV_A);
		4028	pthread_setcanceltype (PTHREAD_CANCEL_ASYNCHRONOUS, 0);
		4029	ev_loop (EV_A_ 0);
		4030	l_release (EV_A);
		4031
		4032	return 0;
		4033	}
		4034
		4035	Instead of invoking all pending watchers, the C<l_invoke> callback will
		4036	signal the main thread via some unspecified mechanism (signals? pipe
		4037	writes? C<Async::Interrupt>?) and then waits until all pending watchers
		4038	have been called (in a while loop because a) spurious wakeups are possible
		4039	and b) skipping inter-thread-communication when there are no pending
		4040	watchers is very beneficial):
		4041
		4042	static void
		4043	l_invoke (EV_P)
		4044	{
		4045	userdata *u = ev_userdata (EV_A);
		4046
		4047	while (ev_pending_count (EV_A))
		4048	{
		4049	wake_up_other_thread_in_some_magic_or_not_so_magic_way ();
		4050	pthread_cond_wait (&u->invoke_cv, &u->lock);
		4051	}
		4052	}
		4053
		4054	Now, whenever the main thread gets told to invoke pending watchers, it
		4055	will grab the lock, call C<ev_invoke_pending> and then signal the loop
		4056	thread to continue:
		4057
		4058	static void
		4059	real_invoke_pending (EV_P)
		4060	{
		4061	userdata *u = ev_userdata (EV_A);
		4062
		4063	pthread_mutex_lock (&u->lock);
		4064	ev_invoke_pending (EV_A);
		4065	pthread_cond_signal (&u->invoke_cv);
		4066	pthread_mutex_unlock (&u->lock);
		4067	}
		4068
		4069	Whenever you want to start/stop a watcher or do other modifications to an
		4070	event loop, you will now have to lock:
		4071
		4072	ev_timer timeout_watcher;
		4073	userdata *u = ev_userdata (EV_A);
		4074
		4075	ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);
		4076
		4077	pthread_mutex_lock (&u->lock);
		4078	ev_timer_start (EV_A_ &timeout_watcher);
		4079	ev_async_send (EV_A_ &u->async_w);
		4080	pthread_mutex_unlock (&u->lock);
		4081
		4082	Note that sending the C<ev_async> watcher is required because otherwise
		4083	an event loop currently blocking in the kernel will have no knowledge
		4084	about the newly added timer. By waking up the loop it will pick up any new
		4085	watchers in the next event loop iteration.
		4086
3371	=head2 COROUTINES	4087	=head3 COROUTINES
3372		4088
3373	Libev is much more accommodating to coroutines ("cooperative threads"):	4089	Libev is very accommodating to coroutines ("cooperative threads"):
3374	libev fully supports nesting calls to it's functions from different	4090	libev fully supports nesting calls to its functions from different
3375	coroutines (e.g. you can call C<ev_loop> on the same loop from two	4091	coroutines (e.g. you can call C<ev_loop> on the same loop from two
3376	different coroutines and switch freely between both coroutines running the	4092	different coroutines, and switch freely between both coroutines running
3377	loop, as long as you don't confuse yourself). The only exception is that	4093	the loop, as long as you don't confuse yourself). The only exception is
3378	you must not do this from C<ev_periodic> reschedule callbacks.	4094	that you must not do this from C<ev_periodic> reschedule callbacks.
3379		4095
3380	Care has been taken to ensure that libev does not keep local state inside	4096	Care has been taken to ensure that libev does not keep local state inside
3381	C<ev_loop>, and other calls do not usually allow coroutine switches.	4097	C<ev_loop>, and other calls do not usually allow for coroutine switches as
		4098	they do not call any callbacks.
3382		4099
		4100	=head2 COMPILER WARNINGS
3383		4101
3384	=head1 COMPLEXITIES	4102	Depending on your compiler and compiler settings, you might get no or a
		4103	lot of warnings when compiling libev code. Some people are apparently
		4104	scared by this.
3385		4105
3386	In this section the complexities of (many of) the algorithms used inside	4106	However, these are unavoidable for many reasons. For one, each compiler
3387	libev will be explained. For complexity discussions about backends see the	4107	has different warnings, and each user has different tastes regarding
3388	documentation for C<ev_default_init>.	4108	warning options. "Warn-free" code therefore cannot be a goal except when
		4109	targeting a specific compiler and compiler-version.
3389		4110
3390	All of the following are about amortised time: If an array needs to be	4111	Another reason is that some compiler warnings require elaborate
3391	extended, libev needs to realloc and move the whole array, but this	4112	workarounds, or other changes to the code that make it less clear and less
3392	happens asymptotically never with higher number of elements, so O(1) might	4113	maintainable.
3393	mean it might do a lengthy realloc operation in rare cases, but on average
3394	it is much faster and asymptotically approaches constant time.
3395		4114
3396	=over 4	4115	And of course, some compiler warnings are just plain stupid, or simply
		4116	wrong (because they don't actually warn about the condition their message
		4117	seems to warn about). For example, certain older gcc versions had some
		4118	warnings that resulted an extreme number of false positives. These have
		4119	been fixed, but some people still insist on making code warn-free with
		4120	such buggy versions.
3397		4121
3398	=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers)	4122	While libev is written to generate as few warnings as possible,
		4123	"warn-free" code is not a goal, and it is recommended not to build libev
		4124	with any compiler warnings enabled unless you are prepared to cope with
		4125	them (e.g. by ignoring them). Remember that warnings are just that:
		4126	warnings, not errors, or proof of bugs.
3399		4127
3400	This means that, when you have a watcher that triggers in one hour and
3401	there are 100 watchers that would trigger before that then inserting will
3402	have to skip roughly seven (C<ld 100>) of these watchers.
3403		4128
3404	=item Changing timer/periodic watchers (by autorepeat or calling again): O(log skipped_other_timers)	4129	=head2 VALGRIND
3405		4130
3406	That means that changing a timer costs less than removing/adding them	4131	Valgrind has a special section here because it is a popular tool that is
3407	as only the relative motion in the event queue has to be paid for.	4132	highly useful. Unfortunately, valgrind reports are very hard to interpret.
3408		4133
3409	=item Starting io/check/prepare/idle/signal/child/fork/async watchers: O(1)	4134	If you think you found a bug (memory leak, uninitialised data access etc.)
		4135	in libev, then check twice: If valgrind reports something like:
3410		4136
3411	These just add the watcher into an array or at the head of a list.	4137	==2274== definitely lost: 0 bytes in 0 blocks.
		4138	==2274== possibly lost: 0 bytes in 0 blocks.
		4139	==2274== still reachable: 256 bytes in 1 blocks.
3412		4140
3413	=item Stopping check/prepare/idle/fork/async watchers: O(1)	4141	Then there is no memory leak, just as memory accounted to global variables
		4142	is not a memleak - the memory is still being referenced, and didn't leak.
3414		4143
3415	=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % EV_PID_HASHSIZE))	4144	Similarly, under some circumstances, valgrind might report kernel bugs
		4145	as if it were a bug in libev (e.g. in realloc or in the poll backend,
		4146	although an acceptable workaround has been found here), or it might be
		4147	confused.
3416		4148
3417	These watchers are stored in lists then need to be walked to find the	4149	Keep in mind that valgrind is a very good tool, but only a tool. Don't
3418	correct watcher to remove. The lists are usually short (you don't usually	4150	make it into some kind of religion.
3419	have many watchers waiting for the same fd or signal).
3420		4151
3421	=item Finding the next timer in each loop iteration: O(1)	4152	If you are unsure about something, feel free to contact the mailing list
		4153	with the full valgrind report and an explanation on why you think this
		4154	is a bug in libev (best check the archives, too :). However, don't be
		4155	annoyed when you get a brisk "this is no bug" answer and take the chance
		4156	of learning how to interpret valgrind properly.
3422		4157
3423	By virtue of using a binary or 4-heap, the next timer is always found at a	4158	If you need, for some reason, empty reports from valgrind for your project
3424	fixed position in the storage array.	4159	I suggest using suppression lists.
3425		4160
3426	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)
3427		4161
3428	A change means an I/O watcher gets started or stopped, which requires	4162	=head1 PORTABILITY NOTES
3429	libev to recalculate its status (and possibly tell the kernel, depending
3430	on backend and whether C<ev_io_set> was used).
3431		4163
3432	=item Activating one watcher (putting it into the pending state): O(1)
3433
3434	=item Priority handling: O(number_of_priorities)
3435
3436	Priorities are implemented by allocating some space for each
3437	priority. When doing priority-based operations, libev usually has to
3438	linearly search all the priorities, but starting/stopping and activating
3439	watchers becomes O(1) with respect to priority handling.
3440
3441	=item Sending an ev_async: O(1)
3442
3443	=item Processing ev_async_send: O(number_of_async_watchers)
3444
3445	=item Processing signals: O(max_signal_number)
3446
3447	Sending involves a system call I<iff> there were no other C<ev_async_send>
3448	calls in the current loop iteration. Checking for async and signal events
3449	involves iterating over all running async watchers or all signal numbers.
3450
3451	=back
3452
3453
3454	=head1 WIN32 PLATFORM LIMITATIONS AND WORKAROUNDS	4164	=head2 WIN32 PLATFORM LIMITATIONS AND WORKAROUNDS
3455		4165
3456	Win32 doesn't support any of the standards (e.g. POSIX) that libev	4166	Win32 doesn't support any of the standards (e.g. POSIX) that libev
3457	requires, and its I/O model is fundamentally incompatible with the POSIX	4167	requires, and its I/O model is fundamentally incompatible with the POSIX
3458	model. Libev still offers limited functionality on this platform in	4168	model. Libev still offers limited functionality on this platform in
3459	the form of the C<EVBACKEND_SELECT> backend, and only supports socket	4169	the form of the C<EVBACKEND_SELECT> backend, and only supports socket
…		…
3466	way (note also that glib is the slowest event library known to man).	4176	way (note also that glib is the slowest event library known to man).
3467		4177
3468	There is no supported compilation method available on windows except	4178	There is no supported compilation method available on windows except
3469	embedding it into other applications.	4179	embedding it into other applications.
3470		4180
		4181	Sensible signal handling is officially unsupported by Microsoft - libev
		4182	tries its best, but under most conditions, signals will simply not work.
		4183
3471	Not a libev limitation but worth mentioning: windows apparently doesn't	4184	Not a libev limitation but worth mentioning: windows apparently doesn't
3472	accept large writes: instead of resulting in a partial write, windows will	4185	accept large writes: instead of resulting in a partial write, windows will
3473	either accept everything or return C<ENOBUFS> if the buffer is too large,	4186	either accept everything or return C<ENOBUFS> if the buffer is too large,
3474	so make sure you only write small amounts into your sockets (less than a	4187	so make sure you only write small amounts into your sockets (less than a
3475	megabyte seems safe, but this apparently depends on the amount of memory	4188	megabyte seems safe, but this apparently depends on the amount of memory
…		…
3479	the abysmal performance of winsockets, using a large number of sockets	4192	the abysmal performance of winsockets, using a large number of sockets
3480	is not recommended (and not reasonable). If your program needs to use	4193	is not recommended (and not reasonable). If your program needs to use
3481	more than a hundred or so sockets, then likely it needs to use a totally	4194	more than a hundred or so sockets, then likely it needs to use a totally
3482	different implementation for windows, as libev offers the POSIX readiness	4195	different implementation for windows, as libev offers the POSIX readiness
3483	notification model, which cannot be implemented efficiently on windows	4196	notification model, which cannot be implemented efficiently on windows
3484	(Microsoft monopoly games).	4197	(due to Microsoft monopoly games).
3485		4198
3486	A typical way to use libev under windows is to embed it (see the embedding	4199	A typical way to use libev under windows is to embed it (see the embedding
3487	section for details) and use the following F<evwrap.h> header file instead	4200	section for details) and use the following F<evwrap.h> header file instead
3488	of F<ev.h>:	4201	of F<ev.h>:
3489		4202
…		…
3525		4238
3526	Early versions of winsocket's select only supported waiting for a maximum	4239	Early versions of winsocket's select only supported waiting for a maximum
3527	of C<64> handles (probably owning to the fact that all windows kernels	4240	of C<64> handles (probably owning to the fact that all windows kernels
3528	can only wait for C<64> things at the same time internally; Microsoft	4241	can only wait for C<64> things at the same time internally; Microsoft
3529	recommends spawning a chain of threads and wait for 63 handles and the	4242	recommends spawning a chain of threads and wait for 63 handles and the
3530	previous thread in each. Great).	4243	previous thread in each. Sounds great!).
3531		4244
3532	Newer versions support more handles, but you need to define C<FD_SETSIZE>	4245	Newer versions support more handles, but you need to define C<FD_SETSIZE>
3533	to some high number (e.g. C<2048>) before compiling the winsocket select	4246	to some high number (e.g. C<2048>) before compiling the winsocket select
3534	call (which might be in libev or elsewhere, for example, perl does its own	4247	call (which might be in libev or elsewhere, for example, perl and many
3535	select emulation on windows).	4248	other interpreters do their own select emulation on windows).
3536		4249
3537	Another limit is the number of file descriptors in the Microsoft runtime	4250	Another limit is the number of file descriptors in the Microsoft runtime
3538	libraries, which by default is C<64> (there must be a hidden I<64> fetish	4251	libraries, which by default is C<64> (there must be a hidden I<64>
3539	or something like this inside Microsoft). You can increase this by calling	4252	fetish or something like this inside Microsoft). You can increase this
3540	C<_setmaxstdio>, which can increase this limit to C<2048> (another	4253	by calling C<_setmaxstdio>, which can increase this limit to C<2048>
3541	arbitrary limit), but is broken in many versions of the Microsoft runtime	4254	(another arbitrary limit), but is broken in many versions of the Microsoft
3542	libraries.
3543
3544	This might get you to about C<512> or C<2048> sockets (depending on	4255	runtime libraries. This might get you to about C<512> or C<2048> sockets
3545	windows version and/or the phase of the moon). To get more, you need to	4256	(depending on windows version and/or the phase of the moon). To get more,
3546	wrap all I/O functions and provide your own fd management, but the cost of	4257	you need to wrap all I/O functions and provide your own fd management, but
3547	calling select (O(n²)) will likely make this unworkable.	4258	the cost of calling select (O(n²)) will likely make this unworkable.
3548		4259
3549	=back	4260	=back
3550		4261
3551
3552	=head1 PORTABILITY REQUIREMENTS	4262	=head2 PORTABILITY REQUIREMENTS
3553		4263
3554	In addition to a working ISO-C implementation, libev relies on a few	4264	In addition to a working ISO-C implementation and of course the
3555	additional extensions:	4265	backend-specific APIs, libev relies on a few additional extensions:
3556		4266
3557	=over 4	4267	=over 4
3558		4268
3559	=item C<void (*)(ev_watcher_type *, int revents)> must have compatible	4269	=item C<void (*)(ev_watcher_type *, int revents)> must have compatible
3560	calling conventions regardless of C<ev_watcher_type *>.	4270	calling conventions regardless of C<ev_watcher_type *>.
…		…
3585	except the initial one, and run the default loop in the initial thread as	4295	except the initial one, and run the default loop in the initial thread as
3586	well.	4296	well.
3587		4297
3588	=item C<long> must be large enough for common memory allocation sizes	4298	=item C<long> must be large enough for common memory allocation sizes
3589		4299
3590	To improve portability and simplify using libev, libev uses C<long>	4300	To improve portability and simplify its API, libev uses C<long> internally
3591	internally instead of C<size_t> when allocating its data structures. On	4301	instead of C<size_t> when allocating its data structures. On non-POSIX
3592	non-POSIX systems (Microsoft...) this might be unexpectedly low, but	4302	systems (Microsoft...) this might be unexpectedly low, but is still at
3593	is still at least 31 bits everywhere, which is enough for hundreds of	4303	least 31 bits everywhere, which is enough for hundreds of millions of
3594	millions of watchers.	4304	watchers.
3595		4305
3596	=item C<double> must hold a time value in seconds with enough accuracy	4306	=item C<double> must hold a time value in seconds with enough accuracy
3597		4307
3598	The type C<double> is used to represent timestamps. It is required to	4308	The type C<double> is used to represent timestamps. It is required to
3599	have at least 51 bits of mantissa (and 9 bits of exponent), which is good	4309	have at least 51 bits of mantissa (and 9 bits of exponent), which is good
3600	enough for at least into the year 4000. This requirement is fulfilled by	4310	enough for at least into the year 4000. This requirement is fulfilled by
3601	implementations implementing IEEE 754 (basically all existing ones).	4311	implementations implementing IEEE 754, which is basically all existing
		4312	ones. With IEEE 754 doubles, you get microsecond accuracy until at least
		4313	2200.
3602		4314
3603	=back	4315	=back
3604		4316
3605	If you know of other additional requirements drop me a note.	4317	If you know of other additional requirements drop me a note.
3606		4318
3607		4319
3608	=head1 COMPILER WARNINGS	4320	=head1 ALGORITHMIC COMPLEXITIES
3609		4321
3610	Depending on your compiler and compiler settings, you might get no or a	4322	In this section the complexities of (many of) the algorithms used inside
3611	lot of warnings when compiling libev code. Some people are apparently	4323	libev will be documented. For complexity discussions about backends see
3612	scared by this.	4324	the documentation for C<ev_default_init>.
3613		4325
3614	However, these are unavoidable for many reasons. For one, each compiler	4326	All of the following are about amortised time: If an array needs to be
3615	has different warnings, and each user has different tastes regarding	4327	extended, libev needs to realloc and move the whole array, but this
3616	warning options. "Warn-free" code therefore cannot be a goal except when	4328	happens asymptotically rarer with higher number of elements, so O(1) might
3617	targeting a specific compiler and compiler-version.	4329	mean that libev does a lengthy realloc operation in rare cases, but on
		4330	average it is much faster and asymptotically approaches constant time.
3618		4331
3619	Another reason is that some compiler warnings require elaborate	4332	=over 4
3620	workarounds, or other changes to the code that make it less clear and less
3621	maintainable.
3622		4333
3623	And of course, some compiler warnings are just plain stupid, or simply	4334	=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers)
3624	wrong (because they don't actually warn about the condition their message
3625	seems to warn about).
3626		4335
3627	While libev is written to generate as few warnings as possible,	4336	This means that, when you have a watcher that triggers in one hour and
3628	"warn-free" code is not a goal, and it is recommended not to build libev	4337	there are 100 watchers that would trigger before that, then inserting will
3629	with any compiler warnings enabled unless you are prepared to cope with	4338	have to skip roughly seven (C<ld 100>) of these watchers.
3630	them (e.g. by ignoring them). Remember that warnings are just that:
3631	warnings, not errors, or proof of bugs.
3632		4339
		4340	=item Changing timer/periodic watchers (by autorepeat or calling again): O(log skipped_other_timers)
3633		4341
3634	=head1 VALGRIND	4342	That means that changing a timer costs less than removing/adding them,
		4343	as only the relative motion in the event queue has to be paid for.
3635		4344
3636	Valgrind has a special section here because it is a popular tool that is	4345	=item Starting io/check/prepare/idle/signal/child/fork/async watchers: O(1)
3637	highly useful, but valgrind reports are very hard to interpret.
3638		4346
3639	If you think you found a bug (memory leak, uninitialised data access etc.)	4347	These just add the watcher into an array or at the head of a list.
3640	in libev, then check twice: If valgrind reports something like:
3641		4348
3642	==2274== definitely lost: 0 bytes in 0 blocks.	4349	=item Stopping check/prepare/idle/fork/async watchers: O(1)
3643	==2274== possibly lost: 0 bytes in 0 blocks.
3644	==2274== still reachable: 256 bytes in 1 blocks.
3645		4350
3646	Then there is no memory leak. Similarly, under some circumstances,	4351	=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % EV_PID_HASHSIZE))
3647	valgrind might report kernel bugs as if it were a bug in libev, or it
3648	might be confused (it is a very good tool, but only a tool).
3649		4352
3650	If you are unsure about something, feel free to contact the mailing list	4353	These watchers are stored in lists, so they need to be walked to find the
3651	with the full valgrind report and an explanation on why you think this is	4354	correct watcher to remove. The lists are usually short (you don't usually
3652	a bug in libev. However, don't be annoyed when you get a brisk "this is	4355	have many watchers waiting for the same fd or signal: one is typical, two
3653	no bug" answer and take the chance of learning how to interpret valgrind	4356	is rare).
3654	properly.
3655		4357
3656	If you need, for some reason, empty reports from valgrind for your project	4358	=item Finding the next timer in each loop iteration: O(1)
3657	I suggest using suppression lists.
3658		4359
		4360	By virtue of using a binary or 4-heap, the next timer is always found at a
		4361	fixed position in the storage array.
		4362
		4363	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)
		4364
		4365	A change means an I/O watcher gets started or stopped, which requires
		4366	libev to recalculate its status (and possibly tell the kernel, depending
		4367	on backend and whether C<ev_io_set> was used).
		4368
		4369	=item Activating one watcher (putting it into the pending state): O(1)
		4370
		4371	=item Priority handling: O(number_of_priorities)
		4372
		4373	Priorities are implemented by allocating some space for each
		4374	priority. When doing priority-based operations, libev usually has to
		4375	linearly search all the priorities, but starting/stopping and activating
		4376	watchers becomes O(1) with respect to priority handling.
		4377
		4378	=item Sending an ev_async: O(1)
		4379
		4380	=item Processing ev_async_send: O(number_of_async_watchers)
		4381
		4382	=item Processing signals: O(max_signal_number)
		4383
		4384	Sending involves a system call I<iff> there were no other C<ev_async_send>
		4385	calls in the current loop iteration. Checking for async and signal events
		4386	involves iterating over all running async watchers or all signal numbers.
		4387
		4388	=back
		4389
		4390
		4391	=head1 GLOSSARY
		4392
		4393	=over 4
		4394
		4395	=item active
		4396
		4397	A watcher is active as long as it has been started (has been attached to
		4398	an event loop) but not yet stopped (disassociated from the event loop).
		4399
		4400	=item application
		4401
		4402	In this document, an application is whatever is using libev.
		4403
		4404	=item callback
		4405
		4406	The address of a function that is called when some event has been
		4407	detected. Callbacks are being passed the event loop, the watcher that
		4408	received the event, and the actual event bitset.
		4409
		4410	=item callback invocation
		4411
		4412	The act of calling the callback associated with a watcher.
		4413
		4414	=item event
		4415
		4416	A change of state of some external event, such as data now being available
		4417	for reading on a file descriptor, time having passed or simply not having
		4418	any other events happening anymore.
		4419
		4420	In libev, events are represented as single bits (such as C<EV_READ> or
		4421	C<EV_TIMEOUT>).
		4422
		4423	=item event library
		4424
		4425	A software package implementing an event model and loop.
		4426
		4427	=item event loop
		4428
		4429	An entity that handles and processes external events and converts them
		4430	into callback invocations.
		4431
		4432	=item event model
		4433
		4434	The model used to describe how an event loop handles and processes
		4435	watchers and events.
		4436
		4437	=item pending
		4438
		4439	A watcher is pending as soon as the corresponding event has been detected,
		4440	and stops being pending as soon as the watcher will be invoked or its
		4441	pending status is explicitly cleared by the application.
		4442
		4443	A watcher can be pending, but not active. Stopping a watcher also clears
		4444	its pending status.
		4445
		4446	=item real time
		4447
		4448	The physical time that is observed. It is apparently strictly monotonic :)
		4449
		4450	=item wall-clock time
		4451
		4452	The time and date as shown on clocks. Unlike real time, it can actually
		4453	be wrong and jump forwards and backwards, e.g. when the you adjust your
		4454	clock.
		4455
		4456	=item watcher
		4457
		4458	A data structure that describes interest in certain events. Watchers need
		4459	to be started (attached to an event loop) before they can receive events.
		4460
		4461	=item watcher invocation
		4462
		4463	The act of calling the callback associated with a watcher.
		4464
		4465	=back
3659		4466
3660	=head1 AUTHOR	4467	=head1 AUTHOR
3661		4468
3662	Marc Lehmann <libev@schmorp.de>.	4469	Marc Lehmann <libev@schmorp.de>, with repeated corrections by Mikael Magnusson.
3663		4470

Diff Legend

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