ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/libev/ev.pod

Comparing libev/ev.pod (file contents):
Revision 1.166 by root, Tue Jun 3 03:48:10 2008 UTC vs.
Revision 1.286 by root, Tue Mar 16 00:26:41 2010 UTC

…		…
9	=head2 EXAMPLE PROGRAM	9	=head2 EXAMPLE PROGRAM
10		10
11	// a single header file is required	11	// a single header file is required
12	#include <ev.h>	12	#include <ev.h>
13		13
		14	#include <stdio.h> // for puts
		15
14	// every watcher type has its own typedef'd struct	16	// every watcher type has its own typedef'd struct
15	// with the name ev_<type>	17	// with the name ev_TYPE
16	ev_io stdin_watcher;	18	ev_io stdin_watcher;
17	ev_timer timeout_watcher;	19	ev_timer timeout_watcher;
18		20
19	// all watcher callbacks have a similar signature	21	// all watcher callbacks have a similar signature
20	// this callback is called when data is readable on stdin	22	// this callback is called when data is readable on stdin
21	static void	23	static void
22	stdin_cb (EV_P_ struct ev_io *w, int revents)	24	stdin_cb (EV_P_ ev_io *w, int revents)
23	{	25	{
24	puts ("stdin ready");	26	puts ("stdin ready");
25	// for one-shot events, one must manually stop the watcher	27	// for one-shot events, one must manually stop the watcher
26	// with its corresponding stop function.	28	// with its corresponding stop function.
27	ev_io_stop (EV_A_ w);	29	ev_io_stop (EV_A_ w);
…		…
30	ev_unloop (EV_A_ EVUNLOOP_ALL);	32	ev_unloop (EV_A_ EVUNLOOP_ALL);
31	}	33	}
32		34
33	// another callback, this time for a time-out	35	// another callback, this time for a time-out
34	static void	36	static void
35	timeout_cb (EV_P_ struct ev_timer *w, int revents)	37	timeout_cb (EV_P_ ev_timer *w, int revents)
36	{	38	{
37	puts ("timeout");	39	puts ("timeout");
38	// this causes the innermost ev_loop to stop iterating	40	// this causes the innermost ev_loop to stop iterating
39	ev_unloop (EV_A_ EVUNLOOP_ONE);	41	ev_unloop (EV_A_ EVUNLOOP_ONE);
40	}	42	}
…		…
60		62
61	// unloop was called, so exit	63	// unloop was called, so exit
62	return 0;	64	return 0;
63	}	65	}
64		66
65	=head1 DESCRIPTION	67	=head1 ABOUT THIS DOCUMENT
		68
		69	This document documents the libev software package.
66		70
67	The newest version of this document is also available as an html-formatted	71	The newest version of this document is also available as an html-formatted
68	web page you might find easier to navigate when reading it for the first	72	web page you might find easier to navigate when reading it for the first
69	time: L<http://pod.tst.eu/http://cvs.schmorp.de/libev/ev.pod>.	73	time: L<http://pod.tst.eu/http://cvs.schmorp.de/libev/ev.pod>.
		74
		75	While this document tries to be as complete as possible in documenting
		76	libev, its usage and the rationale behind its design, it is not a tutorial
		77	on event-based programming, nor will it introduce event-based programming
		78	with libev.
		79
		80	Familarity with event based programming techniques in general is assumed
		81	throughout this document.
		82
		83	=head1 ABOUT LIBEV
70		84
71	Libev is an event loop: you register interest in certain events (such as a	85	Libev is an event loop: you register interest in certain events (such as a
72	file descriptor being readable or a timeout occurring), and it will manage	86	file descriptor being readable or a timeout occurring), and it will manage
73	these event sources and provide your program with events.	87	these event sources and provide your program with events.
74		88
…		…
84	=head2 FEATURES	98	=head2 FEATURES
85		99
86	Libev supports C<select>, C<poll>, the Linux-specific C<epoll>, the	100	Libev supports C<select>, C<poll>, the Linux-specific C<epoll>, the
87	BSD-specific C<kqueue> and the Solaris-specific event port mechanisms	101	BSD-specific C<kqueue> and the Solaris-specific event port mechanisms
88	for file descriptor events (C<ev_io>), the Linux C<inotify> interface	102	for file descriptor events (C<ev_io>), the Linux C<inotify> interface
89	(for C<ev_stat>), relative timers (C<ev_timer>), absolute timers	103	(for C<ev_stat>), Linux eventfd/signalfd (for faster and cleaner
90	with customised rescheduling (C<ev_periodic>), synchronous signals	104	inter-thread wakeup (C<ev_async>)/signal handling (C<ev_signal>)) relative
91	(C<ev_signal>), process status change events (C<ev_child>), and event	105	timers (C<ev_timer>), absolute timers with customised rescheduling
92	watchers dealing with the event loop mechanism itself (C<ev_idle>,	106	(C<ev_periodic>), synchronous signals (C<ev_signal>), process status
93	C<ev_embed>, C<ev_prepare> and C<ev_check> watchers) as well as	107	change events (C<ev_child>), and event watchers dealing with the event
94	file watchers (C<ev_stat>) and even limited support for fork events	108	loop mechanism itself (C<ev_idle>, C<ev_embed>, C<ev_prepare> and
95	(C<ev_fork>).	109	C<ev_check> watchers) as well as file watchers (C<ev_stat>) and even
		110	limited support for fork events (C<ev_fork>).
96		111
97	It also is quite fast (see this	112	It also is quite fast (see this
98	L<benchmark|http://libev.schmorp.de/bench.html> comparing it to libevent	113	L<benchmark|http://libev.schmorp.de/bench.html> comparing it to libevent
99	for example).	114	for example).
100		115
…		…
108	name C<loop> (which is always of type C<struct ev_loop *>) will not have	123	name C<loop> (which is always of type C<struct ev_loop *>) will not have
109	this argument.	124	this argument.
110		125
111	=head2 TIME REPRESENTATION	126	=head2 TIME REPRESENTATION
112		127
113	Libev represents time as a single floating point number, representing the	128	Libev represents time as a single floating point number, representing
114	(fractional) number of seconds since the (POSIX) epoch (somewhere near	129	the (fractional) number of seconds since the (POSIX) epoch (somewhere
115	the beginning of 1970, details are complicated, don't ask). This type is	130	near the beginning of 1970, details are complicated, don't ask). This
116	called C<ev_tstamp>, which is what you should use too. It usually aliases	131	type is called C<ev_tstamp>, which is what you should use too. It usually
117	to the C<double> type in C, and when you need to do any calculations on	132	aliases to the C<double> type in C. When you need to do any calculations
118	it, you should treat it as some floating point value. Unlike the name	133	on it, you should treat it as some floating point value. Unlike the name
119	component C<stamp> might indicate, it is also used for time differences	134	component C<stamp> might indicate, it is also used for time differences
120	throughout libev.	135	throughout libev.
121		136
122	=head1 ERROR HANDLING	137	=head1 ERROR HANDLING
123		138
…		…
214	C<ev_embeddable_backends () & ev_supported_backends ()>, likewise for	229	C<ev_embeddable_backends () & ev_supported_backends ()>, likewise for
215	recommended ones.	230	recommended ones.
216		231
217	See the description of C<ev_embed> watchers for more info.	232	See the description of C<ev_embed> watchers for more info.
218		233
219	=item ev_set_allocator (void *(*cb)(void *ptr, long size))	234	=item ev_set_allocator (void *(*cb)(void *ptr, long size)) [NOT REENTRANT]
220		235
221	Sets the allocation function to use (the prototype is similar - the	236	Sets the allocation function to use (the prototype is similar - the
222	semantics are identical to the C<realloc> C89/SuS/POSIX function). It is	237	semantics are identical to the C<realloc> C89/SuS/POSIX function). It is
223	used to allocate and free memory (no surprises here). If it returns zero	238	used to allocate and free memory (no surprises here). If it returns zero
224	when memory needs to be allocated (C<size != 0>), the library might abort	239	when memory needs to be allocated (C<size != 0>), the library might abort
…		…
250	}	265	}
251		266
252	...	267	...
253	ev_set_allocator (persistent_realloc);	268	ev_set_allocator (persistent_realloc);
254		269
255	=item ev_set_syserr_cb (void (*cb)(const char *msg));	270	=item ev_set_syserr_cb (void (*cb)(const char *msg)); [NOT REENTRANT]
256		271
257	Set the callback function to call on a retryable system call error (such	272	Set the callback function to call on a retryable system call error (such
258	as failed select, poll, epoll_wait). The message is a printable string	273	as failed select, poll, epoll_wait). The message is a printable string
259	indicating the system call or subsystem causing the problem. If this	274	indicating the system call or subsystem causing the problem. If this
260	callback is set, then libev will expect it to remedy the situation, no	275	callback is set, then libev will expect it to remedy the situation, no
…		…
276		291
277	=back	292	=back
278		293
279	=head1 FUNCTIONS CONTROLLING THE EVENT LOOP	294	=head1 FUNCTIONS CONTROLLING THE EVENT LOOP
280		295
281	An event loop is described by a C<struct ev_loop *>. The library knows two	296	An event loop is described by a C<struct ev_loop *> (the C<struct>
282	types of such loops, the I<default> loop, which supports signals and child	297	is I<not> optional in this case, as there is also an C<ev_loop>
283	events, and dynamically created loops which do not.	298	I<function>).
		299
		300	The library knows two types of such loops, the I<default> loop, which
		301	supports signals and child events, and dynamically created loops which do
		302	not.
284		303
285	=over 4	304	=over 4
286		305
287	=item struct ev_loop *ev_default_loop (unsigned int flags)	306	=item struct ev_loop *ev_default_loop (unsigned int flags)
288		307
…		…
294	If you don't know what event loop to use, use the one returned from this	313	If you don't know what event loop to use, use the one returned from this
295	function.	314	function.
296		315
297	Note that this function is I<not> thread-safe, so if you want to use it	316	Note that this function is I<not> thread-safe, so if you want to use it
298	from multiple threads, you have to lock (note also that this is unlikely,	317	from multiple threads, you have to lock (note also that this is unlikely,
299	as loops cannot bes hared easily between threads anyway).	318	as loops cannot be shared easily between threads anyway).
300		319
301	The default loop is the only loop that can handle C<ev_signal> and	320	The default loop is the only loop that can handle C<ev_signal> and
302	C<ev_child> watchers, and to do this, it always registers a handler	321	C<ev_child> watchers, and to do this, it always registers a handler
303	for C<SIGCHLD>. If this is a problem for your application you can either	322	for C<SIGCHLD>. If this is a problem for your application you can either
304	create a dynamic loop with C<ev_loop_new> that doesn't do that, or you	323	create a dynamic loop with C<ev_loop_new> that doesn't do that, or you
…		…
344	flag.	363	flag.
345		364
346	This flag setting cannot be overridden or specified in the C<LIBEV_FLAGS>	365	This flag setting cannot be overridden or specified in the C<LIBEV_FLAGS>
347	environment variable.	366	environment variable.
348		367
		368	=item C<EVFLAG_NOINOTIFY>
		369
		370	When this flag is specified, then libev will not attempt to use the
		371	I<inotify> API for it's C<ev_stat> watchers. Apart from debugging and
		372	testing, this flag can be useful to conserve inotify file descriptors, as
		373	otherwise each loop using C<ev_stat> watchers consumes one inotify handle.
		374
		375	=item C<EVFLAG_SIGNALFD>
		376
		377	When this flag is specified, then libev will attempt to use the
		378	I<signalfd> API for it's C<ev_signal> (and C<ev_child>) watchers. This API
		379	delivers signals synchronously, which makes it both faster and might make
		380	it possible to get the queued signal data. It can also simplify signal
		381	handling with threads, as long as you properly block signals in your
		382	threads that are not interested in handling them.
		383
		384	Signalfd will not be used by default as this changes your signal mask, and
		385	there are a lot of shoddy libraries and programs (glib's threadpool for
		386	example) that can't properly initialise their signal masks.
		387
349	=item C<EVBACKEND_SELECT> (value 1, portable select backend)	388	=item C<EVBACKEND_SELECT> (value 1, portable select backend)
350		389
351	This is your standard select(2) backend. Not I<completely> standard, as	390	This is your standard select(2) backend. Not I<completely> standard, as
352	libev tries to roll its own fd_set with no limits on the number of fds,	391	libev tries to roll its own fd_set with no limits on the number of fds,
353	but if that fails, expect a fairly low limit on the number of fds when	392	but if that fails, expect a fairly low limit on the number of fds when
…		…
359	writing a server, you should C<accept ()> in a loop to accept as many	398	writing a server, you should C<accept ()> in a loop to accept as many
360	connections as possible during one iteration. You might also want to have	399	connections as possible during one iteration. You might also want to have
361	a look at C<ev_set_io_collect_interval ()> to increase the amount of	400	a look at C<ev_set_io_collect_interval ()> to increase the amount of
362	readiness notifications you get per iteration.	401	readiness notifications you get per iteration.
363		402
		403	This backend maps C<EV_READ> to the C<readfds> set and C<EV_WRITE> to the
		404	C<writefds> set (and to work around Microsoft Windows bugs, also onto the
		405	C<exceptfds> set on that platform).
		406
364	=item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows)	407	=item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows)
365		408
366	And this is your standard poll(2) backend. It's more complicated	409	And this is your standard poll(2) backend. It's more complicated
367	than select, but handles sparse fds better and has no artificial	410	than select, but handles sparse fds better and has no artificial
368	limit on the number of fds you can use (except it will slow down	411	limit on the number of fds you can use (except it will slow down
369	considerably with a lot of inactive fds). It scales similarly to select,	412	considerably with a lot of inactive fds). It scales similarly to select,
370	i.e. O(total_fds). See the entry for C<EVBACKEND_SELECT>, above, for	413	i.e. O(total_fds). See the entry for C<EVBACKEND_SELECT>, above, for
371	performance tips.	414	performance tips.
372		415
		416	This backend maps C<EV_READ> to C<POLLIN | POLLERR | POLLHUP>, and
		417	C<EV_WRITE> to C<POLLOUT | POLLERR | POLLHUP>.
		418
373	=item C<EVBACKEND_EPOLL> (value 4, Linux)	419	=item C<EVBACKEND_EPOLL> (value 4, Linux)
		420
		421	Use the linux-specific epoll(7) interface (for both pre- and post-2.6.9
		422	kernels).
374		423
375	For few fds, this backend is a bit little slower than poll and select,	424	For few fds, this backend is a bit little slower than poll and select,
376	but it scales phenomenally better. While poll and select usually scale	425	but it scales phenomenally better. While poll and select usually scale
377	like O(total_fds) where n is the total number of fds (or the highest fd),	426	like O(total_fds) where n is the total number of fds (or the highest fd),
378	epoll scales either O(1) or O(active_fds). The epoll design has a number	427	epoll scales either O(1) or O(active_fds).
379	of shortcomings, such as silently dropping events in some hard-to-detect	428
380	cases and requiring a system call per fd change, no fork support and bad	429	The epoll mechanism deserves honorable mention as the most misdesigned
381	support for dup.	430	of the more advanced event mechanisms: mere annoyances include silently
		431	dropping file descriptors, requiring a system call per change per file
		432	descriptor (and unnecessary guessing of parameters), problems with dup and
		433	so on. The biggest issue is fork races, however - if a program forks then
		434	I<both> parent and child process have to recreate the epoll set, which can
		435	take considerable time (one syscall per file descriptor) and is of course
		436	hard to detect.
		437
		438	Epoll is also notoriously buggy - embedding epoll fds I<should> work, but
		439	of course I<doesn't>, and epoll just loves to report events for totally
		440	I<different> file descriptors (even already closed ones, so one cannot
		441	even remove them from the set) than registered in the set (especially
		442	on SMP systems). Libev tries to counter these spurious notifications by
		443	employing an additional generation counter and comparing that against the
		444	events to filter out spurious ones, recreating the set when required.
382		445
383	While stopping, setting and starting an I/O watcher in the same iteration	446	While stopping, setting and starting an I/O watcher in the same iteration
384	will result in some caching, there is still a system call per such incident	447	will result in some caching, there is still a system call per such
385	(because the fd could point to a different file description now), so its	448	incident (because the same I<file descriptor> could point to a different
386	best to avoid that. Also, C<dup ()>'ed file descriptors might not work	449	I<file description> now), so its best to avoid that. Also, C<dup ()>'ed
387	very well if you register events for both fds.	450	file descriptors might not work very well if you register events for both
388		451	file descriptors.
389	Please note that epoll sometimes generates spurious notifications, so you
390	need to use non-blocking I/O or other means to avoid blocking when no data
391	(or space) is available.
392		452
393	Best performance from this backend is achieved by not unregistering all	453	Best performance from this backend is achieved by not unregistering all
394	watchers for a file descriptor until it has been closed, if possible, i.e.	454	watchers for a file descriptor until it has been closed, if possible,
395	keep at least one watcher active per fd at all times.	455	i.e. keep at least one watcher active per fd at all times. Stopping and
		456	starting a watcher (without re-setting it) also usually doesn't cause
		457	extra overhead. A fork can both result in spurious notifications as well
		458	as in libev having to destroy and recreate the epoll object, which can
		459	take considerable time and thus should be avoided.
		460
		461	All this means that, in practice, C<EVBACKEND_SELECT> can be as fast or
		462	faster than epoll for maybe up to a hundred file descriptors, depending on
		463	the usage. So sad.
396		464
397	While nominally embeddable in other event loops, this feature is broken in	465	While nominally embeddable in other event loops, this feature is broken in
398	all kernel versions tested so far.	466	all kernel versions tested so far.
		467
		468	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
		469	C<EVBACKEND_POLL>.
399		470
400	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)	471	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)
401		472
402	Kqueue deserves special mention, as at the time of this writing, it	473	Kqueue deserves special mention, as at the time of this writing, it
403	was broken on all BSDs except NetBSD (usually it doesn't work reliably	474	was broken on all BSDs except NetBSD (usually it doesn't work reliably
404	with anything but sockets and pipes, except on Darwin, where of course	475	with anything but sockets and pipes, except on Darwin, where of course
405	it's completely useless). For this reason it's not being "auto-detected"	476	it's completely useless). Unlike epoll, however, whose brokenness
		477	is by design, these kqueue bugs can (and eventually will) be fixed
		478	without API changes to existing programs. For this reason it's not being
406	unless you explicitly specify it explicitly in the flags (i.e. using	479	"auto-detected" unless you explicitly specify it in the flags (i.e. using
407	C<EVBACKEND_KQUEUE>) or libev was compiled on a known-to-be-good (-enough)	480	C<EVBACKEND_KQUEUE>) or libev was compiled on a known-to-be-good (-enough)
408	system like NetBSD.	481	system like NetBSD.
409		482
410	You still can embed kqueue into a normal poll or select backend and use it	483	You still can embed kqueue into a normal poll or select backend and use it
411	only for sockets (after having made sure that sockets work with kqueue on	484	only for sockets (after having made sure that sockets work with kqueue on
…		…
413		486
414	It scales in the same way as the epoll backend, but the interface to the	487	It scales in the same way as the epoll backend, but the interface to the
415	kernel is more efficient (which says nothing about its actual speed, of	488	kernel is more efficient (which says nothing about its actual speed, of
416	course). While stopping, setting and starting an I/O watcher does never	489	course). While stopping, setting and starting an I/O watcher does never
417	cause an extra system call as with C<EVBACKEND_EPOLL>, it still adds up to	490	cause an extra system call as with C<EVBACKEND_EPOLL>, it still adds up to
418	two event changes per incident, support for C<fork ()> is very bad and it	491	two event changes per incident. Support for C<fork ()> is very bad (but
419	drops fds silently in similarly hard-to-detect cases.	492	sane, unlike epoll) and it drops fds silently in similarly hard-to-detect
		493	cases
420		494
421	This backend usually performs well under most conditions.	495	This backend usually performs well under most conditions.
422		496
423	While nominally embeddable in other event loops, this doesn't work	497	While nominally embeddable in other event loops, this doesn't work
424	everywhere, so you might need to test for this. And since it is broken	498	everywhere, so you might need to test for this. And since it is broken
425	almost everywhere, you should only use it when you have a lot of sockets	499	almost everywhere, you should only use it when you have a lot of sockets
426	(for which it usually works), by embedding it into another event loop	500	(for which it usually works), by embedding it into another event loop
427	(e.g. C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>) and using it only for	501	(e.g. C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> (but C<poll> is of course
428	sockets.	502	also broken on OS X)) and, did I mention it, using it only for sockets.
		503
		504	This backend maps C<EV_READ> into an C<EVFILT_READ> kevent with
		505	C<NOTE_EOF>, and C<EV_WRITE> into an C<EVFILT_WRITE> kevent with
		506	C<NOTE_EOF>.
429		507
430	=item C<EVBACKEND_DEVPOLL> (value 16, Solaris 8)	508	=item C<EVBACKEND_DEVPOLL> (value 16, Solaris 8)
431		509
432	This is not implemented yet (and might never be, unless you send me an	510	This is not implemented yet (and might never be, unless you send me an
433	implementation). According to reports, C</dev/poll> only supports sockets	511	implementation). According to reports, C</dev/poll> only supports sockets
…		…
446	While this backend scales well, it requires one system call per active	524	While this backend scales well, it requires one system call per active
447	file descriptor per loop iteration. For small and medium numbers of file	525	file descriptor per loop iteration. For small and medium numbers of file
448	descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend	526	descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend
449	might perform better.	527	might perform better.
450		528
451	On the positive side, ignoring the spurious readiness notifications, this	529	On the positive side, with the exception of the spurious readiness
452	backend actually performed to specification in all tests and is fully	530	notifications, this backend actually performed fully to specification
453	embeddable, which is a rare feat among the OS-specific backends.	531	in all tests and is fully embeddable, which is a rare feat among the
		532	OS-specific backends (I vastly prefer correctness over speed hacks).
		533
		534	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
		535	C<EVBACKEND_POLL>.
454		536
455	=item C<EVBACKEND_ALL>	537	=item C<EVBACKEND_ALL>
456		538
457	Try all backends (even potentially broken ones that wouldn't be tried	539	Try all backends (even potentially broken ones that wouldn't be tried
458	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as	540	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as
…		…
460		542
461	It is definitely not recommended to use this flag.	543	It is definitely not recommended to use this flag.
462		544
463	=back	545	=back
464		546
465	If one or more of these are or'ed into the flags value, then only these	547	If one or more of the backend flags are or'ed into the flags value,
466	backends will be tried (in the reverse order as listed here). If none are	548	then only these backends will be tried (in the reverse order as listed
467	specified, all backends in C<ev_recommended_backends ()> will be tried.	549	here). If none are specified, all backends in C<ev_recommended_backends
		550	()> will be tried.
468		551
469	The most typical usage is like this:	552	Example: This is the most typical usage.
470		553
471	if (!ev_default_loop (0))	554	if (!ev_default_loop (0))
472	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");	555	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
473		556
474	Restrict libev to the select and poll backends, and do not allow	557	Example: Restrict libev to the select and poll backends, and do not allow
475	environment settings to be taken into account:	558	environment settings to be taken into account:
476		559
477	ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);	560	ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);
478		561
479	Use whatever libev has to offer, but make sure that kqueue is used if	562	Example: Use whatever libev has to offer, but make sure that kqueue is
480	available (warning, breaks stuff, best use only with your own private	563	used if available (warning, breaks stuff, best use only with your own
481	event loop and only if you know the OS supports your types of fds):	564	private event loop and only if you know the OS supports your types of
		565	fds):
482		566
483	ev_default_loop (ev_recommended_backends () | EVBACKEND_KQUEUE);	567	ev_default_loop (ev_recommended_backends () | EVBACKEND_KQUEUE);
484		568
485	=item struct ev_loop *ev_loop_new (unsigned int flags)	569	=item struct ev_loop *ev_loop_new (unsigned int flags)
486		570
…		…
507	responsibility to either stop all watchers cleanly yourself I<before>	591	responsibility to either stop all watchers cleanly yourself I<before>
508	calling this function, or cope with the fact afterwards (which is usually	592	calling this function, or cope with the fact afterwards (which is usually
509	the easiest thing, you can just ignore the watchers and/or C<free ()> them	593	the easiest thing, you can just ignore the watchers and/or C<free ()> them
510	for example).	594	for example).
511		595
512	Note that certain global state, such as signal state, will not be freed by	596	Note that certain global state, such as signal state (and installed signal
513	this function, and related watchers (such as signal and child watchers)	597	handlers), will not be freed by this function, and related watchers (such
514	would need to be stopped manually.	598	as signal and child watchers) would need to be stopped manually.
515		599
516	In general it is not advisable to call this function except in the	600	In general it is not advisable to call this function except in the
517	rare occasion where you really need to free e.g. the signal handling	601	rare occasion where you really need to free e.g. the signal handling
518	pipe fds. If you need dynamically allocated loops it is better to use	602	pipe fds. If you need dynamically allocated loops it is better to use
519	C<ev_loop_new> and C<ev_loop_destroy>).	603	C<ev_loop_new> and C<ev_loop_destroy>.
520		604
521	=item ev_loop_destroy (loop)	605	=item ev_loop_destroy (loop)
522		606
523	Like C<ev_default_destroy>, but destroys an event loop created by an	607	Like C<ev_default_destroy>, but destroys an event loop created by an
524	earlier call to C<ev_loop_new>.	608	earlier call to C<ev_loop_new>.
…		…
544		628
545	=item ev_loop_fork (loop)	629	=item ev_loop_fork (loop)
546		630
547	Like C<ev_default_fork>, but acts on an event loop created by	631	Like C<ev_default_fork>, but acts on an event loop created by
548	C<ev_loop_new>. Yes, you have to call this on every allocated event loop	632	C<ev_loop_new>. Yes, you have to call this on every allocated event loop
549	after fork, and how you do this is entirely your own problem.	633	after fork that you want to re-use in the child, and how you do this is
		634	entirely your own problem.
550		635
551	=item int ev_is_default_loop (loop)	636	=item int ev_is_default_loop (loop)
552		637
553	Returns true when the given loop actually is the default loop, false otherwise.	638	Returns true when the given loop is, in fact, the default loop, and false
		639	otherwise.
554		640
555	=item unsigned int ev_loop_count (loop)	641	=item unsigned int ev_loop_count (loop)
556		642
557	Returns the count of loop iterations for the loop, which is identical to	643	Returns the count of loop iterations for the loop, which is identical to
558	the number of times libev did poll for new events. It starts at C<0> and	644	the number of times libev did poll for new events. It starts at C<0> and
559	happily wraps around with enough iterations.	645	happily wraps around with enough iterations.
560		646
561	This value can sometimes be useful as a generation counter of sorts (it	647	This value can sometimes be useful as a generation counter of sorts (it
562	"ticks" the number of loop iterations), as it roughly corresponds with	648	"ticks" the number of loop iterations), as it roughly corresponds with
563	C<ev_prepare> and C<ev_check> calls.	649	C<ev_prepare> and C<ev_check> calls.
		650
		651	=item unsigned int ev_loop_depth (loop)
		652
		653	Returns the number of times C<ev_loop> was entered minus the number of
		654	times C<ev_loop> was exited, in other words, the recursion depth.
		655
		656	Outside C<ev_loop>, this number is zero. In a callback, this number is
		657	C<1>, unless C<ev_loop> was invoked recursively (or from another thread),
		658	in which case it is higher.
		659
		660	Leaving C<ev_loop> abnormally (setjmp/longjmp, cancelling the thread
		661	etc.), doesn't count as exit.
564		662
565	=item unsigned int ev_backend (loop)	663	=item unsigned int ev_backend (loop)
566		664
567	Returns one of the C<EVBACKEND_*> flags indicating the event backend in	665	Returns one of the C<EVBACKEND_*> flags indicating the event backend in
568	use.	666	use.
…		…
573	received events and started processing them. This timestamp does not	671	received events and started processing them. This timestamp does not
574	change as long as callbacks are being processed, and this is also the base	672	change as long as callbacks are being processed, and this is also the base
575	time used for relative timers. You can treat it as the timestamp of the	673	time used for relative timers. You can treat it as the timestamp of the
576	event occurring (or more correctly, libev finding out about it).	674	event occurring (or more correctly, libev finding out about it).
577		675
		676	=item ev_now_update (loop)
		677
		678	Establishes the current time by querying the kernel, updating the time
		679	returned by C<ev_now ()> in the progress. This is a costly operation and
		680	is usually done automatically within C<ev_loop ()>.
		681
		682	This function is rarely useful, but when some event callback runs for a
		683	very long time without entering the event loop, updating libev's idea of
		684	the current time is a good idea.
		685
		686	See also L<The special problem of time updates> in the C<ev_timer> section.
		687
		688	=item ev_suspend (loop)
		689
		690	=item ev_resume (loop)
		691
		692	These two functions suspend and resume a loop, for use when the loop is
		693	not used for a while and timeouts should not be processed.
		694
		695	A typical use case would be an interactive program such as a game: When
		696	the user presses C<^Z> to suspend the game and resumes it an hour later it
		697	would be best to handle timeouts as if no time had actually passed while
		698	the program was suspended. This can be achieved by calling C<ev_suspend>
		699	in your C<SIGTSTP> handler, sending yourself a C<SIGSTOP> and calling
		700	C<ev_resume> directly afterwards to resume timer processing.
		701
		702	Effectively, all C<ev_timer> watchers will be delayed by the time spend
		703	between C<ev_suspend> and C<ev_resume>, and all C<ev_periodic> watchers
		704	will be rescheduled (that is, they will lose any events that would have
		705	occured while suspended).
		706
		707	After calling C<ev_suspend> you B<must not> call I<any> function on the
		708	given loop other than C<ev_resume>, and you B<must not> call C<ev_resume>
		709	without a previous call to C<ev_suspend>.
		710
		711	Calling C<ev_suspend>/C<ev_resume> has the side effect of updating the
		712	event loop time (see C<ev_now_update>).
		713
578	=item ev_loop (loop, int flags)	714	=item ev_loop (loop, int flags)
579		715
580	Finally, this is it, the event handler. This function usually is called	716	Finally, this is it, the event handler. This function usually is called
581	after you initialised all your watchers and you want to start handling	717	after you have initialised all your watchers and you want to start
582	events.	718	handling events.
583		719
584	If the flags argument is specified as C<0>, it will not return until	720	If the flags argument is specified as C<0>, it will not return until
585	either no event watchers are active anymore or C<ev_unloop> was called.	721	either no event watchers are active anymore or C<ev_unloop> was called.
586		722
587	Please note that an explicit C<ev_unloop> is usually better than	723	Please note that an explicit C<ev_unloop> is usually better than
588	relying on all watchers to be stopped when deciding when a program has	724	relying on all watchers to be stopped when deciding when a program has
589	finished (especially in interactive programs), but having a program that	725	finished (especially in interactive programs), but having a program
590	automatically loops as long as it has to and no longer by virtue of	726	that automatically loops as long as it has to and no longer by virtue
591	relying on its watchers stopping correctly is a thing of beauty.	727	of relying on its watchers stopping correctly, that is truly a thing of
		728	beauty.
592		729
593	A flags value of C<EVLOOP_NONBLOCK> will look for new events, will handle	730	A flags value of C<EVLOOP_NONBLOCK> will look for new events, will handle
594	those events and any outstanding ones, but will not block your process in	731	those events and any already outstanding ones, but will not block your
595	case there are no events and will return after one iteration of the loop.	732	process in case there are no events and will return after one iteration of
		733	the loop.
596		734
597	A flags value of C<EVLOOP_ONESHOT> will look for new events (waiting if	735	A flags value of C<EVLOOP_ONESHOT> will look for new events (waiting if
598	necessary) and will handle those and any outstanding ones. It will block	736	necessary) and will handle those and any already outstanding ones. It
599	your process until at least one new event arrives, and will return after	737	will block your process until at least one new event arrives (which could
600	one iteration of the loop. This is useful if you are waiting for some	738	be an event internal to libev itself, so there is no guarantee that a
601	external event in conjunction with something not expressible using other	739	user-registered callback will be called), and will return after one
		740	iteration of the loop.
		741
		742	This is useful if you are waiting for some external event in conjunction
		743	with something not expressible using other libev watchers (i.e. "roll your
602	libev watchers. However, a pair of C<ev_prepare>/C<ev_check> watchers is	744	own C<ev_loop>"). However, a pair of C<ev_prepare>/C<ev_check> watchers is
603	usually a better approach for this kind of thing.	745	usually a better approach for this kind of thing.
604		746
605	Here are the gory details of what C<ev_loop> does:	747	Here are the gory details of what C<ev_loop> does:
606		748
607	- Before the first iteration, call any pending watchers.	749	- Before the first iteration, call any pending watchers.
608	* If EVFLAG_FORKCHECK was used, check for a fork.	750	* If EVFLAG_FORKCHECK was used, check for a fork.
609	- If a fork was detected, queue and call all fork watchers.	751	- If a fork was detected (by any means), queue and call all fork watchers.
610	- Queue and call all prepare watchers.	752	- Queue and call all prepare watchers.
611	- If we have been forked, recreate the kernel state.	753	- If we have been forked, detach and recreate the kernel state
		754	as to not disturb the other process.
612	- Update the kernel state with all outstanding changes.	755	- Update the kernel state with all outstanding changes.
613	- Update the "event loop time".	756	- Update the "event loop time" (ev_now ()).
614	- Calculate for how long to sleep or block, if at all	757	- Calculate for how long to sleep or block, if at all
615	(active idle watchers, EVLOOP_NONBLOCK or not having	758	(active idle watchers, EVLOOP_NONBLOCK or not having
616	any active watchers at all will result in not sleeping).	759	any active watchers at all will result in not sleeping).
617	- Sleep if the I/O and timer collect interval say so.	760	- Sleep if the I/O and timer collect interval say so.
618	- Block the process, waiting for any events.	761	- Block the process, waiting for any events.
619	- Queue all outstanding I/O (fd) events.	762	- Queue all outstanding I/O (fd) events.
620	- Update the "event loop time" and do time jump handling.	763	- Update the "event loop time" (ev_now ()), and do time jump adjustments.
621	- Queue all outstanding timers.	764	- Queue all expired timers.
622	- Queue all outstanding periodics.	765	- Queue all expired periodics.
623	- If no events are pending now, queue all idle watchers.	766	- Unless any events are pending now, queue all idle watchers.
624	- Queue all check watchers.	767	- Queue all check watchers.
625	- Call all queued watchers in reverse order (i.e. check watchers first).	768	- Call all queued watchers in reverse order (i.e. check watchers first).
626	Signals and child watchers are implemented as I/O watchers, and will	769	Signals and child watchers are implemented as I/O watchers, and will
627	be handled here by queueing them when their watcher gets executed.	770	be handled here by queueing them when their watcher gets executed.
628	- If ev_unloop has been called, or EVLOOP_ONESHOT or EVLOOP_NONBLOCK	771	- If ev_unloop has been called, or EVLOOP_ONESHOT or EVLOOP_NONBLOCK
…		…
633	anymore.	776	anymore.
634		777
635	... queue jobs here, make sure they register event watchers as long	778	... queue jobs here, make sure they register event watchers as long
636	... as they still have work to do (even an idle watcher will do..)	779	... as they still have work to do (even an idle watcher will do..)
637	ev_loop (my_loop, 0);	780	ev_loop (my_loop, 0);
638	... jobs done. yeah!	781	... jobs done or somebody called unloop. yeah!
639		782
640	=item ev_unloop (loop, how)	783	=item ev_unloop (loop, how)
641		784
642	Can be used to make a call to C<ev_loop> return early (but only after it	785	Can be used to make a call to C<ev_loop> return early (but only after it
643	has processed all outstanding events). The C<how> argument must be either	786	has processed all outstanding events). The C<how> argument must be either
644	C<EVUNLOOP_ONE>, which will make the innermost C<ev_loop> call return, or	787	C<EVUNLOOP_ONE>, which will make the innermost C<ev_loop> call return, or
645	C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.	788	C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.
646		789
647	This "unloop state" will be cleared when entering C<ev_loop> again.	790	This "unloop state" will be cleared when entering C<ev_loop> again.
648		791
		792	It is safe to call C<ev_unloop> from otuside any C<ev_loop> calls.
		793
649	=item ev_ref (loop)	794	=item ev_ref (loop)
650		795
651	=item ev_unref (loop)	796	=item ev_unref (loop)
652		797
653	Ref/unref can be used to add or remove a reference count on the event	798	Ref/unref can be used to add or remove a reference count on the event
654	loop: Every watcher keeps one reference, and as long as the reference	799	loop: Every watcher keeps one reference, and as long as the reference
655	count is nonzero, C<ev_loop> will not return on its own. If you have	800	count is nonzero, C<ev_loop> will not return on its own.
656	a watcher you never unregister that should not keep C<ev_loop> from	801
657	returning, ev_unref() after starting, and ev_ref() before stopping it. For	802	This is useful when you have a watcher that you never intend to
		803	unregister, but that nevertheless should not keep C<ev_loop> from
		804	returning. In such a case, call C<ev_unref> after starting, and C<ev_ref>
		805	before stopping it.
		806
658	example, libev itself uses this for its internal signal pipe: It is not	807	As an example, libev itself uses this for its internal signal pipe: It
659	visible to the libev user and should not keep C<ev_loop> from exiting if	808	is not visible to the libev user and should not keep C<ev_loop> from
660	no event watchers registered by it are active. It is also an excellent	809	exiting if no event watchers registered by it are active. It is also an
661	way to do this for generic recurring timers or from within third-party	810	excellent way to do this for generic recurring timers or from within
662	libraries. Just remember to I<unref after start> and I<ref before stop>	811	third-party libraries. Just remember to I<unref after start> and I<ref
663	(but only if the watcher wasn't active before, or was active before,	812	before stop> (but only if the watcher wasn't active before, or was active
664	respectively).	813	before, respectively. Note also that libev might stop watchers itself
		814	(e.g. non-repeating timers) in which case you have to C<ev_ref>
		815	in the callback).
665		816
666	Example: Create a signal watcher, but keep it from keeping C<ev_loop>	817	Example: Create a signal watcher, but keep it from keeping C<ev_loop>
667	running when nothing else is active.	818	running when nothing else is active.
668		819
669	struct ev_signal exitsig;	820	ev_signal exitsig;
670	ev_signal_init (&exitsig, sig_cb, SIGINT);	821	ev_signal_init (&exitsig, sig_cb, SIGINT);
671	ev_signal_start (loop, &exitsig);	822	ev_signal_start (loop, &exitsig);
672	evf_unref (loop);	823	evf_unref (loop);
673		824
674	Example: For some weird reason, unregister the above signal handler again.	825	Example: For some weird reason, unregister the above signal handler again.
…		…
679	=item ev_set_io_collect_interval (loop, ev_tstamp interval)	830	=item ev_set_io_collect_interval (loop, ev_tstamp interval)
680		831
681	=item ev_set_timeout_collect_interval (loop, ev_tstamp interval)	832	=item ev_set_timeout_collect_interval (loop, ev_tstamp interval)
682		833
683	These advanced functions influence the time that libev will spend waiting	834	These advanced functions influence the time that libev will spend waiting
684	for events. Both are by default C<0>, meaning that libev will try to	835	for events. Both time intervals are by default C<0>, meaning that libev
685	invoke timer/periodic callbacks and I/O callbacks with minimum latency.	836	will try to invoke timer/periodic callbacks and I/O callbacks with minimum
		837	latency.
686		838
687	Setting these to a higher value (the C<interval> I<must> be >= C<0>)	839	Setting these to a higher value (the C<interval> I<must> be >= C<0>)
688	allows libev to delay invocation of I/O and timer/periodic callbacks to	840	allows libev to delay invocation of I/O and timer/periodic callbacks
689	increase efficiency of loop iterations.	841	to increase efficiency of loop iterations (or to increase power-saving
		842	opportunities).
690		843
691	The background is that sometimes your program runs just fast enough to	844	The idea is that sometimes your program runs just fast enough to handle
692	handle one (or very few) event(s) per loop iteration. While this makes	845	one (or very few) event(s) per loop iteration. While this makes the
693	the program responsive, it also wastes a lot of CPU time to poll for new	846	program responsive, it also wastes a lot of CPU time to poll for new
694	events, especially with backends like C<select ()> which have a high	847	events, especially with backends like C<select ()> which have a high
695	overhead for the actual polling but can deliver many events at once.	848	overhead for the actual polling but can deliver many events at once.
696		849
697	By setting a higher I<io collect interval> you allow libev to spend more	850	By setting a higher I<io collect interval> you allow libev to spend more
698	time collecting I/O events, so you can handle more events per iteration,	851	time collecting I/O events, so you can handle more events per iteration,
699	at the cost of increasing latency. Timeouts (both C<ev_periodic> and	852	at the cost of increasing latency. Timeouts (both C<ev_periodic> and
700	C<ev_timer>) will be not affected. Setting this to a non-null value will	853	C<ev_timer>) will be not affected. Setting this to a non-null value will
701	introduce an additional C<ev_sleep ()> call into most loop iterations.	854	introduce an additional C<ev_sleep ()> call into most loop iterations. The
		855	sleep time ensures that libev will not poll for I/O events more often then
		856	once per this interval, on average.
702		857
703	Likewise, by setting a higher I<timeout collect interval> you allow libev	858	Likewise, by setting a higher I<timeout collect interval> you allow libev
704	to spend more time collecting timeouts, at the expense of increased	859	to spend more time collecting timeouts, at the expense of increased
705	latency (the watcher callback will be called later). C<ev_io> watchers	860	latency/jitter/inexactness (the watcher callback will be called
706	will not be affected. Setting this to a non-null value will not introduce	861	later). C<ev_io> watchers will not be affected. Setting this to a non-null
707	any overhead in libev.	862	value will not introduce any overhead in libev.
708		863
709	Many (busy) programs can usually benefit by setting the I/O collect	864	Many (busy) programs can usually benefit by setting the I/O collect
710	interval to a value near C<0.1> or so, which is often enough for	865	interval to a value near C<0.1> or so, which is often enough for
711	interactive servers (of course not for games), likewise for timeouts. It	866	interactive servers (of course not for games), likewise for timeouts. It
712	usually doesn't make much sense to set it to a lower value than C<0.01>,	867	usually doesn't make much sense to set it to a lower value than C<0.01>,
713	as this approaches the timing granularity of most systems.	868	as this approaches the timing granularity of most systems. Note that if
		869	you do transactions with the outside world and you can't increase the
		870	parallelity, then this setting will limit your transaction rate (if you
		871	need to poll once per transaction and the I/O collect interval is 0.01,
		872	then you can't do more than 100 transations per second).
		873
		874	Setting the I<timeout collect interval> can improve the opportunity for
		875	saving power, as the program will "bundle" timer callback invocations that
		876	are "near" in time together, by delaying some, thus reducing the number of
		877	times the process sleeps and wakes up again. Another useful technique to
		878	reduce iterations/wake-ups is to use C<ev_periodic> watchers and make sure
		879	they fire on, say, one-second boundaries only.
		880
		881	Example: we only need 0.1s timeout granularity, and we wish not to poll
		882	more often than 100 times per second:
		883
		884	ev_set_timeout_collect_interval (EV_DEFAULT_UC_ 0.1);
		885	ev_set_io_collect_interval (EV_DEFAULT_UC_ 0.01);
		886
		887	=item ev_invoke_pending (loop)
		888
		889	This call will simply invoke all pending watchers while resetting their
		890	pending state. Normally, C<ev_loop> does this automatically when required,
		891	but when overriding the invoke callback this call comes handy.
		892
		893	=item int ev_pending_count (loop)
		894
		895	Returns the number of pending watchers - zero indicates that no watchers
		896	are pending.
		897
		898	=item ev_set_invoke_pending_cb (loop, void (*invoke_pending_cb)(EV_P))
		899
		900	This overrides the invoke pending functionality of the loop: Instead of
		901	invoking all pending watchers when there are any, C<ev_loop> will call
		902	this callback instead. This is useful, for example, when you want to
		903	invoke the actual watchers inside another context (another thread etc.).
		904
		905	If you want to reset the callback, use C<ev_invoke_pending> as new
		906	callback.
		907
		908	=item ev_set_loop_release_cb (loop, void (*release)(EV_P), void (*acquire)(EV_P))
		909
		910	Sometimes you want to share the same loop between multiple threads. This
		911	can be done relatively simply by putting mutex_lock/unlock calls around
		912	each call to a libev function.
		913
		914	However, C<ev_loop> can run an indefinite time, so it is not feasible to
		915	wait for it to return. One way around this is to wake up the loop via
		916	C<ev_unloop> and C<av_async_send>, another way is to set these I<release>
		917	and I<acquire> callbacks on the loop.
		918
		919	When set, then C<release> will be called just before the thread is
		920	suspended waiting for new events, and C<acquire> is called just
		921	afterwards.
		922
		923	Ideally, C<release> will just call your mutex_unlock function, and
		924	C<acquire> will just call the mutex_lock function again.
		925
		926	While event loop modifications are allowed between invocations of
		927	C<release> and C<acquire> (that's their only purpose after all), no
		928	modifications done will affect the event loop, i.e. adding watchers will
		929	have no effect on the set of file descriptors being watched, or the time
		930	waited. Use an C<ev_async> watcher to wake up C<ev_loop> when you want it
		931	to take note of any changes you made.
		932
		933	In theory, threads executing C<ev_loop> will be async-cancel safe between
		934	invocations of C<release> and C<acquire>.
		935
		936	See also the locking example in the C<THREADS> section later in this
		937	document.
		938
		939	=item ev_set_userdata (loop, void *data)
		940
		941	=item ev_userdata (loop)
		942
		943	Set and retrieve a single C<void *> associated with a loop. When
		944	C<ev_set_userdata> has never been called, then C<ev_userdata> returns
		945	C<0.>
		946
		947	These two functions can be used to associate arbitrary data with a loop,
		948	and are intended solely for the C<invoke_pending_cb>, C<release> and
		949	C<acquire> callbacks described above, but of course can be (ab-)used for
		950	any other purpose as well.
714		951
715	=item ev_loop_verify (loop)	952	=item ev_loop_verify (loop)
716		953
717	This function only does something when C<EV_VERIFY> support has been	954	This function only does something when C<EV_VERIFY> support has been
718	compiled in. It tries to go through all internal structures and checks	955	compiled in, which is the default for non-minimal builds. It tries to go
719	them for validity. If anything is found to be inconsistent, it will print	956	through all internal structures and checks them for validity. If anything
720	an error message to standard error and call C<abort ()>.	957	is found to be inconsistent, it will print an error message to standard
		958	error and call C<abort ()>.
721		959
722	This can be used to catch bugs inside libev itself: under normal	960	This can be used to catch bugs inside libev itself: under normal
723	circumstances, this function will never abort as of course libev keeps its	961	circumstances, this function will never abort as of course libev keeps its
724	data structures consistent.	962	data structures consistent.
725		963
726	=back	964	=back
727		965
728		966
729	=head1 ANATOMY OF A WATCHER	967	=head1 ANATOMY OF A WATCHER
730		968
		969	In the following description, uppercase C<TYPE> in names stands for the
		970	watcher type, e.g. C<ev_TYPE_start> can mean C<ev_timer_start> for timer
		971	watchers and C<ev_io_start> for I/O watchers.
		972
731	A watcher is a structure that you create and register to record your	973	A watcher is a structure that you create and register to record your
732	interest in some event. For instance, if you want to wait for STDIN to	974	interest in some event. For instance, if you want to wait for STDIN to
733	become readable, you would create an C<ev_io> watcher for that:	975	become readable, you would create an C<ev_io> watcher for that:
734		976
735	static void my_cb (struct ev_loop *loop, struct ev_io *w, int revents)	977	static void my_cb (struct ev_loop *loop, ev_io *w, int revents)
736	{	978	{
737	ev_io_stop (w);	979	ev_io_stop (w);
738	ev_unloop (loop, EVUNLOOP_ALL);	980	ev_unloop (loop, EVUNLOOP_ALL);
739	}	981	}
740		982
741	struct ev_loop *loop = ev_default_loop (0);	983	struct ev_loop *loop = ev_default_loop (0);
		984
742	struct ev_io stdin_watcher;	985	ev_io stdin_watcher;
		986
743	ev_init (&stdin_watcher, my_cb);	987	ev_init (&stdin_watcher, my_cb);
744	ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ);	988	ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ);
745	ev_io_start (loop, &stdin_watcher);	989	ev_io_start (loop, &stdin_watcher);
		990
746	ev_loop (loop, 0);	991	ev_loop (loop, 0);
747		992
748	As you can see, you are responsible for allocating the memory for your	993	As you can see, you are responsible for allocating the memory for your
749	watcher structures (and it is usually a bad idea to do this on the stack,	994	watcher structures (and it is I<usually> a bad idea to do this on the
750	although this can sometimes be quite valid).	995	stack).
		996
		997	Each watcher has an associated watcher structure (called C<struct ev_TYPE>
		998	or simply C<ev_TYPE>, as typedefs are provided for all watcher structs).
751		999
752	Each watcher structure must be initialised by a call to C<ev_init	1000	Each watcher structure must be initialised by a call to C<ev_init
753	(watcher *, callback)>, which expects a callback to be provided. This	1001	(watcher *, callback)>, which expects a callback to be provided. This
754	callback gets invoked each time the event occurs (or, in the case of I/O	1002	callback gets invoked each time the event occurs (or, in the case of I/O
755	watchers, each time the event loop detects that the file descriptor given	1003	watchers, each time the event loop detects that the file descriptor given
756	is readable and/or writable).	1004	is readable and/or writable).
757		1005
758	Each watcher type has its own C<< ev_<type>_set (watcher *, ...) >> macro	1006	Each watcher type further has its own C<< ev_TYPE_set (watcher *, ...) >>
759	with arguments specific to this watcher type. There is also a macro	1007	macro to configure it, with arguments specific to the watcher type. There
760	to combine initialisation and setting in one call: C<< ev_<type>_init	1008	is also a macro to combine initialisation and setting in one call: C<<
761	(watcher *, callback, ...) >>.	1009	ev_TYPE_init (watcher *, callback, ...) >>.
762		1010
763	To make the watcher actually watch out for events, you have to start it	1011	To make the watcher actually watch out for events, you have to start it
764	with a watcher-specific start function (C<< ev_<type>_start (loop, watcher	1012	with a watcher-specific start function (C<< ev_TYPE_start (loop, watcher
765	*) >>), and you can stop watching for events at any time by calling the	1013	*) >>), and you can stop watching for events at any time by calling the
766	corresponding stop function (C<< ev_<type>_stop (loop, watcher *) >>.	1014	corresponding stop function (C<< ev_TYPE_stop (loop, watcher *) >>.
767		1015
768	As long as your watcher is active (has been started but not stopped) you	1016	As long as your watcher is active (has been started but not stopped) you
769	must not touch the values stored in it. Most specifically you must never	1017	must not touch the values stored in it. Most specifically you must never
770	reinitialise it or call its C<set> macro.	1018	reinitialise it or call its C<ev_TYPE_set> macro.
771		1019
772	Each and every callback receives the event loop pointer as first, the	1020	Each and every callback receives the event loop pointer as first, the
773	registered watcher structure as second, and a bitset of received events as	1021	registered watcher structure as second, and a bitset of received events as
774	third argument.	1022	third argument.
775		1023
…		…
833		1081
834	=item C<EV_ASYNC>	1082	=item C<EV_ASYNC>
835		1083
836	The given async watcher has been asynchronously notified (see C<ev_async>).	1084	The given async watcher has been asynchronously notified (see C<ev_async>).
837		1085
		1086	=item C<EV_CUSTOM>
		1087
		1088	Not ever sent (or otherwise used) by libev itself, but can be freely used
		1089	by libev users to signal watchers (e.g. via C<ev_feed_event>).
		1090
838	=item C<EV_ERROR>	1091	=item C<EV_ERROR>
839		1092
840	An unspecified error has occurred, the watcher has been stopped. This might	1093	An unspecified error has occurred, the watcher has been stopped. This might
841	happen because the watcher could not be properly started because libev	1094	happen because the watcher could not be properly started because libev
842	ran out of memory, a file descriptor was found to be closed or any other	1095	ran out of memory, a file descriptor was found to be closed or any other
		1096	problem. Libev considers these application bugs.
		1097
843	problem. You best act on it by reporting the problem and somehow coping	1098	You best act on it by reporting the problem and somehow coping with the
844	with the watcher being stopped.	1099	watcher being stopped. Note that well-written programs should not receive
		1100	an error ever, so when your watcher receives it, this usually indicates a
		1101	bug in your program.
845		1102
846	Libev will usually signal a few "dummy" events together with an error,	1103	Libev will usually signal a few "dummy" events together with an error, for
847	for example it might indicate that a fd is readable or writable, and if	1104	example it might indicate that a fd is readable or writable, and if your
848	your callbacks is well-written it can just attempt the operation and cope	1105	callbacks is well-written it can just attempt the operation and cope with
849	with the error from read() or write(). This will not work in multi-threaded	1106	the error from read() or write(). This will not work in multi-threaded
850	programs, though, so beware.	1107	programs, though, as the fd could already be closed and reused for another
		1108	thing, so beware.
851		1109
852	=back	1110	=back
853		1111
854	=head2 GENERIC WATCHER FUNCTIONS	1112	=head2 GENERIC WATCHER FUNCTIONS
855
856	In the following description, C<TYPE> stands for the watcher type,
857	e.g. C<timer> for C<ev_timer> watchers and C<io> for C<ev_io> watchers.
858		1113
859	=over 4	1114	=over 4
860		1115
861	=item C<ev_init> (ev_TYPE *watcher, callback)	1116	=item C<ev_init> (ev_TYPE *watcher, callback)
862		1117
…		…
868	which rolls both calls into one.	1123	which rolls both calls into one.
869		1124
870	You can reinitialise a watcher at any time as long as it has been stopped	1125	You can reinitialise a watcher at any time as long as it has been stopped
871	(or never started) and there are no pending events outstanding.	1126	(or never started) and there are no pending events outstanding.
872		1127
873	The callback is always of type C<void (*)(ev_loop *loop, ev_TYPE *watcher,	1128	The callback is always of type C<void (*)(struct ev_loop *loop, ev_TYPE *watcher,
874	int revents)>.	1129	int revents)>.
875		1130
		1131	Example: Initialise an C<ev_io> watcher in two steps.
		1132
		1133	ev_io w;
		1134	ev_init (&w, my_cb);
		1135	ev_io_set (&w, STDIN_FILENO, EV_READ);
		1136
876	=item C<ev_TYPE_set> (ev_TYPE *, [args])	1137	=item C<ev_TYPE_set> (ev_TYPE *watcher, [args])
877		1138
878	This macro initialises the type-specific parts of a watcher. You need to	1139	This macro initialises the type-specific parts of a watcher. You need to
879	call C<ev_init> at least once before you call this macro, but you can	1140	call C<ev_init> at least once before you call this macro, but you can
880	call C<ev_TYPE_set> any number of times. You must not, however, call this	1141	call C<ev_TYPE_set> any number of times. You must not, however, call this
881	macro on a watcher that is active (it can be pending, however, which is a	1142	macro on a watcher that is active (it can be pending, however, which is a
882	difference to the C<ev_init> macro).	1143	difference to the C<ev_init> macro).
883		1144
884	Although some watcher types do not have type-specific arguments	1145	Although some watcher types do not have type-specific arguments
885	(e.g. C<ev_prepare>) you still need to call its C<set> macro.	1146	(e.g. C<ev_prepare>) you still need to call its C<set> macro.
886		1147
		1148	See C<ev_init>, above, for an example.
		1149
887	=item C<ev_TYPE_init> (ev_TYPE *watcher, callback, [args])	1150	=item C<ev_TYPE_init> (ev_TYPE *watcher, callback, [args])
888		1151
889	This convenience macro rolls both C<ev_init> and C<ev_TYPE_set> macro	1152	This convenience macro rolls both C<ev_init> and C<ev_TYPE_set> macro
890	calls into a single call. This is the most convenient method to initialise	1153	calls into a single call. This is the most convenient method to initialise
891	a watcher. The same limitations apply, of course.	1154	a watcher. The same limitations apply, of course.
892		1155
		1156	Example: Initialise and set an C<ev_io> watcher in one step.
		1157
		1158	ev_io_init (&w, my_cb, STDIN_FILENO, EV_READ);
		1159
893	=item C<ev_TYPE_start> (loop *, ev_TYPE *watcher)	1160	=item C<ev_TYPE_start> (loop, ev_TYPE *watcher)
894		1161
895	Starts (activates) the given watcher. Only active watchers will receive	1162	Starts (activates) the given watcher. Only active watchers will receive
896	events. If the watcher is already active nothing will happen.	1163	events. If the watcher is already active nothing will happen.
897		1164
		1165	Example: Start the C<ev_io> watcher that is being abused as example in this
		1166	whole section.
		1167
		1168	ev_io_start (EV_DEFAULT_UC, &w);
		1169
898	=item C<ev_TYPE_stop> (loop *, ev_TYPE *watcher)	1170	=item C<ev_TYPE_stop> (loop, ev_TYPE *watcher)
899		1171
900	Stops the given watcher again (if active) and clears the pending	1172	Stops the given watcher if active, and clears the pending status (whether
		1173	the watcher was active or not).
		1174
901	status. It is possible that stopped watchers are pending (for example,	1175	It is possible that stopped watchers are pending - for example,
902	non-repeating timers are being stopped when they become pending), but	1176	non-repeating timers are being stopped when they become pending - but
903	C<ev_TYPE_stop> ensures that the watcher is neither active nor pending. If	1177	calling C<ev_TYPE_stop> ensures that the watcher is neither active nor
904	you want to free or reuse the memory used by the watcher it is therefore a	1178	pending. If you want to free or reuse the memory used by the watcher it is
905	good idea to always call its C<ev_TYPE_stop> function.	1179	therefore a good idea to always call its C<ev_TYPE_stop> function.
906		1180
907	=item bool ev_is_active (ev_TYPE *watcher)	1181	=item bool ev_is_active (ev_TYPE *watcher)
908		1182
909	Returns a true value iff the watcher is active (i.e. it has been started	1183	Returns a true value iff the watcher is active (i.e. it has been started
910	and not yet been stopped). As long as a watcher is active you must not modify	1184	and not yet been stopped). As long as a watcher is active you must not modify
…		…
926	=item ev_cb_set (ev_TYPE *watcher, callback)	1200	=item ev_cb_set (ev_TYPE *watcher, callback)
927		1201
928	Change the callback. You can change the callback at virtually any time	1202	Change the callback. You can change the callback at virtually any time
929	(modulo threads).	1203	(modulo threads).
930		1204
931	=item ev_set_priority (ev_TYPE *watcher, priority)	1205	=item ev_set_priority (ev_TYPE *watcher, int priority)
932		1206
933	=item int ev_priority (ev_TYPE *watcher)	1207	=item int ev_priority (ev_TYPE *watcher)
934		1208
935	Set and query the priority of the watcher. The priority is a small	1209	Set and query the priority of the watcher. The priority is a small
936	integer between C<EV_MAXPRI> (default: C<2>) and C<EV_MINPRI>	1210	integer between C<EV_MAXPRI> (default: C<2>) and C<EV_MINPRI>
937	(default: C<-2>). Pending watchers with higher priority will be invoked	1211	(default: C<-2>). Pending watchers with higher priority will be invoked
938	before watchers with lower priority, but priority will not keep watchers	1212	before watchers with lower priority, but priority will not keep watchers
939	from being executed (except for C<ev_idle> watchers).	1213	from being executed (except for C<ev_idle> watchers).
940		1214
941	This means that priorities are I<only> used for ordering callback
942	invocation after new events have been received. This is useful, for
943	example, to reduce latency after idling, or more often, to bind two
944	watchers on the same event and make sure one is called first.
945
946	If you need to suppress invocation when higher priority events are pending	1215	If you need to suppress invocation when higher priority events are pending
947	you need to look at C<ev_idle> watchers, which provide this functionality.	1216	you need to look at C<ev_idle> watchers, which provide this functionality.
948		1217
949	You I<must not> change the priority of a watcher as long as it is active or	1218	You I<must not> change the priority of a watcher as long as it is active or
950	pending.	1219	pending.
951		1220
		1221	Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is
		1222	fine, as long as you do not mind that the priority value you query might
		1223	or might not have been clamped to the valid range.
		1224
952	The default priority used by watchers when no priority has been set is	1225	The default priority used by watchers when no priority has been set is
953	always C<0>, which is supposed to not be too high and not be too low :).	1226	always C<0>, which is supposed to not be too high and not be too low :).
954		1227
955	Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is	1228	See L<WATCHER PRIORITY MODELS>, below, for a more thorough treatment of
956	fine, as long as you do not mind that the priority value you query might	1229	priorities.
957	or might not have been adjusted to be within valid range.
958		1230
959	=item ev_invoke (loop, ev_TYPE *watcher, int revents)	1231	=item ev_invoke (loop, ev_TYPE *watcher, int revents)
960		1232
961	Invoke the C<watcher> with the given C<loop> and C<revents>. Neither	1233	Invoke the C<watcher> with the given C<loop> and C<revents>. Neither
962	C<loop> nor C<revents> need to be valid as long as the watcher callback	1234	C<loop> nor C<revents> need to be valid as long as the watcher callback
963	can deal with that fact.	1235	can deal with that fact, as both are simply passed through to the
		1236	callback.
964		1237
965	=item int ev_clear_pending (loop, ev_TYPE *watcher)	1238	=item int ev_clear_pending (loop, ev_TYPE *watcher)
966		1239
967	If the watcher is pending, this function returns clears its pending status	1240	If the watcher is pending, this function clears its pending status and
968	and returns its C<revents> bitset (as if its callback was invoked). If the	1241	returns its C<revents> bitset (as if its callback was invoked). If the
969	watcher isn't pending it does nothing and returns C<0>.	1242	watcher isn't pending it does nothing and returns C<0>.
970		1243
		1244	Sometimes it can be useful to "poll" a watcher instead of waiting for its
		1245	callback to be invoked, which can be accomplished with this function.
		1246
		1247	=item ev_feed_event (loop, ev_TYPE *watcher, int revents)
		1248
		1249	Feeds the given event set into the event loop, as if the specified event
		1250	had happened for the specified watcher (which must be a pointer to an
		1251	initialised but not necessarily started event watcher). Obviously you must
		1252	not free the watcher as long as it has pending events.
		1253
		1254	Stopping the watcher, letting libev invoke it, or calling
		1255	C<ev_clear_pending> will clear the pending event, even if the watcher was
		1256	not started in the first place.
		1257
		1258	See also C<ev_feed_fd_event> and C<ev_feed_signal_event> for related
		1259	functions that do not need a watcher.
		1260
971	=back	1261	=back
972		1262
973		1263
974	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER	1264	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER
975		1265
976	Each watcher has, by default, a member C<void *data> that you can change	1266	Each watcher has, by default, a member C<void *data> that you can change
977	and read at any time, libev will completely ignore it. This can be used	1267	and read at any time: libev will completely ignore it. This can be used
978	to associate arbitrary data with your watcher. If you need more data and	1268	to associate arbitrary data with your watcher. If you need more data and
979	don't want to allocate memory and store a pointer to it in that data	1269	don't want to allocate memory and store a pointer to it in that data
980	member, you can also "subclass" the watcher type and provide your own	1270	member, you can also "subclass" the watcher type and provide your own
981	data:	1271	data:
982		1272
983	struct my_io	1273	struct my_io
984	{	1274	{
985	struct ev_io io;	1275	ev_io io;
986	int otherfd;	1276	int otherfd;
987	void *somedata;	1277	void *somedata;
988	struct whatever *mostinteresting;	1278	struct whatever *mostinteresting;
989	}	1279	};
		1280
		1281	...
		1282	struct my_io w;
		1283	ev_io_init (&w.io, my_cb, fd, EV_READ);
990		1284
991	And since your callback will be called with a pointer to the watcher, you	1285	And since your callback will be called with a pointer to the watcher, you
992	can cast it back to your own type:	1286	can cast it back to your own type:
993		1287
994	static void my_cb (struct ev_loop *loop, struct ev_io *w_, int revents)	1288	static void my_cb (struct ev_loop *loop, ev_io *w_, int revents)
995	{	1289	{
996	struct my_io *w = (struct my_io *)w_;	1290	struct my_io *w = (struct my_io *)w_;
997	...	1291	...
998	}	1292	}
999		1293
1000	More interesting and less C-conformant ways of casting your callback type	1294	More interesting and less C-conformant ways of casting your callback type
1001	instead have been omitted.	1295	instead have been omitted.
1002		1296
1003	Another common scenario is having some data structure with multiple	1297	Another common scenario is to use some data structure with multiple
1004	watchers:	1298	embedded watchers:
1005		1299
1006	struct my_biggy	1300	struct my_biggy
1007	{	1301	{
1008	int some_data;	1302	int some_data;
1009	ev_timer t1;	1303	ev_timer t1;
1010	ev_timer t2;	1304	ev_timer t2;
1011	}	1305	}
1012		1306
1013	In this case getting the pointer to C<my_biggy> is a bit more complicated,	1307	In this case getting the pointer to C<my_biggy> is a bit more
1014	you need to use C<offsetof>:	1308	complicated: Either you store the address of your C<my_biggy> struct
		1309	in the C<data> member of the watcher (for woozies), or you need to use
		1310	some pointer arithmetic using C<offsetof> inside your watchers (for real
		1311	programmers):
1015		1312
1016	#include <stddef.h>	1313	#include <stddef.h>
1017		1314
1018	static void	1315	static void
1019	t1_cb (EV_P_ struct ev_timer *w, int revents)	1316	t1_cb (EV_P_ ev_timer *w, int revents)
1020	{	1317	{
1021	struct my_biggy big = (struct my_biggy *	1318	struct my_biggy big = (struct my_biggy *)
1022	(((char *)w) - offsetof (struct my_biggy, t1));	1319	(((char *)w) - offsetof (struct my_biggy, t1));
1023	}	1320	}
1024		1321
1025	static void	1322	static void
1026	t2_cb (EV_P_ struct ev_timer *w, int revents)	1323	t2_cb (EV_P_ ev_timer *w, int revents)
1027	{	1324	{
1028	struct my_biggy big = (struct my_biggy *	1325	struct my_biggy big = (struct my_biggy *)
1029	(((char *)w) - offsetof (struct my_biggy, t2));	1326	(((char *)w) - offsetof (struct my_biggy, t2));
1030	}	1327	}
		1328
		1329	=head2 WATCHER PRIORITY MODELS
		1330
		1331	Many event loops support I<watcher priorities>, which are usually small
		1332	integers that influence the ordering of event callback invocation
		1333	between watchers in some way, all else being equal.
		1334
		1335	In libev, Watcher priorities can be set using C<ev_set_priority>. See its
		1336	description for the more technical details such as the actual priority
		1337	range.
		1338
		1339	There are two common ways how these these priorities are being interpreted
		1340	by event loops:
		1341
		1342	In the more common lock-out model, higher priorities "lock out" invocation
		1343	of lower priority watchers, which means as long as higher priority
		1344	watchers receive events, lower priority watchers are not being invoked.
		1345
		1346	The less common only-for-ordering model uses priorities solely to order
		1347	callback invocation within a single event loop iteration: Higher priority
		1348	watchers are invoked before lower priority ones, but they all get invoked
		1349	before polling for new events.
		1350
		1351	Libev uses the second (only-for-ordering) model for all its watchers
		1352	except for idle watchers (which use the lock-out model).
		1353
		1354	The rationale behind this is that implementing the lock-out model for
		1355	watchers is not well supported by most kernel interfaces, and most event
		1356	libraries will just poll for the same events again and again as long as
		1357	their callbacks have not been executed, which is very inefficient in the
		1358	common case of one high-priority watcher locking out a mass of lower
		1359	priority ones.
		1360
		1361	Static (ordering) priorities are most useful when you have two or more
		1362	watchers handling the same resource: a typical usage example is having an
		1363	C<ev_io> watcher to receive data, and an associated C<ev_timer> to handle
		1364	timeouts. Under load, data might be received while the program handles
		1365	other jobs, but since timers normally get invoked first, the timeout
		1366	handler will be executed before checking for data. In that case, giving
		1367	the timer a lower priority than the I/O watcher ensures that I/O will be
		1368	handled first even under adverse conditions (which is usually, but not
		1369	always, what you want).
		1370
		1371	Since idle watchers use the "lock-out" model, meaning that idle watchers
		1372	will only be executed when no same or higher priority watchers have
		1373	received events, they can be used to implement the "lock-out" model when
		1374	required.
		1375
		1376	For example, to emulate how many other event libraries handle priorities,
		1377	you can associate an C<ev_idle> watcher to each such watcher, and in
		1378	the normal watcher callback, you just start the idle watcher. The real
		1379	processing is done in the idle watcher callback. This causes libev to
		1380	continously poll and process kernel event data for the watcher, but when
		1381	the lock-out case is known to be rare (which in turn is rare :), this is
		1382	workable.
		1383
		1384	Usually, however, the lock-out model implemented that way will perform
		1385	miserably under the type of load it was designed to handle. In that case,
		1386	it might be preferable to stop the real watcher before starting the
		1387	idle watcher, so the kernel will not have to process the event in case
		1388	the actual processing will be delayed for considerable time.
		1389
		1390	Here is an example of an I/O watcher that should run at a strictly lower
		1391	priority than the default, and which should only process data when no
		1392	other events are pending:
		1393
		1394	ev_idle idle; // actual processing watcher
		1395	ev_io io; // actual event watcher
		1396
		1397	static void
		1398	io_cb (EV_P_ ev_io *w, int revents)
		1399	{
		1400	// stop the I/O watcher, we received the event, but
		1401	// are not yet ready to handle it.
		1402	ev_io_stop (EV_A_ w);
		1403
		1404	// start the idle watcher to ahndle the actual event.
		1405	// it will not be executed as long as other watchers
		1406	// with the default priority are receiving events.
		1407	ev_idle_start (EV_A_ &idle);
		1408	}
		1409
		1410	static void
		1411	idle_cb (EV_P_ ev_idle *w, int revents)
		1412	{
		1413	// actual processing
		1414	read (STDIN_FILENO, ...);
		1415
		1416	// have to start the I/O watcher again, as
		1417	// we have handled the event
		1418	ev_io_start (EV_P_ &io);
		1419	}
		1420
		1421	// initialisation
		1422	ev_idle_init (&idle, idle_cb);
		1423	ev_io_init (&io, io_cb, STDIN_FILENO, EV_READ);
		1424	ev_io_start (EV_DEFAULT_ &io);
		1425
		1426	In the "real" world, it might also be beneficial to start a timer, so that
		1427	low-priority connections can not be locked out forever under load. This
		1428	enables your program to keep a lower latency for important connections
		1429	during short periods of high load, while not completely locking out less
		1430	important ones.
1031		1431
1032		1432
1033	=head1 WATCHER TYPES	1433	=head1 WATCHER TYPES
1034		1434
1035	This section describes each watcher in detail, but will not repeat	1435	This section describes each watcher in detail, but will not repeat
…		…
1059	In general you can register as many read and/or write event watchers per	1459	In general you can register as many read and/or write event watchers per
1060	fd as you want (as long as you don't confuse yourself). Setting all file	1460	fd as you want (as long as you don't confuse yourself). Setting all file
1061	descriptors to non-blocking mode is also usually a good idea (but not	1461	descriptors to non-blocking mode is also usually a good idea (but not
1062	required if you know what you are doing).	1462	required if you know what you are doing).
1063		1463
1064	If you must do this, then force the use of a known-to-be-good backend	1464	If you cannot use non-blocking mode, then force the use of a
1065	(at the time of this writing, this includes only C<EVBACKEND_SELECT> and	1465	known-to-be-good backend (at the time of this writing, this includes only
1066	C<EVBACKEND_POLL>).	1466	C<EVBACKEND_SELECT> and C<EVBACKEND_POLL>). The same applies to file
		1467	descriptors for which non-blocking operation makes no sense (such as
		1468	files) - libev doesn't guarentee any specific behaviour in that case.
1067		1469
1068	Another thing you have to watch out for is that it is quite easy to	1470	Another thing you have to watch out for is that it is quite easy to
1069	receive "spurious" readiness notifications, that is your callback might	1471	receive "spurious" readiness notifications, that is your callback might
1070	be called with C<EV_READ> but a subsequent C<read>(2) will actually block	1472	be called with C<EV_READ> but a subsequent C<read>(2) will actually block
1071	because there is no data. Not only are some backends known to create a	1473	because there is no data. Not only are some backends known to create a
1072	lot of those (for example Solaris ports), it is very easy to get into	1474	lot of those (for example Solaris ports), it is very easy to get into
1073	this situation even with a relatively standard program structure. Thus	1475	this situation even with a relatively standard program structure. Thus
1074	it is best to always use non-blocking I/O: An extra C<read>(2) returning	1476	it is best to always use non-blocking I/O: An extra C<read>(2) returning
1075	C<EAGAIN> is far preferable to a program hanging until some data arrives.	1477	C<EAGAIN> is far preferable to a program hanging until some data arrives.
1076		1478
1077	If you cannot run the fd in non-blocking mode (for example you should not	1479	If you cannot run the fd in non-blocking mode (for example you should
1078	play around with an Xlib connection), then you have to separately re-test	1480	not play around with an Xlib connection), then you have to separately
1079	whether a file descriptor is really ready with a known-to-be good interface	1481	re-test whether a file descriptor is really ready with a known-to-be good
1080	such as poll (fortunately in our Xlib example, Xlib already does this on	1482	interface such as poll (fortunately in our Xlib example, Xlib already
1081	its own, so its quite safe to use).	1483	does this on its own, so its quite safe to use). Some people additionally
		1484	use C<SIGALRM> and an interval timer, just to be sure you won't block
		1485	indefinitely.
		1486
		1487	But really, best use non-blocking mode.
1082		1488
1083	=head3 The special problem of disappearing file descriptors	1489	=head3 The special problem of disappearing file descriptors
1084		1490
1085	Some backends (e.g. kqueue, epoll) need to be told about closing a file	1491	Some backends (e.g. kqueue, epoll) need to be told about closing a file
1086	descriptor (either by calling C<close> explicitly or by any other means,	1492	descriptor (either due to calling C<close> explicitly or any other means,
1087	such as C<dup>). The reason is that you register interest in some file	1493	such as C<dup2>). The reason is that you register interest in some file
1088	descriptor, but when it goes away, the operating system will silently drop	1494	descriptor, but when it goes away, the operating system will silently drop
1089	this interest. If another file descriptor with the same number then is	1495	this interest. If another file descriptor with the same number then is
1090	registered with libev, there is no efficient way to see that this is, in	1496	registered with libev, there is no efficient way to see that this is, in
1091	fact, a different file descriptor.	1497	fact, a different file descriptor.
1092		1498
…		…
1123	enable C<EVFLAG_FORKCHECK>, or resort to C<EVBACKEND_SELECT> or	1529	enable C<EVFLAG_FORKCHECK>, or resort to C<EVBACKEND_SELECT> or
1124	C<EVBACKEND_POLL>.	1530	C<EVBACKEND_POLL>.
1125		1531
1126	=head3 The special problem of SIGPIPE	1532	=head3 The special problem of SIGPIPE
1127		1533
1128	While not really specific to libev, it is easy to forget about SIGPIPE:	1534	While not really specific to libev, it is easy to forget about C<SIGPIPE>:
1129	when reading from a pipe whose other end has been closed, your program	1535	when writing to a pipe whose other end has been closed, your program gets
1130	gets send a SIGPIPE, which, by default, aborts your program. For most	1536	sent a SIGPIPE, which, by default, aborts your program. For most programs
1131	programs this is sensible behaviour, for daemons, this is usually	1537	this is sensible behaviour, for daemons, this is usually undesirable.
1132	undesirable.
1133		1538
1134	So when you encounter spurious, unexplained daemon exits, make sure you	1539	So when you encounter spurious, unexplained daemon exits, make sure you
1135	ignore SIGPIPE (and maybe make sure you log the exit status of your daemon	1540	ignore SIGPIPE (and maybe make sure you log the exit status of your daemon
1136	somewhere, as that would have given you a big clue).	1541	somewhere, as that would have given you a big clue).
1137		1542
		1543	=head3 The special problem of accept()ing when you can't
		1544
		1545	Many implementations of the POSIX C<accept> function (for example,
		1546	found in port-2004 Linux) have the peculiar behaviour of not removing a
		1547	connection from the pending queue in all error cases.
		1548
		1549	For example, larger servers often run out of file descriptors (because
		1550	of resource limits), causing C<accept> to fail with C<ENFILE> but not
		1551	rejecting the connection, leading to libev signalling readiness on
		1552	the next iteration again (the connection still exists after all), and
		1553	typically causing the program to loop at 100% CPU usage.
		1554
		1555	Unfortunately, the set of errors that cause this issue differs between
		1556	operating systems, there is usually little the app can do to remedy the
		1557	situation, and no known thread-safe method of removing the connection to
		1558	cope with overload is known (to me).
		1559
		1560	One of the easiest ways to handle this situation is to just ignore it
		1561	- when the program encounters an overload, it will just loop until the
		1562	situation is over. While this is a form of busy waiting, no OS offers an
		1563	event-based way to handle this situation, so it's the best one can do.
		1564
		1565	A better way to handle the situation is to log any errors other than
		1566	C<EAGAIN> and C<EWOULDBLOCK>, making sure not to flood the log with such
		1567	messages, and continue as usual, which at least gives the user an idea of
		1568	what could be wrong ("raise the ulimit!"). For extra points one could stop
		1569	the C<ev_io> watcher on the listening fd "for a while", which reduces CPU
		1570	usage.
		1571
		1572	If your program is single-threaded, then you could also keep a dummy file
		1573	descriptor for overload situations (e.g. by opening F</dev/null>), and
		1574	when you run into C<ENFILE> or C<EMFILE>, close it, run C<accept>,
		1575	close that fd, and create a new dummy fd. This will gracefully refuse
		1576	clients under typical overload conditions.
		1577
		1578	The last way to handle it is to simply log the error and C<exit>, as
		1579	is often done with C<malloc> failures, but this results in an easy
		1580	opportunity for a DoS attack.
1138		1581
1139	=head3 Watcher-Specific Functions	1582	=head3 Watcher-Specific Functions
1140		1583
1141	=over 4	1584	=over 4
1142		1585
1143	=item ev_io_init (ev_io *, callback, int fd, int events)	1586	=item ev_io_init (ev_io *, callback, int fd, int events)
1144		1587
1145	=item ev_io_set (ev_io *, int fd, int events)	1588	=item ev_io_set (ev_io *, int fd, int events)
1146		1589
1147	Configures an C<ev_io> watcher. The C<fd> is the file descriptor to	1590	Configures an C<ev_io> watcher. The C<fd> is the file descriptor to
1148	receive events for and events is either C<EV_READ>, C<EV_WRITE> or	1591	receive events for and C<events> is either C<EV_READ>, C<EV_WRITE> or
1149	C<EV_READ | EV_WRITE> to receive the given events.	1592	C<EV_READ | EV_WRITE>, to express the desire to receive the given events.
1150		1593
1151	=item int fd [read-only]	1594	=item int fd [read-only]
1152		1595
1153	The file descriptor being watched.	1596	The file descriptor being watched.
1154		1597
…		…
1163	Example: Call C<stdin_readable_cb> when STDIN_FILENO has become, well	1606	Example: Call C<stdin_readable_cb> when STDIN_FILENO has become, well
1164	readable, but only once. Since it is likely line-buffered, you could	1607	readable, but only once. Since it is likely line-buffered, you could
1165	attempt to read a whole line in the callback.	1608	attempt to read a whole line in the callback.
1166		1609
1167	static void	1610	static void
1168	stdin_readable_cb (struct ev_loop *loop, struct ev_io *w, int revents)	1611	stdin_readable_cb (struct ev_loop *loop, ev_io *w, int revents)
1169	{	1612	{
1170	ev_io_stop (loop, w);	1613	ev_io_stop (loop, w);
1171	.. read from stdin here (or from w->fd) and haqndle any I/O errors	1614	.. read from stdin here (or from w->fd) and handle any I/O errors
1172	}	1615	}
1173		1616
1174	...	1617	...
1175	struct ev_loop *loop = ev_default_init (0);	1618	struct ev_loop *loop = ev_default_init (0);
1176	struct ev_io stdin_readable;	1619	ev_io stdin_readable;
1177	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);	1620	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);
1178	ev_io_start (loop, &stdin_readable);	1621	ev_io_start (loop, &stdin_readable);
1179	ev_loop (loop, 0);	1622	ev_loop (loop, 0);
1180		1623
1181		1624
…		…
1184	Timer watchers are simple relative timers that generate an event after a	1627	Timer watchers are simple relative timers that generate an event after a
1185	given time, and optionally repeating in regular intervals after that.	1628	given time, and optionally repeating in regular intervals after that.
1186		1629
1187	The timers are based on real time, that is, if you register an event that	1630	The timers are based on real time, that is, if you register an event that
1188	times out after an hour and you reset your system clock to January last	1631	times out after an hour and you reset your system clock to January last
1189	year, it will still time out after (roughly) and hour. "Roughly" because	1632	year, it will still time out after (roughly) one hour. "Roughly" because
1190	detecting time jumps is hard, and some inaccuracies are unavoidable (the	1633	detecting time jumps is hard, and some inaccuracies are unavoidable (the
1191	monotonic clock option helps a lot here).	1634	monotonic clock option helps a lot here).
		1635
		1636	The callback is guaranteed to be invoked only I<after> its timeout has
		1637	passed (not I<at>, so on systems with very low-resolution clocks this
		1638	might introduce a small delay). If multiple timers become ready during the
		1639	same loop iteration then the ones with earlier time-out values are invoked
		1640	before ones of the same priority with later time-out values (but this is
		1641	no longer true when a callback calls C<ev_loop> recursively).
		1642
		1643	=head3 Be smart about timeouts
		1644
		1645	Many real-world problems involve some kind of timeout, usually for error
		1646	recovery. A typical example is an HTTP request - if the other side hangs,
		1647	you want to raise some error after a while.
		1648
		1649	What follows are some ways to handle this problem, from obvious and
		1650	inefficient to smart and efficient.
		1651
		1652	In the following, a 60 second activity timeout is assumed - a timeout that
		1653	gets reset to 60 seconds each time there is activity (e.g. each time some
		1654	data or other life sign was received).
		1655
		1656	=over 4
		1657
		1658	=item 1. Use a timer and stop, reinitialise and start it on activity.
		1659
		1660	This is the most obvious, but not the most simple way: In the beginning,
		1661	start the watcher:
		1662
		1663	ev_timer_init (timer, callback, 60., 0.);
		1664	ev_timer_start (loop, timer);
		1665
		1666	Then, each time there is some activity, C<ev_timer_stop> it, initialise it
		1667	and start it again:
		1668
		1669	ev_timer_stop (loop, timer);
		1670	ev_timer_set (timer, 60., 0.);
		1671	ev_timer_start (loop, timer);
		1672
		1673	This is relatively simple to implement, but means that each time there is
		1674	some activity, libev will first have to remove the timer from its internal
		1675	data structure and then add it again. Libev tries to be fast, but it's
		1676	still not a constant-time operation.
		1677
		1678	=item 2. Use a timer and re-start it with C<ev_timer_again> inactivity.
		1679
		1680	This is the easiest way, and involves using C<ev_timer_again> instead of
		1681	C<ev_timer_start>.
		1682
		1683	To implement this, configure an C<ev_timer> with a C<repeat> value
		1684	of C<60> and then call C<ev_timer_again> at start and each time you
		1685	successfully read or write some data. If you go into an idle state where
		1686	you do not expect data to travel on the socket, you can C<ev_timer_stop>
		1687	the timer, and C<ev_timer_again> will automatically restart it if need be.
		1688
		1689	That means you can ignore both the C<ev_timer_start> function and the
		1690	C<after> argument to C<ev_timer_set>, and only ever use the C<repeat>
		1691	member and C<ev_timer_again>.
		1692
		1693	At start:
		1694
		1695	ev_init (timer, callback);
		1696	timer->repeat = 60.;
		1697	ev_timer_again (loop, timer);
		1698
		1699	Each time there is some activity:
		1700
		1701	ev_timer_again (loop, timer);
		1702
		1703	It is even possible to change the time-out on the fly, regardless of
		1704	whether the watcher is active or not:
		1705
		1706	timer->repeat = 30.;
		1707	ev_timer_again (loop, timer);
		1708
		1709	This is slightly more efficient then stopping/starting the timer each time
		1710	you want to modify its timeout value, as libev does not have to completely
		1711	remove and re-insert the timer from/into its internal data structure.
		1712
		1713	It is, however, even simpler than the "obvious" way to do it.
		1714
		1715	=item 3. Let the timer time out, but then re-arm it as required.
		1716
		1717	This method is more tricky, but usually most efficient: Most timeouts are
		1718	relatively long compared to the intervals between other activity - in
		1719	our example, within 60 seconds, there are usually many I/O events with
		1720	associated activity resets.
		1721
		1722	In this case, it would be more efficient to leave the C<ev_timer> alone,
		1723	but remember the time of last activity, and check for a real timeout only
		1724	within the callback:
		1725
		1726	ev_tstamp last_activity; // time of last activity
		1727
		1728	static void
		1729	callback (EV_P_ ev_timer *w, int revents)
		1730	{
		1731	ev_tstamp now = ev_now (EV_A);
		1732	ev_tstamp timeout = last_activity + 60.;
		1733
		1734	// if last_activity + 60. is older than now, we did time out
		1735	if (timeout < now)
		1736	{
		1737	// timeout occured, take action
		1738	}
		1739	else
		1740	{
		1741	// callback was invoked, but there was some activity, re-arm
		1742	// the watcher to fire in last_activity + 60, which is
		1743	// guaranteed to be in the future, so "again" is positive:
		1744	w->repeat = timeout - now;
		1745	ev_timer_again (EV_A_ w);
		1746	}
		1747	}
		1748
		1749	To summarise the callback: first calculate the real timeout (defined
		1750	as "60 seconds after the last activity"), then check if that time has
		1751	been reached, which means something I<did>, in fact, time out. Otherwise
		1752	the callback was invoked too early (C<timeout> is in the future), so
		1753	re-schedule the timer to fire at that future time, to see if maybe we have
		1754	a timeout then.
		1755
		1756	Note how C<ev_timer_again> is used, taking advantage of the
		1757	C<ev_timer_again> optimisation when the timer is already running.
		1758
		1759	This scheme causes more callback invocations (about one every 60 seconds
		1760	minus half the average time between activity), but virtually no calls to
		1761	libev to change the timeout.
		1762
		1763	To start the timer, simply initialise the watcher and set C<last_activity>
		1764	to the current time (meaning we just have some activity :), then call the
		1765	callback, which will "do the right thing" and start the timer:
		1766
		1767	ev_init (timer, callback);
		1768	last_activity = ev_now (loop);
		1769	callback (loop, timer, EV_TIMEOUT);
		1770
		1771	And when there is some activity, simply store the current time in
		1772	C<last_activity>, no libev calls at all:
		1773
		1774	last_actiivty = ev_now (loop);
		1775
		1776	This technique is slightly more complex, but in most cases where the
		1777	time-out is unlikely to be triggered, much more efficient.
		1778
		1779	Changing the timeout is trivial as well (if it isn't hard-coded in the
		1780	callback :) - just change the timeout and invoke the callback, which will
		1781	fix things for you.
		1782
		1783	=item 4. Wee, just use a double-linked list for your timeouts.
		1784
		1785	If there is not one request, but many thousands (millions...), all
		1786	employing some kind of timeout with the same timeout value, then one can
		1787	do even better:
		1788
		1789	When starting the timeout, calculate the timeout value and put the timeout
		1790	at the I<end> of the list.
		1791
		1792	Then use an C<ev_timer> to fire when the timeout at the I<beginning> of
		1793	the list is expected to fire (for example, using the technique #3).
		1794
		1795	When there is some activity, remove the timer from the list, recalculate
		1796	the timeout, append it to the end of the list again, and make sure to
		1797	update the C<ev_timer> if it was taken from the beginning of the list.
		1798
		1799	This way, one can manage an unlimited number of timeouts in O(1) time for
		1800	starting, stopping and updating the timers, at the expense of a major
		1801	complication, and having to use a constant timeout. The constant timeout
		1802	ensures that the list stays sorted.
		1803
		1804	=back
		1805
		1806	So which method the best?
		1807
		1808	Method #2 is a simple no-brain-required solution that is adequate in most
		1809	situations. Method #3 requires a bit more thinking, but handles many cases
		1810	better, and isn't very complicated either. In most case, choosing either
		1811	one is fine, with #3 being better in typical situations.
		1812
		1813	Method #1 is almost always a bad idea, and buys you nothing. Method #4 is
		1814	rather complicated, but extremely efficient, something that really pays
		1815	off after the first million or so of active timers, i.e. it's usually
		1816	overkill :)
		1817
		1818	=head3 The special problem of time updates
		1819
		1820	Establishing the current time is a costly operation (it usually takes at
		1821	least two system calls): EV therefore updates its idea of the current
		1822	time only before and after C<ev_loop> collects new events, which causes a
		1823	growing difference between C<ev_now ()> and C<ev_time ()> when handling
		1824	lots of events in one iteration.
1192		1825
1193	The relative timeouts are calculated relative to the C<ev_now ()>	1826	The relative timeouts are calculated relative to the C<ev_now ()>
1194	time. This is usually the right thing as this timestamp refers to the time	1827	time. This is usually the right thing as this timestamp refers to the time
1195	of the event triggering whatever timeout you are modifying/starting. If	1828	of the event triggering whatever timeout you are modifying/starting. If
1196	you suspect event processing to be delayed and you I<need> to base the timeout	1829	you suspect event processing to be delayed and you I<need> to base the
1197	on the current time, use something like this to adjust for this:	1830	timeout on the current time, use something like this to adjust for this:
1198		1831
1199	ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);	1832	ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);
1200		1833
1201	The callback is guaranteed to be invoked only after its timeout has passed,	1834	If the event loop is suspended for a long time, you can also force an
1202	but if multiple timers become ready during the same loop iteration then	1835	update of the time returned by C<ev_now ()> by calling C<ev_now_update
1203	order of execution is undefined.	1836	()>.
		1837
		1838	=head3 The special problems of suspended animation
		1839
		1840	When you leave the server world it is quite customary to hit machines that
		1841	can suspend/hibernate - what happens to the clocks during such a suspend?
		1842
		1843	Some quick tests made with a Linux 2.6.28 indicate that a suspend freezes
		1844	all processes, while the clocks (C<times>, C<CLOCK_MONOTONIC>) continue
		1845	to run until the system is suspended, but they will not advance while the
		1846	system is suspended. That means, on resume, it will be as if the program
		1847	was frozen for a few seconds, but the suspend time will not be counted
		1848	towards C<ev_timer> when a monotonic clock source is used. The real time
		1849	clock advanced as expected, but if it is used as sole clocksource, then a
		1850	long suspend would be detected as a time jump by libev, and timers would
		1851	be adjusted accordingly.
		1852
		1853	I would not be surprised to see different behaviour in different between
		1854	operating systems, OS versions or even different hardware.
		1855
		1856	The other form of suspend (job control, or sending a SIGSTOP) will see a
		1857	time jump in the monotonic clocks and the realtime clock. If the program
		1858	is suspended for a very long time, and monotonic clock sources are in use,
		1859	then you can expect C<ev_timer>s to expire as the full suspension time
		1860	will be counted towards the timers. When no monotonic clock source is in
		1861	use, then libev will again assume a timejump and adjust accordingly.
		1862
		1863	It might be beneficial for this latter case to call C<ev_suspend>
		1864	and C<ev_resume> in code that handles C<SIGTSTP>, to at least get
		1865	deterministic behaviour in this case (you can do nothing against
		1866	C<SIGSTOP>).
1204		1867
1205	=head3 Watcher-Specific Functions and Data Members	1868	=head3 Watcher-Specific Functions and Data Members
1206		1869
1207	=over 4	1870	=over 4
1208		1871
…		…
1232	If the timer is started but non-repeating, stop it (as if it timed out).	1895	If the timer is started but non-repeating, stop it (as if it timed out).
1233		1896
1234	If the timer is repeating, either start it if necessary (with the	1897	If the timer is repeating, either start it if necessary (with the
1235	C<repeat> value), or reset the running timer to the C<repeat> value.	1898	C<repeat> value), or reset the running timer to the C<repeat> value.
1236		1899
1237	This sounds a bit complicated, but here is a useful and typical	1900	This sounds a bit complicated, see L<Be smart about timeouts>, above, for a
1238	example: Imagine you have a TCP connection and you want a so-called idle	1901	usage example.
1239	timeout, that is, you want to be called when there have been, say, 60
1240	seconds of inactivity on the socket. The easiest way to do this is to
1241	configure an C<ev_timer> with a C<repeat> value of C<60> and then call
1242	C<ev_timer_again> each time you successfully read or write some data. If
1243	you go into an idle state where you do not expect data to travel on the
1244	socket, you can C<ev_timer_stop> the timer, and C<ev_timer_again> will
1245	automatically restart it if need be.
1246		1902
1247	That means you can ignore the C<after> value and C<ev_timer_start>	1903	=item ev_tstamp ev_timer_remaining (loop, ev_timer *)
1248	altogether and only ever use the C<repeat> value and C<ev_timer_again>:
1249		1904
1250	ev_timer_init (timer, callback, 0., 5.);	1905	Returns the remaining time until a timer fires. If the timer is active,
1251	ev_timer_again (loop, timer);	1906	then this time is relative to the current event loop time, otherwise it's
1252	...	1907	the timeout value currently configured.
1253	timer->again = 17.;
1254	ev_timer_again (loop, timer);
1255	...
1256	timer->again = 10.;
1257	ev_timer_again (loop, timer);
1258		1908
1259	This is more slightly efficient then stopping/starting the timer each time	1909	That is, after an C<ev_timer_set (w, 5, 7)>, C<ev_timer_remaining> returns
1260	you want to modify its timeout value.	1910	C<5>. When the timer is started and one second passes, C<ev_timer_remaining>
		1911	will return C<4>. When the timer expires and is restarted, it will return
		1912	roughly C<7> (likely slightly less as callback invocation takes some time,
		1913	too), and so on.
1261		1914
1262	=item ev_tstamp repeat [read-write]	1915	=item ev_tstamp repeat [read-write]
1263		1916
1264	The current C<repeat> value. Will be used each time the watcher times out	1917	The current C<repeat> value. Will be used each time the watcher times out
1265	or C<ev_timer_again> is called and determines the next timeout (if any),	1918	or C<ev_timer_again> is called, and determines the next timeout (if any),
1266	which is also when any modifications are taken into account.	1919	which is also when any modifications are taken into account.
1267		1920
1268	=back	1921	=back
1269		1922
1270	=head3 Examples	1923	=head3 Examples
1271		1924
1272	Example: Create a timer that fires after 60 seconds.	1925	Example: Create a timer that fires after 60 seconds.
1273		1926
1274	static void	1927	static void
1275	one_minute_cb (struct ev_loop *loop, struct ev_timer *w, int revents)	1928	one_minute_cb (struct ev_loop *loop, ev_timer *w, int revents)
1276	{	1929	{
1277	.. one minute over, w is actually stopped right here	1930	.. one minute over, w is actually stopped right here
1278	}	1931	}
1279		1932
1280	struct ev_timer mytimer;	1933	ev_timer mytimer;
1281	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);	1934	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);
1282	ev_timer_start (loop, &mytimer);	1935	ev_timer_start (loop, &mytimer);
1283		1936
1284	Example: Create a timeout timer that times out after 10 seconds of	1937	Example: Create a timeout timer that times out after 10 seconds of
1285	inactivity.	1938	inactivity.
1286		1939
1287	static void	1940	static void
1288	timeout_cb (struct ev_loop *loop, struct ev_timer *w, int revents)	1941	timeout_cb (struct ev_loop *loop, ev_timer *w, int revents)
1289	{	1942	{
1290	.. ten seconds without any activity	1943	.. ten seconds without any activity
1291	}	1944	}
1292		1945
1293	struct ev_timer mytimer;	1946	ev_timer mytimer;
1294	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */	1947	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */
1295	ev_timer_again (&mytimer); /* start timer */	1948	ev_timer_again (&mytimer); /* start timer */
1296	ev_loop (loop, 0);	1949	ev_loop (loop, 0);
1297		1950
1298	// and in some piece of code that gets executed on any "activity":	1951	// and in some piece of code that gets executed on any "activity":
…		…
1303	=head2 C<ev_periodic> - to cron or not to cron?	1956	=head2 C<ev_periodic> - to cron or not to cron?
1304		1957
1305	Periodic watchers are also timers of a kind, but they are very versatile	1958	Periodic watchers are also timers of a kind, but they are very versatile
1306	(and unfortunately a bit complex).	1959	(and unfortunately a bit complex).
1307		1960
1308	Unlike C<ev_timer>'s, they are not based on real time (or relative time)	1961	Unlike C<ev_timer>, periodic watchers are not based on real time (or
1309	but on wall clock time (absolute time). You can tell a periodic watcher	1962	relative time, the physical time that passes) but on wall clock time
1310	to trigger after some specific point in time. For example, if you tell a	1963	(absolute time, the thing you can read on your calender or clock). The
1311	periodic watcher to trigger in 10 seconds (by specifying e.g. C<ev_now ()	1964	difference is that wall clock time can run faster or slower than real
1312	+ 10.>, that is, an absolute time not a delay) and then reset your system	1965	time, and time jumps are not uncommon (e.g. when you adjust your
1313	clock to January of the previous year, then it will take more than year	1966	wrist-watch).
1314	to trigger the event (unlike an C<ev_timer>, which would still trigger
1315	roughly 10 seconds later as it uses a relative timeout).
1316		1967
		1968	You can tell a periodic watcher to trigger after some specific point
		1969	in time: for example, if you tell a periodic watcher to trigger "in 10
		1970	seconds" (by specifying e.g. C<ev_now () + 10.>, that is, an absolute time
		1971	not a delay) and then reset your system clock to January of the previous
		1972	year, then it will take a year or more to trigger the event (unlike an
		1973	C<ev_timer>, which would still trigger roughly 10 seconds after starting
		1974	it, as it uses a relative timeout).
		1975
1317	C<ev_periodic>s can also be used to implement vastly more complex timers,	1976	C<ev_periodic> watchers can also be used to implement vastly more complex
1318	such as triggering an event on each "midnight, local time", or other	1977	timers, such as triggering an event on each "midnight, local time", or
1319	complicated, rules.	1978	other complicated rules. This cannot be done with C<ev_timer> watchers, as
		1979	those cannot react to time jumps.
1320		1980
1321	As with timers, the callback is guaranteed to be invoked only when the	1981	As with timers, the callback is guaranteed to be invoked only when the
1322	time (C<at>) has passed, but if multiple periodic timers become ready	1982	point in time where it is supposed to trigger has passed. If multiple
1323	during the same loop iteration then order of execution is undefined.	1983	timers become ready during the same loop iteration then the ones with
		1984	earlier time-out values are invoked before ones with later time-out values
		1985	(but this is no longer true when a callback calls C<ev_loop> recursively).
1324		1986
1325	=head3 Watcher-Specific Functions and Data Members	1987	=head3 Watcher-Specific Functions and Data Members
1326		1988
1327	=over 4	1989	=over 4
1328		1990
1329	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)	1991	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp offset, ev_tstamp interval, reschedule_cb)
1330		1992
1331	=item ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)	1993	=item ev_periodic_set (ev_periodic *, ev_tstamp offset, ev_tstamp interval, reschedule_cb)
1332		1994
1333	Lots of arguments, lets sort it out... There are basically three modes of	1995	Lots of arguments, let's sort it out... There are basically three modes of
1334	operation, and we will explain them from simplest to complex:	1996	operation, and we will explain them from simplest to most complex:
1335		1997
1336	=over 4	1998	=over 4
1337		1999
1338	=item * absolute timer (at = time, interval = reschedule_cb = 0)	2000	=item * absolute timer (offset = absolute time, interval = 0, reschedule_cb = 0)
1339		2001
1340	In this configuration the watcher triggers an event after the wall clock	2002	In this configuration the watcher triggers an event after the wall clock
1341	time C<at> has passed and doesn't repeat. It will not adjust when a time	2003	time C<offset> has passed. It will not repeat and will not adjust when a
1342	jump occurs, that is, if it is to be run at January 1st 2011 then it will	2004	time jump occurs, that is, if it is to be run at January 1st 2011 then it
1343	run when the system time reaches or surpasses this time.	2005	will be stopped and invoked when the system clock reaches or surpasses
		2006	this point in time.
1344		2007
1345	=item * repeating interval timer (at = offset, interval > 0, reschedule_cb = 0)	2008	=item * repeating interval timer (offset = offset within interval, interval > 0, reschedule_cb = 0)
1346		2009
1347	In this mode the watcher will always be scheduled to time out at the next	2010	In this mode the watcher will always be scheduled to time out at the next
1348	C<at + N * interval> time (for some integer N, which can also be negative)	2011	C<offset + N * interval> time (for some integer N, which can also be
1349	and then repeat, regardless of any time jumps.	2012	negative) and then repeat, regardless of any time jumps. The C<offset>
		2013	argument is merely an offset into the C<interval> periods.
1350		2014
1351	This can be used to create timers that do not drift with respect to system	2015	This can be used to create timers that do not drift with respect to the
1352	time, for example, here is a C<ev_periodic> that triggers each hour, on	2016	system clock, for example, here is an C<ev_periodic> that triggers each
1353	the hour:	2017	hour, on the hour (with respect to UTC):
1354		2018
1355	ev_periodic_set (&periodic, 0., 3600., 0);	2019	ev_periodic_set (&periodic, 0., 3600., 0);
1356		2020
1357	This doesn't mean there will always be 3600 seconds in between triggers,	2021	This doesn't mean there will always be 3600 seconds in between triggers,
1358	but only that the callback will be called when the system time shows a	2022	but only that the callback will be called when the system time shows a
1359	full hour (UTC), or more correctly, when the system time is evenly divisible	2023	full hour (UTC), or more correctly, when the system time is evenly divisible
1360	by 3600.	2024	by 3600.
1361		2025
1362	Another way to think about it (for the mathematically inclined) is that	2026	Another way to think about it (for the mathematically inclined) is that
1363	C<ev_periodic> will try to run the callback in this mode at the next possible	2027	C<ev_periodic> will try to run the callback in this mode at the next possible
1364	time where C<time = at (mod interval)>, regardless of any time jumps.	2028	time where C<time = offset (mod interval)>, regardless of any time jumps.
1365		2029
1366	For numerical stability it is preferable that the C<at> value is near	2030	For numerical stability it is preferable that the C<offset> value is near
1367	C<ev_now ()> (the current time), but there is no range requirement for	2031	C<ev_now ()> (the current time), but there is no range requirement for
1368	this value, and in fact is often specified as zero.	2032	this value, and in fact is often specified as zero.
1369		2033
1370	Note also that there is an upper limit to how often a timer can fire (CPU	2034	Note also that there is an upper limit to how often a timer can fire (CPU
1371	speed for example), so if C<interval> is very small then timing stability	2035	speed for example), so if C<interval> is very small then timing stability
1372	will of course deteriorate. Libev itself tries to be exact to be about one	2036	will of course deteriorate. Libev itself tries to be exact to be about one
1373	millisecond (if the OS supports it and the machine is fast enough).	2037	millisecond (if the OS supports it and the machine is fast enough).
1374		2038
1375	=item * manual reschedule mode (at and interval ignored, reschedule_cb = callback)	2039	=item * manual reschedule mode (offset ignored, interval ignored, reschedule_cb = callback)
1376		2040
1377	In this mode the values for C<interval> and C<at> are both being	2041	In this mode the values for C<interval> and C<offset> are both being
1378	ignored. Instead, each time the periodic watcher gets scheduled, the	2042	ignored. Instead, each time the periodic watcher gets scheduled, the
1379	reschedule callback will be called with the watcher as first, and the	2043	reschedule callback will be called with the watcher as first, and the
1380	current time as second argument.	2044	current time as second argument.
1381		2045
1382	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher,	2046	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher, ever,
1383	ever, or make ANY event loop modifications whatsoever>.	2047	or make ANY other event loop modifications whatsoever, unless explicitly
		2048	allowed by documentation here>.
1384		2049
1385	If you need to stop it, return C<now + 1e30> (or so, fudge fudge) and stop	2050	If you need to stop it, return C<now + 1e30> (or so, fudge fudge) and stop
1386	it afterwards (e.g. by starting an C<ev_prepare> watcher, which is the	2051	it afterwards (e.g. by starting an C<ev_prepare> watcher, which is the
1387	only event loop modification you are allowed to do).	2052	only event loop modification you are allowed to do).
1388		2053
1389	The callback prototype is C<ev_tstamp (*reschedule_cb)(struct ev_periodic	2054	The callback prototype is C<ev_tstamp (*reschedule_cb)(ev_periodic
1390	*w, ev_tstamp now)>, e.g.:	2055	*w, ev_tstamp now)>, e.g.:
1391		2056
		2057	static ev_tstamp
1392	static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now)	2058	my_rescheduler (ev_periodic *w, ev_tstamp now)
1393	{	2059	{
1394	return now + 60.;	2060	return now + 60.;
1395	}	2061	}
1396		2062
1397	It must return the next time to trigger, based on the passed time value	2063	It must return the next time to trigger, based on the passed time value
…		…
1417	a different time than the last time it was called (e.g. in a crond like	2083	a different time than the last time it was called (e.g. in a crond like
1418	program when the crontabs have changed).	2084	program when the crontabs have changed).
1419		2085
1420	=item ev_tstamp ev_periodic_at (ev_periodic *)	2086	=item ev_tstamp ev_periodic_at (ev_periodic *)
1421		2087
1422	When active, returns the absolute time that the watcher is supposed to	2088	When active, returns the absolute time that the watcher is supposed
1423	trigger next.	2089	to trigger next. This is not the same as the C<offset> argument to
		2090	C<ev_periodic_set>, but indeed works even in interval and manual
		2091	rescheduling modes.
1424		2092
1425	=item ev_tstamp offset [read-write]	2093	=item ev_tstamp offset [read-write]
1426		2094
1427	When repeating, this contains the offset value, otherwise this is the	2095	When repeating, this contains the offset value, otherwise this is the
1428	absolute point in time (the C<at> value passed to C<ev_periodic_set>).	2096	absolute point in time (the C<offset> value passed to C<ev_periodic_set>,
		2097	although libev might modify this value for better numerical stability).
1429		2098
1430	Can be modified any time, but changes only take effect when the periodic	2099	Can be modified any time, but changes only take effect when the periodic
1431	timer fires or C<ev_periodic_again> is being called.	2100	timer fires or C<ev_periodic_again> is being called.
1432		2101
1433	=item ev_tstamp interval [read-write]	2102	=item ev_tstamp interval [read-write]
1434		2103
1435	The current interval value. Can be modified any time, but changes only	2104	The current interval value. Can be modified any time, but changes only
1436	take effect when the periodic timer fires or C<ev_periodic_again> is being	2105	take effect when the periodic timer fires or C<ev_periodic_again> is being
1437	called.	2106	called.
1438		2107
1439	=item ev_tstamp (*reschedule_cb)(struct ev_periodic *w, ev_tstamp now) [read-write]	2108	=item ev_tstamp (*reschedule_cb)(ev_periodic *w, ev_tstamp now) [read-write]
1440		2109
1441	The current reschedule callback, or C<0>, if this functionality is	2110	The current reschedule callback, or C<0>, if this functionality is
1442	switched off. Can be changed any time, but changes only take effect when	2111	switched off. Can be changed any time, but changes only take effect when
1443	the periodic timer fires or C<ev_periodic_again> is being called.	2112	the periodic timer fires or C<ev_periodic_again> is being called.
1444		2113
1445	=back	2114	=back
1446		2115
1447	=head3 Examples	2116	=head3 Examples
1448		2117
1449	Example: Call a callback every hour, or, more precisely, whenever the	2118	Example: Call a callback every hour, or, more precisely, whenever the
1450	system clock is divisible by 3600. The callback invocation times have	2119	system time is divisible by 3600. The callback invocation times have
1451	potentially a lot of jitter, but good long-term stability.	2120	potentially a lot of jitter, but good long-term stability.
1452		2121
1453	static void	2122	static void
1454	clock_cb (struct ev_loop *loop, struct ev_io *w, int revents)	2123	clock_cb (struct ev_loop *loop, ev_io *w, int revents)
1455	{	2124	{
1456	... its now a full hour (UTC, or TAI or whatever your clock follows)	2125	... its now a full hour (UTC, or TAI or whatever your clock follows)
1457	}	2126	}
1458		2127
1459	struct ev_periodic hourly_tick;	2128	ev_periodic hourly_tick;
1460	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);	2129	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);
1461	ev_periodic_start (loop, &hourly_tick);	2130	ev_periodic_start (loop, &hourly_tick);
1462		2131
1463	Example: The same as above, but use a reschedule callback to do it:	2132	Example: The same as above, but use a reschedule callback to do it:
1464		2133
1465	#include <math.h>	2134	#include <math.h>
1466		2135
1467	static ev_tstamp	2136	static ev_tstamp
1468	my_scheduler_cb (struct ev_periodic *w, ev_tstamp now)	2137	my_scheduler_cb (ev_periodic *w, ev_tstamp now)
1469	{	2138	{
1470	return fmod (now, 3600.) + 3600.;	2139	return now + (3600. - fmod (now, 3600.));
1471	}	2140	}
1472		2141
1473	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);	2142	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);
1474		2143
1475	Example: Call a callback every hour, starting now:	2144	Example: Call a callback every hour, starting now:
1476		2145
1477	struct ev_periodic hourly_tick;	2146	ev_periodic hourly_tick;
1478	ev_periodic_init (&hourly_tick, clock_cb,	2147	ev_periodic_init (&hourly_tick, clock_cb,
1479	fmod (ev_now (loop), 3600.), 3600., 0);	2148	fmod (ev_now (loop), 3600.), 3600., 0);
1480	ev_periodic_start (loop, &hourly_tick);	2149	ev_periodic_start (loop, &hourly_tick);
1481		2150
1482		2151
…		…
1485	Signal watchers will trigger an event when the process receives a specific	2154	Signal watchers will trigger an event when the process receives a specific
1486	signal one or more times. Even though signals are very asynchronous, libev	2155	signal one or more times. Even though signals are very asynchronous, libev
1487	will try it's best to deliver signals synchronously, i.e. as part of the	2156	will try it's best to deliver signals synchronously, i.e. as part of the
1488	normal event processing, like any other event.	2157	normal event processing, like any other event.
1489		2158
		2159	If you want signals to be delivered truly asynchronously, just use
		2160	C<sigaction> as you would do without libev and forget about sharing
		2161	the signal. You can even use C<ev_async> from a signal handler to
		2162	synchronously wake up an event loop.
		2163
1490	You can configure as many watchers as you like per signal. Only when the	2164	You can configure as many watchers as you like for the same signal, but
		2165	only within the same loop, i.e. you can watch for C<SIGINT> in your
		2166	default loop and for C<SIGIO> in another loop, but you cannot watch for
		2167	C<SIGINT> in both the default loop and another loop at the same time. At
		2168	the moment, C<SIGCHLD> is permanently tied to the default loop.
		2169
1491	first watcher gets started will libev actually register a signal watcher	2170	When the first watcher gets started will libev actually register something
1492	with the kernel (thus it coexists with your own signal handlers as long	2171	with the kernel (thus it coexists with your own signal handlers as long as
1493	as you don't register any with libev). Similarly, when the last signal	2172	you don't register any with libev for the same signal).
1494	watcher for a signal is stopped libev will reset the signal handler to
1495	SIG_DFL (regardless of what it was set to before).
1496		2173
1497	If possible and supported, libev will install its handlers with	2174	If possible and supported, libev will install its handlers with
1498	C<SA_RESTART> behaviour enabled, so system calls should not be unduly	2175	C<SA_RESTART> (or equivalent) behaviour enabled, so system calls should
1499	interrupted. If you have a problem with system calls getting interrupted by	2176	not be unduly interrupted. If you have a problem with system calls getting
1500	signals you can block all signals in an C<ev_check> watcher and unblock	2177	interrupted by signals you can block all signals in an C<ev_check> watcher
1501	them in an C<ev_prepare> watcher.	2178	and unblock them in an C<ev_prepare> watcher.
		2179
		2180	=head3 The special problem of inheritance over fork/execve/pthread_create
		2181
		2182	Both the signal mask (C<sigprocmask>) and the signal disposition
		2183	(C<sigaction>) are unspecified after starting a signal watcher (and after
		2184	stopping it again), that is, libev might or might not block the signal,
		2185	and might or might not set or restore the installed signal handler.
		2186
		2187	While this does not matter for the signal disposition (libev never
		2188	sets signals to C<SIG_IGN>, so handlers will be reset to C<SIG_DFL> on
		2189	C<execve>), this matters for the signal mask: many programs do not expect
		2190	certain signals to be blocked.
		2191
		2192	This means that before calling C<exec> (from the child) you should reset
		2193	the signal mask to whatever "default" you expect (all clear is a good
		2194	choice usually).
		2195
		2196	The simplest way to ensure that the signal mask is reset in the child is
		2197	to install a fork handler with C<pthread_atfork> that resets it. That will
		2198	catch fork calls done by libraries (such as the libc) as well.
		2199
		2200	In current versions of libev, the signal will not be blocked indefinitely
		2201	unless you use the C<signalfd> API (C<EV_SIGNALFD>). While this reduces
		2202	the window of opportunity for problems, it will not go away, as libev
		2203	I<has> to modify the signal mask, at least temporarily.
		2204
		2205	So I can't stress this enough: I<If you do not reset your signal mask when
		2206	you expect it to be empty, you have a race condition in your code>. This
		2207	is not a libev-specific thing, this is true for most event libraries.
1502		2208
1503	=head3 Watcher-Specific Functions and Data Members	2209	=head3 Watcher-Specific Functions and Data Members
1504		2210
1505	=over 4	2211	=over 4
1506		2212
…		…
1517		2223
1518	=back	2224	=back
1519		2225
1520	=head3 Examples	2226	=head3 Examples
1521		2227
1522	Example: Try to exit cleanly on SIGINT and SIGTERM.	2228	Example: Try to exit cleanly on SIGINT.
1523		2229
1524	static void	2230	static void
1525	sigint_cb (struct ev_loop *loop, struct ev_signal *w, int revents)	2231	sigint_cb (struct ev_loop *loop, ev_signal *w, int revents)
1526	{	2232	{
1527	ev_unloop (loop, EVUNLOOP_ALL);	2233	ev_unloop (loop, EVUNLOOP_ALL);
1528	}	2234	}
1529		2235
1530	struct ev_signal signal_watcher;	2236	ev_signal signal_watcher;
1531	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);	2237	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);
1532	ev_signal_start (loop, &sigint_cb);	2238	ev_signal_start (loop, &signal_watcher);
1533		2239
1534		2240
1535	=head2 C<ev_child> - watch out for process status changes	2241	=head2 C<ev_child> - watch out for process status changes
1536		2242
1537	Child watchers trigger when your process receives a SIGCHLD in response to	2243	Child watchers trigger when your process receives a SIGCHLD in response to
1538	some child status changes (most typically when a child of yours dies). It	2244	some child status changes (most typically when a child of yours dies or
1539	is permissible to install a child watcher I<after> the child has been	2245	exits). It is permissible to install a child watcher I<after> the child
1540	forked (which implies it might have already exited), as long as the event	2246	has been forked (which implies it might have already exited), as long
1541	loop isn't entered (or is continued from a watcher).	2247	as the event loop isn't entered (or is continued from a watcher), i.e.,
		2248	forking and then immediately registering a watcher for the child is fine,
		2249	but forking and registering a watcher a few event loop iterations later or
		2250	in the next callback invocation is not.
1542		2251
1543	Only the default event loop is capable of handling signals, and therefore	2252	Only the default event loop is capable of handling signals, and therefore
1544	you can only register child watchers in the default event loop.	2253	you can only register child watchers in the default event loop.
1545		2254
		2255	Due to some design glitches inside libev, child watchers will always be
		2256	handled at maximum priority (their priority is set to C<EV_MAXPRI> by
		2257	libev)
		2258
1546	=head3 Process Interaction	2259	=head3 Process Interaction
1547		2260
1548	Libev grabs C<SIGCHLD> as soon as the default event loop is	2261	Libev grabs C<SIGCHLD> as soon as the default event loop is
1549	initialised. This is necessary to guarantee proper behaviour even if	2262	initialised. This is necessary to guarantee proper behaviour even if the
1550	the first child watcher is started after the child exits. The occurrence	2263	first child watcher is started after the child exits. The occurrence
1551	of C<SIGCHLD> is recorded asynchronously, but child reaping is done	2264	of C<SIGCHLD> is recorded asynchronously, but child reaping is done
1552	synchronously as part of the event loop processing. Libev always reaps all	2265	synchronously as part of the event loop processing. Libev always reaps all
1553	children, even ones not watched.	2266	children, even ones not watched.
1554		2267
1555	=head3 Overriding the Built-In Processing	2268	=head3 Overriding the Built-In Processing
…		…
1559	handler, you can override it easily by installing your own handler for	2272	handler, you can override it easily by installing your own handler for
1560	C<SIGCHLD> after initialising the default loop, and making sure the	2273	C<SIGCHLD> after initialising the default loop, and making sure the
1561	default loop never gets destroyed. You are encouraged, however, to use an	2274	default loop never gets destroyed. You are encouraged, however, to use an
1562	event-based approach to child reaping and thus use libev's support for	2275	event-based approach to child reaping and thus use libev's support for
1563	that, so other libev users can use C<ev_child> watchers freely.	2276	that, so other libev users can use C<ev_child> watchers freely.
		2277
		2278	=head3 Stopping the Child Watcher
		2279
		2280	Currently, the child watcher never gets stopped, even when the
		2281	child terminates, so normally one needs to stop the watcher in the
		2282	callback. Future versions of libev might stop the watcher automatically
		2283	when a child exit is detected (calling C<ev_child_stop> twice is not a
		2284	problem).
1564		2285
1565	=head3 Watcher-Specific Functions and Data Members	2286	=head3 Watcher-Specific Functions and Data Members
1566		2287
1567	=over 4	2288	=over 4
1568		2289
…		…
1600	its completion.	2321	its completion.
1601		2322
1602	ev_child cw;	2323	ev_child cw;
1603		2324
1604	static void	2325	static void
1605	child_cb (EV_P_ struct ev_child *w, int revents)	2326	child_cb (EV_P_ ev_child *w, int revents)
1606	{	2327	{
1607	ev_child_stop (EV_A_ w);	2328	ev_child_stop (EV_A_ w);
1608	printf ("process %d exited with status %x\n", w->rpid, w->rstatus);	2329	printf ("process %d exited with status %x\n", w->rpid, w->rstatus);
1609	}	2330	}
1610		2331
…		…
1625		2346
1626		2347
1627	=head2 C<ev_stat> - did the file attributes just change?	2348	=head2 C<ev_stat> - did the file attributes just change?
1628		2349
1629	This watches a file system path for attribute changes. That is, it calls	2350	This watches a file system path for attribute changes. That is, it calls
1630	C<stat> regularly (or when the OS says it changed) and sees if it changed	2351	C<stat> on that path in regular intervals (or when the OS says it changed)
1631	compared to the last time, invoking the callback if it did.	2352	and sees if it changed compared to the last time, invoking the callback if
		2353	it did.
1632		2354
1633	The path does not need to exist: changing from "path exists" to "path does	2355	The path does not need to exist: changing from "path exists" to "path does
1634	not exist" is a status change like any other. The condition "path does	2356	not exist" is a status change like any other. The condition "path does not
1635	not exist" is signified by the C<st_nlink> field being zero (which is	2357	exist" (or more correctly "path cannot be stat'ed") is signified by the
1636	otherwise always forced to be at least one) and all the other fields of	2358	C<st_nlink> field being zero (which is otherwise always forced to be at
1637	the stat buffer having unspecified contents.	2359	least one) and all the other fields of the stat buffer having unspecified
		2360	contents.
1638		2361
1639	The path I<should> be absolute and I<must not> end in a slash. If it is	2362	The path I<must not> end in a slash or contain special components such as
		2363	C<.> or C<..>. The path I<should> be absolute: If it is relative and
1640	relative and your working directory changes, the behaviour is undefined.	2364	your working directory changes, then the behaviour is undefined.
1641		2365
1642	Since there is no standard to do this, the portable implementation simply	2366	Since there is no portable change notification interface available, the
1643	calls C<stat (2)> regularly on the path to see if it changed somehow. You	2367	portable implementation simply calls C<stat(2)> regularly on the path
1644	can specify a recommended polling interval for this case. If you specify	2368	to see if it changed somehow. You can specify a recommended polling
1645	a polling interval of C<0> (highly recommended!) then a I<suitable,	2369	interval for this case. If you specify a polling interval of C<0> (highly
1646	unspecified default> value will be used (which you can expect to be around	2370	recommended!) then a I<suitable, unspecified default> value will be used
1647	five seconds, although this might change dynamically). Libev will also	2371	(which you can expect to be around five seconds, although this might
1648	impose a minimum interval which is currently around C<0.1>, but thats	2372	change dynamically). Libev will also impose a minimum interval which is
1649	usually overkill.	2373	currently around C<0.1>, but that's usually overkill.
1650		2374
1651	This watcher type is not meant for massive numbers of stat watchers,	2375	This watcher type is not meant for massive numbers of stat watchers,
1652	as even with OS-supported change notifications, this can be	2376	as even with OS-supported change notifications, this can be
1653	resource-intensive.	2377	resource-intensive.
1654		2378
1655	At the time of this writing, only the Linux inotify interface is	2379	At the time of this writing, the only OS-specific interface implemented
1656	implemented (implementing kqueue support is left as an exercise for the	2380	is the Linux inotify interface (implementing kqueue support is left as an
1657	reader, note, however, that the author sees no way of implementing ev_stat	2381	exercise for the reader. Note, however, that the author sees no way of
1658	semantics with kqueue). Inotify will be used to give hints only and should	2382	implementing C<ev_stat> semantics with kqueue, except as a hint).
1659	not change the semantics of C<ev_stat> watchers, which means that libev
1660	sometimes needs to fall back to regular polling again even with inotify,
1661	but changes are usually detected immediately, and if the file exists there
1662	will be no polling.
1663		2383
1664	=head3 ABI Issues (Largefile Support)	2384	=head3 ABI Issues (Largefile Support)
1665		2385
1666	Libev by default (unless the user overrides this) uses the default	2386	Libev by default (unless the user overrides this) uses the default
1667	compilation environment, which means that on systems with optionally	2387	compilation environment, which means that on systems with large file
1668	disabled large file support, you get the 32 bit version of the stat	2388	support disabled by default, you get the 32 bit version of the stat
1669	structure. When using the library from programs that change the ABI to	2389	structure. When using the library from programs that change the ABI to
1670	use 64 bit file offsets the programs will fail. In that case you have to	2390	use 64 bit file offsets the programs will fail. In that case you have to
1671	compile libev with the same flags to get binary compatibility. This is	2391	compile libev with the same flags to get binary compatibility. This is
1672	obviously the case with any flags that change the ABI, but the problem is	2392	obviously the case with any flags that change the ABI, but the problem is
1673	most noticeably with ev_stat and large file support.	2393	most noticeably displayed with ev_stat and large file support.
1674		2394
1675	=head3 Inotify	2395	The solution for this is to lobby your distribution maker to make large
		2396	file interfaces available by default (as e.g. FreeBSD does) and not
		2397	optional. Libev cannot simply switch on large file support because it has
		2398	to exchange stat structures with application programs compiled using the
		2399	default compilation environment.
1676		2400
		2401	=head3 Inotify and Kqueue
		2402
1677	When C<inotify (7)> support has been compiled into libev (generally only	2403	When C<inotify (7)> support has been compiled into libev and present at
1678	available on Linux) and present at runtime, it will be used to speed up	2404	runtime, it will be used to speed up change detection where possible. The
1679	change detection where possible. The inotify descriptor will be created lazily	2405	inotify descriptor will be created lazily when the first C<ev_stat>
1680	when the first C<ev_stat> watcher is being started.	2406	watcher is being started.
1681		2407
1682	Inotify presence does not change the semantics of C<ev_stat> watchers	2408	Inotify presence does not change the semantics of C<ev_stat> watchers
1683	except that changes might be detected earlier, and in some cases, to avoid	2409	except that changes might be detected earlier, and in some cases, to avoid
1684	making regular C<stat> calls. Even in the presence of inotify support	2410	making regular C<stat> calls. Even in the presence of inotify support
1685	there are many cases where libev has to resort to regular C<stat> polling.	2411	there are many cases where libev has to resort to regular C<stat> polling,
		2412	but as long as kernel 2.6.25 or newer is used (2.6.24 and older have too
		2413	many bugs), the path exists (i.e. stat succeeds), and the path resides on
		2414	a local filesystem (libev currently assumes only ext2/3, jfs, reiserfs and
		2415	xfs are fully working) libev usually gets away without polling.
1686		2416
1687	(There is no support for kqueue, as apparently it cannot be used to	2417	There is no support for kqueue, as apparently it cannot be used to
1688	implement this functionality, due to the requirement of having a file	2418	implement this functionality, due to the requirement of having a file
1689	descriptor open on the object at all times).	2419	descriptor open on the object at all times, and detecting renames, unlinks
		2420	etc. is difficult.
		2421
		2422	=head3 C<stat ()> is a synchronous operation
		2423
		2424	Libev doesn't normally do any kind of I/O itself, and so is not blocking
		2425	the process. The exception are C<ev_stat> watchers - those call C<stat
		2426	()>, which is a synchronous operation.
		2427
		2428	For local paths, this usually doesn't matter: unless the system is very
		2429	busy or the intervals between stat's are large, a stat call will be fast,
		2430	as the path data is usually in memory already (except when starting the
		2431	watcher).
		2432
		2433	For networked file systems, calling C<stat ()> can block an indefinite
		2434	time due to network issues, and even under good conditions, a stat call
		2435	often takes multiple milliseconds.
		2436
		2437	Therefore, it is best to avoid using C<ev_stat> watchers on networked
		2438	paths, although this is fully supported by libev.
1690		2439
1691	=head3 The special problem of stat time resolution	2440	=head3 The special problem of stat time resolution
1692		2441
1693	The C<stat ()> system call only supports full-second resolution portably, and	2442	The C<stat ()> system call only supports full-second resolution portably,
1694	even on systems where the resolution is higher, many file systems still	2443	and even on systems where the resolution is higher, most file systems
1695	only support whole seconds.	2444	still only support whole seconds.
1696		2445
1697	That means that, if the time is the only thing that changes, you can	2446	That means that, if the time is the only thing that changes, you can
1698	easily miss updates: on the first update, C<ev_stat> detects a change and	2447	easily miss updates: on the first update, C<ev_stat> detects a change and
1699	calls your callback, which does something. When there is another update	2448	calls your callback, which does something. When there is another update
1700	within the same second, C<ev_stat> will be unable to detect it as the stat	2449	within the same second, C<ev_stat> will be unable to detect unless the
1701	data does not change.	2450	stat data does change in other ways (e.g. file size).
1702		2451
1703	The solution to this is to delay acting on a change for slightly more	2452	The solution to this is to delay acting on a change for slightly more
1704	than a second (or till slightly after the next full second boundary), using	2453	than a second (or till slightly after the next full second boundary), using
1705	a roughly one-second-delay C<ev_timer> (e.g. C<ev_timer_set (w, 0., 1.02);	2454	a roughly one-second-delay C<ev_timer> (e.g. C<ev_timer_set (w, 0., 1.02);
1706	ev_timer_again (loop, w)>).	2455	ev_timer_again (loop, w)>).
…		…
1726	C<path>. The C<interval> is a hint on how quickly a change is expected to	2475	C<path>. The C<interval> is a hint on how quickly a change is expected to
1727	be detected and should normally be specified as C<0> to let libev choose	2476	be detected and should normally be specified as C<0> to let libev choose
1728	a suitable value. The memory pointed to by C<path> must point to the same	2477	a suitable value. The memory pointed to by C<path> must point to the same
1729	path for as long as the watcher is active.	2478	path for as long as the watcher is active.
1730		2479
1731	The callback will receive C<EV_STAT> when a change was detected, relative	2480	The callback will receive an C<EV_STAT> event when a change was detected,
1732	to the attributes at the time the watcher was started (or the last change	2481	relative to the attributes at the time the watcher was started (or the
1733	was detected).	2482	last change was detected).
1734		2483
1735	=item ev_stat_stat (loop, ev_stat *)	2484	=item ev_stat_stat (loop, ev_stat *)
1736		2485
1737	Updates the stat buffer immediately with new values. If you change the	2486	Updates the stat buffer immediately with new values. If you change the
1738	watched path in your callback, you could call this function to avoid	2487	watched path in your callback, you could call this function to avoid
…		…
1821		2570
1822		2571
1823	=head2 C<ev_idle> - when you've got nothing better to do...	2572	=head2 C<ev_idle> - when you've got nothing better to do...
1824		2573
1825	Idle watchers trigger events when no other events of the same or higher	2574	Idle watchers trigger events when no other events of the same or higher
1826	priority are pending (prepare, check and other idle watchers do not	2575	priority are pending (prepare, check and other idle watchers do not count
1827	count).	2576	as receiving "events").
1828		2577
1829	That is, as long as your process is busy handling sockets or timeouts	2578	That is, as long as your process is busy handling sockets or timeouts
1830	(or even signals, imagine) of the same or higher priority it will not be	2579	(or even signals, imagine) of the same or higher priority it will not be
1831	triggered. But when your process is idle (or only lower-priority watchers	2580	triggered. But when your process is idle (or only lower-priority watchers
1832	are pending), the idle watchers are being called once per event loop	2581	are pending), the idle watchers are being called once per event loop
…		…
1843		2592
1844	=head3 Watcher-Specific Functions and Data Members	2593	=head3 Watcher-Specific Functions and Data Members
1845		2594
1846	=over 4	2595	=over 4
1847		2596
1848	=item ev_idle_init (ev_signal *, callback)	2597	=item ev_idle_init (ev_idle *, callback)
1849		2598
1850	Initialises and configures the idle watcher - it has no parameters of any	2599	Initialises and configures the idle watcher - it has no parameters of any
1851	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,	2600	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,
1852	believe me.	2601	believe me.
1853		2602
…		…
1857		2606
1858	Example: Dynamically allocate an C<ev_idle> watcher, start it, and in the	2607	Example: Dynamically allocate an C<ev_idle> watcher, start it, and in the
1859	callback, free it. Also, use no error checking, as usual.	2608	callback, free it. Also, use no error checking, as usual.
1860		2609
1861	static void	2610	static void
1862	idle_cb (struct ev_loop *loop, struct ev_idle *w, int revents)	2611	idle_cb (struct ev_loop *loop, ev_idle *w, int revents)
1863	{	2612	{
1864	free (w);	2613	free (w);
1865	// now do something you wanted to do when the program has	2614	// now do something you wanted to do when the program has
1866	// no longer anything immediate to do.	2615	// no longer anything immediate to do.
1867	}	2616	}
1868		2617
1869	struct ev_idle *idle_watcher = malloc (sizeof (struct ev_idle));	2618	ev_idle *idle_watcher = malloc (sizeof (ev_idle));
1870	ev_idle_init (idle_watcher, idle_cb);	2619	ev_idle_init (idle_watcher, idle_cb);
1871	ev_idle_start (loop, idle_cb);	2620	ev_idle_start (loop, idle_watcher);
1872		2621
1873		2622
1874	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!	2623	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!
1875		2624
1876	Prepare and check watchers are usually (but not always) used in tandem:	2625	Prepare and check watchers are usually (but not always) used in pairs:
1877	prepare watchers get invoked before the process blocks and check watchers	2626	prepare watchers get invoked before the process blocks and check watchers
1878	afterwards.	2627	afterwards.
1879		2628
1880	You I<must not> call C<ev_loop> or similar functions that enter	2629	You I<must not> call C<ev_loop> or similar functions that enter
1881	the current event loop from either C<ev_prepare> or C<ev_check>	2630	the current event loop from either C<ev_prepare> or C<ev_check>
…		…
1884	those watchers, i.e. the sequence will always be C<ev_prepare>, blocking,	2633	those watchers, i.e. the sequence will always be C<ev_prepare>, blocking,
1885	C<ev_check> so if you have one watcher of each kind they will always be	2634	C<ev_check> so if you have one watcher of each kind they will always be
1886	called in pairs bracketing the blocking call.	2635	called in pairs bracketing the blocking call.
1887		2636
1888	Their main purpose is to integrate other event mechanisms into libev and	2637	Their main purpose is to integrate other event mechanisms into libev and
1889	their use is somewhat advanced. This could be used, for example, to track	2638	their use is somewhat advanced. They could be used, for example, to track
1890	variable changes, implement your own watchers, integrate net-snmp or a	2639	variable changes, implement your own watchers, integrate net-snmp or a
1891	coroutine library and lots more. They are also occasionally useful if	2640	coroutine library and lots more. They are also occasionally useful if
1892	you cache some data and want to flush it before blocking (for example,	2641	you cache some data and want to flush it before blocking (for example,
1893	in X programs you might want to do an C<XFlush ()> in an C<ev_prepare>	2642	in X programs you might want to do an C<XFlush ()> in an C<ev_prepare>
1894	watcher).	2643	watcher).
1895		2644
1896	This is done by examining in each prepare call which file descriptors need	2645	This is done by examining in each prepare call which file descriptors
1897	to be watched by the other library, registering C<ev_io> watchers for	2646	need to be watched by the other library, registering C<ev_io> watchers
1898	them and starting an C<ev_timer> watcher for any timeouts (many libraries	2647	for them and starting an C<ev_timer> watcher for any timeouts (many
1899	provide just this functionality). Then, in the check watcher you check for	2648	libraries provide exactly this functionality). Then, in the check watcher,
1900	any events that occurred (by checking the pending status of all watchers	2649	you check for any events that occurred (by checking the pending status
1901	and stopping them) and call back into the library. The I/O and timer	2650	of all watchers and stopping them) and call back into the library. The
1902	callbacks will never actually be called (but must be valid nevertheless,	2651	I/O and timer callbacks will never actually be called (but must be valid
1903	because you never know, you know?).	2652	nevertheless, because you never know, you know?).
1904		2653
1905	As another example, the Perl Coro module uses these hooks to integrate	2654	As another example, the Perl Coro module uses these hooks to integrate
1906	coroutines into libev programs, by yielding to other active coroutines	2655	coroutines into libev programs, by yielding to other active coroutines
1907	during each prepare and only letting the process block if no coroutines	2656	during each prepare and only letting the process block if no coroutines
1908	are ready to run (it's actually more complicated: it only runs coroutines	2657	are ready to run (it's actually more complicated: it only runs coroutines
…		…
1911	loop from blocking if lower-priority coroutines are active, thus mapping	2660	loop from blocking if lower-priority coroutines are active, thus mapping
1912	low-priority coroutines to idle/background tasks).	2661	low-priority coroutines to idle/background tasks).
1913		2662
1914	It is recommended to give C<ev_check> watchers highest (C<EV_MAXPRI>)	2663	It is recommended to give C<ev_check> watchers highest (C<EV_MAXPRI>)
1915	priority, to ensure that they are being run before any other watchers	2664	priority, to ensure that they are being run before any other watchers
		2665	after the poll (this doesn't matter for C<ev_prepare> watchers).
		2666
1916	after the poll. Also, C<ev_check> watchers (and C<ev_prepare> watchers,	2667	Also, C<ev_check> watchers (and C<ev_prepare> watchers, too) should not
1917	too) should not activate ("feed") events into libev. While libev fully	2668	activate ("feed") events into libev. While libev fully supports this, they
1918	supports this, they might get executed before other C<ev_check> watchers	2669	might get executed before other C<ev_check> watchers did their job. As
1919	did their job. As C<ev_check> watchers are often used to embed other	2670	C<ev_check> watchers are often used to embed other (non-libev) event
1920	(non-libev) event loops those other event loops might be in an unusable	2671	loops those other event loops might be in an unusable state until their
1921	state until their C<ev_check> watcher ran (always remind yourself to	2672	C<ev_check> watcher ran (always remind yourself to coexist peacefully with
1922	coexist peacefully with others).	2673	others).
1923		2674
1924	=head3 Watcher-Specific Functions and Data Members	2675	=head3 Watcher-Specific Functions and Data Members
1925		2676
1926	=over 4	2677	=over 4
1927		2678
…		…
1929		2680
1930	=item ev_check_init (ev_check *, callback)	2681	=item ev_check_init (ev_check *, callback)
1931		2682
1932	Initialises and configures the prepare or check watcher - they have no	2683	Initialises and configures the prepare or check watcher - they have no
1933	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>	2684	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>
1934	macros, but using them is utterly, utterly and completely pointless.	2685	macros, but using them is utterly, utterly, utterly and completely
		2686	pointless.
1935		2687
1936	=back	2688	=back
1937		2689
1938	=head3 Examples	2690	=head3 Examples
1939		2691
…		…
1952		2704
1953	static ev_io iow [nfd];	2705	static ev_io iow [nfd];
1954	static ev_timer tw;	2706	static ev_timer tw;
1955		2707
1956	static void	2708	static void
1957	io_cb (ev_loop *loop, ev_io *w, int revents)	2709	io_cb (struct ev_loop *loop, ev_io *w, int revents)
1958	{	2710	{
1959	}	2711	}
1960		2712
1961	// create io watchers for each fd and a timer before blocking	2713	// create io watchers for each fd and a timer before blocking
1962	static void	2714	static void
1963	adns_prepare_cb (ev_loop *loop, ev_prepare *w, int revents)	2715	adns_prepare_cb (struct ev_loop *loop, ev_prepare *w, int revents)
1964	{	2716	{
1965	int timeout = 3600000;	2717	int timeout = 3600000;
1966	struct pollfd fds [nfd];	2718	struct pollfd fds [nfd];
1967	// actual code will need to loop here and realloc etc.	2719	// actual code will need to loop here and realloc etc.
1968	adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));	2720	adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));
1969		2721
1970	/* the callback is illegal, but won't be called as we stop during check */	2722	/* the callback is illegal, but won't be called as we stop during check */
1971	ev_timer_init (&tw, 0, timeout * 1e-3);	2723	ev_timer_init (&tw, 0, timeout * 1e-3, 0.);
1972	ev_timer_start (loop, &tw);	2724	ev_timer_start (loop, &tw);
1973		2725
1974	// create one ev_io per pollfd	2726	// create one ev_io per pollfd
1975	for (int i = 0; i < nfd; ++i)	2727	for (int i = 0; i < nfd; ++i)
1976	{	2728	{
…		…
1983	}	2735	}
1984	}	2736	}
1985		2737
1986	// stop all watchers after blocking	2738	// stop all watchers after blocking
1987	static void	2739	static void
1988	adns_check_cb (ev_loop *loop, ev_check *w, int revents)	2740	adns_check_cb (struct ev_loop *loop, ev_check *w, int revents)
1989	{	2741	{
1990	ev_timer_stop (loop, &tw);	2742	ev_timer_stop (loop, &tw);
1991		2743
1992	for (int i = 0; i < nfd; ++i)	2744	for (int i = 0; i < nfd; ++i)
1993	{	2745	{
…		…
2032	}	2784	}
2033		2785
2034	// do not ever call adns_afterpoll	2786	// do not ever call adns_afterpoll
2035		2787
2036	Method 4: Do not use a prepare or check watcher because the module you	2788	Method 4: Do not use a prepare or check watcher because the module you
2037	want to embed is too inflexible to support it. Instead, you can override	2789	want to embed is not flexible enough to support it. Instead, you can
2038	their poll function. The drawback with this solution is that the main	2790	override their poll function. The drawback with this solution is that the
2039	loop is now no longer controllable by EV. The C<Glib::EV> module does	2791	main loop is now no longer controllable by EV. The C<Glib::EV> module uses
2040	this.	2792	this approach, effectively embedding EV as a client into the horrible
		2793	libglib event loop.
2041		2794
2042	static gint	2795	static gint
2043	event_poll_func (GPollFD *fds, guint nfds, gint timeout)	2796	event_poll_func (GPollFD *fds, guint nfds, gint timeout)
2044	{	2797	{
2045	int got_events = 0;	2798	int got_events = 0;
…		…
2076	prioritise I/O.	2829	prioritise I/O.
2077		2830
2078	As an example for a bug workaround, the kqueue backend might only support	2831	As an example for a bug workaround, the kqueue backend might only support
2079	sockets on some platform, so it is unusable as generic backend, but you	2832	sockets on some platform, so it is unusable as generic backend, but you
2080	still want to make use of it because you have many sockets and it scales	2833	still want to make use of it because you have many sockets and it scales
2081	so nicely. In this case, you would create a kqueue-based loop and embed it	2834	so nicely. In this case, you would create a kqueue-based loop and embed
2082	into your default loop (which might use e.g. poll). Overall operation will	2835	it into your default loop (which might use e.g. poll). Overall operation
2083	be a bit slower because first libev has to poll and then call kevent, but	2836	will be a bit slower because first libev has to call C<poll> and then
2084	at least you can use both at what they are best.	2837	C<kevent>, but at least you can use both mechanisms for what they are
		2838	best: C<kqueue> for scalable sockets and C<poll> if you want it to work :)
2085		2839
2086	As for prioritising I/O: rarely you have the case where some fds have	2840	As for prioritising I/O: under rare circumstances you have the case where
2087	to be watched and handled very quickly (with low latency), and even	2841	some fds have to be watched and handled very quickly (with low latency),
2088	priorities and idle watchers might have too much overhead. In this case	2842	and even priorities and idle watchers might have too much overhead. In
2089	you would put all the high priority stuff in one loop and all the rest in	2843	this case you would put all the high priority stuff in one loop and all
2090	a second one, and embed the second one in the first.	2844	the rest in a second one, and embed the second one in the first.
2091		2845
2092	As long as the watcher is active, the callback will be invoked every time	2846	As long as the watcher is active, the callback will be invoked every
2093	there might be events pending in the embedded loop. The callback must then	2847	time there might be events pending in the embedded loop. The callback
2094	call C<ev_embed_sweep (mainloop, watcher)> to make a single sweep and invoke	2848	must then call C<ev_embed_sweep (mainloop, watcher)> to make a single
2095	their callbacks (you could also start an idle watcher to give the embedded	2849	sweep and invoke their callbacks (the callback doesn't need to invoke the
2096	loop strictly lower priority for example). You can also set the callback	2850	C<ev_embed_sweep> function directly, it could also start an idle watcher
2097	to C<0>, in which case the embed watcher will automatically execute the	2851	to give the embedded loop strictly lower priority for example).
2098	embedded loop sweep.
2099		2852
2100	As long as the watcher is started it will automatically handle events. The	2853	You can also set the callback to C<0>, in which case the embed watcher
2101	callback will be invoked whenever some events have been handled. You can	2854	will automatically execute the embedded loop sweep whenever necessary.
2102	set the callback to C<0> to avoid having to specify one if you are not
2103	interested in that.
2104		2855
2105	Also, there have not currently been made special provisions for forking:	2856	Fork detection will be handled transparently while the C<ev_embed> watcher
2106	when you fork, you not only have to call C<ev_loop_fork> on both loops,	2857	is active, i.e., the embedded loop will automatically be forked when the
2107	but you will also have to stop and restart any C<ev_embed> watchers	2858	embedding loop forks. In other cases, the user is responsible for calling
2108	yourself.	2859	C<ev_loop_fork> on the embedded loop.
2109		2860
2110	Unfortunately, not all backends are embeddable, only the ones returned by	2861	Unfortunately, not all backends are embeddable: only the ones returned by
2111	C<ev_embeddable_backends> are, which, unfortunately, does not include any	2862	C<ev_embeddable_backends> are, which, unfortunately, does not include any
2112	portable one.	2863	portable one.
2113		2864
2114	So when you want to use this feature you will always have to be prepared	2865	So when you want to use this feature you will always have to be prepared
2115	that you cannot get an embeddable loop. The recommended way to get around	2866	that you cannot get an embeddable loop. The recommended way to get around
2116	this is to have a separate variables for your embeddable loop, try to	2867	this is to have a separate variables for your embeddable loop, try to
2117	create it, and if that fails, use the normal loop for everything.	2868	create it, and if that fails, use the normal loop for everything.
		2869
		2870	=head3 C<ev_embed> and fork
		2871
		2872	While the C<ev_embed> watcher is running, forks in the embedding loop will
		2873	automatically be applied to the embedded loop as well, so no special
		2874	fork handling is required in that case. When the watcher is not running,
		2875	however, it is still the task of the libev user to call C<ev_loop_fork ()>
		2876	as applicable.
2118		2877
2119	=head3 Watcher-Specific Functions and Data Members	2878	=head3 Watcher-Specific Functions and Data Members
2120		2879
2121	=over 4	2880	=over 4
2122		2881
…		…
2150	C<loop_lo> (which is C<loop_hi> in the case no embeddable loop can be	2909	C<loop_lo> (which is C<loop_hi> in the case no embeddable loop can be
2151	used).	2910	used).
2152		2911
2153	struct ev_loop *loop_hi = ev_default_init (0);	2912	struct ev_loop *loop_hi = ev_default_init (0);
2154	struct ev_loop *loop_lo = 0;	2913	struct ev_loop *loop_lo = 0;
2155	struct ev_embed embed;	2914	ev_embed embed;
2156		2915
2157	// see if there is a chance of getting one that works	2916	// see if there is a chance of getting one that works
2158	// (remember that a flags value of 0 means autodetection)	2917	// (remember that a flags value of 0 means autodetection)
2159	loop_lo = ev_embeddable_backends () & ev_recommended_backends ()	2918	loop_lo = ev_embeddable_backends () & ev_recommended_backends ()
2160	? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())	2919	? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())
…		…
2174	kqueue implementation). Store the kqueue/socket-only event loop in	2933	kqueue implementation). Store the kqueue/socket-only event loop in
2175	C<loop_socket>. (One might optionally use C<EVFLAG_NOENV>, too).	2934	C<loop_socket>. (One might optionally use C<EVFLAG_NOENV>, too).
2176		2935
2177	struct ev_loop *loop = ev_default_init (0);	2936	struct ev_loop *loop = ev_default_init (0);
2178	struct ev_loop *loop_socket = 0;	2937	struct ev_loop *loop_socket = 0;
2179	struct ev_embed embed;	2938	ev_embed embed;
2180		2939
2181	if (ev_supported_backends () & ~ev_recommended_backends () & EVBACKEND_KQUEUE)	2940	if (ev_supported_backends () & ~ev_recommended_backends () & EVBACKEND_KQUEUE)
2182	if ((loop_socket = ev_loop_new (EVBACKEND_KQUEUE))	2941	if ((loop_socket = ev_loop_new (EVBACKEND_KQUEUE))
2183	{	2942	{
2184	ev_embed_init (&embed, 0, loop_socket);	2943	ev_embed_init (&embed, 0, loop_socket);
…		…
2199	event loop blocks next and before C<ev_check> watchers are being called,	2958	event loop blocks next and before C<ev_check> watchers are being called,
2200	and only in the child after the fork. If whoever good citizen calling	2959	and only in the child after the fork. If whoever good citizen calling
2201	C<ev_default_fork> cheats and calls it in the wrong process, the fork	2960	C<ev_default_fork> cheats and calls it in the wrong process, the fork
2202	handlers will be invoked, too, of course.	2961	handlers will be invoked, too, of course.
2203		2962
		2963	=head3 The special problem of life after fork - how is it possible?
		2964
		2965	Most uses of C<fork()> consist of forking, then some simple calls to ste
		2966	up/change the process environment, followed by a call to C<exec()>. This
		2967	sequence should be handled by libev without any problems.
		2968
		2969	This changes when the application actually wants to do event handling
		2970	in the child, or both parent in child, in effect "continuing" after the
		2971	fork.
		2972
		2973	The default mode of operation (for libev, with application help to detect
		2974	forks) is to duplicate all the state in the child, as would be expected
		2975	when I<either> the parent I<or> the child process continues.
		2976
		2977	When both processes want to continue using libev, then this is usually the
		2978	wrong result. In that case, usually one process (typically the parent) is
		2979	supposed to continue with all watchers in place as before, while the other
		2980	process typically wants to start fresh, i.e. without any active watchers.
		2981
		2982	The cleanest and most efficient way to achieve that with libev is to
		2983	simply create a new event loop, which of course will be "empty", and
		2984	use that for new watchers. This has the advantage of not touching more
		2985	memory than necessary, and thus avoiding the copy-on-write, and the
		2986	disadvantage of having to use multiple event loops (which do not support
		2987	signal watchers).
		2988
		2989	When this is not possible, or you want to use the default loop for
		2990	other reasons, then in the process that wants to start "fresh", call
		2991	C<ev_default_destroy ()> followed by C<ev_default_loop (...)>. Destroying
		2992	the default loop will "orphan" (not stop) all registered watchers, so you
		2993	have to be careful not to execute code that modifies those watchers. Note
		2994	also that in that case, you have to re-register any signal watchers.
		2995
2204	=head3 Watcher-Specific Functions and Data Members	2996	=head3 Watcher-Specific Functions and Data Members
2205		2997
2206	=over 4	2998	=over 4
2207		2999
2208	=item ev_fork_init (ev_signal *, callback)	3000	=item ev_fork_init (ev_signal *, callback)
…		…
2237	=head3 Queueing	3029	=head3 Queueing
2238		3030
2239	C<ev_async> does not support queueing of data in any way. The reason	3031	C<ev_async> does not support queueing of data in any way. The reason
2240	is that the author does not know of a simple (or any) algorithm for a	3032	is that the author does not know of a simple (or any) algorithm for a
2241	multiple-writer-single-reader queue that works in all cases and doesn't	3033	multiple-writer-single-reader queue that works in all cases and doesn't
2242	need elaborate support such as pthreads.	3034	need elaborate support such as pthreads or unportable memory access
		3035	semantics.
2243		3036
2244	That means that if you want to queue data, you have to provide your own	3037	That means that if you want to queue data, you have to provide your own
2245	queue. But at least I can tell you would implement locking around your	3038	queue. But at least I can tell you how to implement locking around your
2246	queue:	3039	queue:
2247		3040
2248	=over 4	3041	=over 4
2249		3042
2250	=item queueing from a signal handler context	3043	=item queueing from a signal handler context
2251		3044
2252	To implement race-free queueing, you simply add to the queue in the signal	3045	To implement race-free queueing, you simply add to the queue in the signal
2253	handler but you block the signal handler in the watcher callback. Here is an example that does that for	3046	handler but you block the signal handler in the watcher callback. Here is
2254	some fictitious SIGUSR1 handler:	3047	an example that does that for some fictitious SIGUSR1 handler:
2255		3048
2256	static ev_async mysig;	3049	static ev_async mysig;
2257		3050
2258	static void	3051	static void
2259	sigusr1_handler (void)	3052	sigusr1_handler (void)
…		…
2325	=over 4	3118	=over 4
2326		3119
2327	=item ev_async_init (ev_async *, callback)	3120	=item ev_async_init (ev_async *, callback)
2328		3121
2329	Initialises and configures the async watcher - it has no parameters of any	3122	Initialises and configures the async watcher - it has no parameters of any
2330	kind. There is a C<ev_asynd_set> macro, but using it is utterly pointless,	3123	kind. There is a C<ev_async_set> macro, but using it is utterly pointless,
2331	believe me.	3124	trust me.
2332		3125
2333	=item ev_async_send (loop, ev_async *)	3126	=item ev_async_send (loop, ev_async *)
2334		3127
2335	Sends/signals/activates the given C<ev_async> watcher, that is, feeds	3128	Sends/signals/activates the given C<ev_async> watcher, that is, feeds
2336	an C<EV_ASYNC> event on the watcher into the event loop. Unlike	3129	an C<EV_ASYNC> event on the watcher into the event loop. Unlike
2337	C<ev_feed_event>, this call is safe to do in other threads, signal or	3130	C<ev_feed_event>, this call is safe to do from other threads, signal or
2338	similar contexts (see the discussion of C<EV_ATOMIC_T> in the embedding	3131	similar contexts (see the discussion of C<EV_ATOMIC_T> in the embedding
2339	section below on what exactly this means).	3132	section below on what exactly this means).
2340		3133
		3134	Note that, as with other watchers in libev, multiple events might get
		3135	compressed into a single callback invocation (another way to look at this
		3136	is that C<ev_async> watchers are level-triggered, set on C<ev_async_send>,
		3137	reset when the event loop detects that).
		3138
2341	This call incurs the overhead of a system call only once per loop iteration,	3139	This call incurs the overhead of a system call only once per event loop
2342	so while the overhead might be noticeable, it doesn't apply to repeated	3140	iteration, so while the overhead might be noticeable, it doesn't apply to
2343	calls to C<ev_async_send>.	3141	repeated calls to C<ev_async_send> for the same event loop.
2344		3142
2345	=item bool = ev_async_pending (ev_async *)	3143	=item bool = ev_async_pending (ev_async *)
2346		3144
2347	Returns a non-zero value when C<ev_async_send> has been called on the	3145	Returns a non-zero value when C<ev_async_send> has been called on the
2348	watcher but the event has not yet been processed (or even noted) by the	3146	watcher but the event has not yet been processed (or even noted) by the
…		…
2351	C<ev_async_send> sets a flag in the watcher and wakes up the loop. When	3149	C<ev_async_send> sets a flag in the watcher and wakes up the loop. When
2352	the loop iterates next and checks for the watcher to have become active,	3150	the loop iterates next and checks for the watcher to have become active,
2353	it will reset the flag again. C<ev_async_pending> can be used to very	3151	it will reset the flag again. C<ev_async_pending> can be used to very
2354	quickly check whether invoking the loop might be a good idea.	3152	quickly check whether invoking the loop might be a good idea.
2355		3153
2356	Not that this does I<not> check whether the watcher itself is pending, only	3154	Not that this does I<not> check whether the watcher itself is pending,
2357	whether it has been requested to make this watcher pending.	3155	only whether it has been requested to make this watcher pending: there
		3156	is a time window between the event loop checking and resetting the async
		3157	notification, and the callback being invoked.
2358		3158
2359	=back	3159	=back
2360		3160
2361		3161
2362	=head1 OTHER FUNCTIONS	3162	=head1 OTHER FUNCTIONS
…		…
2366	=over 4	3166	=over 4
2367		3167
2368	=item ev_once (loop, int fd, int events, ev_tstamp timeout, callback)	3168	=item ev_once (loop, int fd, int events, ev_tstamp timeout, callback)
2369		3169
2370	This function combines a simple timer and an I/O watcher, calls your	3170	This function combines a simple timer and an I/O watcher, calls your
2371	callback on whichever event happens first and automatically stop both	3171	callback on whichever event happens first and automatically stops both
2372	watchers. This is useful if you want to wait for a single event on an fd	3172	watchers. This is useful if you want to wait for a single event on an fd
2373	or timeout without having to allocate/configure/start/stop/free one or	3173	or timeout without having to allocate/configure/start/stop/free one or
2374	more watchers yourself.	3174	more watchers yourself.
2375		3175
2376	If C<fd> is less than 0, then no I/O watcher will be started and events	3176	If C<fd> is less than 0, then no I/O watcher will be started and the
2377	is being ignored. Otherwise, an C<ev_io> watcher for the given C<fd> and	3177	C<events> argument is being ignored. Otherwise, an C<ev_io> watcher for
2378	C<events> set will be created and started.	3178	the given C<fd> and C<events> set will be created and started.
2379		3179
2380	If C<timeout> is less than 0, then no timeout watcher will be	3180	If C<timeout> is less than 0, then no timeout watcher will be
2381	started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and	3181	started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and
2382	repeat = 0) will be started. While C<0> is a valid timeout, it is of	3182	repeat = 0) will be started. C<0> is a valid timeout.
2383	dubious value.
2384		3183
2385	The callback has the type C<void (*cb)(int revents, void *arg)> and gets	3184	The callback has the type C<void (*cb)(int revents, void *arg)> and gets
2386	passed an C<revents> set like normal event callbacks (a combination of	3185	passed an C<revents> set like normal event callbacks (a combination of
2387	C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMEOUT>) and the C<arg>	3186	C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMEOUT>) and the C<arg>
2388	value passed to C<ev_once>:	3187	value passed to C<ev_once>. Note that it is possible to receive I<both>
		3188	a timeout and an io event at the same time - you probably should give io
		3189	events precedence.
		3190
		3191	Example: wait up to ten seconds for data to appear on STDIN_FILENO.
2389		3192
2390	static void stdin_ready (int revents, void *arg)	3193	static void stdin_ready (int revents, void *arg)
2391	{	3194	{
		3195	if (revents & EV_READ)
		3196	/* stdin might have data for us, joy! */;
2392	if (revents & EV_TIMEOUT)	3197	else if (revents & EV_TIMEOUT)
2393	/* doh, nothing entered */;	3198	/* doh, nothing entered */;
2394	else if (revents & EV_READ)
2395	/* stdin might have data for us, joy! */;
2396	}	3199	}
2397		3200
2398	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);	3201	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);
2399		3202
2400	=item ev_feed_event (ev_loop *, watcher *, int revents)
2401
2402	Feeds the given event set into the event loop, as if the specified event
2403	had happened for the specified watcher (which must be a pointer to an
2404	initialised but not necessarily started event watcher).
2405
2406	=item ev_feed_fd_event (ev_loop *, int fd, int revents)	3203	=item ev_feed_fd_event (loop, int fd, int revents)
2407		3204
2408	Feed an event on the given fd, as if a file descriptor backend detected	3205	Feed an event on the given fd, as if a file descriptor backend detected
2409	the given events it.	3206	the given events it.
2410		3207
2411	=item ev_feed_signal_event (ev_loop *loop, int signum)	3208	=item ev_feed_signal_event (loop, int signum)
2412		3209
2413	Feed an event as if the given signal occurred (C<loop> must be the default	3210	Feed an event as if the given signal occurred (C<loop> must be the default
2414	loop!).	3211	loop!).
2415		3212
2416	=back	3213	=back
…		…
2496		3293
2497	=over 4	3294	=over 4
2498		3295
2499	=item ev::TYPE::TYPE ()	3296	=item ev::TYPE::TYPE ()
2500		3297
2501	=item ev::TYPE::TYPE (struct ev_loop *)	3298	=item ev::TYPE::TYPE (loop)
2502		3299
2503	=item ev::TYPE::~TYPE	3300	=item ev::TYPE::~TYPE
2504		3301
2505	The constructor (optionally) takes an event loop to associate the watcher	3302	The constructor (optionally) takes an event loop to associate the watcher
2506	with. If it is omitted, it will use C<EV_DEFAULT>.	3303	with. If it is omitted, it will use C<EV_DEFAULT>.
…		…
2538		3335
2539	myclass obj;	3336	myclass obj;
2540	ev::io iow;	3337	ev::io iow;
2541	iow.set <myclass, &myclass::io_cb> (&obj);	3338	iow.set <myclass, &myclass::io_cb> (&obj);
2542		3339
		3340	=item w->set (object *)
		3341
		3342	This is an B<experimental> feature that might go away in a future version.
		3343
		3344	This is a variation of a method callback - leaving out the method to call
		3345	will default the method to C<operator ()>, which makes it possible to use
		3346	functor objects without having to manually specify the C<operator ()> all
		3347	the time. Incidentally, you can then also leave out the template argument
		3348	list.
		3349
		3350	The C<operator ()> method prototype must be C<void operator ()(watcher &w,
		3351	int revents)>.
		3352
		3353	See the method-C<set> above for more details.
		3354
		3355	Example: use a functor object as callback.
		3356
		3357	struct myfunctor
		3358	{
		3359	void operator() (ev::io &w, int revents)
		3360	{
		3361	...
		3362	}
		3363	}
		3364
		3365	myfunctor f;
		3366
		3367	ev::io w;
		3368	w.set (&f);
		3369
2543	=item w->set<function> (void *data = 0)	3370	=item w->set<function> (void *data = 0)
2544		3371
2545	Also sets a callback, but uses a static method or plain function as	3372	Also sets a callback, but uses a static method or plain function as
2546	callback. The optional C<data> argument will be stored in the watcher's	3373	callback. The optional C<data> argument will be stored in the watcher's
2547	C<data> member and is free for you to use.	3374	C<data> member and is free for you to use.
2548		3375
2549	The prototype of the C<function> must be C<void (*)(ev::TYPE &w, int)>.	3376	The prototype of the C<function> must be C<void (*)(ev::TYPE &w, int)>.
2550		3377
2551	See the method-C<set> above for more details.	3378	See the method-C<set> above for more details.
2552		3379
2553	Example:	3380	Example: Use a plain function as callback.
2554		3381
2555	static void io_cb (ev::io &w, int revents) { }	3382	static void io_cb (ev::io &w, int revents) { }
2556	iow.set <io_cb> ();	3383	iow.set <io_cb> ();
2557		3384
2558	=item w->set (struct ev_loop *)	3385	=item w->set (loop)
2559		3386
2560	Associates a different C<struct ev_loop> with this watcher. You can only	3387	Associates a different C<struct ev_loop> with this watcher. You can only
2561	do this when the watcher is inactive (and not pending either).	3388	do this when the watcher is inactive (and not pending either).
2562		3389
2563	=item w->set ([arguments])	3390	=item w->set ([arguments])
…		…
2596	Example: Define a class with an IO and idle watcher, start one of them in	3423	Example: Define a class with an IO and idle watcher, start one of them in
2597	the constructor.	3424	the constructor.
2598		3425
2599	class myclass	3426	class myclass
2600	{	3427	{
2601	ev::io io; void io_cb (ev::io &w, int revents);	3428	ev::io io ; void io_cb (ev::io &w, int revents);
2602	ev:idle idle void idle_cb (ev::idle &w, int revents);	3429	ev::idle idle; void idle_cb (ev::idle &w, int revents);
2603		3430
2604	myclass (int fd)	3431	myclass (int fd)
2605	{	3432	{
2606	io .set <myclass, &myclass::io_cb > (this);	3433	io .set <myclass, &myclass::io_cb > (this);
2607	idle.set <myclass, &myclass::idle_cb> (this);	3434	idle.set <myclass, &myclass::idle_cb> (this);
…		…
2623	=item Perl	3450	=item Perl
2624		3451
2625	The EV module implements the full libev API and is actually used to test	3452	The EV module implements the full libev API and is actually used to test
2626	libev. EV is developed together with libev. Apart from the EV core module,	3453	libev. EV is developed together with libev. Apart from the EV core module,
2627	there are additional modules that implement libev-compatible interfaces	3454	there are additional modules that implement libev-compatible interfaces
2628	to C<libadns> (C<EV::ADNS>), C<Net::SNMP> (C<Net::SNMP::EV>) and the	3455	to C<libadns> (C<EV::ADNS>, but C<AnyEvent::DNS> is preferred nowadays),
2629	C<libglib> event core (C<Glib::EV> and C<EV::Glib>).	3456	C<Net::SNMP> (C<Net::SNMP::EV>) and the C<libglib> event core (C<Glib::EV>
		3457	and C<EV::Glib>).
2630		3458
2631	It can be found and installed via CPAN, its homepage is at	3459	It can be found and installed via CPAN, its homepage is at
2632	L<http://software.schmorp.de/pkg/EV>.	3460	L<http://software.schmorp.de/pkg/EV>.
2633		3461
2634	=item Python	3462	=item Python
2635		3463
2636	Python bindings can be found at L<http://code.google.com/p/pyev/>. It	3464	Python bindings can be found at L<http://code.google.com/p/pyev/>. It
2637	seems to be quite complete and well-documented. Note, however, that the	3465	seems to be quite complete and well-documented.
2638	patch they require for libev is outright dangerous as it breaks the ABI
2639	for everybody else, and therefore, should never be applied in an installed
2640	libev (if python requires an incompatible ABI then it needs to embed
2641	libev).
2642		3466
2643	=item Ruby	3467	=item Ruby
2644		3468
2645	Tony Arcieri has written a ruby extension that offers access to a subset	3469	Tony Arcieri has written a ruby extension that offers access to a subset
2646	of the libev API and adds file handle abstractions, asynchronous DNS and	3470	of the libev API and adds file handle abstractions, asynchronous DNS and
2647	more on top of it. It can be found via gem servers. Its homepage is at	3471	more on top of it. It can be found via gem servers. Its homepage is at
2648	L<http://rev.rubyforge.org/>.	3472	L<http://rev.rubyforge.org/>.
2649		3473
		3474	Roger Pack reports that using the link order C<-lws2_32 -lmsvcrt-ruby-190>
		3475	makes rev work even on mingw.
		3476
		3477	=item Haskell
		3478
		3479	A haskell binding to libev is available at
		3480	L<http://hackage.haskell.org/cgi-bin/hackage-scripts/package/hlibev>.
		3481
2650	=item D	3482	=item D
2651		3483
2652	Leandro Lucarella has written a D language binding (F<ev.d>) for libev, to	3484	Leandro Lucarella has written a D language binding (F<ev.d>) for libev, to
2653	be found at L<http://git.llucax.com.ar/?p=software/ev.d.git;a=summary>.	3485	be found at L<http://proj.llucax.com.ar/wiki/evd>.
		3486
		3487	=item Ocaml
		3488
		3489	Erkki Seppala has written Ocaml bindings for libev, to be found at
		3490	L<http://modeemi.cs.tut.fi/~flux/software/ocaml-ev/>.
		3491
		3492	=item Lua
		3493
		3494	Brian Maher has written a partial interface to libev for lua (at the
		3495	time of this writing, only C<ev_io> and C<ev_timer>), to be found at
		3496	L<http://github.com/brimworks/lua-ev>.
2654		3497
2655	=back	3498	=back
2656		3499
2657		3500
2658	=head1 MACRO MAGIC	3501	=head1 MACRO MAGIC
…		…
2759		3602
2760	#define EV_STANDALONE 1	3603	#define EV_STANDALONE 1
2761	#include "ev.h"	3604	#include "ev.h"
2762		3605
2763	Both header files and implementation files can be compiled with a C++	3606	Both header files and implementation files can be compiled with a C++
2764	compiler (at least, thats a stated goal, and breakage will be treated	3607	compiler (at least, that's a stated goal, and breakage will be treated
2765	as a bug).	3608	as a bug).
2766		3609
2767	You need the following files in your source tree, or in a directory	3610	You need the following files in your source tree, or in a directory
2768	in your include path (e.g. in libev/ when using -Ilibev):	3611	in your include path (e.g. in libev/ when using -Ilibev):
2769		3612
…		…
2812	libev.m4	3655	libev.m4
2813		3656
2814	=head2 PREPROCESSOR SYMBOLS/MACROS	3657	=head2 PREPROCESSOR SYMBOLS/MACROS
2815		3658
2816	Libev can be configured via a variety of preprocessor symbols you have to	3659	Libev can be configured via a variety of preprocessor symbols you have to
2817	define before including any of its files. The default in the absence of	3660	define before including (or compiling) any of its files. The default in
2818	autoconf is noted for every option.	3661	the absence of autoconf is documented for every option.
		3662
		3663	Symbols marked with "(h)" do not change the ABI, and can have different
		3664	values when compiling libev vs. including F<ev.h>, so it is permissible
		3665	to redefine them before including F<ev.h> without breakign compatibility
		3666	to a compiled library. All other symbols change the ABI, which means all
		3667	users of libev and the libev code itself must be compiled with compatible
		3668	settings.
2819		3669
2820	=over 4	3670	=over 4
2821		3671
2822	=item EV_STANDALONE	3672	=item EV_STANDALONE (h)
2823		3673
2824	Must always be C<1> if you do not use autoconf configuration, which	3674	Must always be C<1> if you do not use autoconf configuration, which
2825	keeps libev from including F<config.h>, and it also defines dummy	3675	keeps libev from including F<config.h>, and it also defines dummy
2826	implementations for some libevent functions (such as logging, which is not	3676	implementations for some libevent functions (such as logging, which is not
2827	supported). It will also not define any of the structs usually found in	3677	supported). It will also not define any of the structs usually found in
2828	F<event.h> that are not directly supported by the libev core alone.	3678	F<event.h> that are not directly supported by the libev core alone.
2829		3679
		3680	In standalone mode, libev will still try to automatically deduce the
		3681	configuration, but has to be more conservative.
		3682
2830	=item EV_USE_MONOTONIC	3683	=item EV_USE_MONOTONIC
2831		3684
2832	If defined to be C<1>, libev will try to detect the availability of the	3685	If defined to be C<1>, libev will try to detect the availability of the
2833	monotonic clock option at both compile time and runtime. Otherwise no use	3686	monotonic clock option at both compile time and runtime. Otherwise no
2834	of the monotonic clock option will be attempted. If you enable this, you	3687	use of the monotonic clock option will be attempted. If you enable this,
2835	usually have to link against librt or something similar. Enabling it when	3688	you usually have to link against librt or something similar. Enabling it
2836	the functionality isn't available is safe, though, although you have	3689	when the functionality isn't available is safe, though, although you have
2837	to make sure you link against any libraries where the C<clock_gettime>	3690	to make sure you link against any libraries where the C<clock_gettime>
2838	function is hiding in (often F<-lrt>).	3691	function is hiding in (often F<-lrt>). See also C<EV_USE_CLOCK_SYSCALL>.
2839		3692
2840	=item EV_USE_REALTIME	3693	=item EV_USE_REALTIME
2841		3694
2842	If defined to be C<1>, libev will try to detect the availability of the	3695	If defined to be C<1>, libev will try to detect the availability of the
2843	real-time clock option at compile time (and assume its availability at	3696	real-time clock option at compile time (and assume its availability
2844	runtime if successful). Otherwise no use of the real-time clock option will	3697	at runtime if successful). Otherwise no use of the real-time clock
2845	be attempted. This effectively replaces C<gettimeofday> by C<clock_get	3698	option will be attempted. This effectively replaces C<gettimeofday>
2846	(CLOCK_REALTIME, ...)> and will not normally affect correctness. See the	3699	by C<clock_get (CLOCK_REALTIME, ...)> and will not normally affect
2847	note about libraries in the description of C<EV_USE_MONOTONIC>, though.	3700	correctness. See the note about libraries in the description of
		3701	C<EV_USE_MONOTONIC>, though. Defaults to the opposite value of
		3702	C<EV_USE_CLOCK_SYSCALL>.
		3703
		3704	=item EV_USE_CLOCK_SYSCALL
		3705
		3706	If defined to be C<1>, libev will try to use a direct syscall instead
		3707	of calling the system-provided C<clock_gettime> function. This option
		3708	exists because on GNU/Linux, C<clock_gettime> is in C<librt>, but C<librt>
		3709	unconditionally pulls in C<libpthread>, slowing down single-threaded
		3710	programs needlessly. Using a direct syscall is slightly slower (in
		3711	theory), because no optimised vdso implementation can be used, but avoids
		3712	the pthread dependency. Defaults to C<1> on GNU/Linux with glibc 2.x or
		3713	higher, as it simplifies linking (no need for C<-lrt>).
2848		3714
2849	=item EV_USE_NANOSLEEP	3715	=item EV_USE_NANOSLEEP
2850		3716
2851	If defined to be C<1>, libev will assume that C<nanosleep ()> is available	3717	If defined to be C<1>, libev will assume that C<nanosleep ()> is available
2852	and will use it for delays. Otherwise it will use C<select ()>.	3718	and will use it for delays. Otherwise it will use C<select ()>.
…		…
2868		3734
2869	=item EV_SELECT_USE_FD_SET	3735	=item EV_SELECT_USE_FD_SET
2870		3736
2871	If defined to C<1>, then the select backend will use the system C<fd_set>	3737	If defined to C<1>, then the select backend will use the system C<fd_set>
2872	structure. This is useful if libev doesn't compile due to a missing	3738	structure. This is useful if libev doesn't compile due to a missing
2873	C<NFDBITS> or C<fd_mask> definition or it mis-guesses the bitset layout on	3739	C<NFDBITS> or C<fd_mask> definition or it mis-guesses the bitset layout
2874	exotic systems. This usually limits the range of file descriptors to some	3740	on exotic systems. This usually limits the range of file descriptors to
2875	low limit such as 1024 or might have other limitations (winsocket only	3741	some low limit such as 1024 or might have other limitations (winsocket
2876	allows 64 sockets). The C<FD_SETSIZE> macro, set before compilation, might	3742	only allows 64 sockets). The C<FD_SETSIZE> macro, set before compilation,
2877	influence the size of the C<fd_set> used.	3743	configures the maximum size of the C<fd_set>.
2878		3744
2879	=item EV_SELECT_IS_WINSOCKET	3745	=item EV_SELECT_IS_WINSOCKET
2880		3746
2881	When defined to C<1>, the select backend will assume that	3747	When defined to C<1>, the select backend will assume that
2882	select/socket/connect etc. don't understand file descriptors but	3748	select/socket/connect etc. don't understand file descriptors but
…		…
2884	be used is the winsock select). This means that it will call	3750	be used is the winsock select). This means that it will call
2885	C<_get_osfhandle> on the fd to convert it to an OS handle. Otherwise,	3751	C<_get_osfhandle> on the fd to convert it to an OS handle. Otherwise,
2886	it is assumed that all these functions actually work on fds, even	3752	it is assumed that all these functions actually work on fds, even
2887	on win32. Should not be defined on non-win32 platforms.	3753	on win32. Should not be defined on non-win32 platforms.
2888		3754
2889	=item EV_FD_TO_WIN32_HANDLE	3755	=item EV_FD_TO_WIN32_HANDLE(fd)
2890		3756
2891	If C<EV_SELECT_IS_WINSOCKET> is enabled, then libev needs a way to map	3757	If C<EV_SELECT_IS_WINSOCKET> is enabled, then libev needs a way to map
2892	file descriptors to socket handles. When not defining this symbol (the	3758	file descriptors to socket handles. When not defining this symbol (the
2893	default), then libev will call C<_get_osfhandle>, which is usually	3759	default), then libev will call C<_get_osfhandle>, which is usually
2894	correct. In some cases, programs use their own file descriptor management,	3760	correct. In some cases, programs use their own file descriptor management,
2895	in which case they can provide this function to map fds to socket handles.	3761	in which case they can provide this function to map fds to socket handles.
		3762
		3763	=item EV_WIN32_HANDLE_TO_FD(handle)
		3764
		3765	If C<EV_SELECT_IS_WINSOCKET> then libev maps handles to file descriptors
		3766	using the standard C<_open_osfhandle> function. For programs implementing
		3767	their own fd to handle mapping, overwriting this function makes it easier
		3768	to do so. This can be done by defining this macro to an appropriate value.
		3769
		3770	=item EV_WIN32_CLOSE_FD(fd)
		3771
		3772	If programs implement their own fd to handle mapping on win32, then this
		3773	macro can be used to override the C<close> function, useful to unregister
		3774	file descriptors again. Note that the replacement function has to close
		3775	the underlying OS handle.
2896		3776
2897	=item EV_USE_POLL	3777	=item EV_USE_POLL
2898		3778
2899	If defined to be C<1>, libev will compile in support for the C<poll>(2)	3779	If defined to be C<1>, libev will compile in support for the C<poll>(2)
2900	backend. Otherwise it will be enabled on non-win32 platforms. It	3780	backend. Otherwise it will be enabled on non-win32 platforms. It
…		…
2947	as well as for signal and thread safety in C<ev_async> watchers.	3827	as well as for signal and thread safety in C<ev_async> watchers.
2948		3828
2949	In the absence of this define, libev will use C<sig_atomic_t volatile>	3829	In the absence of this define, libev will use C<sig_atomic_t volatile>
2950	(from F<signal.h>), which is usually good enough on most platforms.	3830	(from F<signal.h>), which is usually good enough on most platforms.
2951		3831
2952	=item EV_H	3832	=item EV_H (h)
2953		3833
2954	The name of the F<ev.h> header file used to include it. The default if	3834	The name of the F<ev.h> header file used to include it. The default if
2955	undefined is C<"ev.h"> in F<event.h>, F<ev.c> and F<ev++.h>. This can be	3835	undefined is C<"ev.h"> in F<event.h>, F<ev.c> and F<ev++.h>. This can be
2956	used to virtually rename the F<ev.h> header file in case of conflicts.	3836	used to virtually rename the F<ev.h> header file in case of conflicts.
2957		3837
2958	=item EV_CONFIG_H	3838	=item EV_CONFIG_H (h)
2959		3839
2960	If C<EV_STANDALONE> isn't C<1>, this variable can be used to override	3840	If C<EV_STANDALONE> isn't C<1>, this variable can be used to override
2961	F<ev.c>'s idea of where to find the F<config.h> file, similarly to	3841	F<ev.c>'s idea of where to find the F<config.h> file, similarly to
2962	C<EV_H>, above.	3842	C<EV_H>, above.
2963		3843
2964	=item EV_EVENT_H	3844	=item EV_EVENT_H (h)
2965		3845
2966	Similarly to C<EV_H>, this macro can be used to override F<event.c>'s idea	3846	Similarly to C<EV_H>, this macro can be used to override F<event.c>'s idea
2967	of how the F<event.h> header can be found, the default is C<"event.h">.	3847	of how the F<event.h> header can be found, the default is C<"event.h">.
2968		3848
2969	=item EV_PROTOTYPES	3849	=item EV_PROTOTYPES (h)
2970		3850
2971	If defined to be C<0>, then F<ev.h> will not define any function	3851	If defined to be C<0>, then F<ev.h> will not define any function
2972	prototypes, but still define all the structs and other symbols. This is	3852	prototypes, but still define all the structs and other symbols. This is
2973	occasionally useful if you want to provide your own wrapper functions	3853	occasionally useful if you want to provide your own wrapper functions
2974	around libev functions.	3854	around libev functions.
…		…
2993	When doing priority-based operations, libev usually has to linearly search	3873	When doing priority-based operations, libev usually has to linearly search
2994	all the priorities, so having many of them (hundreds) uses a lot of space	3874	all the priorities, so having many of them (hundreds) uses a lot of space
2995	and time, so using the defaults of five priorities (-2 .. +2) is usually	3875	and time, so using the defaults of five priorities (-2 .. +2) is usually
2996	fine.	3876	fine.
2997		3877
2998	If your embedding application does not need any priorities, defining these both to	3878	If your embedding application does not need any priorities, defining these
2999	C<0> will save some memory and CPU.	3879	both to C<0> will save some memory and CPU.
3000		3880
3001	=item EV_PERIODIC_ENABLE	3881	=item EV_PERIODIC_ENABLE, EV_IDLE_ENABLE, EV_EMBED_ENABLE, EV_STAT_ENABLE,
		3882	EV_PREPARE_ENABLE, EV_CHECK_ENABLE, EV_FORK_ENABLE, EV_SIGNAL_ENABLE,
		3883	EV_ASYNC_ENABLE, EV_CHILD_ENABLE.
3002		3884
3003	If undefined or defined to be C<1>, then periodic timers are supported. If	3885	If undefined or defined to be C<1> (and the platform supports it), then
3004	defined to be C<0>, then they are not. Disabling them saves a few kB of	3886	the respective watcher type is supported. If defined to be C<0>, then it
3005	code.	3887	is not. Disabling watcher types mainly saves codesize.
3006		3888
3007	=item EV_IDLE_ENABLE	3889	=item EV_FEATURES
3008
3009	If undefined or defined to be C<1>, then idle watchers are supported. If
3010	defined to be C<0>, then they are not. Disabling them saves a few kB of
3011	code.
3012
3013	=item EV_EMBED_ENABLE
3014
3015	If undefined or defined to be C<1>, then embed watchers are supported. If
3016	defined to be C<0>, then they are not.
3017
3018	=item EV_STAT_ENABLE
3019
3020	If undefined or defined to be C<1>, then stat watchers are supported. If
3021	defined to be C<0>, then they are not.
3022
3023	=item EV_FORK_ENABLE
3024
3025	If undefined or defined to be C<1>, then fork watchers are supported. If
3026	defined to be C<0>, then they are not.
3027
3028	=item EV_ASYNC_ENABLE
3029
3030	If undefined or defined to be C<1>, then async watchers are supported. If
3031	defined to be C<0>, then they are not.
3032
3033	=item EV_MINIMAL
3034		3890
3035	If you need to shave off some kilobytes of code at the expense of some	3891	If you need to shave off some kilobytes of code at the expense of some
3036	speed, define this symbol to C<1>. Currently this is used to override some	3892	speed (but with the full API), you can define this symbol to request
3037	inlining decisions, saves roughly 30% code size on amd64. It also selects a	3893	certain subsets of functionality. The default is to enable all features
3038	much smaller 2-heap for timer management over the default 4-heap.	3894	that can be enabled on the platform.
		3895
		3896	Note that using autoconf will usually override most of the features, so
		3897	using this symbol makes sense mostly when embedding libev.
		3898
		3899	A typical way to use this symbol is to define it to C<0> (or to a bitset
		3900	with some broad features you want) and then selectively re-enable
		3901	additional parts you want, for example if you want everything minimal,
		3902	but multiple event loop support, async and child watchers and the poll
		3903	backend, use this:
		3904
		3905	#define EV_FEATURES 0
		3906	#define EV_MULTIPLICITY 1
		3907	#define EV_USE_POLL 1
		3908	#define EV_CHILD_ENABLE 1
		3909	#define EV_ASYNC_ENABLE 1
		3910
		3911	The actual value is a bitset, it can be a combination of the following
		3912	values:
		3913
		3914	=over 4
		3915
		3916	=item C<1> - faster/larger code
		3917
		3918	Use larger code to speed up some operations.
		3919
		3920	Currently this is used to override some inlining decisions (enlarging the roughly
		3921	30% code size on amd64.
		3922
		3923	When optimising for size, use of compiler flags such as C<-Os> with
		3924	gcc recommended, as well as C<-DNDEBUG>, as libev contains a number of
		3925	assertions.
		3926
		3927	=item C<2> - faster/larger data structures
		3928
		3929	Replaces the small 2-heap for timer management by a faster 4-heap, larger
		3930	hash table sizes and so on. This will usually further increase codesize
		3931	and can additionally have an effect on the size of data structures at
		3932	runtime.
		3933
		3934	=item C<4> - full API configuration
		3935
		3936	This enables priorities (sets C<EV_MAXPRI>=2 and C<EV_MINPRI>=-2), and
		3937	enables multiplicity (C<EV_MULTIPLICITY>=1).
		3938
		3939	It also enables a lot of the "lesser used" core API functions. See C<ev.h>
		3940	for details on which parts of the API are still available without this
		3941	feature, and do not complain if this subset changes over time.
		3942
		3943	=item C<8> - enable all optional watcher types
		3944
		3945	Enables all optional watcher types. If you want to selectively enable
		3946	only some watcher types other than I/O and timers (e.g. prepare,
		3947	embed, async, child...) you can enable them manually by defining
		3948	C<EV_watchertype_ENABLE> to C<1> instead.
		3949
		3950	=item C<16> - enable all backends
		3951
		3952	This enables all backends - without this feature, you need to enable at
		3953	least one backend manually (C<EV_USE_SELECT> is a good choice).
		3954
		3955	=item C<32> - enable OS-specific "helper" APIs
		3956
		3957	Enable inotify, eventfd, signalfd and similar OS-specific helper APIs by
		3958	default.
		3959
		3960	=back
		3961
		3962	Compiling with C<gcc -Os -DEV_STANDALONE -DEV_USE_EPOLL=1 -DEV_FEATURES=0>
		3963	reduces the compiled size of libev from 24.7Kb to 6.5Kb on my GNU/Linux
		3964	amd64 system, while still giving you I/O watchers, timers and monotonic
		3965	clock support.
		3966
		3967	With an intelligent-enough linker (gcc+binutils are intelligent enough
		3968	when you use C<-Wl,--gc-sections -ffunction-sections>) functions unused by
		3969	your program might be left out as well - a binary starting a timer and an
		3970	I/O watcher then might come out at only 5Kb.
		3971
		3972	=item EV_AVOID_STDIO
		3973
		3974	If this is set to C<1> at compiletime, then libev will avoid using stdio
		3975	functions (printf, scanf, perror etc.). This will increase the codesize
		3976	somewhat, but if your program doesn't otherwise depend on stdio and your
		3977	libc allows it, this avoids linking in the stdio library which is quite
		3978	big.
		3979
		3980	Note that error messages might become less precise when this option is
		3981	enabled.
		3982
		3983	=item EV_NSIG
		3984
		3985	The highest supported signal number, +1 (or, the number of
		3986	signals): Normally, libev tries to deduce the maximum number of signals
		3987	automatically, but sometimes this fails, in which case it can be
		3988	specified. Also, using a lower number than detected (C<32> should be
		3989	good for about any system in existance) can save some memory, as libev
		3990	statically allocates some 12-24 bytes per signal number.
3039		3991
3040	=item EV_PID_HASHSIZE	3992	=item EV_PID_HASHSIZE
3041		3993
3042	C<ev_child> watchers use a small hash table to distribute workload by	3994	C<ev_child> watchers use a small hash table to distribute workload by
3043	pid. The default size is C<16> (or C<1> with C<EV_MINIMAL>), usually more	3995	pid. The default size is C<16> (or C<1> with C<EV_FEATURES> disabled),
3044	than enough. If you need to manage thousands of children you might want to	3996	usually more than enough. If you need to manage thousands of children you
3045	increase this value (I<must> be a power of two).	3997	might want to increase this value (I<must> be a power of two).
3046		3998
3047	=item EV_INOTIFY_HASHSIZE	3999	=item EV_INOTIFY_HASHSIZE
3048		4000
3049	C<ev_stat> watchers use a small hash table to distribute workload by	4001	C<ev_stat> watchers use a small hash table to distribute workload by
3050	inotify watch id. The default size is C<16> (or C<1> with C<EV_MINIMAL>),	4002	inotify watch id. The default size is C<16> (or C<1> with C<EV_FEATURES>
3051	usually more than enough. If you need to manage thousands of C<ev_stat>	4003	disabled), usually more than enough. If you need to manage thousands of
3052	watchers you might want to increase this value (I<must> be a power of	4004	C<ev_stat> watchers you might want to increase this value (I<must> be a
3053	two).	4005	power of two).
3054		4006
3055	=item EV_USE_4HEAP	4007	=item EV_USE_4HEAP
3056		4008
3057	Heaps are not very cache-efficient. To improve the cache-efficiency of the	4009	Heaps are not very cache-efficient. To improve the cache-efficiency of the
3058	timer and periodics heap, libev uses a 4-heap when this symbol is defined	4010	timer and periodics heaps, libev uses a 4-heap when this symbol is defined
3059	to C<1>. The 4-heap uses more complicated (longer) code but has	4011	to C<1>. The 4-heap uses more complicated (longer) code but has noticeably
3060	noticeably faster performance with many (thousands) of watchers.	4012	faster performance with many (thousands) of watchers.
3061		4013
3062	The default is C<1> unless C<EV_MINIMAL> is set in which case it is C<0>	4014	The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it
3063	(disabled).	4015	will be C<0>.
3064		4016
3065	=item EV_HEAP_CACHE_AT	4017	=item EV_HEAP_CACHE_AT
3066		4018
3067	Heaps are not very cache-efficient. To improve the cache-efficiency of the	4019	Heaps are not very cache-efficient. To improve the cache-efficiency of the
3068	timer and periodics heap, libev can cache the timestamp (I<at>) within	4020	timer and periodics heaps, libev can cache the timestamp (I<at>) within
3069	the heap structure (selected by defining C<EV_HEAP_CACHE_AT> to C<1>),	4021	the heap structure (selected by defining C<EV_HEAP_CACHE_AT> to C<1>),
3070	which uses 8-12 bytes more per watcher and a few hundred bytes more code,	4022	which uses 8-12 bytes more per watcher and a few hundred bytes more code,
3071	but avoids random read accesses on heap changes. This improves performance	4023	but avoids random read accesses on heap changes. This improves performance
3072	noticeably with with many (hundreds) of watchers.	4024	noticeably with many (hundreds) of watchers.
3073		4025
3074	The default is C<1> unless C<EV_MINIMAL> is set in which case it is C<0>	4026	The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it
3075	(disabled).	4027	will be C<0>.
3076		4028
3077	=item EV_VERIFY	4029	=item EV_VERIFY
3078		4030
3079	Controls how much internal verification (see C<ev_loop_verify ()>) will	4031	Controls how much internal verification (see C<ev_loop_verify ()>) will
3080	be done: If set to C<0>, no internal verification code will be compiled	4032	be done: If set to C<0>, no internal verification code will be compiled
…		…
3082	called. If set to C<2>, then the internal verification code will be	4034	called. If set to C<2>, then the internal verification code will be
3083	called once per loop, which can slow down libev. If set to C<3>, then the	4035	called once per loop, which can slow down libev. If set to C<3>, then the
3084	verification code will be called very frequently, which will slow down	4036	verification code will be called very frequently, which will slow down
3085	libev considerably.	4037	libev considerably.
3086		4038
3087	The default is C<1>, unless C<EV_MINIMAL> is set, in which case it will be	4039	The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it
3088	C<0.>	4040	will be C<0>.
3089		4041
3090	=item EV_COMMON	4042	=item EV_COMMON
3091		4043
3092	By default, all watchers have a C<void *data> member. By redefining	4044	By default, all watchers have a C<void *data> member. By redefining
3093	this macro to a something else you can include more and other types of	4045	this macro to a something else you can include more and other types of
…		…
3110	and the way callbacks are invoked and set. Must expand to a struct member	4062	and the way callbacks are invoked and set. Must expand to a struct member
3111	definition and a statement, respectively. See the F<ev.h> header file for	4063	definition and a statement, respectively. See the F<ev.h> header file for
3112	their default definitions. One possible use for overriding these is to	4064	their default definitions. One possible use for overriding these is to
3113	avoid the C<struct ev_loop *> as first argument in all cases, or to use	4065	avoid the C<struct ev_loop *> as first argument in all cases, or to use
3114	method calls instead of plain function calls in C++.	4066	method calls instead of plain function calls in C++.
		4067
		4068	=back
3115		4069
3116	=head2 EXPORTED API SYMBOLS	4070	=head2 EXPORTED API SYMBOLS
3117		4071
3118	If you need to re-export the API (e.g. via a DLL) and you need a list of	4072	If you need to re-export the API (e.g. via a DLL) and you need a list of
3119	exported symbols, you can use the provided F<Symbol.*> files which list	4073	exported symbols, you can use the provided F<Symbol.*> files which list
…		…
3149	file.	4103	file.
3150		4104
3151	The usage in rxvt-unicode is simpler. It has a F<ev_cpp.h> header file	4105	The usage in rxvt-unicode is simpler. It has a F<ev_cpp.h> header file
3152	that everybody includes and which overrides some configure choices:	4106	that everybody includes and which overrides some configure choices:
3153		4107
3154	#define EV_MINIMAL 1	4108	#define EV_FEATURES 0
3155	#define EV_USE_POLL 0	4109	#define EV_USE_SELECT 1
3156	#define EV_MULTIPLICITY 0
3157	#define EV_PERIODIC_ENABLE 0
3158	#define EV_STAT_ENABLE 0
3159	#define EV_FORK_ENABLE 0
3160	#define EV_CONFIG_H <config.h>	4110	#define EV_CONFIG_H <config.h>
3161	#define EV_MINPRI 0
3162	#define EV_MAXPRI 0
3163		4111
3164	#include "ev++.h"	4112	#include "ev++.h"
3165		4113
3166	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:	4114	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:
3167		4115
3168	#include "ev_cpp.h"	4116	#include "ev_cpp.h"
3169	#include "ev.c"	4117	#include "ev.c"
3170		4118
		4119	=head1 INTERACTION WITH OTHER PROGRAMS OR LIBRARIES
3171		4120
3172	=head1 THREADS AND COROUTINES	4121	=head2 THREADS AND COROUTINES
3173		4122
3174	=head2 THREADS	4123	=head3 THREADS
3175		4124
3176	Libev itself is completely thread-safe, but it uses no locking. This	4125	All libev functions are reentrant and thread-safe unless explicitly
		4126	documented otherwise, but libev implements no locking itself. This means
3177	means that you can use as many loops as you want in parallel, as long as	4127	that you can use as many loops as you want in parallel, as long as there
3178	only one thread ever calls into one libev function with the same loop	4128	are no concurrent calls into any libev function with the same loop
3179	parameter.	4129	parameter (C<ev_default_*> calls have an implicit default loop parameter,
		4130	of course): libev guarantees that different event loops share no data
		4131	structures that need any locking.
3180		4132
3181	Or put differently: calls with different loop parameters can be done in	4133	Or to put it differently: calls with different loop parameters can be done
3182	parallel from multiple threads, calls with the same loop parameter must be	4134	concurrently from multiple threads, calls with the same loop parameter
3183	done serially (but can be done from different threads, as long as only one	4135	must be done serially (but can be done from different threads, as long as
3184	thread ever is inside a call at any point in time, e.g. by using a mutex	4136	only one thread ever is inside a call at any point in time, e.g. by using
3185	per loop).	4137	a mutex per loop).
3186		4138
3187	If you want to know which design is best for your problem, then I cannot	4139	Specifically to support threads (and signal handlers), libev implements
		4140	so-called C<ev_async> watchers, which allow some limited form of
		4141	concurrency on the same event loop, namely waking it up "from the
		4142	outside".
		4143
		4144	If you want to know which design (one loop, locking, or multiple loops
		4145	without or something else still) is best for your problem, then I cannot
3188	help you but by giving some generic advice:	4146	help you, but here is some generic advice:
3189		4147
3190	=over 4	4148	=over 4
3191		4149
3192	=item * most applications have a main thread: use the default libev loop	4150	=item * most applications have a main thread: use the default libev loop
3193	in that thread, or create a separate thread running only the default loop.	4151	in that thread, or create a separate thread running only the default loop.
…		…
3205		4163
3206	Choosing a model is hard - look around, learn, know that usually you can do	4164	Choosing a model is hard - look around, learn, know that usually you can do
3207	better than you currently do :-)	4165	better than you currently do :-)
3208		4166
3209	=item * often you need to talk to some other thread which blocks in the	4167	=item * often you need to talk to some other thread which blocks in the
		4168	event loop.
		4169
3210	event loop - C<ev_async> watchers can be used to wake them up from other	4170	C<ev_async> watchers can be used to wake them up from other threads safely
3211	threads safely (or from signal contexts...).	4171	(or from signal contexts...).
		4172
		4173	An example use would be to communicate signals or other events that only
		4174	work in the default loop by registering the signal watcher with the
		4175	default loop and triggering an C<ev_async> watcher from the default loop
		4176	watcher callback into the event loop interested in the signal.
3212		4177
3213	=back	4178	=back
3214		4179
		4180	=head4 THREAD LOCKING EXAMPLE
		4181
		4182	Here is a fictitious example of how to run an event loop in a different
		4183	thread than where callbacks are being invoked and watchers are
		4184	created/added/removed.
		4185
		4186	For a real-world example, see the C<EV::Loop::Async> perl module,
		4187	which uses exactly this technique (which is suited for many high-level
		4188	languages).
		4189
		4190	The example uses a pthread mutex to protect the loop data, a condition
		4191	variable to wait for callback invocations, an async watcher to notify the
		4192	event loop thread and an unspecified mechanism to wake up the main thread.
		4193
		4194	First, you need to associate some data with the event loop:
		4195
		4196	typedef struct {
		4197	mutex_t lock; /* global loop lock */
		4198	ev_async async_w;
		4199	thread_t tid;
		4200	cond_t invoke_cv;
		4201	} userdata;
		4202
		4203	void prepare_loop (EV_P)
		4204	{
		4205	// for simplicity, we use a static userdata struct.
		4206	static userdata u;
		4207
		4208	ev_async_init (&u->async_w, async_cb);
		4209	ev_async_start (EV_A_ &u->async_w);
		4210
		4211	pthread_mutex_init (&u->lock, 0);
		4212	pthread_cond_init (&u->invoke_cv, 0);
		4213
		4214	// now associate this with the loop
		4215	ev_set_userdata (EV_A_ u);
		4216	ev_set_invoke_pending_cb (EV_A_ l_invoke);
		4217	ev_set_loop_release_cb (EV_A_ l_release, l_acquire);
		4218
		4219	// then create the thread running ev_loop
		4220	pthread_create (&u->tid, 0, l_run, EV_A);
		4221	}
		4222
		4223	The callback for the C<ev_async> watcher does nothing: the watcher is used
		4224	solely to wake up the event loop so it takes notice of any new watchers
		4225	that might have been added:
		4226
		4227	static void
		4228	async_cb (EV_P_ ev_async *w, int revents)
		4229	{
		4230	// just used for the side effects
		4231	}
		4232
		4233	The C<l_release> and C<l_acquire> callbacks simply unlock/lock the mutex
		4234	protecting the loop data, respectively.
		4235
		4236	static void
		4237	l_release (EV_P)
		4238	{
		4239	userdata *u = ev_userdata (EV_A);
		4240	pthread_mutex_unlock (&u->lock);
		4241	}
		4242
		4243	static void
		4244	l_acquire (EV_P)
		4245	{
		4246	userdata *u = ev_userdata (EV_A);
		4247	pthread_mutex_lock (&u->lock);
		4248	}
		4249
		4250	The event loop thread first acquires the mutex, and then jumps straight
		4251	into C<ev_loop>:
		4252
		4253	void *
		4254	l_run (void *thr_arg)
		4255	{
		4256	struct ev_loop *loop = (struct ev_loop *)thr_arg;
		4257
		4258	l_acquire (EV_A);
		4259	pthread_setcanceltype (PTHREAD_CANCEL_ASYNCHRONOUS, 0);
		4260	ev_loop (EV_A_ 0);
		4261	l_release (EV_A);
		4262
		4263	return 0;
		4264	}
		4265
		4266	Instead of invoking all pending watchers, the C<l_invoke> callback will
		4267	signal the main thread via some unspecified mechanism (signals? pipe
		4268	writes? C<Async::Interrupt>?) and then waits until all pending watchers
		4269	have been called (in a while loop because a) spurious wakeups are possible
		4270	and b) skipping inter-thread-communication when there are no pending
		4271	watchers is very beneficial):
		4272
		4273	static void
		4274	l_invoke (EV_P)
		4275	{
		4276	userdata *u = ev_userdata (EV_A);
		4277
		4278	while (ev_pending_count (EV_A))
		4279	{
		4280	wake_up_other_thread_in_some_magic_or_not_so_magic_way ();
		4281	pthread_cond_wait (&u->invoke_cv, &u->lock);
		4282	}
		4283	}
		4284
		4285	Now, whenever the main thread gets told to invoke pending watchers, it
		4286	will grab the lock, call C<ev_invoke_pending> and then signal the loop
		4287	thread to continue:
		4288
		4289	static void
		4290	real_invoke_pending (EV_P)
		4291	{
		4292	userdata *u = ev_userdata (EV_A);
		4293
		4294	pthread_mutex_lock (&u->lock);
		4295	ev_invoke_pending (EV_A);
		4296	pthread_cond_signal (&u->invoke_cv);
		4297	pthread_mutex_unlock (&u->lock);
		4298	}
		4299
		4300	Whenever you want to start/stop a watcher or do other modifications to an
		4301	event loop, you will now have to lock:
		4302
		4303	ev_timer timeout_watcher;
		4304	userdata *u = ev_userdata (EV_A);
		4305
		4306	ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);
		4307
		4308	pthread_mutex_lock (&u->lock);
		4309	ev_timer_start (EV_A_ &timeout_watcher);
		4310	ev_async_send (EV_A_ &u->async_w);
		4311	pthread_mutex_unlock (&u->lock);
		4312
		4313	Note that sending the C<ev_async> watcher is required because otherwise
		4314	an event loop currently blocking in the kernel will have no knowledge
		4315	about the newly added timer. By waking up the loop it will pick up any new
		4316	watchers in the next event loop iteration.
		4317
3215	=head2 COROUTINES	4318	=head3 COROUTINES
3216		4319
3217	Libev is much more accommodating to coroutines ("cooperative threads"):	4320	Libev is very accommodating to coroutines ("cooperative threads"):
3218	libev fully supports nesting calls to it's functions from different	4321	libev fully supports nesting calls to its functions from different
3219	coroutines (e.g. you can call C<ev_loop> on the same loop from two	4322	coroutines (e.g. you can call C<ev_loop> on the same loop from two
3220	different coroutines and switch freely between both coroutines running the	4323	different coroutines, and switch freely between both coroutines running
3221	loop, as long as you don't confuse yourself). The only exception is that	4324	the loop, as long as you don't confuse yourself). The only exception is
3222	you must not do this from C<ev_periodic> reschedule callbacks.	4325	that you must not do this from C<ev_periodic> reschedule callbacks.
3223		4326
3224	Care has been invested into making sure that libev does not keep local	4327	Care has been taken to ensure that libev does not keep local state inside
3225	state inside C<ev_loop>, and other calls do not usually allow coroutine	4328	C<ev_loop>, and other calls do not usually allow for coroutine switches as
3226	switches.	4329	they do not call any callbacks.
3227		4330
		4331	=head2 COMPILER WARNINGS
3228		4332
3229	=head1 COMPLEXITIES	4333	Depending on your compiler and compiler settings, you might get no or a
		4334	lot of warnings when compiling libev code. Some people are apparently
		4335	scared by this.
3230		4336
3231	In this section the complexities of (many of) the algorithms used inside	4337	However, these are unavoidable for many reasons. For one, each compiler
3232	libev will be explained. For complexity discussions about backends see the	4338	has different warnings, and each user has different tastes regarding
3233	documentation for C<ev_default_init>.	4339	warning options. "Warn-free" code therefore cannot be a goal except when
		4340	targeting a specific compiler and compiler-version.
3234		4341
3235	All of the following are about amortised time: If an array needs to be	4342	Another reason is that some compiler warnings require elaborate
3236	extended, libev needs to realloc and move the whole array, but this	4343	workarounds, or other changes to the code that make it less clear and less
3237	happens asymptotically never with higher number of elements, so O(1) might	4344	maintainable.
3238	mean it might do a lengthy realloc operation in rare cases, but on average
3239	it is much faster and asymptotically approaches constant time.
3240		4345
3241	=over 4	4346	And of course, some compiler warnings are just plain stupid, or simply
		4347	wrong (because they don't actually warn about the condition their message
		4348	seems to warn about). For example, certain older gcc versions had some
		4349	warnings that resulted an extreme number of false positives. These have
		4350	been fixed, but some people still insist on making code warn-free with
		4351	such buggy versions.
3242		4352
3243	=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers)	4353	While libev is written to generate as few warnings as possible,
		4354	"warn-free" code is not a goal, and it is recommended not to build libev
		4355	with any compiler warnings enabled unless you are prepared to cope with
		4356	them (e.g. by ignoring them). Remember that warnings are just that:
		4357	warnings, not errors, or proof of bugs.
3244		4358
3245	This means that, when you have a watcher that triggers in one hour and
3246	there are 100 watchers that would trigger before that then inserting will
3247	have to skip roughly seven (C<ld 100>) of these watchers.
3248		4359
3249	=item Changing timer/periodic watchers (by autorepeat or calling again): O(log skipped_other_timers)	4360	=head2 VALGRIND
3250		4361
3251	That means that changing a timer costs less than removing/adding them	4362	Valgrind has a special section here because it is a popular tool that is
3252	as only the relative motion in the event queue has to be paid for.	4363	highly useful. Unfortunately, valgrind reports are very hard to interpret.
3253		4364
3254	=item Starting io/check/prepare/idle/signal/child/fork/async watchers: O(1)	4365	If you think you found a bug (memory leak, uninitialised data access etc.)
		4366	in libev, then check twice: If valgrind reports something like:
3255		4367
3256	These just add the watcher into an array or at the head of a list.	4368	==2274== definitely lost: 0 bytes in 0 blocks.
		4369	==2274== possibly lost: 0 bytes in 0 blocks.
		4370	==2274== still reachable: 256 bytes in 1 blocks.
3257		4371
3258	=item Stopping check/prepare/idle/fork/async watchers: O(1)	4372	Then there is no memory leak, just as memory accounted to global variables
		4373	is not a memleak - the memory is still being referenced, and didn't leak.
3259		4374
3260	=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % EV_PID_HASHSIZE))	4375	Similarly, under some circumstances, valgrind might report kernel bugs
		4376	as if it were a bug in libev (e.g. in realloc or in the poll backend,
		4377	although an acceptable workaround has been found here), or it might be
		4378	confused.
3261		4379
3262	These watchers are stored in lists then need to be walked to find the	4380	Keep in mind that valgrind is a very good tool, but only a tool. Don't
3263	correct watcher to remove. The lists are usually short (you don't usually	4381	make it into some kind of religion.
3264	have many watchers waiting for the same fd or signal).
3265		4382
3266	=item Finding the next timer in each loop iteration: O(1)	4383	If you are unsure about something, feel free to contact the mailing list
		4384	with the full valgrind report and an explanation on why you think this
		4385	is a bug in libev (best check the archives, too :). However, don't be
		4386	annoyed when you get a brisk "this is no bug" answer and take the chance
		4387	of learning how to interpret valgrind properly.
3267		4388
3268	By virtue of using a binary or 4-heap, the next timer is always found at a	4389	If you need, for some reason, empty reports from valgrind for your project
3269	fixed position in the storage array.	4390	I suggest using suppression lists.
3270		4391
3271	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)
3272		4392
3273	A change means an I/O watcher gets started or stopped, which requires	4393	=head1 PORTABILITY NOTES
3274	libev to recalculate its status (and possibly tell the kernel, depending
3275	on backend and whether C<ev_io_set> was used).
3276		4394
3277	=item Activating one watcher (putting it into the pending state): O(1)
3278
3279	=item Priority handling: O(number_of_priorities)
3280
3281	Priorities are implemented by allocating some space for each
3282	priority. When doing priority-based operations, libev usually has to
3283	linearly search all the priorities, but starting/stopping and activating
3284	watchers becomes O(1) w.r.t. priority handling.
3285
3286	=item Sending an ev_async: O(1)
3287
3288	=item Processing ev_async_send: O(number_of_async_watchers)
3289
3290	=item Processing signals: O(max_signal_number)
3291
3292	Sending involves a system call I<iff> there were no other C<ev_async_send>
3293	calls in the current loop iteration. Checking for async and signal events
3294	involves iterating over all running async watchers or all signal numbers.
3295
3296	=back
3297
3298
3299	=head1 WIN32 PLATFORM LIMITATIONS AND WORKAROUNDS	4395	=head2 WIN32 PLATFORM LIMITATIONS AND WORKAROUNDS
3300		4396
3301	Win32 doesn't support any of the standards (e.g. POSIX) that libev	4397	Win32 doesn't support any of the standards (e.g. POSIX) that libev
3302	requires, and its I/O model is fundamentally incompatible with the POSIX	4398	requires, and its I/O model is fundamentally incompatible with the POSIX
3303	model. Libev still offers limited functionality on this platform in	4399	model. Libev still offers limited functionality on this platform in
3304	the form of the C<EVBACKEND_SELECT> backend, and only supports socket	4400	the form of the C<EVBACKEND_SELECT> backend, and only supports socket
…		…
3311	way (note also that glib is the slowest event library known to man).	4407	way (note also that glib is the slowest event library known to man).
3312		4408
3313	There is no supported compilation method available on windows except	4409	There is no supported compilation method available on windows except
3314	embedding it into other applications.	4410	embedding it into other applications.
3315		4411
		4412	Sensible signal handling is officially unsupported by Microsoft - libev
		4413	tries its best, but under most conditions, signals will simply not work.
		4414
3316	Not a libev limitation but worth mentioning: windows apparently doesn't	4415	Not a libev limitation but worth mentioning: windows apparently doesn't
3317	accept large writes: instead of resulting in a partial write, windows will	4416	accept large writes: instead of resulting in a partial write, windows will
3318	either accept everything or return C<ENOBUFS> if the buffer is too large,	4417	either accept everything or return C<ENOBUFS> if the buffer is too large,
3319	so make sure you only write small amounts into your sockets (less than a	4418	so make sure you only write small amounts into your sockets (less than a
3320	megabyte seems safe, but thsi apparently depends on the amount of memory	4419	megabyte seems safe, but this apparently depends on the amount of memory
3321	available).	4420	available).
3322		4421
3323	Due to the many, low, and arbitrary limits on the win32 platform and	4422	Due to the many, low, and arbitrary limits on the win32 platform and
3324	the abysmal performance of winsockets, using a large number of sockets	4423	the abysmal performance of winsockets, using a large number of sockets
3325	is not recommended (and not reasonable). If your program needs to use	4424	is not recommended (and not reasonable). If your program needs to use
3326	more than a hundred or so sockets, then likely it needs to use a totally	4425	more than a hundred or so sockets, then likely it needs to use a totally
3327	different implementation for windows, as libev offers the POSIX readiness	4426	different implementation for windows, as libev offers the POSIX readiness
3328	notification model, which cannot be implemented efficiently on windows	4427	notification model, which cannot be implemented efficiently on windows
3329	(Microsoft monopoly games).	4428	(due to Microsoft monopoly games).
		4429
		4430	A typical way to use libev under windows is to embed it (see the embedding
		4431	section for details) and use the following F<evwrap.h> header file instead
		4432	of F<ev.h>:
		4433
		4434	#define EV_STANDALONE /* keeps ev from requiring config.h */
		4435	#define EV_SELECT_IS_WINSOCKET 1 /* configure libev for windows select */
		4436
		4437	#include "ev.h"
		4438
		4439	And compile the following F<evwrap.c> file into your project (make sure
		4440	you do I<not> compile the F<ev.c> or any other embedded source files!):
		4441
		4442	#include "evwrap.h"
		4443	#include "ev.c"
3330		4444
3331	=over 4	4445	=over 4
3332		4446
3333	=item The winsocket select function	4447	=item The winsocket select function
3334		4448
3335	The winsocket C<select> function doesn't follow POSIX in that it	4449	The winsocket C<select> function doesn't follow POSIX in that it
3336	requires socket I<handles> and not socket I<file descriptors> (it is	4450	requires socket I<handles> and not socket I<file descriptors> (it is
3337	also extremely buggy). This makes select very inefficient, and also	4451	also extremely buggy). This makes select very inefficient, and also
3338	requires a mapping from file descriptors to socket handles. See the	4452	requires a mapping from file descriptors to socket handles (the Microsoft
		4453	C runtime provides the function C<_open_osfhandle> for this). See the
3339	discussion of the C<EV_SELECT_USE_FD_SET>, C<EV_SELECT_IS_WINSOCKET> and	4454	discussion of the C<EV_SELECT_USE_FD_SET>, C<EV_SELECT_IS_WINSOCKET> and
3340	C<EV_FD_TO_WIN32_HANDLE> preprocessor symbols for more info.	4455	C<EV_FD_TO_WIN32_HANDLE> preprocessor symbols for more info.
3341		4456
3342	The configuration for a "naked" win32 using the Microsoft runtime	4457	The configuration for a "naked" win32 using the Microsoft runtime
3343	libraries and raw winsocket select is:	4458	libraries and raw winsocket select is:
…		…
3354		4469
3355	Early versions of winsocket's select only supported waiting for a maximum	4470	Early versions of winsocket's select only supported waiting for a maximum
3356	of C<64> handles (probably owning to the fact that all windows kernels	4471	of C<64> handles (probably owning to the fact that all windows kernels
3357	can only wait for C<64> things at the same time internally; Microsoft	4472	can only wait for C<64> things at the same time internally; Microsoft
3358	recommends spawning a chain of threads and wait for 63 handles and the	4473	recommends spawning a chain of threads and wait for 63 handles and the
3359	previous thread in each. Great).	4474	previous thread in each. Sounds great!).
3360		4475
3361	Newer versions support more handles, but you need to define C<FD_SETSIZE>	4476	Newer versions support more handles, but you need to define C<FD_SETSIZE>
3362	to some high number (e.g. C<2048>) before compiling the winsocket select	4477	to some high number (e.g. C<2048>) before compiling the winsocket select
3363	call (which might be in libev or elsewhere, for example, perl does its own	4478	call (which might be in libev or elsewhere, for example, perl and many
3364	select emulation on windows).	4479	other interpreters do their own select emulation on windows).
3365		4480
3366	Another limit is the number of file descriptors in the Microsoft runtime	4481	Another limit is the number of file descriptors in the Microsoft runtime
3367	libraries, which by default is C<64> (there must be a hidden I<64> fetish	4482	libraries, which by default is C<64> (there must be a hidden I<64>
3368	or something like this inside Microsoft). You can increase this by calling	4483	fetish or something like this inside Microsoft). You can increase this
3369	C<_setmaxstdio>, which can increase this limit to C<2048> (another	4484	by calling C<_setmaxstdio>, which can increase this limit to C<2048>
3370	arbitrary limit), but is broken in many versions of the Microsoft runtime	4485	(another arbitrary limit), but is broken in many versions of the Microsoft
3371	libraries.
3372
3373	This might get you to about C<512> or C<2048> sockets (depending on	4486	runtime libraries. This might get you to about C<512> or C<2048> sockets
3374	windows version and/or the phase of the moon). To get more, you need to	4487	(depending on windows version and/or the phase of the moon). To get more,
3375	wrap all I/O functions and provide your own fd management, but the cost of	4488	you need to wrap all I/O functions and provide your own fd management, but
3376	calling select (O(n²)) will likely make this unworkable.	4489	the cost of calling select (O(n²)) will likely make this unworkable.
3377		4490
3378	=back	4491	=back
3379		4492
3380
3381	=head1 PORTABILITY REQUIREMENTS	4493	=head2 PORTABILITY REQUIREMENTS
3382		4494
3383	In addition to a working ISO-C implementation, libev relies on a few	4495	In addition to a working ISO-C implementation and of course the
3384	additional extensions:	4496	backend-specific APIs, libev relies on a few additional extensions:
3385		4497
3386	=over 4	4498	=over 4
3387		4499
3388	=item C<void (*)(ev_watcher_type *, int revents)> must have compatible	4500	=item C<void (*)(ev_watcher_type *, int revents)> must have compatible
3389	calling conventions regardless of C<ev_watcher_type *>.	4501	calling conventions regardless of C<ev_watcher_type *>.
…		…
3395	calls them using an C<ev_watcher *> internally.	4507	calls them using an C<ev_watcher *> internally.
3396		4508
3397	=item C<sig_atomic_t volatile> must be thread-atomic as well	4509	=item C<sig_atomic_t volatile> must be thread-atomic as well
3398		4510
3399	The type C<sig_atomic_t volatile> (or whatever is defined as	4511	The type C<sig_atomic_t volatile> (or whatever is defined as
3400	C<EV_ATOMIC_T>) must be atomic w.r.t. accesses from different	4512	C<EV_ATOMIC_T>) must be atomic with respect to accesses from different
3401	threads. This is not part of the specification for C<sig_atomic_t>, but is	4513	threads. This is not part of the specification for C<sig_atomic_t>, but is
3402	believed to be sufficiently portable.	4514	believed to be sufficiently portable.
3403		4515
3404	=item C<sigprocmask> must work in a threaded environment	4516	=item C<sigprocmask> must work in a threaded environment
3405		4517
…		…
3414	except the initial one, and run the default loop in the initial thread as	4526	except the initial one, and run the default loop in the initial thread as
3415	well.	4527	well.
3416		4528
3417	=item C<long> must be large enough for common memory allocation sizes	4529	=item C<long> must be large enough for common memory allocation sizes
3418		4530
3419	To improve portability and simplify using libev, libev uses C<long>	4531	To improve portability and simplify its API, libev uses C<long> internally
3420	internally instead of C<size_t> when allocating its data structures. On	4532	instead of C<size_t> when allocating its data structures. On non-POSIX
3421	non-POSIX systems (Microsoft...) this might be unexpectedly low, but	4533	systems (Microsoft...) this might be unexpectedly low, but is still at
3422	is still at least 31 bits everywhere, which is enough for hundreds of	4534	least 31 bits everywhere, which is enough for hundreds of millions of
3423	millions of watchers.	4535	watchers.
3424		4536
3425	=item C<double> must hold a time value in seconds with enough accuracy	4537	=item C<double> must hold a time value in seconds with enough accuracy
3426		4538
3427	The type C<double> is used to represent timestamps. It is required to	4539	The type C<double> is used to represent timestamps. It is required to
3428	have at least 51 bits of mantissa (and 9 bits of exponent), which is good	4540	have at least 51 bits of mantissa (and 9 bits of exponent), which is good
3429	enough for at least into the year 4000. This requirement is fulfilled by	4541	enough for at least into the year 4000. This requirement is fulfilled by
3430	implementations implementing IEEE 754 (basically all existing ones).	4542	implementations implementing IEEE 754, which is basically all existing
		4543	ones. With IEEE 754 doubles, you get microsecond accuracy until at least
		4544	2200.
3431		4545
3432	=back	4546	=back
3433		4547
3434	If you know of other additional requirements drop me a note.	4548	If you know of other additional requirements drop me a note.
3435		4549
3436		4550
3437	=head1 COMPILER WARNINGS	4551	=head1 ALGORITHMIC COMPLEXITIES
3438		4552
3439	Depending on your compiler and compiler settings, you might get no or a	4553	In this section the complexities of (many of) the algorithms used inside
3440	lot of warnings when compiling libev code. Some people are apparently	4554	libev will be documented. For complexity discussions about backends see
3441	scared by this.	4555	the documentation for C<ev_default_init>.
3442		4556
3443	However, these are unavoidable for many reasons. For one, each compiler	4557	All of the following are about amortised time: If an array needs to be
3444	has different warnings, and each user has different tastes regarding	4558	extended, libev needs to realloc and move the whole array, but this
3445	warning options. "Warn-free" code therefore cannot be a goal except when	4559	happens asymptotically rarer with higher number of elements, so O(1) might
3446	targeting a specific compiler and compiler-version.	4560	mean that libev does a lengthy realloc operation in rare cases, but on
		4561	average it is much faster and asymptotically approaches constant time.
3447		4562
3448	Another reason is that some compiler warnings require elaborate	4563	=over 4
3449	workarounds, or other changes to the code that make it less clear and less
3450	maintainable.
3451		4564
3452	And of course, some compiler warnings are just plain stupid, or simply	4565	=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers)
3453	wrong (because they don't actually warn about the condition their message
3454	seems to warn about).
3455		4566
3456	While libev is written to generate as few warnings as possible,	4567	This means that, when you have a watcher that triggers in one hour and
3457	"warn-free" code is not a goal, and it is recommended not to build libev	4568	there are 100 watchers that would trigger before that, then inserting will
3458	with any compiler warnings enabled unless you are prepared to cope with	4569	have to skip roughly seven (C<ld 100>) of these watchers.
3459	them (e.g. by ignoring them). Remember that warnings are just that:
3460	warnings, not errors, or proof of bugs.
3461		4570
		4571	=item Changing timer/periodic watchers (by autorepeat or calling again): O(log skipped_other_timers)
3462		4572
3463	=head1 VALGRIND	4573	That means that changing a timer costs less than removing/adding them,
		4574	as only the relative motion in the event queue has to be paid for.
3464		4575
3465	Valgrind has a special section here because it is a popular tool that is	4576	=item Starting io/check/prepare/idle/signal/child/fork/async watchers: O(1)
3466	highly useful, but valgrind reports are very hard to interpret.
3467		4577
3468	If you think you found a bug (memory leak, uninitialised data access etc.)	4578	These just add the watcher into an array or at the head of a list.
3469	in libev, then check twice: If valgrind reports something like:
3470		4579
3471	==2274== definitely lost: 0 bytes in 0 blocks.	4580	=item Stopping check/prepare/idle/fork/async watchers: O(1)
3472	==2274== possibly lost: 0 bytes in 0 blocks.
3473	==2274== still reachable: 256 bytes in 1 blocks.
3474		4581
3475	Then there is no memory leak. Similarly, under some circumstances,	4582	=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % EV_PID_HASHSIZE))
3476	valgrind might report kernel bugs as if it were a bug in libev, or it
3477	might be confused (it is a very good tool, but only a tool).
3478		4583
3479	If you are unsure about something, feel free to contact the mailing list	4584	These watchers are stored in lists, so they need to be walked to find the
3480	with the full valgrind report and an explanation on why you think this is	4585	correct watcher to remove. The lists are usually short (you don't usually
3481	a bug in libev. However, don't be annoyed when you get a brisk "this is	4586	have many watchers waiting for the same fd or signal: one is typical, two
3482	no bug" answer and take the chance of learning how to interpret valgrind	4587	is rare).
3483	properly.
3484		4588
3485	If you need, for some reason, empty reports from valgrind for your project	4589	=item Finding the next timer in each loop iteration: O(1)
3486	I suggest using suppression lists.
3487		4590
		4591	By virtue of using a binary or 4-heap, the next timer is always found at a
		4592	fixed position in the storage array.
		4593
		4594	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)
		4595
		4596	A change means an I/O watcher gets started or stopped, which requires
		4597	libev to recalculate its status (and possibly tell the kernel, depending
		4598	on backend and whether C<ev_io_set> was used).
		4599
		4600	=item Activating one watcher (putting it into the pending state): O(1)
		4601
		4602	=item Priority handling: O(number_of_priorities)
		4603
		4604	Priorities are implemented by allocating some space for each
		4605	priority. When doing priority-based operations, libev usually has to
		4606	linearly search all the priorities, but starting/stopping and activating
		4607	watchers becomes O(1) with respect to priority handling.
		4608
		4609	=item Sending an ev_async: O(1)
		4610
		4611	=item Processing ev_async_send: O(number_of_async_watchers)
		4612
		4613	=item Processing signals: O(max_signal_number)
		4614
		4615	Sending involves a system call I<iff> there were no other C<ev_async_send>
		4616	calls in the current loop iteration. Checking for async and signal events
		4617	involves iterating over all running async watchers or all signal numbers.
		4618
		4619	=back
		4620
		4621
		4622	=head1 GLOSSARY
		4623
		4624	=over 4
		4625
		4626	=item active
		4627
		4628	A watcher is active as long as it has been started (has been attached to
		4629	an event loop) but not yet stopped (disassociated from the event loop).
		4630
		4631	=item application
		4632
		4633	In this document, an application is whatever is using libev.
		4634
		4635	=item callback
		4636
		4637	The address of a function that is called when some event has been
		4638	detected. Callbacks are being passed the event loop, the watcher that
		4639	received the event, and the actual event bitset.
		4640
		4641	=item callback invocation
		4642
		4643	The act of calling the callback associated with a watcher.
		4644
		4645	=item event
		4646
		4647	A change of state of some external event, such as data now being available
		4648	for reading on a file descriptor, time having passed or simply not having
		4649	any other events happening anymore.
		4650
		4651	In libev, events are represented as single bits (such as C<EV_READ> or
		4652	C<EV_TIMEOUT>).
		4653
		4654	=item event library
		4655
		4656	A software package implementing an event model and loop.
		4657
		4658	=item event loop
		4659
		4660	An entity that handles and processes external events and converts them
		4661	into callback invocations.
		4662
		4663	=item event model
		4664
		4665	The model used to describe how an event loop handles and processes
		4666	watchers and events.
		4667
		4668	=item pending
		4669
		4670	A watcher is pending as soon as the corresponding event has been detected,
		4671	and stops being pending as soon as the watcher will be invoked or its
		4672	pending status is explicitly cleared by the application.
		4673
		4674	A watcher can be pending, but not active. Stopping a watcher also clears
		4675	its pending status.
		4676
		4677	=item real time
		4678
		4679	The physical time that is observed. It is apparently strictly monotonic :)
		4680
		4681	=item wall-clock time
		4682
		4683	The time and date as shown on clocks. Unlike real time, it can actually
		4684	be wrong and jump forwards and backwards, e.g. when the you adjust your
		4685	clock.
		4686
		4687	=item watcher
		4688
		4689	A data structure that describes interest in certain events. Watchers need
		4690	to be started (attached to an event loop) before they can receive events.
		4691
		4692	=item watcher invocation
		4693
		4694	The act of calling the callback associated with a watcher.
		4695
		4696	=back
3488		4697
3489	=head1 AUTHOR	4698	=head1 AUTHOR
3490		4699
3491	Marc Lehmann <libev@schmorp.de>.	4700	Marc Lehmann <libev@schmorp.de>, with repeated corrections by Mikael Magnusson.
3492		4701

Diff Legend

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