ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/libev/ev.pod

Comparing libev/ev.pod (file contents):
Revision 1.27 by root, Wed Nov 14 05:02:07 2007 UTC vs.
Revision 1.404 by root, Sat Apr 28 12:10:07 2012 UTC

…		…
2		2
3	libev - a high performance full-featured event loop written in C	3	libev - a high performance full-featured event loop written in C
4		4
5	=head1 SYNOPSIS	5	=head1 SYNOPSIS
6		6
7	#include <ev.h>	7	#include <ev.h>
8		8
9	=head1 DESCRIPTION	9	=head2 EXAMPLE PROGRAM
		10
		11	// a single header file is required
		12	#include <ev.h>
		13
		14	#include <stdio.h> // for puts
		15
		16	// every watcher type has its own typedef'd struct
		17	// with the name ev_TYPE
		18	ev_io stdin_watcher;
		19	ev_timer timeout_watcher;
		20
		21	// all watcher callbacks have a similar signature
		22	// this callback is called when data is readable on stdin
		23	static void
		24	stdin_cb (EV_P_ ev_io *w, int revents)
		25	{
		26	puts ("stdin ready");
		27	// for one-shot events, one must manually stop the watcher
		28	// with its corresponding stop function.
		29	ev_io_stop (EV_A_ w);
		30
		31	// this causes all nested ev_run's to stop iterating
		32	ev_break (EV_A_ EVBREAK_ALL);
		33	}
		34
		35	// another callback, this time for a time-out
		36	static void
		37	timeout_cb (EV_P_ ev_timer *w, int revents)
		38	{
		39	puts ("timeout");
		40	// this causes the innermost ev_run to stop iterating
		41	ev_break (EV_A_ EVBREAK_ONE);
		42	}
		43
		44	int
		45	main (void)
		46	{
		47	// use the default event loop unless you have special needs
		48	struct ev_loop *loop = EV_DEFAULT;
		49
		50	// initialise an io watcher, then start it
		51	// this one will watch for stdin to become readable
		52	ev_io_init (&stdin_watcher, stdin_cb, /*STDIN_FILENO*/ 0, EV_READ);
		53	ev_io_start (loop, &stdin_watcher);
		54
		55	// initialise a timer watcher, then start it
		56	// simple non-repeating 5.5 second timeout
		57	ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);
		58	ev_timer_start (loop, &timeout_watcher);
		59
		60	// now wait for events to arrive
		61	ev_run (loop, 0);
		62
		63	// break was called, so exit
		64	return 0;
		65	}
		66
		67	=head1 ABOUT THIS DOCUMENT
		68
		69	This document documents the libev software package.
		70
		71	The newest version of this document is also available as an html-formatted
		72	web page you might find easier to navigate when reading it for the first
		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	Familiarity with event based programming techniques in general is assumed
		81	throughout this document.
		82
		83	=head1 WHAT TO READ WHEN IN A HURRY
		84
		85	This manual tries to be very detailed, but unfortunately, this also makes
		86	it very long. If you just want to know the basics of libev, I suggest
		87	reading L<ANATOMY OF A WATCHER>, then the L<EXAMPLE PROGRAM> above and
		88	look up the missing functions in L<GLOBAL FUNCTIONS> and the C<ev_io> and
		89	C<ev_timer> sections in L<WATCHER TYPES>.
		90
		91	=head1 ABOUT LIBEV
10		92
11	Libev is an event loop: you register interest in certain events (such as a	93	Libev is an event loop: you register interest in certain events (such as a
12	file descriptor being readable or a timeout occuring), and it will manage	94	file descriptor being readable or a timeout occurring), and it will manage
13	these event sources and provide your program with events.	95	these event sources and provide your program with events.
14		96
15	To do this, it must take more or less complete control over your process	97	To do this, it must take more or less complete control over your process
16	(or thread) by executing the I<event loop> handler, and will then	98	(or thread) by executing the I<event loop> handler, and will then
17	communicate events via a callback mechanism.	99	communicate events via a callback mechanism.
…		…
19	You register interest in certain events by registering so-called I<event	101	You register interest in certain events by registering so-called I<event
20	watchers>, which are relatively small C structures you initialise with the	102	watchers>, which are relatively small C structures you initialise with the
21	details of the event, and then hand it over to libev by I<starting> the	103	details of the event, and then hand it over to libev by I<starting> the
22	watcher.	104	watcher.
23		105
24	=head1 FEATURES	106	=head2 FEATURES
25		107
26	Libev supports select, poll, the linux-specific epoll and the bsd-specific	108	Libev supports C<select>, C<poll>, the Linux-specific C<epoll>, the
27	kqueue mechanisms for file descriptor events, relative timers, absolute	109	BSD-specific C<kqueue> and the Solaris-specific event port mechanisms
28	timers with customised rescheduling, signal events, process status change	110	for file descriptor events (C<ev_io>), the Linux C<inotify> interface
		111	(for C<ev_stat>), Linux eventfd/signalfd (for faster and cleaner
		112	inter-thread wakeup (C<ev_async>)/signal handling (C<ev_signal>)) relative
		113	timers (C<ev_timer>), absolute timers with customised rescheduling
		114	(C<ev_periodic>), synchronous signals (C<ev_signal>), process status
29	events (related to SIGCHLD), and event watchers dealing with the event	115	change events (C<ev_child>), and event watchers dealing with the event
30	loop mechanism itself (idle, prepare and check watchers). It also is quite	116	loop mechanism itself (C<ev_idle>, C<ev_embed>, C<ev_prepare> and
		117	C<ev_check> watchers) as well as file watchers (C<ev_stat>) and even
		118	limited support for fork events (C<ev_fork>).
		119
		120	It also is quite fast (see this
31	fast (see this L<benchmark|http://libev.schmorp.de/bench.html> comparing	121	L<benchmark|http://libev.schmorp.de/bench.html> comparing it to libevent
32	it to libevent for example).	122	for example).
33		123
34	=head1 CONVENTIONS	124	=head2 CONVENTIONS
35		125
36	Libev is very configurable. In this manual the default configuration	126	Libev is very configurable. In this manual the default (and most common)
37	will be described, which supports multiple event loops. For more info	127	configuration will be described, which supports multiple event loops. For
38	about various configuration options please have a look at the file	128	more info about various configuration options please have a look at
39	F<README.embed> in the libev distribution. If libev was configured without	129	B<EMBED> section in this manual. If libev was configured without support
40	support for multiple event loops, then all functions taking an initial	130	for multiple event loops, then all functions taking an initial argument of
41	argument of name C<loop> (which is always of type C<struct ev_loop *>)	131	name C<loop> (which is always of type C<struct ev_loop *>) will not have
42	will not have this argument.	132	this argument.
43		133
44	=head1 TIME REPRESENTATION	134	=head2 TIME REPRESENTATION
45		135
46	Libev represents time as a single floating point number, representing the	136	Libev represents time as a single floating point number, representing
47	(fractional) number of seconds since the (POSIX) epoch (somewhere near	137	the (fractional) number of seconds since the (POSIX) epoch (in practice
48	the beginning of 1970, details are complicated, don't ask). This type is	138	somewhere near the beginning of 1970, details are complicated, don't
49	called C<ev_tstamp>, which is what you should use too. It usually aliases	139	ask). This type is called C<ev_tstamp>, which is what you should use
50	to the double type in C.	140	too. It usually aliases to the C<double> type in C. When you need to do
		141	any calculations on it, you should treat it as some floating point value.
		142
		143	Unlike the name component C<stamp> might indicate, it is also used for
		144	time differences (e.g. delays) throughout libev.
		145
		146	=head1 ERROR HANDLING
		147
		148	Libev knows three classes of errors: operating system errors, usage errors
		149	and internal errors (bugs).
		150
		151	When libev catches an operating system error it cannot handle (for example
		152	a system call indicating a condition libev cannot fix), it calls the callback
		153	set via C<ev_set_syserr_cb>, which is supposed to fix the problem or
		154	abort. The default is to print a diagnostic message and to call C<abort
		155	()>.
		156
		157	When libev detects a usage error such as a negative timer interval, then
		158	it will print a diagnostic message and abort (via the C<assert> mechanism,
		159	so C<NDEBUG> will disable this checking): these are programming errors in
		160	the libev caller and need to be fixed there.
		161
		162	Libev also has a few internal error-checking C<assert>ions, and also has
		163	extensive consistency checking code. These do not trigger under normal
		164	circumstances, as they indicate either a bug in libev or worse.
		165
51		166
52	=head1 GLOBAL FUNCTIONS	167	=head1 GLOBAL FUNCTIONS
53		168
54	These functions can be called anytime, even before initialising the	169	These functions can be called anytime, even before initialising the
55	library in any way.	170	library in any way.
…		…
58		173
59	=item ev_tstamp ev_time ()	174	=item ev_tstamp ev_time ()
60		175
61	Returns the current time as libev would use it. Please note that the	176	Returns the current time as libev would use it. Please note that the
62	C<ev_now> function is usually faster and also often returns the timestamp	177	C<ev_now> function is usually faster and also often returns the timestamp
63	you actually want to know.	178	you actually want to know. Also interesting is the combination of
		179	C<ev_now_update> and C<ev_now>.
		180
		181	=item ev_sleep (ev_tstamp interval)
		182
		183	Sleep for the given interval: The current thread will be blocked
		184	until either it is interrupted or the given time interval has
		185	passed (approximately - it might return a bit earlier even if not
		186	interrupted). Returns immediately if C<< interval <= 0 >>.
		187
		188	Basically this is a sub-second-resolution C<sleep ()>.
		189
		190	The range of the C<interval> is limited - libev only guarantees to work
		191	with sleep times of up to one day (C<< interval <= 86400 >>).
64		192
65	=item int ev_version_major ()	193	=item int ev_version_major ()
66		194
67	=item int ev_version_minor ()	195	=item int ev_version_minor ()
68		196
69	You can find out the major and minor version numbers of the library	197	You can find out the major and minor ABI version numbers of the library
70	you linked against by calling the functions C<ev_version_major> and	198	you linked against by calling the functions C<ev_version_major> and
71	C<ev_version_minor>. If you want, you can compare against the global	199	C<ev_version_minor>. If you want, you can compare against the global
72	symbols C<EV_VERSION_MAJOR> and C<EV_VERSION_MINOR>, which specify the	200	symbols C<EV_VERSION_MAJOR> and C<EV_VERSION_MINOR>, which specify the
73	version of the library your program was compiled against.	201	version of the library your program was compiled against.
74		202
		203	These version numbers refer to the ABI version of the library, not the
		204	release version.
		205
75	Usually, it's a good idea to terminate if the major versions mismatch,	206	Usually, it's a good idea to terminate if the major versions mismatch,
76	as this indicates an incompatible change. Minor versions are usually	207	as this indicates an incompatible change. Minor versions are usually
77	compatible to older versions, so a larger minor version alone is usually	208	compatible to older versions, so a larger minor version alone is usually
78	not a problem.	209	not a problem.
79		210
		211	Example: Make sure we haven't accidentally been linked against the wrong
		212	version (note, however, that this will not detect other ABI mismatches,
		213	such as LFS or reentrancy).
		214
		215	assert (("libev version mismatch",
		216	ev_version_major () == EV_VERSION_MAJOR
		217	&& ev_version_minor () >= EV_VERSION_MINOR));
		218
		219	=item unsigned int ev_supported_backends ()
		220
		221	Return the set of all backends (i.e. their corresponding C<EV_BACKEND_*>
		222	value) compiled into this binary of libev (independent of their
		223	availability on the system you are running on). See C<ev_default_loop> for
		224	a description of the set values.
		225
		226	Example: make sure we have the epoll method, because yeah this is cool and
		227	a must have and can we have a torrent of it please!!!11
		228
		229	assert (("sorry, no epoll, no sex",
		230	ev_supported_backends () & EVBACKEND_EPOLL));
		231
		232	=item unsigned int ev_recommended_backends ()
		233
		234	Return the set of all backends compiled into this binary of libev and
		235	also recommended for this platform, meaning it will work for most file
		236	descriptor types. This set is often smaller than the one returned by
		237	C<ev_supported_backends>, as for example kqueue is broken on most BSDs
		238	and will not be auto-detected unless you explicitly request it (assuming
		239	you know what you are doing). This is the set of backends that libev will
		240	probe for if you specify no backends explicitly.
		241
		242	=item unsigned int ev_embeddable_backends ()
		243
		244	Returns the set of backends that are embeddable in other event loops. This
		245	value is platform-specific but can include backends not available on the
		246	current system. To find which embeddable backends might be supported on
		247	the current system, you would need to look at C<ev_embeddable_backends ()
		248	& ev_supported_backends ()>, likewise for recommended ones.
		249
		250	See the description of C<ev_embed> watchers for more info.
		251
80	=item ev_set_allocator (void *(*cb)(void *ptr, long size))	252	=item ev_set_allocator (void *(*cb)(void *ptr, long size) throw ())
81		253
82	Sets the allocation function to use (the prototype is similar to the	254	Sets the allocation function to use (the prototype is similar - the
83	realloc C function, the semantics are identical). It is used to allocate	255	semantics are identical to the C<realloc> C89/SuS/POSIX function). It is
84	and free memory (no surprises here). If it returns zero when memory	256	used to allocate and free memory (no surprises here). If it returns zero
85	needs to be allocated, the library might abort or take some potentially	257	when memory needs to be allocated (C<size != 0>), the library might abort
86	destructive action. The default is your system realloc function.	258	or take some potentially destructive action.
		259
		260	Since some systems (at least OpenBSD and Darwin) fail to implement
		261	correct C<realloc> semantics, libev will use a wrapper around the system
		262	C<realloc> and C<free> functions by default.
87		263
88	You could override this function in high-availability programs to, say,	264	You could override this function in high-availability programs to, say,
89	free some memory if it cannot allocate memory, to use a special allocator,	265	free some memory if it cannot allocate memory, to use a special allocator,
90	or even to sleep a while and retry until some memory is available.	266	or even to sleep a while and retry until some memory is available.
91		267
		268	Example: Replace the libev allocator with one that waits a bit and then
		269	retries (example requires a standards-compliant C<realloc>).
		270
		271	static void *
		272	persistent_realloc (void *ptr, size_t size)
		273	{
		274	for (;;)
		275	{
		276	void *newptr = realloc (ptr, size);
		277
		278	if (newptr)
		279	return newptr;
		280
		281	sleep (60);
		282	}
		283	}
		284
		285	...
		286	ev_set_allocator (persistent_realloc);
		287
92	=item ev_set_syserr_cb (void (*cb)(const char *msg));	288	=item ev_set_syserr_cb (void (*cb)(const char *msg) throw ())
93		289
94	Set the callback function to call on a retryable syscall error (such	290	Set the callback function to call on a retryable system call error (such
95	as failed select, poll, epoll_wait). The message is a printable string	291	as failed select, poll, epoll_wait). The message is a printable string
96	indicating the system call or subsystem causing the problem. If this	292	indicating the system call or subsystem causing the problem. If this
97	callback is set, then libev will expect it to remedy the sitution, no	293	callback is set, then libev will expect it to remedy the situation, no
98	matter what, when it returns. That is, libev will generally retry the	294	matter what, when it returns. That is, libev will generally retry the
99	requested operation, or, if the condition doesn't go away, do bad stuff	295	requested operation, or, if the condition doesn't go away, do bad stuff
100	(such as abort).	296	(such as abort).
101		297
		298	Example: This is basically the same thing that libev does internally, too.
		299
		300	static void
		301	fatal_error (const char *msg)
		302	{
		303	perror (msg);
		304	abort ();
		305	}
		306
		307	...
		308	ev_set_syserr_cb (fatal_error);
		309
		310	=item ev_feed_signal (int signum)
		311
		312	This function can be used to "simulate" a signal receive. It is completely
		313	safe to call this function at any time, from any context, including signal
		314	handlers or random threads.
		315
		316	Its main use is to customise signal handling in your process, especially
		317	in the presence of threads. For example, you could block signals
		318	by default in all threads (and specifying C<EVFLAG_NOSIGMASK> when
		319	creating any loops), and in one thread, use C<sigwait> or any other
		320	mechanism to wait for signals, then "deliver" them to libev by calling
		321	C<ev_feed_signal>.
		322
102	=back	323	=back
103		324
104	=head1 FUNCTIONS CONTROLLING THE EVENT LOOP	325	=head1 FUNCTIONS CONTROLLING EVENT LOOPS
105		326
106	An event loop is described by a C<struct ev_loop *>. The library knows two	327	An event loop is described by a C<struct ev_loop *> (the C<struct> is
107	types of such loops, the I<default> loop, which supports signals and child	328	I<not> optional in this case unless libev 3 compatibility is disabled, as
108	events, and dynamically created loops which do not.	329	libev 3 had an C<ev_loop> function colliding with the struct name).
109		330
110	If you use threads, a common model is to run the default event loop	331	The library knows two types of such loops, the I<default> loop, which
111	in your main thread (or in a separate thread) and for each thread you	332	supports child process events, and dynamically created event loops which
112	create, you also create another event loop. Libev itself does no locking	333	do not.
113	whatsoever, so if you mix calls to the same event loop in different
114	threads, make sure you lock (this is usually a bad idea, though, even if
115	done correctly, because it's hideous and inefficient).
116		334
117	=over 4	335	=over 4
118		336
119	=item struct ev_loop *ev_default_loop (unsigned int flags)	337	=item struct ev_loop *ev_default_loop (unsigned int flags)
120		338
121	This will initialise the default event loop if it hasn't been initialised	339	This returns the "default" event loop object, which is what you should
122	yet and return it. If the default loop could not be initialised, returns	340	normally use when you just need "the event loop". Event loop objects and
123	false. If it already was initialised it simply returns it (and ignores the	341	the C<flags> parameter are described in more detail in the entry for
124	flags).	342	C<ev_loop_new>.
		343
		344	If the default loop is already initialised then this function simply
		345	returns it (and ignores the flags. If that is troubling you, check
		346	C<ev_backend ()> afterwards). Otherwise it will create it with the given
		347	flags, which should almost always be C<0>, unless the caller is also the
		348	one calling C<ev_run> or otherwise qualifies as "the main program".
125		349
126	If you don't know what event loop to use, use the one returned from this	350	If you don't know what event loop to use, use the one returned from this
127	function.	351	function (or via the C<EV_DEFAULT> macro).
		352
		353	Note that this function is I<not> thread-safe, so if you want to use it
		354	from multiple threads, you have to employ some kind of mutex (note also
		355	that this case is unlikely, as loops cannot be shared easily between
		356	threads anyway).
		357
		358	The default loop is the only loop that can handle C<ev_child> watchers,
		359	and to do this, it always registers a handler for C<SIGCHLD>. If this is
		360	a problem for your application you can either create a dynamic loop with
		361	C<ev_loop_new> which doesn't do that, or you can simply overwrite the
		362	C<SIGCHLD> signal handler I<after> calling C<ev_default_init>.
		363
		364	Example: This is the most typical usage.
		365
		366	if (!ev_default_loop (0))
		367	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
		368
		369	Example: Restrict libev to the select and poll backends, and do not allow
		370	environment settings to be taken into account:
		371
		372	ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);
		373
		374	=item struct ev_loop *ev_loop_new (unsigned int flags)
		375
		376	This will create and initialise a new event loop object. If the loop
		377	could not be initialised, returns false.
		378
		379	This function is thread-safe, and one common way to use libev with
		380	threads is indeed to create one loop per thread, and using the default
		381	loop in the "main" or "initial" thread.
128		382
129	The flags argument can be used to specify special behaviour or specific	383	The flags argument can be used to specify special behaviour or specific
130	backends to use, and is usually specified as 0 (or EVFLAG_AUTO).	384	backends to use, and is usually specified as C<0> (or C<EVFLAG_AUTO>).
131		385
132	It supports the following flags:	386	The following flags are supported:
133		387
134	=over 4	388	=over 4
135		389
136	=item C<EVFLAG_AUTO>	390	=item C<EVFLAG_AUTO>
137		391
138	The default flags value. Use this if you have no clue (it's the right	392	The default flags value. Use this if you have no clue (it's the right
139	thing, believe me).	393	thing, believe me).
140		394
141	=item C<EVFLAG_NOENV>	395	=item C<EVFLAG_NOENV>
142		396
143	If this flag bit is ored into the flag value (or the program runs setuid	397	If this flag bit is or'ed into the flag value (or the program runs setuid
144	or setgid) then libev will I<not> look at the environment variable	398	or setgid) then libev will I<not> look at the environment variable
145	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will	399	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will
146	override the flags completely if it is found in the environment. This is	400	override the flags completely if it is found in the environment. This is
147	useful to try out specific backends to test their performance, or to work	401	useful to try out specific backends to test their performance, or to work
148	around bugs.	402	around bugs.
149		403
		404	=item C<EVFLAG_FORKCHECK>
		405
		406	Instead of calling C<ev_loop_fork> manually after a fork, you can also
		407	make libev check for a fork in each iteration by enabling this flag.
		408
		409	This works by calling C<getpid ()> on every iteration of the loop,
		410	and thus this might slow down your event loop if you do a lot of loop
		411	iterations and little real work, but is usually not noticeable (on my
		412	GNU/Linux system for example, C<getpid> is actually a simple 5-insn sequence
		413	without a system call and thus I<very> fast, but my GNU/Linux system also has
		414	C<pthread_atfork> which is even faster).
		415
		416	The big advantage of this flag is that you can forget about fork (and
		417	forget about forgetting to tell libev about forking) when you use this
		418	flag.
		419
		420	This flag setting cannot be overridden or specified in the C<LIBEV_FLAGS>
		421	environment variable.
		422
		423	=item C<EVFLAG_NOINOTIFY>
		424
		425	When this flag is specified, then libev will not attempt to use the
		426	I<inotify> API for its C<ev_stat> watchers. Apart from debugging and
		427	testing, this flag can be useful to conserve inotify file descriptors, as
		428	otherwise each loop using C<ev_stat> watchers consumes one inotify handle.
		429
		430	=item C<EVFLAG_SIGNALFD>
		431
		432	When this flag is specified, then libev will attempt to use the
		433	I<signalfd> API for its C<ev_signal> (and C<ev_child>) watchers. This API
		434	delivers signals synchronously, which makes it both faster and might make
		435	it possible to get the queued signal data. It can also simplify signal
		436	handling with threads, as long as you properly block signals in your
		437	threads that are not interested in handling them.
		438
		439	Signalfd will not be used by default as this changes your signal mask, and
		440	there are a lot of shoddy libraries and programs (glib's threadpool for
		441	example) that can't properly initialise their signal masks.
		442
		443	=item C<EVFLAG_NOSIGMASK>
		444
		445	When this flag is specified, then libev will avoid to modify the signal
		446	mask. Specifically, this means you have to make sure signals are unblocked
		447	when you want to receive them.
		448
		449	This behaviour is useful when you want to do your own signal handling, or
		450	want to handle signals only in specific threads and want to avoid libev
		451	unblocking the signals.
		452
		453	It's also required by POSIX in a threaded program, as libev calls
		454	C<sigprocmask>, whose behaviour is officially unspecified.
		455
		456	This flag's behaviour will become the default in future versions of libev.
		457
150	=item C<EVMETHOD_SELECT> (portable select backend)	458	=item C<EVBACKEND_SELECT> (value 1, portable select backend)
151		459
		460	This is your standard select(2) backend. Not I<completely> standard, as
		461	libev tries to roll its own fd_set with no limits on the number of fds,
		462	but if that fails, expect a fairly low limit on the number of fds when
		463	using this backend. It doesn't scale too well (O(highest_fd)), but its
		464	usually the fastest backend for a low number of (low-numbered :) fds.
		465
		466	To get good performance out of this backend you need a high amount of
		467	parallelism (most of the file descriptors should be busy). If you are
		468	writing a server, you should C<accept ()> in a loop to accept as many
		469	connections as possible during one iteration. You might also want to have
		470	a look at C<ev_set_io_collect_interval ()> to increase the amount of
		471	readiness notifications you get per iteration.
		472
		473	This backend maps C<EV_READ> to the C<readfds> set and C<EV_WRITE> to the
		474	C<writefds> set (and to work around Microsoft Windows bugs, also onto the
		475	C<exceptfds> set on that platform).
		476
152	=item C<EVMETHOD_POLL> (poll backend, available everywhere except on windows)	477	=item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows)
153		478
154	=item C<EVMETHOD_EPOLL> (linux only)	479	And this is your standard poll(2) backend. It's more complicated
		480	than select, but handles sparse fds better and has no artificial
		481	limit on the number of fds you can use (except it will slow down
		482	considerably with a lot of inactive fds). It scales similarly to select,
		483	i.e. O(total_fds). See the entry for C<EVBACKEND_SELECT>, above, for
		484	performance tips.
155		485
156	=item C<EVMETHOD_KQUEUE> (some bsds only)	486	This backend maps C<EV_READ> to C<POLLIN | POLLERR | POLLHUP>, and
		487	C<EV_WRITE> to C<POLLOUT | POLLERR | POLLHUP>.
157		488
158	=item C<EVMETHOD_DEVPOLL> (solaris 8 only)	489	=item C<EVBACKEND_EPOLL> (value 4, Linux)
159		490
160	=item C<EVMETHOD_PORT> (solaris 10 only)	491	Use the linux-specific epoll(7) interface (for both pre- and post-2.6.9
		492	kernels).
161		493
162	If one or more of these are ored into the flags value, then only these	494	For few fds, this backend is a bit little slower than poll and select, but
163	backends will be tried (in the reverse order as given here). If one are	495	it scales phenomenally better. While poll and select usually scale like
164	specified, any backend will do.	496	O(total_fds) where total_fds is the total number of fds (or the highest
		497	fd), epoll scales either O(1) or O(active_fds).
		498
		499	The epoll mechanism deserves honorable mention as the most misdesigned
		500	of the more advanced event mechanisms: mere annoyances include silently
		501	dropping file descriptors, requiring a system call per change per file
		502	descriptor (and unnecessary guessing of parameters), problems with dup,
		503	returning before the timeout value, resulting in additional iterations
		504	(and only giving 5ms accuracy while select on the same platform gives
		505	0.1ms) and so on. The biggest issue is fork races, however - if a program
		506	forks then I<both> parent and child process have to recreate the epoll
		507	set, which can take considerable time (one syscall per file descriptor)
		508	and is of course hard to detect.
		509
		510	Epoll is also notoriously buggy - embedding epoll fds I<should> work,
		511	but of course I<doesn't>, and epoll just loves to report events for
		512	totally I<different> file descriptors (even already closed ones, so
		513	one cannot even remove them from the set) than registered in the set
		514	(especially on SMP systems). Libev tries to counter these spurious
		515	notifications by employing an additional generation counter and comparing
		516	that against the events to filter out spurious ones, recreating the set
		517	when required. Epoll also erroneously rounds down timeouts, but gives you
		518	no way to know when and by how much, so sometimes you have to busy-wait
		519	because epoll returns immediately despite a nonzero timeout. And last
		520	not least, it also refuses to work with some file descriptors which work
		521	perfectly fine with C<select> (files, many character devices...).
		522
		523	Epoll is truly the train wreck among event poll mechanisms, a frankenpoll,
		524	cobbled together in a hurry, no thought to design or interaction with
		525	others. Oh, the pain, will it ever stop...
		526
		527	While stopping, setting and starting an I/O watcher in the same iteration
		528	will result in some caching, there is still a system call per such
		529	incident (because the same I<file descriptor> could point to a different
		530	I<file description> now), so its best to avoid that. Also, C<dup ()>'ed
		531	file descriptors might not work very well if you register events for both
		532	file descriptors.
		533
		534	Best performance from this backend is achieved by not unregistering all
		535	watchers for a file descriptor until it has been closed, if possible,
		536	i.e. keep at least one watcher active per fd at all times. Stopping and
		537	starting a watcher (without re-setting it) also usually doesn't cause
		538	extra overhead. A fork can both result in spurious notifications as well
		539	as in libev having to destroy and recreate the epoll object, which can
		540	take considerable time and thus should be avoided.
		541
		542	All this means that, in practice, C<EVBACKEND_SELECT> can be as fast or
		543	faster than epoll for maybe up to a hundred file descriptors, depending on
		544	the usage. So sad.
		545
		546	While nominally embeddable in other event loops, this feature is broken in
		547	all kernel versions tested so far.
		548
		549	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
		550	C<EVBACKEND_POLL>.
		551
		552	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)
		553
		554	Kqueue deserves special mention, as at the time of this writing, it
		555	was broken on all BSDs except NetBSD (usually it doesn't work reliably
		556	with anything but sockets and pipes, except on Darwin, where of course
		557	it's completely useless). Unlike epoll, however, whose brokenness
		558	is by design, these kqueue bugs can (and eventually will) be fixed
		559	without API changes to existing programs. For this reason it's not being
		560	"auto-detected" unless you explicitly specify it in the flags (i.e. using
		561	C<EVBACKEND_KQUEUE>) or libev was compiled on a known-to-be-good (-enough)
		562	system like NetBSD.
		563
		564	You still can embed kqueue into a normal poll or select backend and use it
		565	only for sockets (after having made sure that sockets work with kqueue on
		566	the target platform). See C<ev_embed> watchers for more info.
		567
		568	It scales in the same way as the epoll backend, but the interface to the
		569	kernel is more efficient (which says nothing about its actual speed, of
		570	course). While stopping, setting and starting an I/O watcher does never
		571	cause an extra system call as with C<EVBACKEND_EPOLL>, it still adds up to
		572	two event changes per incident. Support for C<fork ()> is very bad (you
		573	might have to leak fd's on fork, but it's more sane than epoll) and it
		574	drops fds silently in similarly hard-to-detect cases
		575
		576	This backend usually performs well under most conditions.
		577
		578	While nominally embeddable in other event loops, this doesn't work
		579	everywhere, so you might need to test for this. And since it is broken
		580	almost everywhere, you should only use it when you have a lot of sockets
		581	(for which it usually works), by embedding it into another event loop
		582	(e.g. C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> (but C<poll> is of course
		583	also broken on OS X)) and, did I mention it, using it only for sockets.
		584
		585	This backend maps C<EV_READ> into an C<EVFILT_READ> kevent with
		586	C<NOTE_EOF>, and C<EV_WRITE> into an C<EVFILT_WRITE> kevent with
		587	C<NOTE_EOF>.
		588
		589	=item C<EVBACKEND_DEVPOLL> (value 16, Solaris 8)
		590
		591	This is not implemented yet (and might never be, unless you send me an
		592	implementation). According to reports, C</dev/poll> only supports sockets
		593	and is not embeddable, which would limit the usefulness of this backend
		594	immensely.
		595
		596	=item C<EVBACKEND_PORT> (value 32, Solaris 10)
		597
		598	This uses the Solaris 10 event port mechanism. As with everything on Solaris,
		599	it's really slow, but it still scales very well (O(active_fds)).
		600
		601	While this backend scales well, it requires one system call per active
		602	file descriptor per loop iteration. For small and medium numbers of file
		603	descriptors a "slow" C<EVBACKEND_SELECT> or C<EVBACKEND_POLL> backend
		604	might perform better.
		605
		606	On the positive side, this backend actually performed fully to
		607	specification in all tests and is fully embeddable, which is a rare feat
		608	among the OS-specific backends (I vastly prefer correctness over speed
		609	hacks).
		610
		611	On the negative side, the interface is I<bizarre> - so bizarre that
		612	even sun itself gets it wrong in their code examples: The event polling
		613	function sometimes returns events to the caller even though an error
		614	occurred, but with no indication whether it has done so or not (yes, it's
		615	even documented that way) - deadly for edge-triggered interfaces where you
		616	absolutely have to know whether an event occurred or not because you have
		617	to re-arm the watcher.
		618
		619	Fortunately libev seems to be able to work around these idiocies.
		620
		621	This backend maps C<EV_READ> and C<EV_WRITE> in the same way as
		622	C<EVBACKEND_POLL>.
		623
		624	=item C<EVBACKEND_ALL>
		625
		626	Try all backends (even potentially broken ones that wouldn't be tried
		627	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as
		628	C<EVBACKEND_ALL & ~EVBACKEND_KQUEUE>.
		629
		630	It is definitely not recommended to use this flag, use whatever
		631	C<ev_recommended_backends ()> returns, or simply do not specify a backend
		632	at all.
		633
		634	=item C<EVBACKEND_MASK>
		635
		636	Not a backend at all, but a mask to select all backend bits from a
		637	C<flags> value, in case you want to mask out any backends from a flags
		638	value (e.g. when modifying the C<LIBEV_FLAGS> environment variable).
165		639
166	=back	640	=back
167		641
168	=item struct ev_loop *ev_loop_new (unsigned int flags)	642	If one or more of the backend flags are or'ed into the flags value,
		643	then only these backends will be tried (in the reverse order as listed
		644	here). If none are specified, all backends in C<ev_recommended_backends
		645	()> will be tried.
169		646
170	Similar to C<ev_default_loop>, but always creates a new event loop that is	647	Example: Try to create a event loop that uses epoll and nothing else.
171	always distinct from the default loop. Unlike the default loop, it cannot
172	handle signal and child watchers, and attempts to do so will be greeted by
173	undefined behaviour (or a failed assertion if assertions are enabled).
174		648
175	=item ev_default_destroy ()	649	struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV);
		650	if (!epoller)
		651	fatal ("no epoll found here, maybe it hides under your chair");
176		652
177	Destroys the default loop again (frees all memory and kernel state	653	Example: Use whatever libev has to offer, but make sure that kqueue is
178	etc.). This stops all registered event watchers (by not touching them in	654	used if available.
179	any way whatsoever, although you cannot rely on this :).	655
		656	struct ev_loop *loop = ev_loop_new (ev_recommended_backends () | EVBACKEND_KQUEUE);
180		657
181	=item ev_loop_destroy (loop)	658	=item ev_loop_destroy (loop)
182		659
183	Like C<ev_default_destroy>, but destroys an event loop created by an	660	Destroys an event loop object (frees all memory and kernel state
184	earlier call to C<ev_loop_new>.	661	etc.). None of the active event watchers will be stopped in the normal
		662	sense, so e.g. C<ev_is_active> might still return true. It is your
		663	responsibility to either stop all watchers cleanly yourself I<before>
		664	calling this function, or cope with the fact afterwards (which is usually
		665	the easiest thing, you can just ignore the watchers and/or C<free ()> them
		666	for example).
185		667
186	=item ev_default_fork ()	668	Note that certain global state, such as signal state (and installed signal
		669	handlers), will not be freed by this function, and related watchers (such
		670	as signal and child watchers) would need to be stopped manually.
187		671
		672	This function is normally used on loop objects allocated by
		673	C<ev_loop_new>, but it can also be used on the default loop returned by
		674	C<ev_default_loop>, in which case it is not thread-safe.
		675
		676	Note that it is not advisable to call this function on the default loop
		677	except in the rare occasion where you really need to free its resources.
		678	If you need dynamically allocated loops it is better to use C<ev_loop_new>
		679	and C<ev_loop_destroy>.
		680
		681	=item ev_loop_fork (loop)
		682
		683	This function sets a flag that causes subsequent C<ev_run> iterations to
188	This function reinitialises the kernel state for backends that have	684	reinitialise the kernel state for backends that have one. Despite the
189	one. Despite the name, you can call it anytime, but it makes most sense	685	name, you can call it anytime, but it makes most sense after forking, in
190	after forking, in either the parent or child process (or both, but that	686	the child process. You I<must> call it (or use C<EVFLAG_FORKCHECK>) in the
191	again makes little sense).	687	child before resuming or calling C<ev_run>.
192		688
193	You I<must> call this function after forking if and only if you want to	689	Again, you I<have> to call it on I<any> loop that you want to re-use after
194	use the event library in both processes. If you just fork+exec, you don't	690	a fork, I<even if you do not plan to use the loop in the parent>. This is
195	have to call it.	691	because some kernel interfaces *cough* I<kqueue> *cough* do funny things
		692	during fork.
		693
		694	On the other hand, you only need to call this function in the child
		695	process if and only if you want to use the event loop in the child. If
		696	you just fork+exec or create a new loop in the child, you don't have to
		697	call it at all (in fact, C<epoll> is so badly broken that it makes a
		698	difference, but libev will usually detect this case on its own and do a
		699	costly reset of the backend).
196		700
197	The function itself is quite fast and it's usually not a problem to call	701	The function itself is quite fast and it's usually not a problem to call
198	it just in case after a fork. To make this easy, the function will fit in	702	it just in case after a fork.
199	quite nicely into a call to C<pthread_atfork>:
200		703
		704	Example: Automate calling C<ev_loop_fork> on the default loop when
		705	using pthreads.
		706
		707	static void
		708	post_fork_child (void)
		709	{
		710	ev_loop_fork (EV_DEFAULT);
		711	}
		712
		713	...
201	pthread_atfork (0, 0, ev_default_fork);	714	pthread_atfork (0, 0, post_fork_child);
202		715
203	=item ev_loop_fork (loop)	716	=item int ev_is_default_loop (loop)
204		717
205	Like C<ev_default_fork>, but acts on an event loop created by	718	Returns true when the given loop is, in fact, the default loop, and false
206	C<ev_loop_new>. Yes, you have to call this on every allocated event loop	719	otherwise.
207	after fork, and how you do this is entirely your own problem.
208		720
		721	=item unsigned int ev_iteration (loop)
		722
		723	Returns the current iteration count for the event loop, which is identical
		724	to the number of times libev did poll for new events. It starts at C<0>
		725	and happily wraps around with enough iterations.
		726
		727	This value can sometimes be useful as a generation counter of sorts (it
		728	"ticks" the number of loop iterations), as it roughly corresponds with
		729	C<ev_prepare> and C<ev_check> calls - and is incremented between the
		730	prepare and check phases.
		731
209	=item unsigned int ev_method (loop)	732	=item unsigned int ev_depth (loop)
210		733
		734	Returns the number of times C<ev_run> was entered minus the number of
		735	times C<ev_run> was exited normally, in other words, the recursion depth.
		736
		737	Outside C<ev_run>, this number is zero. In a callback, this number is
		738	C<1>, unless C<ev_run> was invoked recursively (or from another thread),
		739	in which case it is higher.
		740
		741	Leaving C<ev_run> abnormally (setjmp/longjmp, cancelling the thread,
		742	throwing an exception etc.), doesn't count as "exit" - consider this
		743	as a hint to avoid such ungentleman-like behaviour unless it's really
		744	convenient, in which case it is fully supported.
		745
		746	=item unsigned int ev_backend (loop)
		747
211	Returns one of the C<EVMETHOD_*> flags indicating the event backend in	748	Returns one of the C<EVBACKEND_*> flags indicating the event backend in
212	use.	749	use.
213		750
214	=item ev_tstamp ev_now (loop)	751	=item ev_tstamp ev_now (loop)
215		752
216	Returns the current "event loop time", which is the time the event loop	753	Returns the current "event loop time", which is the time the event loop
217	got events and started processing them. This timestamp does not change	754	received events and started processing them. This timestamp does not
218	as long as callbacks are being processed, and this is also the base time	755	change as long as callbacks are being processed, and this is also the base
219	used for relative timers. You can treat it as the timestamp of the event	756	time used for relative timers. You can treat it as the timestamp of the
220	occuring (or more correctly, the mainloop finding out about it).	757	event occurring (or more correctly, libev finding out about it).
221		758
		759	=item ev_now_update (loop)
		760
		761	Establishes the current time by querying the kernel, updating the time
		762	returned by C<ev_now ()> in the progress. This is a costly operation and
		763	is usually done automatically within C<ev_run ()>.
		764
		765	This function is rarely useful, but when some event callback runs for a
		766	very long time without entering the event loop, updating libev's idea of
		767	the current time is a good idea.
		768
		769	See also L<The special problem of time updates> in the C<ev_timer> section.
		770
		771	=item ev_suspend (loop)
		772
		773	=item ev_resume (loop)
		774
		775	These two functions suspend and resume an event loop, for use when the
		776	loop is not used for a while and timeouts should not be processed.
		777
		778	A typical use case would be an interactive program such as a game: When
		779	the user presses C<^Z> to suspend the game and resumes it an hour later it
		780	would be best to handle timeouts as if no time had actually passed while
		781	the program was suspended. This can be achieved by calling C<ev_suspend>
		782	in your C<SIGTSTP> handler, sending yourself a C<SIGSTOP> and calling
		783	C<ev_resume> directly afterwards to resume timer processing.
		784
		785	Effectively, all C<ev_timer> watchers will be delayed by the time spend
		786	between C<ev_suspend> and C<ev_resume>, and all C<ev_periodic> watchers
		787	will be rescheduled (that is, they will lose any events that would have
		788	occurred while suspended).
		789
		790	After calling C<ev_suspend> you B<must not> call I<any> function on the
		791	given loop other than C<ev_resume>, and you B<must not> call C<ev_resume>
		792	without a previous call to C<ev_suspend>.
		793
		794	Calling C<ev_suspend>/C<ev_resume> has the side effect of updating the
		795	event loop time (see C<ev_now_update>).
		796
222	=item ev_loop (loop, int flags)	797	=item bool ev_run (loop, int flags)
223		798
224	Finally, this is it, the event handler. This function usually is called	799	Finally, this is it, the event handler. This function usually is called
225	after you initialised all your watchers and you want to start handling	800	after you have initialised all your watchers and you want to start
226	events.	801	handling events. It will ask the operating system for any new events, call
		802	the watcher callbacks, and then repeat the whole process indefinitely: This
		803	is why event loops are called I<loops>.
227		804
228	If the flags argument is specified as 0, it will not return until either	805	If the flags argument is specified as C<0>, it will keep handling events
229	no event watchers are active anymore or C<ev_unloop> was called.	806	until either no event watchers are active anymore or C<ev_break> was
		807	called.
230		808
		809	The return value is false if there are no more active watchers (which
		810	usually means "all jobs done" or "deadlock"), and true in all other cases
		811	(which usually means " you should call C<ev_run> again").
		812
		813	Please note that an explicit C<ev_break> is usually better than
		814	relying on all watchers to be stopped when deciding when a program has
		815	finished (especially in interactive programs), but having a program
		816	that automatically loops as long as it has to and no longer by virtue
		817	of relying on its watchers stopping correctly, that is truly a thing of
		818	beauty.
		819
		820	This function is I<mostly> exception-safe - you can break out of a
		821	C<ev_run> call by calling C<longjmp> in a callback, throwing a C++
		822	exception and so on. This does not decrement the C<ev_depth> value, nor
		823	will it clear any outstanding C<EVBREAK_ONE> breaks.
		824
231	A flags value of C<EVLOOP_NONBLOCK> will look for new events, will handle	825	A flags value of C<EVRUN_NOWAIT> will look for new events, will handle
232	those events and any outstanding ones, but will not block your process in	826	those events and any already outstanding ones, but will not wait and
233	case there are no events and will return after one iteration of the loop.	827	block your process in case there are no events and will return after one
		828	iteration of the loop. This is sometimes useful to poll and handle new
		829	events while doing lengthy calculations, to keep the program responsive.
234		830
235	A flags value of C<EVLOOP_ONESHOT> will look for new events (waiting if	831	A flags value of C<EVRUN_ONCE> will look for new events (waiting if
236	neccessary) and will handle those and any outstanding ones. It will block	832	necessary) and will handle those and any already outstanding ones. It
237	your process until at least one new event arrives, and will return after	833	will block your process until at least one new event arrives (which could
		834	be an event internal to libev itself, so there is no guarantee that a
		835	user-registered callback will be called), and will return after one
238	one iteration of the loop.	836	iteration of the loop.
239		837
240	This flags value could be used to implement alternative looping	838	This is useful if you are waiting for some external event in conjunction
241	constructs, but the C<prepare> and C<check> watchers provide a better and	839	with something not expressible using other libev watchers (i.e. "roll your
242	more generic mechanism.	840	own C<ev_run>"). However, a pair of C<ev_prepare>/C<ev_check> watchers is
		841	usually a better approach for this kind of thing.
243		842
244	Here are the gory details of what ev_loop does:	843	Here are the gory details of what C<ev_run> does (this is for your
		844	understanding, not a guarantee that things will work exactly like this in
		845	future versions):
245		846
246	1. If there are no active watchers (reference count is zero), return.	847	- Increment loop depth.
		848	- Reset the ev_break status.
		849	- Before the first iteration, call any pending watchers.
		850	LOOP:
		851	- If EVFLAG_FORKCHECK was used, check for a fork.
		852	- If a fork was detected (by any means), queue and call all fork watchers.
247	2. Queue and immediately call all prepare watchers.	853	- Queue and call all prepare watchers.
		854	- If ev_break was called, goto FINISH.
248	3. If we have been forked, recreate the kernel state.	855	- If we have been forked, detach and recreate the kernel state
		856	as to not disturb the other process.
249	4. Update the kernel state with all outstanding changes.	857	- Update the kernel state with all outstanding changes.
250	5. Update the "event loop time".	858	- Update the "event loop time" (ev_now ()).
251	6. Calculate for how long to block.	859	- Calculate for how long to sleep or block, if at all
		860	(active idle watchers, EVRUN_NOWAIT or not having
		861	any active watchers at all will result in not sleeping).
		862	- Sleep if the I/O and timer collect interval say so.
		863	- Increment loop iteration counter.
252	7. Block the process, waiting for events.	864	- Block the process, waiting for any events.
		865	- Queue all outstanding I/O (fd) events.
253	8. Update the "event loop time" and do time jump handling.	866	- Update the "event loop time" (ev_now ()), and do time jump adjustments.
254	9. Queue all outstanding timers.	867	- Queue all expired timers.
255	10. Queue all outstanding periodics.	868	- Queue all expired periodics.
256	11. If no events are pending now, queue all idle watchers.	869	- Queue all idle watchers with priority higher than that of pending events.
257	12. Queue all check watchers.	870	- Queue all check watchers.
258	13. Call all queued watchers in reverse order (i.e. check watchers first).	871	- Call all queued watchers in reverse order (i.e. check watchers first).
259	14. If ev_unloop has been called or EVLOOP_ONESHOT or EVLOOP_NONBLOCK	872	Signals and child watchers are implemented as I/O watchers, and will
260	was used, return, otherwise continue with step #1.	873	be handled here by queueing them when their watcher gets executed.
		874	- If ev_break has been called, or EVRUN_ONCE or EVRUN_NOWAIT
		875	were used, or there are no active watchers, goto FINISH, otherwise
		876	continue with step LOOP.
		877	FINISH:
		878	- Reset the ev_break status iff it was EVBREAK_ONE.
		879	- Decrement the loop depth.
		880	- Return.
261		881
		882	Example: Queue some jobs and then loop until no events are outstanding
		883	anymore.
		884
		885	... queue jobs here, make sure they register event watchers as long
		886	... as they still have work to do (even an idle watcher will do..)
		887	ev_run (my_loop, 0);
		888	... jobs done or somebody called break. yeah!
		889
262	=item ev_unloop (loop, how)	890	=item ev_break (loop, how)
263		891
264	Can be used to make a call to C<ev_loop> return early (but only after it	892	Can be used to make a call to C<ev_run> return early (but only after it
265	has processed all outstanding events). The C<how> argument must be either	893	has processed all outstanding events). The C<how> argument must be either
266	C<EVUNLOOP_ONE>, which will make the innermost C<ev_loop> call return, or	894	C<EVBREAK_ONE>, which will make the innermost C<ev_run> call return, or
267	C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.	895	C<EVBREAK_ALL>, which will make all nested C<ev_run> calls return.
		896
		897	This "break state" will be cleared on the next call to C<ev_run>.
		898
		899	It is safe to call C<ev_break> from outside any C<ev_run> calls, too, in
		900	which case it will have no effect.
268		901
269	=item ev_ref (loop)	902	=item ev_ref (loop)
270		903
271	=item ev_unref (loop)	904	=item ev_unref (loop)
272		905
273	Ref/unref can be used to add or remove a reference count on the event	906	Ref/unref can be used to add or remove a reference count on the event
274	loop: Every watcher keeps one reference, and as long as the reference	907	loop: Every watcher keeps one reference, and as long as the reference
275	count is nonzero, C<ev_loop> will not return on its own. If you have	908	count is nonzero, C<ev_run> will not return on its own.
276	a watcher you never unregister that should not keep C<ev_loop> from	909
277	returning, ev_unref() after starting, and ev_ref() before stopping it. For	910	This is useful when you have a watcher that you never intend to
		911	unregister, but that nevertheless should not keep C<ev_run> from
		912	returning. In such a case, call C<ev_unref> after starting, and C<ev_ref>
		913	before stopping it.
		914
278	example, libev itself uses this for its internal signal pipe: It is not	915	As an example, libev itself uses this for its internal signal pipe: It
279	visible to the libev user and should not keep C<ev_loop> from exiting if	916	is not visible to the libev user and should not keep C<ev_run> from
280	no event watchers registered by it are active. It is also an excellent	917	exiting if no event watchers registered by it are active. It is also an
281	way to do this for generic recurring timers or from within third-party	918	excellent way to do this for generic recurring timers or from within
282	libraries. Just remember to I<unref after start> and I<ref before stop>.	919	third-party libraries. Just remember to I<unref after start> and I<ref
		920	before stop> (but only if the watcher wasn't active before, or was active
		921	before, respectively. Note also that libev might stop watchers itself
		922	(e.g. non-repeating timers) in which case you have to C<ev_ref>
		923	in the callback).
		924
		925	Example: Create a signal watcher, but keep it from keeping C<ev_run>
		926	running when nothing else is active.
		927
		928	ev_signal exitsig;
		929	ev_signal_init (&exitsig, sig_cb, SIGINT);
		930	ev_signal_start (loop, &exitsig);
		931	ev_unref (loop);
		932
		933	Example: For some weird reason, unregister the above signal handler again.
		934
		935	ev_ref (loop);
		936	ev_signal_stop (loop, &exitsig);
		937
		938	=item ev_set_io_collect_interval (loop, ev_tstamp interval)
		939
		940	=item ev_set_timeout_collect_interval (loop, ev_tstamp interval)
		941
		942	These advanced functions influence the time that libev will spend waiting
		943	for events. Both time intervals are by default C<0>, meaning that libev
		944	will try to invoke timer/periodic callbacks and I/O callbacks with minimum
		945	latency.
		946
		947	Setting these to a higher value (the C<interval> I<must> be >= C<0>)
		948	allows libev to delay invocation of I/O and timer/periodic callbacks
		949	to increase efficiency of loop iterations (or to increase power-saving
		950	opportunities).
		951
		952	The idea is that sometimes your program runs just fast enough to handle
		953	one (or very few) event(s) per loop iteration. While this makes the
		954	program responsive, it also wastes a lot of CPU time to poll for new
		955	events, especially with backends like C<select ()> which have a high
		956	overhead for the actual polling but can deliver many events at once.
		957
		958	By setting a higher I<io collect interval> you allow libev to spend more
		959	time collecting I/O events, so you can handle more events per iteration,
		960	at the cost of increasing latency. Timeouts (both C<ev_periodic> and
		961	C<ev_timer>) will not be affected. Setting this to a non-null value will
		962	introduce an additional C<ev_sleep ()> call into most loop iterations. The
		963	sleep time ensures that libev will not poll for I/O events more often then
		964	once per this interval, on average (as long as the host time resolution is
		965	good enough).
		966
		967	Likewise, by setting a higher I<timeout collect interval> you allow libev
		968	to spend more time collecting timeouts, at the expense of increased
		969	latency/jitter/inexactness (the watcher callback will be called
		970	later). C<ev_io> watchers will not be affected. Setting this to a non-null
		971	value will not introduce any overhead in libev.
		972
		973	Many (busy) programs can usually benefit by setting the I/O collect
		974	interval to a value near C<0.1> or so, which is often enough for
		975	interactive servers (of course not for games), likewise for timeouts. It
		976	usually doesn't make much sense to set it to a lower value than C<0.01>,
		977	as this approaches the timing granularity of most systems. Note that if
		978	you do transactions with the outside world and you can't increase the
		979	parallelity, then this setting will limit your transaction rate (if you
		980	need to poll once per transaction and the I/O collect interval is 0.01,
		981	then you can't do more than 100 transactions per second).
		982
		983	Setting the I<timeout collect interval> can improve the opportunity for
		984	saving power, as the program will "bundle" timer callback invocations that
		985	are "near" in time together, by delaying some, thus reducing the number of
		986	times the process sleeps and wakes up again. Another useful technique to
		987	reduce iterations/wake-ups is to use C<ev_periodic> watchers and make sure
		988	they fire on, say, one-second boundaries only.
		989
		990	Example: we only need 0.1s timeout granularity, and we wish not to poll
		991	more often than 100 times per second:
		992
		993	ev_set_timeout_collect_interval (EV_DEFAULT_UC_ 0.1);
		994	ev_set_io_collect_interval (EV_DEFAULT_UC_ 0.01);
		995
		996	=item ev_invoke_pending (loop)
		997
		998	This call will simply invoke all pending watchers while resetting their
		999	pending state. Normally, C<ev_run> does this automatically when required,
		1000	but when overriding the invoke callback this call comes handy. This
		1001	function can be invoked from a watcher - this can be useful for example
		1002	when you want to do some lengthy calculation and want to pass further
		1003	event handling to another thread (you still have to make sure only one
		1004	thread executes within C<ev_invoke_pending> or C<ev_run> of course).
		1005
		1006	=item int ev_pending_count (loop)
		1007
		1008	Returns the number of pending watchers - zero indicates that no watchers
		1009	are pending.
		1010
		1011	=item ev_set_invoke_pending_cb (loop, void (*invoke_pending_cb)(EV_P))
		1012
		1013	This overrides the invoke pending functionality of the loop: Instead of
		1014	invoking all pending watchers when there are any, C<ev_run> will call
		1015	this callback instead. This is useful, for example, when you want to
		1016	invoke the actual watchers inside another context (another thread etc.).
		1017
		1018	If you want to reset the callback, use C<ev_invoke_pending> as new
		1019	callback.
		1020
		1021	=item ev_set_loop_release_cb (loop, void (*release)(EV_P) throw (), void (*acquire)(EV_P) throw ())
		1022
		1023	Sometimes you want to share the same loop between multiple threads. This
		1024	can be done relatively simply by putting mutex_lock/unlock calls around
		1025	each call to a libev function.
		1026
		1027	However, C<ev_run> can run an indefinite time, so it is not feasible
		1028	to wait for it to return. One way around this is to wake up the event
		1029	loop via C<ev_break> and C<ev_async_send>, another way is to set these
		1030	I<release> and I<acquire> callbacks on the loop.
		1031
		1032	When set, then C<release> will be called just before the thread is
		1033	suspended waiting for new events, and C<acquire> is called just
		1034	afterwards.
		1035
		1036	Ideally, C<release> will just call your mutex_unlock function, and
		1037	C<acquire> will just call the mutex_lock function again.
		1038
		1039	While event loop modifications are allowed between invocations of
		1040	C<release> and C<acquire> (that's their only purpose after all), no
		1041	modifications done will affect the event loop, i.e. adding watchers will
		1042	have no effect on the set of file descriptors being watched, or the time
		1043	waited. Use an C<ev_async> watcher to wake up C<ev_run> when you want it
		1044	to take note of any changes you made.
		1045
		1046	In theory, threads executing C<ev_run> will be async-cancel safe between
		1047	invocations of C<release> and C<acquire>.
		1048
		1049	See also the locking example in the C<THREADS> section later in this
		1050	document.
		1051
		1052	=item ev_set_userdata (loop, void *data)
		1053
		1054	=item void *ev_userdata (loop)
		1055
		1056	Set and retrieve a single C<void *> associated with a loop. When
		1057	C<ev_set_userdata> has never been called, then C<ev_userdata> returns
		1058	C<0>.
		1059
		1060	These two functions can be used to associate arbitrary data with a loop,
		1061	and are intended solely for the C<invoke_pending_cb>, C<release> and
		1062	C<acquire> callbacks described above, but of course can be (ab-)used for
		1063	any other purpose as well.
		1064
		1065	=item ev_verify (loop)
		1066
		1067	This function only does something when C<EV_VERIFY> support has been
		1068	compiled in, which is the default for non-minimal builds. It tries to go
		1069	through all internal structures and checks them for validity. If anything
		1070	is found to be inconsistent, it will print an error message to standard
		1071	error and call C<abort ()>.
		1072
		1073	This can be used to catch bugs inside libev itself: under normal
		1074	circumstances, this function will never abort as of course libev keeps its
		1075	data structures consistent.
283		1076
284	=back	1077	=back
285		1078
		1079
286	=head1 ANATOMY OF A WATCHER	1080	=head1 ANATOMY OF A WATCHER
287		1081
		1082	In the following description, uppercase C<TYPE> in names stands for the
		1083	watcher type, e.g. C<ev_TYPE_start> can mean C<ev_timer_start> for timer
		1084	watchers and C<ev_io_start> for I/O watchers.
		1085
288	A watcher is a structure that you create and register to record your	1086	A watcher is an opaque structure that you allocate and register to record
289	interest in some event. For instance, if you want to wait for STDIN to	1087	your interest in some event. To make a concrete example, imagine you want
290	become readable, you would create an C<ev_io> watcher for that:	1088	to wait for STDIN to become readable, you would create an C<ev_io> watcher
		1089	for that:
291		1090
292	static void my_cb (struct ev_loop *loop, struct ev_io *w, int revents)	1091	static void my_cb (struct ev_loop *loop, ev_io *w, int revents)
293	{	1092	{
294	ev_io_stop (w);	1093	ev_io_stop (w);
295	ev_unloop (loop, EVUNLOOP_ALL);	1094	ev_break (loop, EVBREAK_ALL);
296	}	1095	}
297		1096
298	struct ev_loop *loop = ev_default_loop (0);	1097	struct ev_loop *loop = ev_default_loop (0);
		1098
299	struct ev_io stdin_watcher;	1099	ev_io stdin_watcher;
		1100
300	ev_init (&stdin_watcher, my_cb);	1101	ev_init (&stdin_watcher, my_cb);
301	ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ);	1102	ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ);
302	ev_io_start (loop, &stdin_watcher);	1103	ev_io_start (loop, &stdin_watcher);
		1104
303	ev_loop (loop, 0);	1105	ev_run (loop, 0);
304		1106
305	As you can see, you are responsible for allocating the memory for your	1107	As you can see, you are responsible for allocating the memory for your
306	watcher structures (and it is usually a bad idea to do this on the stack,	1108	watcher structures (and it is I<usually> a bad idea to do this on the
307	although this can sometimes be quite valid).	1109	stack).
308		1110
		1111	Each watcher has an associated watcher structure (called C<struct ev_TYPE>
		1112	or simply C<ev_TYPE>, as typedefs are provided for all watcher structs).
		1113
309	Each watcher structure must be initialised by a call to C<ev_init	1114	Each watcher structure must be initialised by a call to C<ev_init (watcher
310	(watcher *, callback)>, which expects a callback to be provided. This	1115	*, callback)>, which expects a callback to be provided. This callback is
311	callback gets invoked each time the event occurs (or, in the case of io	1116	invoked each time the event occurs (or, in the case of I/O watchers, each
312	watchers, each time the event loop detects that the file descriptor given	1117	time the event loop detects that the file descriptor given is readable
313	is readable and/or writable).	1118	and/or writable).
314		1119
315	Each watcher type has its own C<< ev_<type>_set (watcher *, ...) >> macro	1120	Each watcher type further has its own C<< ev_TYPE_set (watcher *, ...) >>
316	with arguments specific to this watcher type. There is also a macro	1121	macro to configure it, with arguments specific to the watcher type. There
317	to combine initialisation and setting in one call: C<< ev_<type>_init	1122	is also a macro to combine initialisation and setting in one call: C<<
318	(watcher *, callback, ...) >>.	1123	ev_TYPE_init (watcher *, callback, ...) >>.
319		1124
320	To make the watcher actually watch out for events, you have to start it	1125	To make the watcher actually watch out for events, you have to start it
321	with a watcher-specific start function (C<< ev_<type>_start (loop, watcher	1126	with a watcher-specific start function (C<< ev_TYPE_start (loop, watcher
322	*) >>), and you can stop watching for events at any time by calling the	1127	*) >>), and you can stop watching for events at any time by calling the
323	corresponding stop function (C<< ev_<type>_stop (loop, watcher *) >>.	1128	corresponding stop function (C<< ev_TYPE_stop (loop, watcher *) >>.
324		1129
325	As long as your watcher is active (has been started but not stopped) you	1130	As long as your watcher is active (has been started but not stopped) you
326	must not touch the values stored in it. Most specifically you must never	1131	must not touch the values stored in it. Most specifically you must never
327	reinitialise it or call its set method.	1132	reinitialise it or call its C<ev_TYPE_set> macro.
328
329	You can check whether an event is active by calling the C<ev_is_active
330	(watcher *)> macro. To see whether an event is outstanding (but the
331	callback for it has not been called yet) you can use the C<ev_is_pending
332	(watcher *)> macro.
333		1133
334	Each and every callback receives the event loop pointer as first, the	1134	Each and every callback receives the event loop pointer as first, the
335	registered watcher structure as second, and a bitset of received events as	1135	registered watcher structure as second, and a bitset of received events as
336	third argument.	1136	third argument.
337		1137
…		…
346	=item C<EV_WRITE>	1146	=item C<EV_WRITE>
347		1147
348	The file descriptor in the C<ev_io> watcher has become readable and/or	1148	The file descriptor in the C<ev_io> watcher has become readable and/or
349	writable.	1149	writable.
350		1150
351	=item C<EV_TIMEOUT>	1151	=item C<EV_TIMER>
352		1152
353	The C<ev_timer> watcher has timed out.	1153	The C<ev_timer> watcher has timed out.
354		1154
355	=item C<EV_PERIODIC>	1155	=item C<EV_PERIODIC>
356		1156
…		…
362		1162
363	=item C<EV_CHILD>	1163	=item C<EV_CHILD>
364		1164
365	The pid specified in the C<ev_child> watcher has received a status change.	1165	The pid specified in the C<ev_child> watcher has received a status change.
366		1166
		1167	=item C<EV_STAT>
		1168
		1169	The path specified in the C<ev_stat> watcher changed its attributes somehow.
		1170
367	=item C<EV_IDLE>	1171	=item C<EV_IDLE>
368		1172
369	The C<ev_idle> watcher has determined that you have nothing better to do.	1173	The C<ev_idle> watcher has determined that you have nothing better to do.
370		1174
371	=item C<EV_PREPARE>	1175	=item C<EV_PREPARE>
372		1176
373	=item C<EV_CHECK>	1177	=item C<EV_CHECK>
374		1178
375	All C<ev_prepare> watchers are invoked just I<before> C<ev_loop> starts	1179	All C<ev_prepare> watchers are invoked just I<before> C<ev_run> starts
376	to gather new events, and all C<ev_check> watchers are invoked just after	1180	to gather new events, and all C<ev_check> watchers are invoked just after
377	C<ev_loop> has gathered them, but before it invokes any callbacks for any	1181	C<ev_run> has gathered them, but before it invokes any callbacks for any
378	received events. Callbacks of both watcher types can start and stop as	1182	received events. Callbacks of both watcher types can start and stop as
379	many watchers as they want, and all of them will be taken into account	1183	many watchers as they want, and all of them will be taken into account
380	(for example, a C<ev_prepare> watcher might start an idle watcher to keep	1184	(for example, a C<ev_prepare> watcher might start an idle watcher to keep
381	C<ev_loop> from blocking).	1185	C<ev_run> from blocking).
		1186
		1187	=item C<EV_EMBED>
		1188
		1189	The embedded event loop specified in the C<ev_embed> watcher needs attention.
		1190
		1191	=item C<EV_FORK>
		1192
		1193	The event loop has been resumed in the child process after fork (see
		1194	C<ev_fork>).
		1195
		1196	=item C<EV_CLEANUP>
		1197
		1198	The event loop is about to be destroyed (see C<ev_cleanup>).
		1199
		1200	=item C<EV_ASYNC>
		1201
		1202	The given async watcher has been asynchronously notified (see C<ev_async>).
		1203
		1204	=item C<EV_CUSTOM>
		1205
		1206	Not ever sent (or otherwise used) by libev itself, but can be freely used
		1207	by libev users to signal watchers (e.g. via C<ev_feed_event>).
382		1208
383	=item C<EV_ERROR>	1209	=item C<EV_ERROR>
384		1210
385	An unspecified error has occured, the watcher has been stopped. This might	1211	An unspecified error has occurred, the watcher has been stopped. This might
386	happen because the watcher could not be properly started because libev	1212	happen because the watcher could not be properly started because libev
387	ran out of memory, a file descriptor was found to be closed or any other	1213	ran out of memory, a file descriptor was found to be closed or any other
		1214	problem. Libev considers these application bugs.
		1215
388	problem. You best act on it by reporting the problem and somehow coping	1216	You best act on it by reporting the problem and somehow coping with the
389	with the watcher being stopped.	1217	watcher being stopped. Note that well-written programs should not receive
		1218	an error ever, so when your watcher receives it, this usually indicates a
		1219	bug in your program.
390		1220
391	Libev will usually signal a few "dummy" events together with an error,	1221	Libev will usually signal a few "dummy" events together with an error, for
392	for example it might indicate that a fd is readable or writable, and if	1222	example it might indicate that a fd is readable or writable, and if your
393	your callbacks is well-written it can just attempt the operation and cope	1223	callbacks is well-written it can just attempt the operation and cope with
394	with the error from read() or write(). This will not work in multithreaded	1224	the error from read() or write(). This will not work in multi-threaded
395	programs, though, so beware.	1225	programs, though, as the fd could already be closed and reused for another
		1226	thing, so beware.
396		1227
397	=back	1228	=back
398		1229
399	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER	1230	=head2 GENERIC WATCHER FUNCTIONS
400		1231
401	Each watcher has, by default, a member C<void *data> that you can change	1232	=over 4
402	and read at any time, libev will completely ignore it. This can be used
403	to associate arbitrary data with your watcher. If you need more data and
404	don't want to allocate memory and store a pointer to it in that data
405	member, you can also "subclass" the watcher type and provide your own
406	data:
407		1233
408	struct my_io	1234	=item C<ev_init> (ev_TYPE *watcher, callback)
		1235
		1236	This macro initialises the generic portion of a watcher. The contents
		1237	of the watcher object can be arbitrary (so C<malloc> will do). Only
		1238	the generic parts of the watcher are initialised, you I<need> to call
		1239	the type-specific C<ev_TYPE_set> macro afterwards to initialise the
		1240	type-specific parts. For each type there is also a C<ev_TYPE_init> macro
		1241	which rolls both calls into one.
		1242
		1243	You can reinitialise a watcher at any time as long as it has been stopped
		1244	(or never started) and there are no pending events outstanding.
		1245
		1246	The callback is always of type C<void (*)(struct ev_loop *loop, ev_TYPE *watcher,
		1247	int revents)>.
		1248
		1249	Example: Initialise an C<ev_io> watcher in two steps.
		1250
		1251	ev_io w;
		1252	ev_init (&w, my_cb);
		1253	ev_io_set (&w, STDIN_FILENO, EV_READ);
		1254
		1255	=item C<ev_TYPE_set> (ev_TYPE *watcher, [args])
		1256
		1257	This macro initialises the type-specific parts of a watcher. You need to
		1258	call C<ev_init> at least once before you call this macro, but you can
		1259	call C<ev_TYPE_set> any number of times. You must not, however, call this
		1260	macro on a watcher that is active (it can be pending, however, which is a
		1261	difference to the C<ev_init> macro).
		1262
		1263	Although some watcher types do not have type-specific arguments
		1264	(e.g. C<ev_prepare>) you still need to call its C<set> macro.
		1265
		1266	See C<ev_init>, above, for an example.
		1267
		1268	=item C<ev_TYPE_init> (ev_TYPE *watcher, callback, [args])
		1269
		1270	This convenience macro rolls both C<ev_init> and C<ev_TYPE_set> macro
		1271	calls into a single call. This is the most convenient method to initialise
		1272	a watcher. The same limitations apply, of course.
		1273
		1274	Example: Initialise and set an C<ev_io> watcher in one step.
		1275
		1276	ev_io_init (&w, my_cb, STDIN_FILENO, EV_READ);
		1277
		1278	=item C<ev_TYPE_start> (loop, ev_TYPE *watcher)
		1279
		1280	Starts (activates) the given watcher. Only active watchers will receive
		1281	events. If the watcher is already active nothing will happen.
		1282
		1283	Example: Start the C<ev_io> watcher that is being abused as example in this
		1284	whole section.
		1285
		1286	ev_io_start (EV_DEFAULT_UC, &w);
		1287
		1288	=item C<ev_TYPE_stop> (loop, ev_TYPE *watcher)
		1289
		1290	Stops the given watcher if active, and clears the pending status (whether
		1291	the watcher was active or not).
		1292
		1293	It is possible that stopped watchers are pending - for example,
		1294	non-repeating timers are being stopped when they become pending - but
		1295	calling C<ev_TYPE_stop> ensures that the watcher is neither active nor
		1296	pending. If you want to free or reuse the memory used by the watcher it is
		1297	therefore a good idea to always call its C<ev_TYPE_stop> function.
		1298
		1299	=item bool ev_is_active (ev_TYPE *watcher)
		1300
		1301	Returns a true value iff the watcher is active (i.e. it has been started
		1302	and not yet been stopped). As long as a watcher is active you must not modify
		1303	it.
		1304
		1305	=item bool ev_is_pending (ev_TYPE *watcher)
		1306
		1307	Returns a true value iff the watcher is pending, (i.e. it has outstanding
		1308	events but its callback has not yet been invoked). As long as a watcher
		1309	is pending (but not active) you must not call an init function on it (but
		1310	C<ev_TYPE_set> is safe), you must not change its priority, and you must
		1311	make sure the watcher is available to libev (e.g. you cannot C<free ()>
		1312	it).
		1313
		1314	=item callback ev_cb (ev_TYPE *watcher)
		1315
		1316	Returns the callback currently set on the watcher.
		1317
		1318	=item ev_cb_set (ev_TYPE *watcher, callback)
		1319
		1320	Change the callback. You can change the callback at virtually any time
		1321	(modulo threads).
		1322
		1323	=item ev_set_priority (ev_TYPE *watcher, int priority)
		1324
		1325	=item int ev_priority (ev_TYPE *watcher)
		1326
		1327	Set and query the priority of the watcher. The priority is a small
		1328	integer between C<EV_MAXPRI> (default: C<2>) and C<EV_MINPRI>
		1329	(default: C<-2>). Pending watchers with higher priority will be invoked
		1330	before watchers with lower priority, but priority will not keep watchers
		1331	from being executed (except for C<ev_idle> watchers).
		1332
		1333	If you need to suppress invocation when higher priority events are pending
		1334	you need to look at C<ev_idle> watchers, which provide this functionality.
		1335
		1336	You I<must not> change the priority of a watcher as long as it is active or
		1337	pending.
		1338
		1339	Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is
		1340	fine, as long as you do not mind that the priority value you query might
		1341	or might not have been clamped to the valid range.
		1342
		1343	The default priority used by watchers when no priority has been set is
		1344	always C<0>, which is supposed to not be too high and not be too low :).
		1345
		1346	See L<WATCHER PRIORITY MODELS>, below, for a more thorough treatment of
		1347	priorities.
		1348
		1349	=item ev_invoke (loop, ev_TYPE *watcher, int revents)
		1350
		1351	Invoke the C<watcher> with the given C<loop> and C<revents>. Neither
		1352	C<loop> nor C<revents> need to be valid as long as the watcher callback
		1353	can deal with that fact, as both are simply passed through to the
		1354	callback.
		1355
		1356	=item int ev_clear_pending (loop, ev_TYPE *watcher)
		1357
		1358	If the watcher is pending, this function clears its pending status and
		1359	returns its C<revents> bitset (as if its callback was invoked). If the
		1360	watcher isn't pending it does nothing and returns C<0>.
		1361
		1362	Sometimes it can be useful to "poll" a watcher instead of waiting for its
		1363	callback to be invoked, which can be accomplished with this function.
		1364
		1365	=item ev_feed_event (loop, ev_TYPE *watcher, int revents)
		1366
		1367	Feeds the given event set into the event loop, as if the specified event
		1368	had happened for the specified watcher (which must be a pointer to an
		1369	initialised but not necessarily started event watcher). Obviously you must
		1370	not free the watcher as long as it has pending events.
		1371
		1372	Stopping the watcher, letting libev invoke it, or calling
		1373	C<ev_clear_pending> will clear the pending event, even if the watcher was
		1374	not started in the first place.
		1375
		1376	See also C<ev_feed_fd_event> and C<ev_feed_signal_event> for related
		1377	functions that do not need a watcher.
		1378
		1379	=back
		1380
		1381	See also the L<ASSOCIATING CUSTOM DATA WITH A WATCHER> and L<BUILDING YOUR
		1382	OWN COMPOSITE WATCHERS> idioms.
		1383
		1384	=head2 WATCHER STATES
		1385
		1386	There are various watcher states mentioned throughout this manual -
		1387	active, pending and so on. In this section these states and the rules to
		1388	transition between them will be described in more detail - and while these
		1389	rules might look complicated, they usually do "the right thing".
		1390
		1391	=over 4
		1392
		1393	=item initialiased
		1394
		1395	Before a watcher can be registered with the event loop it has to be
		1396	initialised. This can be done with a call to C<ev_TYPE_init>, or calls to
		1397	C<ev_init> followed by the watcher-specific C<ev_TYPE_set> function.
		1398
		1399	In this state it is simply some block of memory that is suitable for
		1400	use in an event loop. It can be moved around, freed, reused etc. at
		1401	will - as long as you either keep the memory contents intact, or call
		1402	C<ev_TYPE_init> again.
		1403
		1404	=item started/running/active
		1405
		1406	Once a watcher has been started with a call to C<ev_TYPE_start> it becomes
		1407	property of the event loop, and is actively waiting for events. While in
		1408	this state it cannot be accessed (except in a few documented ways), moved,
		1409	freed or anything else - the only legal thing is to keep a pointer to it,
		1410	and call libev functions on it that are documented to work on active watchers.
		1411
		1412	=item pending
		1413
		1414	If a watcher is active and libev determines that an event it is interested
		1415	in has occurred (such as a timer expiring), it will become pending. It will
		1416	stay in this pending state until either it is stopped or its callback is
		1417	about to be invoked, so it is not normally pending inside the watcher
		1418	callback.
		1419
		1420	The watcher might or might not be active while it is pending (for example,
		1421	an expired non-repeating timer can be pending but no longer active). If it
		1422	is stopped, it can be freely accessed (e.g. by calling C<ev_TYPE_set>),
		1423	but it is still property of the event loop at this time, so cannot be
		1424	moved, freed or reused. And if it is active the rules described in the
		1425	previous item still apply.
		1426
		1427	It is also possible to feed an event on a watcher that is not active (e.g.
		1428	via C<ev_feed_event>), in which case it becomes pending without being
		1429	active.
		1430
		1431	=item stopped
		1432
		1433	A watcher can be stopped implicitly by libev (in which case it might still
		1434	be pending), or explicitly by calling its C<ev_TYPE_stop> function. The
		1435	latter will clear any pending state the watcher might be in, regardless
		1436	of whether it was active or not, so stopping a watcher explicitly before
		1437	freeing it is often a good idea.
		1438
		1439	While stopped (and not pending) the watcher is essentially in the
		1440	initialised state, that is, it can be reused, moved, modified in any way
		1441	you wish (but when you trash the memory block, you need to C<ev_TYPE_init>
		1442	it again).
		1443
		1444	=back
		1445
		1446	=head2 WATCHER PRIORITY MODELS
		1447
		1448	Many event loops support I<watcher priorities>, which are usually small
		1449	integers that influence the ordering of event callback invocation
		1450	between watchers in some way, all else being equal.
		1451
		1452	In libev, Watcher priorities can be set using C<ev_set_priority>. See its
		1453	description for the more technical details such as the actual priority
		1454	range.
		1455
		1456	There are two common ways how these these priorities are being interpreted
		1457	by event loops:
		1458
		1459	In the more common lock-out model, higher priorities "lock out" invocation
		1460	of lower priority watchers, which means as long as higher priority
		1461	watchers receive events, lower priority watchers are not being invoked.
		1462
		1463	The less common only-for-ordering model uses priorities solely to order
		1464	callback invocation within a single event loop iteration: Higher priority
		1465	watchers are invoked before lower priority ones, but they all get invoked
		1466	before polling for new events.
		1467
		1468	Libev uses the second (only-for-ordering) model for all its watchers
		1469	except for idle watchers (which use the lock-out model).
		1470
		1471	The rationale behind this is that implementing the lock-out model for
		1472	watchers is not well supported by most kernel interfaces, and most event
		1473	libraries will just poll for the same events again and again as long as
		1474	their callbacks have not been executed, which is very inefficient in the
		1475	common case of one high-priority watcher locking out a mass of lower
		1476	priority ones.
		1477
		1478	Static (ordering) priorities are most useful when you have two or more
		1479	watchers handling the same resource: a typical usage example is having an
		1480	C<ev_io> watcher to receive data, and an associated C<ev_timer> to handle
		1481	timeouts. Under load, data might be received while the program handles
		1482	other jobs, but since timers normally get invoked first, the timeout
		1483	handler will be executed before checking for data. In that case, giving
		1484	the timer a lower priority than the I/O watcher ensures that I/O will be
		1485	handled first even under adverse conditions (which is usually, but not
		1486	always, what you want).
		1487
		1488	Since idle watchers use the "lock-out" model, meaning that idle watchers
		1489	will only be executed when no same or higher priority watchers have
		1490	received events, they can be used to implement the "lock-out" model when
		1491	required.
		1492
		1493	For example, to emulate how many other event libraries handle priorities,
		1494	you can associate an C<ev_idle> watcher to each such watcher, and in
		1495	the normal watcher callback, you just start the idle watcher. The real
		1496	processing is done in the idle watcher callback. This causes libev to
		1497	continuously poll and process kernel event data for the watcher, but when
		1498	the lock-out case is known to be rare (which in turn is rare :), this is
		1499	workable.
		1500
		1501	Usually, however, the lock-out model implemented that way will perform
		1502	miserably under the type of load it was designed to handle. In that case,
		1503	it might be preferable to stop the real watcher before starting the
		1504	idle watcher, so the kernel will not have to process the event in case
		1505	the actual processing will be delayed for considerable time.
		1506
		1507	Here is an example of an I/O watcher that should run at a strictly lower
		1508	priority than the default, and which should only process data when no
		1509	other events are pending:
		1510
		1511	ev_idle idle; // actual processing watcher
		1512	ev_io io; // actual event watcher
		1513
		1514	static void
		1515	io_cb (EV_P_ ev_io *w, int revents)
409	{	1516	{
410	struct ev_io io;	1517	// stop the I/O watcher, we received the event, but
411	int otherfd;	1518	// are not yet ready to handle it.
412	void *somedata;	1519	ev_io_stop (EV_A_ w);
413	struct whatever *mostinteresting;	1520
		1521	// start the idle watcher to handle the actual event.
		1522	// it will not be executed as long as other watchers
		1523	// with the default priority are receiving events.
		1524	ev_idle_start (EV_A_ &idle);
414	}	1525	}
415		1526
416	And since your callback will be called with a pointer to the watcher, you	1527	static void
417	can cast it back to your own type:	1528	idle_cb (EV_P_ ev_idle *w, int revents)
418
419	static void my_cb (struct ev_loop *loop, struct ev_io *w_, int revents)
420	{	1529	{
421	struct my_io *w = (struct my_io *)w_;	1530	// actual processing
422	...	1531	read (STDIN_FILENO, ...);
		1532
		1533	// have to start the I/O watcher again, as
		1534	// we have handled the event
		1535	ev_io_start (EV_P_ &io);
423	}	1536	}
424		1537
425	More interesting and less C-conformant ways of catsing your callback type	1538	// initialisation
426	have been omitted....	1539	ev_idle_init (&idle, idle_cb);
		1540	ev_io_init (&io, io_cb, STDIN_FILENO, EV_READ);
		1541	ev_io_start (EV_DEFAULT_ &io);
		1542
		1543	In the "real" world, it might also be beneficial to start a timer, so that
		1544	low-priority connections can not be locked out forever under load. This
		1545	enables your program to keep a lower latency for important connections
		1546	during short periods of high load, while not completely locking out less
		1547	important ones.
427		1548
428		1549
429	=head1 WATCHER TYPES	1550	=head1 WATCHER TYPES
430		1551
431	This section describes each watcher in detail, but will not repeat	1552	This section describes each watcher in detail, but will not repeat
432	information given in the last section.	1553	information given in the last section. Any initialisation/set macros,
		1554	functions and members specific to the watcher type are explained.
433		1555
		1556	Members are additionally marked with either I<[read-only]>, meaning that,
		1557	while the watcher is active, you can look at the member and expect some
		1558	sensible content, but you must not modify it (you can modify it while the
		1559	watcher is stopped to your hearts content), or I<[read-write]>, which
		1560	means you can expect it to have some sensible content while the watcher
		1561	is active, but you can also modify it. Modifying it may not do something
		1562	sensible or take immediate effect (or do anything at all), but libev will
		1563	not crash or malfunction in any way.
		1564
		1565
434	=head2 C<ev_io> - is this file descriptor readable or writable	1566	=head2 C<ev_io> - is this file descriptor readable or writable?
435		1567
436	I/O watchers check whether a file descriptor is readable or writable	1568	I/O watchers check whether a file descriptor is readable or writable
437	in each iteration of the event loop (This behaviour is called	1569	in each iteration of the event loop, or, more precisely, when reading
438	level-triggering because you keep receiving events as long as the	1570	would not block the process and writing would at least be able to write
439	condition persists. Remember you can stop the watcher if you don't want to	1571	some data. This behaviour is called level-triggering because you keep
440	act on the event and neither want to receive future events).	1572	receiving events as long as the condition persists. Remember you can stop
		1573	the watcher if you don't want to act on the event and neither want to
		1574	receive future events.
441		1575
442	In general you can register as many read and/or write event watchers per	1576	In general you can register as many read and/or write event watchers per
443	fd as you want (as long as you don't confuse yourself). Setting all file	1577	fd as you want (as long as you don't confuse yourself). Setting all file
444	descriptors to non-blocking mode is also usually a good idea (but not	1578	descriptors to non-blocking mode is also usually a good idea (but not
445	required if you know what you are doing).	1579	required if you know what you are doing).
446		1580
447	You have to be careful with dup'ed file descriptors, though. Some backends	1581	Another thing you have to watch out for is that it is quite easy to
448	(the linux epoll backend is a notable example) cannot handle dup'ed file	1582	receive "spurious" readiness notifications, that is, your callback might
449	descriptors correctly if you register interest in two or more fds pointing	1583	be called with C<EV_READ> but a subsequent C<read>(2) will actually block
450	to the same underlying file/socket etc. description (that is, they share	1584	because there is no data. It is very easy to get into this situation even
451	the same underlying "file open").	1585	with a relatively standard program structure. Thus it is best to always
		1586	use non-blocking I/O: An extra C<read>(2) returning C<EAGAIN> is far
		1587	preferable to a program hanging until some data arrives.
452		1588
453	If you must do this, then force the use of a known-to-be-good backend	1589	If you cannot run the fd in non-blocking mode (for example you should
454	(at the time of this writing, this includes only EVMETHOD_SELECT and	1590	not play around with an Xlib connection), then you have to separately
455	EVMETHOD_POLL).	1591	re-test whether a file descriptor is really ready with a known-to-be good
		1592	interface such as poll (fortunately in the case of Xlib, it already does
		1593	this on its own, so its quite safe to use). Some people additionally
		1594	use C<SIGALRM> and an interval timer, just to be sure you won't block
		1595	indefinitely.
		1596
		1597	But really, best use non-blocking mode.
		1598
		1599	=head3 The special problem of disappearing file descriptors
		1600
		1601	Some backends (e.g. kqueue, epoll) need to be told about closing a file
		1602	descriptor (either due to calling C<close> explicitly or any other means,
		1603	such as C<dup2>). The reason is that you register interest in some file
		1604	descriptor, but when it goes away, the operating system will silently drop
		1605	this interest. If another file descriptor with the same number then is
		1606	registered with libev, there is no efficient way to see that this is, in
		1607	fact, a different file descriptor.
		1608
		1609	To avoid having to explicitly tell libev about such cases, libev follows
		1610	the following policy: Each time C<ev_io_set> is being called, libev
		1611	will assume that this is potentially a new file descriptor, otherwise
		1612	it is assumed that the file descriptor stays the same. That means that
		1613	you I<have> to call C<ev_io_set> (or C<ev_io_init>) when you change the
		1614	descriptor even if the file descriptor number itself did not change.
		1615
		1616	This is how one would do it normally anyway, the important point is that
		1617	the libev application should not optimise around libev but should leave
		1618	optimisations to libev.
		1619
		1620	=head3 The special problem of dup'ed file descriptors
		1621
		1622	Some backends (e.g. epoll), cannot register events for file descriptors,
		1623	but only events for the underlying file descriptions. That means when you
		1624	have C<dup ()>'ed file descriptors or weirder constellations, and register
		1625	events for them, only one file descriptor might actually receive events.
		1626
		1627	There is no workaround possible except not registering events
		1628	for potentially C<dup ()>'ed file descriptors, or to resort to
		1629	C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>.
		1630
		1631	=head3 The special problem of files
		1632
		1633	Many people try to use C<select> (or libev) on file descriptors
		1634	representing files, and expect it to become ready when their program
		1635	doesn't block on disk accesses (which can take a long time on their own).
		1636
		1637	However, this cannot ever work in the "expected" way - you get a readiness
		1638	notification as soon as the kernel knows whether and how much data is
		1639	there, and in the case of open files, that's always the case, so you
		1640	always get a readiness notification instantly, and your read (or possibly
		1641	write) will still block on the disk I/O.
		1642
		1643	Another way to view it is that in the case of sockets, pipes, character
		1644	devices and so on, there is another party (the sender) that delivers data
		1645	on its own, but in the case of files, there is no such thing: the disk
		1646	will not send data on its own, simply because it doesn't know what you
		1647	wish to read - you would first have to request some data.
		1648
		1649	Since files are typically not-so-well supported by advanced notification
		1650	mechanism, libev tries hard to emulate POSIX behaviour with respect
		1651	to files, even though you should not use it. The reason for this is
		1652	convenience: sometimes you want to watch STDIN or STDOUT, which is
		1653	usually a tty, often a pipe, but also sometimes files or special devices
		1654	(for example, C<epoll> on Linux works with F</dev/random> but not with
		1655	F</dev/urandom>), and even though the file might better be served with
		1656	asynchronous I/O instead of with non-blocking I/O, it is still useful when
		1657	it "just works" instead of freezing.
		1658
		1659	So avoid file descriptors pointing to files when you know it (e.g. use
		1660	libeio), but use them when it is convenient, e.g. for STDIN/STDOUT, or
		1661	when you rarely read from a file instead of from a socket, and want to
		1662	reuse the same code path.
		1663
		1664	=head3 The special problem of fork
		1665
		1666	Some backends (epoll, kqueue) do not support C<fork ()> at all or exhibit
		1667	useless behaviour. Libev fully supports fork, but needs to be told about
		1668	it in the child if you want to continue to use it in the child.
		1669
		1670	To support fork in your child processes, you have to call C<ev_loop_fork
		1671	()> after a fork in the child, enable C<EVFLAG_FORKCHECK>, or resort to
		1672	C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>.
		1673
		1674	=head3 The special problem of SIGPIPE
		1675
		1676	While not really specific to libev, it is easy to forget about C<SIGPIPE>:
		1677	when writing to a pipe whose other end has been closed, your program gets
		1678	sent a SIGPIPE, which, by default, aborts your program. For most programs
		1679	this is sensible behaviour, for daemons, this is usually undesirable.
		1680
		1681	So when you encounter spurious, unexplained daemon exits, make sure you
		1682	ignore SIGPIPE (and maybe make sure you log the exit status of your daemon
		1683	somewhere, as that would have given you a big clue).
		1684
		1685	=head3 The special problem of accept()ing when you can't
		1686
		1687	Many implementations of the POSIX C<accept> function (for example,
		1688	found in post-2004 Linux) have the peculiar behaviour of not removing a
		1689	connection from the pending queue in all error cases.
		1690
		1691	For example, larger servers often run out of file descriptors (because
		1692	of resource limits), causing C<accept> to fail with C<ENFILE> but not
		1693	rejecting the connection, leading to libev signalling readiness on
		1694	the next iteration again (the connection still exists after all), and
		1695	typically causing the program to loop at 100% CPU usage.
		1696
		1697	Unfortunately, the set of errors that cause this issue differs between
		1698	operating systems, there is usually little the app can do to remedy the
		1699	situation, and no known thread-safe method of removing the connection to
		1700	cope with overload is known (to me).
		1701
		1702	One of the easiest ways to handle this situation is to just ignore it
		1703	- when the program encounters an overload, it will just loop until the
		1704	situation is over. While this is a form of busy waiting, no OS offers an
		1705	event-based way to handle this situation, so it's the best one can do.
		1706
		1707	A better way to handle the situation is to log any errors other than
		1708	C<EAGAIN> and C<EWOULDBLOCK>, making sure not to flood the log with such
		1709	messages, and continue as usual, which at least gives the user an idea of
		1710	what could be wrong ("raise the ulimit!"). For extra points one could stop
		1711	the C<ev_io> watcher on the listening fd "for a while", which reduces CPU
		1712	usage.
		1713
		1714	If your program is single-threaded, then you could also keep a dummy file
		1715	descriptor for overload situations (e.g. by opening F</dev/null>), and
		1716	when you run into C<ENFILE> or C<EMFILE>, close it, run C<accept>,
		1717	close that fd, and create a new dummy fd. This will gracefully refuse
		1718	clients under typical overload conditions.
		1719
		1720	The last way to handle it is to simply log the error and C<exit>, as
		1721	is often done with C<malloc> failures, but this results in an easy
		1722	opportunity for a DoS attack.
		1723
		1724	=head3 Watcher-Specific Functions
456		1725
457	=over 4	1726	=over 4
458		1727
459	=item ev_io_init (ev_io *, callback, int fd, int events)	1728	=item ev_io_init (ev_io *, callback, int fd, int events)
460		1729
461	=item ev_io_set (ev_io *, int fd, int events)	1730	=item ev_io_set (ev_io *, int fd, int events)
462		1731
463	Configures an C<ev_io> watcher. The fd is the file descriptor to rceeive	1732	Configures an C<ev_io> watcher. The C<fd> is the file descriptor to
464	events for and events is either C<EV_READ>, C<EV_WRITE> or C<EV_READ |	1733	receive events for and C<events> is either C<EV_READ>, C<EV_WRITE> or
465	EV_WRITE> to receive the given events.	1734	C<EV_READ | EV_WRITE>, to express the desire to receive the given events.
		1735
		1736	=item int fd [read-only]
		1737
		1738	The file descriptor being watched.
		1739
		1740	=item int events [read-only]
		1741
		1742	The events being watched.
466		1743
467	=back	1744	=back
468		1745
		1746	=head3 Examples
		1747
		1748	Example: Call C<stdin_readable_cb> when STDIN_FILENO has become, well
		1749	readable, but only once. Since it is likely line-buffered, you could
		1750	attempt to read a whole line in the callback.
		1751
		1752	static void
		1753	stdin_readable_cb (struct ev_loop *loop, ev_io *w, int revents)
		1754	{
		1755	ev_io_stop (loop, w);
		1756	.. read from stdin here (or from w->fd) and handle any I/O errors
		1757	}
		1758
		1759	...
		1760	struct ev_loop *loop = ev_default_init (0);
		1761	ev_io stdin_readable;
		1762	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);
		1763	ev_io_start (loop, &stdin_readable);
		1764	ev_run (loop, 0);
		1765
		1766
469	=head2 C<ev_timer> - relative and optionally recurring timeouts	1767	=head2 C<ev_timer> - relative and optionally repeating timeouts
470		1768
471	Timer watchers are simple relative timers that generate an event after a	1769	Timer watchers are simple relative timers that generate an event after a
472	given time, and optionally repeating in regular intervals after that.	1770	given time, and optionally repeating in regular intervals after that.
473		1771
474	The timers are based on real time, that is, if you register an event that	1772	The timers are based on real time, that is, if you register an event that
475	times out after an hour and you reset your system clock to last years	1773	times out after an hour and you reset your system clock to January last
476	time, it will still time out after (roughly) and hour. "Roughly" because	1774	year, it will still time out after (roughly) one hour. "Roughly" because
477	detecting time jumps is hard, and soem inaccuracies are unavoidable (the	1775	detecting time jumps is hard, and some inaccuracies are unavoidable (the
478	monotonic clock option helps a lot here).	1776	monotonic clock option helps a lot here).
		1777
		1778	The callback is guaranteed to be invoked only I<after> its timeout has
		1779	passed (not I<at>, so on systems with very low-resolution clocks this
		1780	might introduce a small delay, see "the special problem of being too
		1781	early", below). If multiple timers become ready during the same loop
		1782	iteration then the ones with earlier time-out values are invoked before
		1783	ones of the same priority with later time-out values (but this is no
		1784	longer true when a callback calls C<ev_run> recursively).
		1785
		1786	=head3 Be smart about timeouts
		1787
		1788	Many real-world problems involve some kind of timeout, usually for error
		1789	recovery. A typical example is an HTTP request - if the other side hangs,
		1790	you want to raise some error after a while.
		1791
		1792	What follows are some ways to handle this problem, from obvious and
		1793	inefficient to smart and efficient.
		1794
		1795	In the following, a 60 second activity timeout is assumed - a timeout that
		1796	gets reset to 60 seconds each time there is activity (e.g. each time some
		1797	data or other life sign was received).
		1798
		1799	=over 4
		1800
		1801	=item 1. Use a timer and stop, reinitialise and start it on activity.
		1802
		1803	This is the most obvious, but not the most simple way: In the beginning,
		1804	start the watcher:
		1805
		1806	ev_timer_init (timer, callback, 60., 0.);
		1807	ev_timer_start (loop, timer);
		1808
		1809	Then, each time there is some activity, C<ev_timer_stop> it, initialise it
		1810	and start it again:
		1811
		1812	ev_timer_stop (loop, timer);
		1813	ev_timer_set (timer, 60., 0.);
		1814	ev_timer_start (loop, timer);
		1815
		1816	This is relatively simple to implement, but means that each time there is
		1817	some activity, libev will first have to remove the timer from its internal
		1818	data structure and then add it again. Libev tries to be fast, but it's
		1819	still not a constant-time operation.
		1820
		1821	=item 2. Use a timer and re-start it with C<ev_timer_again> inactivity.
		1822
		1823	This is the easiest way, and involves using C<ev_timer_again> instead of
		1824	C<ev_timer_start>.
		1825
		1826	To implement this, configure an C<ev_timer> with a C<repeat> value
		1827	of C<60> and then call C<ev_timer_again> at start and each time you
		1828	successfully read or write some data. If you go into an idle state where
		1829	you do not expect data to travel on the socket, you can C<ev_timer_stop>
		1830	the timer, and C<ev_timer_again> will automatically restart it if need be.
		1831
		1832	That means you can ignore both the C<ev_timer_start> function and the
		1833	C<after> argument to C<ev_timer_set>, and only ever use the C<repeat>
		1834	member and C<ev_timer_again>.
		1835
		1836	At start:
		1837
		1838	ev_init (timer, callback);
		1839	timer->repeat = 60.;
		1840	ev_timer_again (loop, timer);
		1841
		1842	Each time there is some activity:
		1843
		1844	ev_timer_again (loop, timer);
		1845
		1846	It is even possible to change the time-out on the fly, regardless of
		1847	whether the watcher is active or not:
		1848
		1849	timer->repeat = 30.;
		1850	ev_timer_again (loop, timer);
		1851
		1852	This is slightly more efficient then stopping/starting the timer each time
		1853	you want to modify its timeout value, as libev does not have to completely
		1854	remove and re-insert the timer from/into its internal data structure.
		1855
		1856	It is, however, even simpler than the "obvious" way to do it.
		1857
		1858	=item 3. Let the timer time out, but then re-arm it as required.
		1859
		1860	This method is more tricky, but usually most efficient: Most timeouts are
		1861	relatively long compared to the intervals between other activity - in
		1862	our example, within 60 seconds, there are usually many I/O events with
		1863	associated activity resets.
		1864
		1865	In this case, it would be more efficient to leave the C<ev_timer> alone,
		1866	but remember the time of last activity, and check for a real timeout only
		1867	within the callback:
		1868
		1869	ev_tstamp timeout = 60.;
		1870	ev_tstamp last_activity; // time of last activity
		1871	ev_timer timer;
		1872
		1873	static void
		1874	callback (EV_P_ ev_timer *w, int revents)
		1875	{
		1876	// calculate when the timeout would happen
		1877	ev_tstamp after = last_activity - ev_now (EV_A) + timeout;
		1878
		1879	// if negative, it means we the timeout already occurred
		1880	if (after < 0.)
		1881	{
		1882	// timeout occurred, take action
		1883	}
		1884	else
		1885	{
		1886	// callback was invoked, but there was some recent
		1887	// activity. simply restart the timer to time out
		1888	// after "after" seconds, which is the earliest time
		1889	// the timeout can occur.
		1890	ev_timer_set (w, after, 0.);
		1891	ev_timer_start (EV_A_ w);
		1892	}
		1893	}
		1894
		1895	To summarise the callback: first calculate in how many seconds the
		1896	timeout will occur (by calculating the absolute time when it would occur,
		1897	C<last_activity + timeout>, and subtracting the current time, C<ev_now
		1898	(EV_A)> from that).
		1899
		1900	If this value is negative, then we are already past the timeout, i.e. we
		1901	timed out, and need to do whatever is needed in this case.
		1902
		1903	Otherwise, we now the earliest time at which the timeout would trigger,
		1904	and simply start the timer with this timeout value.
		1905
		1906	In other words, each time the callback is invoked it will check whether
		1907	the timeout occurred. If not, it will simply reschedule itself to check
		1908	again at the earliest time it could time out. Rinse. Repeat.
		1909
		1910	This scheme causes more callback invocations (about one every 60 seconds
		1911	minus half the average time between activity), but virtually no calls to
		1912	libev to change the timeout.
		1913
		1914	To start the machinery, simply initialise the watcher and set
		1915	C<last_activity> to the current time (meaning there was some activity just
		1916	now), then call the callback, which will "do the right thing" and start
		1917	the timer:
		1918
		1919	last_activity = ev_now (EV_A);
		1920	ev_init (&timer, callback);
		1921	callback (EV_A_ &timer, 0);
		1922
		1923	When there is some activity, simply store the current time in
		1924	C<last_activity>, no libev calls at all:
		1925
		1926	if (activity detected)
		1927	last_activity = ev_now (EV_A);
		1928
		1929	When your timeout value changes, then the timeout can be changed by simply
		1930	providing a new value, stopping the timer and calling the callback, which
		1931	will again do the right thing (for example, time out immediately :).
		1932
		1933	timeout = new_value;
		1934	ev_timer_stop (EV_A_ &timer);
		1935	callback (EV_A_ &timer, 0);
		1936
		1937	This technique is slightly more complex, but in most cases where the
		1938	time-out is unlikely to be triggered, much more efficient.
		1939
		1940	=item 4. Wee, just use a double-linked list for your timeouts.
		1941
		1942	If there is not one request, but many thousands (millions...), all
		1943	employing some kind of timeout with the same timeout value, then one can
		1944	do even better:
		1945
		1946	When starting the timeout, calculate the timeout value and put the timeout
		1947	at the I<end> of the list.
		1948
		1949	Then use an C<ev_timer> to fire when the timeout at the I<beginning> of
		1950	the list is expected to fire (for example, using the technique #3).
		1951
		1952	When there is some activity, remove the timer from the list, recalculate
		1953	the timeout, append it to the end of the list again, and make sure to
		1954	update the C<ev_timer> if it was taken from the beginning of the list.
		1955
		1956	This way, one can manage an unlimited number of timeouts in O(1) time for
		1957	starting, stopping and updating the timers, at the expense of a major
		1958	complication, and having to use a constant timeout. The constant timeout
		1959	ensures that the list stays sorted.
		1960
		1961	=back
		1962
		1963	So which method the best?
		1964
		1965	Method #2 is a simple no-brain-required solution that is adequate in most
		1966	situations. Method #3 requires a bit more thinking, but handles many cases
		1967	better, and isn't very complicated either. In most case, choosing either
		1968	one is fine, with #3 being better in typical situations.
		1969
		1970	Method #1 is almost always a bad idea, and buys you nothing. Method #4 is
		1971	rather complicated, but extremely efficient, something that really pays
		1972	off after the first million or so of active timers, i.e. it's usually
		1973	overkill :)
		1974
		1975	=head3 The special problem of being too early
		1976
		1977	If you ask a timer to call your callback after three seconds, then
		1978	you expect it to be invoked after three seconds - but of course, this
		1979	cannot be guaranteed to infinite precision. Less obviously, it cannot be
		1980	guaranteed to any precision by libev - imagine somebody suspending the
		1981	process with a STOP signal for a few hours for example.
		1982
		1983	So, libev tries to invoke your callback as soon as possible I<after> the
		1984	delay has occurred, but cannot guarantee this.
		1985
		1986	A less obvious failure mode is calling your callback too early: many event
		1987	loops compare timestamps with a "elapsed delay >= requested delay", but
		1988	this can cause your callback to be invoked much earlier than you would
		1989	expect.
		1990
		1991	To see why, imagine a system with a clock that only offers full second
		1992	resolution (think windows if you can't come up with a broken enough OS
		1993	yourself). If you schedule a one-second timer at the time 500.9, then the
		1994	event loop will schedule your timeout to elapse at a system time of 500
		1995	(500.9 truncated to the resolution) + 1, or 501.
		1996
		1997	If an event library looks at the timeout 0.1s later, it will see "501 >=
		1998	501" and invoke the callback 0.1s after it was started, even though a
		1999	one-second delay was requested - this is being "too early", despite best
		2000	intentions.
		2001
		2002	This is the reason why libev will never invoke the callback if the elapsed
		2003	delay equals the requested delay, but only when the elapsed delay is
		2004	larger than the requested delay. In the example above, libev would only invoke
		2005	the callback at system time 502, or 1.1s after the timer was started.
		2006
		2007	So, while libev cannot guarantee that your callback will be invoked
		2008	exactly when requested, it I<can> and I<does> guarantee that the requested
		2009	delay has actually elapsed, or in other words, it always errs on the "too
		2010	late" side of things.
		2011
		2012	=head3 The special problem of time updates
		2013
		2014	Establishing the current time is a costly operation (it usually takes
		2015	at least one system call): EV therefore updates its idea of the current
		2016	time only before and after C<ev_run> collects new events, which causes a
		2017	growing difference between C<ev_now ()> and C<ev_time ()> when handling
		2018	lots of events in one iteration.
479		2019
480	The relative timeouts are calculated relative to the C<ev_now ()>	2020	The relative timeouts are calculated relative to the C<ev_now ()>
481	time. This is usually the right thing as this timestamp refers to the time	2021	time. This is usually the right thing as this timestamp refers to the time
482	of the event triggering whatever timeout you are modifying/starting. If	2022	of the event triggering whatever timeout you are modifying/starting. If
483	you suspect event processing to be delayed and you *need* to base the timeout	2023	you suspect event processing to be delayed and you I<need> to base the
484	on the current time, use something like this to adjust for this:	2024	timeout on the current time, use something like this to adjust for this:
485		2025
486	ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);	2026	ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);
487		2027
		2028	If the event loop is suspended for a long time, you can also force an
		2029	update of the time returned by C<ev_now ()> by calling C<ev_now_update
		2030	()>.
		2031
		2032	=head3 The special problem of unsynchronised clocks
		2033
		2034	Modern systems have a variety of clocks - libev itself uses the normal
		2035	"wall clock" clock and, if available, the monotonic clock (to avoid time
		2036	jumps).
		2037
		2038	Neither of these clocks is synchronised with each other or any other clock
		2039	on the system, so C<ev_time ()> might return a considerably different time
		2040	than C<gettimeofday ()> or C<time ()>. On a GNU/Linux system, for example,
		2041	a call to C<gettimeofday> might return a second count that is one higher
		2042	than a directly following call to C<time>.
		2043
		2044	The moral of this is to only compare libev-related timestamps with
		2045	C<ev_time ()> and C<ev_now ()>, at least if you want better precision than
		2046	a second or so.
		2047
		2048	One more problem arises due to this lack of synchronisation: if libev uses
		2049	the system monotonic clock and you compare timestamps from C<ev_time>
		2050	or C<ev_now> from when you started your timer and when your callback is
		2051	invoked, you will find that sometimes the callback is a bit "early".
		2052
		2053	This is because C<ev_timer>s work in real time, not wall clock time, so
		2054	libev makes sure your callback is not invoked before the delay happened,
		2055	I<measured according to the real time>, not the system clock.
		2056
		2057	If your timeouts are based on a physical timescale (e.g. "time out this
		2058	connection after 100 seconds") then this shouldn't bother you as it is
		2059	exactly the right behaviour.
		2060
		2061	If you want to compare wall clock/system timestamps to your timers, then
		2062	you need to use C<ev_periodic>s, as these are based on the wall clock
		2063	time, where your comparisons will always generate correct results.
		2064
		2065	=head3 The special problems of suspended animation
		2066
		2067	When you leave the server world it is quite customary to hit machines that
		2068	can suspend/hibernate - what happens to the clocks during such a suspend?
		2069
		2070	Some quick tests made with a Linux 2.6.28 indicate that a suspend freezes
		2071	all processes, while the clocks (C<times>, C<CLOCK_MONOTONIC>) continue
		2072	to run until the system is suspended, but they will not advance while the
		2073	system is suspended. That means, on resume, it will be as if the program
		2074	was frozen for a few seconds, but the suspend time will not be counted
		2075	towards C<ev_timer> when a monotonic clock source is used. The real time
		2076	clock advanced as expected, but if it is used as sole clocksource, then a
		2077	long suspend would be detected as a time jump by libev, and timers would
		2078	be adjusted accordingly.
		2079
		2080	I would not be surprised to see different behaviour in different between
		2081	operating systems, OS versions or even different hardware.
		2082
		2083	The other form of suspend (job control, or sending a SIGSTOP) will see a
		2084	time jump in the monotonic clocks and the realtime clock. If the program
		2085	is suspended for a very long time, and monotonic clock sources are in use,
		2086	then you can expect C<ev_timer>s to expire as the full suspension time
		2087	will be counted towards the timers. When no monotonic clock source is in
		2088	use, then libev will again assume a timejump and adjust accordingly.
		2089
		2090	It might be beneficial for this latter case to call C<ev_suspend>
		2091	and C<ev_resume> in code that handles C<SIGTSTP>, to at least get
		2092	deterministic behaviour in this case (you can do nothing against
		2093	C<SIGSTOP>).
		2094
		2095	=head3 Watcher-Specific Functions and Data Members
		2096
488	=over 4	2097	=over 4
489		2098
490	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)	2099	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)
491		2100
492	=item ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)	2101	=item ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)
493		2102
494	Configure the timer to trigger after C<after> seconds. If C<repeat> is	2103	Configure the timer to trigger after C<after> seconds. If C<repeat>
495	C<0.>, then it will automatically be stopped. If it is positive, then the	2104	is C<0.>, then it will automatically be stopped once the timeout is
496	timer will automatically be configured to trigger again C<repeat> seconds	2105	reached. If it is positive, then the timer will automatically be
497	later, again, and again, until stopped manually.	2106	configured to trigger again C<repeat> seconds later, again, and again,
		2107	until stopped manually.
498		2108
499	The timer itself will do a best-effort at avoiding drift, that is, if you	2109	The timer itself will do a best-effort at avoiding drift, that is, if
500	configure a timer to trigger every 10 seconds, then it will trigger at	2110	you configure a timer to trigger every 10 seconds, then it will normally
501	exactly 10 second intervals. If, however, your program cannot keep up with	2111	trigger at exactly 10 second intervals. If, however, your program cannot
502	the timer (because it takes longer than those 10 seconds to do stuff) the	2112	keep up with the timer (because it takes longer than those 10 seconds to
503	timer will not fire more than once per event loop iteration.	2113	do stuff) the timer will not fire more than once per event loop iteration.
504		2114
505	=item ev_timer_again (loop)	2115	=item ev_timer_again (loop, ev_timer *)
506		2116
507	This will act as if the timer timed out and restart it again if it is	2117	This will act as if the timer timed out, and restarts it again if it is
508	repeating. The exact semantics are:	2118	repeating. It basically works like calling C<ev_timer_stop>, updating the
		2119	timeout to the C<repeat> value and calling C<ev_timer_start>.
509		2120
		2121	The exact semantics are as in the following rules, all of which will be
		2122	applied to the watcher:
		2123
		2124	=over 4
		2125
		2126	=item If the timer is pending, the pending status is always cleared.
		2127
510	If the timer is started but nonrepeating, stop it.	2128	=item If the timer is started but non-repeating, stop it (as if it timed
		2129	out, without invoking it).
511		2130
512	If the timer is repeating, either start it if necessary (with the repeat	2131	=item If the timer is repeating, make the C<repeat> value the new timeout
513	value), or reset the running timer to the repeat value.	2132	and start the timer, if necessary.
514
515	This sounds a bit complicated, but here is a useful and typical
516	example: Imagine you have a tcp connection and you want a so-called idle
517	timeout, that is, you want to be called when there have been, say, 60
518	seconds of inactivity on the socket. The easiest way to do this is to
519	configure an C<ev_timer> with after=repeat=60 and calling ev_timer_again each
520	time you successfully read or write some data. If you go into an idle
521	state where you do not expect data to travel on the socket, you can stop
522	the timer, and again will automatically restart it if need be.
523		2133
524	=back	2134	=back
525		2135
		2136	This sounds a bit complicated, see L<Be smart about timeouts>, above, for a
		2137	usage example.
		2138
		2139	=item ev_tstamp ev_timer_remaining (loop, ev_timer *)
		2140
		2141	Returns the remaining time until a timer fires. If the timer is active,
		2142	then this time is relative to the current event loop time, otherwise it's
		2143	the timeout value currently configured.
		2144
		2145	That is, after an C<ev_timer_set (w, 5, 7)>, C<ev_timer_remaining> returns
		2146	C<5>. When the timer is started and one second passes, C<ev_timer_remaining>
		2147	will return C<4>. When the timer expires and is restarted, it will return
		2148	roughly C<7> (likely slightly less as callback invocation takes some time,
		2149	too), and so on.
		2150
		2151	=item ev_tstamp repeat [read-write]
		2152
		2153	The current C<repeat> value. Will be used each time the watcher times out
		2154	or C<ev_timer_again> is called, and determines the next timeout (if any),
		2155	which is also when any modifications are taken into account.
		2156
		2157	=back
		2158
		2159	=head3 Examples
		2160
		2161	Example: Create a timer that fires after 60 seconds.
		2162
		2163	static void
		2164	one_minute_cb (struct ev_loop *loop, ev_timer *w, int revents)
		2165	{
		2166	.. one minute over, w is actually stopped right here
		2167	}
		2168
		2169	ev_timer mytimer;
		2170	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);
		2171	ev_timer_start (loop, &mytimer);
		2172
		2173	Example: Create a timeout timer that times out after 10 seconds of
		2174	inactivity.
		2175
		2176	static void
		2177	timeout_cb (struct ev_loop *loop, ev_timer *w, int revents)
		2178	{
		2179	.. ten seconds without any activity
		2180	}
		2181
		2182	ev_timer mytimer;
		2183	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */
		2184	ev_timer_again (&mytimer); /* start timer */
		2185	ev_run (loop, 0);
		2186
		2187	// and in some piece of code that gets executed on any "activity":
		2188	// reset the timeout to start ticking again at 10 seconds
		2189	ev_timer_again (&mytimer);
		2190
		2191
526	=head2 C<ev_periodic> - to cron or not to cron	2192	=head2 C<ev_periodic> - to cron or not to cron?
527		2193
528	Periodic watchers are also timers of a kind, but they are very versatile	2194	Periodic watchers are also timers of a kind, but they are very versatile
529	(and unfortunately a bit complex).	2195	(and unfortunately a bit complex).
530		2196
531	Unlike C<ev_timer>'s, they are not based on real time (or relative time)	2197	Unlike C<ev_timer>, periodic watchers are not based on real time (or
532	but on wallclock time (absolute time). You can tell a periodic watcher	2198	relative time, the physical time that passes) but on wall clock time
533	to trigger "at" some specific point in time. For example, if you tell a	2199	(absolute time, the thing you can read on your calender or clock). The
534	periodic watcher to trigger in 10 seconds (by specifiying e.g. c<ev_now ()	2200	difference is that wall clock time can run faster or slower than real
535	+ 10.>) and then reset your system clock to the last year, then it will	2201	time, and time jumps are not uncommon (e.g. when you adjust your
536	take a year to trigger the event (unlike an C<ev_timer>, which would trigger	2202	wrist-watch).
537	roughly 10 seconds later and of course not if you reset your system time
538	again).
539		2203
540	They can also be used to implement vastly more complex timers, such as	2204	You can tell a periodic watcher to trigger after some specific point
		2205	in time: for example, if you tell a periodic watcher to trigger "in 10
		2206	seconds" (by specifying e.g. C<ev_now () + 10.>, that is, an absolute time
		2207	not a delay) and then reset your system clock to January of the previous
		2208	year, then it will take a year or more to trigger the event (unlike an
		2209	C<ev_timer>, which would still trigger roughly 10 seconds after starting
		2210	it, as it uses a relative timeout).
		2211
		2212	C<ev_periodic> watchers can also be used to implement vastly more complex
541	triggering an event on eahc midnight, local time.	2213	timers, such as triggering an event on each "midnight, local time", or
		2214	other complicated rules. This cannot be done with C<ev_timer> watchers, as
		2215	those cannot react to time jumps.
		2216
		2217	As with timers, the callback is guaranteed to be invoked only when the
		2218	point in time where it is supposed to trigger has passed. If multiple
		2219	timers become ready during the same loop iteration then the ones with
		2220	earlier time-out values are invoked before ones with later time-out values
		2221	(but this is no longer true when a callback calls C<ev_run> recursively).
		2222
		2223	=head3 Watcher-Specific Functions and Data Members
542		2224
543	=over 4	2225	=over 4
544		2226
545	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)	2227	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp offset, ev_tstamp interval, reschedule_cb)
546		2228
547	=item ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)	2229	=item ev_periodic_set (ev_periodic *, ev_tstamp offset, ev_tstamp interval, reschedule_cb)
548		2230
549	Lots of arguments, lets sort it out... There are basically three modes of	2231	Lots of arguments, let's sort it out... There are basically three modes of
550	operation, and we will explain them from simplest to complex:	2232	operation, and we will explain them from simplest to most complex:
551
552		2233
553	=over 4	2234	=over 4
554		2235
555	=item * absolute timer (interval = reschedule_cb = 0)	2236	=item * absolute timer (offset = absolute time, interval = 0, reschedule_cb = 0)
556		2237
557	In this configuration the watcher triggers an event at the wallclock time	2238	In this configuration the watcher triggers an event after the wall clock
558	C<at> and doesn't repeat. It will not adjust when a time jump occurs,	2239	time C<offset> has passed. It will not repeat and will not adjust when a
559	that is, if it is to be run at January 1st 2011 then it will run when the	2240	time jump occurs, that is, if it is to be run at January 1st 2011 then it
560	system time reaches or surpasses this time.	2241	will be stopped and invoked when the system clock reaches or surpasses
		2242	this point in time.
561		2243
562	=item * non-repeating interval timer (interval > 0, reschedule_cb = 0)	2244	=item * repeating interval timer (offset = offset within interval, interval > 0, reschedule_cb = 0)
563		2245
564	In this mode the watcher will always be scheduled to time out at the next	2246	In this mode the watcher will always be scheduled to time out at the next
565	C<at + N * interval> time (for some integer N) and then repeat, regardless	2247	C<offset + N * interval> time (for some integer N, which can also be
566	of any time jumps.	2248	negative) and then repeat, regardless of any time jumps. The C<offset>
		2249	argument is merely an offset into the C<interval> periods.
567		2250
568	This can be used to create timers that do not drift with respect to system	2251	This can be used to create timers that do not drift with respect to the
569	time:	2252	system clock, for example, here is an C<ev_periodic> that triggers each
		2253	hour, on the hour (with respect to UTC):
570		2254
571	ev_periodic_set (&periodic, 0., 3600., 0);	2255	ev_periodic_set (&periodic, 0., 3600., 0);
572		2256
573	This doesn't mean there will always be 3600 seconds in between triggers,	2257	This doesn't mean there will always be 3600 seconds in between triggers,
574	but only that the the callback will be called when the system time shows a	2258	but only that the callback will be called when the system time shows a
575	full hour (UTC), or more correctly, when the system time is evenly divisible	2259	full hour (UTC), or more correctly, when the system time is evenly divisible
576	by 3600.	2260	by 3600.
577		2261
578	Another way to think about it (for the mathematically inclined) is that	2262	Another way to think about it (for the mathematically inclined) is that
579	C<ev_periodic> will try to run the callback in this mode at the next possible	2263	C<ev_periodic> will try to run the callback in this mode at the next possible
580	time where C<time = at (mod interval)>, regardless of any time jumps.	2264	time where C<time = offset (mod interval)>, regardless of any time jumps.
581		2265
		2266	The C<interval> I<MUST> be positive, and for numerical stability, the
		2267	interval value should be higher than C<1/8192> (which is around 100
		2268	microseconds) and C<offset> should be higher than C<0> and should have
		2269	at most a similar magnitude as the current time (say, within a factor of
		2270	ten). Typical values for offset are, in fact, C<0> or something between
		2271	C<0> and C<interval>, which is also the recommended range.
		2272
		2273	Note also that there is an upper limit to how often a timer can fire (CPU
		2274	speed for example), so if C<interval> is very small then timing stability
		2275	will of course deteriorate. Libev itself tries to be exact to be about one
		2276	millisecond (if the OS supports it and the machine is fast enough).
		2277
582	=item * manual reschedule mode (reschedule_cb = callback)	2278	=item * manual reschedule mode (offset ignored, interval ignored, reschedule_cb = callback)
583		2279
584	In this mode the values for C<interval> and C<at> are both being	2280	In this mode the values for C<interval> and C<offset> are both being
585	ignored. Instead, each time the periodic watcher gets scheduled, the	2281	ignored. Instead, each time the periodic watcher gets scheduled, the
586	reschedule callback will be called with the watcher as first, and the	2282	reschedule callback will be called with the watcher as first, and the
587	current time as second argument.	2283	current time as second argument.
588		2284
589	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher,	2285	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher, ever,
590	ever, or make any event loop modifications>. If you need to stop it,	2286	or make ANY other event loop modifications whatsoever, unless explicitly
591	return C<now + 1e30> (or so, fudge fudge) and stop it afterwards (e.g. by	2287	allowed by documentation here>.
592	starting a prepare watcher).
593		2288
		2289	If you need to stop it, return C<now + 1e30> (or so, fudge fudge) and stop
		2290	it afterwards (e.g. by starting an C<ev_prepare> watcher, which is the
		2291	only event loop modification you are allowed to do).
		2292
594	Its prototype is C<ev_tstamp (*reschedule_cb)(struct ev_periodic *w,	2293	The callback prototype is C<ev_tstamp (*reschedule_cb)(ev_periodic
595	ev_tstamp now)>, e.g.:	2294	*w, ev_tstamp now)>, e.g.:
596		2295
		2296	static ev_tstamp
597	static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now)	2297	my_rescheduler (ev_periodic *w, ev_tstamp now)
598	{	2298	{
599	return now + 60.;	2299	return now + 60.;
600	}	2300	}
601		2301
602	It must return the next time to trigger, based on the passed time value	2302	It must return the next time to trigger, based on the passed time value
603	(that is, the lowest time value larger than to the second argument). It	2303	(that is, the lowest time value larger than to the second argument). It
604	will usually be called just before the callback will be triggered, but	2304	will usually be called just before the callback will be triggered, but
605	might be called at other times, too.	2305	might be called at other times, too.
606		2306
607	NOTE: I<< This callback must always return a time that is later than the	2307	NOTE: I<< This callback must always return a time that is higher than or
608	passed C<now> value >>. Not even C<now> itself will do, it I<must> be larger.	2308	equal to the passed C<now> value >>.
609		2309
610	This can be used to create very complex timers, such as a timer that	2310	This can be used to create very complex timers, such as a timer that
611	triggers on each midnight, local time. To do this, you would calculate the	2311	triggers on "next midnight, local time". To do this, you would calculate the
612	next midnight after C<now> and return the timestamp value for this. How	2312	next midnight after C<now> and return the timestamp value for this. How
613	you do this is, again, up to you (but it is not trivial, which is the main	2313	you do this is, again, up to you (but it is not trivial, which is the main
614	reason I omitted it as an example).	2314	reason I omitted it as an example).
615		2315
616	=back	2316	=back
…		…
620	Simply stops and restarts the periodic watcher again. This is only useful	2320	Simply stops and restarts the periodic watcher again. This is only useful
621	when you changed some parameters or the reschedule callback would return	2321	when you changed some parameters or the reschedule callback would return
622	a different time than the last time it was called (e.g. in a crond like	2322	a different time than the last time it was called (e.g. in a crond like
623	program when the crontabs have changed).	2323	program when the crontabs have changed).
624		2324
		2325	=item ev_tstamp ev_periodic_at (ev_periodic *)
		2326
		2327	When active, returns the absolute time that the watcher is supposed
		2328	to trigger next. This is not the same as the C<offset> argument to
		2329	C<ev_periodic_set>, but indeed works even in interval and manual
		2330	rescheduling modes.
		2331
		2332	=item ev_tstamp offset [read-write]
		2333
		2334	When repeating, this contains the offset value, otherwise this is the
		2335	absolute point in time (the C<offset> value passed to C<ev_periodic_set>,
		2336	although libev might modify this value for better numerical stability).
		2337
		2338	Can be modified any time, but changes only take effect when the periodic
		2339	timer fires or C<ev_periodic_again> is being called.
		2340
		2341	=item ev_tstamp interval [read-write]
		2342
		2343	The current interval value. Can be modified any time, but changes only
		2344	take effect when the periodic timer fires or C<ev_periodic_again> is being
		2345	called.
		2346
		2347	=item ev_tstamp (*reschedule_cb)(ev_periodic *w, ev_tstamp now) [read-write]
		2348
		2349	The current reschedule callback, or C<0>, if this functionality is
		2350	switched off. Can be changed any time, but changes only take effect when
		2351	the periodic timer fires or C<ev_periodic_again> is being called.
		2352
625	=back	2353	=back
626		2354
		2355	=head3 Examples
		2356
		2357	Example: Call a callback every hour, or, more precisely, whenever the
		2358	system time is divisible by 3600. The callback invocation times have
		2359	potentially a lot of jitter, but good long-term stability.
		2360
		2361	static void
		2362	clock_cb (struct ev_loop *loop, ev_periodic *w, int revents)
		2363	{
		2364	... its now a full hour (UTC, or TAI or whatever your clock follows)
		2365	}
		2366
		2367	ev_periodic hourly_tick;
		2368	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);
		2369	ev_periodic_start (loop, &hourly_tick);
		2370
		2371	Example: The same as above, but use a reschedule callback to do it:
		2372
		2373	#include <math.h>
		2374
		2375	static ev_tstamp
		2376	my_scheduler_cb (ev_periodic *w, ev_tstamp now)
		2377	{
		2378	return now + (3600. - fmod (now, 3600.));
		2379	}
		2380
		2381	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);
		2382
		2383	Example: Call a callback every hour, starting now:
		2384
		2385	ev_periodic hourly_tick;
		2386	ev_periodic_init (&hourly_tick, clock_cb,
		2387	fmod (ev_now (loop), 3600.), 3600., 0);
		2388	ev_periodic_start (loop, &hourly_tick);
		2389
		2390
627	=head2 C<ev_signal> - signal me when a signal gets signalled	2391	=head2 C<ev_signal> - signal me when a signal gets signalled!
628		2392
629	Signal watchers will trigger an event when the process receives a specific	2393	Signal watchers will trigger an event when the process receives a specific
630	signal one or more times. Even though signals are very asynchronous, libev	2394	signal one or more times. Even though signals are very asynchronous, libev
631	will try it's best to deliver signals synchronously, i.e. as part of the	2395	will try its best to deliver signals synchronously, i.e. as part of the
632	normal event processing, like any other event.	2396	normal event processing, like any other event.
633		2397
		2398	If you want signals to be delivered truly asynchronously, just use
		2399	C<sigaction> as you would do without libev and forget about sharing
		2400	the signal. You can even use C<ev_async> from a signal handler to
		2401	synchronously wake up an event loop.
		2402
634	You can configure as many watchers as you like per signal. Only when the	2403	You can configure as many watchers as you like for the same signal, but
		2404	only within the same loop, i.e. you can watch for C<SIGINT> in your
		2405	default loop and for C<SIGIO> in another loop, but you cannot watch for
		2406	C<SIGINT> in both the default loop and another loop at the same time. At
		2407	the moment, C<SIGCHLD> is permanently tied to the default loop.
		2408
635	first watcher gets started will libev actually register a signal watcher	2409	When the first watcher gets started will libev actually register something
636	with the kernel (thus it coexists with your own signal handlers as long	2410	with the kernel (thus it coexists with your own signal handlers as long as
637	as you don't register any with libev). Similarly, when the last signal	2411	you don't register any with libev for the same signal).
638	watcher for a signal is stopped libev will reset the signal handler to	2412
639	SIG_DFL (regardless of what it was set to before).	2413	If possible and supported, libev will install its handlers with
		2414	C<SA_RESTART> (or equivalent) behaviour enabled, so system calls should
		2415	not be unduly interrupted. If you have a problem with system calls getting
		2416	interrupted by signals you can block all signals in an C<ev_check> watcher
		2417	and unblock them in an C<ev_prepare> watcher.
		2418
		2419	=head3 The special problem of inheritance over fork/execve/pthread_create
		2420
		2421	Both the signal mask (C<sigprocmask>) and the signal disposition
		2422	(C<sigaction>) are unspecified after starting a signal watcher (and after
		2423	stopping it again), that is, libev might or might not block the signal,
		2424	and might or might not set or restore the installed signal handler (but
		2425	see C<EVFLAG_NOSIGMASK>).
		2426
		2427	While this does not matter for the signal disposition (libev never
		2428	sets signals to C<SIG_IGN>, so handlers will be reset to C<SIG_DFL> on
		2429	C<execve>), this matters for the signal mask: many programs do not expect
		2430	certain signals to be blocked.
		2431
		2432	This means that before calling C<exec> (from the child) you should reset
		2433	the signal mask to whatever "default" you expect (all clear is a good
		2434	choice usually).
		2435
		2436	The simplest way to ensure that the signal mask is reset in the child is
		2437	to install a fork handler with C<pthread_atfork> that resets it. That will
		2438	catch fork calls done by libraries (such as the libc) as well.
		2439
		2440	In current versions of libev, the signal will not be blocked indefinitely
		2441	unless you use the C<signalfd> API (C<EV_SIGNALFD>). While this reduces
		2442	the window of opportunity for problems, it will not go away, as libev
		2443	I<has> to modify the signal mask, at least temporarily.
		2444
		2445	So I can't stress this enough: I<If you do not reset your signal mask when
		2446	you expect it to be empty, you have a race condition in your code>. This
		2447	is not a libev-specific thing, this is true for most event libraries.
		2448
		2449	=head3 The special problem of threads signal handling
		2450
		2451	POSIX threads has problematic signal handling semantics, specifically,
		2452	a lot of functionality (sigfd, sigwait etc.) only really works if all
		2453	threads in a process block signals, which is hard to achieve.
		2454
		2455	When you want to use sigwait (or mix libev signal handling with your own
		2456	for the same signals), you can tackle this problem by globally blocking
		2457	all signals before creating any threads (or creating them with a fully set
		2458	sigprocmask) and also specifying the C<EVFLAG_NOSIGMASK> when creating
		2459	loops. Then designate one thread as "signal receiver thread" which handles
		2460	these signals. You can pass on any signals that libev might be interested
		2461	in by calling C<ev_feed_signal>.
		2462
		2463	=head3 Watcher-Specific Functions and Data Members
640		2464
641	=over 4	2465	=over 4
642		2466
643	=item ev_signal_init (ev_signal *, callback, int signum)	2467	=item ev_signal_init (ev_signal *, callback, int signum)
644		2468
645	=item ev_signal_set (ev_signal *, int signum)	2469	=item ev_signal_set (ev_signal *, int signum)
646		2470
647	Configures the watcher to trigger on the given signal number (usually one	2471	Configures the watcher to trigger on the given signal number (usually one
648	of the C<SIGxxx> constants).	2472	of the C<SIGxxx> constants).
649		2473
		2474	=item int signum [read-only]
		2475
		2476	The signal the watcher watches out for.
		2477
650	=back	2478	=back
651		2479
		2480	=head3 Examples
		2481
		2482	Example: Try to exit cleanly on SIGINT.
		2483
		2484	static void
		2485	sigint_cb (struct ev_loop *loop, ev_signal *w, int revents)
		2486	{
		2487	ev_break (loop, EVBREAK_ALL);
		2488	}
		2489
		2490	ev_signal signal_watcher;
		2491	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);
		2492	ev_signal_start (loop, &signal_watcher);
		2493
		2494
652	=head2 C<ev_child> - wait for pid status changes	2495	=head2 C<ev_child> - watch out for process status changes
653		2496
654	Child watchers trigger when your process receives a SIGCHLD in response to	2497	Child watchers trigger when your process receives a SIGCHLD in response to
655	some child status changes (most typically when a child of yours dies).	2498	some child status changes (most typically when a child of yours dies or
		2499	exits). It is permissible to install a child watcher I<after> the child
		2500	has been forked (which implies it might have already exited), as long
		2501	as the event loop isn't entered (or is continued from a watcher), i.e.,
		2502	forking and then immediately registering a watcher for the child is fine,
		2503	but forking and registering a watcher a few event loop iterations later or
		2504	in the next callback invocation is not.
		2505
		2506	Only the default event loop is capable of handling signals, and therefore
		2507	you can only register child watchers in the default event loop.
		2508
		2509	Due to some design glitches inside libev, child watchers will always be
		2510	handled at maximum priority (their priority is set to C<EV_MAXPRI> by
		2511	libev)
		2512
		2513	=head3 Process Interaction
		2514
		2515	Libev grabs C<SIGCHLD> as soon as the default event loop is
		2516	initialised. This is necessary to guarantee proper behaviour even if the
		2517	first child watcher is started after the child exits. The occurrence
		2518	of C<SIGCHLD> is recorded asynchronously, but child reaping is done
		2519	synchronously as part of the event loop processing. Libev always reaps all
		2520	children, even ones not watched.
		2521
		2522	=head3 Overriding the Built-In Processing
		2523
		2524	Libev offers no special support for overriding the built-in child
		2525	processing, but if your application collides with libev's default child
		2526	handler, you can override it easily by installing your own handler for
		2527	C<SIGCHLD> after initialising the default loop, and making sure the
		2528	default loop never gets destroyed. You are encouraged, however, to use an
		2529	event-based approach to child reaping and thus use libev's support for
		2530	that, so other libev users can use C<ev_child> watchers freely.
		2531
		2532	=head3 Stopping the Child Watcher
		2533
		2534	Currently, the child watcher never gets stopped, even when the
		2535	child terminates, so normally one needs to stop the watcher in the
		2536	callback. Future versions of libev might stop the watcher automatically
		2537	when a child exit is detected (calling C<ev_child_stop> twice is not a
		2538	problem).
		2539
		2540	=head3 Watcher-Specific Functions and Data Members
656		2541
657	=over 4	2542	=over 4
658		2543
659	=item ev_child_init (ev_child *, callback, int pid)	2544	=item ev_child_init (ev_child *, callback, int pid, int trace)
660		2545
661	=item ev_child_set (ev_child *, int pid)	2546	=item ev_child_set (ev_child *, int pid, int trace)
662		2547
663	Configures the watcher to wait for status changes of process C<pid> (or	2548	Configures the watcher to wait for status changes of process C<pid> (or
664	I<any> process if C<pid> is specified as C<0>). The callback can look	2549	I<any> process if C<pid> is specified as C<0>). The callback can look
665	at the C<rstatus> member of the C<ev_child> watcher structure to see	2550	at the C<rstatus> member of the C<ev_child> watcher structure to see
666	the status word (use the macros from C<sys/wait.h> and see your systems	2551	the status word (use the macros from C<sys/wait.h> and see your systems
667	C<waitpid> documentation). The C<rpid> member contains the pid of the	2552	C<waitpid> documentation). The C<rpid> member contains the pid of the
668	process causing the status change.	2553	process causing the status change. C<trace> must be either C<0> (only
		2554	activate the watcher when the process terminates) or C<1> (additionally
		2555	activate the watcher when the process is stopped or continued).
		2556
		2557	=item int pid [read-only]
		2558
		2559	The process id this watcher watches out for, or C<0>, meaning any process id.
		2560
		2561	=item int rpid [read-write]
		2562
		2563	The process id that detected a status change.
		2564
		2565	=item int rstatus [read-write]
		2566
		2567	The process exit/trace status caused by C<rpid> (see your systems
		2568	C<waitpid> and C<sys/wait.h> documentation for details).
669		2569
670	=back	2570	=back
671		2571
		2572	=head3 Examples
		2573
		2574	Example: C<fork()> a new process and install a child handler to wait for
		2575	its completion.
		2576
		2577	ev_child cw;
		2578
		2579	static void
		2580	child_cb (EV_P_ ev_child *w, int revents)
		2581	{
		2582	ev_child_stop (EV_A_ w);
		2583	printf ("process %d exited with status %x\n", w->rpid, w->rstatus);
		2584	}
		2585
		2586	pid_t pid = fork ();
		2587
		2588	if (pid < 0)
		2589	// error
		2590	else if (pid == 0)
		2591	{
		2592	// the forked child executes here
		2593	exit (1);
		2594	}
		2595	else
		2596	{
		2597	ev_child_init (&cw, child_cb, pid, 0);
		2598	ev_child_start (EV_DEFAULT_ &cw);
		2599	}
		2600
		2601
		2602	=head2 C<ev_stat> - did the file attributes just change?
		2603
		2604	This watches a file system path for attribute changes. That is, it calls
		2605	C<stat> on that path in regular intervals (or when the OS says it changed)
		2606	and sees if it changed compared to the last time, invoking the callback if
		2607	it did.
		2608
		2609	The path does not need to exist: changing from "path exists" to "path does
		2610	not exist" is a status change like any other. The condition "path does not
		2611	exist" (or more correctly "path cannot be stat'ed") is signified by the
		2612	C<st_nlink> field being zero (which is otherwise always forced to be at
		2613	least one) and all the other fields of the stat buffer having unspecified
		2614	contents.
		2615
		2616	The path I<must not> end in a slash or contain special components such as
		2617	C<.> or C<..>. The path I<should> be absolute: If it is relative and
		2618	your working directory changes, then the behaviour is undefined.
		2619
		2620	Since there is no portable change notification interface available, the
		2621	portable implementation simply calls C<stat(2)> regularly on the path
		2622	to see if it changed somehow. You can specify a recommended polling
		2623	interval for this case. If you specify a polling interval of C<0> (highly
		2624	recommended!) then a I<suitable, unspecified default> value will be used
		2625	(which you can expect to be around five seconds, although this might
		2626	change dynamically). Libev will also impose a minimum interval which is
		2627	currently around C<0.1>, but that's usually overkill.
		2628
		2629	This watcher type is not meant for massive numbers of stat watchers,
		2630	as even with OS-supported change notifications, this can be
		2631	resource-intensive.
		2632
		2633	At the time of this writing, the only OS-specific interface implemented
		2634	is the Linux inotify interface (implementing kqueue support is left as an
		2635	exercise for the reader. Note, however, that the author sees no way of
		2636	implementing C<ev_stat> semantics with kqueue, except as a hint).
		2637
		2638	=head3 ABI Issues (Largefile Support)
		2639
		2640	Libev by default (unless the user overrides this) uses the default
		2641	compilation environment, which means that on systems with large file
		2642	support disabled by default, you get the 32 bit version of the stat
		2643	structure. When using the library from programs that change the ABI to
		2644	use 64 bit file offsets the programs will fail. In that case you have to
		2645	compile libev with the same flags to get binary compatibility. This is
		2646	obviously the case with any flags that change the ABI, but the problem is
		2647	most noticeably displayed with ev_stat and large file support.
		2648
		2649	The solution for this is to lobby your distribution maker to make large
		2650	file interfaces available by default (as e.g. FreeBSD does) and not
		2651	optional. Libev cannot simply switch on large file support because it has
		2652	to exchange stat structures with application programs compiled using the
		2653	default compilation environment.
		2654
		2655	=head3 Inotify and Kqueue
		2656
		2657	When C<inotify (7)> support has been compiled into libev and present at
		2658	runtime, it will be used to speed up change detection where possible. The
		2659	inotify descriptor will be created lazily when the first C<ev_stat>
		2660	watcher is being started.
		2661
		2662	Inotify presence does not change the semantics of C<ev_stat> watchers
		2663	except that changes might be detected earlier, and in some cases, to avoid
		2664	making regular C<stat> calls. Even in the presence of inotify support
		2665	there are many cases where libev has to resort to regular C<stat> polling,
		2666	but as long as kernel 2.6.25 or newer is used (2.6.24 and older have too
		2667	many bugs), the path exists (i.e. stat succeeds), and the path resides on
		2668	a local filesystem (libev currently assumes only ext2/3, jfs, reiserfs and
		2669	xfs are fully working) libev usually gets away without polling.
		2670
		2671	There is no support for kqueue, as apparently it cannot be used to
		2672	implement this functionality, due to the requirement of having a file
		2673	descriptor open on the object at all times, and detecting renames, unlinks
		2674	etc. is difficult.
		2675
		2676	=head3 C<stat ()> is a synchronous operation
		2677
		2678	Libev doesn't normally do any kind of I/O itself, and so is not blocking
		2679	the process. The exception are C<ev_stat> watchers - those call C<stat
		2680	()>, which is a synchronous operation.
		2681
		2682	For local paths, this usually doesn't matter: unless the system is very
		2683	busy or the intervals between stat's are large, a stat call will be fast,
		2684	as the path data is usually in memory already (except when starting the
		2685	watcher).
		2686
		2687	For networked file systems, calling C<stat ()> can block an indefinite
		2688	time due to network issues, and even under good conditions, a stat call
		2689	often takes multiple milliseconds.
		2690
		2691	Therefore, it is best to avoid using C<ev_stat> watchers on networked
		2692	paths, although this is fully supported by libev.
		2693
		2694	=head3 The special problem of stat time resolution
		2695
		2696	The C<stat ()> system call only supports full-second resolution portably,
		2697	and even on systems where the resolution is higher, most file systems
		2698	still only support whole seconds.
		2699
		2700	That means that, if the time is the only thing that changes, you can
		2701	easily miss updates: on the first update, C<ev_stat> detects a change and
		2702	calls your callback, which does something. When there is another update
		2703	within the same second, C<ev_stat> will be unable to detect unless the
		2704	stat data does change in other ways (e.g. file size).
		2705
		2706	The solution to this is to delay acting on a change for slightly more
		2707	than a second (or till slightly after the next full second boundary), using
		2708	a roughly one-second-delay C<ev_timer> (e.g. C<ev_timer_set (w, 0., 1.02);
		2709	ev_timer_again (loop, w)>).
		2710
		2711	The C<.02> offset is added to work around small timing inconsistencies
		2712	of some operating systems (where the second counter of the current time
		2713	might be be delayed. One such system is the Linux kernel, where a call to
		2714	C<gettimeofday> might return a timestamp with a full second later than
		2715	a subsequent C<time> call - if the equivalent of C<time ()> is used to
		2716	update file times then there will be a small window where the kernel uses
		2717	the previous second to update file times but libev might already execute
		2718	the timer callback).
		2719
		2720	=head3 Watcher-Specific Functions and Data Members
		2721
		2722	=over 4
		2723
		2724	=item ev_stat_init (ev_stat *, callback, const char *path, ev_tstamp interval)
		2725
		2726	=item ev_stat_set (ev_stat *, const char *path, ev_tstamp interval)
		2727
		2728	Configures the watcher to wait for status changes of the given
		2729	C<path>. The C<interval> is a hint on how quickly a change is expected to
		2730	be detected and should normally be specified as C<0> to let libev choose
		2731	a suitable value. The memory pointed to by C<path> must point to the same
		2732	path for as long as the watcher is active.
		2733
		2734	The callback will receive an C<EV_STAT> event when a change was detected,
		2735	relative to the attributes at the time the watcher was started (or the
		2736	last change was detected).
		2737
		2738	=item ev_stat_stat (loop, ev_stat *)
		2739
		2740	Updates the stat buffer immediately with new values. If you change the
		2741	watched path in your callback, you could call this function to avoid
		2742	detecting this change (while introducing a race condition if you are not
		2743	the only one changing the path). Can also be useful simply to find out the
		2744	new values.
		2745
		2746	=item ev_statdata attr [read-only]
		2747
		2748	The most-recently detected attributes of the file. Although the type is
		2749	C<ev_statdata>, this is usually the (or one of the) C<struct stat> types
		2750	suitable for your system, but you can only rely on the POSIX-standardised
		2751	members to be present. If the C<st_nlink> member is C<0>, then there was
		2752	some error while C<stat>ing the file.
		2753
		2754	=item ev_statdata prev [read-only]
		2755
		2756	The previous attributes of the file. The callback gets invoked whenever
		2757	C<prev> != C<attr>, or, more precisely, one or more of these members
		2758	differ: C<st_dev>, C<st_ino>, C<st_mode>, C<st_nlink>, C<st_uid>,
		2759	C<st_gid>, C<st_rdev>, C<st_size>, C<st_atime>, C<st_mtime>, C<st_ctime>.
		2760
		2761	=item ev_tstamp interval [read-only]
		2762
		2763	The specified interval.
		2764
		2765	=item const char *path [read-only]
		2766
		2767	The file system path that is being watched.
		2768
		2769	=back
		2770
		2771	=head3 Examples
		2772
		2773	Example: Watch C</etc/passwd> for attribute changes.
		2774
		2775	static void
		2776	passwd_cb (struct ev_loop *loop, ev_stat *w, int revents)
		2777	{
		2778	/* /etc/passwd changed in some way */
		2779	if (w->attr.st_nlink)
		2780	{
		2781	printf ("passwd current size %ld\n", (long)w->attr.st_size);
		2782	printf ("passwd current atime %ld\n", (long)w->attr.st_mtime);
		2783	printf ("passwd current mtime %ld\n", (long)w->attr.st_mtime);
		2784	}
		2785	else
		2786	/* you shalt not abuse printf for puts */
		2787	puts ("wow, /etc/passwd is not there, expect problems. "
		2788	"if this is windows, they already arrived\n");
		2789	}
		2790
		2791	...
		2792	ev_stat passwd;
		2793
		2794	ev_stat_init (&passwd, passwd_cb, "/etc/passwd", 0.);
		2795	ev_stat_start (loop, &passwd);
		2796
		2797	Example: Like above, but additionally use a one-second delay so we do not
		2798	miss updates (however, frequent updates will delay processing, too, so
		2799	one might do the work both on C<ev_stat> callback invocation I<and> on
		2800	C<ev_timer> callback invocation).
		2801
		2802	static ev_stat passwd;
		2803	static ev_timer timer;
		2804
		2805	static void
		2806	timer_cb (EV_P_ ev_timer *w, int revents)
		2807	{
		2808	ev_timer_stop (EV_A_ w);
		2809
		2810	/* now it's one second after the most recent passwd change */
		2811	}
		2812
		2813	static void
		2814	stat_cb (EV_P_ ev_stat *w, int revents)
		2815	{
		2816	/* reset the one-second timer */
		2817	ev_timer_again (EV_A_ &timer);
		2818	}
		2819
		2820	...
		2821	ev_stat_init (&passwd, stat_cb, "/etc/passwd", 0.);
		2822	ev_stat_start (loop, &passwd);
		2823	ev_timer_init (&timer, timer_cb, 0., 1.02);
		2824
		2825
672	=head2 C<ev_idle> - when you've got nothing better to do	2826	=head2 C<ev_idle> - when you've got nothing better to do...
673		2827
674	Idle watchers trigger events when there are no other events are pending	2828	Idle watchers trigger events when no other events of the same or higher
675	(prepare, check and other idle watchers do not count). That is, as long	2829	priority are pending (prepare, check and other idle watchers do not count
676	as your process is busy handling sockets or timeouts (or even signals,	2830	as receiving "events").
677	imagine) it will not be triggered. But when your process is idle all idle	2831
678	watchers are being called again and again, once per event loop iteration -	2832	That is, as long as your process is busy handling sockets or timeouts
		2833	(or even signals, imagine) of the same or higher priority it will not be
		2834	triggered. But when your process is idle (or only lower-priority watchers
		2835	are pending), the idle watchers are being called once per event loop
679	until stopped, that is, or your process receives more events and becomes	2836	iteration - until stopped, that is, or your process receives more events
680	busy.	2837	and becomes busy again with higher priority stuff.
681		2838
682	The most noteworthy effect is that as long as any idle watchers are	2839	The most noteworthy effect is that as long as any idle watchers are
683	active, the process will not block when waiting for new events.	2840	active, the process will not block when waiting for new events.
684		2841
685	Apart from keeping your process non-blocking (which is a useful	2842	Apart from keeping your process non-blocking (which is a useful
686	effect on its own sometimes), idle watchers are a good place to do	2843	effect on its own sometimes), idle watchers are a good place to do
687	"pseudo-background processing", or delay processing stuff to after the	2844	"pseudo-background processing", or delay processing stuff to after the
688	event loop has handled all outstanding events.	2845	event loop has handled all outstanding events.
689		2846
		2847	=head3 Watcher-Specific Functions and Data Members
		2848
690	=over 4	2849	=over 4
691		2850
692	=item ev_idle_init (ev_signal *, callback)	2851	=item ev_idle_init (ev_idle *, callback)
693		2852
694	Initialises and configures the idle watcher - it has no parameters of any	2853	Initialises and configures the idle watcher - it has no parameters of any
695	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,	2854	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,
696	believe me.	2855	believe me.
697		2856
698	=back	2857	=back
699		2858
		2859	=head3 Examples
		2860
		2861	Example: Dynamically allocate an C<ev_idle> watcher, start it, and in the
		2862	callback, free it. Also, use no error checking, as usual.
		2863
		2864	static void
		2865	idle_cb (struct ev_loop *loop, ev_idle *w, int revents)
		2866	{
		2867	free (w);
		2868	// now do something you wanted to do when the program has
		2869	// no longer anything immediate to do.
		2870	}
		2871
		2872	ev_idle *idle_watcher = malloc (sizeof (ev_idle));
		2873	ev_idle_init (idle_watcher, idle_cb);
		2874	ev_idle_start (loop, idle_watcher);
		2875
		2876
700	=head2 C<ev_prepare> and C<ev_check> - customise your event loop	2877	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!
701		2878
702	Prepare and check watchers are usually (but not always) used in tandem:	2879	Prepare and check watchers are usually (but not always) used in pairs:
703	prepare watchers get invoked before the process blocks and check watchers	2880	prepare watchers get invoked before the process blocks and check watchers
704	afterwards.	2881	afterwards.
705		2882
		2883	You I<must not> call C<ev_run> or similar functions that enter
		2884	the current event loop from either C<ev_prepare> or C<ev_check>
		2885	watchers. Other loops than the current one are fine, however. The
		2886	rationale behind this is that you do not need to check for recursion in
		2887	those watchers, i.e. the sequence will always be C<ev_prepare>, blocking,
		2888	C<ev_check> so if you have one watcher of each kind they will always be
		2889	called in pairs bracketing the blocking call.
		2890
706	Their main purpose is to integrate other event mechanisms into libev. This	2891	Their main purpose is to integrate other event mechanisms into libev and
707	could be used, for example, to track variable changes, implement your own	2892	their use is somewhat advanced. They could be used, for example, to track
708	watchers, integrate net-snmp or a coroutine library and lots more.	2893	variable changes, implement your own watchers, integrate net-snmp or a
		2894	coroutine library and lots more. They are also occasionally useful if
		2895	you cache some data and want to flush it before blocking (for example,
		2896	in X programs you might want to do an C<XFlush ()> in an C<ev_prepare>
		2897	watcher).
709		2898
710	This is done by examining in each prepare call which file descriptors need	2899	This is done by examining in each prepare call which file descriptors
711	to be watched by the other library, registering C<ev_io> watchers for	2900	need to be watched by the other library, registering C<ev_io> watchers
712	them and starting an C<ev_timer> watcher for any timeouts (many libraries	2901	for them and starting an C<ev_timer> watcher for any timeouts (many
713	provide just this functionality). Then, in the check watcher you check for	2902	libraries provide exactly this functionality). Then, in the check watcher,
714	any events that occured (by checking the pending status of all watchers	2903	you check for any events that occurred (by checking the pending status
715	and stopping them) and call back into the library. The I/O and timer	2904	of all watchers and stopping them) and call back into the library. The
716	callbacks will never actually be called (but must be valid nevertheless,	2905	I/O and timer callbacks will never actually be called (but must be valid
717	because you never know, you know?).	2906	nevertheless, because you never know, you know?).
718		2907
719	As another example, the Perl Coro module uses these hooks to integrate	2908	As another example, the Perl Coro module uses these hooks to integrate
720	coroutines into libev programs, by yielding to other active coroutines	2909	coroutines into libev programs, by yielding to other active coroutines
721	during each prepare and only letting the process block if no coroutines	2910	during each prepare and only letting the process block if no coroutines
722	are ready to run (it's actually more complicated: it only runs coroutines	2911	are ready to run (it's actually more complicated: it only runs coroutines
723	with priority higher than or equal to the event loop and one coroutine	2912	with priority higher than or equal to the event loop and one coroutine
724	of lower priority, but only once, using idle watchers to keep the event	2913	of lower priority, but only once, using idle watchers to keep the event
725	loop from blocking if lower-priority coroutines are active, thus mapping	2914	loop from blocking if lower-priority coroutines are active, thus mapping
726	low-priority coroutines to idle/background tasks).	2915	low-priority coroutines to idle/background tasks).
727		2916
		2917	It is recommended to give C<ev_check> watchers highest (C<EV_MAXPRI>)
		2918	priority, to ensure that they are being run before any other watchers
		2919	after the poll (this doesn't matter for C<ev_prepare> watchers).
		2920
		2921	Also, C<ev_check> watchers (and C<ev_prepare> watchers, too) should not
		2922	activate ("feed") events into libev. While libev fully supports this, they
		2923	might get executed before other C<ev_check> watchers did their job. As
		2924	C<ev_check> watchers are often used to embed other (non-libev) event
		2925	loops those other event loops might be in an unusable state until their
		2926	C<ev_check> watcher ran (always remind yourself to coexist peacefully with
		2927	others).
		2928
		2929	=head3 Watcher-Specific Functions and Data Members
		2930
728	=over 4	2931	=over 4
729		2932
730	=item ev_prepare_init (ev_prepare *, callback)	2933	=item ev_prepare_init (ev_prepare *, callback)
731		2934
732	=item ev_check_init (ev_check *, callback)	2935	=item ev_check_init (ev_check *, callback)
733		2936
734	Initialises and configures the prepare or check watcher - they have no	2937	Initialises and configures the prepare or check watcher - they have no
735	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>	2938	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>
736	macros, but using them is utterly, utterly and completely pointless.	2939	macros, but using them is utterly, utterly, utterly and completely
		2940	pointless.
737		2941
738	=back	2942	=back
739		2943
		2944	=head3 Examples
		2945
		2946	There are a number of principal ways to embed other event loops or modules
		2947	into libev. Here are some ideas on how to include libadns into libev
		2948	(there is a Perl module named C<EV::ADNS> that does this, which you could
		2949	use as a working example. Another Perl module named C<EV::Glib> embeds a
		2950	Glib main context into libev, and finally, C<Glib::EV> embeds EV into the
		2951	Glib event loop).
		2952
		2953	Method 1: Add IO watchers and a timeout watcher in a prepare handler,
		2954	and in a check watcher, destroy them and call into libadns. What follows
		2955	is pseudo-code only of course. This requires you to either use a low
		2956	priority for the check watcher or use C<ev_clear_pending> explicitly, as
		2957	the callbacks for the IO/timeout watchers might not have been called yet.
		2958
		2959	static ev_io iow [nfd];
		2960	static ev_timer tw;
		2961
		2962	static void
		2963	io_cb (struct ev_loop *loop, ev_io *w, int revents)
		2964	{
		2965	}
		2966
		2967	// create io watchers for each fd and a timer before blocking
		2968	static void
		2969	adns_prepare_cb (struct ev_loop *loop, ev_prepare *w, int revents)
		2970	{
		2971	int timeout = 3600000;
		2972	struct pollfd fds [nfd];
		2973	// actual code will need to loop here and realloc etc.
		2974	adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));
		2975
		2976	/* the callback is illegal, but won't be called as we stop during check */
		2977	ev_timer_init (&tw, 0, timeout * 1e-3, 0.);
		2978	ev_timer_start (loop, &tw);
		2979
		2980	// create one ev_io per pollfd
		2981	for (int i = 0; i < nfd; ++i)
		2982	{
		2983	ev_io_init (iow + i, io_cb, fds [i].fd,
		2984	((fds [i].events & POLLIN ? EV_READ : 0)
		2985	| (fds [i].events & POLLOUT ? EV_WRITE : 0)));
		2986
		2987	fds [i].revents = 0;
		2988	ev_io_start (loop, iow + i);
		2989	}
		2990	}
		2991
		2992	// stop all watchers after blocking
		2993	static void
		2994	adns_check_cb (struct ev_loop *loop, ev_check *w, int revents)
		2995	{
		2996	ev_timer_stop (loop, &tw);
		2997
		2998	for (int i = 0; i < nfd; ++i)
		2999	{
		3000	// set the relevant poll flags
		3001	// could also call adns_processreadable etc. here
		3002	struct pollfd *fd = fds + i;
		3003	int revents = ev_clear_pending (iow + i);
		3004	if (revents & EV_READ ) fd->revents |= fd->events & POLLIN;
		3005	if (revents & EV_WRITE) fd->revents |= fd->events & POLLOUT;
		3006
		3007	// now stop the watcher
		3008	ev_io_stop (loop, iow + i);
		3009	}
		3010
		3011	adns_afterpoll (adns, fds, nfd, timeval_from (ev_now (loop));
		3012	}
		3013
		3014	Method 2: This would be just like method 1, but you run C<adns_afterpoll>
		3015	in the prepare watcher and would dispose of the check watcher.
		3016
		3017	Method 3: If the module to be embedded supports explicit event
		3018	notification (libadns does), you can also make use of the actual watcher
		3019	callbacks, and only destroy/create the watchers in the prepare watcher.
		3020
		3021	static void
		3022	timer_cb (EV_P_ ev_timer *w, int revents)
		3023	{
		3024	adns_state ads = (adns_state)w->data;
		3025	update_now (EV_A);
		3026
		3027	adns_processtimeouts (ads, &tv_now);
		3028	}
		3029
		3030	static void
		3031	io_cb (EV_P_ ev_io *w, int revents)
		3032	{
		3033	adns_state ads = (adns_state)w->data;
		3034	update_now (EV_A);
		3035
		3036	if (revents & EV_READ ) adns_processreadable (ads, w->fd, &tv_now);
		3037	if (revents & EV_WRITE) adns_processwriteable (ads, w->fd, &tv_now);
		3038	}
		3039
		3040	// do not ever call adns_afterpoll
		3041
		3042	Method 4: Do not use a prepare or check watcher because the module you
		3043	want to embed is not flexible enough to support it. Instead, you can
		3044	override their poll function. The drawback with this solution is that the
		3045	main loop is now no longer controllable by EV. The C<Glib::EV> module uses
		3046	this approach, effectively embedding EV as a client into the horrible
		3047	libglib event loop.
		3048
		3049	static gint
		3050	event_poll_func (GPollFD *fds, guint nfds, gint timeout)
		3051	{
		3052	int got_events = 0;
		3053
		3054	for (n = 0; n < nfds; ++n)
		3055	// create/start io watcher that sets the relevant bits in fds[n] and increment got_events
		3056
		3057	if (timeout >= 0)
		3058	// create/start timer
		3059
		3060	// poll
		3061	ev_run (EV_A_ 0);
		3062
		3063	// stop timer again
		3064	if (timeout >= 0)
		3065	ev_timer_stop (EV_A_ &to);
		3066
		3067	// stop io watchers again - their callbacks should have set
		3068	for (n = 0; n < nfds; ++n)
		3069	ev_io_stop (EV_A_ iow [n]);
		3070
		3071	return got_events;
		3072	}
		3073
		3074
		3075	=head2 C<ev_embed> - when one backend isn't enough...
		3076
		3077	This is a rather advanced watcher type that lets you embed one event loop
		3078	into another (currently only C<ev_io> events are supported in the embedded
		3079	loop, other types of watchers might be handled in a delayed or incorrect
		3080	fashion and must not be used).
		3081
		3082	There are primarily two reasons you would want that: work around bugs and
		3083	prioritise I/O.
		3084
		3085	As an example for a bug workaround, the kqueue backend might only support
		3086	sockets on some platform, so it is unusable as generic backend, but you
		3087	still want to make use of it because you have many sockets and it scales
		3088	so nicely. In this case, you would create a kqueue-based loop and embed
		3089	it into your default loop (which might use e.g. poll). Overall operation
		3090	will be a bit slower because first libev has to call C<poll> and then
		3091	C<kevent>, but at least you can use both mechanisms for what they are
		3092	best: C<kqueue> for scalable sockets and C<poll> if you want it to work :)
		3093
		3094	As for prioritising I/O: under rare circumstances you have the case where
		3095	some fds have to be watched and handled very quickly (with low latency),
		3096	and even priorities and idle watchers might have too much overhead. In
		3097	this case you would put all the high priority stuff in one loop and all
		3098	the rest in a second one, and embed the second one in the first.
		3099
		3100	As long as the watcher is active, the callback will be invoked every
		3101	time there might be events pending in the embedded loop. The callback
		3102	must then call C<ev_embed_sweep (mainloop, watcher)> to make a single
		3103	sweep and invoke their callbacks (the callback doesn't need to invoke the
		3104	C<ev_embed_sweep> function directly, it could also start an idle watcher
		3105	to give the embedded loop strictly lower priority for example).
		3106
		3107	You can also set the callback to C<0>, in which case the embed watcher
		3108	will automatically execute the embedded loop sweep whenever necessary.
		3109
		3110	Fork detection will be handled transparently while the C<ev_embed> watcher
		3111	is active, i.e., the embedded loop will automatically be forked when the
		3112	embedding loop forks. In other cases, the user is responsible for calling
		3113	C<ev_loop_fork> on the embedded loop.
		3114
		3115	Unfortunately, not all backends are embeddable: only the ones returned by
		3116	C<ev_embeddable_backends> are, which, unfortunately, does not include any
		3117	portable one.
		3118
		3119	So when you want to use this feature you will always have to be prepared
		3120	that you cannot get an embeddable loop. The recommended way to get around
		3121	this is to have a separate variables for your embeddable loop, try to
		3122	create it, and if that fails, use the normal loop for everything.
		3123
		3124	=head3 C<ev_embed> and fork
		3125
		3126	While the C<ev_embed> watcher is running, forks in the embedding loop will
		3127	automatically be applied to the embedded loop as well, so no special
		3128	fork handling is required in that case. When the watcher is not running,
		3129	however, it is still the task of the libev user to call C<ev_loop_fork ()>
		3130	as applicable.
		3131
		3132	=head3 Watcher-Specific Functions and Data Members
		3133
		3134	=over 4
		3135
		3136	=item ev_embed_init (ev_embed *, callback, struct ev_loop *embedded_loop)
		3137
		3138	=item ev_embed_set (ev_embed *, callback, struct ev_loop *embedded_loop)
		3139
		3140	Configures the watcher to embed the given loop, which must be
		3141	embeddable. If the callback is C<0>, then C<ev_embed_sweep> will be
		3142	invoked automatically, otherwise it is the responsibility of the callback
		3143	to invoke it (it will continue to be called until the sweep has been done,
		3144	if you do not want that, you need to temporarily stop the embed watcher).
		3145
		3146	=item ev_embed_sweep (loop, ev_embed *)
		3147
		3148	Make a single, non-blocking sweep over the embedded loop. This works
		3149	similarly to C<ev_run (embedded_loop, EVRUN_NOWAIT)>, but in the most
		3150	appropriate way for embedded loops.
		3151
		3152	=item struct ev_loop *other [read-only]
		3153
		3154	The embedded event loop.
		3155
		3156	=back
		3157
		3158	=head3 Examples
		3159
		3160	Example: Try to get an embeddable event loop and embed it into the default
		3161	event loop. If that is not possible, use the default loop. The default
		3162	loop is stored in C<loop_hi>, while the embeddable loop is stored in
		3163	C<loop_lo> (which is C<loop_hi> in the case no embeddable loop can be
		3164	used).
		3165
		3166	struct ev_loop *loop_hi = ev_default_init (0);
		3167	struct ev_loop *loop_lo = 0;
		3168	ev_embed embed;
		3169
		3170	// see if there is a chance of getting one that works
		3171	// (remember that a flags value of 0 means autodetection)
		3172	loop_lo = ev_embeddable_backends () & ev_recommended_backends ()
		3173	? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())
		3174	: 0;
		3175
		3176	// if we got one, then embed it, otherwise default to loop_hi
		3177	if (loop_lo)
		3178	{
		3179	ev_embed_init (&embed, 0, loop_lo);
		3180	ev_embed_start (loop_hi, &embed);
		3181	}
		3182	else
		3183	loop_lo = loop_hi;
		3184
		3185	Example: Check if kqueue is available but not recommended and create
		3186	a kqueue backend for use with sockets (which usually work with any
		3187	kqueue implementation). Store the kqueue/socket-only event loop in
		3188	C<loop_socket>. (One might optionally use C<EVFLAG_NOENV>, too).
		3189
		3190	struct ev_loop *loop = ev_default_init (0);
		3191	struct ev_loop *loop_socket = 0;
		3192	ev_embed embed;
		3193
		3194	if (ev_supported_backends () & ~ev_recommended_backends () & EVBACKEND_KQUEUE)
		3195	if ((loop_socket = ev_loop_new (EVBACKEND_KQUEUE))
		3196	{
		3197	ev_embed_init (&embed, 0, loop_socket);
		3198	ev_embed_start (loop, &embed);
		3199	}
		3200
		3201	if (!loop_socket)
		3202	loop_socket = loop;
		3203
		3204	// now use loop_socket for all sockets, and loop for everything else
		3205
		3206
		3207	=head2 C<ev_fork> - the audacity to resume the event loop after a fork
		3208
		3209	Fork watchers are called when a C<fork ()> was detected (usually because
		3210	whoever is a good citizen cared to tell libev about it by calling
		3211	C<ev_default_fork> or C<ev_loop_fork>). The invocation is done before the
		3212	event loop blocks next and before C<ev_check> watchers are being called,
		3213	and only in the child after the fork. If whoever good citizen calling
		3214	C<ev_default_fork> cheats and calls it in the wrong process, the fork
		3215	handlers will be invoked, too, of course.
		3216
		3217	=head3 The special problem of life after fork - how is it possible?
		3218
		3219	Most uses of C<fork()> consist of forking, then some simple calls to set
		3220	up/change the process environment, followed by a call to C<exec()>. This
		3221	sequence should be handled by libev without any problems.
		3222
		3223	This changes when the application actually wants to do event handling
		3224	in the child, or both parent in child, in effect "continuing" after the
		3225	fork.
		3226
		3227	The default mode of operation (for libev, with application help to detect
		3228	forks) is to duplicate all the state in the child, as would be expected
		3229	when I<either> the parent I<or> the child process continues.
		3230
		3231	When both processes want to continue using libev, then this is usually the
		3232	wrong result. In that case, usually one process (typically the parent) is
		3233	supposed to continue with all watchers in place as before, while the other
		3234	process typically wants to start fresh, i.e. without any active watchers.
		3235
		3236	The cleanest and most efficient way to achieve that with libev is to
		3237	simply create a new event loop, which of course will be "empty", and
		3238	use that for new watchers. This has the advantage of not touching more
		3239	memory than necessary, and thus avoiding the copy-on-write, and the
		3240	disadvantage of having to use multiple event loops (which do not support
		3241	signal watchers).
		3242
		3243	When this is not possible, or you want to use the default loop for
		3244	other reasons, then in the process that wants to start "fresh", call
		3245	C<ev_loop_destroy (EV_DEFAULT)> followed by C<ev_default_loop (...)>.
		3246	Destroying the default loop will "orphan" (not stop) all registered
		3247	watchers, so you have to be careful not to execute code that modifies
		3248	those watchers. Note also that in that case, you have to re-register any
		3249	signal watchers.
		3250
		3251	=head3 Watcher-Specific Functions and Data Members
		3252
		3253	=over 4
		3254
		3255	=item ev_fork_init (ev_fork *, callback)
		3256
		3257	Initialises and configures the fork watcher - it has no parameters of any
		3258	kind. There is a C<ev_fork_set> macro, but using it is utterly pointless,
		3259	really.
		3260
		3261	=back
		3262
		3263
		3264	=head2 C<ev_cleanup> - even the best things end
		3265
		3266	Cleanup watchers are called just before the event loop is being destroyed
		3267	by a call to C<ev_loop_destroy>.
		3268
		3269	While there is no guarantee that the event loop gets destroyed, cleanup
		3270	watchers provide a convenient method to install cleanup hooks for your
		3271	program, worker threads and so on - you just to make sure to destroy the
		3272	loop when you want them to be invoked.
		3273
		3274	Cleanup watchers are invoked in the same way as any other watcher. Unlike
		3275	all other watchers, they do not keep a reference to the event loop (which
		3276	makes a lot of sense if you think about it). Like all other watchers, you
		3277	can call libev functions in the callback, except C<ev_cleanup_start>.
		3278
		3279	=head3 Watcher-Specific Functions and Data Members
		3280
		3281	=over 4
		3282
		3283	=item ev_cleanup_init (ev_cleanup *, callback)
		3284
		3285	Initialises and configures the cleanup watcher - it has no parameters of
		3286	any kind. There is a C<ev_cleanup_set> macro, but using it is utterly
		3287	pointless, I assure you.
		3288
		3289	=back
		3290
		3291	Example: Register an atexit handler to destroy the default loop, so any
		3292	cleanup functions are called.
		3293
		3294	static void
		3295	program_exits (void)
		3296	{
		3297	ev_loop_destroy (EV_DEFAULT_UC);
		3298	}
		3299
		3300	...
		3301	atexit (program_exits);
		3302
		3303
		3304	=head2 C<ev_async> - how to wake up an event loop
		3305
		3306	In general, you cannot use an C<ev_loop> from multiple threads or other
		3307	asynchronous sources such as signal handlers (as opposed to multiple event
		3308	loops - those are of course safe to use in different threads).
		3309
		3310	Sometimes, however, you need to wake up an event loop you do not control,
		3311	for example because it belongs to another thread. This is what C<ev_async>
		3312	watchers do: as long as the C<ev_async> watcher is active, you can signal
		3313	it by calling C<ev_async_send>, which is thread- and signal safe.
		3314
		3315	This functionality is very similar to C<ev_signal> watchers, as signals,
		3316	too, are asynchronous in nature, and signals, too, will be compressed
		3317	(i.e. the number of callback invocations may be less than the number of
		3318	C<ev_async_send> calls). In fact, you could use signal watchers as a kind
		3319	of "global async watchers" by using a watcher on an otherwise unused
		3320	signal, and C<ev_feed_signal> to signal this watcher from another thread,
		3321	even without knowing which loop owns the signal.
		3322
		3323	=head3 Queueing
		3324
		3325	C<ev_async> does not support queueing of data in any way. The reason
		3326	is that the author does not know of a simple (or any) algorithm for a
		3327	multiple-writer-single-reader queue that works in all cases and doesn't
		3328	need elaborate support such as pthreads or unportable memory access
		3329	semantics.
		3330
		3331	That means that if you want to queue data, you have to provide your own
		3332	queue. But at least I can tell you how to implement locking around your
		3333	queue:
		3334
		3335	=over 4
		3336
		3337	=item queueing from a signal handler context
		3338
		3339	To implement race-free queueing, you simply add to the queue in the signal
		3340	handler but you block the signal handler in the watcher callback. Here is
		3341	an example that does that for some fictitious SIGUSR1 handler:
		3342
		3343	static ev_async mysig;
		3344
		3345	static void
		3346	sigusr1_handler (void)
		3347	{
		3348	sometype data;
		3349
		3350	// no locking etc.
		3351	queue_put (data);
		3352	ev_async_send (EV_DEFAULT_ &mysig);
		3353	}
		3354
		3355	static void
		3356	mysig_cb (EV_P_ ev_async *w, int revents)
		3357	{
		3358	sometype data;
		3359	sigset_t block, prev;
		3360
		3361	sigemptyset (&block);
		3362	sigaddset (&block, SIGUSR1);
		3363	sigprocmask (SIG_BLOCK, &block, &prev);
		3364
		3365	while (queue_get (&data))
		3366	process (data);
		3367
		3368	if (sigismember (&prev, SIGUSR1)
		3369	sigprocmask (SIG_UNBLOCK, &block, 0);
		3370	}
		3371
		3372	(Note: pthreads in theory requires you to use C<pthread_setmask>
		3373	instead of C<sigprocmask> when you use threads, but libev doesn't do it
		3374	either...).
		3375
		3376	=item queueing from a thread context
		3377
		3378	The strategy for threads is different, as you cannot (easily) block
		3379	threads but you can easily preempt them, so to queue safely you need to
		3380	employ a traditional mutex lock, such as in this pthread example:
		3381
		3382	static ev_async mysig;
		3383	static pthread_mutex_t mymutex = PTHREAD_MUTEX_INITIALIZER;
		3384
		3385	static void
		3386	otherthread (void)
		3387	{
		3388	// only need to lock the actual queueing operation
		3389	pthread_mutex_lock (&mymutex);
		3390	queue_put (data);
		3391	pthread_mutex_unlock (&mymutex);
		3392
		3393	ev_async_send (EV_DEFAULT_ &mysig);
		3394	}
		3395
		3396	static void
		3397	mysig_cb (EV_P_ ev_async *w, int revents)
		3398	{
		3399	pthread_mutex_lock (&mymutex);
		3400
		3401	while (queue_get (&data))
		3402	process (data);
		3403
		3404	pthread_mutex_unlock (&mymutex);
		3405	}
		3406
		3407	=back
		3408
		3409
		3410	=head3 Watcher-Specific Functions and Data Members
		3411
		3412	=over 4
		3413
		3414	=item ev_async_init (ev_async *, callback)
		3415
		3416	Initialises and configures the async watcher - it has no parameters of any
		3417	kind. There is a C<ev_async_set> macro, but using it is utterly pointless,
		3418	trust me.
		3419
		3420	=item ev_async_send (loop, ev_async *)
		3421
		3422	Sends/signals/activates the given C<ev_async> watcher, that is, feeds
		3423	an C<EV_ASYNC> event on the watcher into the event loop, and instantly
		3424	returns.
		3425
		3426	Unlike C<ev_feed_event>, this call is safe to do from other threads,
		3427	signal or similar contexts (see the discussion of C<EV_ATOMIC_T> in the
		3428	embedding section below on what exactly this means).
		3429
		3430	Note that, as with other watchers in libev, multiple events might get
		3431	compressed into a single callback invocation (another way to look at
		3432	this is that C<ev_async> watchers are level-triggered: they are set on
		3433	C<ev_async_send>, reset when the event loop detects that).
		3434
		3435	This call incurs the overhead of at most one extra system call per event
		3436	loop iteration, if the event loop is blocked, and no syscall at all if
		3437	the event loop (or your program) is processing events. That means that
		3438	repeated calls are basically free (there is no need to avoid calls for
		3439	performance reasons) and that the overhead becomes smaller (typically
		3440	zero) under load.
		3441
		3442	=item bool = ev_async_pending (ev_async *)
		3443
		3444	Returns a non-zero value when C<ev_async_send> has been called on the
		3445	watcher but the event has not yet been processed (or even noted) by the
		3446	event loop.
		3447
		3448	C<ev_async_send> sets a flag in the watcher and wakes up the loop. When
		3449	the loop iterates next and checks for the watcher to have become active,
		3450	it will reset the flag again. C<ev_async_pending> can be used to very
		3451	quickly check whether invoking the loop might be a good idea.
		3452
		3453	Not that this does I<not> check whether the watcher itself is pending,
		3454	only whether it has been requested to make this watcher pending: there
		3455	is a time window between the event loop checking and resetting the async
		3456	notification, and the callback being invoked.
		3457
		3458	=back
		3459
		3460
740	=head1 OTHER FUNCTIONS	3461	=head1 OTHER FUNCTIONS
741		3462
742	There are some other functions of possible interest. Described. Here. Now.	3463	There are some other functions of possible interest. Described. Here. Now.
743		3464
744	=over 4	3465	=over 4
745		3466
746	=item ev_once (loop, int fd, int events, ev_tstamp timeout, callback)	3467	=item ev_once (loop, int fd, int events, ev_tstamp timeout, callback)
747		3468
748	This function combines a simple timer and an I/O watcher, calls your	3469	This function combines a simple timer and an I/O watcher, calls your
749	callback on whichever event happens first and automatically stop both	3470	callback on whichever event happens first and automatically stops both
750	watchers. This is useful if you want to wait for a single event on an fd	3471	watchers. This is useful if you want to wait for a single event on an fd
751	or timeout without having to allocate/configure/start/stop/free one or	3472	or timeout without having to allocate/configure/start/stop/free one or
752	more watchers yourself.	3473	more watchers yourself.
753		3474
754	If C<fd> is less than 0, then no I/O watcher will be started and events	3475	If C<fd> is less than 0, then no I/O watcher will be started and the
755	is being ignored. Otherwise, an C<ev_io> watcher for the given C<fd> and	3476	C<events> argument is being ignored. Otherwise, an C<ev_io> watcher for
756	C<events> set will be craeted and started.	3477	the given C<fd> and C<events> set will be created and started.
757		3478
758	If C<timeout> is less than 0, then no timeout watcher will be	3479	If C<timeout> is less than 0, then no timeout watcher will be
759	started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and	3480	started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and
760	repeat = 0) will be started. While C<0> is a valid timeout, it is of	3481	repeat = 0) will be started. C<0> is a valid timeout.
761	dubious value.
762		3482
763	The callback has the type C<void (*cb)(int revents, void *arg)> and gets	3483	The callback has the type C<void (*cb)(int revents, void *arg)> and is
764	passed an C<revents> set like normal event callbacks (a combination of	3484	passed an C<revents> set like normal event callbacks (a combination of
765	C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMEOUT>) and the C<arg>	3485	C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMER>) and the C<arg>
766	value passed to C<ev_once>:	3486	value passed to C<ev_once>. Note that it is possible to receive I<both>
		3487	a timeout and an io event at the same time - you probably should give io
		3488	events precedence.
767		3489
		3490	Example: wait up to ten seconds for data to appear on STDIN_FILENO.
		3491
768	static void stdin_ready (int revents, void *arg)	3492	static void stdin_ready (int revents, void *arg)
		3493	{
		3494	if (revents & EV_READ)
		3495	/* stdin might have data for us, joy! */;
		3496	else if (revents & EV_TIMER)
		3497	/* doh, nothing entered */;
		3498	}
		3499
		3500	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);
		3501
		3502	=item ev_feed_fd_event (loop, int fd, int revents)
		3503
		3504	Feed an event on the given fd, as if a file descriptor backend detected
		3505	the given events.
		3506
		3507	=item ev_feed_signal_event (loop, int signum)
		3508
		3509	Feed an event as if the given signal occurred. See also C<ev_feed_signal>,
		3510	which is async-safe.
		3511
		3512	=back
		3513
		3514
		3515	=head1 COMMON OR USEFUL IDIOMS (OR BOTH)
		3516
		3517	This section explains some common idioms that are not immediately
		3518	obvious. Note that examples are sprinkled over the whole manual, and this
		3519	section only contains stuff that wouldn't fit anywhere else.
		3520
		3521	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER
		3522
		3523	Each watcher has, by default, a C<void *data> member that you can read
		3524	or modify at any time: libev will completely ignore it. This can be used
		3525	to associate arbitrary data with your watcher. If you need more data and
		3526	don't want to allocate memory separately and store a pointer to it in that
		3527	data member, you can also "subclass" the watcher type and provide your own
		3528	data:
		3529
		3530	struct my_io
		3531	{
		3532	ev_io io;
		3533	int otherfd;
		3534	void *somedata;
		3535	struct whatever *mostinteresting;
		3536	};
		3537
		3538	...
		3539	struct my_io w;
		3540	ev_io_init (&w.io, my_cb, fd, EV_READ);
		3541
		3542	And since your callback will be called with a pointer to the watcher, you
		3543	can cast it back to your own type:
		3544
		3545	static void my_cb (struct ev_loop *loop, ev_io *w_, int revents)
		3546	{
		3547	struct my_io *w = (struct my_io *)w_;
		3548	...
		3549	}
		3550
		3551	More interesting and less C-conformant ways of casting your callback
		3552	function type instead have been omitted.
		3553
		3554	=head2 BUILDING YOUR OWN COMPOSITE WATCHERS
		3555
		3556	Another common scenario is to use some data structure with multiple
		3557	embedded watchers, in effect creating your own watcher that combines
		3558	multiple libev event sources into one "super-watcher":
		3559
		3560	struct my_biggy
		3561	{
		3562	int some_data;
		3563	ev_timer t1;
		3564	ev_timer t2;
		3565	}
		3566
		3567	In this case getting the pointer to C<my_biggy> is a bit more
		3568	complicated: Either you store the address of your C<my_biggy> struct in
		3569	the C<data> member of the watcher (for woozies or C++ coders), or you need
		3570	to use some pointer arithmetic using C<offsetof> inside your watchers (for
		3571	real programmers):
		3572
		3573	#include <stddef.h>
		3574
		3575	static void
		3576	t1_cb (EV_P_ ev_timer *w, int revents)
		3577	{
		3578	struct my_biggy big = (struct my_biggy *)
		3579	(((char *)w) - offsetof (struct my_biggy, t1));
		3580	}
		3581
		3582	static void
		3583	t2_cb (EV_P_ ev_timer *w, int revents)
		3584	{
		3585	struct my_biggy big = (struct my_biggy *)
		3586	(((char *)w) - offsetof (struct my_biggy, t2));
		3587	}
		3588
		3589	=head2 AVOIDING FINISHING BEFORE RETURNING
		3590
		3591	Often you have structures like this in event-based programs:
		3592
		3593	callback ()
769	{	3594	{
770	if (revents & EV_TIMEOUT)	3595	free (request);
771	/* doh, nothing entered */;
772	else if (revents & EV_READ)
773	/* stdin might have data for us, joy! */;
774	}	3596	}
775		3597
776	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);	3598	request = start_new_request (..., callback);
777		3599
778	=item ev_feed_event (loop, watcher, int events)	3600	The intent is to start some "lengthy" operation. The C<request> could be
		3601	used to cancel the operation, or do other things with it.
779		3602
780	Feeds the given event set into the event loop, as if the specified event	3603	It's not uncommon to have code paths in C<start_new_request> that
781	had happened for the specified watcher (which must be a pointer to an	3604	immediately invoke the callback, for example, to report errors. Or you add
782	initialised but not necessarily started event watcher).	3605	some caching layer that finds that it can skip the lengthy aspects of the
		3606	operation and simply invoke the callback with the result.
783		3607
784	=item ev_feed_fd_event (loop, int fd, int revents)	3608	The problem here is that this will happen I<before> C<start_new_request>
		3609	has returned, so C<request> is not set.
785		3610
786	Feed an event on the given fd, as if a file descriptor backend detected	3611	Even if you pass the request by some safer means to the callback, you
787	the given events it.	3612	might want to do something to the request after starting it, such as
		3613	canceling it, which probably isn't working so well when the callback has
		3614	already been invoked.
788		3615
789	=item ev_feed_signal_event (loop, int signum)	3616	A common way around all these issues is to make sure that
		3617	C<start_new_request> I<always> returns before the callback is invoked. If
		3618	C<start_new_request> immediately knows the result, it can artificially
		3619	delay invoking the callback by e.g. using a C<prepare> or C<idle> watcher
		3620	for example, or more sneakily, by reusing an existing (stopped) watcher
		3621	and pushing it into the pending queue:
790		3622
791	Feed an event as if the given signal occured (loop must be the default loop!).	3623	ev_set_cb (watcher, callback);
		3624	ev_feed_event (EV_A_ watcher, 0);
792		3625
793	=back	3626	This way, C<start_new_request> can safely return before the callback is
		3627	invoked, while not delaying callback invocation too much.
		3628
		3629	=head2 MODEL/NESTED EVENT LOOP INVOCATIONS AND EXIT CONDITIONS
		3630
		3631	Often (especially in GUI toolkits) there are places where you have
		3632	I<modal> interaction, which is most easily implemented by recursively
		3633	invoking C<ev_run>.
		3634
		3635	This brings the problem of exiting - a callback might want to finish the
		3636	main C<ev_run> call, but not the nested one (e.g. user clicked "Quit", but
		3637	a modal "Are you sure?" dialog is still waiting), or just the nested one
		3638	and not the main one (e.g. user clocked "Ok" in a modal dialog), or some
		3639	other combination: In these cases, C<ev_break> will not work alone.
		3640
		3641	The solution is to maintain "break this loop" variable for each C<ev_run>
		3642	invocation, and use a loop around C<ev_run> until the condition is
		3643	triggered, using C<EVRUN_ONCE>:
		3644
		3645	// main loop
		3646	int exit_main_loop = 0;
		3647
		3648	while (!exit_main_loop)
		3649	ev_run (EV_DEFAULT_ EVRUN_ONCE);
		3650
		3651	// in a modal watcher
		3652	int exit_nested_loop = 0;
		3653
		3654	while (!exit_nested_loop)
		3655	ev_run (EV_A_ EVRUN_ONCE);
		3656
		3657	To exit from any of these loops, just set the corresponding exit variable:
		3658
		3659	// exit modal loop
		3660	exit_nested_loop = 1;
		3661
		3662	// exit main program, after modal loop is finished
		3663	exit_main_loop = 1;
		3664
		3665	// exit both
		3666	exit_main_loop = exit_nested_loop = 1;
		3667
		3668	=head2 THREAD LOCKING EXAMPLE
		3669
		3670	Here is a fictitious example of how to run an event loop in a different
		3671	thread from where callbacks are being invoked and watchers are
		3672	created/added/removed.
		3673
		3674	For a real-world example, see the C<EV::Loop::Async> perl module,
		3675	which uses exactly this technique (which is suited for many high-level
		3676	languages).
		3677
		3678	The example uses a pthread mutex to protect the loop data, a condition
		3679	variable to wait for callback invocations, an async watcher to notify the
		3680	event loop thread and an unspecified mechanism to wake up the main thread.
		3681
		3682	First, you need to associate some data with the event loop:
		3683
		3684	typedef struct {
		3685	mutex_t lock; /* global loop lock */
		3686	ev_async async_w;
		3687	thread_t tid;
		3688	cond_t invoke_cv;
		3689	} userdata;
		3690
		3691	void prepare_loop (EV_P)
		3692	{
		3693	// for simplicity, we use a static userdata struct.
		3694	static userdata u;
		3695
		3696	ev_async_init (&u->async_w, async_cb);
		3697	ev_async_start (EV_A_ &u->async_w);
		3698
		3699	pthread_mutex_init (&u->lock, 0);
		3700	pthread_cond_init (&u->invoke_cv, 0);
		3701
		3702	// now associate this with the loop
		3703	ev_set_userdata (EV_A_ u);
		3704	ev_set_invoke_pending_cb (EV_A_ l_invoke);
		3705	ev_set_loop_release_cb (EV_A_ l_release, l_acquire);
		3706
		3707	// then create the thread running ev_run
		3708	pthread_create (&u->tid, 0, l_run, EV_A);
		3709	}
		3710
		3711	The callback for the C<ev_async> watcher does nothing: the watcher is used
		3712	solely to wake up the event loop so it takes notice of any new watchers
		3713	that might have been added:
		3714
		3715	static void
		3716	async_cb (EV_P_ ev_async *w, int revents)
		3717	{
		3718	// just used for the side effects
		3719	}
		3720
		3721	The C<l_release> and C<l_acquire> callbacks simply unlock/lock the mutex
		3722	protecting the loop data, respectively.
		3723
		3724	static void
		3725	l_release (EV_P)
		3726	{
		3727	userdata *u = ev_userdata (EV_A);
		3728	pthread_mutex_unlock (&u->lock);
		3729	}
		3730
		3731	static void
		3732	l_acquire (EV_P)
		3733	{
		3734	userdata *u = ev_userdata (EV_A);
		3735	pthread_mutex_lock (&u->lock);
		3736	}
		3737
		3738	The event loop thread first acquires the mutex, and then jumps straight
		3739	into C<ev_run>:
		3740
		3741	void *
		3742	l_run (void *thr_arg)
		3743	{
		3744	struct ev_loop *loop = (struct ev_loop *)thr_arg;
		3745
		3746	l_acquire (EV_A);
		3747	pthread_setcanceltype (PTHREAD_CANCEL_ASYNCHRONOUS, 0);
		3748	ev_run (EV_A_ 0);
		3749	l_release (EV_A);
		3750
		3751	return 0;
		3752	}
		3753
		3754	Instead of invoking all pending watchers, the C<l_invoke> callback will
		3755	signal the main thread via some unspecified mechanism (signals? pipe
		3756	writes? C<Async::Interrupt>?) and then waits until all pending watchers
		3757	have been called (in a while loop because a) spurious wakeups are possible
		3758	and b) skipping inter-thread-communication when there are no pending
		3759	watchers is very beneficial):
		3760
		3761	static void
		3762	l_invoke (EV_P)
		3763	{
		3764	userdata *u = ev_userdata (EV_A);
		3765
		3766	while (ev_pending_count (EV_A))
		3767	{
		3768	wake_up_other_thread_in_some_magic_or_not_so_magic_way ();
		3769	pthread_cond_wait (&u->invoke_cv, &u->lock);
		3770	}
		3771	}
		3772
		3773	Now, whenever the main thread gets told to invoke pending watchers, it
		3774	will grab the lock, call C<ev_invoke_pending> and then signal the loop
		3775	thread to continue:
		3776
		3777	static void
		3778	real_invoke_pending (EV_P)
		3779	{
		3780	userdata *u = ev_userdata (EV_A);
		3781
		3782	pthread_mutex_lock (&u->lock);
		3783	ev_invoke_pending (EV_A);
		3784	pthread_cond_signal (&u->invoke_cv);
		3785	pthread_mutex_unlock (&u->lock);
		3786	}
		3787
		3788	Whenever you want to start/stop a watcher or do other modifications to an
		3789	event loop, you will now have to lock:
		3790
		3791	ev_timer timeout_watcher;
		3792	userdata *u = ev_userdata (EV_A);
		3793
		3794	ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);
		3795
		3796	pthread_mutex_lock (&u->lock);
		3797	ev_timer_start (EV_A_ &timeout_watcher);
		3798	ev_async_send (EV_A_ &u->async_w);
		3799	pthread_mutex_unlock (&u->lock);
		3800
		3801	Note that sending the C<ev_async> watcher is required because otherwise
		3802	an event loop currently blocking in the kernel will have no knowledge
		3803	about the newly added timer. By waking up the loop it will pick up any new
		3804	watchers in the next event loop iteration.
		3805
		3806	=head2 THREADS, COROUTINES, CONTINUATIONS, QUEUES... INSTEAD OF CALLBACKS
		3807
		3808	While the overhead of a callback that e.g. schedules a thread is small, it
		3809	is still an overhead. If you embed libev, and your main usage is with some
		3810	kind of threads or coroutines, you might want to customise libev so that
		3811	doesn't need callbacks anymore.
		3812
		3813	Imagine you have coroutines that you can switch to using a function
		3814	C<switch_to (coro)>, that libev runs in a coroutine called C<libev_coro>
		3815	and that due to some magic, the currently active coroutine is stored in a
		3816	global called C<current_coro>. Then you can build your own "wait for libev
		3817	event" primitive by changing C<EV_CB_DECLARE> and C<EV_CB_INVOKE> (note
		3818	the differing C<;> conventions):
		3819
		3820	#define EV_CB_DECLARE(type) struct my_coro *cb;
		3821	#define EV_CB_INVOKE(watcher) switch_to ((watcher)->cb)
		3822
		3823	That means instead of having a C callback function, you store the
		3824	coroutine to switch to in each watcher, and instead of having libev call
		3825	your callback, you instead have it switch to that coroutine.
		3826
		3827	A coroutine might now wait for an event with a function called
		3828	C<wait_for_event>. (the watcher needs to be started, as always, but it doesn't
		3829	matter when, or whether the watcher is active or not when this function is
		3830	called):
		3831
		3832	void
		3833	wait_for_event (ev_watcher *w)
		3834	{
		3835	ev_cb_set (w) = current_coro;
		3836	switch_to (libev_coro);
		3837	}
		3838
		3839	That basically suspends the coroutine inside C<wait_for_event> and
		3840	continues the libev coroutine, which, when appropriate, switches back to
		3841	this or any other coroutine.
		3842
		3843	You can do similar tricks if you have, say, threads with an event queue -
		3844	instead of storing a coroutine, you store the queue object and instead of
		3845	switching to a coroutine, you push the watcher onto the queue and notify
		3846	any waiters.
		3847
		3848	To embed libev, see L<EMBEDDING>, but in short, it's easiest to create two
		3849	files, F<my_ev.h> and F<my_ev.c> that include the respective libev files:
		3850
		3851	// my_ev.h
		3852	#define EV_CB_DECLARE(type) struct my_coro *cb;
		3853	#define EV_CB_INVOKE(watcher) switch_to ((watcher)->cb);
		3854	#include "../libev/ev.h"
		3855
		3856	// my_ev.c
		3857	#define EV_H "my_ev.h"
		3858	#include "../libev/ev.c"
		3859
		3860	And then use F<my_ev.h> when you would normally use F<ev.h>, and compile
		3861	F<my_ev.c> into your project. When properly specifying include paths, you
		3862	can even use F<ev.h> as header file name directly.
		3863
794		3864
795	=head1 LIBEVENT EMULATION	3865	=head1 LIBEVENT EMULATION
796		3866
797	Libev offers a compatibility emulation layer for libevent. It cannot	3867	Libev offers a compatibility emulation layer for libevent. It cannot
798	emulate the internals of libevent, so here are some usage hints:	3868	emulate the internals of libevent, so here are some usage hints:
799		3869
800	=over 4	3870	=over 4
		3871
		3872	=item * Only the libevent-1.4.1-beta API is being emulated.
		3873
		3874	This was the newest libevent version available when libev was implemented,
		3875	and is still mostly unchanged in 2010.
801		3876
802	=item * Use it by including <event.h>, as usual.	3877	=item * Use it by including <event.h>, as usual.
803		3878
804	=item * The following members are fully supported: ev_base, ev_callback,	3879	=item * The following members are fully supported: ev_base, ev_callback,
805	ev_arg, ev_fd, ev_res, ev_events.	3880	ev_arg, ev_fd, ev_res, ev_events.
…		…
810		3885
811	=item * Priorities are not currently supported. Initialising priorities	3886	=item * Priorities are not currently supported. Initialising priorities
812	will fail and all watchers will have the same priority, even though there	3887	will fail and all watchers will have the same priority, even though there
813	is an ev_pri field.	3888	is an ev_pri field.
814		3889
		3890	=item * In libevent, the last base created gets the signals, in libev, the
		3891	base that registered the signal gets the signals.
		3892
815	=item * Other members are not supported.	3893	=item * Other members are not supported.
816		3894
817	=item * The libev emulation is I<not> ABI compatible to libevent, you need	3895	=item * The libev emulation is I<not> ABI compatible to libevent, you need
818	to use the libev header file and library.	3896	to use the libev header file and library.
819		3897
820	=back	3898	=back
821		3899
822	=head1 C++ SUPPORT	3900	=head1 C++ SUPPORT
823		3901
824	TBD.	3902	=head2 C API
		3903
		3904	The normal C API should work fine when used from C++: both ev.h and the
		3905	libev sources can be compiled as C++. Therefore, code that uses the C API
		3906	will work fine.
		3907
		3908	Proper exception specifications might have to be added to callbacks passed
		3909	to libev: exceptions may be thrown only from watcher callbacks, all
		3910	other callbacks (allocator, syserr, loop acquire/release and periodioc
		3911	reschedule callbacks) must not throw exceptions, and might need a C<throw
		3912	()> specification. If you have code that needs to be compiled as both C
		3913	and C++ you can use the C<EV_THROW> macro for this:
		3914
		3915	static void
		3916	fatal_error (const char *msg) EV_THROW
		3917	{
		3918	perror (msg);
		3919	abort ();
		3920	}
		3921
		3922	...
		3923	ev_set_syserr_cb (fatal_error);
		3924
		3925	The only API functions that can currently throw exceptions are C<ev_run>,
		3926	C<ev_invoke>, C<ev_invoke_pending> and C<ev_loop_destroy> (the latter
		3927	because it runs cleanup watchers).
		3928
		3929	Throwing exceptions in watcher callbacks is only supported if libev itself
		3930	is compiled with a C++ compiler or your C and C++ environments allow
		3931	throwing exceptions through C libraries (most do).
		3932
		3933	=head2 C++ API
		3934
		3935	Libev comes with some simplistic wrapper classes for C++ that mainly allow
		3936	you to use some convenience methods to start/stop watchers and also change
		3937	the callback model to a model using method callbacks on objects.
		3938
		3939	To use it,
		3940
		3941	#include <ev++.h>
		3942
		3943	This automatically includes F<ev.h> and puts all of its definitions (many
		3944	of them macros) into the global namespace. All C++ specific things are
		3945	put into the C<ev> namespace. It should support all the same embedding
		3946	options as F<ev.h>, most notably C<EV_MULTIPLICITY>.
		3947
		3948	Care has been taken to keep the overhead low. The only data member the C++
		3949	classes add (compared to plain C-style watchers) is the event loop pointer
		3950	that the watcher is associated with (or no additional members at all if
		3951	you disable C<EV_MULTIPLICITY> when embedding libev).
		3952
		3953	Currently, functions, static and non-static member functions and classes
		3954	with C<operator ()> can be used as callbacks. Other types should be easy
		3955	to add as long as they only need one additional pointer for context. If
		3956	you need support for other types of functors please contact the author
		3957	(preferably after implementing it).
		3958
		3959	For all this to work, your C++ compiler either has to use the same calling
		3960	conventions as your C compiler (for static member functions), or you have
		3961	to embed libev and compile libev itself as C++.
		3962
		3963	Here is a list of things available in the C<ev> namespace:
		3964
		3965	=over 4
		3966
		3967	=item C<ev::READ>, C<ev::WRITE> etc.
		3968
		3969	These are just enum values with the same values as the C<EV_READ> etc.
		3970	macros from F<ev.h>.
		3971
		3972	=item C<ev::tstamp>, C<ev::now>
		3973
		3974	Aliases to the same types/functions as with the C<ev_> prefix.
		3975
		3976	=item C<ev::io>, C<ev::timer>, C<ev::periodic>, C<ev::idle>, C<ev::sig> etc.
		3977
		3978	For each C<ev_TYPE> watcher in F<ev.h> there is a corresponding class of
		3979	the same name in the C<ev> namespace, with the exception of C<ev_signal>
		3980	which is called C<ev::sig> to avoid clashes with the C<signal> macro
		3981	defined by many implementations.
		3982
		3983	All of those classes have these methods:
		3984
		3985	=over 4
		3986
		3987	=item ev::TYPE::TYPE ()
		3988
		3989	=item ev::TYPE::TYPE (loop)
		3990
		3991	=item ev::TYPE::~TYPE
		3992
		3993	The constructor (optionally) takes an event loop to associate the watcher
		3994	with. If it is omitted, it will use C<EV_DEFAULT>.
		3995
		3996	The constructor calls C<ev_init> for you, which means you have to call the
		3997	C<set> method before starting it.
		3998
		3999	It will not set a callback, however: You have to call the templated C<set>
		4000	method to set a callback before you can start the watcher.
		4001
		4002	(The reason why you have to use a method is a limitation in C++ which does
		4003	not allow explicit template arguments for constructors).
		4004
		4005	The destructor automatically stops the watcher if it is active.
		4006
		4007	=item w->set<class, &class::method> (object *)
		4008
		4009	This method sets the callback method to call. The method has to have a
		4010	signature of C<void (*)(ev_TYPE &, int)>, it receives the watcher as
		4011	first argument and the C<revents> as second. The object must be given as
		4012	parameter and is stored in the C<data> member of the watcher.
		4013
		4014	This method synthesizes efficient thunking code to call your method from
		4015	the C callback that libev requires. If your compiler can inline your
		4016	callback (i.e. it is visible to it at the place of the C<set> call and
		4017	your compiler is good :), then the method will be fully inlined into the
		4018	thunking function, making it as fast as a direct C callback.
		4019
		4020	Example: simple class declaration and watcher initialisation
		4021
		4022	struct myclass
		4023	{
		4024	void io_cb (ev::io &w, int revents) { }
		4025	}
		4026
		4027	myclass obj;
		4028	ev::io iow;
		4029	iow.set <myclass, &myclass::io_cb> (&obj);
		4030
		4031	=item w->set (object *)
		4032
		4033	This is a variation of a method callback - leaving out the method to call
		4034	will default the method to C<operator ()>, which makes it possible to use
		4035	functor objects without having to manually specify the C<operator ()> all
		4036	the time. Incidentally, you can then also leave out the template argument
		4037	list.
		4038
		4039	The C<operator ()> method prototype must be C<void operator ()(watcher &w,
		4040	int revents)>.
		4041
		4042	See the method-C<set> above for more details.
		4043
		4044	Example: use a functor object as callback.
		4045
		4046	struct myfunctor
		4047	{
		4048	void operator() (ev::io &w, int revents)
		4049	{
		4050	...
		4051	}
		4052	}
		4053
		4054	myfunctor f;
		4055
		4056	ev::io w;
		4057	w.set (&f);
		4058
		4059	=item w->set<function> (void *data = 0)
		4060
		4061	Also sets a callback, but uses a static method or plain function as
		4062	callback. The optional C<data> argument will be stored in the watcher's
		4063	C<data> member and is free for you to use.
		4064
		4065	The prototype of the C<function> must be C<void (*)(ev::TYPE &w, int)>.
		4066
		4067	See the method-C<set> above for more details.
		4068
		4069	Example: Use a plain function as callback.
		4070
		4071	static void io_cb (ev::io &w, int revents) { }
		4072	iow.set <io_cb> ();
		4073
		4074	=item w->set (loop)
		4075
		4076	Associates a different C<struct ev_loop> with this watcher. You can only
		4077	do this when the watcher is inactive (and not pending either).
		4078
		4079	=item w->set ([arguments])
		4080
		4081	Basically the same as C<ev_TYPE_set>, with the same arguments. Either this
		4082	method or a suitable start method must be called at least once. Unlike the
		4083	C counterpart, an active watcher gets automatically stopped and restarted
		4084	when reconfiguring it with this method.
		4085
		4086	=item w->start ()
		4087
		4088	Starts the watcher. Note that there is no C<loop> argument, as the
		4089	constructor already stores the event loop.
		4090
		4091	=item w->start ([arguments])
		4092
		4093	Instead of calling C<set> and C<start> methods separately, it is often
		4094	convenient to wrap them in one call. Uses the same type of arguments as
		4095	the configure C<set> method of the watcher.
		4096
		4097	=item w->stop ()
		4098
		4099	Stops the watcher if it is active. Again, no C<loop> argument.
		4100
		4101	=item w->again () (C<ev::timer>, C<ev::periodic> only)
		4102
		4103	For C<ev::timer> and C<ev::periodic>, this invokes the corresponding
		4104	C<ev_TYPE_again> function.
		4105
		4106	=item w->sweep () (C<ev::embed> only)
		4107
		4108	Invokes C<ev_embed_sweep>.
		4109
		4110	=item w->update () (C<ev::stat> only)
		4111
		4112	Invokes C<ev_stat_stat>.
		4113
		4114	=back
		4115
		4116	=back
		4117
		4118	Example: Define a class with two I/O and idle watchers, start the I/O
		4119	watchers in the constructor.
		4120
		4121	class myclass
		4122	{
		4123	ev::io io ; void io_cb (ev::io &w, int revents);
		4124	ev::io io2 ; void io2_cb (ev::io &w, int revents);
		4125	ev::idle idle; void idle_cb (ev::idle &w, int revents);
		4126
		4127	myclass (int fd)
		4128	{
		4129	io .set <myclass, &myclass::io_cb > (this);
		4130	io2 .set <myclass, &myclass::io2_cb > (this);
		4131	idle.set <myclass, &myclass::idle_cb> (this);
		4132
		4133	io.set (fd, ev::WRITE); // configure the watcher
		4134	io.start (); // start it whenever convenient
		4135
		4136	io2.start (fd, ev::READ); // set + start in one call
		4137	}
		4138	};
		4139
		4140
		4141	=head1 OTHER LANGUAGE BINDINGS
		4142
		4143	Libev does not offer other language bindings itself, but bindings for a
		4144	number of languages exist in the form of third-party packages. If you know
		4145	any interesting language binding in addition to the ones listed here, drop
		4146	me a note.
		4147
		4148	=over 4
		4149
		4150	=item Perl
		4151
		4152	The EV module implements the full libev API and is actually used to test
		4153	libev. EV is developed together with libev. Apart from the EV core module,
		4154	there are additional modules that implement libev-compatible interfaces
		4155	to C<libadns> (C<EV::ADNS>, but C<AnyEvent::DNS> is preferred nowadays),
		4156	C<Net::SNMP> (C<Net::SNMP::EV>) and the C<libglib> event core (C<Glib::EV>
		4157	and C<EV::Glib>).
		4158
		4159	It can be found and installed via CPAN, its homepage is at
		4160	L<http://software.schmorp.de/pkg/EV>.
		4161
		4162	=item Python
		4163
		4164	Python bindings can be found at L<http://code.google.com/p/pyev/>. It
		4165	seems to be quite complete and well-documented.
		4166
		4167	=item Ruby
		4168
		4169	Tony Arcieri has written a ruby extension that offers access to a subset
		4170	of the libev API and adds file handle abstractions, asynchronous DNS and
		4171	more on top of it. It can be found via gem servers. Its homepage is at
		4172	L<http://rev.rubyforge.org/>.
		4173
		4174	Roger Pack reports that using the link order C<-lws2_32 -lmsvcrt-ruby-190>
		4175	makes rev work even on mingw.
		4176
		4177	=item Haskell
		4178
		4179	A haskell binding to libev is available at
		4180	L<http://hackage.haskell.org/cgi-bin/hackage-scripts/package/hlibev>.
		4181
		4182	=item D
		4183
		4184	Leandro Lucarella has written a D language binding (F<ev.d>) for libev, to
		4185	be found at L<http://www.llucax.com.ar/proj/ev.d/index.html>.
		4186
		4187	=item Ocaml
		4188
		4189	Erkki Seppala has written Ocaml bindings for libev, to be found at
		4190	L<http://modeemi.cs.tut.fi/~flux/software/ocaml-ev/>.
		4191
		4192	=item Lua
		4193
		4194	Brian Maher has written a partial interface to libev for lua (at the
		4195	time of this writing, only C<ev_io> and C<ev_timer>), to be found at
		4196	L<http://github.com/brimworks/lua-ev>.
		4197
		4198	=back
		4199
		4200
		4201	=head1 MACRO MAGIC
		4202
		4203	Libev can be compiled with a variety of options, the most fundamental
		4204	of which is C<EV_MULTIPLICITY>. This option determines whether (most)
		4205	functions and callbacks have an initial C<struct ev_loop *> argument.
		4206
		4207	To make it easier to write programs that cope with either variant, the
		4208	following macros are defined:
		4209
		4210	=over 4
		4211
		4212	=item C<EV_A>, C<EV_A_>
		4213
		4214	This provides the loop I<argument> for functions, if one is required ("ev
		4215	loop argument"). The C<EV_A> form is used when this is the sole argument,
		4216	C<EV_A_> is used when other arguments are following. Example:
		4217
		4218	ev_unref (EV_A);
		4219	ev_timer_add (EV_A_ watcher);
		4220	ev_run (EV_A_ 0);
		4221
		4222	It assumes the variable C<loop> of type C<struct ev_loop *> is in scope,
		4223	which is often provided by the following macro.
		4224
		4225	=item C<EV_P>, C<EV_P_>
		4226
		4227	This provides the loop I<parameter> for functions, if one is required ("ev
		4228	loop parameter"). The C<EV_P> form is used when this is the sole parameter,
		4229	C<EV_P_> is used when other parameters are following. Example:
		4230
		4231	// this is how ev_unref is being declared
		4232	static void ev_unref (EV_P);
		4233
		4234	// this is how you can declare your typical callback
		4235	static void cb (EV_P_ ev_timer *w, int revents)
		4236
		4237	It declares a parameter C<loop> of type C<struct ev_loop *>, quite
		4238	suitable for use with C<EV_A>.
		4239
		4240	=item C<EV_DEFAULT>, C<EV_DEFAULT_>
		4241
		4242	Similar to the other two macros, this gives you the value of the default
		4243	loop, if multiple loops are supported ("ev loop default"). The default loop
		4244	will be initialised if it isn't already initialised.
		4245
		4246	For non-multiplicity builds, these macros do nothing, so you always have
		4247	to initialise the loop somewhere.
		4248
		4249	=item C<EV_DEFAULT_UC>, C<EV_DEFAULT_UC_>
		4250
		4251	Usage identical to C<EV_DEFAULT> and C<EV_DEFAULT_>, but requires that the
		4252	default loop has been initialised (C<UC> == unchecked). Their behaviour
		4253	is undefined when the default loop has not been initialised by a previous
		4254	execution of C<EV_DEFAULT>, C<EV_DEFAULT_> or C<ev_default_init (...)>.
		4255
		4256	It is often prudent to use C<EV_DEFAULT> when initialising the first
		4257	watcher in a function but use C<EV_DEFAULT_UC> afterwards.
		4258
		4259	=back
		4260
		4261	Example: Declare and initialise a check watcher, utilising the above
		4262	macros so it will work regardless of whether multiple loops are supported
		4263	or not.
		4264
		4265	static void
		4266	check_cb (EV_P_ ev_timer *w, int revents)
		4267	{
		4268	ev_check_stop (EV_A_ w);
		4269	}
		4270
		4271	ev_check check;
		4272	ev_check_init (&check, check_cb);
		4273	ev_check_start (EV_DEFAULT_ &check);
		4274	ev_run (EV_DEFAULT_ 0);
		4275
		4276	=head1 EMBEDDING
		4277
		4278	Libev can (and often is) directly embedded into host
		4279	applications. Examples of applications that embed it include the Deliantra
		4280	Game Server, the EV perl module, the GNU Virtual Private Ethernet (gvpe)
		4281	and rxvt-unicode.
		4282
		4283	The goal is to enable you to just copy the necessary files into your
		4284	source directory without having to change even a single line in them, so
		4285	you can easily upgrade by simply copying (or having a checked-out copy of
		4286	libev somewhere in your source tree).
		4287
		4288	=head2 FILESETS
		4289
		4290	Depending on what features you need you need to include one or more sets of files
		4291	in your application.
		4292
		4293	=head3 CORE EVENT LOOP
		4294
		4295	To include only the libev core (all the C<ev_*> functions), with manual
		4296	configuration (no autoconf):
		4297
		4298	#define EV_STANDALONE 1
		4299	#include "ev.c"
		4300
		4301	This will automatically include F<ev.h>, too, and should be done in a
		4302	single C source file only to provide the function implementations. To use
		4303	it, do the same for F<ev.h> in all files wishing to use this API (best
		4304	done by writing a wrapper around F<ev.h> that you can include instead and
		4305	where you can put other configuration options):
		4306
		4307	#define EV_STANDALONE 1
		4308	#include "ev.h"
		4309
		4310	Both header files and implementation files can be compiled with a C++
		4311	compiler (at least, that's a stated goal, and breakage will be treated
		4312	as a bug).
		4313
		4314	You need the following files in your source tree, or in a directory
		4315	in your include path (e.g. in libev/ when using -Ilibev):
		4316
		4317	ev.h
		4318	ev.c
		4319	ev_vars.h
		4320	ev_wrap.h
		4321
		4322	ev_win32.c required on win32 platforms only
		4323
		4324	ev_select.c only when select backend is enabled (which is enabled by default)
		4325	ev_poll.c only when poll backend is enabled (disabled by default)
		4326	ev_epoll.c only when the epoll backend is enabled (disabled by default)
		4327	ev_kqueue.c only when the kqueue backend is enabled (disabled by default)
		4328	ev_port.c only when the solaris port backend is enabled (disabled by default)
		4329
		4330	F<ev.c> includes the backend files directly when enabled, so you only need
		4331	to compile this single file.
		4332
		4333	=head3 LIBEVENT COMPATIBILITY API
		4334
		4335	To include the libevent compatibility API, also include:
		4336
		4337	#include "event.c"
		4338
		4339	in the file including F<ev.c>, and:
		4340
		4341	#include "event.h"
		4342
		4343	in the files that want to use the libevent API. This also includes F<ev.h>.
		4344
		4345	You need the following additional files for this:
		4346
		4347	event.h
		4348	event.c
		4349
		4350	=head3 AUTOCONF SUPPORT
		4351
		4352	Instead of using C<EV_STANDALONE=1> and providing your configuration in
		4353	whatever way you want, you can also C<m4_include([libev.m4])> in your
		4354	F<configure.ac> and leave C<EV_STANDALONE> undefined. F<ev.c> will then
		4355	include F<config.h> and configure itself accordingly.
		4356
		4357	For this of course you need the m4 file:
		4358
		4359	libev.m4
		4360
		4361	=head2 PREPROCESSOR SYMBOLS/MACROS
		4362
		4363	Libev can be configured via a variety of preprocessor symbols you have to
		4364	define before including (or compiling) any of its files. The default in
		4365	the absence of autoconf is documented for every option.
		4366
		4367	Symbols marked with "(h)" do not change the ABI, and can have different
		4368	values when compiling libev vs. including F<ev.h>, so it is permissible
		4369	to redefine them before including F<ev.h> without breaking compatibility
		4370	to a compiled library. All other symbols change the ABI, which means all
		4371	users of libev and the libev code itself must be compiled with compatible
		4372	settings.
		4373
		4374	=over 4
		4375
		4376	=item EV_COMPAT3 (h)
		4377
		4378	Backwards compatibility is a major concern for libev. This is why this
		4379	release of libev comes with wrappers for the functions and symbols that
		4380	have been renamed between libev version 3 and 4.
		4381
		4382	You can disable these wrappers (to test compatibility with future
		4383	versions) by defining C<EV_COMPAT3> to C<0> when compiling your
		4384	sources. This has the additional advantage that you can drop the C<struct>
		4385	from C<struct ev_loop> declarations, as libev will provide an C<ev_loop>
		4386	typedef in that case.
		4387
		4388	In some future version, the default for C<EV_COMPAT3> will become C<0>,
		4389	and in some even more future version the compatibility code will be
		4390	removed completely.
		4391
		4392	=item EV_STANDALONE (h)
		4393
		4394	Must always be C<1> if you do not use autoconf configuration, which
		4395	keeps libev from including F<config.h>, and it also defines dummy
		4396	implementations for some libevent functions (such as logging, which is not
		4397	supported). It will also not define any of the structs usually found in
		4398	F<event.h> that are not directly supported by the libev core alone.
		4399
		4400	In standalone mode, libev will still try to automatically deduce the
		4401	configuration, but has to be more conservative.
		4402
		4403	=item EV_USE_FLOOR
		4404
		4405	If defined to be C<1>, libev will use the C<floor ()> function for its
		4406	periodic reschedule calculations, otherwise libev will fall back on a
		4407	portable (slower) implementation. If you enable this, you usually have to
		4408	link against libm or something equivalent. Enabling this when the C<floor>
		4409	function is not available will fail, so the safe default is to not enable
		4410	this.
		4411
		4412	=item EV_USE_MONOTONIC
		4413
		4414	If defined to be C<1>, libev will try to detect the availability of the
		4415	monotonic clock option at both compile time and runtime. Otherwise no
		4416	use of the monotonic clock option will be attempted. If you enable this,
		4417	you usually have to link against librt or something similar. Enabling it
		4418	when the functionality isn't available is safe, though, although you have
		4419	to make sure you link against any libraries where the C<clock_gettime>
		4420	function is hiding in (often F<-lrt>). See also C<EV_USE_CLOCK_SYSCALL>.
		4421
		4422	=item EV_USE_REALTIME
		4423
		4424	If defined to be C<1>, libev will try to detect the availability of the
		4425	real-time clock option at compile time (and assume its availability
		4426	at runtime if successful). Otherwise no use of the real-time clock
		4427	option will be attempted. This effectively replaces C<gettimeofday>
		4428	by C<clock_get (CLOCK_REALTIME, ...)> and will not normally affect
		4429	correctness. See the note about libraries in the description of
		4430	C<EV_USE_MONOTONIC>, though. Defaults to the opposite value of
		4431	C<EV_USE_CLOCK_SYSCALL>.
		4432
		4433	=item EV_USE_CLOCK_SYSCALL
		4434
		4435	If defined to be C<1>, libev will try to use a direct syscall instead
		4436	of calling the system-provided C<clock_gettime> function. This option
		4437	exists because on GNU/Linux, C<clock_gettime> is in C<librt>, but C<librt>
		4438	unconditionally pulls in C<libpthread>, slowing down single-threaded
		4439	programs needlessly. Using a direct syscall is slightly slower (in
		4440	theory), because no optimised vdso implementation can be used, but avoids
		4441	the pthread dependency. Defaults to C<1> on GNU/Linux with glibc 2.x or
		4442	higher, as it simplifies linking (no need for C<-lrt>).
		4443
		4444	=item EV_USE_NANOSLEEP
		4445
		4446	If defined to be C<1>, libev will assume that C<nanosleep ()> is available
		4447	and will use it for delays. Otherwise it will use C<select ()>.
		4448
		4449	=item EV_USE_EVENTFD
		4450
		4451	If defined to be C<1>, then libev will assume that C<eventfd ()> is
		4452	available and will probe for kernel support at runtime. This will improve
		4453	C<ev_signal> and C<ev_async> performance and reduce resource consumption.
		4454	If undefined, it will be enabled if the headers indicate GNU/Linux + Glibc
		4455	2.7 or newer, otherwise disabled.
		4456
		4457	=item EV_USE_SELECT
		4458
		4459	If undefined or defined to be C<1>, libev will compile in support for the
		4460	C<select>(2) backend. No attempt at auto-detection will be done: if no
		4461	other method takes over, select will be it. Otherwise the select backend
		4462	will not be compiled in.
		4463
		4464	=item EV_SELECT_USE_FD_SET
		4465
		4466	If defined to C<1>, then the select backend will use the system C<fd_set>
		4467	structure. This is useful if libev doesn't compile due to a missing
		4468	C<NFDBITS> or C<fd_mask> definition or it mis-guesses the bitset layout
		4469	on exotic systems. This usually limits the range of file descriptors to
		4470	some low limit such as 1024 or might have other limitations (winsocket
		4471	only allows 64 sockets). The C<FD_SETSIZE> macro, set before compilation,
		4472	configures the maximum size of the C<fd_set>.
		4473
		4474	=item EV_SELECT_IS_WINSOCKET
		4475
		4476	When defined to C<1>, the select backend will assume that
		4477	select/socket/connect etc. don't understand file descriptors but
		4478	wants osf handles on win32 (this is the case when the select to
		4479	be used is the winsock select). This means that it will call
		4480	C<_get_osfhandle> on the fd to convert it to an OS handle. Otherwise,
		4481	it is assumed that all these functions actually work on fds, even
		4482	on win32. Should not be defined on non-win32 platforms.
		4483
		4484	=item EV_FD_TO_WIN32_HANDLE(fd)
		4485
		4486	If C<EV_SELECT_IS_WINSOCKET> is enabled, then libev needs a way to map
		4487	file descriptors to socket handles. When not defining this symbol (the
		4488	default), then libev will call C<_get_osfhandle>, which is usually
		4489	correct. In some cases, programs use their own file descriptor management,
		4490	in which case they can provide this function to map fds to socket handles.
		4491
		4492	=item EV_WIN32_HANDLE_TO_FD(handle)
		4493
		4494	If C<EV_SELECT_IS_WINSOCKET> then libev maps handles to file descriptors
		4495	using the standard C<_open_osfhandle> function. For programs implementing
		4496	their own fd to handle mapping, overwriting this function makes it easier
		4497	to do so. This can be done by defining this macro to an appropriate value.
		4498
		4499	=item EV_WIN32_CLOSE_FD(fd)
		4500
		4501	If programs implement their own fd to handle mapping on win32, then this
		4502	macro can be used to override the C<close> function, useful to unregister
		4503	file descriptors again. Note that the replacement function has to close
		4504	the underlying OS handle.
		4505
		4506	=item EV_USE_POLL
		4507
		4508	If defined to be C<1>, libev will compile in support for the C<poll>(2)
		4509	backend. Otherwise it will be enabled on non-win32 platforms. It
		4510	takes precedence over select.
		4511
		4512	=item EV_USE_EPOLL
		4513
		4514	If defined to be C<1>, libev will compile in support for the Linux
		4515	C<epoll>(7) backend. Its availability will be detected at runtime,
		4516	otherwise another method will be used as fallback. This is the preferred
		4517	backend for GNU/Linux systems. If undefined, it will be enabled if the
		4518	headers indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.
		4519
		4520	=item EV_USE_KQUEUE
		4521
		4522	If defined to be C<1>, libev will compile in support for the BSD style
		4523	C<kqueue>(2) backend. Its actual availability will be detected at runtime,
		4524	otherwise another method will be used as fallback. This is the preferred
		4525	backend for BSD and BSD-like systems, although on most BSDs kqueue only
		4526	supports some types of fds correctly (the only platform we found that
		4527	supports ptys for example was NetBSD), so kqueue might be compiled in, but
		4528	not be used unless explicitly requested. The best way to use it is to find
		4529	out whether kqueue supports your type of fd properly and use an embedded
		4530	kqueue loop.
		4531
		4532	=item EV_USE_PORT
		4533
		4534	If defined to be C<1>, libev will compile in support for the Solaris
		4535	10 port style backend. Its availability will be detected at runtime,
		4536	otherwise another method will be used as fallback. This is the preferred
		4537	backend for Solaris 10 systems.
		4538
		4539	=item EV_USE_DEVPOLL
		4540
		4541	Reserved for future expansion, works like the USE symbols above.
		4542
		4543	=item EV_USE_INOTIFY
		4544
		4545	If defined to be C<1>, libev will compile in support for the Linux inotify
		4546	interface to speed up C<ev_stat> watchers. Its actual availability will
		4547	be detected at runtime. If undefined, it will be enabled if the headers
		4548	indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.
		4549
		4550	=item EV_NO_SMP
		4551
		4552	If defined to be C<1>, libev will assume that memory is always coherent
		4553	between threads, that is, threads can be used, but threads never run on
		4554	different cpus (or different cpu cores). This reduces dependencies
		4555	and makes libev faster.
		4556
		4557	=item EV_NO_THREADS
		4558
		4559	If defined to be C<1>, libev will assume that it will never be called
		4560	from different threads, which is a stronger assumption than C<EV_NO_SMP>,
		4561	above. This reduces dependencies and makes libev faster.
		4562
		4563	=item EV_ATOMIC_T
		4564
		4565	Libev requires an integer type (suitable for storing C<0> or C<1>) whose
		4566	access is atomic and serialised with respect to other threads or signal
		4567	contexts. No such type is easily found in the C language, so you can
		4568	provide your own type that you know is safe for your purposes. It is used
		4569	both for signal handler "locking" as well as for signal and thread safety
		4570	in C<ev_async> watchers.
		4571
		4572	In the absence of this define, libev will use C<sig_atomic_t volatile>
		4573	(from F<signal.h>), which is usually good enough on most platforms,
		4574	although strictly speaking using a type that also implies a memory fence
		4575	is required.
		4576
		4577	=item EV_H (h)
		4578
		4579	The name of the F<ev.h> header file used to include it. The default if
		4580	undefined is C<"ev.h"> in F<event.h>, F<ev.c> and F<ev++.h>. This can be
		4581	used to virtually rename the F<ev.h> header file in case of conflicts.
		4582
		4583	=item EV_CONFIG_H (h)
		4584
		4585	If C<EV_STANDALONE> isn't C<1>, this variable can be used to override
		4586	F<ev.c>'s idea of where to find the F<config.h> file, similarly to
		4587	C<EV_H>, above.
		4588
		4589	=item EV_EVENT_H (h)
		4590
		4591	Similarly to C<EV_H>, this macro can be used to override F<event.c>'s idea
		4592	of how the F<event.h> header can be found, the default is C<"event.h">.
		4593
		4594	=item EV_PROTOTYPES (h)
		4595
		4596	If defined to be C<0>, then F<ev.h> will not define any function
		4597	prototypes, but still define all the structs and other symbols. This is
		4598	occasionally useful if you want to provide your own wrapper functions
		4599	around libev functions.
		4600
		4601	=item EV_MULTIPLICITY
		4602
		4603	If undefined or defined to C<1>, then all event-loop-specific functions
		4604	will have the C<struct ev_loop *> as first argument, and you can create
		4605	additional independent event loops. Otherwise there will be no support
		4606	for multiple event loops and there is no first event loop pointer
		4607	argument. Instead, all functions act on the single default loop.
		4608
		4609	Note that C<EV_DEFAULT> and C<EV_DEFAULT_> will no longer provide a
		4610	default loop when multiplicity is switched off - you always have to
		4611	initialise the loop manually in this case.
		4612
		4613	=item EV_MINPRI
		4614
		4615	=item EV_MAXPRI
		4616
		4617	The range of allowed priorities. C<EV_MINPRI> must be smaller or equal to
		4618	C<EV_MAXPRI>, but otherwise there are no non-obvious limitations. You can
		4619	provide for more priorities by overriding those symbols (usually defined
		4620	to be C<-2> and C<2>, respectively).
		4621
		4622	When doing priority-based operations, libev usually has to linearly search
		4623	all the priorities, so having many of them (hundreds) uses a lot of space
		4624	and time, so using the defaults of five priorities (-2 .. +2) is usually
		4625	fine.
		4626
		4627	If your embedding application does not need any priorities, defining these
		4628	both to C<0> will save some memory and CPU.
		4629
		4630	=item EV_PERIODIC_ENABLE, EV_IDLE_ENABLE, EV_EMBED_ENABLE, EV_STAT_ENABLE,
		4631	EV_PREPARE_ENABLE, EV_CHECK_ENABLE, EV_FORK_ENABLE, EV_SIGNAL_ENABLE,
		4632	EV_ASYNC_ENABLE, EV_CHILD_ENABLE.
		4633
		4634	If undefined or defined to be C<1> (and the platform supports it), then
		4635	the respective watcher type is supported. If defined to be C<0>, then it
		4636	is not. Disabling watcher types mainly saves code size.
		4637
		4638	=item EV_FEATURES
		4639
		4640	If you need to shave off some kilobytes of code at the expense of some
		4641	speed (but with the full API), you can define this symbol to request
		4642	certain subsets of functionality. The default is to enable all features
		4643	that can be enabled on the platform.
		4644
		4645	A typical way to use this symbol is to define it to C<0> (or to a bitset
		4646	with some broad features you want) and then selectively re-enable
		4647	additional parts you want, for example if you want everything minimal,
		4648	but multiple event loop support, async and child watchers and the poll
		4649	backend, use this:
		4650
		4651	#define EV_FEATURES 0
		4652	#define EV_MULTIPLICITY 1
		4653	#define EV_USE_POLL 1
		4654	#define EV_CHILD_ENABLE 1
		4655	#define EV_ASYNC_ENABLE 1
		4656
		4657	The actual value is a bitset, it can be a combination of the following
		4658	values (by default, all of these are enabled):
		4659
		4660	=over 4
		4661
		4662	=item C<1> - faster/larger code
		4663
		4664	Use larger code to speed up some operations.
		4665
		4666	Currently this is used to override some inlining decisions (enlarging the
		4667	code size by roughly 30% on amd64).
		4668
		4669	When optimising for size, use of compiler flags such as C<-Os> with
		4670	gcc is recommended, as well as C<-DNDEBUG>, as libev contains a number of
		4671	assertions.
		4672
		4673	The default is off when C<__OPTIMIZE_SIZE__> is defined by your compiler
		4674	(e.g. gcc with C<-Os>).
		4675
		4676	=item C<2> - faster/larger data structures
		4677
		4678	Replaces the small 2-heap for timer management by a faster 4-heap, larger
		4679	hash table sizes and so on. This will usually further increase code size
		4680	and can additionally have an effect on the size of data structures at
		4681	runtime.
		4682
		4683	The default is off when C<__OPTIMIZE_SIZE__> is defined by your compiler
		4684	(e.g. gcc with C<-Os>).
		4685
		4686	=item C<4> - full API configuration
		4687
		4688	This enables priorities (sets C<EV_MAXPRI>=2 and C<EV_MINPRI>=-2), and
		4689	enables multiplicity (C<EV_MULTIPLICITY>=1).
		4690
		4691	=item C<8> - full API
		4692
		4693	This enables a lot of the "lesser used" API functions. See C<ev.h> for
		4694	details on which parts of the API are still available without this
		4695	feature, and do not complain if this subset changes over time.
		4696
		4697	=item C<16> - enable all optional watcher types
		4698
		4699	Enables all optional watcher types. If you want to selectively enable
		4700	only some watcher types other than I/O and timers (e.g. prepare,
		4701	embed, async, child...) you can enable them manually by defining
		4702	C<EV_watchertype_ENABLE> to C<1> instead.
		4703
		4704	=item C<32> - enable all backends
		4705
		4706	This enables all backends - without this feature, you need to enable at
		4707	least one backend manually (C<EV_USE_SELECT> is a good choice).
		4708
		4709	=item C<64> - enable OS-specific "helper" APIs
		4710
		4711	Enable inotify, eventfd, signalfd and similar OS-specific helper APIs by
		4712	default.
		4713
		4714	=back
		4715
		4716	Compiling with C<gcc -Os -DEV_STANDALONE -DEV_USE_EPOLL=1 -DEV_FEATURES=0>
		4717	reduces the compiled size of libev from 24.7Kb code/2.8Kb data to 6.5Kb
		4718	code/0.3Kb data on my GNU/Linux amd64 system, while still giving you I/O
		4719	watchers, timers and monotonic clock support.
		4720
		4721	With an intelligent-enough linker (gcc+binutils are intelligent enough
		4722	when you use C<-Wl,--gc-sections -ffunction-sections>) functions unused by
		4723	your program might be left out as well - a binary starting a timer and an
		4724	I/O watcher then might come out at only 5Kb.
		4725
		4726	=item EV_API_STATIC
		4727
		4728	If this symbol is defined (by default it is not), then all identifiers
		4729	will have static linkage. This means that libev will not export any
		4730	identifiers, and you cannot link against libev anymore. This can be useful
		4731	when you embed libev, only want to use libev functions in a single file,
		4732	and do not want its identifiers to be visible.
		4733
		4734	To use this, define C<EV_API_STATIC> and include F<ev.c> in the file that
		4735	wants to use libev.
		4736
		4737	This option only works when libev is compiled with a C compiler, as C++
		4738	doesn't support the required declaration syntax.
		4739
		4740	=item EV_AVOID_STDIO
		4741
		4742	If this is set to C<1> at compiletime, then libev will avoid using stdio
		4743	functions (printf, scanf, perror etc.). This will increase the code size
		4744	somewhat, but if your program doesn't otherwise depend on stdio and your
		4745	libc allows it, this avoids linking in the stdio library which is quite
		4746	big.
		4747
		4748	Note that error messages might become less precise when this option is
		4749	enabled.
		4750
		4751	=item EV_NSIG
		4752
		4753	The highest supported signal number, +1 (or, the number of
		4754	signals): Normally, libev tries to deduce the maximum number of signals
		4755	automatically, but sometimes this fails, in which case it can be
		4756	specified. Also, using a lower number than detected (C<32> should be
		4757	good for about any system in existence) can save some memory, as libev
		4758	statically allocates some 12-24 bytes per signal number.
		4759
		4760	=item EV_PID_HASHSIZE
		4761
		4762	C<ev_child> watchers use a small hash table to distribute workload by
		4763	pid. The default size is C<16> (or C<1> with C<EV_FEATURES> disabled),
		4764	usually more than enough. If you need to manage thousands of children you
		4765	might want to increase this value (I<must> be a power of two).
		4766
		4767	=item EV_INOTIFY_HASHSIZE
		4768
		4769	C<ev_stat> watchers use a small hash table to distribute workload by
		4770	inotify watch id. The default size is C<16> (or C<1> with C<EV_FEATURES>
		4771	disabled), usually more than enough. If you need to manage thousands of
		4772	C<ev_stat> watchers you might want to increase this value (I<must> be a
		4773	power of two).
		4774
		4775	=item EV_USE_4HEAP
		4776
		4777	Heaps are not very cache-efficient. To improve the cache-efficiency of the
		4778	timer and periodics heaps, libev uses a 4-heap when this symbol is defined
		4779	to C<1>. The 4-heap uses more complicated (longer) code but has noticeably
		4780	faster performance with many (thousands) of watchers.
		4781
		4782	The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it
		4783	will be C<0>.
		4784
		4785	=item EV_HEAP_CACHE_AT
		4786
		4787	Heaps are not very cache-efficient. To improve the cache-efficiency of the
		4788	timer and periodics heaps, libev can cache the timestamp (I<at>) within
		4789	the heap structure (selected by defining C<EV_HEAP_CACHE_AT> to C<1>),
		4790	which uses 8-12 bytes more per watcher and a few hundred bytes more code,
		4791	but avoids random read accesses on heap changes. This improves performance
		4792	noticeably with many (hundreds) of watchers.
		4793
		4794	The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it
		4795	will be C<0>.
		4796
		4797	=item EV_VERIFY
		4798
		4799	Controls how much internal verification (see C<ev_verify ()>) will
		4800	be done: If set to C<0>, no internal verification code will be compiled
		4801	in. If set to C<1>, then verification code will be compiled in, but not
		4802	called. If set to C<2>, then the internal verification code will be
		4803	called once per loop, which can slow down libev. If set to C<3>, then the
		4804	verification code will be called very frequently, which will slow down
		4805	libev considerably.
		4806
		4807	The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it
		4808	will be C<0>.
		4809
		4810	=item EV_COMMON
		4811
		4812	By default, all watchers have a C<void *data> member. By redefining
		4813	this macro to something else you can include more and other types of
		4814	members. You have to define it each time you include one of the files,
		4815	though, and it must be identical each time.
		4816
		4817	For example, the perl EV module uses something like this:
		4818
		4819	#define EV_COMMON \
		4820	SV *self; /* contains this struct */ \
		4821	SV *cb_sv, *fh /* note no trailing ";" */
		4822
		4823	=item EV_CB_DECLARE (type)
		4824
		4825	=item EV_CB_INVOKE (watcher, revents)
		4826
		4827	=item ev_set_cb (ev, cb)
		4828
		4829	Can be used to change the callback member declaration in each watcher,
		4830	and the way callbacks are invoked and set. Must expand to a struct member
		4831	definition and a statement, respectively. See the F<ev.h> header file for
		4832	their default definitions. One possible use for overriding these is to
		4833	avoid the C<struct ev_loop *> as first argument in all cases, or to use
		4834	method calls instead of plain function calls in C++.
		4835
		4836	=back
		4837
		4838	=head2 EXPORTED API SYMBOLS
		4839
		4840	If you need to re-export the API (e.g. via a DLL) and you need a list of
		4841	exported symbols, you can use the provided F<Symbol.*> files which list
		4842	all public symbols, one per line:
		4843
		4844	Symbols.ev for libev proper
		4845	Symbols.event for the libevent emulation
		4846
		4847	This can also be used to rename all public symbols to avoid clashes with
		4848	multiple versions of libev linked together (which is obviously bad in
		4849	itself, but sometimes it is inconvenient to avoid this).
		4850
		4851	A sed command like this will create wrapper C<#define>'s that you need to
		4852	include before including F<ev.h>:
		4853
		4854	<Symbols.ev sed -e "s/.*/#define & myprefix_&/" >wrap.h
		4855
		4856	This would create a file F<wrap.h> which essentially looks like this:
		4857
		4858	#define ev_backend myprefix_ev_backend
		4859	#define ev_check_start myprefix_ev_check_start
		4860	#define ev_check_stop myprefix_ev_check_stop
		4861	...
		4862
		4863	=head2 EXAMPLES
		4864
		4865	For a real-world example of a program the includes libev
		4866	verbatim, you can have a look at the EV perl module
		4867	(L<http://software.schmorp.de/pkg/EV.html>). It has the libev files in
		4868	the F<libev/> subdirectory and includes them in the F<EV/EVAPI.h> (public
		4869	interface) and F<EV.xs> (implementation) files. Only the F<EV.xs> file
		4870	will be compiled. It is pretty complex because it provides its own header
		4871	file.
		4872
		4873	The usage in rxvt-unicode is simpler. It has a F<ev_cpp.h> header file
		4874	that everybody includes and which overrides some configure choices:
		4875
		4876	#define EV_FEATURES 8
		4877	#define EV_USE_SELECT 1
		4878	#define EV_PREPARE_ENABLE 1
		4879	#define EV_IDLE_ENABLE 1
		4880	#define EV_SIGNAL_ENABLE 1
		4881	#define EV_CHILD_ENABLE 1
		4882	#define EV_USE_STDEXCEPT 0
		4883	#define EV_CONFIG_H <config.h>
		4884
		4885	#include "ev++.h"
		4886
		4887	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:
		4888
		4889	#include "ev_cpp.h"
		4890	#include "ev.c"
		4891
		4892	=head1 INTERACTION WITH OTHER PROGRAMS, LIBRARIES OR THE ENVIRONMENT
		4893
		4894	=head2 THREADS AND COROUTINES
		4895
		4896	=head3 THREADS
		4897
		4898	All libev functions are reentrant and thread-safe unless explicitly
		4899	documented otherwise, but libev implements no locking itself. This means
		4900	that you can use as many loops as you want in parallel, as long as there
		4901	are no concurrent calls into any libev function with the same loop
		4902	parameter (C<ev_default_*> calls have an implicit default loop parameter,
		4903	of course): libev guarantees that different event loops share no data
		4904	structures that need any locking.
		4905
		4906	Or to put it differently: calls with different loop parameters can be done
		4907	concurrently from multiple threads, calls with the same loop parameter
		4908	must be done serially (but can be done from different threads, as long as
		4909	only one thread ever is inside a call at any point in time, e.g. by using
		4910	a mutex per loop).
		4911
		4912	Specifically to support threads (and signal handlers), libev implements
		4913	so-called C<ev_async> watchers, which allow some limited form of
		4914	concurrency on the same event loop, namely waking it up "from the
		4915	outside".
		4916
		4917	If you want to know which design (one loop, locking, or multiple loops
		4918	without or something else still) is best for your problem, then I cannot
		4919	help you, but here is some generic advice:
		4920
		4921	=over 4
		4922
		4923	=item * most applications have a main thread: use the default libev loop
		4924	in that thread, or create a separate thread running only the default loop.
		4925
		4926	This helps integrating other libraries or software modules that use libev
		4927	themselves and don't care/know about threading.
		4928
		4929	=item * one loop per thread is usually a good model.
		4930
		4931	Doing this is almost never wrong, sometimes a better-performance model
		4932	exists, but it is always a good start.
		4933
		4934	=item * other models exist, such as the leader/follower pattern, where one
		4935	loop is handed through multiple threads in a kind of round-robin fashion.
		4936
		4937	Choosing a model is hard - look around, learn, know that usually you can do
		4938	better than you currently do :-)
		4939
		4940	=item * often you need to talk to some other thread which blocks in the
		4941	event loop.
		4942
		4943	C<ev_async> watchers can be used to wake them up from other threads safely
		4944	(or from signal contexts...).
		4945
		4946	An example use would be to communicate signals or other events that only
		4947	work in the default loop by registering the signal watcher with the
		4948	default loop and triggering an C<ev_async> watcher from the default loop
		4949	watcher callback into the event loop interested in the signal.
		4950
		4951	=back
		4952
		4953	See also L<THREAD LOCKING EXAMPLE>.
		4954
		4955	=head3 COROUTINES
		4956
		4957	Libev is very accommodating to coroutines ("cooperative threads"):
		4958	libev fully supports nesting calls to its functions from different
		4959	coroutines (e.g. you can call C<ev_run> on the same loop from two
		4960	different coroutines, and switch freely between both coroutines running
		4961	the loop, as long as you don't confuse yourself). The only exception is
		4962	that you must not do this from C<ev_periodic> reschedule callbacks.
		4963
		4964	Care has been taken to ensure that libev does not keep local state inside
		4965	C<ev_run>, and other calls do not usually allow for coroutine switches as
		4966	they do not call any callbacks.
		4967
		4968	=head2 COMPILER WARNINGS
		4969
		4970	Depending on your compiler and compiler settings, you might get no or a
		4971	lot of warnings when compiling libev code. Some people are apparently
		4972	scared by this.
		4973
		4974	However, these are unavoidable for many reasons. For one, each compiler
		4975	has different warnings, and each user has different tastes regarding
		4976	warning options. "Warn-free" code therefore cannot be a goal except when
		4977	targeting a specific compiler and compiler-version.
		4978
		4979	Another reason is that some compiler warnings require elaborate
		4980	workarounds, or other changes to the code that make it less clear and less
		4981	maintainable.
		4982
		4983	And of course, some compiler warnings are just plain stupid, or simply
		4984	wrong (because they don't actually warn about the condition their message
		4985	seems to warn about). For example, certain older gcc versions had some
		4986	warnings that resulted in an extreme number of false positives. These have
		4987	been fixed, but some people still insist on making code warn-free with
		4988	such buggy versions.
		4989
		4990	While libev is written to generate as few warnings as possible,
		4991	"warn-free" code is not a goal, and it is recommended not to build libev
		4992	with any compiler warnings enabled unless you are prepared to cope with
		4993	them (e.g. by ignoring them). Remember that warnings are just that:
		4994	warnings, not errors, or proof of bugs.
		4995
		4996
		4997	=head2 VALGRIND
		4998
		4999	Valgrind has a special section here because it is a popular tool that is
		5000	highly useful. Unfortunately, valgrind reports are very hard to interpret.
		5001
		5002	If you think you found a bug (memory leak, uninitialised data access etc.)
		5003	in libev, then check twice: If valgrind reports something like:
		5004
		5005	==2274== definitely lost: 0 bytes in 0 blocks.
		5006	==2274== possibly lost: 0 bytes in 0 blocks.
		5007	==2274== still reachable: 256 bytes in 1 blocks.
		5008
		5009	Then there is no memory leak, just as memory accounted to global variables
		5010	is not a memleak - the memory is still being referenced, and didn't leak.
		5011
		5012	Similarly, under some circumstances, valgrind might report kernel bugs
		5013	as if it were a bug in libev (e.g. in realloc or in the poll backend,
		5014	although an acceptable workaround has been found here), or it might be
		5015	confused.
		5016
		5017	Keep in mind that valgrind is a very good tool, but only a tool. Don't
		5018	make it into some kind of religion.
		5019
		5020	If you are unsure about something, feel free to contact the mailing list
		5021	with the full valgrind report and an explanation on why you think this
		5022	is a bug in libev (best check the archives, too :). However, don't be
		5023	annoyed when you get a brisk "this is no bug" answer and take the chance
		5024	of learning how to interpret valgrind properly.
		5025
		5026	If you need, for some reason, empty reports from valgrind for your project
		5027	I suggest using suppression lists.
		5028
		5029
		5030	=head1 PORTABILITY NOTES
		5031
		5032	=head2 GNU/LINUX 32 BIT LIMITATIONS
		5033
		5034	GNU/Linux is the only common platform that supports 64 bit file/large file
		5035	interfaces but I<disables> them by default.
		5036
		5037	That means that libev compiled in the default environment doesn't support
		5038	files larger than 2GiB or so, which mainly affects C<ev_stat> watchers.
		5039
		5040	Unfortunately, many programs try to work around this GNU/Linux issue
		5041	by enabling the large file API, which makes them incompatible with the
		5042	standard libev compiled for their system.
		5043
		5044	Likewise, libev cannot enable the large file API itself as this would
		5045	suddenly make it incompatible to the default compile time environment,
		5046	i.e. all programs not using special compile switches.
		5047
		5048	=head2 OS/X AND DARWIN BUGS
		5049
		5050	The whole thing is a bug if you ask me - basically any system interface
		5051	you touch is broken, whether it is locales, poll, kqueue or even the
		5052	OpenGL drivers.
		5053
		5054	=head3 C<kqueue> is buggy
		5055
		5056	The kqueue syscall is broken in all known versions - most versions support
		5057	only sockets, many support pipes.
		5058
		5059	Libev tries to work around this by not using C<kqueue> by default on this
		5060	rotten platform, but of course you can still ask for it when creating a
		5061	loop - embedding a socket-only kqueue loop into a select-based one is
		5062	probably going to work well.
		5063
		5064	=head3 C<poll> is buggy
		5065
		5066	Instead of fixing C<kqueue>, Apple replaced their (working) C<poll>
		5067	implementation by something calling C<kqueue> internally around the 10.5.6
		5068	release, so now C<kqueue> I<and> C<poll> are broken.
		5069
		5070	Libev tries to work around this by not using C<poll> by default on
		5071	this rotten platform, but of course you can still ask for it when creating
		5072	a loop.
		5073
		5074	=head3 C<select> is buggy
		5075
		5076	All that's left is C<select>, and of course Apple found a way to fuck this
		5077	one up as well: On OS/X, C<select> actively limits the number of file
		5078	descriptors you can pass in to 1024 - your program suddenly crashes when
		5079	you use more.
		5080
		5081	There is an undocumented "workaround" for this - defining
		5082	C<_DARWIN_UNLIMITED_SELECT>, which libev tries to use, so select I<should>
		5083	work on OS/X.
		5084
		5085	=head2 SOLARIS PROBLEMS AND WORKAROUNDS
		5086
		5087	=head3 C<errno> reentrancy
		5088
		5089	The default compile environment on Solaris is unfortunately so
		5090	thread-unsafe that you can't even use components/libraries compiled
		5091	without C<-D_REENTRANT> in a threaded program, which, of course, isn't
		5092	defined by default. A valid, if stupid, implementation choice.
		5093
		5094	If you want to use libev in threaded environments you have to make sure
		5095	it's compiled with C<_REENTRANT> defined.
		5096
		5097	=head3 Event port backend
		5098
		5099	The scalable event interface for Solaris is called "event
		5100	ports". Unfortunately, this mechanism is very buggy in all major
		5101	releases. If you run into high CPU usage, your program freezes or you get
		5102	a large number of spurious wakeups, make sure you have all the relevant
		5103	and latest kernel patches applied. No, I don't know which ones, but there
		5104	are multiple ones to apply, and afterwards, event ports actually work
		5105	great.
		5106
		5107	If you can't get it to work, you can try running the program by setting
		5108	the environment variable C<LIBEV_FLAGS=3> to only allow C<poll> and
		5109	C<select> backends.
		5110
		5111	=head2 AIX POLL BUG
		5112
		5113	AIX unfortunately has a broken C<poll.h> header. Libev works around
		5114	this by trying to avoid the poll backend altogether (i.e. it's not even
		5115	compiled in), which normally isn't a big problem as C<select> works fine
		5116	with large bitsets on AIX, and AIX is dead anyway.
		5117
		5118	=head2 WIN32 PLATFORM LIMITATIONS AND WORKAROUNDS
		5119
		5120	=head3 General issues
		5121
		5122	Win32 doesn't support any of the standards (e.g. POSIX) that libev
		5123	requires, and its I/O model is fundamentally incompatible with the POSIX
		5124	model. Libev still offers limited functionality on this platform in
		5125	the form of the C<EVBACKEND_SELECT> backend, and only supports socket
		5126	descriptors. This only applies when using Win32 natively, not when using
		5127	e.g. cygwin. Actually, it only applies to the microsofts own compilers,
		5128	as every compiler comes with a slightly differently broken/incompatible
		5129	environment.
		5130
		5131	Lifting these limitations would basically require the full
		5132	re-implementation of the I/O system. If you are into this kind of thing,
		5133	then note that glib does exactly that for you in a very portable way (note
		5134	also that glib is the slowest event library known to man).
		5135
		5136	There is no supported compilation method available on windows except
		5137	embedding it into other applications.
		5138
		5139	Sensible signal handling is officially unsupported by Microsoft - libev
		5140	tries its best, but under most conditions, signals will simply not work.
		5141
		5142	Not a libev limitation but worth mentioning: windows apparently doesn't
		5143	accept large writes: instead of resulting in a partial write, windows will
		5144	either accept everything or return C<ENOBUFS> if the buffer is too large,
		5145	so make sure you only write small amounts into your sockets (less than a
		5146	megabyte seems safe, but this apparently depends on the amount of memory
		5147	available).
		5148
		5149	Due to the many, low, and arbitrary limits on the win32 platform and
		5150	the abysmal performance of winsockets, using a large number of sockets
		5151	is not recommended (and not reasonable). If your program needs to use
		5152	more than a hundred or so sockets, then likely it needs to use a totally
		5153	different implementation for windows, as libev offers the POSIX readiness
		5154	notification model, which cannot be implemented efficiently on windows
		5155	(due to Microsoft monopoly games).
		5156
		5157	A typical way to use libev under windows is to embed it (see the embedding
		5158	section for details) and use the following F<evwrap.h> header file instead
		5159	of F<ev.h>:
		5160
		5161	#define EV_STANDALONE /* keeps ev from requiring config.h */
		5162	#define EV_SELECT_IS_WINSOCKET 1 /* configure libev for windows select */
		5163
		5164	#include "ev.h"
		5165
		5166	And compile the following F<evwrap.c> file into your project (make sure
		5167	you do I<not> compile the F<ev.c> or any other embedded source files!):
		5168
		5169	#include "evwrap.h"
		5170	#include "ev.c"
		5171
		5172	=head3 The winsocket C<select> function
		5173
		5174	The winsocket C<select> function doesn't follow POSIX in that it
		5175	requires socket I<handles> and not socket I<file descriptors> (it is
		5176	also extremely buggy). This makes select very inefficient, and also
		5177	requires a mapping from file descriptors to socket handles (the Microsoft
		5178	C runtime provides the function C<_open_osfhandle> for this). See the
		5179	discussion of the C<EV_SELECT_USE_FD_SET>, C<EV_SELECT_IS_WINSOCKET> and
		5180	C<EV_FD_TO_WIN32_HANDLE> preprocessor symbols for more info.
		5181
		5182	The configuration for a "naked" win32 using the Microsoft runtime
		5183	libraries and raw winsocket select is:
		5184
		5185	#define EV_USE_SELECT 1
		5186	#define EV_SELECT_IS_WINSOCKET 1 /* forces EV_SELECT_USE_FD_SET, too */
		5187
		5188	Note that winsockets handling of fd sets is O(n), so you can easily get a
		5189	complexity in the O(n²) range when using win32.
		5190
		5191	=head3 Limited number of file descriptors
		5192
		5193	Windows has numerous arbitrary (and low) limits on things.
		5194
		5195	Early versions of winsocket's select only supported waiting for a maximum
		5196	of C<64> handles (probably owning to the fact that all windows kernels
		5197	can only wait for C<64> things at the same time internally; Microsoft
		5198	recommends spawning a chain of threads and wait for 63 handles and the
		5199	previous thread in each. Sounds great!).
		5200
		5201	Newer versions support more handles, but you need to define C<FD_SETSIZE>
		5202	to some high number (e.g. C<2048>) before compiling the winsocket select
		5203	call (which might be in libev or elsewhere, for example, perl and many
		5204	other interpreters do their own select emulation on windows).
		5205
		5206	Another limit is the number of file descriptors in the Microsoft runtime
		5207	libraries, which by default is C<64> (there must be a hidden I<64>
		5208	fetish or something like this inside Microsoft). You can increase this
		5209	by calling C<_setmaxstdio>, which can increase this limit to C<2048>
		5210	(another arbitrary limit), but is broken in many versions of the Microsoft
		5211	runtime libraries. This might get you to about C<512> or C<2048> sockets
		5212	(depending on windows version and/or the phase of the moon). To get more,
		5213	you need to wrap all I/O functions and provide your own fd management, but
		5214	the cost of calling select (O(n²)) will likely make this unworkable.
		5215
		5216	=head2 PORTABILITY REQUIREMENTS
		5217
		5218	In addition to a working ISO-C implementation and of course the
		5219	backend-specific APIs, libev relies on a few additional extensions:
		5220
		5221	=over 4
		5222
		5223	=item C<void (*)(ev_watcher_type *, int revents)> must have compatible
		5224	calling conventions regardless of C<ev_watcher_type *>.
		5225
		5226	Libev assumes not only that all watcher pointers have the same internal
		5227	structure (guaranteed by POSIX but not by ISO C for example), but it also
		5228	assumes that the same (machine) code can be used to call any watcher
		5229	callback: The watcher callbacks have different type signatures, but libev
		5230	calls them using an C<ev_watcher *> internally.
		5231
		5232	=item pointer accesses must be thread-atomic
		5233
		5234	Accessing a pointer value must be atomic, it must both be readable and
		5235	writable in one piece - this is the case on all current architectures.
		5236
		5237	=item C<sig_atomic_t volatile> must be thread-atomic as well
		5238
		5239	The type C<sig_atomic_t volatile> (or whatever is defined as
		5240	C<EV_ATOMIC_T>) must be atomic with respect to accesses from different
		5241	threads. This is not part of the specification for C<sig_atomic_t>, but is
		5242	believed to be sufficiently portable.
		5243
		5244	=item C<sigprocmask> must work in a threaded environment
		5245
		5246	Libev uses C<sigprocmask> to temporarily block signals. This is not
		5247	allowed in a threaded program (C<pthread_sigmask> has to be used). Typical
		5248	pthread implementations will either allow C<sigprocmask> in the "main
		5249	thread" or will block signals process-wide, both behaviours would
		5250	be compatible with libev. Interaction between C<sigprocmask> and
		5251	C<pthread_sigmask> could complicate things, however.
		5252
		5253	The most portable way to handle signals is to block signals in all threads
		5254	except the initial one, and run the default loop in the initial thread as
		5255	well.
		5256
		5257	=item C<long> must be large enough for common memory allocation sizes
		5258
		5259	To improve portability and simplify its API, libev uses C<long> internally
		5260	instead of C<size_t> when allocating its data structures. On non-POSIX
		5261	systems (Microsoft...) this might be unexpectedly low, but is still at
		5262	least 31 bits everywhere, which is enough for hundreds of millions of
		5263	watchers.
		5264
		5265	=item C<double> must hold a time value in seconds with enough accuracy
		5266
		5267	The type C<double> is used to represent timestamps. It is required to
		5268	have at least 51 bits of mantissa (and 9 bits of exponent), which is
		5269	good enough for at least into the year 4000 with millisecond accuracy
		5270	(the design goal for libev). This requirement is overfulfilled by
		5271	implementations using IEEE 754, which is basically all existing ones.
		5272
		5273	With IEEE 754 doubles, you get microsecond accuracy until at least the
		5274	year 2255 (and millisecond accuracy till the year 287396 - by then, libev
		5275	is either obsolete or somebody patched it to use C<long double> or
		5276	something like that, just kidding).
		5277
		5278	=back
		5279
		5280	If you know of other additional requirements drop me a note.
		5281
		5282
		5283	=head1 ALGORITHMIC COMPLEXITIES
		5284
		5285	In this section the complexities of (many of) the algorithms used inside
		5286	libev will be documented. For complexity discussions about backends see
		5287	the documentation for C<ev_default_init>.
		5288
		5289	All of the following are about amortised time: If an array needs to be
		5290	extended, libev needs to realloc and move the whole array, but this
		5291	happens asymptotically rarer with higher number of elements, so O(1) might
		5292	mean that libev does a lengthy realloc operation in rare cases, but on
		5293	average it is much faster and asymptotically approaches constant time.
		5294
		5295	=over 4
		5296
		5297	=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers)
		5298
		5299	This means that, when you have a watcher that triggers in one hour and
		5300	there are 100 watchers that would trigger before that, then inserting will
		5301	have to skip roughly seven (C<ld 100>) of these watchers.
		5302
		5303	=item Changing timer/periodic watchers (by autorepeat or calling again): O(log skipped_other_timers)
		5304
		5305	That means that changing a timer costs less than removing/adding them,
		5306	as only the relative motion in the event queue has to be paid for.
		5307
		5308	=item Starting io/check/prepare/idle/signal/child/fork/async watchers: O(1)
		5309
		5310	These just add the watcher into an array or at the head of a list.
		5311
		5312	=item Stopping check/prepare/idle/fork/async watchers: O(1)
		5313
		5314	=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % EV_PID_HASHSIZE))
		5315
		5316	These watchers are stored in lists, so they need to be walked to find the
		5317	correct watcher to remove. The lists are usually short (you don't usually
		5318	have many watchers waiting for the same fd or signal: one is typical, two
		5319	is rare).
		5320
		5321	=item Finding the next timer in each loop iteration: O(1)
		5322
		5323	By virtue of using a binary or 4-heap, the next timer is always found at a
		5324	fixed position in the storage array.
		5325
		5326	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)
		5327
		5328	A change means an I/O watcher gets started or stopped, which requires
		5329	libev to recalculate its status (and possibly tell the kernel, depending
		5330	on backend and whether C<ev_io_set> was used).
		5331
		5332	=item Activating one watcher (putting it into the pending state): O(1)
		5333
		5334	=item Priority handling: O(number_of_priorities)
		5335
		5336	Priorities are implemented by allocating some space for each
		5337	priority. When doing priority-based operations, libev usually has to
		5338	linearly search all the priorities, but starting/stopping and activating
		5339	watchers becomes O(1) with respect to priority handling.
		5340
		5341	=item Sending an ev_async: O(1)
		5342
		5343	=item Processing ev_async_send: O(number_of_async_watchers)
		5344
		5345	=item Processing signals: O(max_signal_number)
		5346
		5347	Sending involves a system call I<iff> there were no other C<ev_async_send>
		5348	calls in the current loop iteration and the loop is currently
		5349	blocked. Checking for async and signal events involves iterating over all
		5350	running async watchers or all signal numbers.
		5351
		5352	=back
		5353
		5354
		5355	=head1 PORTING FROM LIBEV 3.X TO 4.X
		5356
		5357	The major version 4 introduced some incompatible changes to the API.
		5358
		5359	At the moment, the C<ev.h> header file provides compatibility definitions
		5360	for all changes, so most programs should still compile. The compatibility
		5361	layer might be removed in later versions of libev, so better update to the
		5362	new API early than late.
		5363
		5364	=over 4
		5365
		5366	=item C<EV_COMPAT3> backwards compatibility mechanism
		5367
		5368	The backward compatibility mechanism can be controlled by
		5369	C<EV_COMPAT3>. See L<PREPROCESSOR SYMBOLS/MACROS> in the L<EMBEDDING>
		5370	section.
		5371
		5372	=item C<ev_default_destroy> and C<ev_default_fork> have been removed
		5373
		5374	These calls can be replaced easily by their C<ev_loop_xxx> counterparts:
		5375
		5376	ev_loop_destroy (EV_DEFAULT_UC);
		5377	ev_loop_fork (EV_DEFAULT);
		5378
		5379	=item function/symbol renames
		5380
		5381	A number of functions and symbols have been renamed:
		5382
		5383	ev_loop => ev_run
		5384	EVLOOP_NONBLOCK => EVRUN_NOWAIT
		5385	EVLOOP_ONESHOT => EVRUN_ONCE
		5386
		5387	ev_unloop => ev_break
		5388	EVUNLOOP_CANCEL => EVBREAK_CANCEL
		5389	EVUNLOOP_ONE => EVBREAK_ONE
		5390	EVUNLOOP_ALL => EVBREAK_ALL
		5391
		5392	EV_TIMEOUT => EV_TIMER
		5393
		5394	ev_loop_count => ev_iteration
		5395	ev_loop_depth => ev_depth
		5396	ev_loop_verify => ev_verify
		5397
		5398	Most functions working on C<struct ev_loop> objects don't have an
		5399	C<ev_loop_> prefix, so it was removed; C<ev_loop>, C<ev_unloop> and
		5400	associated constants have been renamed to not collide with the C<struct
		5401	ev_loop> anymore and C<EV_TIMER> now follows the same naming scheme
		5402	as all other watcher types. Note that C<ev_loop_fork> is still called
		5403	C<ev_loop_fork> because it would otherwise clash with the C<ev_fork>
		5404	typedef.
		5405
		5406	=item C<EV_MINIMAL> mechanism replaced by C<EV_FEATURES>
		5407
		5408	The preprocessor symbol C<EV_MINIMAL> has been replaced by a different
		5409	mechanism, C<EV_FEATURES>. Programs using C<EV_MINIMAL> usually compile
		5410	and work, but the library code will of course be larger.
		5411
		5412	=back
		5413
		5414
		5415	=head1 GLOSSARY
		5416
		5417	=over 4
		5418
		5419	=item active
		5420
		5421	A watcher is active as long as it has been started and not yet stopped.
		5422	See L<WATCHER STATES> for details.
		5423
		5424	=item application
		5425
		5426	In this document, an application is whatever is using libev.
		5427
		5428	=item backend
		5429
		5430	The part of the code dealing with the operating system interfaces.
		5431
		5432	=item callback
		5433
		5434	The address of a function that is called when some event has been
		5435	detected. Callbacks are being passed the event loop, the watcher that
		5436	received the event, and the actual event bitset.
		5437
		5438	=item callback/watcher invocation
		5439
		5440	The act of calling the callback associated with a watcher.
		5441
		5442	=item event
		5443
		5444	A change of state of some external event, such as data now being available
		5445	for reading on a file descriptor, time having passed or simply not having
		5446	any other events happening anymore.
		5447
		5448	In libev, events are represented as single bits (such as C<EV_READ> or
		5449	C<EV_TIMER>).
		5450
		5451	=item event library
		5452
		5453	A software package implementing an event model and loop.
		5454
		5455	=item event loop
		5456
		5457	An entity that handles and processes external events and converts them
		5458	into callback invocations.
		5459
		5460	=item event model
		5461
		5462	The model used to describe how an event loop handles and processes
		5463	watchers and events.
		5464
		5465	=item pending
		5466
		5467	A watcher is pending as soon as the corresponding event has been
		5468	detected. See L<WATCHER STATES> for details.
		5469
		5470	=item real time
		5471
		5472	The physical time that is observed. It is apparently strictly monotonic :)
		5473
		5474	=item wall-clock time
		5475
		5476	The time and date as shown on clocks. Unlike real time, it can actually
		5477	be wrong and jump forwards and backwards, e.g. when you adjust your
		5478	clock.
		5479
		5480	=item watcher
		5481
		5482	A data structure that describes interest in certain events. Watchers need
		5483	to be started (attached to an event loop) before they can receive events.
		5484
		5485	=back
825		5486
826	=head1 AUTHOR	5487	=head1 AUTHOR
827		5488
828	Marc Lehmann <libev@schmorp.de>.	5489	Marc Lehmann <libev@schmorp.de>, with repeated corrections by Mikael
		5490	Magnusson and Emanuele Giaquinta, and minor corrections by many others.
829		5491

Diff Legend

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