ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/libev/ev.pod

Comparing libev/ev.pod (file contents):
Revision 1.42 by root, Sat Nov 24 16:31:45 2007 UTC vs.
Revision 1.371 by root, Sat Jun 4 05:25:03 2011 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 C<double> type in C, and when you need to do any calculations on	140	too. It usually aliases to the C<double> type in C. When you need to do
51	it, you should treat it as such.	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.
52		165
53		166
54	=head1 GLOBAL FUNCTIONS	167	=head1 GLOBAL FUNCTIONS
55		168
56	These functions can be called anytime, even before initialising the	169	These functions can be called anytime, even before initialising the
…		…
60		173
61	=item ev_tstamp ev_time ()	174	=item ev_tstamp ev_time ()
62		175
63	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
64	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
65	you actually want to know.	178	you actually want to know. Also interesting is the combination of
		179	C<ev_update_now> 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 >>).
66		192
67	=item int ev_version_major ()	193	=item int ev_version_major ()
68		194
69	=item int ev_version_minor ()	195	=item int ev_version_minor ()
70		196
71	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
72	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
73	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
74	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
75	version of the library your program was compiled against.	201	version of the library your program was compiled against.
76		202
		203	These version numbers refer to the ABI version of the library, not the
		204	release version.
		205
77	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,
78	as this indicates an incompatible change. Minor versions are usually	207	as this indicates an incompatible change. Minor versions are usually
79	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
80	not a problem.	209	not a problem.
81		210
82	Example: make sure we haven't accidentally been linked against the wrong	211	Example: Make sure we haven't accidentally been linked against the wrong
83	version:	212	version (note, however, that this will not detect other ABI mismatches,
		213	such as LFS or reentrancy).
84		214
85	assert (("libev version mismatch",	215	assert (("libev version mismatch",
86	ev_version_major () == EV_VERSION_MAJOR	216	ev_version_major () == EV_VERSION_MAJOR
87	&& ev_version_minor () >= EV_VERSION_MINOR));	217	&& ev_version_minor () >= EV_VERSION_MINOR));
88		218
89	=item unsigned int ev_supported_backends ()	219	=item unsigned int ev_supported_backends ()
90		220
91	Return the set of all backends (i.e. their corresponding C<EV_BACKEND_*>	221	Return the set of all backends (i.e. their corresponding C<EV_BACKEND_*>
92	value) compiled into this binary of libev (independent of their	222	value) compiled into this binary of libev (independent of their
…		…
94	a description of the set values.	224	a description of the set values.
95		225
96	Example: make sure we have the epoll method, because yeah this is cool and	226	Example: make sure we have the epoll method, because yeah this is cool and
97	a must have and can we have a torrent of it please!!!11	227	a must have and can we have a torrent of it please!!!11
98		228
99	assert (("sorry, no epoll, no sex",	229	assert (("sorry, no epoll, no sex",
100	ev_supported_backends () & EVBACKEND_EPOLL));	230	ev_supported_backends () & EVBACKEND_EPOLL));
101		231
102	=item unsigned int ev_recommended_backends ()	232	=item unsigned int ev_recommended_backends ()
103		233
104	Return the set of all backends compiled into this binary of libev and also	234	Return the set of all backends compiled into this binary of libev and
105	recommended for this platform. This set is often smaller than the one	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
106	returned by C<ev_supported_backends>, as for example kqueue is broken on	237	C<ev_supported_backends>, as for example kqueue is broken on most BSDs
107	most BSDs and will not be autodetected unless you explicitly request it	238	and will not be auto-detected unless you explicitly request it (assuming
108	(assuming you know what you are doing). This is the set of backends that	239	you know what you are doing). This is the set of backends that libev will
109	libev will probe for if you specify no backends explicitly.	240	probe for if you specify no backends explicitly.
110		241
111	=item unsigned int ev_embeddable_backends ()	242	=item unsigned int ev_embeddable_backends ()
112		243
113	Returns the set of backends that are embeddable in other event loops. This	244	Returns the set of backends that are embeddable in other event loops. This
114	is the theoretical, all-platform, value. To find which backends	245	value is platform-specific but can include backends not available on the
115	might be supported on the current system, you would need to look at	246	current system. To find which embeddable backends might be supported on
116	C<ev_embeddable_backends () & ev_supported_backends ()>, likewise for	247	the current system, you would need to look at C<ev_embeddable_backends ()
117	recommended ones.	248	& ev_supported_backends ()>, likewise for recommended ones.
118		249
119	See the description of C<ev_embed> watchers for more info.	250	See the description of C<ev_embed> watchers for more info.
120		251
121	=item ev_set_allocator (void *(*cb)(void *ptr, long size))	252	=item ev_set_allocator (void *(*cb)(void *ptr, long size))
122		253
123	Sets the allocation function to use (the prototype is similar to the	254	Sets the allocation function to use (the prototype is similar - the
124	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
125	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
126	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
127	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.
128		263
129	You could override this function in high-availability programs to, say,	264	You could override this function in high-availability programs to, say,
130	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,
131	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.
132		267
133	Example: replace the libev allocator with one that waits a bit and then	268	Example: Replace the libev allocator with one that waits a bit and then
134	retries: better than mine).	269	retries (example requires a standards-compliant C<realloc>).
135		270
136	static void *	271	static void *
137	persistent_realloc (void *ptr, long size)	272	persistent_realloc (void *ptr, size_t size)
138	{	273	{
139	for (;;)	274	for (;;)
140	{	275	{
141	void *newptr = realloc (ptr, size);	276	void *newptr = realloc (ptr, size);
142		277
…		…
148	}	283	}
149		284
150	...	285	...
151	ev_set_allocator (persistent_realloc);	286	ev_set_allocator (persistent_realloc);
152		287
153	=item ev_set_syserr_cb (void (*cb)(const char *msg));	288	=item ev_set_syserr_cb (void (*cb)(const char *msg))
154		289
155	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
156	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
157	indicating the system call or subsystem causing the problem. If this	292	indicating the system call or subsystem causing the problem. If this
158	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
159	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
160	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
161	(such as abort).	296	(such as abort).
162		297
163	Example: do the same thing as libev does internally:	298	Example: This is basically the same thing that libev does internally, too.
164		299
165	static void	300	static void
166	fatal_error (const char *msg)	301	fatal_error (const char *msg)
167	{	302	{
168	perror (msg);	303	perror (msg);
…		…
170	}	305	}
171		306
172	...	307	...
173	ev_set_syserr_cb (fatal_error);	308	ev_set_syserr_cb (fatal_error);
174		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
175	=back	323	=back
176		324
177	=head1 FUNCTIONS CONTROLLING THE EVENT LOOP	325	=head1 FUNCTIONS CONTROLLING EVENT LOOPS
178		326
179	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
180	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
181	events, and dynamically created loops which do not.	329	libev 3 had an C<ev_loop> function colliding with the struct name).
182		330
183	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
184	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
185	create, you also create another event loop. Libev itself does no locking	333	do not.
186	whatsoever, so if you mix calls to the same event loop in different
187	threads, make sure you lock (this is usually a bad idea, though, even if
188	done correctly, because it's hideous and inefficient).
189		334
190	=over 4	335	=over 4
191		336
192	=item struct ev_loop *ev_default_loop (unsigned int flags)	337	=item struct ev_loop *ev_default_loop (unsigned int flags)
193		338
194	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
195	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
196	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
197	flags. If that is troubling you, check C<ev_backend ()> afterwards).	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".
198		349
199	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
200	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.
201		382
202	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
203	backends to use, and is usually specified as C<0> (or C<EVFLAG_AUTO>).	384	backends to use, and is usually specified as C<0> (or C<EVFLAG_AUTO>).
204		385
205	The following flags are supported:	386	The following flags are supported:
…		…
211	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
212	thing, believe me).	393	thing, believe me).
213		394
214	=item C<EVFLAG_NOENV>	395	=item C<EVFLAG_NOENV>
215		396
216	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
217	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
218	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will	399	C<LIBEV_FLAGS>. Otherwise (the default), this environment variable will
219	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
220	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
221	around bugs.	402	around bugs.
222		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 ahve 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
223	=item C<EVBACKEND_SELECT> (value 1, portable select backend)	458	=item C<EVBACKEND_SELECT> (value 1, portable select backend)
224		459
225	This is your standard select(2) backend. Not I<completely> standard, as	460	This is your standard select(2) backend. Not I<completely> standard, as
226	libev tries to roll its own fd_set with no limits on the number of fds,	461	libev tries to roll its own fd_set with no limits on the number of fds,
227	but if that fails, expect a fairly low limit on the number of fds when	462	but if that fails, expect a fairly low limit on the number of fds when
228	using this backend. It doesn't scale too well (O(highest_fd)), but its usually	463	using this backend. It doesn't scale too well (O(highest_fd)), but its
229	the fastest backend for a low number of fds.	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).
230		476
231	=item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows)	477	=item C<EVBACKEND_POLL> (value 2, poll backend, available everywhere except on windows)
232		478
233	And this is your standard poll(2) backend. It's more complicated than	479	And this is your standard poll(2) backend. It's more complicated
234	select, but handles sparse fds better and has no artificial limit on the	480	than select, but handles sparse fds better and has no artificial
235	number of fds you can use (except it will slow down considerably with a	481	limit on the number of fds you can use (except it will slow down
236	lot of inactive fds). It scales similarly to select, i.e. O(total_fds).	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.
		485
		486	This backend maps C<EV_READ> to C<POLLIN | POLLERR | POLLHUP>, and
		487	C<EV_WRITE> to C<POLLOUT | POLLERR | POLLHUP>.
237		488
238	=item C<EVBACKEND_EPOLL> (value 4, Linux)	489	=item C<EVBACKEND_EPOLL> (value 4, Linux)
239		490
		491	Use the linux-specific epoll(7) interface (for both pre- and post-2.6.9
		492	kernels).
		493
240	For few fds, this backend is a bit little slower than poll and select,	494	For few fds, this backend is a bit little slower than poll and select, but
241	but it scales phenomenally better. While poll and select usually scale like	495	it scales phenomenally better. While poll and select usually scale like
242	O(total_fds) where n is the total number of fds (or the highest fd), epoll scales	496	O(total_fds) where total_fds is the total number of fds (or the highest
243	either O(1) or O(active_fds).	497	fd), epoll scales either O(1) or O(active_fds).
244		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 errornously 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
245	While stopping and starting an I/O watcher in the same iteration will	527	While stopping, setting and starting an I/O watcher in the same iteration
246	result in some caching, there is still a syscall per such incident	528	will result in some caching, there is still a system call per such
247	(because the fd could point to a different file description now), so its	529	incident (because the same I<file descriptor> could point to a different
248	best to avoid that. Also, dup()ed file descriptors might not work very	530	I<file description> now), so its best to avoid that. Also, C<dup ()>'ed
249	well if you register events for both fds.	531	file descriptors might not work very well if you register events for both
		532	file descriptors.
250		533
251	Please note that epoll sometimes generates spurious notifications, so you	534	Best performance from this backend is achieved by not unregistering all
252	need to use non-blocking I/O or other means to avoid blocking when no data	535	watchers for a file descriptor until it has been closed, if possible,
253	(or space) is available.	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>.
254		551
255	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)	552	=item C<EVBACKEND_KQUEUE> (value 8, most BSD clones)
256		553
257	Kqueue deserves special mention, as at the time of this writing, it	554	Kqueue deserves special mention, as at the time of this writing, it
258	was broken on all BSDs except NetBSD (usually it doesn't work with	555	was broken on all BSDs except NetBSD (usually it doesn't work reliably
259	anything but sockets and pipes, except on Darwin, where of course its	556	with anything but sockets and pipes, except on Darwin, where of course
260	completely useless). For this reason its not being "autodetected"	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
261	unless you explicitly specify it explicitly in the flags (i.e. using	560	"auto-detected" unless you explicitly specify it in the flags (i.e. using
262	C<EVBACKEND_KQUEUE>).	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.
263		567
264	It scales in the same way as the epoll backend, but the interface to the	568	It scales in the same way as the epoll backend, but the interface to the
265	kernel is more efficient (which says nothing about its actual speed, of	569	kernel is more efficient (which says nothing about its actual speed, of
266	course). While starting and stopping an I/O watcher does not cause an	570	course). While stopping, setting and starting an I/O watcher does never
267	extra syscall as with epoll, it still adds up to four event changes per	571	cause an extra system call as with C<EVBACKEND_EPOLL>, it still adds up to
268	incident, so its best to avoid that.	572	two event changes per incident. Support for C<fork ()> is very bad (but
		573	sane, unlike epoll) and it drops fds silently in similarly hard-to-detect
		574	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>.
269		588
270	=item C<EVBACKEND_DEVPOLL> (value 16, Solaris 8)	589	=item C<EVBACKEND_DEVPOLL> (value 16, Solaris 8)
271		590
272	This is not implemented yet (and might never be).	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.
273		595
274	=item C<EVBACKEND_PORT> (value 32, Solaris 10)	596	=item C<EVBACKEND_PORT> (value 32, Solaris 10)
275		597
276	This uses the Solaris 10 port mechanism. As with everything on Solaris,	598	This uses the Solaris 10 event port mechanism. As with everything on Solaris,
277	it's really slow, but it still scales very well (O(active_fds)).	599	it's really slow, but it still scales very well (O(active_fds)).
278		600
279	Please note that solaris ports can result in a lot of spurious	601	While this backend scales well, it requires one system call per active
280	notifications, so you need to use non-blocking I/O or other means to avoid	602	file descriptor per loop iteration. For small and medium numbers of file
281	blocking when no data (or space) is available.	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 returning 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
		616	you absolutely have to know whether an event occurred or not because you
		617	have 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>.
282		623
283	=item C<EVBACKEND_ALL>	624	=item C<EVBACKEND_ALL>
284		625
285	Try all backends (even potentially broken ones that wouldn't be tried	626	Try all backends (even potentially broken ones that wouldn't be tried
286	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as	627	with C<EVFLAG_AUTO>). Since this is a mask, you can do stuff such as
287	C<EVBACKEND_ALL & ~EVBACKEND_KQUEUE>.	628	C<EVBACKEND_ALL & ~EVBACKEND_KQUEUE>.
288		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).
		639
289	=back	640	=back
290		641
291	If one or more of these are ored into the flags value, then only these	642	If one or more of the backend flags are or'ed into the flags value,
292	backends will be tried (in the reverse order as given here). If none are	643	then only these backends will be tried (in the reverse order as listed
293	specified, most compiled-in backend will be tried, usually in reverse	644	here). If none are specified, all backends in C<ev_recommended_backends
294	order of their flag values :)	645	()> will be tried.
295		646
296	The most typical usage is like this:
297
298	if (!ev_default_loop (0))
299	fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?");
300
301	Restrict libev to the select and poll backends, and do not allow
302	environment settings to be taken into account:
303
304	ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV);
305
306	Use whatever libev has to offer, but make sure that kqueue is used if
307	available (warning, breaks stuff, best use only with your own private
308	event loop and only if you know the OS supports your types of fds):
309
310	ev_default_loop (ev_recommended_backends () | EVBACKEND_KQUEUE);
311
312	=item struct ev_loop *ev_loop_new (unsigned int flags)
313
314	Similar to C<ev_default_loop>, but always creates a new event loop that is
315	always distinct from the default loop. Unlike the default loop, it cannot
316	handle signal and child watchers, and attempts to do so will be greeted by
317	undefined behaviour (or a failed assertion if assertions are enabled).
318
319	Example: try to create a event loop that uses epoll and nothing else.	647	Example: Try to create a event loop that uses epoll and nothing else.
320		648
321	struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV);	649	struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV);
322	if (!epoller)	650	if (!epoller)
323	fatal ("no epoll found here, maybe it hides under your chair");	651	fatal ("no epoll found here, maybe it hides under your chair");
324		652
		653	Example: Use whatever libev has to offer, but make sure that kqueue is
		654	used if available.
		655
		656	struct ev_loop *loop = ev_loop_new (ev_recommended_backends () | EVBACKEND_KQUEUE);
		657
325	=item ev_default_destroy ()	658	=item ev_loop_destroy (loop)
326		659
327	Destroys the default loop again (frees all memory and kernel state	660	Destroys an event loop object (frees all memory and kernel state
328	etc.). None of the active event watchers will be stopped in the normal	661	etc.). None of the active event watchers will be stopped in the normal
329	sense, so e.g. C<ev_is_active> might still return true. It is your	662	sense, so e.g. C<ev_is_active> might still return true. It is your
330	responsibility to either stop all watchers cleanly yoursef I<before>	663	responsibility to either stop all watchers cleanly yourself I<before>
331	calling this function, or cope with the fact afterwards (which is usually	664	calling this function, or cope with the fact afterwards (which is usually
332	the easiest thing, youc na just ignore the watchers and/or C<free ()> them	665	the easiest thing, you can just ignore the watchers and/or C<free ()> them
333	for example).	666	for example).
334		667
		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.
		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
335	=item ev_loop_destroy (loop)	681	=item ev_loop_fork (loop)
336		682
337	Like C<ev_default_destroy>, but destroys an event loop created by an	683	This function sets a flag that causes subsequent C<ev_run> iterations to
338	earlier call to C<ev_loop_new>.
339
340	=item ev_default_fork ()
341
342	This function reinitialises the kernel state for backends that have	684	reinitialise the kernel state for backends that have one. Despite the
343	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
344	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
345	again makes little sense).	687	child before resuming or calling C<ev_run>.
346		688
347	You I<must> call this function in the child process after forking if and	689	Again, you I<have> to call it on I<any> loop that you want to re-use after
348	only if you want to use the event library in both processes. If you just	690	a fork, I<even if you do not plan to use the loop in the parent>. This is
349	fork+exec, you don't 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).
350		700
351	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
352	it just in case after a fork. To make this easy, the function will fit in	702	it just in case after a fork.
353	quite nicely into a call to C<pthread_atfork>:
354		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	...
355	pthread_atfork (0, 0, ev_default_fork);	714	pthread_atfork (0, 0, post_fork_child);
356		715
357	At the moment, C<EVBACKEND_SELECT> and C<EVBACKEND_POLL> are safe to use	716	=item int ev_is_default_loop (loop)
358	without calling this function, so if you force one of those backends you
359	do not need to care.
360		717
361	=item ev_loop_fork (loop)	718	Returns true when the given loop is, in fact, the default loop, and false
		719	otherwise.
362		720
363	Like C<ev_default_fork>, but acts on an event loop created by	721	=item unsigned int ev_iteration (loop)
364	C<ev_loop_new>. Yes, you have to call this on every allocated event loop	722
365	after fork, and how you do this is entirely your own problem.	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
		732	=item unsigned int ev_depth (loop)
		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.
366		745
367	=item unsigned int ev_backend (loop)	746	=item unsigned int ev_backend (loop)
368		747
369	Returns one of the C<EVBACKEND_*> flags indicating the event backend in	748	Returns one of the C<EVBACKEND_*> flags indicating the event backend in
370	use.	749	use.
…		…
373		752
374	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
375	received events and started processing them. This timestamp does not	754	received events and started processing them. This timestamp does not
376	change as long as callbacks are being processed, and this is also the base	755	change as long as callbacks are being processed, and this is also the base
377	time used for relative timers. You can treat it as the timestamp of the	756	time used for relative timers. You can treat it as the timestamp of the
378	event occuring (or more correctly, libev finding out about it).	757	event occurring (or more correctly, libev finding out about it).
379		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
380	=item ev_loop (loop, int flags)	797	=item ev_run (loop, int flags)
381		798
382	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
383	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
384	events.	801	handling events. It will ask the operating system for any new events, call
		802	the watcher callbacks, an then repeat the whole process indefinitely: This
		803	is why event loops are called I<loops>.
385		804
386	If the flags argument is specified as C<0>, it will not return until	805	If the flags argument is specified as C<0>, it will keep handling events
387	either 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.
388		808
389	Please note that an explicit C<ev_unloop> is usually better than	809	Please note that an explicit C<ev_break> is usually better than
390	relying on all watchers to be stopped when deciding when a program has	810	relying on all watchers to be stopped when deciding when a program has
391	finished (especially in interactive programs), but having a program that	811	finished (especially in interactive programs), but having a program
392	automatically loops as long as it has to and no longer by virtue of	812	that automatically loops as long as it has to and no longer by virtue
393	relying on its watchers stopping correctly is a thing of beauty.	813	of relying on its watchers stopping correctly, that is truly a thing of
		814	beauty.
394		815
		816	This function is also I<mostly> exception-safe - you can break out of
		817	a C<ev_run> call by calling C<longjmp> in a callback, throwing a C++
		818	exception and so on. This does not decrement the C<ev_depth> value, nor
		819	will it clear any outstanding C<EVBREAK_ONE> breaks.
		820
395	A flags value of C<EVLOOP_NONBLOCK> will look for new events, will handle	821	A flags value of C<EVRUN_NOWAIT> will look for new events, will handle
396	those events and any outstanding ones, but will not block your process in	822	those events and any already outstanding ones, but will not wait and
397	case there are no events and will return after one iteration of the loop.	823	block your process in case there are no events and will return after one
		824	iteration of the loop. This is sometimes useful to poll and handle new
		825	events while doing lengthy calculations, to keep the program responsive.
398		826
399	A flags value of C<EVLOOP_ONESHOT> will look for new events (waiting if	827	A flags value of C<EVRUN_ONCE> will look for new events (waiting if
400	neccessary) and will handle those and any outstanding ones. It will block	828	necessary) and will handle those and any already outstanding ones. It
401	your process until at least one new event arrives, and will return after	829	will block your process until at least one new event arrives (which could
402	one iteration of the loop. This is useful if you are waiting for some	830	be an event internal to libev itself, so there is no guarantee that a
403	external event in conjunction with something not expressible using other	831	user-registered callback will be called), and will return after one
		832	iteration of the loop.
		833
		834	This is useful if you are waiting for some external event in conjunction
		835	with something not expressible using other libev watchers (i.e. "roll your
404	libev watchers. However, a pair of C<ev_prepare>/C<ev_check> watchers is	836	own C<ev_run>"). However, a pair of C<ev_prepare>/C<ev_check> watchers is
405	usually a better approach for this kind of thing.	837	usually a better approach for this kind of thing.
406		838
407	Here are the gory details of what C<ev_loop> does:	839	Here are the gory details of what C<ev_run> does (this is for your
		840	understanding, not a guarantee that things will work exactly like this in
		841	future versions):
408		842
409	* If there are no active watchers (reference count is zero), return.	843	- Increment loop depth.
410	- Queue prepare watchers and then call all outstanding watchers.	844	- Reset the ev_break status.
		845	- Before the first iteration, call any pending watchers.
		846	LOOP:
		847	- If EVFLAG_FORKCHECK was used, check for a fork.
		848	- If a fork was detected (by any means), queue and call all fork watchers.
		849	- Queue and call all prepare watchers.
		850	- If ev_break was called, goto FINISH.
411	- If we have been forked, recreate the kernel state.	851	- If we have been forked, detach and recreate the kernel state
		852	as to not disturb the other process.
412	- Update the kernel state with all outstanding changes.	853	- Update the kernel state with all outstanding changes.
413	- Update the "event loop time".	854	- Update the "event loop time" (ev_now ()).
414	- Calculate for how long to block.	855	- Calculate for how long to sleep or block, if at all
		856	(active idle watchers, EVRUN_NOWAIT or not having
		857	any active watchers at all will result in not sleeping).
		858	- Sleep if the I/O and timer collect interval say so.
		859	- Increment loop iteration counter.
415	- Block the process, waiting for any events.	860	- Block the process, waiting for any events.
416	- Queue all outstanding I/O (fd) events.	861	- Queue all outstanding I/O (fd) events.
417	- Update the "event loop time" and do time jump handling.	862	- Update the "event loop time" (ev_now ()), and do time jump adjustments.
418	- Queue all outstanding timers.	863	- Queue all expired timers.
419	- Queue all outstanding periodics.	864	- Queue all expired periodics.
420	- If no events are pending now, queue all idle watchers.	865	- Queue all idle watchers with priority higher than that of pending events.
421	- Queue all check watchers.	866	- Queue all check watchers.
422	- Call all queued watchers in reverse order (i.e. check watchers first).	867	- Call all queued watchers in reverse order (i.e. check watchers first).
423	Signals and child watchers are implemented as I/O watchers, and will	868	Signals and child watchers are implemented as I/O watchers, and will
424	be handled here by queueing them when their watcher gets executed.	869	be handled here by queueing them when their watcher gets executed.
425	- If ev_unloop has been called or EVLOOP_ONESHOT or EVLOOP_NONBLOCK	870	- If ev_break has been called, or EVRUN_ONCE or EVRUN_NOWAIT
426	were used, return, otherwise continue with step *.	871	were used, or there are no active watchers, goto FINISH, otherwise
		872	continue with step LOOP.
		873	FINISH:
		874	- Reset the ev_break status iff it was EVBREAK_ONE.
		875	- Decrement the loop depth.
		876	- Return.
427		877
428	Example: queue some jobs and then loop until no events are outsanding	878	Example: Queue some jobs and then loop until no events are outstanding
429	anymore.	879	anymore.
430		880
431	... queue jobs here, make sure they register event watchers as long	881	... queue jobs here, make sure they register event watchers as long
432	... as they still have work to do (even an idle watcher will do..)	882	... as they still have work to do (even an idle watcher will do..)
433	ev_loop (my_loop, 0);	883	ev_run (my_loop, 0);
434	... jobs done. yeah!	884	... jobs done or somebody called break. yeah!
435		885
436	=item ev_unloop (loop, how)	886	=item ev_break (loop, how)
437		887
438	Can be used to make a call to C<ev_loop> return early (but only after it	888	Can be used to make a call to C<ev_run> return early (but only after it
439	has processed all outstanding events). The C<how> argument must be either	889	has processed all outstanding events). The C<how> argument must be either
440	C<EVUNLOOP_ONE>, which will make the innermost C<ev_loop> call return, or	890	C<EVBREAK_ONE>, which will make the innermost C<ev_run> call return, or
441	C<EVUNLOOP_ALL>, which will make all nested C<ev_loop> calls return.	891	C<EVBREAK_ALL>, which will make all nested C<ev_run> calls return.
		892
		893	This "break state" will be cleared on the next call to C<ev_run>.
		894
		895	It is safe to call C<ev_break> from outside any C<ev_run> calls, too, in
		896	which case it will have no effect.
442		897
443	=item ev_ref (loop)	898	=item ev_ref (loop)
444		899
445	=item ev_unref (loop)	900	=item ev_unref (loop)
446		901
447	Ref/unref can be used to add or remove a reference count on the event	902	Ref/unref can be used to add or remove a reference count on the event
448	loop: Every watcher keeps one reference, and as long as the reference	903	loop: Every watcher keeps one reference, and as long as the reference
449	count is nonzero, C<ev_loop> will not return on its own. If you have	904	count is nonzero, C<ev_run> will not return on its own.
450	a watcher you never unregister that should not keep C<ev_loop> from	905
451	returning, ev_unref() after starting, and ev_ref() before stopping it. For	906	This is useful when you have a watcher that you never intend to
		907	unregister, but that nevertheless should not keep C<ev_run> from
		908	returning. In such a case, call C<ev_unref> after starting, and C<ev_ref>
		909	before stopping it.
		910
452	example, libev itself uses this for its internal signal pipe: It is not	911	As an example, libev itself uses this for its internal signal pipe: It
453	visible to the libev user and should not keep C<ev_loop> from exiting if	912	is not visible to the libev user and should not keep C<ev_run> from
454	no event watchers registered by it are active. It is also an excellent	913	exiting if no event watchers registered by it are active. It is also an
455	way to do this for generic recurring timers or from within third-party	914	excellent way to do this for generic recurring timers or from within
456	libraries. Just remember to I<unref after start> and I<ref before stop>.	915	third-party libraries. Just remember to I<unref after start> and I<ref
		916	before stop> (but only if the watcher wasn't active before, or was active
		917	before, respectively. Note also that libev might stop watchers itself
		918	(e.g. non-repeating timers) in which case you have to C<ev_ref>
		919	in the callback).
457		920
458	Example: create a signal watcher, but keep it from keeping C<ev_loop>	921	Example: Create a signal watcher, but keep it from keeping C<ev_run>
459	running when nothing else is active.	922	running when nothing else is active.
460		923
461	struct dv_signal exitsig;	924	ev_signal exitsig;
462	ev_signal_init (&exitsig, sig_cb, SIGINT);	925	ev_signal_init (&exitsig, sig_cb, SIGINT);
463	ev_signal_start (myloop, &exitsig);	926	ev_signal_start (loop, &exitsig);
464	evf_unref (myloop);	927	ev_unref (loop);
465		928
466	Example: for some weird reason, unregister the above signal handler again.	929	Example: For some weird reason, unregister the above signal handler again.
467		930
468	ev_ref (myloop);	931	ev_ref (loop);
469	ev_signal_stop (myloop, &exitsig);	932	ev_signal_stop (loop, &exitsig);
		933
		934	=item ev_set_io_collect_interval (loop, ev_tstamp interval)
		935
		936	=item ev_set_timeout_collect_interval (loop, ev_tstamp interval)
		937
		938	These advanced functions influence the time that libev will spend waiting
		939	for events. Both time intervals are by default C<0>, meaning that libev
		940	will try to invoke timer/periodic callbacks and I/O callbacks with minimum
		941	latency.
		942
		943	Setting these to a higher value (the C<interval> I<must> be >= C<0>)
		944	allows libev to delay invocation of I/O and timer/periodic callbacks
		945	to increase efficiency of loop iterations (or to increase power-saving
		946	opportunities).
		947
		948	The idea is that sometimes your program runs just fast enough to handle
		949	one (or very few) event(s) per loop iteration. While this makes the
		950	program responsive, it also wastes a lot of CPU time to poll for new
		951	events, especially with backends like C<select ()> which have a high
		952	overhead for the actual polling but can deliver many events at once.
		953
		954	By setting a higher I<io collect interval> you allow libev to spend more
		955	time collecting I/O events, so you can handle more events per iteration,
		956	at the cost of increasing latency. Timeouts (both C<ev_periodic> and
		957	C<ev_timer>) will be not affected. Setting this to a non-null value will
		958	introduce an additional C<ev_sleep ()> call into most loop iterations. The
		959	sleep time ensures that libev will not poll for I/O events more often then
		960	once per this interval, on average.
		961
		962	Likewise, by setting a higher I<timeout collect interval> you allow libev
		963	to spend more time collecting timeouts, at the expense of increased
		964	latency/jitter/inexactness (the watcher callback will be called
		965	later). C<ev_io> watchers will not be affected. Setting this to a non-null
		966	value will not introduce any overhead in libev.
		967
		968	Many (busy) programs can usually benefit by setting the I/O collect
		969	interval to a value near C<0.1> or so, which is often enough for
		970	interactive servers (of course not for games), likewise for timeouts. It
		971	usually doesn't make much sense to set it to a lower value than C<0.01>,
		972	as this approaches the timing granularity of most systems. Note that if
		973	you do transactions with the outside world and you can't increase the
		974	parallelity, then this setting will limit your transaction rate (if you
		975	need to poll once per transaction and the I/O collect interval is 0.01,
		976	then you can't do more than 100 transactions per second).
		977
		978	Setting the I<timeout collect interval> can improve the opportunity for
		979	saving power, as the program will "bundle" timer callback invocations that
		980	are "near" in time together, by delaying some, thus reducing the number of
		981	times the process sleeps and wakes up again. Another useful technique to
		982	reduce iterations/wake-ups is to use C<ev_periodic> watchers and make sure
		983	they fire on, say, one-second boundaries only.
		984
		985	Example: we only need 0.1s timeout granularity, and we wish not to poll
		986	more often than 100 times per second:
		987
		988	ev_set_timeout_collect_interval (EV_DEFAULT_UC_ 0.1);
		989	ev_set_io_collect_interval (EV_DEFAULT_UC_ 0.01);
		990
		991	=item ev_invoke_pending (loop)
		992
		993	This call will simply invoke all pending watchers while resetting their
		994	pending state. Normally, C<ev_run> does this automatically when required,
		995	but when overriding the invoke callback this call comes handy. This
		996	function can be invoked from a watcher - this can be useful for example
		997	when you want to do some lengthy calculation and want to pass further
		998	event handling to another thread (you still have to make sure only one
		999	thread executes within C<ev_invoke_pending> or C<ev_run> of course).
		1000
		1001	=item int ev_pending_count (loop)
		1002
		1003	Returns the number of pending watchers - zero indicates that no watchers
		1004	are pending.
		1005
		1006	=item ev_set_invoke_pending_cb (loop, void (*invoke_pending_cb)(EV_P))
		1007
		1008	This overrides the invoke pending functionality of the loop: Instead of
		1009	invoking all pending watchers when there are any, C<ev_run> will call
		1010	this callback instead. This is useful, for example, when you want to
		1011	invoke the actual watchers inside another context (another thread etc.).
		1012
		1013	If you want to reset the callback, use C<ev_invoke_pending> as new
		1014	callback.
		1015
		1016	=item ev_set_loop_release_cb (loop, void (*release)(EV_P), void (*acquire)(EV_P))
		1017
		1018	Sometimes you want to share the same loop between multiple threads. This
		1019	can be done relatively simply by putting mutex_lock/unlock calls around
		1020	each call to a libev function.
		1021
		1022	However, C<ev_run> can run an indefinite time, so it is not feasible
		1023	to wait for it to return. One way around this is to wake up the event
		1024	loop via C<ev_break> and C<av_async_send>, another way is to set these
		1025	I<release> and I<acquire> callbacks on the loop.
		1026
		1027	When set, then C<release> will be called just before the thread is
		1028	suspended waiting for new events, and C<acquire> is called just
		1029	afterwards.
		1030
		1031	Ideally, C<release> will just call your mutex_unlock function, and
		1032	C<acquire> will just call the mutex_lock function again.
		1033
		1034	While event loop modifications are allowed between invocations of
		1035	C<release> and C<acquire> (that's their only purpose after all), no
		1036	modifications done will affect the event loop, i.e. adding watchers will
		1037	have no effect on the set of file descriptors being watched, or the time
		1038	waited. Use an C<ev_async> watcher to wake up C<ev_run> when you want it
		1039	to take note of any changes you made.
		1040
		1041	In theory, threads executing C<ev_run> will be async-cancel safe between
		1042	invocations of C<release> and C<acquire>.
		1043
		1044	See also the locking example in the C<THREADS> section later in this
		1045	document.
		1046
		1047	=item ev_set_userdata (loop, void *data)
		1048
		1049	=item void *ev_userdata (loop)
		1050
		1051	Set and retrieve a single C<void *> associated with a loop. When
		1052	C<ev_set_userdata> has never been called, then C<ev_userdata> returns
		1053	C<0>.
		1054
		1055	These two functions can be used to associate arbitrary data with a loop,
		1056	and are intended solely for the C<invoke_pending_cb>, C<release> and
		1057	C<acquire> callbacks described above, but of course can be (ab-)used for
		1058	any other purpose as well.
		1059
		1060	=item ev_verify (loop)
		1061
		1062	This function only does something when C<EV_VERIFY> support has been
		1063	compiled in, which is the default for non-minimal builds. It tries to go
		1064	through all internal structures and checks them for validity. If anything
		1065	is found to be inconsistent, it will print an error message to standard
		1066	error and call C<abort ()>.
		1067
		1068	This can be used to catch bugs inside libev itself: under normal
		1069	circumstances, this function will never abort as of course libev keeps its
		1070	data structures consistent.
470		1071
471	=back	1072	=back
472		1073
473		1074
474	=head1 ANATOMY OF A WATCHER	1075	=head1 ANATOMY OF A WATCHER
475		1076
		1077	In the following description, uppercase C<TYPE> in names stands for the
		1078	watcher type, e.g. C<ev_TYPE_start> can mean C<ev_timer_start> for timer
		1079	watchers and C<ev_io_start> for I/O watchers.
		1080
476	A watcher is a structure that you create and register to record your	1081	A watcher is an opaque structure that you allocate and register to record
477	interest in some event. For instance, if you want to wait for STDIN to	1082	your interest in some event. To make a concrete example, imagine you want
478	become readable, you would create an C<ev_io> watcher for that:	1083	to wait for STDIN to become readable, you would create an C<ev_io> watcher
		1084	for that:
479		1085
480	static void my_cb (struct ev_loop *loop, struct ev_io *w, int revents)	1086	static void my_cb (struct ev_loop *loop, ev_io *w, int revents)
481	{	1087	{
482	ev_io_stop (w);	1088	ev_io_stop (w);
483	ev_unloop (loop, EVUNLOOP_ALL);	1089	ev_break (loop, EVBREAK_ALL);
484	}	1090	}
485		1091
486	struct ev_loop *loop = ev_default_loop (0);	1092	struct ev_loop *loop = ev_default_loop (0);
		1093
487	struct ev_io stdin_watcher;	1094	ev_io stdin_watcher;
		1095
488	ev_init (&stdin_watcher, my_cb);	1096	ev_init (&stdin_watcher, my_cb);
489	ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ);	1097	ev_io_set (&stdin_watcher, STDIN_FILENO, EV_READ);
490	ev_io_start (loop, &stdin_watcher);	1098	ev_io_start (loop, &stdin_watcher);
		1099
491	ev_loop (loop, 0);	1100	ev_run (loop, 0);
492		1101
493	As you can see, you are responsible for allocating the memory for your	1102	As you can see, you are responsible for allocating the memory for your
494	watcher structures (and it is usually a bad idea to do this on the stack,	1103	watcher structures (and it is I<usually> a bad idea to do this on the
495	although this can sometimes be quite valid).	1104	stack).
496		1105
		1106	Each watcher has an associated watcher structure (called C<struct ev_TYPE>
		1107	or simply C<ev_TYPE>, as typedefs are provided for all watcher structs).
		1108
497	Each watcher structure must be initialised by a call to C<ev_init	1109	Each watcher structure must be initialised by a call to C<ev_init (watcher
498	(watcher *, callback)>, which expects a callback to be provided. This	1110	*, callback)>, which expects a callback to be provided. This callback is
499	callback gets invoked each time the event occurs (or, in the case of io	1111	invoked each time the event occurs (or, in the case of I/O watchers, each
500	watchers, each time the event loop detects that the file descriptor given	1112	time the event loop detects that the file descriptor given is readable
501	is readable and/or writable).	1113	and/or writable).
502		1114
503	Each watcher type has its own C<< ev_<type>_set (watcher *, ...) >> macro	1115	Each watcher type further has its own C<< ev_TYPE_set (watcher *, ...) >>
504	with arguments specific to this watcher type. There is also a macro	1116	macro to configure it, with arguments specific to the watcher type. There
505	to combine initialisation and setting in one call: C<< ev_<type>_init	1117	is also a macro to combine initialisation and setting in one call: C<<
506	(watcher *, callback, ...) >>.	1118	ev_TYPE_init (watcher *, callback, ...) >>.
507		1119
508	To make the watcher actually watch out for events, you have to start it	1120	To make the watcher actually watch out for events, you have to start it
509	with a watcher-specific start function (C<< ev_<type>_start (loop, watcher	1121	with a watcher-specific start function (C<< ev_TYPE_start (loop, watcher
510	*) >>), and you can stop watching for events at any time by calling the	1122	*) >>), and you can stop watching for events at any time by calling the
511	corresponding stop function (C<< ev_<type>_stop (loop, watcher *) >>.	1123	corresponding stop function (C<< ev_TYPE_stop (loop, watcher *) >>.
512		1124
513	As long as your watcher is active (has been started but not stopped) you	1125	As long as your watcher is active (has been started but not stopped) you
514	must not touch the values stored in it. Most specifically you must never	1126	must not touch the values stored in it. Most specifically you must never
515	reinitialise it or call its C<set> macro.	1127	reinitialise it or call its C<ev_TYPE_set> macro.
516		1128
517	Each and every callback receives the event loop pointer as first, the	1129	Each and every callback receives the event loop pointer as first, the
518	registered watcher structure as second, and a bitset of received events as	1130	registered watcher structure as second, and a bitset of received events as
519	third argument.	1131	third argument.
520		1132
…		…
529	=item C<EV_WRITE>	1141	=item C<EV_WRITE>
530		1142
531	The file descriptor in the C<ev_io> watcher has become readable and/or	1143	The file descriptor in the C<ev_io> watcher has become readable and/or
532	writable.	1144	writable.
533		1145
534	=item C<EV_TIMEOUT>	1146	=item C<EV_TIMER>
535		1147
536	The C<ev_timer> watcher has timed out.	1148	The C<ev_timer> watcher has timed out.
537		1149
538	=item C<EV_PERIODIC>	1150	=item C<EV_PERIODIC>
539		1151
…		…
545		1157
546	=item C<EV_CHILD>	1158	=item C<EV_CHILD>
547		1159
548	The pid specified in the C<ev_child> watcher has received a status change.	1160	The pid specified in the C<ev_child> watcher has received a status change.
549		1161
		1162	=item C<EV_STAT>
		1163
		1164	The path specified in the C<ev_stat> watcher changed its attributes somehow.
		1165
550	=item C<EV_IDLE>	1166	=item C<EV_IDLE>
551		1167
552	The C<ev_idle> watcher has determined that you have nothing better to do.	1168	The C<ev_idle> watcher has determined that you have nothing better to do.
553		1169
554	=item C<EV_PREPARE>	1170	=item C<EV_PREPARE>
555		1171
556	=item C<EV_CHECK>	1172	=item C<EV_CHECK>
557		1173
558	All C<ev_prepare> watchers are invoked just I<before> C<ev_loop> starts	1174	All C<ev_prepare> watchers are invoked just I<before> C<ev_run> starts
559	to gather new events, and all C<ev_check> watchers are invoked just after	1175	to gather new events, and all C<ev_check> watchers are invoked just after
560	C<ev_loop> has gathered them, but before it invokes any callbacks for any	1176	C<ev_run> has gathered them, but before it invokes any callbacks for any
561	received events. Callbacks of both watcher types can start and stop as	1177	received events. Callbacks of both watcher types can start and stop as
562	many watchers as they want, and all of them will be taken into account	1178	many watchers as they want, and all of them will be taken into account
563	(for example, a C<ev_prepare> watcher might start an idle watcher to keep	1179	(for example, a C<ev_prepare> watcher might start an idle watcher to keep
564	C<ev_loop> from blocking).	1180	C<ev_run> from blocking).
		1181
		1182	=item C<EV_EMBED>
		1183
		1184	The embedded event loop specified in the C<ev_embed> watcher needs attention.
		1185
		1186	=item C<EV_FORK>
		1187
		1188	The event loop has been resumed in the child process after fork (see
		1189	C<ev_fork>).
		1190
		1191	=item C<EV_CLEANUP>
		1192
		1193	The event loop is about to be destroyed (see C<ev_cleanup>).
		1194
		1195	=item C<EV_ASYNC>
		1196
		1197	The given async watcher has been asynchronously notified (see C<ev_async>).
		1198
		1199	=item C<EV_CUSTOM>
		1200
		1201	Not ever sent (or otherwise used) by libev itself, but can be freely used
		1202	by libev users to signal watchers (e.g. via C<ev_feed_event>).
565		1203
566	=item C<EV_ERROR>	1204	=item C<EV_ERROR>
567		1205
568	An unspecified error has occured, the watcher has been stopped. This might	1206	An unspecified error has occurred, the watcher has been stopped. This might
569	happen because the watcher could not be properly started because libev	1207	happen because the watcher could not be properly started because libev
570	ran out of memory, a file descriptor was found to be closed or any other	1208	ran out of memory, a file descriptor was found to be closed or any other
		1209	problem. Libev considers these application bugs.
		1210
571	problem. You best act on it by reporting the problem and somehow coping	1211	You best act on it by reporting the problem and somehow coping with the
572	with the watcher being stopped.	1212	watcher being stopped. Note that well-written programs should not receive
		1213	an error ever, so when your watcher receives it, this usually indicates a
		1214	bug in your program.
573		1215
574	Libev will usually signal a few "dummy" events together with an error,	1216	Libev will usually signal a few "dummy" events together with an error, for
575	for example it might indicate that a fd is readable or writable, and if	1217	example it might indicate that a fd is readable or writable, and if your
576	your callbacks is well-written it can just attempt the operation and cope	1218	callbacks is well-written it can just attempt the operation and cope with
577	with the error from read() or write(). This will not work in multithreaded	1219	the error from read() or write(). This will not work in multi-threaded
578	programs, though, so beware.	1220	programs, though, as the fd could already be closed and reused for another
		1221	thing, so beware.
579		1222
580	=back	1223	=back
581		1224
582	=head2 GENERIC WATCHER FUNCTIONS	1225	=head2 GENERIC WATCHER FUNCTIONS
583
584	In the following description, C<TYPE> stands for the watcher type,
585	e.g. C<timer> for C<ev_timer> watchers and C<io> for C<ev_io> watchers.
586		1226
587	=over 4	1227	=over 4
588		1228
589	=item C<ev_init> (ev_TYPE *watcher, callback)	1229	=item C<ev_init> (ev_TYPE *watcher, callback)
590		1230
…		…
596	which rolls both calls into one.	1236	which rolls both calls into one.
597		1237
598	You can reinitialise a watcher at any time as long as it has been stopped	1238	You can reinitialise a watcher at any time as long as it has been stopped
599	(or never started) and there are no pending events outstanding.	1239	(or never started) and there are no pending events outstanding.
600		1240
601	The callback is always of type C<void (*)(ev_loop *loop, ev_TYPE *watcher,	1241	The callback is always of type C<void (*)(struct ev_loop *loop, ev_TYPE *watcher,
602	int revents)>.	1242	int revents)>.
603		1243
		1244	Example: Initialise an C<ev_io> watcher in two steps.
		1245
		1246	ev_io w;
		1247	ev_init (&w, my_cb);
		1248	ev_io_set (&w, STDIN_FILENO, EV_READ);
		1249
604	=item C<ev_TYPE_set> (ev_TYPE *, [args])	1250	=item C<ev_TYPE_set> (ev_TYPE *watcher, [args])
605		1251
606	This macro initialises the type-specific parts of a watcher. You need to	1252	This macro initialises the type-specific parts of a watcher. You need to
607	call C<ev_init> at least once before you call this macro, but you can	1253	call C<ev_init> at least once before you call this macro, but you can
608	call C<ev_TYPE_set> any number of times. You must not, however, call this	1254	call C<ev_TYPE_set> any number of times. You must not, however, call this
609	macro on a watcher that is active (it can be pending, however, which is a	1255	macro on a watcher that is active (it can be pending, however, which is a
610	difference to the C<ev_init> macro).	1256	difference to the C<ev_init> macro).
611		1257
612	Although some watcher types do not have type-specific arguments	1258	Although some watcher types do not have type-specific arguments
613	(e.g. C<ev_prepare>) you still need to call its C<set> macro.	1259	(e.g. C<ev_prepare>) you still need to call its C<set> macro.
614		1260
		1261	See C<ev_init>, above, for an example.
		1262
615	=item C<ev_TYPE_init> (ev_TYPE *watcher, callback, [args])	1263	=item C<ev_TYPE_init> (ev_TYPE *watcher, callback, [args])
616		1264
617	This convinience macro rolls both C<ev_init> and C<ev_TYPE_set> macro	1265	This convenience macro rolls both C<ev_init> and C<ev_TYPE_set> macro
618	calls into a single call. This is the most convinient method to initialise	1266	calls into a single call. This is the most convenient method to initialise
619	a watcher. The same limitations apply, of course.	1267	a watcher. The same limitations apply, of course.
620		1268
		1269	Example: Initialise and set an C<ev_io> watcher in one step.
		1270
		1271	ev_io_init (&w, my_cb, STDIN_FILENO, EV_READ);
		1272
621	=item C<ev_TYPE_start> (loop *, ev_TYPE *watcher)	1273	=item C<ev_TYPE_start> (loop, ev_TYPE *watcher)
622		1274
623	Starts (activates) the given watcher. Only active watchers will receive	1275	Starts (activates) the given watcher. Only active watchers will receive
624	events. If the watcher is already active nothing will happen.	1276	events. If the watcher is already active nothing will happen.
625		1277
		1278	Example: Start the C<ev_io> watcher that is being abused as example in this
		1279	whole section.
		1280
		1281	ev_io_start (EV_DEFAULT_UC, &w);
		1282
626	=item C<ev_TYPE_stop> (loop *, ev_TYPE *watcher)	1283	=item C<ev_TYPE_stop> (loop, ev_TYPE *watcher)
627		1284
628	Stops the given watcher again (if active) and clears the pending	1285	Stops the given watcher if active, and clears the pending status (whether
		1286	the watcher was active or not).
		1287
629	status. It is possible that stopped watchers are pending (for example,	1288	It is possible that stopped watchers are pending - for example,
630	non-repeating timers are being stopped when they become pending), but	1289	non-repeating timers are being stopped when they become pending - but
631	C<ev_TYPE_stop> ensures that the watcher is neither active nor pending. If	1290	calling C<ev_TYPE_stop> ensures that the watcher is neither active nor
632	you want to free or reuse the memory used by the watcher it is therefore a	1291	pending. If you want to free or reuse the memory used by the watcher it is
633	good idea to always call its C<ev_TYPE_stop> function.	1292	therefore a good idea to always call its C<ev_TYPE_stop> function.
634		1293
635	=item bool ev_is_active (ev_TYPE *watcher)	1294	=item bool ev_is_active (ev_TYPE *watcher)
636		1295
637	Returns a true value iff the watcher is active (i.e. it has been started	1296	Returns a true value iff the watcher is active (i.e. it has been started
638	and not yet been stopped). As long as a watcher is active you must not modify	1297	and not yet been stopped). As long as a watcher is active you must not modify
…		…
641	=item bool ev_is_pending (ev_TYPE *watcher)	1300	=item bool ev_is_pending (ev_TYPE *watcher)
642		1301
643	Returns a true value iff the watcher is pending, (i.e. it has outstanding	1302	Returns a true value iff the watcher is pending, (i.e. it has outstanding
644	events but its callback has not yet been invoked). As long as a watcher	1303	events but its callback has not yet been invoked). As long as a watcher
645	is pending (but not active) you must not call an init function on it (but	1304	is pending (but not active) you must not call an init function on it (but
646	C<ev_TYPE_set> is safe) and you must make sure the watcher is available to	1305	C<ev_TYPE_set> is safe), you must not change its priority, and you must
647	libev (e.g. you cnanot C<free ()> it).	1306	make sure the watcher is available to libev (e.g. you cannot C<free ()>
		1307	it).
648		1308
649	=item callback = ev_cb (ev_TYPE *watcher)	1309	=item callback ev_cb (ev_TYPE *watcher)
650		1310
651	Returns the callback currently set on the watcher.	1311	Returns the callback currently set on the watcher.
652		1312
653	=item ev_cb_set (ev_TYPE *watcher, callback)	1313	=item ev_cb_set (ev_TYPE *watcher, callback)
654		1314
655	Change the callback. You can change the callback at virtually any time	1315	Change the callback. You can change the callback at virtually any time
656	(modulo threads).	1316	(modulo threads).
657		1317
		1318	=item ev_set_priority (ev_TYPE *watcher, int priority)
		1319
		1320	=item int ev_priority (ev_TYPE *watcher)
		1321
		1322	Set and query the priority of the watcher. The priority is a small
		1323	integer between C<EV_MAXPRI> (default: C<2>) and C<EV_MINPRI>
		1324	(default: C<-2>). Pending watchers with higher priority will be invoked
		1325	before watchers with lower priority, but priority will not keep watchers
		1326	from being executed (except for C<ev_idle> watchers).
		1327
		1328	If you need to suppress invocation when higher priority events are pending
		1329	you need to look at C<ev_idle> watchers, which provide this functionality.
		1330
		1331	You I<must not> change the priority of a watcher as long as it is active or
		1332	pending.
		1333
		1334	Setting a priority outside the range of C<EV_MINPRI> to C<EV_MAXPRI> is
		1335	fine, as long as you do not mind that the priority value you query might
		1336	or might not have been clamped to the valid range.
		1337
		1338	The default priority used by watchers when no priority has been set is
		1339	always C<0>, which is supposed to not be too high and not be too low :).
		1340
		1341	See L<WATCHER PRIORITY MODELS>, below, for a more thorough treatment of
		1342	priorities.
		1343
		1344	=item ev_invoke (loop, ev_TYPE *watcher, int revents)
		1345
		1346	Invoke the C<watcher> with the given C<loop> and C<revents>. Neither
		1347	C<loop> nor C<revents> need to be valid as long as the watcher callback
		1348	can deal with that fact, as both are simply passed through to the
		1349	callback.
		1350
		1351	=item int ev_clear_pending (loop, ev_TYPE *watcher)
		1352
		1353	If the watcher is pending, this function clears its pending status and
		1354	returns its C<revents> bitset (as if its callback was invoked). If the
		1355	watcher isn't pending it does nothing and returns C<0>.
		1356
		1357	Sometimes it can be useful to "poll" a watcher instead of waiting for its
		1358	callback to be invoked, which can be accomplished with this function.
		1359
		1360	=item ev_feed_event (loop, ev_TYPE *watcher, int revents)
		1361
		1362	Feeds the given event set into the event loop, as if the specified event
		1363	had happened for the specified watcher (which must be a pointer to an
		1364	initialised but not necessarily started event watcher). Obviously you must
		1365	not free the watcher as long as it has pending events.
		1366
		1367	Stopping the watcher, letting libev invoke it, or calling
		1368	C<ev_clear_pending> will clear the pending event, even if the watcher was
		1369	not started in the first place.
		1370
		1371	See also C<ev_feed_fd_event> and C<ev_feed_signal_event> for related
		1372	functions that do not need a watcher.
		1373
658	=back	1374	=back
659		1375
		1376	See also the L<ASSOCIATING CUSTOM DATA WITH A WATCHER> and L<BUILDING YOUR
		1377	OWN COMPOSITE WATCHERS> idioms.
660		1378
661	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER	1379	=head2 WATCHER STATES
662		1380
663	Each watcher has, by default, a member C<void *data> that you can change	1381	There are various watcher states mentioned throughout this manual -
664	and read at any time, libev will completely ignore it. This can be used	1382	active, pending and so on. In this section these states and the rules to
665	to associate arbitrary data with your watcher. If you need more data and	1383	transition between them will be described in more detail - and while these
666	don't want to allocate memory and store a pointer to it in that data	1384	rules might look complicated, they usually do "the right thing".
667	member, you can also "subclass" the watcher type and provide your own
668	data:
669		1385
670	struct my_io	1386	=over 4
		1387
		1388	=item initialiased
		1389
		1390	Before a watcher can be registered with the event looop it has to be
		1391	initialised. This can be done with a call to C<ev_TYPE_init>, or calls to
		1392	C<ev_init> followed by the watcher-specific C<ev_TYPE_set> function.
		1393
		1394	In this state it is simply some block of memory that is suitable for
		1395	use in an event loop. It can be moved around, freed, reused etc. at
		1396	will - as long as you either keep the memory contents intact, or call
		1397	C<ev_TYPE_init> again.
		1398
		1399	=item started/running/active
		1400
		1401	Once a watcher has been started with a call to C<ev_TYPE_start> it becomes
		1402	property of the event loop, and is actively waiting for events. While in
		1403	this state it cannot be accessed (except in a few documented ways), moved,
		1404	freed or anything else - the only legal thing is to keep a pointer to it,
		1405	and call libev functions on it that are documented to work on active watchers.
		1406
		1407	=item pending
		1408
		1409	If a watcher is active and libev determines that an event it is interested
		1410	in has occurred (such as a timer expiring), it will become pending. It will
		1411	stay in this pending state until either it is stopped or its callback is
		1412	about to be invoked, so it is not normally pending inside the watcher
		1413	callback.
		1414
		1415	The watcher might or might not be active while it is pending (for example,
		1416	an expired non-repeating timer can be pending but no longer active). If it
		1417	is stopped, it can be freely accessed (e.g. by calling C<ev_TYPE_set>),
		1418	but it is still property of the event loop at this time, so cannot be
		1419	moved, freed or reused. And if it is active the rules described in the
		1420	previous item still apply.
		1421
		1422	It is also possible to feed an event on a watcher that is not active (e.g.
		1423	via C<ev_feed_event>), in which case it becomes pending without being
		1424	active.
		1425
		1426	=item stopped
		1427
		1428	A watcher can be stopped implicitly by libev (in which case it might still
		1429	be pending), or explicitly by calling its C<ev_TYPE_stop> function. The
		1430	latter will clear any pending state the watcher might be in, regardless
		1431	of whether it was active or not, so stopping a watcher explicitly before
		1432	freeing it is often a good idea.
		1433
		1434	While stopped (and not pending) the watcher is essentially in the
		1435	initialised state, that is, it can be reused, moved, modified in any way
		1436	you wish (but when you trash the memory block, you need to C<ev_TYPE_init>
		1437	it again).
		1438
		1439	=back
		1440
		1441	=head2 WATCHER PRIORITY MODELS
		1442
		1443	Many event loops support I<watcher priorities>, which are usually small
		1444	integers that influence the ordering of event callback invocation
		1445	between watchers in some way, all else being equal.
		1446
		1447	In libev, Watcher priorities can be set using C<ev_set_priority>. See its
		1448	description for the more technical details such as the actual priority
		1449	range.
		1450
		1451	There are two common ways how these these priorities are being interpreted
		1452	by event loops:
		1453
		1454	In the more common lock-out model, higher priorities "lock out" invocation
		1455	of lower priority watchers, which means as long as higher priority
		1456	watchers receive events, lower priority watchers are not being invoked.
		1457
		1458	The less common only-for-ordering model uses priorities solely to order
		1459	callback invocation within a single event loop iteration: Higher priority
		1460	watchers are invoked before lower priority ones, but they all get invoked
		1461	before polling for new events.
		1462
		1463	Libev uses the second (only-for-ordering) model for all its watchers
		1464	except for idle watchers (which use the lock-out model).
		1465
		1466	The rationale behind this is that implementing the lock-out model for
		1467	watchers is not well supported by most kernel interfaces, and most event
		1468	libraries will just poll for the same events again and again as long as
		1469	their callbacks have not been executed, which is very inefficient in the
		1470	common case of one high-priority watcher locking out a mass of lower
		1471	priority ones.
		1472
		1473	Static (ordering) priorities are most useful when you have two or more
		1474	watchers handling the same resource: a typical usage example is having an
		1475	C<ev_io> watcher to receive data, and an associated C<ev_timer> to handle
		1476	timeouts. Under load, data might be received while the program handles
		1477	other jobs, but since timers normally get invoked first, the timeout
		1478	handler will be executed before checking for data. In that case, giving
		1479	the timer a lower priority than the I/O watcher ensures that I/O will be
		1480	handled first even under adverse conditions (which is usually, but not
		1481	always, what you want).
		1482
		1483	Since idle watchers use the "lock-out" model, meaning that idle watchers
		1484	will only be executed when no same or higher priority watchers have
		1485	received events, they can be used to implement the "lock-out" model when
		1486	required.
		1487
		1488	For example, to emulate how many other event libraries handle priorities,
		1489	you can associate an C<ev_idle> watcher to each such watcher, and in
		1490	the normal watcher callback, you just start the idle watcher. The real
		1491	processing is done in the idle watcher callback. This causes libev to
		1492	continuously poll and process kernel event data for the watcher, but when
		1493	the lock-out case is known to be rare (which in turn is rare :), this is
		1494	workable.
		1495
		1496	Usually, however, the lock-out model implemented that way will perform
		1497	miserably under the type of load it was designed to handle. In that case,
		1498	it might be preferable to stop the real watcher before starting the
		1499	idle watcher, so the kernel will not have to process the event in case
		1500	the actual processing will be delayed for considerable time.
		1501
		1502	Here is an example of an I/O watcher that should run at a strictly lower
		1503	priority than the default, and which should only process data when no
		1504	other events are pending:
		1505
		1506	ev_idle idle; // actual processing watcher
		1507	ev_io io; // actual event watcher
		1508
		1509	static void
		1510	io_cb (EV_P_ ev_io *w, int revents)
671	{	1511	{
672	struct ev_io io;	1512	// stop the I/O watcher, we received the event, but
673	int otherfd;	1513	// are not yet ready to handle it.
674	void *somedata;	1514	ev_io_stop (EV_A_ w);
675	struct whatever *mostinteresting;	1515
		1516	// start the idle watcher to handle the actual event.
		1517	// it will not be executed as long as other watchers
		1518	// with the default priority are receiving events.
		1519	ev_idle_start (EV_A_ &idle);
676	}	1520	}
677		1521
678	And since your callback will be called with a pointer to the watcher, you	1522	static void
679	can cast it back to your own type:	1523	idle_cb (EV_P_ ev_idle *w, int revents)
680
681	static void my_cb (struct ev_loop *loop, struct ev_io *w_, int revents)
682	{	1524	{
683	struct my_io *w = (struct my_io *)w_;	1525	// actual processing
684	...	1526	read (STDIN_FILENO, ...);
		1527
		1528	// have to start the I/O watcher again, as
		1529	// we have handled the event
		1530	ev_io_start (EV_P_ &io);
685	}	1531	}
686		1532
687	More interesting and less C-conformant ways of catsing your callback type	1533	// initialisation
688	have been omitted....	1534	ev_idle_init (&idle, idle_cb);
		1535	ev_io_init (&io, io_cb, STDIN_FILENO, EV_READ);
		1536	ev_io_start (EV_DEFAULT_ &io);
		1537
		1538	In the "real" world, it might also be beneficial to start a timer, so that
		1539	low-priority connections can not be locked out forever under load. This
		1540	enables your program to keep a lower latency for important connections
		1541	during short periods of high load, while not completely locking out less
		1542	important ones.
689		1543
690		1544
691	=head1 WATCHER TYPES	1545	=head1 WATCHER TYPES
692		1546
693	This section describes each watcher in detail, but will not repeat	1547	This section describes each watcher in detail, but will not repeat
694	information given in the last section.	1548	information given in the last section. Any initialisation/set macros,
		1549	functions and members specific to the watcher type are explained.
		1550
		1551	Members are additionally marked with either I<[read-only]>, meaning that,
		1552	while the watcher is active, you can look at the member and expect some
		1553	sensible content, but you must not modify it (you can modify it while the
		1554	watcher is stopped to your hearts content), or I<[read-write]>, which
		1555	means you can expect it to have some sensible content while the watcher
		1556	is active, but you can also modify it. Modifying it may not do something
		1557	sensible or take immediate effect (or do anything at all), but libev will
		1558	not crash or malfunction in any way.
695		1559
696		1560
697	=head2 C<ev_io> - is this file descriptor readable or writable?	1561	=head2 C<ev_io> - is this file descriptor readable or writable?
698		1562
699	I/O watchers check whether a file descriptor is readable or writable	1563	I/O watchers check whether a file descriptor is readable or writable
…		…
707	In general you can register as many read and/or write event watchers per	1571	In general you can register as many read and/or write event watchers per
708	fd as you want (as long as you don't confuse yourself). Setting all file	1572	fd as you want (as long as you don't confuse yourself). Setting all file
709	descriptors to non-blocking mode is also usually a good idea (but not	1573	descriptors to non-blocking mode is also usually a good idea (but not
710	required if you know what you are doing).	1574	required if you know what you are doing).
711		1575
712	You have to be careful with dup'ed file descriptors, though. Some backends
713	(the linux epoll backend is a notable example) cannot handle dup'ed file
714	descriptors correctly if you register interest in two or more fds pointing
715	to the same underlying file/socket/etc. description (that is, they share
716	the same underlying "file open").
717
718	If you must do this, then force the use of a known-to-be-good backend
719	(at the time of this writing, this includes only C<EVBACKEND_SELECT> and
720	C<EVBACKEND_POLL>).
721
722	Another thing you have to watch out for is that it is quite easy to	1576	Another thing you have to watch out for is that it is quite easy to
723	receive "spurious" readyness notifications, that is your callback might	1577	receive "spurious" readiness notifications, that is, your callback might
724	be called with C<EV_READ> but a subsequent C<read>(2) will actually block	1578	be called with C<EV_READ> but a subsequent C<read>(2) will actually block
725	because there is no data. Not only are some backends known to create a	1579	because there is no data. It is very easy to get into this situation even
726	lot of those (for example solaris ports), it is very easy to get into	1580	with a relatively standard program structure. Thus it is best to always
727	this situation even with a relatively standard program structure. Thus	1581	use non-blocking I/O: An extra C<read>(2) returning C<EAGAIN> is far
728	it is best to always use non-blocking I/O: An extra C<read>(2) returning
729	C<EAGAIN> is far preferable to a program hanging until some data arrives.	1582	preferable to a program hanging until some data arrives.
730		1583
731	If you cannot run the fd in non-blocking mode (for example you should not	1584	If you cannot run the fd in non-blocking mode (for example you should
732	play around with an Xlib connection), then you have to seperately re-test	1585	not play around with an Xlib connection), then you have to separately
733	wether a file descriptor is really ready with a known-to-be good interface	1586	re-test whether a file descriptor is really ready with a known-to-be good
734	such as poll (fortunately in our Xlib example, Xlib already does this on	1587	interface such as poll (fortunately in the case of Xlib, it already does
735	its own, so its quite safe to use).	1588	this on its own, so its quite safe to use). Some people additionally
		1589	use C<SIGALRM> and an interval timer, just to be sure you won't block
		1590	indefinitely.
		1591
		1592	But really, best use non-blocking mode.
		1593
		1594	=head3 The special problem of disappearing file descriptors
		1595
		1596	Some backends (e.g. kqueue, epoll) need to be told about closing a file
		1597	descriptor (either due to calling C<close> explicitly or any other means,
		1598	such as C<dup2>). The reason is that you register interest in some file
		1599	descriptor, but when it goes away, the operating system will silently drop
		1600	this interest. If another file descriptor with the same number then is
		1601	registered with libev, there is no efficient way to see that this is, in
		1602	fact, a different file descriptor.
		1603
		1604	To avoid having to explicitly tell libev about such cases, libev follows
		1605	the following policy: Each time C<ev_io_set> is being called, libev
		1606	will assume that this is potentially a new file descriptor, otherwise
		1607	it is assumed that the file descriptor stays the same. That means that
		1608	you I<have> to call C<ev_io_set> (or C<ev_io_init>) when you change the
		1609	descriptor even if the file descriptor number itself did not change.
		1610
		1611	This is how one would do it normally anyway, the important point is that
		1612	the libev application should not optimise around libev but should leave
		1613	optimisations to libev.
		1614
		1615	=head3 The special problem of dup'ed file descriptors
		1616
		1617	Some backends (e.g. epoll), cannot register events for file descriptors,
		1618	but only events for the underlying file descriptions. That means when you
		1619	have C<dup ()>'ed file descriptors or weirder constellations, and register
		1620	events for them, only one file descriptor might actually receive events.
		1621
		1622	There is no workaround possible except not registering events
		1623	for potentially C<dup ()>'ed file descriptors, or to resort to
		1624	C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>.
		1625
		1626	=head3 The special problem of files
		1627
		1628	Many people try to use C<select> (or libev) on file descriptors
		1629	representing files, and expect it to become ready when their program
		1630	doesn't block on disk accesses (which can take a long time on their own).
		1631
		1632	However, this cannot ever work in the "expected" way - you get a readiness
		1633	notification as soon as the kernel knows whether and how much data is
		1634	there, and in the case of open files, that's always the case, so you
		1635	always get a readiness notification instantly, and your read (or possibly
		1636	write) will still block on the disk I/O.
		1637
		1638	Another way to view it is that in the case of sockets, pipes, character
		1639	devices and so on, there is another party (the sender) that delivers data
		1640	on its own, but in the case of files, there is no such thing: the disk
		1641	will not send data on its own, simply because it doesn't know what you
		1642	wish to read - you would first have to request some data.
		1643
		1644	Since files are typically not-so-well supported by advanced notification
		1645	mechanism, libev tries hard to emulate POSIX behaviour with respect
		1646	to files, even though you should not use it. The reason for this is
		1647	convenience: sometimes you want to watch STDIN or STDOUT, which is
		1648	usually a tty, often a pipe, but also sometimes files or special devices
		1649	(for example, C<epoll> on Linux works with F</dev/random> but not with
		1650	F</dev/urandom>), and even though the file might better be served with
		1651	asynchronous I/O instead of with non-blocking I/O, it is still useful when
		1652	it "just works" instead of freezing.
		1653
		1654	So avoid file descriptors pointing to files when you know it (e.g. use
		1655	libeio), but use them when it is convenient, e.g. for STDIN/STDOUT, or
		1656	when you rarely read from a file instead of from a socket, and want to
		1657	reuse the same code path.
		1658
		1659	=head3 The special problem of fork
		1660
		1661	Some backends (epoll, kqueue) do not support C<fork ()> at all or exhibit
		1662	useless behaviour. Libev fully supports fork, but needs to be told about
		1663	it in the child if you want to continue to use it in the child.
		1664
		1665	To support fork in your child processes, you have to call C<ev_loop_fork
		1666	()> after a fork in the child, enable C<EVFLAG_FORKCHECK>, or resort to
		1667	C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>.
		1668
		1669	=head3 The special problem of SIGPIPE
		1670
		1671	While not really specific to libev, it is easy to forget about C<SIGPIPE>:
		1672	when writing to a pipe whose other end has been closed, your program gets
		1673	sent a SIGPIPE, which, by default, aborts your program. For most programs
		1674	this is sensible behaviour, for daemons, this is usually undesirable.
		1675
		1676	So when you encounter spurious, unexplained daemon exits, make sure you
		1677	ignore SIGPIPE (and maybe make sure you log the exit status of your daemon
		1678	somewhere, as that would have given you a big clue).
		1679
		1680	=head3 The special problem of accept()ing when you can't
		1681
		1682	Many implementations of the POSIX C<accept> function (for example,
		1683	found in post-2004 Linux) have the peculiar behaviour of not removing a
		1684	connection from the pending queue in all error cases.
		1685
		1686	For example, larger servers often run out of file descriptors (because
		1687	of resource limits), causing C<accept> to fail with C<ENFILE> but not
		1688	rejecting the connection, leading to libev signalling readiness on
		1689	the next iteration again (the connection still exists after all), and
		1690	typically causing the program to loop at 100% CPU usage.
		1691
		1692	Unfortunately, the set of errors that cause this issue differs between
		1693	operating systems, there is usually little the app can do to remedy the
		1694	situation, and no known thread-safe method of removing the connection to
		1695	cope with overload is known (to me).
		1696
		1697	One of the easiest ways to handle this situation is to just ignore it
		1698	- when the program encounters an overload, it will just loop until the
		1699	situation is over. While this is a form of busy waiting, no OS offers an
		1700	event-based way to handle this situation, so it's the best one can do.
		1701
		1702	A better way to handle the situation is to log any errors other than
		1703	C<EAGAIN> and C<EWOULDBLOCK>, making sure not to flood the log with such
		1704	messages, and continue as usual, which at least gives the user an idea of
		1705	what could be wrong ("raise the ulimit!"). For extra points one could stop
		1706	the C<ev_io> watcher on the listening fd "for a while", which reduces CPU
		1707	usage.
		1708
		1709	If your program is single-threaded, then you could also keep a dummy file
		1710	descriptor for overload situations (e.g. by opening F</dev/null>), and
		1711	when you run into C<ENFILE> or C<EMFILE>, close it, run C<accept>,
		1712	close that fd, and create a new dummy fd. This will gracefully refuse
		1713	clients under typical overload conditions.
		1714
		1715	The last way to handle it is to simply log the error and C<exit>, as
		1716	is often done with C<malloc> failures, but this results in an easy
		1717	opportunity for a DoS attack.
		1718
		1719	=head3 Watcher-Specific Functions
736		1720
737	=over 4	1721	=over 4
738		1722
739	=item ev_io_init (ev_io *, callback, int fd, int events)	1723	=item ev_io_init (ev_io *, callback, int fd, int events)
740		1724
741	=item ev_io_set (ev_io *, int fd, int events)	1725	=item ev_io_set (ev_io *, int fd, int events)
742		1726
743	Configures an C<ev_io> watcher. The C<fd> is the file descriptor to	1727	Configures an C<ev_io> watcher. The C<fd> is the file descriptor to
744	rceeive events for and events is either C<EV_READ>, C<EV_WRITE> or	1728	receive events for and C<events> is either C<EV_READ>, C<EV_WRITE> or
745	C<EV_READ | EV_WRITE> to receive the given events.	1729	C<EV_READ | EV_WRITE>, to express the desire to receive the given events.
		1730
		1731	=item int fd [read-only]
		1732
		1733	The file descriptor being watched.
		1734
		1735	=item int events [read-only]
		1736
		1737	The events being watched.
746		1738
747	=back	1739	=back
748		1740
		1741	=head3 Examples
		1742
749	Example: call C<stdin_readable_cb> when STDIN_FILENO has become, well	1743	Example: Call C<stdin_readable_cb> when STDIN_FILENO has become, well
750	readable, but only once. Since it is likely line-buffered, you could	1744	readable, but only once. Since it is likely line-buffered, you could
751	attempt to read a whole line in the callback:	1745	attempt to read a whole line in the callback.
752		1746
753	static void	1747	static void
754	stdin_readable_cb (struct ev_loop *loop, struct ev_io *w, int revents)	1748	stdin_readable_cb (struct ev_loop *loop, ev_io *w, int revents)
755	{	1749	{
756	ev_io_stop (loop, w);	1750	ev_io_stop (loop, w);
757	.. read from stdin here (or from w->fd) and haqndle any I/O errors	1751	.. read from stdin here (or from w->fd) and handle any I/O errors
758	}	1752	}
759		1753
760	...	1754	...
761	struct ev_loop *loop = ev_default_init (0);	1755	struct ev_loop *loop = ev_default_init (0);
762	struct ev_io stdin_readable;	1756	ev_io stdin_readable;
763	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);	1757	ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ);
764	ev_io_start (loop, &stdin_readable);	1758	ev_io_start (loop, &stdin_readable);
765	ev_loop (loop, 0);	1759	ev_run (loop, 0);
766		1760
767		1761
768	=head2 C<ev_timer> - relative and optionally repeating timeouts	1762	=head2 C<ev_timer> - relative and optionally repeating timeouts
769		1763
770	Timer watchers are simple relative timers that generate an event after a	1764	Timer watchers are simple relative timers that generate an event after a
771	given time, and optionally repeating in regular intervals after that.	1765	given time, and optionally repeating in regular intervals after that.
772		1766
773	The timers are based on real time, that is, if you register an event that	1767	The timers are based on real time, that is, if you register an event that
774	times out after an hour and you reset your system clock to last years	1768	times out after an hour and you reset your system clock to January last
775	time, it will still time out after (roughly) and hour. "Roughly" because	1769	year, it will still time out after (roughly) one hour. "Roughly" because
776	detecting time jumps is hard, and some inaccuracies are unavoidable (the	1770	detecting time jumps is hard, and some inaccuracies are unavoidable (the
777	monotonic clock option helps a lot here).	1771	monotonic clock option helps a lot here).
		1772
		1773	The callback is guaranteed to be invoked only I<after> its timeout has
		1774	passed (not I<at>, so on systems with very low-resolution clocks this
		1775	might introduce a small delay). If multiple timers become ready during the
		1776	same loop iteration then the ones with earlier time-out values are invoked
		1777	before ones of the same priority with later time-out values (but this is
		1778	no longer true when a callback calls C<ev_run> recursively).
		1779
		1780	=head3 Be smart about timeouts
		1781
		1782	Many real-world problems involve some kind of timeout, usually for error
		1783	recovery. A typical example is an HTTP request - if the other side hangs,
		1784	you want to raise some error after a while.
		1785
		1786	What follows are some ways to handle this problem, from obvious and
		1787	inefficient to smart and efficient.
		1788
		1789	In the following, a 60 second activity timeout is assumed - a timeout that
		1790	gets reset to 60 seconds each time there is activity (e.g. each time some
		1791	data or other life sign was received).
		1792
		1793	=over 4
		1794
		1795	=item 1. Use a timer and stop, reinitialise and start it on activity.
		1796
		1797	This is the most obvious, but not the most simple way: In the beginning,
		1798	start the watcher:
		1799
		1800	ev_timer_init (timer, callback, 60., 0.);
		1801	ev_timer_start (loop, timer);
		1802
		1803	Then, each time there is some activity, C<ev_timer_stop> it, initialise it
		1804	and start it again:
		1805
		1806	ev_timer_stop (loop, timer);
		1807	ev_timer_set (timer, 60., 0.);
		1808	ev_timer_start (loop, timer);
		1809
		1810	This is relatively simple to implement, but means that each time there is
		1811	some activity, libev will first have to remove the timer from its internal
		1812	data structure and then add it again. Libev tries to be fast, but it's
		1813	still not a constant-time operation.
		1814
		1815	=item 2. Use a timer and re-start it with C<ev_timer_again> inactivity.
		1816
		1817	This is the easiest way, and involves using C<ev_timer_again> instead of
		1818	C<ev_timer_start>.
		1819
		1820	To implement this, configure an C<ev_timer> with a C<repeat> value
		1821	of C<60> and then call C<ev_timer_again> at start and each time you
		1822	successfully read or write some data. If you go into an idle state where
		1823	you do not expect data to travel on the socket, you can C<ev_timer_stop>
		1824	the timer, and C<ev_timer_again> will automatically restart it if need be.
		1825
		1826	That means you can ignore both the C<ev_timer_start> function and the
		1827	C<after> argument to C<ev_timer_set>, and only ever use the C<repeat>
		1828	member and C<ev_timer_again>.
		1829
		1830	At start:
		1831
		1832	ev_init (timer, callback);
		1833	timer->repeat = 60.;
		1834	ev_timer_again (loop, timer);
		1835
		1836	Each time there is some activity:
		1837
		1838	ev_timer_again (loop, timer);
		1839
		1840	It is even possible to change the time-out on the fly, regardless of
		1841	whether the watcher is active or not:
		1842
		1843	timer->repeat = 30.;
		1844	ev_timer_again (loop, timer);
		1845
		1846	This is slightly more efficient then stopping/starting the timer each time
		1847	you want to modify its timeout value, as libev does not have to completely
		1848	remove and re-insert the timer from/into its internal data structure.
		1849
		1850	It is, however, even simpler than the "obvious" way to do it.
		1851
		1852	=item 3. Let the timer time out, but then re-arm it as required.
		1853
		1854	This method is more tricky, but usually most efficient: Most timeouts are
		1855	relatively long compared to the intervals between other activity - in
		1856	our example, within 60 seconds, there are usually many I/O events with
		1857	associated activity resets.
		1858
		1859	In this case, it would be more efficient to leave the C<ev_timer> alone,
		1860	but remember the time of last activity, and check for a real timeout only
		1861	within the callback:
		1862
		1863	ev_tstamp last_activity; // time of last activity
		1864
		1865	static void
		1866	callback (EV_P_ ev_timer *w, int revents)
		1867	{
		1868	ev_tstamp now = ev_now (EV_A);
		1869	ev_tstamp timeout = last_activity + 60.;
		1870
		1871	// if last_activity + 60. is older than now, we did time out
		1872	if (timeout < now)
		1873	{
		1874	// timeout occurred, take action
		1875	}
		1876	else
		1877	{
		1878	// callback was invoked, but there was some activity, re-arm
		1879	// the watcher to fire in last_activity + 60, which is
		1880	// guaranteed to be in the future, so "again" is positive:
		1881	w->repeat = timeout - now;
		1882	ev_timer_again (EV_A_ w);
		1883	}
		1884	}
		1885
		1886	To summarise the callback: first calculate the real timeout (defined
		1887	as "60 seconds after the last activity"), then check if that time has
		1888	been reached, which means something I<did>, in fact, time out. Otherwise
		1889	the callback was invoked too early (C<timeout> is in the future), so
		1890	re-schedule the timer to fire at that future time, to see if maybe we have
		1891	a timeout then.
		1892
		1893	Note how C<ev_timer_again> is used, taking advantage of the
		1894	C<ev_timer_again> optimisation when the timer is already running.
		1895
		1896	This scheme causes more callback invocations (about one every 60 seconds
		1897	minus half the average time between activity), but virtually no calls to
		1898	libev to change the timeout.
		1899
		1900	To start the timer, simply initialise the watcher and set C<last_activity>
		1901	to the current time (meaning we just have some activity :), then call the
		1902	callback, which will "do the right thing" and start the timer:
		1903
		1904	ev_init (timer, callback);
		1905	last_activity = ev_now (loop);
		1906	callback (loop, timer, EV_TIMER);
		1907
		1908	And when there is some activity, simply store the current time in
		1909	C<last_activity>, no libev calls at all:
		1910
		1911	last_activity = ev_now (loop);
		1912
		1913	This technique is slightly more complex, but in most cases where the
		1914	time-out is unlikely to be triggered, much more efficient.
		1915
		1916	Changing the timeout is trivial as well (if it isn't hard-coded in the
		1917	callback :) - just change the timeout and invoke the callback, which will
		1918	fix things for you.
		1919
		1920	=item 4. Wee, just use a double-linked list for your timeouts.
		1921
		1922	If there is not one request, but many thousands (millions...), all
		1923	employing some kind of timeout with the same timeout value, then one can
		1924	do even better:
		1925
		1926	When starting the timeout, calculate the timeout value and put the timeout
		1927	at the I<end> of the list.
		1928
		1929	Then use an C<ev_timer> to fire when the timeout at the I<beginning> of
		1930	the list is expected to fire (for example, using the technique #3).
		1931
		1932	When there is some activity, remove the timer from the list, recalculate
		1933	the timeout, append it to the end of the list again, and make sure to
		1934	update the C<ev_timer> if it was taken from the beginning of the list.
		1935
		1936	This way, one can manage an unlimited number of timeouts in O(1) time for
		1937	starting, stopping and updating the timers, at the expense of a major
		1938	complication, and having to use a constant timeout. The constant timeout
		1939	ensures that the list stays sorted.
		1940
		1941	=back
		1942
		1943	So which method the best?
		1944
		1945	Method #2 is a simple no-brain-required solution that is adequate in most
		1946	situations. Method #3 requires a bit more thinking, but handles many cases
		1947	better, and isn't very complicated either. In most case, choosing either
		1948	one is fine, with #3 being better in typical situations.
		1949
		1950	Method #1 is almost always a bad idea, and buys you nothing. Method #4 is
		1951	rather complicated, but extremely efficient, something that really pays
		1952	off after the first million or so of active timers, i.e. it's usually
		1953	overkill :)
		1954
		1955	=head3 The special problem of time updates
		1956
		1957	Establishing the current time is a costly operation (it usually takes at
		1958	least two system calls): EV therefore updates its idea of the current
		1959	time only before and after C<ev_run> collects new events, which causes a
		1960	growing difference between C<ev_now ()> and C<ev_time ()> when handling
		1961	lots of events in one iteration.
778		1962
779	The relative timeouts are calculated relative to the C<ev_now ()>	1963	The relative timeouts are calculated relative to the C<ev_now ()>
780	time. This is usually the right thing as this timestamp refers to the time	1964	time. This is usually the right thing as this timestamp refers to the time
781	of the event triggering whatever timeout you are modifying/starting. If	1965	of the event triggering whatever timeout you are modifying/starting. If
782	you suspect event processing to be delayed and you I<need> to base the timeout	1966	you suspect event processing to be delayed and you I<need> to base the
783	on the current time, use something like this to adjust for this:	1967	timeout on the current time, use something like this to adjust for this:
784		1968
785	ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);	1969	ev_timer_set (&timer, after + ev_now () - ev_time (), 0.);
786		1970
787	The callback is guarenteed to be invoked only when its timeout has passed,	1971	If the event loop is suspended for a long time, you can also force an
788	but if multiple timers become ready during the same loop iteration then	1972	update of the time returned by C<ev_now ()> by calling C<ev_now_update
789	order of execution is undefined.	1973	()>.
		1974
		1975	=head3 The special problems of suspended animation
		1976
		1977	When you leave the server world it is quite customary to hit machines that
		1978	can suspend/hibernate - what happens to the clocks during such a suspend?
		1979
		1980	Some quick tests made with a Linux 2.6.28 indicate that a suspend freezes
		1981	all processes, while the clocks (C<times>, C<CLOCK_MONOTONIC>) continue
		1982	to run until the system is suspended, but they will not advance while the
		1983	system is suspended. That means, on resume, it will be as if the program
		1984	was frozen for a few seconds, but the suspend time will not be counted
		1985	towards C<ev_timer> when a monotonic clock source is used. The real time
		1986	clock advanced as expected, but if it is used as sole clocksource, then a
		1987	long suspend would be detected as a time jump by libev, and timers would
		1988	be adjusted accordingly.
		1989
		1990	I would not be surprised to see different behaviour in different between
		1991	operating systems, OS versions or even different hardware.
		1992
		1993	The other form of suspend (job control, or sending a SIGSTOP) will see a
		1994	time jump in the monotonic clocks and the realtime clock. If the program
		1995	is suspended for a very long time, and monotonic clock sources are in use,
		1996	then you can expect C<ev_timer>s to expire as the full suspension time
		1997	will be counted towards the timers. When no monotonic clock source is in
		1998	use, then libev will again assume a timejump and adjust accordingly.
		1999
		2000	It might be beneficial for this latter case to call C<ev_suspend>
		2001	and C<ev_resume> in code that handles C<SIGTSTP>, to at least get
		2002	deterministic behaviour in this case (you can do nothing against
		2003	C<SIGSTOP>).
		2004
		2005	=head3 Watcher-Specific Functions and Data Members
790		2006
791	=over 4	2007	=over 4
792		2008
793	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)	2009	=item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat)
794		2010
795	=item ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)	2011	=item ev_timer_set (ev_timer *, ev_tstamp after, ev_tstamp repeat)
796		2012
797	Configure the timer to trigger after C<after> seconds. If C<repeat> is	2013	Configure the timer to trigger after C<after> seconds. If C<repeat>
798	C<0.>, then it will automatically be stopped. If it is positive, then the	2014	is C<0.>, then it will automatically be stopped once the timeout is
799	timer will automatically be configured to trigger again C<repeat> seconds	2015	reached. If it is positive, then the timer will automatically be
800	later, again, and again, until stopped manually.	2016	configured to trigger again C<repeat> seconds later, again, and again,
		2017	until stopped manually.
801		2018
802	The timer itself will do a best-effort at avoiding drift, that is, if you	2019	The timer itself will do a best-effort at avoiding drift, that is, if
803	configure a timer to trigger every 10 seconds, then it will trigger at	2020	you configure a timer to trigger every 10 seconds, then it will normally
804	exactly 10 second intervals. If, however, your program cannot keep up with	2021	trigger at exactly 10 second intervals. If, however, your program cannot
805	the timer (because it takes longer than those 10 seconds to do stuff) the	2022	keep up with the timer (because it takes longer than those 10 seconds to
806	timer will not fire more than once per event loop iteration.	2023	do stuff) the timer will not fire more than once per event loop iteration.
807		2024
808	=item ev_timer_again (loop)	2025	=item ev_timer_again (loop, ev_timer *)
809		2026
810	This will act as if the timer timed out and restart it again if it is	2027	This will act as if the timer timed out and restart it again if it is
811	repeating. The exact semantics are:	2028	repeating. The exact semantics are:
812		2029
		2030	If the timer is pending, its pending status is cleared.
		2031
813	If the timer is started but nonrepeating, stop it.	2032	If the timer is started but non-repeating, stop it (as if it timed out).
814		2033
815	If the timer is repeating, either start it if necessary (with the repeat	2034	If the timer is repeating, either start it if necessary (with the
816	value), or reset the running timer to the repeat value.	2035	C<repeat> value), or reset the running timer to the C<repeat> value.
817		2036
818	This sounds a bit complicated, but here is a useful and typical	2037	This sounds a bit complicated, see L<Be smart about timeouts>, above, for a
819	example: Imagine you have a tcp connection and you want a so-called idle	2038	usage example.
820	timeout, that is, you want to be called when there have been, say, 60	2039
821	seconds of inactivity on the socket. The easiest way to do this is to	2040	=item ev_tstamp ev_timer_remaining (loop, ev_timer *)
822	configure an C<ev_timer> with after=repeat=60 and calling ev_timer_again each	2041
823	time you successfully read or write some data. If you go into an idle	2042	Returns the remaining time until a timer fires. If the timer is active,
824	state where you do not expect data to travel on the socket, you can stop	2043	then this time is relative to the current event loop time, otherwise it's
825	the timer, and again will automatically restart it if need be.	2044	the timeout value currently configured.
		2045
		2046	That is, after an C<ev_timer_set (w, 5, 7)>, C<ev_timer_remaining> returns
		2047	C<5>. When the timer is started and one second passes, C<ev_timer_remaining>
		2048	will return C<4>. When the timer expires and is restarted, it will return
		2049	roughly C<7> (likely slightly less as callback invocation takes some time,
		2050	too), and so on.
		2051
		2052	=item ev_tstamp repeat [read-write]
		2053
		2054	The current C<repeat> value. Will be used each time the watcher times out
		2055	or C<ev_timer_again> is called, and determines the next timeout (if any),
		2056	which is also when any modifications are taken into account.
826		2057
827	=back	2058	=back
828		2059
		2060	=head3 Examples
		2061
829	Example: create a timer that fires after 60 seconds.	2062	Example: Create a timer that fires after 60 seconds.
830		2063
831	static void	2064	static void
832	one_minute_cb (struct ev_loop *loop, struct ev_timer *w, int revents)	2065	one_minute_cb (struct ev_loop *loop, ev_timer *w, int revents)
833	{	2066	{
834	.. one minute over, w is actually stopped right here	2067	.. one minute over, w is actually stopped right here
835	}	2068	}
836		2069
837	struct ev_timer mytimer;	2070	ev_timer mytimer;
838	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);	2071	ev_timer_init (&mytimer, one_minute_cb, 60., 0.);
839	ev_timer_start (loop, &mytimer);	2072	ev_timer_start (loop, &mytimer);
840		2073
841	Example: create a timeout timer that times out after 10 seconds of	2074	Example: Create a timeout timer that times out after 10 seconds of
842	inactivity.	2075	inactivity.
843		2076
844	static void	2077	static void
845	timeout_cb (struct ev_loop *loop, struct ev_timer *w, int revents)	2078	timeout_cb (struct ev_loop *loop, ev_timer *w, int revents)
846	{	2079	{
847	.. ten seconds without any activity	2080	.. ten seconds without any activity
848	}	2081	}
849		2082
850	struct ev_timer mytimer;	2083	ev_timer mytimer;
851	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */	2084	ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */
852	ev_timer_again (&mytimer); /* start timer */	2085	ev_timer_again (&mytimer); /* start timer */
853	ev_loop (loop, 0);	2086	ev_run (loop, 0);
854		2087
855	// and in some piece of code that gets executed on any "activity":	2088	// and in some piece of code that gets executed on any "activity":
856	// reset the timeout to start ticking again at 10 seconds	2089	// reset the timeout to start ticking again at 10 seconds
857	ev_timer_again (&mytimer);	2090	ev_timer_again (&mytimer);
858		2091
859		2092
860	=head2 C<ev_periodic> - to cron or not to cron?	2093	=head2 C<ev_periodic> - to cron or not to cron?
861		2094
862	Periodic watchers are also timers of a kind, but they are very versatile	2095	Periodic watchers are also timers of a kind, but they are very versatile
863	(and unfortunately a bit complex).	2096	(and unfortunately a bit complex).
864		2097
865	Unlike C<ev_timer>'s, they are not based on real time (or relative time)	2098	Unlike C<ev_timer>, periodic watchers are not based on real time (or
866	but on wallclock time (absolute time). You can tell a periodic watcher	2099	relative time, the physical time that passes) but on wall clock time
867	to trigger "at" some specific point in time. For example, if you tell a	2100	(absolute time, the thing you can read on your calender or clock). The
868	periodic watcher to trigger in 10 seconds (by specifiying e.g. C<ev_now ()	2101	difference is that wall clock time can run faster or slower than real
869	+ 10.>) and then reset your system clock to the last year, then it will	2102	time, and time jumps are not uncommon (e.g. when you adjust your
870	take a year to trigger the event (unlike an C<ev_timer>, which would trigger	2103	wrist-watch).
871	roughly 10 seconds later and of course not if you reset your system time
872	again).
873		2104
874	They can also be used to implement vastly more complex timers, such as	2105	You can tell a periodic watcher to trigger after some specific point
		2106	in time: for example, if you tell a periodic watcher to trigger "in 10
		2107	seconds" (by specifying e.g. C<ev_now () + 10.>, that is, an absolute time
		2108	not a delay) and then reset your system clock to January of the previous
		2109	year, then it will take a year or more to trigger the event (unlike an
		2110	C<ev_timer>, which would still trigger roughly 10 seconds after starting
		2111	it, as it uses a relative timeout).
		2112
		2113	C<ev_periodic> watchers can also be used to implement vastly more complex
875	triggering an event on eahc midnight, local time.	2114	timers, such as triggering an event on each "midnight, local time", or
		2115	other complicated rules. This cannot be done with C<ev_timer> watchers, as
		2116	those cannot react to time jumps.
876		2117
877	As with timers, the callback is guarenteed to be invoked only when the	2118	As with timers, the callback is guaranteed to be invoked only when the
878	time (C<at>) has been passed, but if multiple periodic timers become ready	2119	point in time where it is supposed to trigger has passed. If multiple
879	during the same loop iteration then order of execution is undefined.	2120	timers become ready during the same loop iteration then the ones with
		2121	earlier time-out values are invoked before ones with later time-out values
		2122	(but this is no longer true when a callback calls C<ev_run> recursively).
		2123
		2124	=head3 Watcher-Specific Functions and Data Members
880		2125
881	=over 4	2126	=over 4
882		2127
883	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb)	2128	=item ev_periodic_init (ev_periodic *, callback, ev_tstamp offset, ev_tstamp interval, reschedule_cb)
884		2129
885	=item ev_periodic_set (ev_periodic *, ev_tstamp after, ev_tstamp repeat, reschedule_cb)	2130	=item ev_periodic_set (ev_periodic *, ev_tstamp offset, ev_tstamp interval, reschedule_cb)
886		2131
887	Lots of arguments, lets sort it out... There are basically three modes of	2132	Lots of arguments, let's sort it out... There are basically three modes of
888	operation, and we will explain them from simplest to complex:	2133	operation, and we will explain them from simplest to most complex:
889		2134
890	=over 4	2135	=over 4
891		2136
892	=item * absolute timer (interval = reschedule_cb = 0)	2137	=item * absolute timer (offset = absolute time, interval = 0, reschedule_cb = 0)
893		2138
894	In this configuration the watcher triggers an event at the wallclock time	2139	In this configuration the watcher triggers an event after the wall clock
895	C<at> and doesn't repeat. It will not adjust when a time jump occurs,	2140	time C<offset> has passed. It will not repeat and will not adjust when a
896	that is, if it is to be run at January 1st 2011 then it will run when the	2141	time jump occurs, that is, if it is to be run at January 1st 2011 then it
897	system time reaches or surpasses this time.	2142	will be stopped and invoked when the system clock reaches or surpasses
		2143	this point in time.
898		2144
899	=item * non-repeating interval timer (interval > 0, reschedule_cb = 0)	2145	=item * repeating interval timer (offset = offset within interval, interval > 0, reschedule_cb = 0)
900		2146
901	In this mode the watcher will always be scheduled to time out at the next	2147	In this mode the watcher will always be scheduled to time out at the next
902	C<at + N * interval> time (for some integer N) and then repeat, regardless	2148	C<offset + N * interval> time (for some integer N, which can also be
903	of any time jumps.	2149	negative) and then repeat, regardless of any time jumps. The C<offset>
		2150	argument is merely an offset into the C<interval> periods.
904		2151
905	This can be used to create timers that do not drift with respect to system	2152	This can be used to create timers that do not drift with respect to the
906	time:	2153	system clock, for example, here is an C<ev_periodic> that triggers each
		2154	hour, on the hour (with respect to UTC):
907		2155
908	ev_periodic_set (&periodic, 0., 3600., 0);	2156	ev_periodic_set (&periodic, 0., 3600., 0);
909		2157
910	This doesn't mean there will always be 3600 seconds in between triggers,	2158	This doesn't mean there will always be 3600 seconds in between triggers,
911	but only that the the callback will be called when the system time shows a	2159	but only that the callback will be called when the system time shows a
912	full hour (UTC), or more correctly, when the system time is evenly divisible	2160	full hour (UTC), or more correctly, when the system time is evenly divisible
913	by 3600.	2161	by 3600.
914		2162
915	Another way to think about it (for the mathematically inclined) is that	2163	Another way to think about it (for the mathematically inclined) is that
916	C<ev_periodic> will try to run the callback in this mode at the next possible	2164	C<ev_periodic> will try to run the callback in this mode at the next possible
917	time where C<time = at (mod interval)>, regardless of any time jumps.	2165	time where C<time = offset (mod interval)>, regardless of any time jumps.
918		2166
		2167	The C<interval> I<MUST> be positive, and for numerical stability, the
		2168	interval value should be higher than C<1/8192> (which is around 100
		2169	microseconds) and C<offset> should be higher than C<0> and should have
		2170	at most a similar magnitude as the current time (say, within a factor of
		2171	ten). Typical values for offset are, in fact, C<0> or something between
		2172	C<0> and C<interval>, which is also the recommended range.
		2173
		2174	Note also that there is an upper limit to how often a timer can fire (CPU
		2175	speed for example), so if C<interval> is very small then timing stability
		2176	will of course deteriorate. Libev itself tries to be exact to be about one
		2177	millisecond (if the OS supports it and the machine is fast enough).
		2178
919	=item * manual reschedule mode (reschedule_cb = callback)	2179	=item * manual reschedule mode (offset ignored, interval ignored, reschedule_cb = callback)
920		2180
921	In this mode the values for C<interval> and C<at> are both being	2181	In this mode the values for C<interval> and C<offset> are both being
922	ignored. Instead, each time the periodic watcher gets scheduled, the	2182	ignored. Instead, each time the periodic watcher gets scheduled, the
923	reschedule callback will be called with the watcher as first, and the	2183	reschedule callback will be called with the watcher as first, and the
924	current time as second argument.	2184	current time as second argument.
925		2185
926	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher,	2186	NOTE: I<This callback MUST NOT stop or destroy any periodic watcher, ever,
927	ever, or make any event loop modifications>. If you need to stop it,	2187	or make ANY other event loop modifications whatsoever, unless explicitly
928	return C<now + 1e30> (or so, fudge fudge) and stop it afterwards (e.g. by	2188	allowed by documentation here>.
929	starting a prepare watcher).
930		2189
		2190	If you need to stop it, return C<now + 1e30> (or so, fudge fudge) and stop
		2191	it afterwards (e.g. by starting an C<ev_prepare> watcher, which is the
		2192	only event loop modification you are allowed to do).
		2193
931	Its prototype is C<ev_tstamp (*reschedule_cb)(struct ev_periodic *w,	2194	The callback prototype is C<ev_tstamp (*reschedule_cb)(ev_periodic
932	ev_tstamp now)>, e.g.:	2195	*w, ev_tstamp now)>, e.g.:
933		2196
		2197	static ev_tstamp
934	static ev_tstamp my_rescheduler (struct ev_periodic *w, ev_tstamp now)	2198	my_rescheduler (ev_periodic *w, ev_tstamp now)
935	{	2199	{
936	return now + 60.;	2200	return now + 60.;
937	}	2201	}
938		2202
939	It must return the next time to trigger, based on the passed time value	2203	It must return the next time to trigger, based on the passed time value
940	(that is, the lowest time value larger than to the second argument). It	2204	(that is, the lowest time value larger than to the second argument). It
941	will usually be called just before the callback will be triggered, but	2205	will usually be called just before the callback will be triggered, but
942	might be called at other times, too.	2206	might be called at other times, too.
943		2207
944	NOTE: I<< This callback must always return a time that is later than the	2208	NOTE: I<< This callback must always return a time that is higher than or
945	passed C<now> value >>. Not even C<now> itself will do, it I<must> be larger.	2209	equal to the passed C<now> value >>.
946		2210
947	This can be used to create very complex timers, such as a timer that	2211	This can be used to create very complex timers, such as a timer that
948	triggers on each midnight, local time. To do this, you would calculate the	2212	triggers on "next midnight, local time". To do this, you would calculate the
949	next midnight after C<now> and return the timestamp value for this. How	2213	next midnight after C<now> and return the timestamp value for this. How
950	you do this is, again, up to you (but it is not trivial, which is the main	2214	you do this is, again, up to you (but it is not trivial, which is the main
951	reason I omitted it as an example).	2215	reason I omitted it as an example).
952		2216
953	=back	2217	=back
…		…
957	Simply stops and restarts the periodic watcher again. This is only useful	2221	Simply stops and restarts the periodic watcher again. This is only useful
958	when you changed some parameters or the reschedule callback would return	2222	when you changed some parameters or the reschedule callback would return
959	a different time than the last time it was called (e.g. in a crond like	2223	a different time than the last time it was called (e.g. in a crond like
960	program when the crontabs have changed).	2224	program when the crontabs have changed).
961		2225
		2226	=item ev_tstamp ev_periodic_at (ev_periodic *)
		2227
		2228	When active, returns the absolute time that the watcher is supposed
		2229	to trigger next. This is not the same as the C<offset> argument to
		2230	C<ev_periodic_set>, but indeed works even in interval and manual
		2231	rescheduling modes.
		2232
		2233	=item ev_tstamp offset [read-write]
		2234
		2235	When repeating, this contains the offset value, otherwise this is the
		2236	absolute point in time (the C<offset> value passed to C<ev_periodic_set>,
		2237	although libev might modify this value for better numerical stability).
		2238
		2239	Can be modified any time, but changes only take effect when the periodic
		2240	timer fires or C<ev_periodic_again> is being called.
		2241
		2242	=item ev_tstamp interval [read-write]
		2243
		2244	The current interval value. Can be modified any time, but changes only
		2245	take effect when the periodic timer fires or C<ev_periodic_again> is being
		2246	called.
		2247
		2248	=item ev_tstamp (*reschedule_cb)(ev_periodic *w, ev_tstamp now) [read-write]
		2249
		2250	The current reschedule callback, or C<0>, if this functionality is
		2251	switched off. Can be changed any time, but changes only take effect when
		2252	the periodic timer fires or C<ev_periodic_again> is being called.
		2253
962	=back	2254	=back
963		2255
		2256	=head3 Examples
		2257
964	Example: call a callback every hour, or, more precisely, whenever the	2258	Example: Call a callback every hour, or, more precisely, whenever the
965	system clock is divisible by 3600. The callback invocation times have	2259	system time is divisible by 3600. The callback invocation times have
966	potentially a lot of jittering, but good long-term stability.	2260	potentially a lot of jitter, but good long-term stability.
967		2261
968	static void	2262	static void
969	clock_cb (struct ev_loop *loop, struct ev_io *w, int revents)	2263	clock_cb (struct ev_loop *loop, ev_periodic *w, int revents)
970	{	2264	{
971	... its now a full hour (UTC, or TAI or whatever your clock follows)	2265	... its now a full hour (UTC, or TAI or whatever your clock follows)
972	}	2266	}
973		2267
974	struct ev_periodic hourly_tick;	2268	ev_periodic hourly_tick;
975	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);	2269	ev_periodic_init (&hourly_tick, clock_cb, 0., 3600., 0);
976	ev_periodic_start (loop, &hourly_tick);	2270	ev_periodic_start (loop, &hourly_tick);
977		2271
978	Example: the same as above, but use a reschedule callback to do it:	2272	Example: The same as above, but use a reschedule callback to do it:
979		2273
980	#include <math.h>	2274	#include <math.h>
981		2275
982	static ev_tstamp	2276	static ev_tstamp
983	my_scheduler_cb (struct ev_periodic *w, ev_tstamp now)	2277	my_scheduler_cb (ev_periodic *w, ev_tstamp now)
984	{	2278	{
985	return fmod (now, 3600.) + 3600.;	2279	return now + (3600. - fmod (now, 3600.));
986	}	2280	}
987		2281
988	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);	2282	ev_periodic_init (&hourly_tick, clock_cb, 0., 0., my_scheduler_cb);
989		2283
990	Example: call a callback every hour, starting now:	2284	Example: Call a callback every hour, starting now:
991		2285
992	struct ev_periodic hourly_tick;	2286	ev_periodic hourly_tick;
993	ev_periodic_init (&hourly_tick, clock_cb,	2287	ev_periodic_init (&hourly_tick, clock_cb,
994	fmod (ev_now (loop), 3600.), 3600., 0);	2288	fmod (ev_now (loop), 3600.), 3600., 0);
995	ev_periodic_start (loop, &hourly_tick);	2289	ev_periodic_start (loop, &hourly_tick);
996		2290
997		2291
998	=head2 C<ev_signal> - signal me when a signal gets signalled!	2292	=head2 C<ev_signal> - signal me when a signal gets signalled!
999		2293
1000	Signal watchers will trigger an event when the process receives a specific	2294	Signal watchers will trigger an event when the process receives a specific
1001	signal one or more times. Even though signals are very asynchronous, libev	2295	signal one or more times. Even though signals are very asynchronous, libev
1002	will try it's best to deliver signals synchronously, i.e. as part of the	2296	will try its best to deliver signals synchronously, i.e. as part of the
1003	normal event processing, like any other event.	2297	normal event processing, like any other event.
1004		2298
		2299	If you want signals to be delivered truly asynchronously, just use
		2300	C<sigaction> as you would do without libev and forget about sharing
		2301	the signal. You can even use C<ev_async> from a signal handler to
		2302	synchronously wake up an event loop.
		2303
1005	You can configure as many watchers as you like per signal. Only when the	2304	You can configure as many watchers as you like for the same signal, but
		2305	only within the same loop, i.e. you can watch for C<SIGINT> in your
		2306	default loop and for C<SIGIO> in another loop, but you cannot watch for
		2307	C<SIGINT> in both the default loop and another loop at the same time. At
		2308	the moment, C<SIGCHLD> is permanently tied to the default loop.
		2309
1006	first watcher gets started will libev actually register a signal watcher	2310	When the first watcher gets started will libev actually register something
1007	with the kernel (thus it coexists with your own signal handlers as long	2311	with the kernel (thus it coexists with your own signal handlers as long as
1008	as you don't register any with libev). Similarly, when the last signal	2312	you don't register any with libev for the same signal).
1009	watcher for a signal is stopped libev will reset the signal handler to	2313
1010	SIG_DFL (regardless of what it was set to before).	2314	If possible and supported, libev will install its handlers with
		2315	C<SA_RESTART> (or equivalent) behaviour enabled, so system calls should
		2316	not be unduly interrupted. If you have a problem with system calls getting
		2317	interrupted by signals you can block all signals in an C<ev_check> watcher
		2318	and unblock them in an C<ev_prepare> watcher.
		2319
		2320	=head3 The special problem of inheritance over fork/execve/pthread_create
		2321
		2322	Both the signal mask (C<sigprocmask>) and the signal disposition
		2323	(C<sigaction>) are unspecified after starting a signal watcher (and after
		2324	stopping it again), that is, libev might or might not block the signal,
		2325	and might or might not set or restore the installed signal handler (but
		2326	see C<EVFLAG_NOSIGMASK>).
		2327
		2328	While this does not matter for the signal disposition (libev never
		2329	sets signals to C<SIG_IGN>, so handlers will be reset to C<SIG_DFL> on
		2330	C<execve>), this matters for the signal mask: many programs do not expect
		2331	certain signals to be blocked.
		2332
		2333	This means that before calling C<exec> (from the child) you should reset
		2334	the signal mask to whatever "default" you expect (all clear is a good
		2335	choice usually).
		2336
		2337	The simplest way to ensure that the signal mask is reset in the child is
		2338	to install a fork handler with C<pthread_atfork> that resets it. That will
		2339	catch fork calls done by libraries (such as the libc) as well.
		2340
		2341	In current versions of libev, the signal will not be blocked indefinitely
		2342	unless you use the C<signalfd> API (C<EV_SIGNALFD>). While this reduces
		2343	the window of opportunity for problems, it will not go away, as libev
		2344	I<has> to modify the signal mask, at least temporarily.
		2345
		2346	So I can't stress this enough: I<If you do not reset your signal mask when
		2347	you expect it to be empty, you have a race condition in your code>. This
		2348	is not a libev-specific thing, this is true for most event libraries.
		2349
		2350	=head3 The special problem of threads signal handling
		2351
		2352	POSIX threads has problematic signal handling semantics, specifically,
		2353	a lot of functionality (sigfd, sigwait etc.) only really works if all
		2354	threads in a process block signals, which is hard to achieve.
		2355
		2356	When you want to use sigwait (or mix libev signal handling with your own
		2357	for the same signals), you can tackle this problem by globally blocking
		2358	all signals before creating any threads (or creating them with a fully set
		2359	sigprocmask) and also specifying the C<EVFLAG_NOSIGMASK> when creating
		2360	loops. Then designate one thread as "signal receiver thread" which handles
		2361	these signals. You can pass on any signals that libev might be interested
		2362	in by calling C<ev_feed_signal>.
		2363
		2364	=head3 Watcher-Specific Functions and Data Members
1011		2365
1012	=over 4	2366	=over 4
1013		2367
1014	=item ev_signal_init (ev_signal *, callback, int signum)	2368	=item ev_signal_init (ev_signal *, callback, int signum)
1015		2369
1016	=item ev_signal_set (ev_signal *, int signum)	2370	=item ev_signal_set (ev_signal *, int signum)
1017		2371
1018	Configures the watcher to trigger on the given signal number (usually one	2372	Configures the watcher to trigger on the given signal number (usually one
1019	of the C<SIGxxx> constants).	2373	of the C<SIGxxx> constants).
1020		2374
		2375	=item int signum [read-only]
		2376
		2377	The signal the watcher watches out for.
		2378
1021	=back	2379	=back
1022		2380
		2381	=head3 Examples
		2382
		2383	Example: Try to exit cleanly on SIGINT.
		2384
		2385	static void
		2386	sigint_cb (struct ev_loop *loop, ev_signal *w, int revents)
		2387	{
		2388	ev_break (loop, EVBREAK_ALL);
		2389	}
		2390
		2391	ev_signal signal_watcher;
		2392	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);
		2393	ev_signal_start (loop, &signal_watcher);
		2394
1023		2395
1024	=head2 C<ev_child> - watch out for process status changes	2396	=head2 C<ev_child> - watch out for process status changes
1025		2397
1026	Child watchers trigger when your process receives a SIGCHLD in response to	2398	Child watchers trigger when your process receives a SIGCHLD in response to
1027	some child status changes (most typically when a child of yours dies).	2399	some child status changes (most typically when a child of yours dies or
		2400	exits). It is permissible to install a child watcher I<after> the child
		2401	has been forked (which implies it might have already exited), as long
		2402	as the event loop isn't entered (or is continued from a watcher), i.e.,
		2403	forking and then immediately registering a watcher for the child is fine,
		2404	but forking and registering a watcher a few event loop iterations later or
		2405	in the next callback invocation is not.
		2406
		2407	Only the default event loop is capable of handling signals, and therefore
		2408	you can only register child watchers in the default event loop.
		2409
		2410	Due to some design glitches inside libev, child watchers will always be
		2411	handled at maximum priority (their priority is set to C<EV_MAXPRI> by
		2412	libev)
		2413
		2414	=head3 Process Interaction
		2415
		2416	Libev grabs C<SIGCHLD> as soon as the default event loop is
		2417	initialised. This is necessary to guarantee proper behaviour even if the
		2418	first child watcher is started after the child exits. The occurrence
		2419	of C<SIGCHLD> is recorded asynchronously, but child reaping is done
		2420	synchronously as part of the event loop processing. Libev always reaps all
		2421	children, even ones not watched.
		2422
		2423	=head3 Overriding the Built-In Processing
		2424
		2425	Libev offers no special support for overriding the built-in child
		2426	processing, but if your application collides with libev's default child
		2427	handler, you can override it easily by installing your own handler for
		2428	C<SIGCHLD> after initialising the default loop, and making sure the
		2429	default loop never gets destroyed. You are encouraged, however, to use an
		2430	event-based approach to child reaping and thus use libev's support for
		2431	that, so other libev users can use C<ev_child> watchers freely.
		2432
		2433	=head3 Stopping the Child Watcher
		2434
		2435	Currently, the child watcher never gets stopped, even when the
		2436	child terminates, so normally one needs to stop the watcher in the
		2437	callback. Future versions of libev might stop the watcher automatically
		2438	when a child exit is detected (calling C<ev_child_stop> twice is not a
		2439	problem).
		2440
		2441	=head3 Watcher-Specific Functions and Data Members
1028		2442
1029	=over 4	2443	=over 4
1030		2444
1031	=item ev_child_init (ev_child *, callback, int pid)	2445	=item ev_child_init (ev_child *, callback, int pid, int trace)
1032		2446
1033	=item ev_child_set (ev_child *, int pid)	2447	=item ev_child_set (ev_child *, int pid, int trace)
1034		2448
1035	Configures the watcher to wait for status changes of process C<pid> (or	2449	Configures the watcher to wait for status changes of process C<pid> (or
1036	I<any> process if C<pid> is specified as C<0>). The callback can look	2450	I<any> process if C<pid> is specified as C<0>). The callback can look
1037	at the C<rstatus> member of the C<ev_child> watcher structure to see	2451	at the C<rstatus> member of the C<ev_child> watcher structure to see
1038	the status word (use the macros from C<sys/wait.h> and see your systems	2452	the status word (use the macros from C<sys/wait.h> and see your systems
1039	C<waitpid> documentation). The C<rpid> member contains the pid of the	2453	C<waitpid> documentation). The C<rpid> member contains the pid of the
1040	process causing the status change.	2454	process causing the status change. C<trace> must be either C<0> (only
		2455	activate the watcher when the process terminates) or C<1> (additionally
		2456	activate the watcher when the process is stopped or continued).
		2457
		2458	=item int pid [read-only]
		2459
		2460	The process id this watcher watches out for, or C<0>, meaning any process id.
		2461
		2462	=item int rpid [read-write]
		2463
		2464	The process id that detected a status change.
		2465
		2466	=item int rstatus [read-write]
		2467
		2468	The process exit/trace status caused by C<rpid> (see your systems
		2469	C<waitpid> and C<sys/wait.h> documentation for details).
1041		2470
1042	=back	2471	=back
1043		2472
1044	Example: try to exit cleanly on SIGINT and SIGTERM.	2473	=head3 Examples
1045		2474
		2475	Example: C<fork()> a new process and install a child handler to wait for
		2476	its completion.
		2477
		2478	ev_child cw;
		2479
1046	static void	2480	static void
1047	sigint_cb (struct ev_loop *loop, struct ev_signal *w, int revents)	2481	child_cb (EV_P_ ev_child *w, int revents)
1048	{	2482	{
1049	ev_unloop (loop, EVUNLOOP_ALL);	2483	ev_child_stop (EV_A_ w);
		2484	printf ("process %d exited with status %x\n", w->rpid, w->rstatus);
1050	}	2485	}
1051		2486
1052	struct ev_signal signal_watcher;	2487	pid_t pid = fork ();
1053	ev_signal_init (&signal_watcher, sigint_cb, SIGINT);	2488
1054	ev_signal_start (loop, &sigint_cb);	2489	if (pid < 0)
		2490	// error
		2491	else if (pid == 0)
		2492	{
		2493	// the forked child executes here
		2494	exit (1);
		2495	}
		2496	else
		2497	{
		2498	ev_child_init (&cw, child_cb, pid, 0);
		2499	ev_child_start (EV_DEFAULT_ &cw);
		2500	}
		2501
		2502
		2503	=head2 C<ev_stat> - did the file attributes just change?
		2504
		2505	This watches a file system path for attribute changes. That is, it calls
		2506	C<stat> on that path in regular intervals (or when the OS says it changed)
		2507	and sees if it changed compared to the last time, invoking the callback if
		2508	it did.
		2509
		2510	The path does not need to exist: changing from "path exists" to "path does
		2511	not exist" is a status change like any other. The condition "path does not
		2512	exist" (or more correctly "path cannot be stat'ed") is signified by the
		2513	C<st_nlink> field being zero (which is otherwise always forced to be at
		2514	least one) and all the other fields of the stat buffer having unspecified
		2515	contents.
		2516
		2517	The path I<must not> end in a slash or contain special components such as
		2518	C<.> or C<..>. The path I<should> be absolute: If it is relative and
		2519	your working directory changes, then the behaviour is undefined.
		2520
		2521	Since there is no portable change notification interface available, the
		2522	portable implementation simply calls C<stat(2)> regularly on the path
		2523	to see if it changed somehow. You can specify a recommended polling
		2524	interval for this case. If you specify a polling interval of C<0> (highly
		2525	recommended!) then a I<suitable, unspecified default> value will be used
		2526	(which you can expect to be around five seconds, although this might
		2527	change dynamically). Libev will also impose a minimum interval which is
		2528	currently around C<0.1>, but that's usually overkill.
		2529
		2530	This watcher type is not meant for massive numbers of stat watchers,
		2531	as even with OS-supported change notifications, this can be
		2532	resource-intensive.
		2533
		2534	At the time of this writing, the only OS-specific interface implemented
		2535	is the Linux inotify interface (implementing kqueue support is left as an
		2536	exercise for the reader. Note, however, that the author sees no way of
		2537	implementing C<ev_stat> semantics with kqueue, except as a hint).
		2538
		2539	=head3 ABI Issues (Largefile Support)
		2540
		2541	Libev by default (unless the user overrides this) uses the default
		2542	compilation environment, which means that on systems with large file
		2543	support disabled by default, you get the 32 bit version of the stat
		2544	structure. When using the library from programs that change the ABI to
		2545	use 64 bit file offsets the programs will fail. In that case you have to
		2546	compile libev with the same flags to get binary compatibility. This is
		2547	obviously the case with any flags that change the ABI, but the problem is
		2548	most noticeably displayed with ev_stat and large file support.
		2549
		2550	The solution for this is to lobby your distribution maker to make large
		2551	file interfaces available by default (as e.g. FreeBSD does) and not
		2552	optional. Libev cannot simply switch on large file support because it has
		2553	to exchange stat structures with application programs compiled using the
		2554	default compilation environment.
		2555
		2556	=head3 Inotify and Kqueue
		2557
		2558	When C<inotify (7)> support has been compiled into libev and present at
		2559	runtime, it will be used to speed up change detection where possible. The
		2560	inotify descriptor will be created lazily when the first C<ev_stat>
		2561	watcher is being started.
		2562
		2563	Inotify presence does not change the semantics of C<ev_stat> watchers
		2564	except that changes might be detected earlier, and in some cases, to avoid
		2565	making regular C<stat> calls. Even in the presence of inotify support
		2566	there are many cases where libev has to resort to regular C<stat> polling,
		2567	but as long as kernel 2.6.25 or newer is used (2.6.24 and older have too
		2568	many bugs), the path exists (i.e. stat succeeds), and the path resides on
		2569	a local filesystem (libev currently assumes only ext2/3, jfs, reiserfs and
		2570	xfs are fully working) libev usually gets away without polling.
		2571
		2572	There is no support for kqueue, as apparently it cannot be used to
		2573	implement this functionality, due to the requirement of having a file
		2574	descriptor open on the object at all times, and detecting renames, unlinks
		2575	etc. is difficult.
		2576
		2577	=head3 C<stat ()> is a synchronous operation
		2578
		2579	Libev doesn't normally do any kind of I/O itself, and so is not blocking
		2580	the process. The exception are C<ev_stat> watchers - those call C<stat
		2581	()>, which is a synchronous operation.
		2582
		2583	For local paths, this usually doesn't matter: unless the system is very
		2584	busy or the intervals between stat's are large, a stat call will be fast,
		2585	as the path data is usually in memory already (except when starting the
		2586	watcher).
		2587
		2588	For networked file systems, calling C<stat ()> can block an indefinite
		2589	time due to network issues, and even under good conditions, a stat call
		2590	often takes multiple milliseconds.
		2591
		2592	Therefore, it is best to avoid using C<ev_stat> watchers on networked
		2593	paths, although this is fully supported by libev.
		2594
		2595	=head3 The special problem of stat time resolution
		2596
		2597	The C<stat ()> system call only supports full-second resolution portably,
		2598	and even on systems where the resolution is higher, most file systems
		2599	still only support whole seconds.
		2600
		2601	That means that, if the time is the only thing that changes, you can
		2602	easily miss updates: on the first update, C<ev_stat> detects a change and
		2603	calls your callback, which does something. When there is another update
		2604	within the same second, C<ev_stat> will be unable to detect unless the
		2605	stat data does change in other ways (e.g. file size).
		2606
		2607	The solution to this is to delay acting on a change for slightly more
		2608	than a second (or till slightly after the next full second boundary), using
		2609	a roughly one-second-delay C<ev_timer> (e.g. C<ev_timer_set (w, 0., 1.02);
		2610	ev_timer_again (loop, w)>).
		2611
		2612	The C<.02> offset is added to work around small timing inconsistencies
		2613	of some operating systems (where the second counter of the current time
		2614	might be be delayed. One such system is the Linux kernel, where a call to
		2615	C<gettimeofday> might return a timestamp with a full second later than
		2616	a subsequent C<time> call - if the equivalent of C<time ()> is used to
		2617	update file times then there will be a small window where the kernel uses
		2618	the previous second to update file times but libev might already execute
		2619	the timer callback).
		2620
		2621	=head3 Watcher-Specific Functions and Data Members
		2622
		2623	=over 4
		2624
		2625	=item ev_stat_init (ev_stat *, callback, const char *path, ev_tstamp interval)
		2626
		2627	=item ev_stat_set (ev_stat *, const char *path, ev_tstamp interval)
		2628
		2629	Configures the watcher to wait for status changes of the given
		2630	C<path>. The C<interval> is a hint on how quickly a change is expected to
		2631	be detected and should normally be specified as C<0> to let libev choose
		2632	a suitable value. The memory pointed to by C<path> must point to the same
		2633	path for as long as the watcher is active.
		2634
		2635	The callback will receive an C<EV_STAT> event when a change was detected,
		2636	relative to the attributes at the time the watcher was started (or the
		2637	last change was detected).
		2638
		2639	=item ev_stat_stat (loop, ev_stat *)
		2640
		2641	Updates the stat buffer immediately with new values. If you change the
		2642	watched path in your callback, you could call this function to avoid
		2643	detecting this change (while introducing a race condition if you are not
		2644	the only one changing the path). Can also be useful simply to find out the
		2645	new values.
		2646
		2647	=item ev_statdata attr [read-only]
		2648
		2649	The most-recently detected attributes of the file. Although the type is
		2650	C<ev_statdata>, this is usually the (or one of the) C<struct stat> types
		2651	suitable for your system, but you can only rely on the POSIX-standardised
		2652	members to be present. If the C<st_nlink> member is C<0>, then there was
		2653	some error while C<stat>ing the file.
		2654
		2655	=item ev_statdata prev [read-only]
		2656
		2657	The previous attributes of the file. The callback gets invoked whenever
		2658	C<prev> != C<attr>, or, more precisely, one or more of these members
		2659	differ: C<st_dev>, C<st_ino>, C<st_mode>, C<st_nlink>, C<st_uid>,
		2660	C<st_gid>, C<st_rdev>, C<st_size>, C<st_atime>, C<st_mtime>, C<st_ctime>.
		2661
		2662	=item ev_tstamp interval [read-only]
		2663
		2664	The specified interval.
		2665
		2666	=item const char *path [read-only]
		2667
		2668	The file system path that is being watched.
		2669
		2670	=back
		2671
		2672	=head3 Examples
		2673
		2674	Example: Watch C</etc/passwd> for attribute changes.
		2675
		2676	static void
		2677	passwd_cb (struct ev_loop *loop, ev_stat *w, int revents)
		2678	{
		2679	/* /etc/passwd changed in some way */
		2680	if (w->attr.st_nlink)
		2681	{
		2682	printf ("passwd current size %ld\n", (long)w->attr.st_size);
		2683	printf ("passwd current atime %ld\n", (long)w->attr.st_mtime);
		2684	printf ("passwd current mtime %ld\n", (long)w->attr.st_mtime);
		2685	}
		2686	else
		2687	/* you shalt not abuse printf for puts */
		2688	puts ("wow, /etc/passwd is not there, expect problems. "
		2689	"if this is windows, they already arrived\n");
		2690	}
		2691
		2692	...
		2693	ev_stat passwd;
		2694
		2695	ev_stat_init (&passwd, passwd_cb, "/etc/passwd", 0.);
		2696	ev_stat_start (loop, &passwd);
		2697
		2698	Example: Like above, but additionally use a one-second delay so we do not
		2699	miss updates (however, frequent updates will delay processing, too, so
		2700	one might do the work both on C<ev_stat> callback invocation I<and> on
		2701	C<ev_timer> callback invocation).
		2702
		2703	static ev_stat passwd;
		2704	static ev_timer timer;
		2705
		2706	static void
		2707	timer_cb (EV_P_ ev_timer *w, int revents)
		2708	{
		2709	ev_timer_stop (EV_A_ w);
		2710
		2711	/* now it's one second after the most recent passwd change */
		2712	}
		2713
		2714	static void
		2715	stat_cb (EV_P_ ev_stat *w, int revents)
		2716	{
		2717	/* reset the one-second timer */
		2718	ev_timer_again (EV_A_ &timer);
		2719	}
		2720
		2721	...
		2722	ev_stat_init (&passwd, stat_cb, "/etc/passwd", 0.);
		2723	ev_stat_start (loop, &passwd);
		2724	ev_timer_init (&timer, timer_cb, 0., 1.02);
1055		2725
1056		2726
1057	=head2 C<ev_idle> - when you've got nothing better to do...	2727	=head2 C<ev_idle> - when you've got nothing better to do...
1058		2728
1059	Idle watchers trigger events when there are no other events are pending	2729	Idle watchers trigger events when no other events of the same or higher
1060	(prepare, check and other idle watchers do not count). That is, as long	2730	priority are pending (prepare, check and other idle watchers do not count
1061	as your process is busy handling sockets or timeouts (or even signals,	2731	as receiving "events").
1062	imagine) it will not be triggered. But when your process is idle all idle	2732
1063	watchers are being called again and again, once per event loop iteration -	2733	That is, as long as your process is busy handling sockets or timeouts
		2734	(or even signals, imagine) of the same or higher priority it will not be
		2735	triggered. But when your process is idle (or only lower-priority watchers
		2736	are pending), the idle watchers are being called once per event loop
1064	until stopped, that is, or your process receives more events and becomes	2737	iteration - until stopped, that is, or your process receives more events
1065	busy.	2738	and becomes busy again with higher priority stuff.
1066		2739
1067	The most noteworthy effect is that as long as any idle watchers are	2740	The most noteworthy effect is that as long as any idle watchers are
1068	active, the process will not block when waiting for new events.	2741	active, the process will not block when waiting for new events.
1069		2742
1070	Apart from keeping your process non-blocking (which is a useful	2743	Apart from keeping your process non-blocking (which is a useful
1071	effect on its own sometimes), idle watchers are a good place to do	2744	effect on its own sometimes), idle watchers are a good place to do
1072	"pseudo-background processing", or delay processing stuff to after the	2745	"pseudo-background processing", or delay processing stuff to after the
1073	event loop has handled all outstanding events.	2746	event loop has handled all outstanding events.
1074		2747
		2748	=head3 Watcher-Specific Functions and Data Members
		2749
1075	=over 4	2750	=over 4
1076		2751
1077	=item ev_idle_init (ev_signal *, callback)	2752	=item ev_idle_init (ev_idle *, callback)
1078		2753
1079	Initialises and configures the idle watcher - it has no parameters of any	2754	Initialises and configures the idle watcher - it has no parameters of any
1080	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,	2755	kind. There is a C<ev_idle_set> macro, but using it is utterly pointless,
1081	believe me.	2756	believe me.
1082		2757
1083	=back	2758	=back
1084		2759
		2760	=head3 Examples
		2761
1085	Example: dynamically allocate an C<ev_idle>, start it, and in the	2762	Example: Dynamically allocate an C<ev_idle> watcher, start it, and in the
1086	callback, free it. Alos, use no error checking, as usual.	2763	callback, free it. Also, use no error checking, as usual.
1087		2764
1088	static void	2765	static void
1089	idle_cb (struct ev_loop *loop, struct ev_idle *w, int revents)	2766	idle_cb (struct ev_loop *loop, ev_idle *w, int revents)
1090	{	2767	{
1091	free (w);	2768	free (w);
1092	// now do something you wanted to do when the program has	2769	// now do something you wanted to do when the program has
1093	// no longer asnything immediate to do.	2770	// no longer anything immediate to do.
1094	}	2771	}
1095		2772
1096	struct ev_idle *idle_watcher = malloc (sizeof (struct ev_idle));	2773	ev_idle *idle_watcher = malloc (sizeof (ev_idle));
1097	ev_idle_init (idle_watcher, idle_cb);	2774	ev_idle_init (idle_watcher, idle_cb);
1098	ev_idle_start (loop, idle_cb);	2775	ev_idle_start (loop, idle_watcher);
1099		2776
1100		2777
1101	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!	2778	=head2 C<ev_prepare> and C<ev_check> - customise your event loop!
1102		2779
1103	Prepare and check watchers are usually (but not always) used in tandem:	2780	Prepare and check watchers are usually (but not always) used in pairs:
1104	prepare watchers get invoked before the process blocks and check watchers	2781	prepare watchers get invoked before the process blocks and check watchers
1105	afterwards.	2782	afterwards.
1106		2783
		2784	You I<must not> call C<ev_run> or similar functions that enter
		2785	the current event loop from either C<ev_prepare> or C<ev_check>
		2786	watchers. Other loops than the current one are fine, however. The
		2787	rationale behind this is that you do not need to check for recursion in
		2788	those watchers, i.e. the sequence will always be C<ev_prepare>, blocking,
		2789	C<ev_check> so if you have one watcher of each kind they will always be
		2790	called in pairs bracketing the blocking call.
		2791
1107	Their main purpose is to integrate other event mechanisms into libev and	2792	Their main purpose is to integrate other event mechanisms into libev and
1108	their use is somewhat advanced. This could be used, for example, to track	2793	their use is somewhat advanced. They could be used, for example, to track
1109	variable changes, implement your own watchers, integrate net-snmp or a	2794	variable changes, implement your own watchers, integrate net-snmp or a
1110	coroutine library and lots more.	2795	coroutine library and lots more. They are also occasionally useful if
		2796	you cache some data and want to flush it before blocking (for example,
		2797	in X programs you might want to do an C<XFlush ()> in an C<ev_prepare>
		2798	watcher).
1111		2799
1112	This is done by examining in each prepare call which file descriptors need	2800	This is done by examining in each prepare call which file descriptors
1113	to be watched by the other library, registering C<ev_io> watchers for	2801	need to be watched by the other library, registering C<ev_io> watchers
1114	them and starting an C<ev_timer> watcher for any timeouts (many libraries	2802	for them and starting an C<ev_timer> watcher for any timeouts (many
1115	provide just this functionality). Then, in the check watcher you check for	2803	libraries provide exactly this functionality). Then, in the check watcher,
1116	any events that occured (by checking the pending status of all watchers	2804	you check for any events that occurred (by checking the pending status
1117	and stopping them) and call back into the library. The I/O and timer	2805	of all watchers and stopping them) and call back into the library. The
1118	callbacks will never actually be called (but must be valid nevertheless,	2806	I/O and timer callbacks will never actually be called (but must be valid
1119	because you never know, you know?).	2807	nevertheless, because you never know, you know?).
1120		2808
1121	As another example, the Perl Coro module uses these hooks to integrate	2809	As another example, the Perl Coro module uses these hooks to integrate
1122	coroutines into libev programs, by yielding to other active coroutines	2810	coroutines into libev programs, by yielding to other active coroutines
1123	during each prepare and only letting the process block if no coroutines	2811	during each prepare and only letting the process block if no coroutines
1124	are ready to run (it's actually more complicated: it only runs coroutines	2812	are ready to run (it's actually more complicated: it only runs coroutines
1125	with priority higher than or equal to the event loop and one coroutine	2813	with priority higher than or equal to the event loop and one coroutine
1126	of lower priority, but only once, using idle watchers to keep the event	2814	of lower priority, but only once, using idle watchers to keep the event
1127	loop from blocking if lower-priority coroutines are active, thus mapping	2815	loop from blocking if lower-priority coroutines are active, thus mapping
1128	low-priority coroutines to idle/background tasks).	2816	low-priority coroutines to idle/background tasks).
1129		2817
		2818	It is recommended to give C<ev_check> watchers highest (C<EV_MAXPRI>)
		2819	priority, to ensure that they are being run before any other watchers
		2820	after the poll (this doesn't matter for C<ev_prepare> watchers).
		2821
		2822	Also, C<ev_check> watchers (and C<ev_prepare> watchers, too) should not
		2823	activate ("feed") events into libev. While libev fully supports this, they
		2824	might get executed before other C<ev_check> watchers did their job. As
		2825	C<ev_check> watchers are often used to embed other (non-libev) event
		2826	loops those other event loops might be in an unusable state until their
		2827	C<ev_check> watcher ran (always remind yourself to coexist peacefully with
		2828	others).
		2829
		2830	=head3 Watcher-Specific Functions and Data Members
		2831
1130	=over 4	2832	=over 4
1131		2833
1132	=item ev_prepare_init (ev_prepare *, callback)	2834	=item ev_prepare_init (ev_prepare *, callback)
1133		2835
1134	=item ev_check_init (ev_check *, callback)	2836	=item ev_check_init (ev_check *, callback)
1135		2837
1136	Initialises and configures the prepare or check watcher - they have no	2838	Initialises and configures the prepare or check watcher - they have no
1137	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>	2839	parameters of any kind. There are C<ev_prepare_set> and C<ev_check_set>
1138	macros, but using them is utterly, utterly and completely pointless.	2840	macros, but using them is utterly, utterly, utterly and completely
		2841	pointless.
1139		2842
1140	=back	2843	=back
1141		2844
1142	Example: *TODO*.	2845	=head3 Examples
		2846
		2847	There are a number of principal ways to embed other event loops or modules
		2848	into libev. Here are some ideas on how to include libadns into libev
		2849	(there is a Perl module named C<EV::ADNS> that does this, which you could
		2850	use as a working example. Another Perl module named C<EV::Glib> embeds a
		2851	Glib main context into libev, and finally, C<Glib::EV> embeds EV into the
		2852	Glib event loop).
		2853
		2854	Method 1: Add IO watchers and a timeout watcher in a prepare handler,
		2855	and in a check watcher, destroy them and call into libadns. What follows
		2856	is pseudo-code only of course. This requires you to either use a low
		2857	priority for the check watcher or use C<ev_clear_pending> explicitly, as
		2858	the callbacks for the IO/timeout watchers might not have been called yet.
		2859
		2860	static ev_io iow [nfd];
		2861	static ev_timer tw;
		2862
		2863	static void
		2864	io_cb (struct ev_loop *loop, ev_io *w, int revents)
		2865	{
		2866	}
		2867
		2868	// create io watchers for each fd and a timer before blocking
		2869	static void
		2870	adns_prepare_cb (struct ev_loop *loop, ev_prepare *w, int revents)
		2871	{
		2872	int timeout = 3600000;
		2873	struct pollfd fds [nfd];
		2874	// actual code will need to loop here and realloc etc.
		2875	adns_beforepoll (ads, fds, &nfd, &timeout, timeval_from (ev_time ()));
		2876
		2877	/* the callback is illegal, but won't be called as we stop during check */
		2878	ev_timer_init (&tw, 0, timeout * 1e-3, 0.);
		2879	ev_timer_start (loop, &tw);
		2880
		2881	// create one ev_io per pollfd
		2882	for (int i = 0; i < nfd; ++i)
		2883	{
		2884	ev_io_init (iow + i, io_cb, fds [i].fd,
		2885	((fds [i].events & POLLIN ? EV_READ : 0)
		2886	| (fds [i].events & POLLOUT ? EV_WRITE : 0)));
		2887
		2888	fds [i].revents = 0;
		2889	ev_io_start (loop, iow + i);
		2890	}
		2891	}
		2892
		2893	// stop all watchers after blocking
		2894	static void
		2895	adns_check_cb (struct ev_loop *loop, ev_check *w, int revents)
		2896	{
		2897	ev_timer_stop (loop, &tw);
		2898
		2899	for (int i = 0; i < nfd; ++i)
		2900	{
		2901	// set the relevant poll flags
		2902	// could also call adns_processreadable etc. here
		2903	struct pollfd *fd = fds + i;
		2904	int revents = ev_clear_pending (iow + i);
		2905	if (revents & EV_READ ) fd->revents |= fd->events & POLLIN;
		2906	if (revents & EV_WRITE) fd->revents |= fd->events & POLLOUT;
		2907
		2908	// now stop the watcher
		2909	ev_io_stop (loop, iow + i);
		2910	}
		2911
		2912	adns_afterpoll (adns, fds, nfd, timeval_from (ev_now (loop));
		2913	}
		2914
		2915	Method 2: This would be just like method 1, but you run C<adns_afterpoll>
		2916	in the prepare watcher and would dispose of the check watcher.
		2917
		2918	Method 3: If the module to be embedded supports explicit event
		2919	notification (libadns does), you can also make use of the actual watcher
		2920	callbacks, and only destroy/create the watchers in the prepare watcher.
		2921
		2922	static void
		2923	timer_cb (EV_P_ ev_timer *w, int revents)
		2924	{
		2925	adns_state ads = (adns_state)w->data;
		2926	update_now (EV_A);
		2927
		2928	adns_processtimeouts (ads, &tv_now);
		2929	}
		2930
		2931	static void
		2932	io_cb (EV_P_ ev_io *w, int revents)
		2933	{
		2934	adns_state ads = (adns_state)w->data;
		2935	update_now (EV_A);
		2936
		2937	if (revents & EV_READ ) adns_processreadable (ads, w->fd, &tv_now);
		2938	if (revents & EV_WRITE) adns_processwriteable (ads, w->fd, &tv_now);
		2939	}
		2940
		2941	// do not ever call adns_afterpoll
		2942
		2943	Method 4: Do not use a prepare or check watcher because the module you
		2944	want to embed is not flexible enough to support it. Instead, you can
		2945	override their poll function. The drawback with this solution is that the
		2946	main loop is now no longer controllable by EV. The C<Glib::EV> module uses
		2947	this approach, effectively embedding EV as a client into the horrible
		2948	libglib event loop.
		2949
		2950	static gint
		2951	event_poll_func (GPollFD *fds, guint nfds, gint timeout)
		2952	{
		2953	int got_events = 0;
		2954
		2955	for (n = 0; n < nfds; ++n)
		2956	// create/start io watcher that sets the relevant bits in fds[n] and increment got_events
		2957
		2958	if (timeout >= 0)
		2959	// create/start timer
		2960
		2961	// poll
		2962	ev_run (EV_A_ 0);
		2963
		2964	// stop timer again
		2965	if (timeout >= 0)
		2966	ev_timer_stop (EV_A_ &to);
		2967
		2968	// stop io watchers again - their callbacks should have set
		2969	for (n = 0; n < nfds; ++n)
		2970	ev_io_stop (EV_A_ iow [n]);
		2971
		2972	return got_events;
		2973	}
1143		2974
1144		2975
1145	=head2 C<ev_embed> - when one backend isn't enough...	2976	=head2 C<ev_embed> - when one backend isn't enough...
1146		2977
1147	This is a rather advanced watcher type that lets you embed one event loop	2978	This is a rather advanced watcher type that lets you embed one event loop
…		…
1153	prioritise I/O.	2984	prioritise I/O.
1154		2985
1155	As an example for a bug workaround, the kqueue backend might only support	2986	As an example for a bug workaround, the kqueue backend might only support
1156	sockets on some platform, so it is unusable as generic backend, but you	2987	sockets on some platform, so it is unusable as generic backend, but you
1157	still want to make use of it because you have many sockets and it scales	2988	still want to make use of it because you have many sockets and it scales
1158	so nicely. In this case, you would create a kqueue-based loop and embed it	2989	so nicely. In this case, you would create a kqueue-based loop and embed
1159	into your default loop (which might use e.g. poll). Overall operation will	2990	it into your default loop (which might use e.g. poll). Overall operation
1160	be a bit slower because first libev has to poll and then call kevent, but	2991	will be a bit slower because first libev has to call C<poll> and then
1161	at least you can use both at what they are best.	2992	C<kevent>, but at least you can use both mechanisms for what they are
		2993	best: C<kqueue> for scalable sockets and C<poll> if you want it to work :)
1162		2994
1163	As for prioritising I/O: rarely you have the case where some fds have	2995	As for prioritising I/O: under rare circumstances you have the case where
1164	to be watched and handled very quickly (with low latency), and even	2996	some fds have to be watched and handled very quickly (with low latency),
1165	priorities and idle watchers might have too much overhead. In this case	2997	and even priorities and idle watchers might have too much overhead. In
1166	you would put all the high priority stuff in one loop and all the rest in	2998	this case you would put all the high priority stuff in one loop and all
1167	a second one, and embed the second one in the first.	2999	the rest in a second one, and embed the second one in the first.
1168		3000
1169	As long as the watcher is active, the callback will be invoked every time	3001	As long as the watcher is active, the callback will be invoked every
1170	there might be events pending in the embedded loop. The callback must then	3002	time there might be events pending in the embedded loop. The callback
1171	call C<ev_embed_sweep (mainloop, watcher)> to make a single sweep and invoke	3003	must then call C<ev_embed_sweep (mainloop, watcher)> to make a single
1172	their callbacks (you could also start an idle watcher to give the embedded	3004	sweep and invoke their callbacks (the callback doesn't need to invoke the
1173	loop strictly lower priority for example). You can also set the callback	3005	C<ev_embed_sweep> function directly, it could also start an idle watcher
1174	to C<0>, in which case the embed watcher will automatically execute the	3006	to give the embedded loop strictly lower priority for example).
1175	embedded loop sweep.
1176		3007
1177	As long as the watcher is started it will automatically handle events. The	3008	You can also set the callback to C<0>, in which case the embed watcher
1178	callback will be invoked whenever some events have been handled. You can	3009	will automatically execute the embedded loop sweep whenever necessary.
1179	set the callback to C<0> to avoid having to specify one if you are not
1180	interested in that.
1181		3010
1182	Also, there have not currently been made special provisions for forking:	3011	Fork detection will be handled transparently while the C<ev_embed> watcher
1183	when you fork, you not only have to call C<ev_loop_fork> on both loops,	3012	is active, i.e., the embedded loop will automatically be forked when the
1184	but you will also have to stop and restart any C<ev_embed> watchers	3013	embedding loop forks. In other cases, the user is responsible for calling
1185	yourself.	3014	C<ev_loop_fork> on the embedded loop.
1186		3015
1187	Unfortunately, not all backends are embeddable, only the ones returned by	3016	Unfortunately, not all backends are embeddable: only the ones returned by
1188	C<ev_embeddable_backends> are, which, unfortunately, does not include any	3017	C<ev_embeddable_backends> are, which, unfortunately, does not include any
1189	portable one.	3018	portable one.
1190		3019
1191	So when you want to use this feature you will always have to be prepared	3020	So when you want to use this feature you will always have to be prepared
1192	that you cannot get an embeddable loop. The recommended way to get around	3021	that you cannot get an embeddable loop. The recommended way to get around
1193	this is to have a separate variables for your embeddable loop, try to	3022	this is to have a separate variables for your embeddable loop, try to
1194	create it, and if that fails, use the normal loop for everything:	3023	create it, and if that fails, use the normal loop for everything.
1195		3024
1196	struct ev_loop *loop_hi = ev_default_init (0);	3025	=head3 C<ev_embed> and fork
1197	struct ev_loop *loop_lo = 0;
1198	struct ev_embed embed;
1199
1200	// see if there is a chance of getting one that works
1201	// (remember that a flags value of 0 means autodetection)
1202	loop_lo = ev_embeddable_backends () & ev_recommended_backends ()
1203	? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())
1204	: 0;
1205		3026
1206	// if we got one, then embed it, otherwise default to loop_hi	3027	While the C<ev_embed> watcher is running, forks in the embedding loop will
1207	if (loop_lo)	3028	automatically be applied to the embedded loop as well, so no special
1208	{	3029	fork handling is required in that case. When the watcher is not running,
1209	ev_embed_init (&embed, 0, loop_lo);	3030	however, it is still the task of the libev user to call C<ev_loop_fork ()>
1210	ev_embed_start (loop_hi, &embed);	3031	as applicable.
1211	}	3032
1212	else	3033	=head3 Watcher-Specific Functions and Data Members
1213	loop_lo = loop_hi;
1214		3034
1215	=over 4	3035	=over 4
1216		3036
1217	=item ev_embed_init (ev_embed *, callback, struct ev_loop *embedded_loop)	3037	=item ev_embed_init (ev_embed *, callback, struct ev_loop *embedded_loop)
1218		3038
…		…
1220		3040
1221	Configures the watcher to embed the given loop, which must be	3041	Configures the watcher to embed the given loop, which must be
1222	embeddable. If the callback is C<0>, then C<ev_embed_sweep> will be	3042	embeddable. If the callback is C<0>, then C<ev_embed_sweep> will be
1223	invoked automatically, otherwise it is the responsibility of the callback	3043	invoked automatically, otherwise it is the responsibility of the callback
1224	to invoke it (it will continue to be called until the sweep has been done,	3044	to invoke it (it will continue to be called until the sweep has been done,
1225	if you do not want thta, you need to temporarily stop the embed watcher).	3045	if you do not want that, you need to temporarily stop the embed watcher).
1226		3046
1227	=item ev_embed_sweep (loop, ev_embed *)	3047	=item ev_embed_sweep (loop, ev_embed *)
1228		3048
1229	Make a single, non-blocking sweep over the embedded loop. This works	3049	Make a single, non-blocking sweep over the embedded loop. This works
1230	similarly to C<ev_loop (embedded_loop, EVLOOP_NONBLOCK)>, but in the most	3050	similarly to C<ev_run (embedded_loop, EVRUN_NOWAIT)>, but in the most
1231	apropriate way for embedded loops.	3051	appropriate way for embedded loops.
		3052
		3053	=item struct ev_loop *other [read-only]
		3054
		3055	The embedded event loop.
1232		3056
1233	=back	3057	=back
1234		3058
		3059	=head3 Examples
		3060
		3061	Example: Try to get an embeddable event loop and embed it into the default
		3062	event loop. If that is not possible, use the default loop. The default
		3063	loop is stored in C<loop_hi>, while the embeddable loop is stored in
		3064	C<loop_lo> (which is C<loop_hi> in the case no embeddable loop can be
		3065	used).
		3066
		3067	struct ev_loop *loop_hi = ev_default_init (0);
		3068	struct ev_loop *loop_lo = 0;
		3069	ev_embed embed;
		3070
		3071	// see if there is a chance of getting one that works
		3072	// (remember that a flags value of 0 means autodetection)
		3073	loop_lo = ev_embeddable_backends () & ev_recommended_backends ()
		3074	? ev_loop_new (ev_embeddable_backends () & ev_recommended_backends ())
		3075	: 0;
		3076
		3077	// if we got one, then embed it, otherwise default to loop_hi
		3078	if (loop_lo)
		3079	{
		3080	ev_embed_init (&embed, 0, loop_lo);
		3081	ev_embed_start (loop_hi, &embed);
		3082	}
		3083	else
		3084	loop_lo = loop_hi;
		3085
		3086	Example: Check if kqueue is available but not recommended and create
		3087	a kqueue backend for use with sockets (which usually work with any
		3088	kqueue implementation). Store the kqueue/socket-only event loop in
		3089	C<loop_socket>. (One might optionally use C<EVFLAG_NOENV>, too).
		3090
		3091	struct ev_loop *loop = ev_default_init (0);
		3092	struct ev_loop *loop_socket = 0;
		3093	ev_embed embed;
		3094
		3095	if (ev_supported_backends () & ~ev_recommended_backends () & EVBACKEND_KQUEUE)
		3096	if ((loop_socket = ev_loop_new (EVBACKEND_KQUEUE))
		3097	{
		3098	ev_embed_init (&embed, 0, loop_socket);
		3099	ev_embed_start (loop, &embed);
		3100	}
		3101
		3102	if (!loop_socket)
		3103	loop_socket = loop;
		3104
		3105	// now use loop_socket for all sockets, and loop for everything else
		3106
		3107
		3108	=head2 C<ev_fork> - the audacity to resume the event loop after a fork
		3109
		3110	Fork watchers are called when a C<fork ()> was detected (usually because
		3111	whoever is a good citizen cared to tell libev about it by calling
		3112	C<ev_default_fork> or C<ev_loop_fork>). The invocation is done before the
		3113	event loop blocks next and before C<ev_check> watchers are being called,
		3114	and only in the child after the fork. If whoever good citizen calling
		3115	C<ev_default_fork> cheats and calls it in the wrong process, the fork
		3116	handlers will be invoked, too, of course.
		3117
		3118	=head3 The special problem of life after fork - how is it possible?
		3119
		3120	Most uses of C<fork()> consist of forking, then some simple calls to set
		3121	up/change the process environment, followed by a call to C<exec()>. This
		3122	sequence should be handled by libev without any problems.
		3123
		3124	This changes when the application actually wants to do event handling
		3125	in the child, or both parent in child, in effect "continuing" after the
		3126	fork.
		3127
		3128	The default mode of operation (for libev, with application help to detect
		3129	forks) is to duplicate all the state in the child, as would be expected
		3130	when I<either> the parent I<or> the child process continues.
		3131
		3132	When both processes want to continue using libev, then this is usually the
		3133	wrong result. In that case, usually one process (typically the parent) is
		3134	supposed to continue with all watchers in place as before, while the other
		3135	process typically wants to start fresh, i.e. without any active watchers.
		3136
		3137	The cleanest and most efficient way to achieve that with libev is to
		3138	simply create a new event loop, which of course will be "empty", and
		3139	use that for new watchers. This has the advantage of not touching more
		3140	memory than necessary, and thus avoiding the copy-on-write, and the
		3141	disadvantage of having to use multiple event loops (which do not support
		3142	signal watchers).
		3143
		3144	When this is not possible, or you want to use the default loop for
		3145	other reasons, then in the process that wants to start "fresh", call
		3146	C<ev_loop_destroy (EV_DEFAULT)> followed by C<ev_default_loop (...)>.
		3147	Destroying the default loop will "orphan" (not stop) all registered
		3148	watchers, so you have to be careful not to execute code that modifies
		3149	those watchers. Note also that in that case, you have to re-register any
		3150	signal watchers.
		3151
		3152	=head3 Watcher-Specific Functions and Data Members
		3153
		3154	=over 4
		3155
		3156	=item ev_fork_init (ev_fork *, callback)
		3157
		3158	Initialises and configures the fork watcher - it has no parameters of any
		3159	kind. There is a C<ev_fork_set> macro, but using it is utterly pointless,
		3160	really.
		3161
		3162	=back
		3163
		3164
		3165	=head2 C<ev_cleanup> - even the best things end
		3166
		3167	Cleanup watchers are called just before the event loop is being destroyed
		3168	by a call to C<ev_loop_destroy>.
		3169
		3170	While there is no guarantee that the event loop gets destroyed, cleanup
		3171	watchers provide a convenient method to install cleanup hooks for your
		3172	program, worker threads and so on - you just to make sure to destroy the
		3173	loop when you want them to be invoked.
		3174
		3175	Cleanup watchers are invoked in the same way as any other watcher. Unlike
		3176	all other watchers, they do not keep a reference to the event loop (which
		3177	makes a lot of sense if you think about it). Like all other watchers, you
		3178	can call libev functions in the callback, except C<ev_cleanup_start>.
		3179
		3180	=head3 Watcher-Specific Functions and Data Members
		3181
		3182	=over 4
		3183
		3184	=item ev_cleanup_init (ev_cleanup *, callback)
		3185
		3186	Initialises and configures the cleanup watcher - it has no parameters of
		3187	any kind. There is a C<ev_cleanup_set> macro, but using it is utterly
		3188	pointless, I assure you.
		3189
		3190	=back
		3191
		3192	Example: Register an atexit handler to destroy the default loop, so any
		3193	cleanup functions are called.
		3194
		3195	static void
		3196	program_exits (void)
		3197	{
		3198	ev_loop_destroy (EV_DEFAULT_UC);
		3199	}
		3200
		3201	...
		3202	atexit (program_exits);
		3203
		3204
		3205	=head2 C<ev_async> - how to wake up an event loop
		3206
		3207	In general, you cannot use an C<ev_loop> from multiple threads or other
		3208	asynchronous sources such as signal handlers (as opposed to multiple event
		3209	loops - those are of course safe to use in different threads).
		3210
		3211	Sometimes, however, you need to wake up an event loop you do not control,
		3212	for example because it belongs to another thread. This is what C<ev_async>
		3213	watchers do: as long as the C<ev_async> watcher is active, you can signal
		3214	it by calling C<ev_async_send>, which is thread- and signal safe.
		3215
		3216	This functionality is very similar to C<ev_signal> watchers, as signals,
		3217	too, are asynchronous in nature, and signals, too, will be compressed
		3218	(i.e. the number of callback invocations may be less than the number of
		3219	C<ev_async_sent> calls). In fact, you could use signal watchers as a kind
		3220	of "global async watchers" by using a watcher on an otherwise unused
		3221	signal, and C<ev_feed_signal> to signal this watcher from another thread,
		3222	even without knowing which loop owns the signal.
		3223
		3224	Unlike C<ev_signal> watchers, C<ev_async> works with any event loop, not
		3225	just the default loop.
		3226
		3227	=head3 Queueing
		3228
		3229	C<ev_async> does not support queueing of data in any way. The reason
		3230	is that the author does not know of a simple (or any) algorithm for a
		3231	multiple-writer-single-reader queue that works in all cases and doesn't
		3232	need elaborate support such as pthreads or unportable memory access
		3233	semantics.
		3234
		3235	That means that if you want to queue data, you have to provide your own
		3236	queue. But at least I can tell you how to implement locking around your
		3237	queue:
		3238
		3239	=over 4
		3240
		3241	=item queueing from a signal handler context
		3242
		3243	To implement race-free queueing, you simply add to the queue in the signal
		3244	handler but you block the signal handler in the watcher callback. Here is
		3245	an example that does that for some fictitious SIGUSR1 handler:
		3246
		3247	static ev_async mysig;
		3248
		3249	static void
		3250	sigusr1_handler (void)
		3251	{
		3252	sometype data;
		3253
		3254	// no locking etc.
		3255	queue_put (data);
		3256	ev_async_send (EV_DEFAULT_ &mysig);
		3257	}
		3258
		3259	static void
		3260	mysig_cb (EV_P_ ev_async *w, int revents)
		3261	{
		3262	sometype data;
		3263	sigset_t block, prev;
		3264
		3265	sigemptyset (&block);
		3266	sigaddset (&block, SIGUSR1);
		3267	sigprocmask (SIG_BLOCK, &block, &prev);
		3268
		3269	while (queue_get (&data))
		3270	process (data);
		3271
		3272	if (sigismember (&prev, SIGUSR1)
		3273	sigprocmask (SIG_UNBLOCK, &block, 0);
		3274	}
		3275
		3276	(Note: pthreads in theory requires you to use C<pthread_setmask>
		3277	instead of C<sigprocmask> when you use threads, but libev doesn't do it
		3278	either...).
		3279
		3280	=item queueing from a thread context
		3281
		3282	The strategy for threads is different, as you cannot (easily) block
		3283	threads but you can easily preempt them, so to queue safely you need to
		3284	employ a traditional mutex lock, such as in this pthread example:
		3285
		3286	static ev_async mysig;
		3287	static pthread_mutex_t mymutex = PTHREAD_MUTEX_INITIALIZER;
		3288
		3289	static void
		3290	otherthread (void)
		3291	{
		3292	// only need to lock the actual queueing operation
		3293	pthread_mutex_lock (&mymutex);
		3294	queue_put (data);
		3295	pthread_mutex_unlock (&mymutex);
		3296
		3297	ev_async_send (EV_DEFAULT_ &mysig);
		3298	}
		3299
		3300	static void
		3301	mysig_cb (EV_P_ ev_async *w, int revents)
		3302	{
		3303	pthread_mutex_lock (&mymutex);
		3304
		3305	while (queue_get (&data))
		3306	process (data);
		3307
		3308	pthread_mutex_unlock (&mymutex);
		3309	}
		3310
		3311	=back
		3312
		3313
		3314	=head3 Watcher-Specific Functions and Data Members
		3315
		3316	=over 4
		3317
		3318	=item ev_async_init (ev_async *, callback)
		3319
		3320	Initialises and configures the async watcher - it has no parameters of any
		3321	kind. There is a C<ev_async_set> macro, but using it is utterly pointless,
		3322	trust me.
		3323
		3324	=item ev_async_send (loop, ev_async *)
		3325
		3326	Sends/signals/activates the given C<ev_async> watcher, that is, feeds
		3327	an C<EV_ASYNC> event on the watcher into the event loop, and instantly
		3328	returns.
		3329
		3330	Unlike C<ev_feed_event>, this call is safe to do from other threads,
		3331	signal or similar contexts (see the discussion of C<EV_ATOMIC_T> in the
		3332	embedding section below on what exactly this means).
		3333
		3334	Note that, as with other watchers in libev, multiple events might get
		3335	compressed into a single callback invocation (another way to look at this
		3336	is that C<ev_async> watchers are level-triggered, set on C<ev_async_send>,
		3337	reset when the event loop detects that).
		3338
		3339	This call incurs the overhead of a system call only once per event loop
		3340	iteration, so while the overhead might be noticeable, it doesn't apply to
		3341	repeated calls to C<ev_async_send> for the same event loop.
		3342
		3343	=item bool = ev_async_pending (ev_async *)
		3344
		3345	Returns a non-zero value when C<ev_async_send> has been called on the
		3346	watcher but the event has not yet been processed (or even noted) by the
		3347	event loop.
		3348
		3349	C<ev_async_send> sets a flag in the watcher and wakes up the loop. When
		3350	the loop iterates next and checks for the watcher to have become active,
		3351	it will reset the flag again. C<ev_async_pending> can be used to very
		3352	quickly check whether invoking the loop might be a good idea.
		3353
		3354	Not that this does I<not> check whether the watcher itself is pending,
		3355	only whether it has been requested to make this watcher pending: there
		3356	is a time window between the event loop checking and resetting the async
		3357	notification, and the callback being invoked.
		3358
		3359	=back
		3360
1235		3361
1236	=head1 OTHER FUNCTIONS	3362	=head1 OTHER FUNCTIONS
1237		3363
1238	There are some other functions of possible interest. Described. Here. Now.	3364	There are some other functions of possible interest. Described. Here. Now.
1239		3365
1240	=over 4	3366	=over 4
1241		3367
1242	=item ev_once (loop, int fd, int events, ev_tstamp timeout, callback)	3368	=item ev_once (loop, int fd, int events, ev_tstamp timeout, callback)
1243		3369
1244	This function combines a simple timer and an I/O watcher, calls your	3370	This function combines a simple timer and an I/O watcher, calls your
1245	callback on whichever event happens first and automatically stop both	3371	callback on whichever event happens first and automatically stops both
1246	watchers. This is useful if you want to wait for a single event on an fd	3372	watchers. This is useful if you want to wait for a single event on an fd
1247	or timeout without having to allocate/configure/start/stop/free one or	3373	or timeout without having to allocate/configure/start/stop/free one or
1248	more watchers yourself.	3374	more watchers yourself.
1249		3375
1250	If C<fd> is less than 0, then no I/O watcher will be started and events	3376	If C<fd> is less than 0, then no I/O watcher will be started and the
1251	is being ignored. Otherwise, an C<ev_io> watcher for the given C<fd> and	3377	C<events> argument is being ignored. Otherwise, an C<ev_io> watcher for
1252	C<events> set will be craeted and started.	3378	the given C<fd> and C<events> set will be created and started.
1253		3379
1254	If C<timeout> is less than 0, then no timeout watcher will be	3380	If C<timeout> is less than 0, then no timeout watcher will be
1255	started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and	3381	started. Otherwise an C<ev_timer> watcher with after = C<timeout> (and
1256	repeat = 0) will be started. While C<0> is a valid timeout, it is of	3382	repeat = 0) will be started. C<0> is a valid timeout.
1257	dubious value.
1258		3383
1259	The callback has the type C<void (*cb)(int revents, void *arg)> and gets	3384	The callback has the type C<void (*cb)(int revents, void *arg)> and is
1260	passed an C<revents> set like normal event callbacks (a combination of	3385	passed an C<revents> set like normal event callbacks (a combination of
1261	C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMEOUT>) and the C<arg>	3386	C<EV_ERROR>, C<EV_READ>, C<EV_WRITE> or C<EV_TIMER>) and the C<arg>
1262	value passed to C<ev_once>:	3387	value passed to C<ev_once>. Note that it is possible to receive I<both>
		3388	a timeout and an io event at the same time - you probably should give io
		3389	events precedence.
1263		3390
		3391	Example: wait up to ten seconds for data to appear on STDIN_FILENO.
		3392
1264	static void stdin_ready (int revents, void *arg)	3393	static void stdin_ready (int revents, void *arg)
1265	{	3394	{
1266	if (revents & EV_TIMEOUT)
1267	/* doh, nothing entered */;
1268	else if (revents & EV_READ)	3395	if (revents & EV_READ)
1269	/* stdin might have data for us, joy! */;	3396	/* stdin might have data for us, joy! */;
		3397	else if (revents & EV_TIMER)
		3398	/* doh, nothing entered */;
1270	}	3399	}
1271		3400
1272	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);	3401	ev_once (STDIN_FILENO, EV_READ, 10., stdin_ready, 0);
1273		3402
1274	=item ev_feed_event (ev_loop *, watcher *, int revents)
1275
1276	Feeds the given event set into the event loop, as if the specified event
1277	had happened for the specified watcher (which must be a pointer to an
1278	initialised but not necessarily started event watcher).
1279
1280	=item ev_feed_fd_event (ev_loop *, int fd, int revents)	3403	=item ev_feed_fd_event (loop, int fd, int revents)
1281		3404
1282	Feed an event on the given fd, as if a file descriptor backend detected	3405	Feed an event on the given fd, as if a file descriptor backend detected
1283	the given events it.	3406	the given events it.
1284		3407
1285	=item ev_feed_signal_event (ev_loop *loop, int signum)	3408	=item ev_feed_signal_event (loop, int signum)
1286		3409
1287	Feed an event as if the given signal occured (C<loop> must be the default	3410	Feed an event as if the given signal occurred. See also C<ev_feed_signal>,
1288	loop!).	3411	which is async-safe.
1289		3412
1290	=back	3413	=back
		3414
		3415
		3416	=head1 COMMON OR USEFUL IDIOMS (OR BOTH)
		3417
		3418	This section explains some common idioms that are not immediately
		3419	obvious. Note that examples are sprinkled over the whole manual, and this
		3420	section only contains stuff that wouldn't fit anywhere else.
		3421
		3422	=head2 ASSOCIATING CUSTOM DATA WITH A WATCHER
		3423
		3424	Each watcher has, by default, a C<void *data> member that you can read
		3425	or modify at any time: libev will completely ignore it. This can be used
		3426	to associate arbitrary data with your watcher. If you need more data and
		3427	don't want to allocate memory separately and store a pointer to it in that
		3428	data member, you can also "subclass" the watcher type and provide your own
		3429	data:
		3430
		3431	struct my_io
		3432	{
		3433	ev_io io;
		3434	int otherfd;
		3435	void *somedata;
		3436	struct whatever *mostinteresting;
		3437	};
		3438
		3439	...
		3440	struct my_io w;
		3441	ev_io_init (&w.io, my_cb, fd, EV_READ);
		3442
		3443	And since your callback will be called with a pointer to the watcher, you
		3444	can cast it back to your own type:
		3445
		3446	static void my_cb (struct ev_loop *loop, ev_io *w_, int revents)
		3447	{
		3448	struct my_io *w = (struct my_io *)w_;
		3449	...
		3450	}
		3451
		3452	More interesting and less C-conformant ways of casting your callback
		3453	function type instead have been omitted.
		3454
		3455	=head2 BUILDING YOUR OWN COMPOSITE WATCHERS
		3456
		3457	Another common scenario is to use some data structure with multiple
		3458	embedded watchers, in effect creating your own watcher that combines
		3459	multiple libev event sources into one "super-watcher":
		3460
		3461	struct my_biggy
		3462	{
		3463	int some_data;
		3464	ev_timer t1;
		3465	ev_timer t2;
		3466	}
		3467
		3468	In this case getting the pointer to C<my_biggy> is a bit more
		3469	complicated: Either you store the address of your C<my_biggy> struct in
		3470	the C<data> member of the watcher (for woozies or C++ coders), or you need
		3471	to use some pointer arithmetic using C<offsetof> inside your watchers (for
		3472	real programmers):
		3473
		3474	#include <stddef.h>
		3475
		3476	static void
		3477	t1_cb (EV_P_ ev_timer *w, int revents)
		3478	{
		3479	struct my_biggy big = (struct my_biggy *)
		3480	(((char *)w) - offsetof (struct my_biggy, t1));
		3481	}
		3482
		3483	static void
		3484	t2_cb (EV_P_ ev_timer *w, int revents)
		3485	{
		3486	struct my_biggy big = (struct my_biggy *)
		3487	(((char *)w) - offsetof (struct my_biggy, t2));
		3488	}
		3489
		3490	=head2 MODEL/NESTED EVENT LOOP INVOCATIONS AND EXIT CONDITIONS
		3491
		3492	Often (especially in GUI toolkits) there are places where you have
		3493	I<modal> interaction, which is most easily implemented by recursively
		3494	invoking C<ev_run>.
		3495
		3496	This brings the problem of exiting - a callback might want to finish the
		3497	main C<ev_run> call, but not the nested one (e.g. user clicked "Quit", but
		3498	a modal "Are you sure?" dialog is still waiting), or just the nested one
		3499	and not the main one (e.g. user clocked "Ok" in a modal dialog), or some
		3500	other combination: In these cases, C<ev_break> will not work alone.
		3501
		3502	The solution is to maintain "break this loop" variable for each C<ev_run>
		3503	invocation, and use a loop around C<ev_run> until the condition is
		3504	triggered, using C<EVRUN_ONCE>:
		3505
		3506	// main loop
		3507	int exit_main_loop = 0;
		3508
		3509	while (!exit_main_loop)
		3510	ev_run (EV_DEFAULT_ EVRUN_ONCE);
		3511
		3512	// in a model watcher
		3513	int exit_nested_loop = 0;
		3514
		3515	while (!exit_nested_loop)
		3516	ev_run (EV_A_ EVRUN_ONCE);
		3517
		3518	To exit from any of these loops, just set the corresponding exit variable:
		3519
		3520	// exit modal loop
		3521	exit_nested_loop = 1;
		3522
		3523	// exit main program, after modal loop is finished
		3524	exit_main_loop = 1;
		3525
		3526	// exit both
		3527	exit_main_loop = exit_nested_loop = 1;
		3528
		3529	=head2 THREAD LOCKING EXAMPLE
		3530
		3531	Here is a fictitious example of how to run an event loop in a different
		3532	thread from where callbacks are being invoked and watchers are
		3533	created/added/removed.
		3534
		3535	For a real-world example, see the C<EV::Loop::Async> perl module,
		3536	which uses exactly this technique (which is suited for many high-level
		3537	languages).
		3538
		3539	The example uses a pthread mutex to protect the loop data, a condition
		3540	variable to wait for callback invocations, an async watcher to notify the
		3541	event loop thread and an unspecified mechanism to wake up the main thread.
		3542
		3543	First, you need to associate some data with the event loop:
		3544
		3545	typedef struct {
		3546	mutex_t lock; /* global loop lock */
		3547	ev_async async_w;
		3548	thread_t tid;
		3549	cond_t invoke_cv;
		3550	} userdata;
		3551
		3552	void prepare_loop (EV_P)
		3553	{
		3554	// for simplicity, we use a static userdata struct.
		3555	static userdata u;
		3556
		3557	ev_async_init (&u->async_w, async_cb);
		3558	ev_async_start (EV_A_ &u->async_w);
		3559
		3560	pthread_mutex_init (&u->lock, 0);
		3561	pthread_cond_init (&u->invoke_cv, 0);
		3562
		3563	// now associate this with the loop
		3564	ev_set_userdata (EV_A_ u);
		3565	ev_set_invoke_pending_cb (EV_A_ l_invoke);
		3566	ev_set_loop_release_cb (EV_A_ l_release, l_acquire);
		3567
		3568	// then create the thread running ev_run
		3569	pthread_create (&u->tid, 0, l_run, EV_A);
		3570	}
		3571
		3572	The callback for the C<ev_async> watcher does nothing: the watcher is used
		3573	solely to wake up the event loop so it takes notice of any new watchers
		3574	that might have been added:
		3575
		3576	static void
		3577	async_cb (EV_P_ ev_async *w, int revents)
		3578	{
		3579	// just used for the side effects
		3580	}
		3581
		3582	The C<l_release> and C<l_acquire> callbacks simply unlock/lock the mutex
		3583	protecting the loop data, respectively.
		3584
		3585	static void
		3586	l_release (EV_P)
		3587	{
		3588	userdata *u = ev_userdata (EV_A);
		3589	pthread_mutex_unlock (&u->lock);
		3590	}
		3591
		3592	static void
		3593	l_acquire (EV_P)
		3594	{
		3595	userdata *u = ev_userdata (EV_A);
		3596	pthread_mutex_lock (&u->lock);
		3597	}
		3598
		3599	The event loop thread first acquires the mutex, and then jumps straight
		3600	into C<ev_run>:
		3601
		3602	void *
		3603	l_run (void *thr_arg)
		3604	{
		3605	struct ev_loop *loop = (struct ev_loop *)thr_arg;
		3606
		3607	l_acquire (EV_A);
		3608	pthread_setcanceltype (PTHREAD_CANCEL_ASYNCHRONOUS, 0);
		3609	ev_run (EV_A_ 0);
		3610	l_release (EV_A);
		3611
		3612	return 0;
		3613	}
		3614
		3615	Instead of invoking all pending watchers, the C<l_invoke> callback will
		3616	signal the main thread via some unspecified mechanism (signals? pipe
		3617	writes? C<Async::Interrupt>?) and then waits until all pending watchers
		3618	have been called (in a while loop because a) spurious wakeups are possible
		3619	and b) skipping inter-thread-communication when there are no pending
		3620	watchers is very beneficial):
		3621
		3622	static void
		3623	l_invoke (EV_P)
		3624	{
		3625	userdata *u = ev_userdata (EV_A);
		3626
		3627	while (ev_pending_count (EV_A))
		3628	{
		3629	wake_up_other_thread_in_some_magic_or_not_so_magic_way ();
		3630	pthread_cond_wait (&u->invoke_cv, &u->lock);
		3631	}
		3632	}
		3633
		3634	Now, whenever the main thread gets told to invoke pending watchers, it
		3635	will grab the lock, call C<ev_invoke_pending> and then signal the loop
		3636	thread to continue:
		3637
		3638	static void
		3639	real_invoke_pending (EV_P)
		3640	{
		3641	userdata *u = ev_userdata (EV_A);
		3642
		3643	pthread_mutex_lock (&u->lock);
		3644	ev_invoke_pending (EV_A);
		3645	pthread_cond_signal (&u->invoke_cv);
		3646	pthread_mutex_unlock (&u->lock);
		3647	}
		3648
		3649	Whenever you want to start/stop a watcher or do other modifications to an
		3650	event loop, you will now have to lock:
		3651
		3652	ev_timer timeout_watcher;
		3653	userdata *u = ev_userdata (EV_A);
		3654
		3655	ev_timer_init (&timeout_watcher, timeout_cb, 5.5, 0.);
		3656
		3657	pthread_mutex_lock (&u->lock);
		3658	ev_timer_start (EV_A_ &timeout_watcher);
		3659	ev_async_send (EV_A_ &u->async_w);
		3660	pthread_mutex_unlock (&u->lock);
		3661
		3662	Note that sending the C<ev_async> watcher is required because otherwise
		3663	an event loop currently blocking in the kernel will have no knowledge
		3664	about the newly added timer. By waking up the loop it will pick up any new
		3665	watchers in the next event loop iteration.
		3666
		3667	=head2 THREADS, COROUTINES, CONTINUATIONS, QUEUES... INSTEAD OF CALLBACKS
		3668
		3669	While the overhead of a callback that e.g. schedules a thread is small, it
		3670	is still an overhead. If you embed libev, and your main usage is with some
		3671	kind of threads or coroutines, you might want to customise libev so that
		3672	doesn't need callbacks anymore.
		3673
		3674	Imagine you have coroutines that you can switch to using a function
		3675	C<switch_to (coro)>, that libev runs in a coroutine called C<libev_coro>
		3676	and that due to some magic, the currently active coroutine is stored in a
		3677	global called C<current_coro>. Then you can build your own "wait for libev
		3678	event" primitive by changing C<EV_CB_DECLARE> and C<EV_CB_INVOKE> (note
		3679	the differing C<;> conventions):
		3680
		3681	#define EV_CB_DECLARE(type) struct my_coro *cb;
		3682	#define EV_CB_INVOKE(watcher) switch_to ((watcher)->cb)
		3683
		3684	That means instead of having a C callback function, you store the
		3685	coroutine to switch to in each watcher, and instead of having libev call
		3686	your callback, you instead have it switch to that coroutine.
		3687
		3688	A coroutine might now wait for an event with a function called
		3689	C<wait_for_event>. (the watcher needs to be started, as always, but it doesn't
		3690	matter when, or whether the watcher is active or not when this function is
		3691	called):
		3692
		3693	void
		3694	wait_for_event (ev_watcher *w)
		3695	{
		3696	ev_cb_set (w) = current_coro;
		3697	switch_to (libev_coro);
		3698	}
		3699
		3700	That basically suspends the coroutine inside C<wait_for_event> and
		3701	continues the libev coroutine, which, when appropriate, switches back to
		3702	this or any other coroutine. I am sure if you sue this your own :)
		3703
		3704	You can do similar tricks if you have, say, threads with an event queue -
		3705	instead of storing a coroutine, you store the queue object and instead of
		3706	switching to a coroutine, you push the watcher onto the queue and notify
		3707	any waiters.
		3708
		3709	To embed libev, see L<EMBEDDING>, but in short, it's easiest to create two
		3710	files, F<my_ev.h> and F<my_ev.c> that include the respective libev files:
		3711
		3712	// my_ev.h
		3713	#define EV_CB_DECLARE(type) struct my_coro *cb;
		3714	#define EV_CB_INVOKE(watcher) switch_to ((watcher)->cb);
		3715	#include "../libev/ev.h"
		3716
		3717	// my_ev.c
		3718	#define EV_H "my_ev.h"
		3719	#include "../libev/ev.c"
		3720
		3721	And then use F<my_ev.h> when you would normally use F<ev.h>, and compile
		3722	F<my_ev.c> into your project. When properly specifying include paths, you
		3723	can even use F<ev.h> as header file name directly.
1291		3724
1292		3725
1293	=head1 LIBEVENT EMULATION	3726	=head1 LIBEVENT EMULATION
1294		3727
1295	Libev offers a compatibility emulation layer for libevent. It cannot	3728	Libev offers a compatibility emulation layer for libevent. It cannot
1296	emulate the internals of libevent, so here are some usage hints:	3729	emulate the internals of libevent, so here are some usage hints:
1297		3730
1298	=over 4	3731	=over 4
		3732
		3733	=item * Only the libevent-1.4.1-beta API is being emulated.
		3734
		3735	This was the newest libevent version available when libev was implemented,
		3736	and is still mostly unchanged in 2010.
1299		3737
1300	=item * Use it by including <event.h>, as usual.	3738	=item * Use it by including <event.h>, as usual.
1301		3739
1302	=item * The following members are fully supported: ev_base, ev_callback,	3740	=item * The following members are fully supported: ev_base, ev_callback,
1303	ev_arg, ev_fd, ev_res, ev_events.	3741	ev_arg, ev_fd, ev_res, ev_events.
…		…
1308		3746
1309	=item * Priorities are not currently supported. Initialising priorities	3747	=item * Priorities are not currently supported. Initialising priorities
1310	will fail and all watchers will have the same priority, even though there	3748	will fail and all watchers will have the same priority, even though there
1311	is an ev_pri field.	3749	is an ev_pri field.
1312		3750
		3751	=item * In libevent, the last base created gets the signals, in libev, the
		3752	base that registered the signal gets the signals.
		3753
1313	=item * Other members are not supported.	3754	=item * Other members are not supported.
1314		3755
1315	=item * The libev emulation is I<not> ABI compatible to libevent, you need	3756	=item * The libev emulation is I<not> ABI compatible to libevent, you need
1316	to use the libev header file and library.	3757	to use the libev header file and library.
1317		3758
1318	=back	3759	=back
1319		3760
1320	=head1 C++ SUPPORT	3761	=head1 C++ SUPPORT
1321		3762
1322	Libev comes with some simplistic wrapper classes for C++ that mainly allow	3763	Libev comes with some simplistic wrapper classes for C++ that mainly allow
1323	you to use some convinience methods to start/stop watchers and also change	3764	you to use some convenience methods to start/stop watchers and also change
1324	the callback model to a model using method callbacks on objects.	3765	the callback model to a model using method callbacks on objects.
1325		3766
1326	To use it,	3767	To use it,
1327		3768
1328	#include <ev++.h>	3769	#include <ev++.h>
1329		3770
1330	(it is not installed by default). This automatically includes F<ev.h>	3771	This automatically includes F<ev.h> and puts all of its definitions (many
1331	and puts all of its definitions (many of them macros) into the global	3772	of them macros) into the global namespace. All C++ specific things are
1332	namespace. All C++ specific things are put into the C<ev> namespace.	3773	put into the C<ev> namespace. It should support all the same embedding
		3774	options as F<ev.h>, most notably C<EV_MULTIPLICITY>.
1333		3775
1334	It should support all the same embedding options as F<ev.h>, most notably	3776	Care has been taken to keep the overhead low. The only data member the C++
1335	C<EV_MULTIPLICITY>.	3777	classes add (compared to plain C-style watchers) is the event loop pointer
		3778	that the watcher is associated with (or no additional members at all if
		3779	you disable C<EV_MULTIPLICITY> when embedding libev).
		3780
		3781	Currently, functions, static and non-static member functions and classes
		3782	with C<operator ()> can be used as callbacks. Other types should be easy
		3783	to add as long as they only need one additional pointer for context. If
		3784	you need support for other types of functors please contact the author
		3785	(preferably after implementing it).
1336		3786
1337	Here is a list of things available in the C<ev> namespace:	3787	Here is a list of things available in the C<ev> namespace:
1338		3788
1339	=over 4	3789	=over 4
1340		3790
…		…
1356		3806
1357	All of those classes have these methods:	3807	All of those classes have these methods:
1358		3808
1359	=over 4	3809	=over 4
1360		3810
1361	=item ev::TYPE::TYPE (object *, object::method *)	3811	=item ev::TYPE::TYPE ()
1362		3812
1363	=item ev::TYPE::TYPE (object *, object::method *, struct ev_loop *)	3813	=item ev::TYPE::TYPE (loop)
1364		3814
1365	=item ev::TYPE::~TYPE	3815	=item ev::TYPE::~TYPE
1366		3816
1367	The constructor takes a pointer to an object and a method pointer to	3817	The constructor (optionally) takes an event loop to associate the watcher
1368	the event handler callback to call in this class. The constructor calls	3818	with. If it is omitted, it will use C<EV_DEFAULT>.
1369	C<ev_init> for you, which means you have to call the C<set> method	3819
1370	before starting it. If you do not specify a loop then the constructor	3820	The constructor calls C<ev_init> for you, which means you have to call the
1371	automatically associates the default loop with this watcher.	3821	C<set> method before starting it.
		3822
		3823	It will not set a callback, however: You have to call the templated C<set>
		3824	method to set a callback before you can start the watcher.
		3825
		3826	(The reason why you have to use a method is a limitation in C++ which does
		3827	not allow explicit template arguments for constructors).
1372		3828
1373	The destructor automatically stops the watcher if it is active.	3829	The destructor automatically stops the watcher if it is active.
1374		3830
		3831	=item w->set<class, &class::method> (object *)
		3832
		3833	This method sets the callback method to call. The method has to have a
		3834	signature of C<void (*)(ev_TYPE &, int)>, it receives the watcher as
		3835	first argument and the C<revents> as second. The object must be given as
		3836	parameter and is stored in the C<data> member of the watcher.
		3837
		3838	This method synthesizes efficient thunking code to call your method from
		3839	the C callback that libev requires. If your compiler can inline your
		3840	callback (i.e. it is visible to it at the place of the C<set> call and
		3841	your compiler is good :), then the method will be fully inlined into the
		3842	thunking function, making it as fast as a direct C callback.
		3843
		3844	Example: simple class declaration and watcher initialisation
		3845
		3846	struct myclass
		3847	{
		3848	void io_cb (ev::io &w, int revents) { }
		3849	}
		3850
		3851	myclass obj;
		3852	ev::io iow;
		3853	iow.set <myclass, &myclass::io_cb> (&obj);
		3854
		3855	=item w->set (object *)
		3856
		3857	This is a variation of a method callback - leaving out the method to call
		3858	will default the method to C<operator ()>, which makes it possible to use
		3859	functor objects without having to manually specify the C<operator ()> all
		3860	the time. Incidentally, you can then also leave out the template argument
		3861	list.
		3862
		3863	The C<operator ()> method prototype must be C<void operator ()(watcher &w,
		3864	int revents)>.
		3865
		3866	See the method-C<set> above for more details.
		3867
		3868	Example: use a functor object as callback.
		3869
		3870	struct myfunctor
		3871	{
		3872	void operator() (ev::io &w, int revents)
		3873	{
		3874	...
		3875	}
		3876	}
		3877
		3878	myfunctor f;
		3879
		3880	ev::io w;
		3881	w.set (&f);
		3882
		3883	=item w->set<function> (void *data = 0)
		3884
		3885	Also sets a callback, but uses a static method or plain function as
		3886	callback. The optional C<data> argument will be stored in the watcher's
		3887	C<data> member and is free for you to use.
		3888
		3889	The prototype of the C<function> must be C<void (*)(ev::TYPE &w, int)>.
		3890
		3891	See the method-C<set> above for more details.
		3892
		3893	Example: Use a plain function as callback.
		3894
		3895	static void io_cb (ev::io &w, int revents) { }
		3896	iow.set <io_cb> ();
		3897
1375	=item w->set (struct ev_loop *)	3898	=item w->set (loop)
1376		3899
1377	Associates a different C<struct ev_loop> with this watcher. You can only	3900	Associates a different C<struct ev_loop> with this watcher. You can only
1378	do this when the watcher is inactive (and not pending either).	3901	do this when the watcher is inactive (and not pending either).
1379		3902
1380	=item w->set ([args])	3903	=item w->set ([arguments])
1381		3904
1382	Basically the same as C<ev_TYPE_set>, with the same args. Must be	3905	Basically the same as C<ev_TYPE_set>, with the same arguments. Either this
1383	called at least once. Unlike the C counterpart, an active watcher gets	3906	method or a suitable start method must be called at least once. Unlike the
1384	automatically stopped and restarted.	3907	C counterpart, an active watcher gets automatically stopped and restarted
		3908	when reconfiguring it with this method.
1385		3909
1386	=item w->start ()	3910	=item w->start ()
1387		3911
1388	Starts the watcher. Note that there is no C<loop> argument as the	3912	Starts the watcher. Note that there is no C<loop> argument, as the
1389	constructor already takes the loop.	3913	constructor already stores the event loop.
		3914
		3915	=item w->start ([arguments])
		3916
		3917	Instead of calling C<set> and C<start> methods separately, it is often
		3918	convenient to wrap them in one call. Uses the same type of arguments as
		3919	the configure C<set> method of the watcher.
1390		3920
1391	=item w->stop ()	3921	=item w->stop ()
1392		3922
1393	Stops the watcher if it is active. Again, no C<loop> argument.	3923	Stops the watcher if it is active. Again, no C<loop> argument.
1394		3924
1395	=item w->again () C<ev::timer>, C<ev::periodic> only	3925	=item w->again () (C<ev::timer>, C<ev::periodic> only)
1396		3926
1397	For C<ev::timer> and C<ev::periodic>, this invokes the corresponding	3927	For C<ev::timer> and C<ev::periodic>, this invokes the corresponding
1398	C<ev_TYPE_again> function.	3928	C<ev_TYPE_again> function.
1399		3929
1400	=item w->sweep () C<ev::embed> only	3930	=item w->sweep () (C<ev::embed> only)
1401		3931
1402	Invokes C<ev_embed_sweep>.	3932	Invokes C<ev_embed_sweep>.
1403		3933
		3934	=item w->update () (C<ev::stat> only)
		3935
		3936	Invokes C<ev_stat_stat>.
		3937
1404	=back	3938	=back
1405		3939
1406	=back	3940	=back
1407		3941
1408	Example: Define a class with an IO and idle watcher, start one of them in	3942	Example: Define a class with two I/O and idle watchers, start the I/O
1409	the constructor.	3943	watchers in the constructor.
1410		3944
1411	class myclass	3945	class myclass
1412	{	3946	{
1413	ev_io io; void io_cb (ev::io &w, int revents);	3947	ev::io io ; void io_cb (ev::io &w, int revents);
		3948	ev::io2 io2 ; void io2_cb (ev::io &w, int revents);
1414	ev_idle idle void idle_cb (ev::idle &w, int revents);	3949	ev::idle idle; void idle_cb (ev::idle &w, int revents);
1415		3950
1416	myclass ();	3951	myclass (int fd)
		3952	{
		3953	io .set <myclass, &myclass::io_cb > (this);
		3954	io2 .set <myclass, &myclass::io2_cb > (this);
		3955	idle.set <myclass, &myclass::idle_cb> (this);
		3956
		3957	io.set (fd, ev::WRITE); // configure the watcher
		3958	io.start (); // start it whenever convenient
		3959
		3960	io2.start (fd, ev::READ); // set + start in one call
		3961	}
1417	}	3962	};
1418		3963
1419	myclass::myclass (int fd)	3964
1420	: io (this, &myclass::io_cb),	3965	=head1 OTHER LANGUAGE BINDINGS
1421	idle (this, &myclass::idle_cb)	3966
		3967	Libev does not offer other language bindings itself, but bindings for a
		3968	number of languages exist in the form of third-party packages. If you know
		3969	any interesting language binding in addition to the ones listed here, drop
		3970	me a note.
		3971
		3972	=over 4
		3973
		3974	=item Perl
		3975
		3976	The EV module implements the full libev API and is actually used to test
		3977	libev. EV is developed together with libev. Apart from the EV core module,
		3978	there are additional modules that implement libev-compatible interfaces
		3979	to C<libadns> (C<EV::ADNS>, but C<AnyEvent::DNS> is preferred nowadays),
		3980	C<Net::SNMP> (C<Net::SNMP::EV>) and the C<libglib> event core (C<Glib::EV>
		3981	and C<EV::Glib>).
		3982
		3983	It can be found and installed via CPAN, its homepage is at
		3984	L<http://software.schmorp.de/pkg/EV>.
		3985
		3986	=item Python
		3987
		3988	Python bindings can be found at L<http://code.google.com/p/pyev/>. It
		3989	seems to be quite complete and well-documented.
		3990
		3991	=item Ruby
		3992
		3993	Tony Arcieri has written a ruby extension that offers access to a subset
		3994	of the libev API and adds file handle abstractions, asynchronous DNS and
		3995	more on top of it. It can be found via gem servers. Its homepage is at
		3996	L<http://rev.rubyforge.org/>.
		3997
		3998	Roger Pack reports that using the link order C<-lws2_32 -lmsvcrt-ruby-190>
		3999	makes rev work even on mingw.
		4000
		4001	=item Haskell
		4002
		4003	A haskell binding to libev is available at
		4004	L<http://hackage.haskell.org/cgi-bin/hackage-scripts/package/hlibev>.
		4005
		4006	=item D
		4007
		4008	Leandro Lucarella has written a D language binding (F<ev.d>) for libev, to
		4009	be found at L<http://proj.llucax.com.ar/wiki/evd>.
		4010
		4011	=item Ocaml
		4012
		4013	Erkki Seppala has written Ocaml bindings for libev, to be found at
		4014	L<http://modeemi.cs.tut.fi/~flux/software/ocaml-ev/>.
		4015
		4016	=item Lua
		4017
		4018	Brian Maher has written a partial interface to libev for lua (at the
		4019	time of this writing, only C<ev_io> and C<ev_timer>), to be found at
		4020	L<http://github.com/brimworks/lua-ev>.
		4021
		4022	=back
		4023
		4024
		4025	=head1 MACRO MAGIC
		4026
		4027	Libev can be compiled with a variety of options, the most fundamental
		4028	of which is C<EV_MULTIPLICITY>. This option determines whether (most)
		4029	functions and callbacks have an initial C<struct ev_loop *> argument.
		4030
		4031	To make it easier to write programs that cope with either variant, the
		4032	following macros are defined:
		4033
		4034	=over 4
		4035
		4036	=item C<EV_A>, C<EV_A_>
		4037
		4038	This provides the loop I<argument> for functions, if one is required ("ev
		4039	loop argument"). The C<EV_A> form is used when this is the sole argument,
		4040	C<EV_A_> is used when other arguments are following. Example:
		4041
		4042	ev_unref (EV_A);
		4043	ev_timer_add (EV_A_ watcher);
		4044	ev_run (EV_A_ 0);
		4045
		4046	It assumes the variable C<loop> of type C<struct ev_loop *> is in scope,
		4047	which is often provided by the following macro.
		4048
		4049	=item C<EV_P>, C<EV_P_>
		4050
		4051	This provides the loop I<parameter> for functions, if one is required ("ev
		4052	loop parameter"). The C<EV_P> form is used when this is the sole parameter,
		4053	C<EV_P_> is used when other parameters are following. Example:
		4054
		4055	// this is how ev_unref is being declared
		4056	static void ev_unref (EV_P);
		4057
		4058	// this is how you can declare your typical callback
		4059	static void cb (EV_P_ ev_timer *w, int revents)
		4060
		4061	It declares a parameter C<loop> of type C<struct ev_loop *>, quite
		4062	suitable for use with C<EV_A>.
		4063
		4064	=item C<EV_DEFAULT>, C<EV_DEFAULT_>
		4065
		4066	Similar to the other two macros, this gives you the value of the default
		4067	loop, if multiple loops are supported ("ev loop default").
		4068
		4069	=item C<EV_DEFAULT_UC>, C<EV_DEFAULT_UC_>
		4070
		4071	Usage identical to C<EV_DEFAULT> and C<EV_DEFAULT_>, but requires that the
		4072	default loop has been initialised (C<UC> == unchecked). Their behaviour
		4073	is undefined when the default loop has not been initialised by a previous
		4074	execution of C<EV_DEFAULT>, C<EV_DEFAULT_> or C<ev_default_init (...)>.
		4075
		4076	It is often prudent to use C<EV_DEFAULT> when initialising the first
		4077	watcher in a function but use C<EV_DEFAULT_UC> afterwards.
		4078
		4079	=back
		4080
		4081	Example: Declare and initialise a check watcher, utilising the above
		4082	macros so it will work regardless of whether multiple loops are supported
		4083	or not.
		4084
		4085	static void
		4086	check_cb (EV_P_ ev_timer *w, int revents)
1422	{	4087	{
1423	io.start (fd, ev::READ);	4088	ev_check_stop (EV_A_ w);
1424	}	4089	}
		4090
		4091	ev_check check;
		4092	ev_check_init (&check, check_cb);
		4093	ev_check_start (EV_DEFAULT_ &check);
		4094	ev_run (EV_DEFAULT_ 0);
1425		4095
1426	=head1 EMBEDDING	4096	=head1 EMBEDDING
1427		4097
1428	Libev can (and often is) directly embedded into host	4098	Libev can (and often is) directly embedded into host
1429	applications. Examples of applications that embed it include the Deliantra	4099	applications. Examples of applications that embed it include the Deliantra
1430	Game Server, the EV perl module, the GNU Virtual Private Ethernet (gvpe)	4100	Game Server, the EV perl module, the GNU Virtual Private Ethernet (gvpe)
1431	and rxvt-unicode.	4101	and rxvt-unicode.
1432		4102
1433	The goal is to enable you to just copy the neecssary files into your	4103	The goal is to enable you to just copy the necessary files into your
1434	source directory without having to change even a single line in them, so	4104	source directory without having to change even a single line in them, so
1435	you can easily upgrade by simply copying (or having a checked-out copy of	4105	you can easily upgrade by simply copying (or having a checked-out copy of
1436	libev somewhere in your source tree).	4106	libev somewhere in your source tree).
1437		4107
1438	=head2 FILESETS	4108	=head2 FILESETS
1439		4109
1440	Depending on what features you need you need to include one or more sets of files	4110	Depending on what features you need you need to include one or more sets of files
1441	in your app.	4111	in your application.
1442		4112
1443	=head3 CORE EVENT LOOP	4113	=head3 CORE EVENT LOOP
1444		4114
1445	To include only the libev core (all the C<ev_*> functions), with manual	4115	To include only the libev core (all the C<ev_*> functions), with manual
1446	configuration (no autoconf):	4116	configuration (no autoconf):
1447		4117
1448	#define EV_STANDALONE 1	4118	#define EV_STANDALONE 1
1449	#include "ev.c"	4119	#include "ev.c"
1450		4120
1451	This will automatically include F<ev.h>, too, and should be done in a	4121	This will automatically include F<ev.h>, too, and should be done in a
1452	single C source file only to provide the function implementations. To use	4122	single C source file only to provide the function implementations. To use
1453	it, do the same for F<ev.h> in all files wishing to use this API (best	4123	it, do the same for F<ev.h> in all files wishing to use this API (best
1454	done by writing a wrapper around F<ev.h> that you can include instead and	4124	done by writing a wrapper around F<ev.h> that you can include instead and
1455	where you can put other configuration options):	4125	where you can put other configuration options):
1456		4126
1457	#define EV_STANDALONE 1	4127	#define EV_STANDALONE 1
1458	#include "ev.h"	4128	#include "ev.h"
1459		4129
1460	Both header files and implementation files can be compiled with a C++	4130	Both header files and implementation files can be compiled with a C++
1461	compiler (at least, thats a stated goal, and breakage will be treated	4131	compiler (at least, that's a stated goal, and breakage will be treated
1462	as a bug).	4132	as a bug).
1463		4133
1464	You need the following files in your source tree, or in a directory	4134	You need the following files in your source tree, or in a directory
1465	in your include path (e.g. in libev/ when using -Ilibev):	4135	in your include path (e.g. in libev/ when using -Ilibev):
1466		4136
1467	ev.h	4137	ev.h
1468	ev.c	4138	ev.c
1469	ev_vars.h	4139	ev_vars.h
1470	ev_wrap.h	4140	ev_wrap.h
1471		4141
1472	ev_win32.c required on win32 platforms only	4142	ev_win32.c required on win32 platforms only
1473		4143
1474	ev_select.c only when select backend is enabled (which is is by default)	4144	ev_select.c only when select backend is enabled (which is enabled by default)
1475	ev_poll.c only when poll backend is enabled (disabled by default)	4145	ev_poll.c only when poll backend is enabled (disabled by default)
1476	ev_epoll.c only when the epoll backend is enabled (disabled by default)	4146	ev_epoll.c only when the epoll backend is enabled (disabled by default)
1477	ev_kqueue.c only when the kqueue backend is enabled (disabled by default)	4147	ev_kqueue.c only when the kqueue backend is enabled (disabled by default)
1478	ev_port.c only when the solaris port backend is enabled (disabled by default)	4148	ev_port.c only when the solaris port backend is enabled (disabled by default)
1479		4149
1480	F<ev.c> includes the backend files directly when enabled, so you only need	4150	F<ev.c> includes the backend files directly when enabled, so you only need
1481	to compile a single file.	4151	to compile this single file.
1482		4152
1483	=head3 LIBEVENT COMPATIBILITY API	4153	=head3 LIBEVENT COMPATIBILITY API
1484		4154
1485	To include the libevent compatibility API, also include:	4155	To include the libevent compatibility API, also include:
1486		4156
1487	#include "event.c"	4157	#include "event.c"
1488		4158
1489	in the file including F<ev.c>, and:	4159	in the file including F<ev.c>, and:
1490		4160
1491	#include "event.h"	4161	#include "event.h"
1492		4162
1493	in the files that want to use the libevent API. This also includes F<ev.h>.	4163	in the files that want to use the libevent API. This also includes F<ev.h>.
1494		4164
1495	You need the following additional files for this:	4165	You need the following additional files for this:
1496		4166
1497	event.h	4167	event.h
1498	event.c	4168	event.c
1499		4169
1500	=head3 AUTOCONF SUPPORT	4170	=head3 AUTOCONF SUPPORT
1501		4171
1502	Instead of using C<EV_STANDALONE=1> and providing your config in	4172	Instead of using C<EV_STANDALONE=1> and providing your configuration in
1503	whatever way you want, you can also C<m4_include([libev.m4])> in your	4173	whatever way you want, you can also C<m4_include([libev.m4])> in your
1504	F<configure.ac> and leave C<EV_STANDALONE> off. F<ev.c> will then include	4174	F<configure.ac> and leave C<EV_STANDALONE> undefined. F<ev.c> will then
1505	F<config.h> and configure itself accordingly.	4175	include F<config.h> and configure itself accordingly.
1506		4176
1507	For this of course you need the m4 file:	4177	For this of course you need the m4 file:
1508		4178
1509	libev.m4	4179	libev.m4
1510		4180
1511	=head2 PREPROCESSOR SYMBOLS/MACROS	4181	=head2 PREPROCESSOR SYMBOLS/MACROS
1512		4182
1513	Libev can be configured via a variety of preprocessor symbols you have to define	4183	Libev can be configured via a variety of preprocessor symbols you have to
1514	before including any of its files. The default is not to build for multiplicity	4184	define before including (or compiling) any of its files. The default in
1515	and only include the select backend.	4185	the absence of autoconf is documented for every option.
		4186
		4187	Symbols marked with "(h)" do not change the ABI, and can have different
		4188	values when compiling libev vs. including F<ev.h>, so it is permissible
		4189	to redefine them before including F<ev.h> without breaking compatibility
		4190	to a compiled library. All other symbols change the ABI, which means all
		4191	users of libev and the libev code itself must be compiled with compatible
		4192	settings.
1516		4193
1517	=over 4	4194	=over 4
1518		4195
		4196	=item EV_COMPAT3 (h)
		4197
		4198	Backwards compatibility is a major concern for libev. This is why this
		4199	release of libev comes with wrappers for the functions and symbols that
		4200	have been renamed between libev version 3 and 4.
		4201
		4202	You can disable these wrappers (to test compatibility with future
		4203	versions) by defining C<EV_COMPAT3> to C<0> when compiling your
		4204	sources. This has the additional advantage that you can drop the C<struct>
		4205	from C<struct ev_loop> declarations, as libev will provide an C<ev_loop>
		4206	typedef in that case.
		4207
		4208	In some future version, the default for C<EV_COMPAT3> will become C<0>,
		4209	and in some even more future version the compatibility code will be
		4210	removed completely.
		4211
1519	=item EV_STANDALONE	4212	=item EV_STANDALONE (h)
1520		4213
1521	Must always be C<1> if you do not use autoconf configuration, which	4214	Must always be C<1> if you do not use autoconf configuration, which
1522	keeps libev from including F<config.h>, and it also defines dummy	4215	keeps libev from including F<config.h>, and it also defines dummy
1523	implementations for some libevent functions (such as logging, which is not	4216	implementations for some libevent functions (such as logging, which is not
1524	supported). It will also not define any of the structs usually found in	4217	supported). It will also not define any of the structs usually found in
1525	F<event.h> that are not directly supported by the libev core alone.	4218	F<event.h> that are not directly supported by the libev core alone.
1526		4219
		4220	In standalone mode, libev will still try to automatically deduce the
		4221	configuration, but has to be more conservative.
		4222
		4223	=item EV_USE_FLOOR
		4224
		4225	If defined to be C<1>, libev will use the C<floor ()> function for its
		4226	periodic reschedule calculations, otherwise libev will fall back on a
		4227	portable (slower) implementation. If you enable this, you usually have to
		4228	link against libm or something equivalent. Enabling this when the C<floor>
		4229	function is not available will fail, so the safe default is to not enable
		4230	this.
		4231
1527	=item EV_USE_MONOTONIC	4232	=item EV_USE_MONOTONIC
1528		4233
1529	If defined to be C<1>, libev will try to detect the availability of the	4234	If defined to be C<1>, libev will try to detect the availability of the
1530	monotonic clock option at both compiletime and runtime. Otherwise no use	4235	monotonic clock option at both compile time and runtime. Otherwise no
1531	of the monotonic clock option will be attempted. If you enable this, you	4236	use of the monotonic clock option will be attempted. If you enable this,
1532	usually have to link against librt or something similar. Enabling it when	4237	you usually have to link against librt or something similar. Enabling it
1533	the functionality isn't available is safe, though, althoguh you have	4238	when the functionality isn't available is safe, though, although you have
1534	to make sure you link against any libraries where the C<clock_gettime>	4239	to make sure you link against any libraries where the C<clock_gettime>
1535	function is hiding in (often F<-lrt>).	4240	function is hiding in (often F<-lrt>). See also C<EV_USE_CLOCK_SYSCALL>.
1536		4241
1537	=item EV_USE_REALTIME	4242	=item EV_USE_REALTIME
1538		4243
1539	If defined to be C<1>, libev will try to detect the availability of the	4244	If defined to be C<1>, libev will try to detect the availability of the
1540	realtime clock option at compiletime (and assume its availability at	4245	real-time clock option at compile time (and assume its availability
1541	runtime if successful). Otherwise no use of the realtime clock option will	4246	at runtime if successful). Otherwise no use of the real-time clock
1542	be attempted. This effectively replaces C<gettimeofday> by C<clock_get	4247	option will be attempted. This effectively replaces C<gettimeofday>
1543	(CLOCK_REALTIME, ...)> and will not normally affect correctness. See tzhe note about libraries	4248	by C<clock_get (CLOCK_REALTIME, ...)> and will not normally affect
1544	in the description of C<EV_USE_MONOTONIC>, though.	4249	correctness. See the note about libraries in the description of
		4250	C<EV_USE_MONOTONIC>, though. Defaults to the opposite value of
		4251	C<EV_USE_CLOCK_SYSCALL>.
		4252
		4253	=item EV_USE_CLOCK_SYSCALL
		4254
		4255	If defined to be C<1>, libev will try to use a direct syscall instead
		4256	of calling the system-provided C<clock_gettime> function. This option
		4257	exists because on GNU/Linux, C<clock_gettime> is in C<librt>, but C<librt>
		4258	unconditionally pulls in C<libpthread>, slowing down single-threaded
		4259	programs needlessly. Using a direct syscall is slightly slower (in
		4260	theory), because no optimised vdso implementation can be used, but avoids
		4261	the pthread dependency. Defaults to C<1> on GNU/Linux with glibc 2.x or
		4262	higher, as it simplifies linking (no need for C<-lrt>).
		4263
		4264	=item EV_USE_NANOSLEEP
		4265
		4266	If defined to be C<1>, libev will assume that C<nanosleep ()> is available
		4267	and will use it for delays. Otherwise it will use C<select ()>.
		4268
		4269	=item EV_USE_EVENTFD
		4270
		4271	If defined to be C<1>, then libev will assume that C<eventfd ()> is
		4272	available and will probe for kernel support at runtime. This will improve
		4273	C<ev_signal> and C<ev_async> performance and reduce resource consumption.
		4274	If undefined, it will be enabled if the headers indicate GNU/Linux + Glibc
		4275	2.7 or newer, otherwise disabled.
1545		4276
1546	=item EV_USE_SELECT	4277	=item EV_USE_SELECT
1547		4278
1548	If undefined or defined to be C<1>, libev will compile in support for the	4279	If undefined or defined to be C<1>, libev will compile in support for the
1549	C<select>(2) backend. No attempt at autodetection will be done: if no	4280	C<select>(2) backend. No attempt at auto-detection will be done: if no
1550	other method takes over, select will be it. Otherwise the select backend	4281	other method takes over, select will be it. Otherwise the select backend
1551	will not be compiled in.	4282	will not be compiled in.
1552		4283
1553	=item EV_SELECT_USE_FD_SET	4284	=item EV_SELECT_USE_FD_SET
1554		4285
1555	If defined to C<1>, then the select backend will use the system C<fd_set>	4286	If defined to C<1>, then the select backend will use the system C<fd_set>
1556	structure. This is useful if libev doesn't compile due to a missing	4287	structure. This is useful if libev doesn't compile due to a missing
1557	C<NFDBITS> or C<fd_mask> definition or it misguesses the bitset layout on	4288	C<NFDBITS> or C<fd_mask> definition or it mis-guesses the bitset layout
1558	exotic systems. This usually limits the range of file descriptors to some	4289	on exotic systems. This usually limits the range of file descriptors to
1559	low limit such as 1024 or might have other limitations (winsocket only	4290	some low limit such as 1024 or might have other limitations (winsocket
1560	allows 64 sockets). The C<FD_SETSIZE> macro, set before compilation, might	4291	only allows 64 sockets). The C<FD_SETSIZE> macro, set before compilation,
1561	influence the size of the C<fd_set> used.	4292	configures the maximum size of the C<fd_set>.
1562		4293
1563	=item EV_SELECT_IS_WINSOCKET	4294	=item EV_SELECT_IS_WINSOCKET
1564		4295
1565	When defined to C<1>, the select backend will assume that	4296	When defined to C<1>, the select backend will assume that
1566	select/socket/connect etc. don't understand file descriptors but	4297	select/socket/connect etc. don't understand file descriptors but
…		…
1568	be used is the winsock select). This means that it will call	4299	be used is the winsock select). This means that it will call
1569	C<_get_osfhandle> on the fd to convert it to an OS handle. Otherwise,	4300	C<_get_osfhandle> on the fd to convert it to an OS handle. Otherwise,
1570	it is assumed that all these functions actually work on fds, even	4301	it is assumed that all these functions actually work on fds, even
1571	on win32. Should not be defined on non-win32 platforms.	4302	on win32. Should not be defined on non-win32 platforms.
1572		4303
		4304	=item EV_FD_TO_WIN32_HANDLE(fd)
		4305
		4306	If C<EV_SELECT_IS_WINSOCKET> is enabled, then libev needs a way to map
		4307	file descriptors to socket handles. When not defining this symbol (the
		4308	default), then libev will call C<_get_osfhandle>, which is usually
		4309	correct. In some cases, programs use their own file descriptor management,
		4310	in which case they can provide this function to map fds to socket handles.
		4311
		4312	=item EV_WIN32_HANDLE_TO_FD(handle)
		4313
		4314	If C<EV_SELECT_IS_WINSOCKET> then libev maps handles to file descriptors
		4315	using the standard C<_open_osfhandle> function. For programs implementing
		4316	their own fd to handle mapping, overwriting this function makes it easier
		4317	to do so. This can be done by defining this macro to an appropriate value.
		4318
		4319	=item EV_WIN32_CLOSE_FD(fd)
		4320
		4321	If programs implement their own fd to handle mapping on win32, then this
		4322	macro can be used to override the C<close> function, useful to unregister
		4323	file descriptors again. Note that the replacement function has to close
		4324	the underlying OS handle.
		4325
1573	=item EV_USE_POLL	4326	=item EV_USE_POLL
1574		4327
1575	If defined to be C<1>, libev will compile in support for the C<poll>(2)	4328	If defined to be C<1>, libev will compile in support for the C<poll>(2)
1576	backend. Otherwise it will be enabled on non-win32 platforms. It	4329	backend. Otherwise it will be enabled on non-win32 platforms. It
1577	takes precedence over select.	4330	takes precedence over select.
1578		4331
1579	=item EV_USE_EPOLL	4332	=item EV_USE_EPOLL
1580		4333
1581	If defined to be C<1>, libev will compile in support for the Linux	4334	If defined to be C<1>, libev will compile in support for the Linux
1582	C<epoll>(7) backend. Its availability will be detected at runtime,	4335	C<epoll>(7) backend. Its availability will be detected at runtime,
1583	otherwise another method will be used as fallback. This is the	4336	otherwise another method will be used as fallback. This is the preferred
1584	preferred backend for GNU/Linux systems.	4337	backend for GNU/Linux systems. If undefined, it will be enabled if the
		4338	headers indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.
1585		4339
1586	=item EV_USE_KQUEUE	4340	=item EV_USE_KQUEUE
1587		4341
1588	If defined to be C<1>, libev will compile in support for the BSD style	4342	If defined to be C<1>, libev will compile in support for the BSD style
1589	C<kqueue>(2) backend. Its actual availability will be detected at runtime,	4343	C<kqueue>(2) backend. Its actual availability will be detected at runtime,
…		…
1602	otherwise another method will be used as fallback. This is the preferred	4356	otherwise another method will be used as fallback. This is the preferred
1603	backend for Solaris 10 systems.	4357	backend for Solaris 10 systems.
1604		4358
1605	=item EV_USE_DEVPOLL	4359	=item EV_USE_DEVPOLL
1606		4360
1607	reserved for future expansion, works like the USE symbols above.	4361	Reserved for future expansion, works like the USE symbols above.
1608		4362
		4363	=item EV_USE_INOTIFY
		4364
		4365	If defined to be C<1>, libev will compile in support for the Linux inotify
		4366	interface to speed up C<ev_stat> watchers. Its actual availability will
		4367	be detected at runtime. If undefined, it will be enabled if the headers
		4368	indicate GNU/Linux + Glibc 2.4 or newer, otherwise disabled.
		4369
		4370	=item EV_ATOMIC_T
		4371
		4372	Libev requires an integer type (suitable for storing C<0> or C<1>) whose
		4373	access is atomic with respect to other threads or signal contexts. No such
		4374	type is easily found in the C language, so you can provide your own type
		4375	that you know is safe for your purposes. It is used both for signal handler "locking"
		4376	as well as for signal and thread safety in C<ev_async> watchers.
		4377
		4378	In the absence of this define, libev will use C<sig_atomic_t volatile>
		4379	(from F<signal.h>), which is usually good enough on most platforms.
		4380
1609	=item EV_H	4381	=item EV_H (h)
1610		4382
1611	The name of the F<ev.h> header file used to include it. The default if	4383	The name of the F<ev.h> header file used to include it. The default if
1612	undefined is C<< <ev.h> >> in F<event.h> and C<"ev.h"> in F<ev.c>. This	4384	undefined is C<"ev.h"> in F<event.h>, F<ev.c> and F<ev++.h>. This can be
1613	can be used to virtually rename the F<ev.h> header file in case of conflicts.	4385	used to virtually rename the F<ev.h> header file in case of conflicts.
1614		4386
1615	=item EV_CONFIG_H	4387	=item EV_CONFIG_H (h)
1616		4388
1617	If C<EV_STANDALONE> isn't C<1>, this variable can be used to override	4389	If C<EV_STANDALONE> isn't C<1>, this variable can be used to override
1618	F<ev.c>'s idea of where to find the F<config.h> file, similarly to	4390	F<ev.c>'s idea of where to find the F<config.h> file, similarly to
1619	C<EV_H>, above.	4391	C<EV_H>, above.
1620		4392
1621	=item EV_EVENT_H	4393	=item EV_EVENT_H (h)
1622		4394
1623	Similarly to C<EV_H>, this macro can be used to override F<event.c>'s idea	4395	Similarly to C<EV_H>, this macro can be used to override F<event.c>'s idea
1624	of how the F<event.h> header can be found.	4396	of how the F<event.h> header can be found, the default is C<"event.h">.
1625		4397
1626	=item EV_PROTOTYPES	4398	=item EV_PROTOTYPES (h)
1627		4399
1628	If defined to be C<0>, then F<ev.h> will not define any function	4400	If defined to be C<0>, then F<ev.h> will not define any function
1629	prototypes, but still define all the structs and other symbols. This is	4401	prototypes, but still define all the structs and other symbols. This is
1630	occasionally useful if you want to provide your own wrapper functions	4402	occasionally useful if you want to provide your own wrapper functions
1631	around libev functions.	4403	around libev functions.
…		…
1636	will have the C<struct ev_loop *> as first argument, and you can create	4408	will have the C<struct ev_loop *> as first argument, and you can create
1637	additional independent event loops. Otherwise there will be no support	4409	additional independent event loops. Otherwise there will be no support
1638	for multiple event loops and there is no first event loop pointer	4410	for multiple event loops and there is no first event loop pointer
1639	argument. Instead, all functions act on the single default loop.	4411	argument. Instead, all functions act on the single default loop.
1640		4412
1641	=item EV_PERIODICS	4413	=item EV_MINPRI
1642		4414
1643	If undefined or defined to be C<1>, then periodic timers are supported,	4415	=item EV_MAXPRI
1644	otherwise not. This saves a few kb of code.	4416
		4417	The range of allowed priorities. C<EV_MINPRI> must be smaller or equal to
		4418	C<EV_MAXPRI>, but otherwise there are no non-obvious limitations. You can
		4419	provide for more priorities by overriding those symbols (usually defined
		4420	to be C<-2> and C<2>, respectively).
		4421
		4422	When doing priority-based operations, libev usually has to linearly search
		4423	all the priorities, so having many of them (hundreds) uses a lot of space
		4424	and time, so using the defaults of five priorities (-2 .. +2) is usually
		4425	fine.
		4426
		4427	If your embedding application does not need any priorities, defining these
		4428	both to C<0> will save some memory and CPU.
		4429
		4430	=item EV_PERIODIC_ENABLE, EV_IDLE_ENABLE, EV_EMBED_ENABLE, EV_STAT_ENABLE,
		4431	EV_PREPARE_ENABLE, EV_CHECK_ENABLE, EV_FORK_ENABLE, EV_SIGNAL_ENABLE,
		4432	EV_ASYNC_ENABLE, EV_CHILD_ENABLE.
		4433
		4434	If undefined or defined to be C<1> (and the platform supports it), then
		4435	the respective watcher type is supported. If defined to be C<0>, then it
		4436	is not. Disabling watcher types mainly saves code size.
		4437
		4438	=item EV_FEATURES
		4439
		4440	If you need to shave off some kilobytes of code at the expense of some
		4441	speed (but with the full API), you can define this symbol to request
		4442	certain subsets of functionality. The default is to enable all features
		4443	that can be enabled on the platform.
		4444
		4445	A typical way to use this symbol is to define it to C<0> (or to a bitset
		4446	with some broad features you want) and then selectively re-enable
		4447	additional parts you want, for example if you want everything minimal,
		4448	but multiple event loop support, async and child watchers and the poll
		4449	backend, use this:
		4450
		4451	#define EV_FEATURES 0
		4452	#define EV_MULTIPLICITY 1
		4453	#define EV_USE_POLL 1
		4454	#define EV_CHILD_ENABLE 1
		4455	#define EV_ASYNC_ENABLE 1
		4456
		4457	The actual value is a bitset, it can be a combination of the following
		4458	values:
		4459
		4460	=over 4
		4461
		4462	=item C<1> - faster/larger code
		4463
		4464	Use larger code to speed up some operations.
		4465
		4466	Currently this is used to override some inlining decisions (enlarging the
		4467	code size by roughly 30% on amd64).
		4468
		4469	When optimising for size, use of compiler flags such as C<-Os> with
		4470	gcc is recommended, as well as C<-DNDEBUG>, as libev contains a number of
		4471	assertions.
		4472
		4473	=item C<2> - faster/larger data structures
		4474
		4475	Replaces the small 2-heap for timer management by a faster 4-heap, larger
		4476	hash table sizes and so on. This will usually further increase code size
		4477	and can additionally have an effect on the size of data structures at
		4478	runtime.
		4479
		4480	=item C<4> - full API configuration
		4481
		4482	This enables priorities (sets C<EV_MAXPRI>=2 and C<EV_MINPRI>=-2), and
		4483	enables multiplicity (C<EV_MULTIPLICITY>=1).
		4484
		4485	=item C<8> - full API
		4486
		4487	This enables a lot of the "lesser used" API functions. See C<ev.h> for
		4488	details on which parts of the API are still available without this
		4489	feature, and do not complain if this subset changes over time.
		4490
		4491	=item C<16> - enable all optional watcher types
		4492
		4493	Enables all optional watcher types. If you want to selectively enable
		4494	only some watcher types other than I/O and timers (e.g. prepare,
		4495	embed, async, child...) you can enable them manually by defining
		4496	C<EV_watchertype_ENABLE> to C<1> instead.
		4497
		4498	=item C<32> - enable all backends
		4499
		4500	This enables all backends - without this feature, you need to enable at
		4501	least one backend manually (C<EV_USE_SELECT> is a good choice).
		4502
		4503	=item C<64> - enable OS-specific "helper" APIs
		4504
		4505	Enable inotify, eventfd, signalfd and similar OS-specific helper APIs by
		4506	default.
		4507
		4508	=back
		4509
		4510	Compiling with C<gcc -Os -DEV_STANDALONE -DEV_USE_EPOLL=1 -DEV_FEATURES=0>
		4511	reduces the compiled size of libev from 24.7Kb code/2.8Kb data to 6.5Kb
		4512	code/0.3Kb data on my GNU/Linux amd64 system, while still giving you I/O
		4513	watchers, timers and monotonic clock support.
		4514
		4515	With an intelligent-enough linker (gcc+binutils are intelligent enough
		4516	when you use C<-Wl,--gc-sections -ffunction-sections>) functions unused by
		4517	your program might be left out as well - a binary starting a timer and an
		4518	I/O watcher then might come out at only 5Kb.
		4519
		4520	=item EV_AVOID_STDIO
		4521
		4522	If this is set to C<1> at compiletime, then libev will avoid using stdio
		4523	functions (printf, scanf, perror etc.). This will increase the code size
		4524	somewhat, but if your program doesn't otherwise depend on stdio and your
		4525	libc allows it, this avoids linking in the stdio library which is quite
		4526	big.
		4527
		4528	Note that error messages might become less precise when this option is
		4529	enabled.
		4530
		4531	=item EV_NSIG
		4532
		4533	The highest supported signal number, +1 (or, the number of
		4534	signals): Normally, libev tries to deduce the maximum number of signals
		4535	automatically, but sometimes this fails, in which case it can be
		4536	specified. Also, using a lower number than detected (C<32> should be
		4537	good for about any system in existence) can save some memory, as libev
		4538	statically allocates some 12-24 bytes per signal number.
		4539
		4540	=item EV_PID_HASHSIZE
		4541
		4542	C<ev_child> watchers use a small hash table to distribute workload by
		4543	pid. The default size is C<16> (or C<1> with C<EV_FEATURES> disabled),
		4544	usually more than enough. If you need to manage thousands of children you
		4545	might want to increase this value (I<must> be a power of two).
		4546
		4547	=item EV_INOTIFY_HASHSIZE
		4548
		4549	C<ev_stat> watchers use a small hash table to distribute workload by
		4550	inotify watch id. The default size is C<16> (or C<1> with C<EV_FEATURES>
		4551	disabled), usually more than enough. If you need to manage thousands of
		4552	C<ev_stat> watchers you might want to increase this value (I<must> be a
		4553	power of two).
		4554
		4555	=item EV_USE_4HEAP
		4556
		4557	Heaps are not very cache-efficient. To improve the cache-efficiency of the
		4558	timer and periodics heaps, libev uses a 4-heap when this symbol is defined
		4559	to C<1>. The 4-heap uses more complicated (longer) code but has noticeably
		4560	faster performance with many (thousands) of watchers.
		4561
		4562	The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it
		4563	will be C<0>.
		4564
		4565	=item EV_HEAP_CACHE_AT
		4566
		4567	Heaps are not very cache-efficient. To improve the cache-efficiency of the
		4568	timer and periodics heaps, libev can cache the timestamp (I<at>) within
		4569	the heap structure (selected by defining C<EV_HEAP_CACHE_AT> to C<1>),
		4570	which uses 8-12 bytes more per watcher and a few hundred bytes more code,
		4571	but avoids random read accesses on heap changes. This improves performance
		4572	noticeably with many (hundreds) of watchers.
		4573
		4574	The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it
		4575	will be C<0>.
		4576
		4577	=item EV_VERIFY
		4578
		4579	Controls how much internal verification (see C<ev_verify ()>) will
		4580	be done: If set to C<0>, no internal verification code will be compiled
		4581	in. If set to C<1>, then verification code will be compiled in, but not
		4582	called. If set to C<2>, then the internal verification code will be
		4583	called once per loop, which can slow down libev. If set to C<3>, then the
		4584	verification code will be called very frequently, which will slow down
		4585	libev considerably.
		4586
		4587	The default is C<1>, unless C<EV_FEATURES> overrides it, in which case it
		4588	will be C<0>.
1645		4589
1646	=item EV_COMMON	4590	=item EV_COMMON
1647		4591
1648	By default, all watchers have a C<void *data> member. By redefining	4592	By default, all watchers have a C<void *data> member. By redefining
1649	this macro to a something else you can include more and other types of	4593	this macro to something else you can include more and other types of
1650	members. You have to define it each time you include one of the files,	4594	members. You have to define it each time you include one of the files,
1651	though, and it must be identical each time.	4595	though, and it must be identical each time.
1652		4596
1653	For example, the perl EV module uses something like this:	4597	For example, the perl EV module uses something like this:
1654		4598
1655	#define EV_COMMON \	4599	#define EV_COMMON \
1656	SV *self; /* contains this struct */ \	4600	SV *self; /* contains this struct */ \
1657	SV *cb_sv, *fh /* note no trailing ";" */	4601	SV *cb_sv, *fh /* note no trailing ";" */
1658		4602
1659	=item EV_CB_DECLARE(type)	4603	=item EV_CB_DECLARE (type)
1660		4604
1661	=item EV_CB_INVOKE(watcher,revents)	4605	=item EV_CB_INVOKE (watcher, revents)
1662		4606
1663	=item ev_set_cb(ev,cb)	4607	=item ev_set_cb (ev, cb)
1664		4608
1665	Can be used to change the callback member declaration in each watcher,	4609	Can be used to change the callback member declaration in each watcher,
1666	and the way callbacks are invoked and set. Must expand to a struct member	4610	and the way callbacks are invoked and set. Must expand to a struct member
1667	definition and a statement, respectively. See the F<ev.v> header file for	4611	definition and a statement, respectively. See the F<ev.h> header file for
1668	their default definitions. One possible use for overriding these is to	4612	their default definitions. One possible use for overriding these is to
1669	avoid the ev_loop pointer as first argument in all cases, or to use method	4613	avoid the C<struct ev_loop *> as first argument in all cases, or to use
1670	calls instead of plain function calls in C++.	4614	method calls instead of plain function calls in C++.
		4615
		4616	=back
		4617
		4618	=head2 EXPORTED API SYMBOLS
		4619
		4620	If you need to re-export the API (e.g. via a DLL) and you need a list of
		4621	exported symbols, you can use the provided F<Symbol.*> files which list
		4622	all public symbols, one per line:
		4623
		4624	Symbols.ev for libev proper
		4625	Symbols.event for the libevent emulation
		4626
		4627	This can also be used to rename all public symbols to avoid clashes with
		4628	multiple versions of libev linked together (which is obviously bad in
		4629	itself, but sometimes it is inconvenient to avoid this).
		4630
		4631	A sed command like this will create wrapper C<#define>'s that you need to
		4632	include before including F<ev.h>:
		4633
		4634	<Symbols.ev sed -e "s/.*/#define & myprefix_&/" >wrap.h
		4635
		4636	This would create a file F<wrap.h> which essentially looks like this:
		4637
		4638	#define ev_backend myprefix_ev_backend
		4639	#define ev_check_start myprefix_ev_check_start
		4640	#define ev_check_stop myprefix_ev_check_stop
		4641	...
1671		4642
1672	=head2 EXAMPLES	4643	=head2 EXAMPLES
1673		4644
1674	For a real-world example of a program the includes libev	4645	For a real-world example of a program the includes libev
1675	verbatim, you can have a look at the EV perl module	4646	verbatim, you can have a look at the EV perl module
…		…
1678	interface) and F<EV.xs> (implementation) files. Only the F<EV.xs> file	4649	interface) and F<EV.xs> (implementation) files. Only the F<EV.xs> file
1679	will be compiled. It is pretty complex because it provides its own header	4650	will be compiled. It is pretty complex because it provides its own header
1680	file.	4651	file.
1681		4652
1682	The usage in rxvt-unicode is simpler. It has a F<ev_cpp.h> header file	4653	The usage in rxvt-unicode is simpler. It has a F<ev_cpp.h> header file
1683	that everybody includes and which overrides some autoconf choices:	4654	that everybody includes and which overrides some configure choices:
1684		4655
		4656	#define EV_FEATURES 8
		4657	#define EV_USE_SELECT 1
		4658	#define EV_PREPARE_ENABLE 1
		4659	#define EV_IDLE_ENABLE 1
		4660	#define EV_SIGNAL_ENABLE 1
		4661	#define EV_CHILD_ENABLE 1
1685	#define EV_USE_POLL 0	4662	#define EV_USE_STDEXCEPT 0
1686	#define EV_MULTIPLICITY 0
1687	#define EV_PERIODICS 0
1688	#define EV_CONFIG_H <config.h>	4663	#define EV_CONFIG_H <config.h>
1689		4664
1690	#include "ev++.h"	4665	#include "ev++.h"
1691		4666
1692	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:	4667	And a F<ev_cpp.C> implementation file that contains libev proper and is compiled:
1693		4668
1694	#include "ev_cpp.h"	4669	#include "ev_cpp.h"
1695	#include "ev.c"	4670	#include "ev.c"
		4671
		4672	=head1 INTERACTION WITH OTHER PROGRAMS, LIBRARIES OR THE ENVIRONMENT
		4673
		4674	=head2 THREADS AND COROUTINES
		4675
		4676	=head3 THREADS
		4677
		4678	All libev functions are reentrant and thread-safe unless explicitly
		4679	documented otherwise, but libev implements no locking itself. This means
		4680	that you can use as many loops as you want in parallel, as long as there
		4681	are no concurrent calls into any libev function with the same loop
		4682	parameter (C<ev_default_*> calls have an implicit default loop parameter,
		4683	of course): libev guarantees that different event loops share no data
		4684	structures that need any locking.
		4685
		4686	Or to put it differently: calls with different loop parameters can be done
		4687	concurrently from multiple threads, calls with the same loop parameter
		4688	must be done serially (but can be done from different threads, as long as
		4689	only one thread ever is inside a call at any point in time, e.g. by using
		4690	a mutex per loop).
		4691
		4692	Specifically to support threads (and signal handlers), libev implements
		4693	so-called C<ev_async> watchers, which allow some limited form of
		4694	concurrency on the same event loop, namely waking it up "from the
		4695	outside".
		4696
		4697	If you want to know which design (one loop, locking, or multiple loops
		4698	without or something else still) is best for your problem, then I cannot
		4699	help you, but here is some generic advice:
		4700
		4701	=over 4
		4702
		4703	=item * most applications have a main thread: use the default libev loop
		4704	in that thread, or create a separate thread running only the default loop.
		4705
		4706	This helps integrating other libraries or software modules that use libev
		4707	themselves and don't care/know about threading.
		4708
		4709	=item * one loop per thread is usually a good model.
		4710
		4711	Doing this is almost never wrong, sometimes a better-performance model
		4712	exists, but it is always a good start.
		4713
		4714	=item * other models exist, such as the leader/follower pattern, where one
		4715	loop is handed through multiple threads in a kind of round-robin fashion.
		4716
		4717	Choosing a model is hard - look around, learn, know that usually you can do
		4718	better than you currently do :-)
		4719
		4720	=item * often you need to talk to some other thread which blocks in the
		4721	event loop.
		4722
		4723	C<ev_async> watchers can be used to wake them up from other threads safely
		4724	(or from signal contexts...).
		4725
		4726	An example use would be to communicate signals or other events that only
		4727	work in the default loop by registering the signal watcher with the
		4728	default loop and triggering an C<ev_async> watcher from the default loop
		4729	watcher callback into the event loop interested in the signal.
		4730
		4731	=back
		4732
		4733	See also L<THREAD LOCKING EXAMPLE>.
		4734
		4735	=head3 COROUTINES
		4736
		4737	Libev is very accommodating to coroutines ("cooperative threads"):
		4738	libev fully supports nesting calls to its functions from different
		4739	coroutines (e.g. you can call C<ev_run> on the same loop from two
		4740	different coroutines, and switch freely between both coroutines running
		4741	the loop, as long as you don't confuse yourself). The only exception is
		4742	that you must not do this from C<ev_periodic> reschedule callbacks.
		4743
		4744	Care has been taken to ensure that libev does not keep local state inside
		4745	C<ev_run>, and other calls do not usually allow for coroutine switches as
		4746	they do not call any callbacks.
		4747
		4748	=head2 COMPILER WARNINGS
		4749
		4750	Depending on your compiler and compiler settings, you might get no or a
		4751	lot of warnings when compiling libev code. Some people are apparently
		4752	scared by this.
		4753
		4754	However, these are unavoidable for many reasons. For one, each compiler
		4755	has different warnings, and each user has different tastes regarding
		4756	warning options. "Warn-free" code therefore cannot be a goal except when
		4757	targeting a specific compiler and compiler-version.
		4758
		4759	Another reason is that some compiler warnings require elaborate
		4760	workarounds, or other changes to the code that make it less clear and less
		4761	maintainable.
		4762
		4763	And of course, some compiler warnings are just plain stupid, or simply
		4764	wrong (because they don't actually warn about the condition their message
		4765	seems to warn about). For example, certain older gcc versions had some
		4766	warnings that resulted in an extreme number of false positives. These have
		4767	been fixed, but some people still insist on making code warn-free with
		4768	such buggy versions.
		4769
		4770	While libev is written to generate as few warnings as possible,
		4771	"warn-free" code is not a goal, and it is recommended not to build libev
		4772	with any compiler warnings enabled unless you are prepared to cope with
		4773	them (e.g. by ignoring them). Remember that warnings are just that:
		4774	warnings, not errors, or proof of bugs.
		4775
		4776
		4777	=head2 VALGRIND
		4778
		4779	Valgrind has a special section here because it is a popular tool that is
		4780	highly useful. Unfortunately, valgrind reports are very hard to interpret.
		4781
		4782	If you think you found a bug (memory leak, uninitialised data access etc.)
		4783	in libev, then check twice: If valgrind reports something like:
		4784
		4785	==2274== definitely lost: 0 bytes in 0 blocks.
		4786	==2274== possibly lost: 0 bytes in 0 blocks.
		4787	==2274== still reachable: 256 bytes in 1 blocks.
		4788
		4789	Then there is no memory leak, just as memory accounted to global variables
		4790	is not a memleak - the memory is still being referenced, and didn't leak.
		4791
		4792	Similarly, under some circumstances, valgrind might report kernel bugs
		4793	as if it were a bug in libev (e.g. in realloc or in the poll backend,
		4794	although an acceptable workaround has been found here), or it might be
		4795	confused.
		4796
		4797	Keep in mind that valgrind is a very good tool, but only a tool. Don't
		4798	make it into some kind of religion.
		4799
		4800	If you are unsure about something, feel free to contact the mailing list
		4801	with the full valgrind report and an explanation on why you think this
		4802	is a bug in libev (best check the archives, too :). However, don't be
		4803	annoyed when you get a brisk "this is no bug" answer and take the chance
		4804	of learning how to interpret valgrind properly.
		4805
		4806	If you need, for some reason, empty reports from valgrind for your project
		4807	I suggest using suppression lists.
		4808
		4809
		4810	=head1 PORTABILITY NOTES
		4811
		4812	=head2 GNU/LINUX 32 BIT LIMITATIONS
		4813
		4814	GNU/Linux is the only common platform that supports 64 bit file/large file
		4815	interfaces but I<disables> them by default.
		4816
		4817	That means that libev compiled in the default environment doesn't support
		4818	files larger than 2GiB or so, which mainly affects C<ev_stat> watchers.
		4819
		4820	Unfortunately, many programs try to work around this GNU/Linux issue
		4821	by enabling the large file API, which makes them incompatible with the
		4822	standard libev compiled for their system.
		4823
		4824	Likewise, libev cannot enable the large file API itself as this would
		4825	suddenly make it incompatible to the default compile time environment,
		4826	i.e. all programs not using special compile switches.
		4827
		4828	=head2 OS/X AND DARWIN BUGS
		4829
		4830	The whole thing is a bug if you ask me - basically any system interface
		4831	you touch is broken, whether it is locales, poll, kqueue or even the
		4832	OpenGL drivers.
		4833
		4834	=head3 C<kqueue> is buggy
		4835
		4836	The kqueue syscall is broken in all known versions - most versions support
		4837	only sockets, many support pipes.
		4838
		4839	Libev tries to work around this by not using C<kqueue> by default on this
		4840	rotten platform, but of course you can still ask for it when creating a
		4841	loop - embedding a socket-only kqueue loop into a select-based one is
		4842	probably going to work well.
		4843
		4844	=head3 C<poll> is buggy
		4845
		4846	Instead of fixing C<kqueue>, Apple replaced their (working) C<poll>
		4847	implementation by something calling C<kqueue> internally around the 10.5.6
		4848	release, so now C<kqueue> I<and> C<poll> are broken.
		4849
		4850	Libev tries to work around this by not using C<poll> by default on
		4851	this rotten platform, but of course you can still ask for it when creating
		4852	a loop.
		4853
		4854	=head3 C<select> is buggy
		4855
		4856	All that's left is C<select>, and of course Apple found a way to fuck this
		4857	one up as well: On OS/X, C<select> actively limits the number of file
		4858	descriptors you can pass in to 1024 - your program suddenly crashes when
		4859	you use more.
		4860
		4861	There is an undocumented "workaround" for this - defining
		4862	C<_DARWIN_UNLIMITED_SELECT>, which libev tries to use, so select I<should>
		4863	work on OS/X.
		4864
		4865	=head2 SOLARIS PROBLEMS AND WORKAROUNDS
		4866
		4867	=head3 C<errno> reentrancy
		4868
		4869	The default compile environment on Solaris is unfortunately so
		4870	thread-unsafe that you can't even use components/libraries compiled
		4871	without C<-D_REENTRANT> in a threaded program, which, of course, isn't
		4872	defined by default. A valid, if stupid, implementation choice.
		4873
		4874	If you want to use libev in threaded environments you have to make sure
		4875	it's compiled with C<_REENTRANT> defined.
		4876
		4877	=head3 Event port backend
		4878
		4879	The scalable event interface for Solaris is called "event
		4880	ports". Unfortunately, this mechanism is very buggy in all major
		4881	releases. If you run into high CPU usage, your program freezes or you get
		4882	a large number of spurious wakeups, make sure you have all the relevant
		4883	and latest kernel patches applied. No, I don't know which ones, but there
		4884	are multiple ones to apply, and afterwards, event ports actually work
		4885	great.
		4886
		4887	If you can't get it to work, you can try running the program by setting
		4888	the environment variable C<LIBEV_FLAGS=3> to only allow C<poll> and
		4889	C<select> backends.
		4890
		4891	=head2 AIX POLL BUG
		4892
		4893	AIX unfortunately has a broken C<poll.h> header. Libev works around
		4894	this by trying to avoid the poll backend altogether (i.e. it's not even
		4895	compiled in), which normally isn't a big problem as C<select> works fine
		4896	with large bitsets on AIX, and AIX is dead anyway.
		4897
		4898	=head2 WIN32 PLATFORM LIMITATIONS AND WORKAROUNDS
		4899
		4900	=head3 General issues
		4901
		4902	Win32 doesn't support any of the standards (e.g. POSIX) that libev
		4903	requires, and its I/O model is fundamentally incompatible with the POSIX
		4904	model. Libev still offers limited functionality on this platform in
		4905	the form of the C<EVBACKEND_SELECT> backend, and only supports socket
		4906	descriptors. This only applies when using Win32 natively, not when using
		4907	e.g. cygwin. Actually, it only applies to the microsofts own compilers,
		4908	as every compielr comes with a slightly differently broken/incompatible
		4909	environment.
		4910
		4911	Lifting these limitations would basically require the full
		4912	re-implementation of the I/O system. If you are into this kind of thing,
		4913	then note that glib does exactly that for you in a very portable way (note
		4914	also that glib is the slowest event library known to man).
		4915
		4916	There is no supported compilation method available on windows except
		4917	embedding it into other applications.
		4918
		4919	Sensible signal handling is officially unsupported by Microsoft - libev
		4920	tries its best, but under most conditions, signals will simply not work.
		4921
		4922	Not a libev limitation but worth mentioning: windows apparently doesn't
		4923	accept large writes: instead of resulting in a partial write, windows will
		4924	either accept everything or return C<ENOBUFS> if the buffer is too large,
		4925	so make sure you only write small amounts into your sockets (less than a
		4926	megabyte seems safe, but this apparently depends on the amount of memory
		4927	available).
		4928
		4929	Due to the many, low, and arbitrary limits on the win32 platform and
		4930	the abysmal performance of winsockets, using a large number of sockets
		4931	is not recommended (and not reasonable). If your program needs to use
		4932	more than a hundred or so sockets, then likely it needs to use a totally
		4933	different implementation for windows, as libev offers the POSIX readiness
		4934	notification model, which cannot be implemented efficiently on windows
		4935	(due to Microsoft monopoly games).
		4936
		4937	A typical way to use libev under windows is to embed it (see the embedding
		4938	section for details) and use the following F<evwrap.h> header file instead
		4939	of F<ev.h>:
		4940
		4941	#define EV_STANDALONE /* keeps ev from requiring config.h */
		4942	#define EV_SELECT_IS_WINSOCKET 1 /* configure libev for windows select */
		4943
		4944	#include "ev.h"
		4945
		4946	And compile the following F<evwrap.c> file into your project (make sure
		4947	you do I<not> compile the F<ev.c> or any other embedded source files!):
		4948
		4949	#include "evwrap.h"
		4950	#include "ev.c"
		4951
		4952	=head3 The winsocket C<select> function
		4953
		4954	The winsocket C<select> function doesn't follow POSIX in that it
		4955	requires socket I<handles> and not socket I<file descriptors> (it is
		4956	also extremely buggy). This makes select very inefficient, and also
		4957	requires a mapping from file descriptors to socket handles (the Microsoft
		4958	C runtime provides the function C<_open_osfhandle> for this). See the
		4959	discussion of the C<EV_SELECT_USE_FD_SET>, C<EV_SELECT_IS_WINSOCKET> and
		4960	C<EV_FD_TO_WIN32_HANDLE> preprocessor symbols for more info.
		4961
		4962	The configuration for a "naked" win32 using the Microsoft runtime
		4963	libraries and raw winsocket select is:
		4964
		4965	#define EV_USE_SELECT 1
		4966	#define EV_SELECT_IS_WINSOCKET 1 /* forces EV_SELECT_USE_FD_SET, too */
		4967
		4968	Note that winsockets handling of fd sets is O(n), so you can easily get a
		4969	complexity in the O(n²) range when using win32.
		4970
		4971	=head3 Limited number of file descriptors
		4972
		4973	Windows has numerous arbitrary (and low) limits on things.
		4974
		4975	Early versions of winsocket's select only supported waiting for a maximum
		4976	of C<64> handles (probably owning to the fact that all windows kernels
		4977	can only wait for C<64> things at the same time internally; Microsoft
		4978	recommends spawning a chain of threads and wait for 63 handles and the
		4979	previous thread in each. Sounds great!).
		4980
		4981	Newer versions support more handles, but you need to define C<FD_SETSIZE>
		4982	to some high number (e.g. C<2048>) before compiling the winsocket select
		4983	call (which might be in libev or elsewhere, for example, perl and many
		4984	other interpreters do their own select emulation on windows).
		4985
		4986	Another limit is the number of file descriptors in the Microsoft runtime
		4987	libraries, which by default is C<64> (there must be a hidden I<64>
		4988	fetish or something like this inside Microsoft). You can increase this
		4989	by calling C<_setmaxstdio>, which can increase this limit to C<2048>
		4990	(another arbitrary limit), but is broken in many versions of the Microsoft
		4991	runtime libraries. This might get you to about C<512> or C<2048> sockets
		4992	(depending on windows version and/or the phase of the moon). To get more,
		4993	you need to wrap all I/O functions and provide your own fd management, but
		4994	the cost of calling select (O(n²)) will likely make this unworkable.
		4995
		4996	=head2 PORTABILITY REQUIREMENTS
		4997
		4998	In addition to a working ISO-C implementation and of course the
		4999	backend-specific APIs, libev relies on a few additional extensions:
		5000
		5001	=over 4
		5002
		5003	=item C<void (*)(ev_watcher_type *, int revents)> must have compatible
		5004	calling conventions regardless of C<ev_watcher_type *>.
		5005
		5006	Libev assumes not only that all watcher pointers have the same internal
		5007	structure (guaranteed by POSIX but not by ISO C for example), but it also
		5008	assumes that the same (machine) code can be used to call any watcher
		5009	callback: The watcher callbacks have different type signatures, but libev
		5010	calls them using an C<ev_watcher *> internally.
		5011
		5012	=item pointer accesses must be thread-atomic
		5013
		5014	Accessing a pointer value must be atomic, it must both be readable and
		5015	writable in one piece - this is the case on all current architectures.
		5016
		5017	=item C<sig_atomic_t volatile> must be thread-atomic as well
		5018
		5019	The type C<sig_atomic_t volatile> (or whatever is defined as
		5020	C<EV_ATOMIC_T>) must be atomic with respect to accesses from different
		5021	threads. This is not part of the specification for C<sig_atomic_t>, but is
		5022	believed to be sufficiently portable.
		5023
		5024	=item C<sigprocmask> must work in a threaded environment
		5025
		5026	Libev uses C<sigprocmask> to temporarily block signals. This is not
		5027	allowed in a threaded program (C<pthread_sigmask> has to be used). Typical
		5028	pthread implementations will either allow C<sigprocmask> in the "main
		5029	thread" or will block signals process-wide, both behaviours would
		5030	be compatible with libev. Interaction between C<sigprocmask> and
		5031	C<pthread_sigmask> could complicate things, however.
		5032
		5033	The most portable way to handle signals is to block signals in all threads
		5034	except the initial one, and run the default loop in the initial thread as
		5035	well.
		5036
		5037	=item C<long> must be large enough for common memory allocation sizes
		5038
		5039	To improve portability and simplify its API, libev uses C<long> internally
		5040	instead of C<size_t> when allocating its data structures. On non-POSIX
		5041	systems (Microsoft...) this might be unexpectedly low, but is still at
		5042	least 31 bits everywhere, which is enough for hundreds of millions of
		5043	watchers.
		5044
		5045	=item C<double> must hold a time value in seconds with enough accuracy
		5046
		5047	The type C<double> is used to represent timestamps. It is required to
		5048	have at least 51 bits of mantissa (and 9 bits of exponent), which is
		5049	good enough for at least into the year 4000 with millisecond accuracy
		5050	(the design goal for libev). This requirement is overfulfilled by
		5051	implementations using IEEE 754, which is basically all existing ones. With
		5052	IEEE 754 doubles, you get microsecond accuracy until at least 2200.
		5053
		5054	=back
		5055
		5056	If you know of other additional requirements drop me a note.
		5057
		5058
		5059	=head1 ALGORITHMIC COMPLEXITIES
		5060
		5061	In this section the complexities of (many of) the algorithms used inside
		5062	libev will be documented. For complexity discussions about backends see
		5063	the documentation for C<ev_default_init>.
		5064
		5065	All of the following are about amortised time: If an array needs to be
		5066	extended, libev needs to realloc and move the whole array, but this
		5067	happens asymptotically rarer with higher number of elements, so O(1) might
		5068	mean that libev does a lengthy realloc operation in rare cases, but on
		5069	average it is much faster and asymptotically approaches constant time.
		5070
		5071	=over 4
		5072
		5073	=item Starting and stopping timer/periodic watchers: O(log skipped_other_timers)
		5074
		5075	This means that, when you have a watcher that triggers in one hour and
		5076	there are 100 watchers that would trigger before that, then inserting will
		5077	have to skip roughly seven (C<ld 100>) of these watchers.
		5078
		5079	=item Changing timer/periodic watchers (by autorepeat or calling again): O(log skipped_other_timers)
		5080
		5081	That means that changing a timer costs less than removing/adding them,
		5082	as only the relative motion in the event queue has to be paid for.
		5083
		5084	=item Starting io/check/prepare/idle/signal/child/fork/async watchers: O(1)
		5085
		5086	These just add the watcher into an array or at the head of a list.
		5087
		5088	=item Stopping check/prepare/idle/fork/async watchers: O(1)
		5089
		5090	=item Stopping an io/signal/child watcher: O(number_of_watchers_for_this_(fd/signal/pid % EV_PID_HASHSIZE))
		5091
		5092	These watchers are stored in lists, so they need to be walked to find the
		5093	correct watcher to remove. The lists are usually short (you don't usually
		5094	have many watchers waiting for the same fd or signal: one is typical, two
		5095	is rare).
		5096
		5097	=item Finding the next timer in each loop iteration: O(1)
		5098
		5099	By virtue of using a binary or 4-heap, the next timer is always found at a
		5100	fixed position in the storage array.
		5101
		5102	=item Each change on a file descriptor per loop iteration: O(number_of_watchers_for_this_fd)
		5103
		5104	A change means an I/O watcher gets started or stopped, which requires
		5105	libev to recalculate its status (and possibly tell the kernel, depending
		5106	on backend and whether C<ev_io_set> was used).
		5107
		5108	=item Activating one watcher (putting it into the pending state): O(1)
		5109
		5110	=item Priority handling: O(number_of_priorities)
		5111
		5112	Priorities are implemented by allocating some space for each
		5113	priority. When doing priority-based operations, libev usually has to
		5114	linearly search all the priorities, but starting/stopping and activating
		5115	watchers becomes O(1) with respect to priority handling.
		5116
		5117	=item Sending an ev_async: O(1)
		5118
		5119	=item Processing ev_async_send: O(number_of_async_watchers)
		5120
		5121	=item Processing signals: O(max_signal_number)
		5122
		5123	Sending involves a system call I<iff> there were no other C<ev_async_send>
		5124	calls in the current loop iteration. Checking for async and signal events
		5125	involves iterating over all running async watchers or all signal numbers.
		5126
		5127	=back
		5128
		5129
		5130	=head1 PORTING FROM LIBEV 3.X TO 4.X
		5131
		5132	The major version 4 introduced some incompatible changes to the API.
		5133
		5134	At the moment, the C<ev.h> header file provides compatibility definitions
		5135	for all changes, so most programs should still compile. The compatibility
		5136	layer might be removed in later versions of libev, so better update to the
		5137	new API early than late.
		5138
		5139	=over 4
		5140
		5141	=item C<EV_COMPAT3> backwards compatibility mechanism
		5142
		5143	The backward compatibility mechanism can be controlled by
		5144	C<EV_COMPAT3>. See L<PREPROCESSOR SYMBOLS/MACROS> in the L<EMBEDDING>
		5145	section.
		5146
		5147	=item C<ev_default_destroy> and C<ev_default_fork> have been removed
		5148
		5149	These calls can be replaced easily by their C<ev_loop_xxx> counterparts:
		5150
		5151	ev_loop_destroy (EV_DEFAULT_UC);
		5152	ev_loop_fork (EV_DEFAULT);
		5153
		5154	=item function/symbol renames
		5155
		5156	A number of functions and symbols have been renamed:
		5157
		5158	ev_loop => ev_run
		5159	EVLOOP_NONBLOCK => EVRUN_NOWAIT
		5160	EVLOOP_ONESHOT => EVRUN_ONCE
		5161
		5162	ev_unloop => ev_break
		5163	EVUNLOOP_CANCEL => EVBREAK_CANCEL
		5164	EVUNLOOP_ONE => EVBREAK_ONE
		5165	EVUNLOOP_ALL => EVBREAK_ALL
		5166
		5167	EV_TIMEOUT => EV_TIMER
		5168
		5169	ev_loop_count => ev_iteration
		5170	ev_loop_depth => ev_depth
		5171	ev_loop_verify => ev_verify
		5172
		5173	Most functions working on C<struct ev_loop> objects don't have an
		5174	C<ev_loop_> prefix, so it was removed; C<ev_loop>, C<ev_unloop> and
		5175	associated constants have been renamed to not collide with the C<struct
		5176	ev_loop> anymore and C<EV_TIMER> now follows the same naming scheme
		5177	as all other watcher types. Note that C<ev_loop_fork> is still called
		5178	C<ev_loop_fork> because it would otherwise clash with the C<ev_fork>
		5179	typedef.
		5180
		5181	=item C<EV_MINIMAL> mechanism replaced by C<EV_FEATURES>
		5182
		5183	The preprocessor symbol C<EV_MINIMAL> has been replaced by a different
		5184	mechanism, C<EV_FEATURES>. Programs using C<EV_MINIMAL> usually compile
		5185	and work, but the library code will of course be larger.
		5186
		5187	=back
		5188
		5189
		5190	=head1 GLOSSARY
		5191
		5192	=over 4
		5193
		5194	=item active
		5195
		5196	A watcher is active as long as it has been started and not yet stopped.
		5197	See L<WATCHER STATES> for details.
		5198
		5199	=item application
		5200
		5201	In this document, an application is whatever is using libev.
		5202
		5203	=item backend
		5204
		5205	The part of the code dealing with the operating system interfaces.
		5206
		5207	=item callback
		5208
		5209	The address of a function that is called when some event has been
		5210	detected. Callbacks are being passed the event loop, the watcher that
		5211	received the event, and the actual event bitset.
		5212
		5213	=item callback/watcher invocation
		5214
		5215	The act of calling the callback associated with a watcher.
		5216
		5217	=item event
		5218
		5219	A change of state of some external event, such as data now being available
		5220	for reading on a file descriptor, time having passed or simply not having
		5221	any other events happening anymore.
		5222
		5223	In libev, events are represented as single bits (such as C<EV_READ> or
		5224	C<EV_TIMER>).
		5225
		5226	=item event library
		5227
		5228	A software package implementing an event model and loop.
		5229
		5230	=item event loop
		5231
		5232	An entity that handles and processes external events and converts them
		5233	into callback invocations.
		5234
		5235	=item event model
		5236
		5237	The model used to describe how an event loop handles and processes
		5238	watchers and events.
		5239
		5240	=item pending
		5241
		5242	A watcher is pending as soon as the corresponding event has been
		5243	detected. See L<WATCHER STATES> for details.
		5244
		5245	=item real time
		5246
		5247	The physical time that is observed. It is apparently strictly monotonic :)
		5248
		5249	=item wall-clock time
		5250
		5251	The time and date as shown on clocks. Unlike real time, it can actually
		5252	be wrong and jump forwards and backwards, e.g. when you adjust your
		5253	clock.
		5254
		5255	=item watcher
		5256
		5257	A data structure that describes interest in certain events. Watchers need
		5258	to be started (attached to an event loop) before they can receive events.
		5259
		5260	=back
1696		5261
1697	=head1 AUTHOR	5262	=head1 AUTHOR
1698		5263
1699	Marc Lehmann <libev@schmorp.de>.	5264	Marc Lehmann <libev@schmorp.de>, with repeated corrections by Mikael
		5265	Magnusson and Emanuele Giaquinta, and minor corrections by many others.
1700		5266

Diff Legend

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